fridge_m viscam3 FINAL flush at 300 trajs

fridge_m/chunk_008/_logs/b1_ep00000000.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern minimalist kitchen with smooth matte light gray walls and a light yellow tile floor; a tall stainless steel single-door fridge with a stainless steel handle stands alone against the wall, no other appliances or clutter in frame, soft daylight from above. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000000/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000000/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000000/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000000/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:22,  2.84s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.28s/it]
 10%|█         | 3/30 [00:10<01:32,  3.42s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.48s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.52s/it]
 20%|██        | 6/30 [00:20<01:24,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.56s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.57s/it]
 30%|███       | 9/30 [00:31<01:15,  3.58s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.59s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.60s/it]
 40%|████      | 12/30 [00:42<01:04,  3.60s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.60s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.60s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.61s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.61s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.61s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.61s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.61s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.61s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.61s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.61s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.61s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.61s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.61s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.61s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.61s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.61s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:30,  3.11s/it]
  7%|▋         | 2/30 [00:07<01:40,  3.58s/it]
 10%|█         | 3/30 [00:10<01:40,  3.73s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.80s/it]
 17%|█▋        | 5/30 [00:18<01:36,  3.84s/it]
 20%|██        | 6/30 [00:22<01:33,  3.88s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.88s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:42<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.88s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.88s/it]
 93%|██████��██▎| 28/30 [01:48<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.79s/it]
 17%|█▋        | 5/30 [00:18<01:36,  3.85s/it]
 20%|██        | 6/30 [00:22<01:32,  3.87s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.90s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.89s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:42<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.88s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:36,  3.86s/it]
 20%|██        | 6/30 [00:22<01:33,  3.91s/it]
 23%|██▎       | 7/30 [00:26<01:30,  3.95s/it]
 27%|██▋       | 8/30 [00:30<01:26,  3.94s/it]
 30%|███       | 9/30 [00:34<01:22,  3.94s/it]
 33%|███▎      | 10/30 [00:38<01:19,  3.95s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.95s/it]
 40%|████      | 12/30 [00:46<01:11,  3.94s/it]
 43%|████▎     | 13/30 [00:50<01:06,  3.93s/it]
 47%|████▋     | 14/30 [00:54<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:58<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:21<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000000/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000000/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000000/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000001.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real clean residential kitchen with painted white walls and a wide-plank pale gray laminate floor; a tall stainless steel single-door fridge with a stainless steel slab handle is centered against the wall, bare countertops in the periphery, even ambient lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000001/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000001/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000001/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000001/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.79s/it]
  7%|▋         | 2/30 [00:06<01:29,  3.21s/it]
 10%|█         | 3/30 [00:09<01:30,  3.35s/it]
 13%|█▎        | 4/30 [00:13<01:28,  3.42s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.46s/it]
 20%|██        | 6/30 [00:20<01:23,  3.48s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.49s/it]
 27%|██▋       | 8/30 [00:27<01:17,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.52s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.52s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.53s/it]
 40%|████      | 12/30 [00:41<01:03,  3.53s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.53s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.53s/it]
 50%|█████     | 15/30 [00:52<00:52,  3.53s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.54s/it]
 57%|█████▋    | 17/30 [00:59<00:45,  3.53s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.53s/it]
 63%|██████▎   | 19/30 [01:06<00:38,  3.53s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.53s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.54s/it]
 73%|███████▎  | 22/30 [01:16<00:28,  3.54s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.54s/it]
 80%|████████  | 24/30 [01:24<00:21,  3.53s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.53s/it]
 87%|████████▋ | 26/30 [01:31<00:14,  3.53s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.53s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.53s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.53s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:25,  2.96s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.47s/it]
 10%|█         | 3/30 [00:10<01:38,  3.64s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.70s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.74s/it]
 20%|██        | 6/30 [00:22<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.77s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.80s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|███████��█▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.47s/it]
 10%|█         | 3/30 [00:10<01:37,  3.62s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.72s/it]
 20%|██        | 6/30 [00:21<01:30,  3.75s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.80s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.80s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.81s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.81s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:34,  3.26s/it]
  7%|▋         | 2/30 [00:07<01:42,  3.65s/it]
 10%|█         | 3/30 [00:11<01:40,  3.73s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:30,  3.78s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:30<01:23,  3.79s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:41<01:11,  3.78s/it]
 40%|████      | 12/30 [00:45<01:08,  3.78s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:04<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:23<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000001/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000001/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000001/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000002.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real Scandinavian-style kitchen with chalk white walls and warm cream-yellow vinyl flooring; a tall stainless steel single-door fridge with a stainless steel handle is the only appliance visible, plain white skirting board, soft diffuse daylight. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000002/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000002/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000002/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000002/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:22,  2.84s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.27s/it]
 10%|█         | 3/30 [00:09<01:32,  3.42s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.49s/it]
 17%|█▋        | 5/30 [00:17<01:28,  3.53s/it]
 20%|██        | 6/30 [00:20<01:25,  3.57s/it]
 23%|██▎       | 7/30 [00:24<01:22,  3.60s/it]
 27%|██▋       | 8/30 [00:28<01:19,  3.61s/it]
 30%|███       | 9/30 [00:31<01:16,  3.64s/it]
 33%|███▎      | 10/30 [00:35<01:12,  3.64s/it]
 37%|███▋      | 11/30 [00:39<01:09,  3.64s/it]
 40%|████      | 12/30 [00:42<01:05,  3.65s/it]
 43%|████▎     | 13/30 [00:46<01:02,  3.65s/it]
 47%|████▋     | 14/30 [00:50<00:58,  3.65s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.65s/it]
 53%|█████▎    | 16/30 [00:57<00:51,  3.66s/it]
 57%|█████▋    | 17/30 [01:01<00:47,  3.65s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.65s/it]
 63%|██████▎   | 19/30 [01:08<00:40,  3.65s/it]
 67%|██████▋   | 20/30 [01:12<00:36,  3.65s/it]
 70%|███████   | 21/30 [01:15<00:32,  3.65s/it]
 73%|███████▎  | 22/30 [01:19<00:29,  3.65s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.65s/it]
 80%|████████  | 24/30 [01:26<00:21,  3.67s/it]
 83%|████████▎ | 25/30 [01:30<00:18,  3.66s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.65s/it]
 90%|█████████ | 27/30 [01:37<00:10,  3.66s/it]
 93%|█████████▎| 28/30 [01:41<00:07,  3.67s/it]
 97%|█████████▋| 29/30 [01:45<00:03,  3.69s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:40,  3.72s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.79s/it]
 17%|█▋        | 5/30 [00:18<01:36,  3.85s/it]
 20%|██        | 6/30 [00:22<01:32,  3.87s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.89s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.90s/it]
 30%|███       | 9/30 [00:34<01:22,  3.91s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.92s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.92s/it]
 40%|████      | 12/30 [00:46<01:10,  3.93s/it]
 43%|████▎     | 13/30 [00:50<01:06,  3.93s/it]
 47%|████▋     | 14/30 [00:54<01:02,  3.93s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.92s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.92s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.92s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:37<00:19,  3.95s/it]
 87%|████████▋ | 26/30 [01:41<00:15,  3.94s/it]
 90%|█████████ | 27/30 [01:45<00:11,  3.96s/it]
 93%|█████████▎| 28/30 [01:49<00:07,  3.96s/it]
 97%|█████████▋| 29/30 [01:53<00:03,  3.97s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.90s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.91s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.92s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.92s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.92s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.93s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.93s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.93s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.94s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.93s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.94s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.94s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.91s/it]
 37%|███▋      | 11/30 [00:41<01:14,  3.92s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.91s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.92s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.92s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.93s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.94s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.96s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.94s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.93s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.94s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.93s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.95s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.94s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000002/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000002/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000002/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000003.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real industrial loft kitchen with light cool gray concrete-look walls and a polished light gray concrete floor; a tall stainless steel single-door fridge with a stainless steel handle stands flush to the wall, no upper cabinets in frame, neutral overhead lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000003/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000003/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000003/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000003/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:22,  2.84s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.28s/it]
 10%|█         | 3/30 [00:09<01:31,  3.39s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.44s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.47s/it]
 20%|██        | 6/30 [00:20<01:23,  3.49s/it]
 23%|██▎       | 7/30 [00:24<01:20,  3.52s/it]
 27%|██▋       | 8/30 [00:27<01:17,  3.54s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.56s/it]
 40%|████      | 12/30 [00:41<01:04,  3.57s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.57s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.57s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.58s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.58s/it]
 57%|█████▋    | 17/30 [00:59<00:46,  3.58s/it]
 60%|██████    | 18/30 [01:03<00:42,  3.58s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.58s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.59s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.59s/it]
 73%|███████▎  | 22/30 [01:17<00:28,  3.60s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.60s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:28<00:17,  3.60s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.61s/it]
 90%|█████████ | 27/30 [01:35<00:10,  3.61s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.60s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.60s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.73s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.82s/it]
 27%|██▋       | 8/30 [00:29<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.87s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:31<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|███████���█▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:33<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.73s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.90s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000003/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000003/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000003/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000004.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real compact apartment kitchen with smooth eggshell white walls and a light beige-yellow LVP floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, no other appliances visible, gentle natural light from off-camera. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000004/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000004/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000004/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000004/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.82s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.25s/it]
 10%|█         | 3/30 [00:09<01:31,  3.39s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.46s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.56s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.58s/it]
 30%|███       | 9/30 [00:31<01:15,  3.62s/it]
 33%|███▎      | 10/30 [00:35<01:12,  3.63s/it]
 37%|███▋      | 11/30 [00:38<01:09,  3.64s/it]
 40%|████      | 12/30 [00:42<01:05,  3.63s/it]
 43%|████▎     | 13/30 [00:46<01:01,  3.62s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.61s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.61s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.61s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.61s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.61s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.62s/it]
 70%|███████   | 21/30 [01:15<00:32,  3.62s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.61s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.61s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.61s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.61s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.61s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.61s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.60s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.60s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.05s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:40,  3.71s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.80s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.86s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.88s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.91s/it]
 30%|███       | 9/30 [00:34<01:22,  3.93s/it]
 33%|███▎      | 10/30 [00:38<01:19,  3.96s/it]
 37%|███▋      | 11/30 [00:42<01:15,  3.96s/it]
 40%|████      | 12/30 [00:46<01:11,  3.96s/it]
 43%|████▎     | 13/30 [00:50<01:07,  3.94s/it]
 47%|████▋     | 14/30 [00:54<01:02,  3.92s/it]
 50%|█████     | 15/30 [00:58<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:21<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:25<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.87s/it]
 93%|████████��▎| 28/30 [01:48<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.86s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:40,  3.74s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.80s/it]
 17%|█▋        | 5/30 [00:18<01:36,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.86s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.90s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.91s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.92s/it]
 40%|████      | 12/30 [00:46<01:10,  3.90s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.86s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.86s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.06s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.80s/it]
 17%|█▋        | 5/30 [00:18<01:36,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.86s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.88s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.89s/it]
 30%|███       | 9/30 [00:34<01:22,  3.91s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.92s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.91s/it]
 40%|████      | 12/30 [00:46<01:11,  3.94s/it]
 43%|████▎     | 13/30 [00:50<01:06,  3.93s/it]
 47%|████▋     | 14/30 [00:54<01:02,  3.94s/it]
 50%|█████     | 15/30 [00:58<00:58,  3.93s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.92s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.89s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:21<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.87s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000004/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000004/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000004/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000005.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real small contemporary kitchen with pale dove gray walls and a white matte tile floor; a tall stainless steel single-door fridge with a stainless steel pull handle sits centered against the wall, empty surroundings, soft even fill light. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000005/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000005/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000005/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000005/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.76s/it]
  7%|▋         | 2/30 [00:06<01:29,  3.20s/it]
 10%|█         | 3/30 [00:09<01:30,  3.34s/it]
 13%|█▎        | 4/30 [00:13<01:28,  3.42s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.47s/it]
 20%|██        | 6/30 [00:20<01:23,  3.48s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.49s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.50s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.51s/it]
 37%|███▋      | 11/30 [00:37<01:06,  3.51s/it]
 40%|████      | 12/30 [00:41<01:03,  3.51s/it]
 43%|████▎     | 13/30 [00:45<00:59,  3.52s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.52s/it]
 50%|█████     | 15/30 [00:52<00:52,  3.53s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.53s/it]
 57%|█████▋    | 17/30 [00:59<00:45,  3.53s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.53s/it]
 63%|██████▎   | 19/30 [01:06<00:38,  3.53s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.53s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.53s/it]
 73%|███████▎  | 22/30 [01:16<00:28,  3.53s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.53s/it]
 80%|████████  | 24/30 [01:23<00:21,  3.53s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.53s/it]
 87%|████████▋ | 26/30 [01:30<00:14,  3.53s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.53s/it]
 93%|█████████▎| 28/30 [01:37<00:07,  3.55s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.54s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:25,  2.96s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.63s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.70s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.74s/it]
 20%|██        | 6/30 [00:22<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.77s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.78s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:25,  2.96s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.44s/it]
 10%|█         | 3/30 [00:10<01:37,  3.59s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.67s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.75s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:33<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.44s/it]
 10%|█         | 3/30 [00:10<01:37,  3.60s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.67s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.71s/it]
 20%|██        | 6/30 [00:21<01:29,  3.73s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.75s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:40<01:12,  3.80s/it]
 40%|████      | 12/30 [00:44<01:08,  3.80s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.81s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.81s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.80s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000005/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000005/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000005/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000006.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real bright open-plan kitchen with light gray painted walls and a creamy off-white engineered wood floor; a stainless steel single-door fridge with a stainless steel handle stands alone, no decor on walls, soft balanced daylight illumination. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 295 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000006/ext1/depth/original_depth.mp4
+  ext1 mask: 295 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000006/ext1/mask/rgb_robot.mp4
+  wrist depth: 295 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000006/wrist/depth/original_depth.mp4
+  wrist mask: 295 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000006/wrist/mask/rgb_robot.mp4
+Total source frames: 295
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.80s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.38s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.45s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.49s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.54s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:15,  3.58s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.58s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.58s/it]
 40%|████      | 12/30 [00:42<01:04,  3.58s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.58s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.58s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.58s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.59s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.59s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.59s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.59s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.59s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.59s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.59s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:28<00:18,  3.61s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.61s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.62s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.63s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.63s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.72s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.76s/it]
 20%|██        | 6/30 [00:22<01:30,  3.79s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.81s/it]
 27%|██▋       | 8/30 [00:29<01:24,  3.82s/it]
 30%|███       | 9/30 [00:33<01:20,  3.83s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.84s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.84s/it]
 40%|████      | 12/30 [00:45<01:09,  3.85s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.87s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.86s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.86s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.86s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.86s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.86s/it]
 73%|███████▎  | 22/30 [01:23<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:27<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:31<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.92s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.84s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.84s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.85s/it]
 40%|████      | 12/30 [00:45<01:09,  3.85s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.85s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.85s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.85s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.86s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.86s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.86s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.86s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.86s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:27<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:31<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.05s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.84s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.85s/it]
 40%|████      | 12/30 [00:45<01:09,  3.85s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.85s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.85s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.85s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.86s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.86s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.86s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:31<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+Saving full-length videos (295 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000006/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000006/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000006/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000007.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real neutral utility kitchen with smooth warm white walls and a light yellow linoleum floor; a stainless steel single-door fridge with a stainless steel slab handle is the focal point, plain wall behind it, even shadowless lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000007/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000007/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000007/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000007/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.81s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.39s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.47s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.51s/it]
 20%|██        | 6/30 [00:20<01:24,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.56s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.58s/it]
 30%|███       | 9/30 [00:31<01:15,  3.59s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.60s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.60s/it]
 40%|████      | 12/30 [00:42<01:04,  3.61s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.61s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.61s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.62s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.62s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.62s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.62s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.62s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.62s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.63s/it]
 73%|███████▎  | 22/30 [01:18<00:29,  3.64s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.64s/it]
 80%|████████  | 24/30 [01:26<00:21,  3.66s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.67s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.67s/it]
 90%|█████████ | 27/30 [01:37<00:11,  3.68s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.70s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.70s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:33<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.93s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.95s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.96s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.96s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.96s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.95s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.97s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.96s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.94s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:09,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.92s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.93s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.94s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.95s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.94s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.94s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.95s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.94s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.94s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.94s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|���▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:32,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:09,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.92s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.94s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.93s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.93s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000007/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000007/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000007/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000008.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real minimalist studio kitchen with light gray plaster walls and a pale yellow oak engineered floor; a tall stainless steel single-door fridge with stainless steel handle stands solo against the wall, no upper cabinets, soft daylight from a side window. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000008/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000008/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000008/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000008/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:32,  3.31s/it]
 10%|█         | 3/30 [00:10<01:32,  3.44s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.48s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.51s/it]
 20%|██        | 6/30 [00:20<01:24,  3.53s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.54s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.57s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.58s/it]
 40%|████      | 12/30 [00:42<01:04,  3.58s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.59s/it]
 50%|█████     | 15/30 [00:53<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.62s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.62s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.62s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.61s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.61s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.61s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.61s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.61s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.61s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.61s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.61s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.61s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.61s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.61s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.05s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:32,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.89s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.55s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.89s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.86s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.89s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000008/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000008/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000008/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000009.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern home kitchen with crisp white painted walls and a smooth light gray ceramic tile floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, surrounding area empty and uncluttered, neutral white lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000009/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000009/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000009/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000009/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.80s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.22s/it]
 10%|█         | 3/30 [00:09<01:30,  3.35s/it]
 13%|█▎        | 4/30 [00:13<01:28,  3.42s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.46s/it]
 20%|██        | 6/30 [00:20<01:23,  3.47s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.49s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.51s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.53s/it]
 37%|███▋      | 11/30 [00:38<01:06,  3.53s/it]
 40%|████      | 12/30 [00:41<01:03,  3.53s/it]
 43%|████▎     | 13/30 [00:45<00:59,  3.53s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.54s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.56s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.56s/it]
 57%|█████▋    | 17/30 [00:59<00:46,  3.56s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.55s/it]
 63%|██████▎   | 19/30 [01:06<00:39,  3.55s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.54s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.54s/it]
 73%|███████▎  | 22/30 [01:17<00:28,  3.55s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.55s/it]
 80%|████████  | 24/30 [01:24<00:21,  3.56s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.57s/it]
 87%|████████▋ | 26/30 [01:31<00:14,  3.56s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.57s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.57s/it]
 97%|█████████▋| 29/30 [01:42<00:03,  3.56s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.62s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.69s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.74s/it]
 20%|██        | 6/30 [00:22<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.77s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.81s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.84s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:48<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:52<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:56<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.85s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.83s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.82s/it]
 73%|███████▎  | 22/30 [01:23<00:30,  3.81s/it]
 77%|███████▋  | 23/30 [01:27<00:26,  3.81s/it]
 80%|████████  | 24/30 [01:31<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.80s/it]
 93%|█████████��| 28/30 [01:46<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:50<00:03,  3.79s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.47s/it]
 10%|█         | 3/30 [00:10<01:37,  3.63s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.70s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.75s/it]
 20%|██        | 6/30 [00:22<01:30,  3.77s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.78s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.80s/it]
 30%|███       | 9/30 [00:33<01:19,  3.80s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.80s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.81s/it]
 40%|████      | 12/30 [00:45<01:08,  3.83s/it]
 43%|████▎     | 13/30 [00:48<01:05,  3.83s/it]
 47%|████▋     | 14/30 [00:52<01:01,  3.84s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.84s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.84s/it]
 57%|█████▋    | 17/30 [01:04<00:49,  3.83s/it]
 60%|██████    | 18/30 [01:08<00:45,  3.82s/it]
 63%|██████▎   | 19/30 [01:11<00:42,  3.82s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.81s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.81s/it]
 73%|███████▎  | 22/30 [01:23<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.80s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.79s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.81s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.47s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.77s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.80s/it]
 47%|████▋     | 14/30 [00:52<01:01,  3.82s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.82s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.84s/it]
 57%|█████▋    | 17/30 [01:03<00:50,  3.85s/it]
 60%|██████    | 18/30 [01:07<00:46,  3.84s/it]
 63%|██████▎   | 19/30 [01:11<00:42,  3.83s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.82s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.81s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.81s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000009/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000009/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000009/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000010.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern minimalist kitchen with smooth matte light gray walls and a light yellow tile floor; a tall stainless steel single-door fridge with a stainless steel handle stands alone against the wall, no other appliances or clutter in frame, soft daylight from above. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000010/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000010/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000010/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000010/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.80s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.23s/it]
 10%|█         | 3/30 [00:09<01:30,  3.37s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.44s/it]
 17%|█▋        | 5/30 [00:16<01:27,  3.49s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.54s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:15,  3.58s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.60s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.60s/it]
 40%|████      | 12/30 [00:42<01:04,  3.60s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.60s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.60s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.60s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.65s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.65s/it]
 63%|██████▎   | 19/30 [01:07<00:40,  3.66s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.66s/it]
 70%|███████   | 21/30 [01:15<00:32,  3.66s/it]
 73%|███████▎  | 22/30 [01:18<00:29,  3.65s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.66s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.65s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.63s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.63s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.62s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.62s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.61s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.10s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.55s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.96s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.95s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.95s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.94s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.93s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|██████��██▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.10s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.86s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.86s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.86s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.86s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.86s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.86s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:33<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.86s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.86s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.86s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.86s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.86s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.86s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000010/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000010/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000010/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b1_ep00000011.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real clean residential kitchen with painted white walls and a wide-plank pale gray laminate floor; a tall stainless steel single-door fridge with a stainless steel slab handle is centered against the wall, bare countertops in the periphery, even ambient lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000011/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000011/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000011/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b1/h5_v4/episode_00000011/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.81s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.38s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.46s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.56s/it]
 27%|██▋       | 8/30 [00:27<01:19,  3.60s/it]
 30%|███       | 9/30 [00:31<01:16,  3.62s/it]
 33%|███▎      | 10/30 [00:35<01:12,  3.63s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.63s/it]
 40%|████      | 12/30 [00:42<01:05,  3.63s/it]
 43%|████▎     | 13/30 [00:46<01:01,  3.63s/it]
 47%|████▋     | 14/30 [00:49<00:58,  3.63s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.63s/it]
 53%|█████▎    | 16/30 [00:57<00:51,  3.65s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.66s/it]
 60%|██████    | 18/30 [01:04<00:44,  3.67s/it]
 63%|██████▎   | 19/30 [01:08<00:40,  3.69s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.70s/it]
 70%|███████   | 21/30 [01:15<00:33,  3.68s/it]
 73%|███████▎  | 22/30 [01:19<00:29,  3.69s/it]
 77%|███████▋  | 23/30 [01:23<00:25,  3.70s/it]
 80%|████████  | 24/30 [01:26<00:22,  3.69s/it]
 83%|████████▎ | 25/30 [01:30<00:18,  3.68s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.67s/it]
 90%|█████████ | 27/30 [01:37<00:10,  3.67s/it]
 93%|█████████▎| 28/30 [01:41<00:07,  3.66s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.67s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:14,  3.90s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:59,  3.94s/it]
 53%|█████▎    | 16/30 [01:01<00:55,  3.94s/it]
 57%|█████▋    | 17/30 [01:05<00:51,  3.95s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.99s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.99s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.97s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.98s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.96s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.96s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.94s/it]
 83%|████████▎ | 25/30 [01:37<00:19,  3.93s/it]
 87%|████████▋ | 26/30 [01:41<00:15,  3.93s/it]
 90%|█████████ | 27/30 [01:45<00:11,  3.92s/it]
 93%|███████��█▎| 28/30 [01:48<00:07,  3.92s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.92s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.93s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.92s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:51,  3.93s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.92s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.92s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.92s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.93s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.92s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000011/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000011/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b1_ep00000011/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000000.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real Scandinavian-style kitchen with chalk white walls and warm cream-yellow vinyl flooring; a tall stainless steel single-door fridge with a stainless steel handle is the only appliance visible, plain white skirting board, soft diffuse daylight. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 282 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000000/ext1/depth/original_depth.mp4
+  ext1 mask: 282 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000000/ext1/mask/rgb_robot.mp4
+  wrist depth: 282 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000000/wrist/depth/original_depth.mp4
+  wrist mask: 282 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000000/wrist/mask/rgb_robot.mp4
+Total source frames: 282
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.23s/it]
 10%|█         | 3/30 [00:09<01:31,  3.38s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.45s/it]
 17%|█▋        | 5/30 [00:16<01:27,  3.49s/it]
 20%|██        | 6/30 [00:20<01:24,  3.53s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:04,  3.57s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.58s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.58s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.58s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.58s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.58s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.59s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.59s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.59s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.61s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.64s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.63s/it]
 83%|████████▎ | 25/30 [01:28<00:18,  3.62s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.62s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.61s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.61s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.61s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.55s/it]
 10%|█         | 3/30 [00:10<01:40,  3.72s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.80s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.06s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.93s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.08s/it]
  7%|▋         | 2/30 [00:07<01:41,  3.62s/it]
 10%|█         | 3/30 [00:10<01:41,  3.75s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.79s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.90s/it]
+
+
+Saving full-length videos (282 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000000/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000000/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000000/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000001.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real industrial loft kitchen with light cool gray concrete-look walls and a polished light gray concrete floor; a tall stainless steel single-door fridge with a stainless steel handle stands flush to the wall, no upper cabinets in frame, neutral overhead lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000001/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000001/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000001/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000001/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.22s/it]
 10%|█         | 3/30 [00:09<01:31,  3.39s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.43s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.46s/it]
 20%|██        | 6/30 [00:20<01:23,  3.48s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.49s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.49s/it]
 30%|███       | 9/30 [00:30<01:13,  3.50s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.50s/it]
 37%|███▋      | 11/30 [00:37<01:06,  3.51s/it]
 40%|████      | 12/30 [00:41<01:03,  3.51s/it]
 43%|████▎     | 13/30 [00:45<00:59,  3.51s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.51s/it]
 50%|█████     | 15/30 [00:52<00:52,  3.52s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.52s/it]
 57%|█████▋    | 17/30 [00:59<00:45,  3.52s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.54s/it]
 63%|██████▎   | 19/30 [01:06<00:39,  3.57s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.56s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.55s/it]
 73%|███████▎  | 22/30 [01:16<00:28,  3.54s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.54s/it]
 80%|████████  | 24/30 [01:23<00:21,  3.54s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.54s/it]
 87%|████████▋ | 26/30 [01:31<00:14,  3.54s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.54s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.54s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.53s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.77s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.78s/it]
 40%|████      | 12/30 [00:44<01:08,  3.78s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.78s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.81s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.83s/it]
 63%|██████▎   | 19/30 [01:11<00:42,  3.85s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.83s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.82s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.81s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.81s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|███████���█▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.81s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.84s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.84s/it]
 77%|███████▋  | 23/30 [01:26<00:27,  3.86s/it]
 80%|████████  | 24/30 [01:30<00:23,  3.86s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.85s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.84s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.83s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.83s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.83s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.77s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.78s/it]
 40%|████      | 12/30 [00:44<01:08,  3.78s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:14<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.82s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.82s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.81s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.82s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.82s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.81s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000001/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000001/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000001/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000002.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real compact apartment kitchen with smooth eggshell white walls and a light beige-yellow LVP floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, no other appliances visible, gentle natural light from off-camera. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000002/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000002/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000002/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000002/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.81s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.38s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.46s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.53s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.57s/it]
 30%|███       | 9/30 [00:31<01:15,  3.58s/it]
 33%|███▎      | 10/30 [00:35<01:12,  3.61s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.63s/it]
 40%|████      | 12/30 [00:42<01:05,  3.63s/it]
 43%|████▎     | 13/30 [00:46<01:02,  3.65s/it]
 47%|████▋     | 14/30 [00:49<00:58,  3.64s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.66s/it]
 53%|█████▎    | 16/30 [00:57<00:51,  3.66s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.65s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.65s/it]
 63%|██████▎   | 19/30 [01:08<00:40,  3.64s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.64s/it]
 70%|███████   | 21/30 [01:15<00:32,  3.64s/it]
 73%|███████▎  | 22/30 [01:18<00:29,  3.64s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.64s/it]
 80%|████████  | 24/30 [01:26<00:21,  3.64s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.64s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.64s/it]
 90%|█████████ | 27/30 [01:37<00:10,  3.64s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.64s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.64s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:37<01:18,  3.90s/it]
 37%|███▋      | 11/30 [00:41<01:14,  3.91s/it]
 40%|████      | 12/30 [00:45<01:10,  3.93s/it]
 43%|████▎     | 13/30 [00:49<01:07,  3.95s/it]
 47%|████▋     | 14/30 [00:53<01:03,  3.95s/it]
 50%|█████     | 15/30 [00:57<00:59,  3.94s/it]
 53%|█████▎    | 16/30 [01:01<00:55,  3.93s/it]
 57%|█████▋    | 17/30 [01:05<00:51,  3.93s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.92s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.92s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.92s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.92s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.93s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|████████��▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.89s/it]
 30%|███       | 9/30 [00:34<01:22,  3.92s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.93s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.94s/it]
 40%|████      | 12/30 [00:46<01:10,  3.94s/it]
 43%|████▎     | 13/30 [00:50<01:06,  3.94s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.92s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.92s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.91s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.92s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.92s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.92s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.86s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.89s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.90s/it]
 30%|███       | 9/30 [00:34<01:21,  3.90s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.90s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.90s/it]
 40%|████      | 12/30 [00:46<01:10,  3.90s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.90s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.93s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000002/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000002/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000002/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000003.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real small contemporary kitchen with pale dove gray walls and a white matte tile floor; a tall stainless steel single-door fridge with a stainless steel pull handle sits centered against the wall, empty surroundings, soft even fill light. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000003/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000003/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000003/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000003/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.22s/it]
 10%|█         | 3/30 [00:09<01:30,  3.37s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.44s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.48s/it]
 20%|██        | 6/30 [00:20<01:24,  3.51s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.53s/it]
 27%|██▋       | 8/30 [00:27<01:17,  3.54s/it]
 30%|███       | 9/30 [00:31<01:14,  3.55s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.58s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.59s/it]
 40%|████      | 12/30 [00:42<01:04,  3.60s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.62s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.62s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.63s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.62s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.61s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.61s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:11<00:35,  3.60s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.60s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.60s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.59s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.59s/it]
 83%|████████▎ | 25/30 [01:28<00:17,  3.59s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.59s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.59s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.60s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.60s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.48s/it]
 10%|█         | 3/30 [00:10<01:38,  3.64s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.72s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.77s/it]
 20%|██        | 6/30 [00:22<01:31,  3.80s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.82s/it]
 27%|██▋       | 8/30 [00:29<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:14,  3.90s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.86s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.86s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.48s/it]
 10%|█         | 3/30 [00:10<01:38,  3.64s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.72s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.77s/it]
 20%|██        | 6/30 [00:22<01:31,  3.80s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.82s/it]
 27%|██▋       | 8/30 [00:29<01:24,  3.85s/it]
 30%|███       | 9/30 [00:33<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.48s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.86s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.86s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.86s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.86s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.86s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.86s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.86s/it]
 80%|████████  | 24/30 [01:31<00:23,  3.86s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.86s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.86s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.85s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.85s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.85s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000003/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000003/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000003/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000004.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real bright open-plan kitchen with light gray painted walls and a creamy off-white engineered wood floor; a stainless steel single-door fridge with a stainless steel handle stands alone, no decor on walls, soft balanced daylight illumination. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000004/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000004/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000004/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000004/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.82s/it]
  7%|▋         | 2/30 [00:06<01:32,  3.29s/it]
 10%|█         | 3/30 [00:09<01:32,  3.41s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.47s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.54s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:04,  3.57s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.60s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.62s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.60s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.60s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.59s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.59s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.59s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.59s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:28<00:17,  3.60s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.60s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.60s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.60s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.60s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.55s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:40,  3.58s/it]
 10%|█         | 3/30 [00:10<01:40,  3.72s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:20,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.10s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.55s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000004/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000004/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000004/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000005.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real neutral utility kitchen with smooth warm white walls and a light yellow linoleum floor; a stainless steel single-door fridge with a stainless steel slab handle is the focal point, plain wall behind it, even shadowless lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000005/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000005/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000005/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000005/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.22s/it]
 10%|█         | 3/30 [00:09<01:30,  3.36s/it]
 13%|█▎        | 4/30 [00:13<01:28,  3.42s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.45s/it]
 20%|██        | 6/30 [00:20<01:23,  3.47s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.49s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.49s/it]
 30%|███       | 9/30 [00:30<01:13,  3.51s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.51s/it]
 37%|███▋      | 11/30 [00:37<01:06,  3.52s/it]
 40%|████      | 12/30 [00:41<01:03,  3.54s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.57s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.56s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.55s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.54s/it]
 57%|█████▋    | 17/30 [00:59<00:45,  3.53s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.53s/it]
 63%|██████▎   | 19/30 [01:06<00:38,  3.53s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.53s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.53s/it]
 73%|███████▎  | 22/30 [01:16<00:28,  3.53s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.53s/it]
 80%|████████  | 24/30 [01:24<00:21,  3.53s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.53s/it]
 87%|████████▋ | 26/30 [01:31<00:14,  3.53s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.53s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.53s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.53s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.75s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:14<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:33<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.81s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.82s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.83s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.84s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.69s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.74s/it]
 20%|██        | 6/30 [00:21<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.78s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.80s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.80s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.80s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.80s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.80s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.69s/it]
 17%|���▋        | 5/30 [00:18<01:33,  3.73s/it]
 20%|██        | 6/30 [00:21<01:29,  3.75s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.77s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.78s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:33<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.81s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.81s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000005/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000005/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000005/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000006.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real minimalist studio kitchen with light gray plaster walls and a pale yellow oak engineered floor; a tall stainless steel single-door fridge with stainless steel handle stands solo against the wall, no upper cabinets, soft daylight from a side window. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000006/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000006/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000006/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000006/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:23,  2.86s/it]
  7%|▋         | 2/30 [00:06<01:33,  3.35s/it]
 10%|█         | 3/30 [00:10<01:33,  3.47s/it]
 13%|█▎        | 4/30 [00:13<01:31,  3.52s/it]
 17%|█▋        | 5/30 [00:17<01:29,  3.57s/it]
 20%|██        | 6/30 [00:20<01:25,  3.57s/it]
 23%|██▎       | 7/30 [00:24<01:22,  3.58s/it]
 27%|██▋       | 8/30 [00:28<01:19,  3.59s/it]
 30%|███       | 9/30 [00:31<01:15,  3.59s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.60s/it]
 37%|███▋      | 11/30 [00:39<01:08,  3.60s/it]
 40%|████      | 12/30 [00:42<01:04,  3.61s/it]
 43%|████▎     | 13/30 [00:46<01:01,  3.61s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.62s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.63s/it]
 53%|█████▎    | 16/30 [00:57<00:50,  3.64s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.64s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.64s/it]
 63%|██████▎   | 19/30 [01:08<00:40,  3.65s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.66s/it]
 70%|███████   | 21/30 [01:15<00:32,  3.65s/it]
 73%|███████▎  | 22/30 [01:19<00:29,  3.64s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.64s/it]
 80%|████████  | 24/30 [01:26<00:21,  3.64s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.63s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.63s/it]
 90%|█████████ | 27/30 [01:37<00:10,  3.63s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.63s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.63s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.08s/it]
  7%|▋         | 2/30 [00:07<01:40,  3.58s/it]
 10%|█         | 3/30 [00:10<01:40,  3.74s/it]
 13%|█▎        | 4/30 [00:14<01:39,  3.81s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.86s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.90s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.90s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.90s/it]
 40%|████      | 12/30 [00:46<01:10,  3.90s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.90s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.92s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.88s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.92s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.92s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.90s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.91s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:50<01:07,  3.95s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.93s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.93s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.92s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.92s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.92s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.91s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.92s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.92s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.92s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000006/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000006/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000006/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000007.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern home kitchen with crisp white painted walls and a smooth light gray ceramic tile floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, surrounding area empty and uncluttered, neutral white lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000007/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000007/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000007/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000007/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.25s/it]
 10%|█         | 3/30 [00:09<01:32,  3.43s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.47s/it]
 17%|█▋        | 5/30 [00:17<01:28,  3.53s/it]
 20%|██        | 6/30 [00:20<01:24,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:04,  3.58s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.59s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.59s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.59s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.59s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.59s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.59s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.59s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.59s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.59s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.59s/it]
 83%|████████▎ | 25/30 [01:28<00:17,  3.59s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.59s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.59s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.59s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.59s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:30,  3.11s/it]
  7%|▋         | 2/30 [00:07<01:40,  3.60s/it]
 10%|█         | 3/30 [00:10<01:40,  3.73s/it]
 13%|█▎        | 4/30 [00:14<01:39,  3.82s/it]
 17%|█▋        | 5/30 [00:18<01:36,  3.86s/it]
 20%|██        | 6/30 [00:22<01:33,  3.88s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.91s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.90s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:42<01:13,  3.88s/it]
 40%|████      | 12/30 [00:46<01:09,  3.88s/it]
 43%|████▎     | 13/30 [00:50<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.86s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.86s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.86s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.86s/it]
 93%|█████████��| 28/30 [01:47<00:07,  3.86s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.86s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:33,  3.90s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.91s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.90s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:42<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.88s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.86s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.86s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.86s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.86s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.48s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:33<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000007/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000007/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000007/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000008.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern minimalist kitchen with smooth matte light gray walls and a light yellow tile floor; a tall stainless steel single-door fridge with a stainless steel handle stands alone against the wall, no other appliances or clutter in frame, soft daylight from above. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000008/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000008/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000008/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000008/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.81s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.25s/it]
 10%|█         | 3/30 [00:09<01:31,  3.40s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.46s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.54s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:04,  3.57s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.58s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.59s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.59s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.59s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.59s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.59s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.61s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.60s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:28<00:18,  3.60s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.60s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.60s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.60s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.60s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.90s/it]
 93%|██████��██▎| 28/30 [01:47<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:33<01:20,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.55s/it]
 10%|█         | 3/30 [00:10<01:40,  3.72s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.85s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:42<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:09,  3.88s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.89s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000008/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000008/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000008/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000009.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real clean residential kitchen with painted white walls and a wide-plank pale gray laminate floor; a tall stainless steel single-door fridge with a stainless steel slab handle is centered against the wall, bare countertops in the periphery, even ambient lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 248 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000009/ext1/depth/original_depth.mp4
+  ext1 mask: 248 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000009/ext1/mask/rgb_robot.mp4
+  wrist depth: 248 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000009/wrist/depth/original_depth.mp4
+  wrist mask: 248 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000009/wrist/mask/rgb_robot.mp4
+Total source frames: 248
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.79s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.22s/it]
 10%|█         | 3/30 [00:09<01:30,  3.36s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.43s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.46s/it]
 20%|██        | 6/30 [00:20<01:23,  3.48s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.50s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.51s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.51s/it]
 37%|███▋      | 11/30 [00:38<01:06,  3.52s/it]
 40%|████      | 12/30 [00:41<01:03,  3.52s/it]
 43%|████▎     | 13/30 [00:45<00:59,  3.52s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.54s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.54s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.53s/it]
 57%|█████▋    | 17/30 [00:59<00:45,  3.53s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.53s/it]
 63%|██████▎   | 19/30 [01:06<00:38,  3.53s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.54s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.55s/it]
 73%|███████▎  | 22/30 [01:16<00:28,  3.55s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.55s/it]
 80%|████████  | 24/30 [01:24<00:21,  3.55s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.55s/it]
 87%|████████▋ | 26/30 [01:31<00:14,  3.55s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.55s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.54s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.54s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.47s/it]
 10%|█         | 3/30 [00:10<01:38,  3.64s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.70s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.73s/it]
 20%|██        | 6/30 [00:22<01:30,  3.75s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.77s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.80s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.80s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.80s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.80s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.82s/it]
 93%|███████��█▎| 28/30 [01:45<00:07,  3.81s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.81s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.74s/it]
 20%|██        | 6/30 [00:21<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.77s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.72s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.75s/it]
 20%|██        | 6/30 [00:22<01:30,  3.78s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.80s/it]
 30%|███       | 9/30 [00:33<01:19,  3.80s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.81s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.81s/it]
 40%|████      | 12/30 [00:45<01:08,  3.80s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.80s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.80s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.80s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.82s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.82s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.82s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.82s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.81s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+Saving full-length videos (248 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000009/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000009/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000009/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000010.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real Scandinavian-style kitchen with chalk white walls and warm cream-yellow vinyl flooring; a tall stainless steel single-door fridge with a stainless steel handle is the only appliance visible, plain white skirting board, soft diffuse daylight. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000010/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000010/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000010/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000010/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.82s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.28s/it]
 10%|█         | 3/30 [00:10<01:32,  3.43s/it]
 13%|█▎        | 4/30 [00:13<01:31,  3.51s/it]
 17%|█▋        | 5/30 [00:17<01:28,  3.54s/it]
 20%|██        | 6/30 [00:20<01:25,  3.55s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.56s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:14,  3.57s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.57s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:04,  3.58s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.59s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.59s/it]
 50%|█████     | 15/30 [00:53<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.59s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.59s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.59s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.59s/it]
 67%|██████▋   | 20/30 [01:11<00:35,  3.60s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.60s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.60s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.60s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:29<00:17,  3.60s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.60s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.60s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.60s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.60s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.83s/it]
 20%|██        | 6/30 [00:22<01:33,  3.88s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.90s/it]
 27%|██▋       | 8/30 [00:30<01:26,  3.92s/it]
 30%|███       | 9/30 [00:34<01:22,  3.92s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.90s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.89s/it]
 40%|████      | 12/30 [00:45<01:09,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.86s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.72s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.77s/it]
 20%|██        | 6/30 [00:22<01:31,  3.79s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.90s/it]
 30%|███       | 9/30 [00:34<01:22,  3.91s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.94s/it]
 37%|███▋      | 11/30 [00:42<01:15,  3.96s/it]
 40%|████      | 12/30 [00:46<01:11,  3.96s/it]
 43%|████▎     | 13/30 [00:50<01:07,  3.96s/it]
 47%|████▋     | 14/30 [00:53<01:03,  3.96s/it]
 50%|█████     | 15/30 [00:57<00:59,  3.94s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.93s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.92s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.92s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.93s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:37<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:14,  3.94s/it]
 40%|████      | 12/30 [00:45<01:11,  3.95s/it]
 43%|████▎     | 13/30 [00:49<01:07,  3.94s/it]
 47%|████▋     | 14/30 [00:53<01:03,  3.94s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.93s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.89s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.93s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000010/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000010/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000010/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000011.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real industrial loft kitchen with light cool gray concrete-look walls and a polished light gray concrete floor; a tall stainless steel single-door fridge with a stainless steel handle stands flush to the wall, no upper cabinets in frame, neutral overhead lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000011/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000011/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000011/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000011/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.23s/it]
 10%|█         | 3/30 [00:09<01:31,  3.38s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.45s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:15,  3.58s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.59s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.60s/it]
 40%|████      | 12/30 [00:42<01:04,  3.60s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.61s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.62s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.63s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.63s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.63s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.63s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.63s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.63s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.63s/it]
 73%|███████▎  | 22/30 [01:18<00:29,  3.63s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.63s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.63s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.63s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.63s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.64s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.64s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.64s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.91s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.92s/it]
 93%|███████���█▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.05s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.93s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.93s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.92s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.88s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.91s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.91s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000011/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000011/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000011/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b2_ep00000012.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real compact apartment kitchen with smooth eggshell white walls and a light beige-yellow LVP floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, no other appliances visible, gentle natural light from off-camera. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000012/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000012/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000012/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b2/h5_v4/episode_00000012/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.83s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.23s/it]
 10%|█         | 3/30 [00:09<01:30,  3.37s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.43s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.46s/it]
 20%|██        | 6/30 [00:20<01:23,  3.48s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.49s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.50s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.51s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.53s/it]
 40%|████      | 12/30 [00:41<01:03,  3.53s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.54s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.54s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.54s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.54s/it]
 57%|█████▋    | 17/30 [00:59<00:46,  3.56s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.56s/it]
 63%|██████▎   | 19/30 [01:06<00:39,  3.57s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.57s/it]
 70%|███████   | 21/30 [01:13<00:32,  3.57s/it]
 73%|███████▎  | 22/30 [01:17<00:28,  3.58s/it]
 77%|███████▋  | 23/30 [01:20<00:25,  3.58s/it]
 80%|████████  | 24/30 [01:24<00:21,  3.58s/it]
 83%|████████▎ | 25/30 [01:28<00:17,  3.59s/it]
 87%|████████▋ | 26/30 [01:31<00:14,  3.58s/it]
 90%|█████████ | 27/30 [01:35<00:10,  3.58s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.57s/it]
 97%|█████████▋| 29/30 [01:42<00:03,  3.57s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.71s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.74s/it]
 20%|██        | 6/30 [00:22<01:30,  3.77s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.78s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.79s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.80s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.82s/it]
 47%|████▋     | 14/30 [00:52<01:01,  3.84s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.83s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.84s/it]
 57%|█████▋    | 17/30 [01:04<00:49,  3.84s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.84s/it]
 63%|██████▎   | 19/30 [01:11<00:42,  3.83s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.82s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.83s/it]
 73%|███████▎  | 22/30 [01:23<00:30,  3.83s/it]
 77%|███████▋  | 23/30 [01:27<00:26,  3.82s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.82s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.84s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.84s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.83s/it]
 93%|████████��▎| 28/30 [01:46<00:07,  3.83s/it]
 97%|█████████▋| 29/30 [01:50<00:03,  3.83s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.72s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.75s/it]
 20%|██        | 6/30 [00:22<01:30,  3.78s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.79s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.82s/it]
 40%|████      | 12/30 [00:45<01:08,  3.81s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.81s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.81s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.80s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:04<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.80s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.81s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.98s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.63s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.71s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.75s/it]
 20%|██        | 6/30 [00:22<01:30,  3.77s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.80s/it]
 30%|███       | 9/30 [00:33<01:20,  3.81s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.82s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.82s/it]
 40%|████      | 12/30 [00:45<01:09,  3.84s/it]
 43%|████▎     | 13/30 [00:48<01:05,  3.83s/it]
 47%|████▋     | 14/30 [00:52<01:01,  3.85s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.84s/it]
 57%|█████▋    | 17/30 [01:04<00:49,  3.83s/it]
 60%|██████    | 18/30 [01:08<00:45,  3.83s/it]
 63%|██████▎   | 19/30 [01:11<00:42,  3.82s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.82s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.82s/it]
 73%|███████▎  | 22/30 [01:23<00:30,  3.81s/it]
 77%|███████▋  | 23/30 [01:27<00:26,  3.81s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.81s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.81s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.81s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.81s/it]
 93%|█████████▎| 28/30 [01:46<00:07,  3.81s/it]
 97%|█████████▋| 29/30 [01:50<00:03,  3.81s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000012/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000012/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b2_ep00000012/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000000.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real small contemporary kitchen with pale dove gray walls and a white matte tile floor; a tall stainless steel single-door fridge with a stainless steel pull handle sits centered against the wall, empty surroundings, soft even fill light. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000000/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000000/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000000/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000000/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:22,  2.84s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.27s/it]
 10%|█         | 3/30 [00:09<01:31,  3.40s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.50s/it]
 17%|█▋        | 5/30 [00:17<01:28,  3.53s/it]
 20%|██        | 6/30 [00:20<01:25,  3.55s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.56s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.58s/it]
 30%|███       | 9/30 [00:31<01:15,  3.58s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.59s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.59s/it]
 40%|████      | 12/30 [00:42<01:04,  3.59s/it]
 43%|████▎     | 13/30 [00:46<01:01,  3.61s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.60s/it]
 50%|█████     | 15/30 [00:53<00:53,  3.60s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.59s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.60s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.60s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.62s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.62s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.62s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.61s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.61s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.61s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.60s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.61s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.61s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.62s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.62s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.56s/it]
 10%|█         | 3/30 [00:10<01:40,  3.73s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.80s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.92s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.92s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.10s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.55s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:14,  3.90s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.92s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.93s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.93s/it]
 57%|█████▋    | 17/30 [01:05<00:51,  3.92s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.92s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.93s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.93s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.92s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000000/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000000/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000000/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000001.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real bright open-plan kitchen with light gray painted walls and a creamy off-white engineered wood floor; a stainless steel single-door fridge with a stainless steel handle stands alone, no decor on walls, soft balanced daylight illumination. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000001/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000001/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000001/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000001/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.80s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.40s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.47s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.52s/it]
 20%|██        | 6/30 [00:20<01:24,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.57s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:05,  3.61s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.64s/it]
 47%|████▋     | 14/30 [00:49<00:58,  3.63s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.64s/it]
 53%|█████▎    | 16/30 [00:56<00:51,  3.66s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.67s/it]
 60%|██████    | 18/30 [01:04<00:44,  3.69s/it]
 63%|██████▎   | 19/30 [01:08<00:40,  3.69s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.68s/it]
 70%|███████   | 21/30 [01:15<00:33,  3.67s/it]
 73%|███████▎  | 22/30 [01:19<00:29,  3.66s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.65s/it]
 80%|████████  | 24/30 [01:26<00:21,  3.65s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.64s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.65s/it]
 90%|█████████ | 27/30 [01:37<00:10,  3.65s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.64s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.63s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.48s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.73s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.77s/it]
 20%|██        | 6/30 [00:22<01:31,  3.80s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.82s/it]
 27%|██▋       | 8/30 [00:29<01:24,  3.83s/it]
 30%|███       | 9/30 [00:33<01:20,  3.84s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.85s/it]
 40%|████      | 12/30 [00:45<01:09,  3.85s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.85s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:47,  3.97s/it]
 63%|██████▎   | 19/30 [01:12<00:43,  3.96s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.94s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.94s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.93s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.93s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.72s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.77s/it]
 20%|██        | 6/30 [00:22<01:31,  3.79s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.81s/it]
 27%|██▋       | 8/30 [00:29<01:24,  3.82s/it]
 30%|███       | 9/30 [00:33<01:20,  3.83s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.84s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.85s/it]
 40%|████      | 12/30 [00:45<01:09,  3.85s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.85s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.85s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.73s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.93s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.94s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000001/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000001/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000001/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000002.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real neutral utility kitchen with smooth warm white walls and a light yellow linoleum floor; a stainless steel single-door fridge with a stainless steel slab handle is the focal point, plain wall behind it, even shadowless lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000002/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000002/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000002/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000002/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:35,  3.41s/it]
 10%|█         | 3/30 [00:10<01:36,  3.56s/it]
 13%|█▎        | 4/30 [00:14<01:32,  3.57s/it]
 17%|█▋        | 5/30 [00:17<01:29,  3.58s/it]
 20%|██        | 6/30 [00:21<01:26,  3.59s/it]
 23%|██▎       | 7/30 [00:24<01:22,  3.59s/it]
 27%|██▋       | 8/30 [00:28<01:19,  3.59s/it]
 30%|███       | 9/30 [00:32<01:15,  3.60s/it]
 33%|███▎      | 10/30 [00:35<01:12,  3.60s/it]
 37%|███▋      | 11/30 [00:39<01:08,  3.61s/it]
 40%|████      | 12/30 [00:42<01:05,  3.61s/it]
 43%|████▎     | 13/30 [00:46<01:01,  3.61s/it]
 47%|████▋     | 14/30 [00:50<00:57,  3.62s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.63s/it]
 53%|█████▎    | 16/30 [00:57<00:50,  3.63s/it]
 57%|█████▋    | 17/30 [01:01<00:47,  3.65s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.66s/it]
 63%|██████▎   | 19/30 [01:08<00:40,  3.67s/it]
 67%|██████▋   | 20/30 [01:12<00:36,  3.67s/it]
 70%|███████   | 21/30 [01:15<00:32,  3.66s/it]
 73%|███████▎  | 22/30 [01:19<00:29,  3.66s/it]
 77%|███████▋  | 23/30 [01:23<00:25,  3.68s/it]
 80%|████████  | 24/30 [01:26<00:22,  3.69s/it]
 83%|████████▎ | 25/30 [01:30<00:18,  3.69s/it]
 87%|████████▋ | 26/30 [01:34<00:14,  3.68s/it]
 90%|█████████ | 27/30 [01:37<00:11,  3.67s/it]
 93%|█████████▎| 28/30 [01:41<00:07,  3.67s/it]
 97%|█████████▋| 29/30 [01:45<00:03,  3.67s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.09s/it]
  7%|▋         | 2/30 [00:06<01:40,  3.57s/it]
 10%|█         | 3/30 [00:10<01:40,  3.71s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.79s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.87s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.88s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.89s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.90s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.90s/it]
 40%|████      | 12/30 [00:46<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:30,  3.10s/it]
  7%|▋         | 2/30 [00:07<01:40,  3.59s/it]
 10%|█         | 3/30 [00:10<01:40,  3.74s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.79s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.85s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:42<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.91s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.92s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.92s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.93s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.93s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.93s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.92s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.95s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.95s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:40,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|���▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.85s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.91s/it]
 40%|████      | 12/30 [00:45<01:10,  3.90s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.92s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.92s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.93s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.94s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.94s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.93s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.94s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.95s/it]
 83%|████████▎ | 25/30 [01:37<00:19,  3.94s/it]
 87%|████████▋ | 26/30 [01:41<00:15,  3.94s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.94s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.94s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.94s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000002/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000002/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000002/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000003.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real minimalist studio kitchen with light gray plaster walls and a pale yellow oak engineered floor; a tall stainless steel single-door fridge with stainless steel handle stands solo against the wall, no upper cabinets, soft daylight from a side window. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000003/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000003/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000003/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000003/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.81s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.39s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.46s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.49s/it]
 20%|██        | 6/30 [00:20<01:24,  3.53s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.57s/it]
 30%|███       | 9/30 [00:31<01:14,  3.57s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.60s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.61s/it]
 40%|████      | 12/30 [00:42<01:05,  3.62s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.62s/it]
 47%|████▋     | 14/30 [00:49<00:58,  3.63s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.63s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.62s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.61s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.60s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:11<00:35,  3.59s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.59s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.59s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.59s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:29<00:17,  3.60s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.60s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.60s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.60s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.60s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.05s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:09,  3.88s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.82s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:33<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.89s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000003/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000003/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000003/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000004.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern home kitchen with crisp white painted walls and a smooth light gray ceramic tile floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, surrounding area empty and uncluttered, neutral white lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000004/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000004/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000004/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000004/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:22,  2.83s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.23s/it]
 10%|█         | 3/30 [00:09<01:30,  3.36s/it]
 13%|█▎        | 4/30 [00:13<01:28,  3.42s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.45s/it]
 20%|██        | 6/30 [00:20<01:23,  3.47s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.49s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.50s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.52s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.54s/it]
 40%|████      | 12/30 [00:41<01:03,  3.54s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.55s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.55s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.56s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.56s/it]
 57%|█████▋    | 17/30 [00:59<00:46,  3.55s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.55s/it]
 63%|██████▎   | 19/30 [01:06<00:38,  3.54s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.54s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.54s/it]
 73%|███████▎  | 22/30 [01:17<00:28,  3.54s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.53s/it]
 80%|████████  | 24/30 [01:24<00:21,  3.53s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.53s/it]
 87%|████████▋ | 26/30 [01:31<00:14,  3.53s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.53s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.53s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.53s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.62s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.71s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.74s/it]
 20%|██        | 6/30 [00:22<01:30,  3.77s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.78s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.79s/it]
 30%|███       | 9/30 [00:33<01:20,  3.82s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.83s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.84s/it]
 40%|████      | 12/30 [00:45<01:08,  3.83s/it]
 43%|████▎     | 13/30 [00:48<01:05,  3.83s/it]
 47%|████▋     | 14/30 [00:52<01:01,  3.84s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.85s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.84s/it]
 57%|█████▋    | 17/30 [01:04<00:49,  3.84s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.84s/it]
 63%|██████▎   | 19/30 [01:11<00:42,  3.84s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.83s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.84s/it]
 73%|███████▎  | 22/30 [01:23<00:30,  3.84s/it]
 77%|███████▋  | 23/30 [01:27<00:26,  3.84s/it]
 80%|████████  | 24/30 [01:31<00:23,  3.84s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.83s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.82s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.82s/it]
 93%|█████████��| 28/30 [01:46<00:07,  3.82s/it]
 97%|█████████▋| 29/30 [01:50<00:03,  3.82s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.98s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:37,  3.62s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.69s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.73s/it]
 20%|██        | 6/30 [00:22<01:30,  3.77s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.78s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.79s/it]
 30%|███       | 9/30 [00:33<01:20,  3.82s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.83s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.82s/it]
 40%|████      | 12/30 [00:44<01:08,  3.81s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.80s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.81s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.81s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.07s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.71s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.74s/it]
 20%|██        | 6/30 [00:22<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.80s/it]
 30%|███       | 9/30 [00:33<01:19,  3.80s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.81s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.82s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:48<01:05,  3.84s/it]
 47%|████▋     | 14/30 [00:52<01:01,  3.83s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.83s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.82s/it]
 57%|█████▋    | 17/30 [01:04<00:49,  3.82s/it]
 60%|██████    | 18/30 [01:08<00:45,  3.82s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.82s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.81s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.81s/it]
 73%|███████▎  | 22/30 [01:23<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:27<00:26,  3.81s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.81s/it]
 90%|█████████ | 27/30 [01:42<00:11,  3.81s/it]
 93%|█████████▎| 28/30 [01:46<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.81s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000004/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000004/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000004/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000005.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern minimalist kitchen with smooth matte light gray walls and a light yellow tile floor; a tall stainless steel single-door fridge with a stainless steel handle stands alone against the wall, no other appliances or clutter in frame, soft daylight from above. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000005/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000005/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000005/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000005/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:23,  2.88s/it]
  7%|▋         | 2/30 [00:06<01:32,  3.29s/it]
 10%|█         | 3/30 [00:10<01:32,  3.42s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.47s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.51s/it]
 20%|██        | 6/30 [00:20<01:24,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:14,  3.57s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.58s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.59s/it]
 40%|████      | 12/30 [00:42<01:04,  3.60s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.60s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.60s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.60s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.61s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.62s/it]
 63%|██████▎   | 19/30 [01:07<00:40,  3.65s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.69s/it]
 70%|███████   | 21/30 [01:15<00:33,  3.70s/it]
 73%|███████▎  | 22/30 [01:19<00:29,  3.73s/it]
 77%|███████▋  | 23/30 [01:22<00:26,  3.75s/it]
 80%|████████  | 24/30 [01:26<00:22,  3.73s/it]
 83%|████████▎ | 25/30 [01:30<00:18,  3.70s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.68s/it]
 90%|█████████ | 27/30 [01:37<00:11,  3.67s/it]
 93%|█████████▎| 28/30 [01:41<00:07,  3.66s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.65s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.96s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.96s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.99s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.98s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.98s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.97s/it]
 83%|████████▎ | 25/30 [01:37<00:19,  3.99s/it]
 87%|████████▋ | 26/30 [01:41<00:15,  3.96s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.94s/it]
 93%|██████��██▎| 28/30 [01:48<00:07,  3.92s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.99s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.49s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.73s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.80s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:29<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.89s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.94s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.93s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000005/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000005/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000005/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000006.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real clean residential kitchen with painted white walls and a wide-plank pale gray laminate floor; a tall stainless steel single-door fridge with a stainless steel slab handle is centered against the wall, bare countertops in the periphery, even ambient lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000006/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000006/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000006/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000006/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.81s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.38s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.45s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.54s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.57s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.58s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.58s/it]
 40%|████      | 12/30 [00:42<01:04,  3.59s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.60s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.60s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.61s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.62s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.62s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.62s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.62s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.63s/it]
 73%|███████▎  | 22/30 [01:18<00:29,  3.63s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.63s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.63s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.63s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.63s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.64s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.64s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.64s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:40,  3.71s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.85s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|███████��█▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.90s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.92s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.92s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.92s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.91s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.80s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.89s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.89s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.91s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.91s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.92s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.93s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.93s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000006/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000006/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000006/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000007.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real Scandinavian-style kitchen with chalk white walls and warm cream-yellow vinyl flooring; a tall stainless steel single-door fridge with a stainless steel handle is the only appliance visible, plain white skirting board, soft diffuse daylight. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000007/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000007/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000007/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000007/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:23,  2.88s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.28s/it]
 10%|█         | 3/30 [00:10<01:32,  3.41s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.48s/it]
 17%|█▋        | 5/30 [00:17<01:28,  3.53s/it]
 20%|██        | 6/30 [00:20<01:25,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:15,  3.58s/it]
 33%|███▎      | 10/30 [00:35<01:12,  3.61s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.63s/it]
 40%|████      | 12/30 [00:42<01:05,  3.62s/it]
 43%|████▎     | 13/30 [00:46<01:01,  3.61s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.61s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.62s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.61s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.61s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.60s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.61s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.61s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.61s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.61s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.61s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.61s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.61s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.61s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.63s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.89s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.89s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.92s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:32,  3.86s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.89s/it]
 30%|███       | 9/30 [00:34<01:22,  3.90s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.91s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.91s/it]
 40%|████      | 12/30 [00:45<01:10,  3.90s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.87s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.78s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:32,  3.85s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.86s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.88s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000007/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000007/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000007/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000008.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real industrial loft kitchen with light cool gray concrete-look walls and a polished light gray concrete floor; a tall stainless steel single-door fridge with a stainless steel handle stands flush to the wall, no upper cabinets in frame, neutral overhead lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000008/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000008/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000008/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000008/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:29,  3.20s/it]
 10%|█         | 3/30 [00:09<01:30,  3.34s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.43s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.47s/it]
 20%|██        | 6/30 [00:20<01:23,  3.50s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.50s/it]
 27%|██▋       | 8/30 [00:27<01:17,  3.51s/it]
 30%|███       | 9/30 [00:31<01:13,  3.52s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.54s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.56s/it]
 40%|████      | 12/30 [00:41<01:04,  3.57s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.56s/it]
 47%|████▋     | 14/30 [00:48<00:57,  3.57s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.56s/it]
 53%|█████▎    | 16/30 [00:56<00:49,  3.56s/it]
 57%|█████▋    | 17/30 [00:59<00:46,  3.56s/it]
 60%|██████    | 18/30 [01:03<00:42,  3.55s/it]
 63%|██████▎   | 19/30 [01:06<00:39,  3.55s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.54s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.54s/it]
 73%|███████▎  | 22/30 [01:17<00:28,  3.53s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.53s/it]
 80%|████████  | 24/30 [01:24<00:21,  3.53s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.53s/it]
 87%|████████▋ | 26/30 [01:31<00:14,  3.53s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.53s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.53s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.54s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.47s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:40<01:12,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.81s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.82s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.81s/it]
 93%|███████���█▎| 28/30 [01:45<00:07,  3.81s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.81s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.75s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.79s/it]
 30%|███       | 9/30 [00:33<01:19,  3.81s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.81s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.81s/it]
 40%|████      | 12/30 [00:44<01:08,  3.81s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.80s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.80s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.80s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.80s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.81s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.83s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.83s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.82s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.48s/it]
 10%|█         | 3/30 [00:10<01:38,  3.65s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.72s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.75s/it]
 20%|██        | 6/30 [00:22<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.78s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.80s/it]
 30%|███       | 9/30 [00:33<01:19,  3.81s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.81s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.81s/it]
 40%|████      | 12/30 [00:44<01:08,  3.81s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.81s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.80s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.80s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.81s/it]
 57%|█████▋    | 17/30 [01:04<00:49,  3.81s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.80s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.80s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.81s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000008/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000008/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000008/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000009.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real compact apartment kitchen with smooth eggshell white walls and a light beige-yellow LVP floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, no other appliances visible, gentle natural light from off-camera. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 298 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000009/ext1/depth/original_depth.mp4
+  ext1 mask: 298 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000009/ext1/mask/rgb_robot.mp4
+  wrist depth: 298 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000009/wrist/depth/original_depth.mp4
+  wrist mask: 298 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000009/wrist/mask/rgb_robot.mp4
+Total source frames: 298
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.79s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.22s/it]
 10%|█         | 3/30 [00:09<01:30,  3.36s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.44s/it]
 17%|█▋        | 5/30 [00:16<01:27,  3.48s/it]
 20%|██        | 6/30 [00:20<01:24,  3.51s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.53s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:41<01:04,  3.57s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.58s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.58s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.58s/it]
 57%|█████▋    | 17/30 [00:59<00:46,  3.60s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.60s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.60s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.60s/it]
 73%|███████▎  | 22/30 [01:17<00:28,  3.61s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.63s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.64s/it]
 83%|████████▎ | 25/30 [01:28<00:18,  3.64s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.66s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.64s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.63s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.62s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.85s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.89s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.91s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.91s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.91s/it]
 93%|████████��▎| 28/30 [01:47<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.73s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.73s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.82s/it]
 27%|██▋       | 8/30 [00:29<01:24,  3.83s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.87s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.87s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+Saving full-length videos (298 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000009/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000009/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000009/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b3_ep00000010.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real small contemporary kitchen with pale dove gray walls and a white matte tile floor; a tall stainless steel single-door fridge with a stainless steel pull handle sits centered against the wall, empty surroundings, soft even fill light. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000010/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000010/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000010/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b3/h5_v4/episode_00000010/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.79s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.38s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.47s/it]
 17%|█▋        | 5/30 [00:17<01:29,  3.56s/it]
 20%|██        | 6/30 [00:20<01:25,  3.58s/it]
 23%|██▎       | 7/30 [00:24<01:23,  3.61s/it]
 27%|██▋       | 8/30 [00:28<01:19,  3.60s/it]
 30%|███       | 9/30 [00:31<01:15,  3.60s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.60s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.60s/it]
 40%|████      | 12/30 [00:42<01:04,  3.60s/it]
 43%|████▎     | 13/30 [00:46<01:01,  3.60s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.61s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.63s/it]
 53%|█████▎    | 16/30 [00:57<00:50,  3.62s/it]
 57%|█████▋    | 17/30 [01:00<00:47,  3.63s/it]
 60%|██████    | 18/30 [01:04<00:43,  3.63s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.63s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.63s/it]
 70%|███████   | 21/30 [01:15<00:32,  3.63s/it]
 73%|███████▎  | 22/30 [01:18<00:29,  3.64s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.63s/it]
 80%|████████  | 24/30 [01:26<00:21,  3.63s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.64s/it]
 87%|████████▋ | 26/30 [01:33<00:14,  3.64s/it]
 90%|█████████ | 27/30 [01:37<00:10,  3.64s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.65s/it]
 97%|█████████▋| 29/30 [01:44<00:03,  3.64s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.05s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:26,  3.92s/it]
 30%|███       | 9/30 [00:34<01:22,  3.91s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.91s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.91s/it]
 40%|████      | 12/30 [00:46<01:10,  3.92s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.93s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.92s/it]
 57%|█████▋    | 17/30 [01:05<00:51,  3.94s/it]
 60%|██████    | 18/30 [01:09<00:47,  3.94s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.93s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.94s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.93s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.93s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:37<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:41<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:14,  3.90s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.92s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.92s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.92s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.93s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.93s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.93s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.92s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.92s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.83s/it]
 20%|██        | 6/30 [00:22<01:32,  3.85s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:26,  3.91s/it]
 30%|███       | 9/30 [00:34<01:22,  3.91s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.91s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.91s/it]
 40%|████      | 12/30 [00:46<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.91s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.92s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.91s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.93s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.94s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.93s/it]
 77%|███████▋  | 23/30 [01:29<00:27,  3.94s/it]
 80%|████████  | 24/30 [01:33<00:23,  3.93s/it]
 83%|████████▎ | 25/30 [01:37<00:19,  3.95s/it]
 87%|████████▋ | 26/30 [01:41<00:15,  3.95s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.94s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.95s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.95s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000010/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000010/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b3_ep00000010/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000000.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real bright open-plan kitchen with light gray painted walls and a creamy off-white engineered wood floor; a stainless steel single-door fridge with a stainless steel handle stands alone, no decor on walls, soft balanced daylight illumination. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000000/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000000/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000000/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000000/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:29,  3.20s/it]
 10%|█         | 3/30 [00:09<01:30,  3.35s/it]
 13%|█▎        | 4/30 [00:13<01:28,  3.41s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.45s/it]
 20%|██        | 6/30 [00:20<01:23,  3.47s/it]
 23%|██▎       | 7/30 [00:23<01:20,  3.48s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.51s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.51s/it]
 37%|███▋      | 11/30 [00:37<01:06,  3.51s/it]
 40%|████      | 12/30 [00:41<01:03,  3.52s/it]
 43%|████▎     | 13/30 [00:44<00:59,  3.52s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.52s/it]
 50%|█████     | 15/30 [00:52<00:52,  3.52s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.52s/it]
 57%|█████▋    | 17/30 [00:59<00:45,  3.52s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.52s/it]
 63%|██████▎   | 19/30 [01:06<00:38,  3.52s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.52s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.52s/it]
 73%|███████▎  | 22/30 [01:16<00:28,  3.53s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.53s/it]
 80%|████████  | 24/30 [01:23<00:21,  3.53s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.52s/it]
 87%|████████▋ | 26/30 [01:30<00:14,  3.53s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.53s/it]
 93%|█████████▎| 28/30 [01:37<00:07,  3.53s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.53s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.62s/it]
 13%|█▎        | 4/30 [00:14<01:36,  3.71s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.76s/it]
 20%|██        | 6/30 [00:22<01:30,  3.77s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.80s/it]
 30%|███       | 9/30 [00:33<01:19,  3.80s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.80s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.69s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.73s/it]
 20%|██        | 6/30 [00:21<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.80s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.80s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.80s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.80s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:25,  2.96s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.44s/it]
 10%|█         | 3/30 [00:10<01:37,  3.60s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.67s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.71s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.75s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.79s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.78s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.78s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.78s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.78s/it]
 53%|█████▎    | 16/30 [00:59<00:52,  3.78s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.78s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.78s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.78s/it]
 67%|██████▋   | 20/30 [01:14<00:37,  3.78s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.78s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.78s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:33<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.78s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.78s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.78s/it]
 97%|█████████▋| 29/30 [01:48<00:03,  3.78s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000000/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000000/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000000/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000001.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real neutral utility kitchen with smooth warm white walls and a light yellow linoleum floor; a stainless steel single-door fridge with a stainless steel slab handle is the focal point, plain wall behind it, even shadowless lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000001/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000001/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000001/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000001/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.82s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.26s/it]
 10%|█         | 3/30 [00:09<01:31,  3.40s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.47s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.53s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.54s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.57s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.57s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:04,  3.57s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.58s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.59s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.60s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.60s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.60s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.60s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.60s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.60s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:28<00:17,  3.60s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.60s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.62s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.63s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.64s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.57s/it]
 10%|█         | 3/30 [00:10<01:40,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:39,  3.83s/it]
 17%|█▋        | 5/30 [00:18<01:36,  3.84s/it]
 20%|██        | 6/30 [00:22<01:32,  3.86s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.87s/it]
 30%|███       | 9/30 [00:34<01:21,  3.87s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.91s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.56s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.91s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.91s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.56s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.77s/it]
 17%|���▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:20,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000001/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000001/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000001/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000002.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real minimalist studio kitchen with light gray plaster walls and a pale yellow oak engineered floor; a tall stainless steel single-door fridge with stainless steel handle stands solo against the wall, no upper cabinets, soft daylight from a side window. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000002/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000002/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000002/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000002/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.82s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.38s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.45s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.54s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.57s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:04,  3.58s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.59s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.59s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.59s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.59s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.60s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.60s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.60s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.60s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.61s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.61s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.63s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.62s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.62s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.61s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.61s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:31<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:27,  3.82s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.92s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.92s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.92s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.94s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.92s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.00s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:38,  3.66s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.86s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.89s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.92s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.89s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000002/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000002/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000002/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000003.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern home kitchen with crisp white painted walls and a smooth light gray ceramic tile floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, surrounding area empty and uncluttered, neutral white lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000003/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000003/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000003/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000003/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:22,  2.86s/it]
  7%|▋         | 2/30 [00:06<01:31,  3.26s/it]
 10%|█         | 3/30 [00:09<01:31,  3.39s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.46s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.54s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:14,  3.57s/it]
 33%|███▎      | 10/30 [00:35<01:11,  3.59s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.60s/it]
 40%|████      | 12/30 [00:42<01:04,  3.61s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.61s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.61s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.60s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.61s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.61s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.61s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.62s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.62s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.62s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.62s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.62s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.62s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.62s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.62s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.62s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.62s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.89s/it]
 30%|███       | 9/30 [00:34<01:22,  3.91s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.91s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.91s/it]
 40%|████      | 12/30 [00:45<01:10,  3.90s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.89s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.89s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.89s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.89s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.89s/it]
 93%|█████████��| 28/30 [01:48<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.55s/it]
 10%|█         | 3/30 [00:10<01:40,  3.72s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.79s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.85s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.91s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.92s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.92s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.91s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.91s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:38,  3.90s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.89s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.89s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.08s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.89s/it]
 30%|███       | 9/30 [00:34<01:21,  3.90s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.90s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.89s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.89s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.89s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.89s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.89s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.91s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.91s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.93s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.93s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.93s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.93s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.93s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.93s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000003/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000003/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000003/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000004.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real modern minimalist kitchen with smooth matte light gray walls and a light yellow tile floor; a tall stainless steel single-door fridge with a stainless steel handle stands alone against the wall, no other appliances or clutter in frame, soft daylight from above. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000004/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000004/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000004/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000004/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.78s/it]
  7%|▋         | 2/30 [00:06<01:29,  3.20s/it]
 10%|█         | 3/30 [00:09<01:30,  3.34s/it]
 13%|█▎        | 4/30 [00:13<01:28,  3.40s/it]
 17%|█▋        | 5/30 [00:16<01:25,  3.44s/it]
 20%|██        | 6/30 [00:20<01:23,  3.46s/it]
 23%|██▎       | 7/30 [00:23<01:19,  3.48s/it]
 27%|██▋       | 8/30 [00:27<01:17,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.51s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.52s/it]
 37%|███▋      | 11/30 [00:37<01:06,  3.53s/it]
 40%|████      | 12/30 [00:41<01:03,  3.53s/it]
 43%|████▎     | 13/30 [00:45<00:59,  3.52s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.53s/it]
 50%|█████     | 15/30 [00:52<00:52,  3.53s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.53s/it]
 57%|█████▋    | 17/30 [00:59<00:45,  3.53s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.53s/it]
 63%|██████▎   | 19/30 [01:06<00:39,  3.55s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.54s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.54s/it]
 73%|███████▎  | 22/30 [01:16<00:28,  3.54s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.53s/it]
 80%|████████  | 24/30 [01:23<00:21,  3.53s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.53s/it]
 87%|████████▋ | 26/30 [01:30<00:14,  3.53s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.53s/it]
 93%|█████████▎| 28/30 [01:38<00:07,  3.53s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.53s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.98s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.48s/it]
 10%|█         | 3/30 [00:10<01:37,  3.62s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:30,  3.76s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.79s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.80s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.80s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.80s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.78s/it]
 93%|██████��██▎| 28/30 [01:45<00:07,  3.78s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.60s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:27,  3.79s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.80s/it]
 30%|███       | 9/30 [00:33<01:19,  3.80s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.82s/it]
 37%|███▋      | 11/30 [00:41<01:12,  3.83s/it]
 40%|████      | 12/30 [00:44<01:08,  3.82s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.81s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.80s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.80s/it]
 53%|█████▎    | 16/30 [01:00<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.80s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.79s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.79s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.79s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.77s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.80s/it]
 37%|███▋      | 11/30 [00:40<01:12,  3.80s/it]
 40%|████      | 12/30 [00:44<01:08,  3.80s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.80s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.79s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.79s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.79s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.79s/it]
 83%|████████▎ | 25/30 [01:34<00:19,  3.80s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.79s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000004/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000004/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000004/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000005.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real clean residential kitchen with painted white walls and a wide-plank pale gray laminate floor; a tall stainless steel single-door fridge with a stainless steel slab handle is centered against the wall, bare countertops in the periphery, even ambient lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000005/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000005/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000005/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000005/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.81s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.40s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.46s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.53s/it]
 27%|██▋       | 8/30 [00:27<01:17,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.58s/it]
 40%|████      | 12/30 [00:42<01:04,  3.58s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.59s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.59s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.59s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.60s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.60s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.60s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.60s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.60s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:28<00:18,  3.62s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.62s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.62s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.62s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.62s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:29,  3.09s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.54s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.86s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.86s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.86s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.86s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.86s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.87s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.87s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.90s/it]
 93%|███████��█▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.89s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.07s/it]
  7%|▋         | 2/30 [00:06<01:39,  3.56s/it]
 10%|█         | 3/30 [00:10<01:39,  3.70s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:34<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.86s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:32,  3.18s/it]
  7%|▋         | 2/30 [00:07<01:40,  3.59s/it]
 10%|█         | 3/30 [00:10<01:40,  3.71s/it]
 13%|█▎        | 4/30 [00:14<01:38,  3.77s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.86s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.86s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.86s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.86s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.87s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.91s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000005/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000005/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000005/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000006.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real Scandinavian-style kitchen with chalk white walls and warm cream-yellow vinyl flooring; a tall stainless steel single-door fridge with a stainless steel handle is the only appliance visible, plain white skirting board, soft diffuse daylight. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000006/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000006/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000006/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000006/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:20,  2.77s/it]
  7%|▋         | 2/30 [00:06<01:29,  3.21s/it]
 10%|█         | 3/30 [00:09<01:30,  3.36s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.43s/it]
 17%|█▋        | 5/30 [00:16<01:26,  3.47s/it]
 20%|██        | 6/30 [00:20<01:24,  3.50s/it]
 23%|██▎       | 7/30 [00:24<01:20,  3.52s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:41<01:04,  3.58s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.59s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.59s/it]
 57%|█████▋    | 17/30 [00:59<00:46,  3.59s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.59s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.59s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.59s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.59s/it]
 73%|███████▎  | 22/30 [01:17<00:28,  3.60s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.61s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:28<00:18,  3.61s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.62s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.63s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.62s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.62s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.50s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.78s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.86s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.86s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.86s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.86s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.86s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.86s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.86s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.89s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.89s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:38,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.74s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:31,  3.81s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.84s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.85s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.85s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:57,  3.87s/it]
 53%|█████▎    | 16/30 [01:00<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.87s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000006/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000006/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000006/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000007.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real industrial loft kitchen with light cool gray concrete-look walls and a polished light gray concrete floor; a tall stainless steel single-door fridge with a stainless steel handle stands flush to the wall, no upper cabinets in frame, neutral overhead lighting. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000007/ext1/depth/original_depth.mp4
+  ext1 mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000007/ext1/mask/rgb_robot.mp4
+  wrist depth: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000007/wrist/depth/original_depth.mp4
+  wrist mask: 301 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000007/wrist/mask/rgb_robot.mp4
+Total source frames: 301
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.80s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.23s/it]
 10%|█         | 3/30 [00:09<01:31,  3.37s/it]
 13%|█▎        | 4/30 [00:13<01:30,  3.47s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.50s/it]
 20%|██        | 6/30 [00:20<01:24,  3.53s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.55s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.56s/it]
 30%|███       | 9/30 [00:31<01:14,  3.57s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.58s/it]
 37%|███▋      | 11/30 [00:38<01:08,  3.58s/it]
 40%|████      | 12/30 [00:42<01:04,  3.59s/it]
 43%|████▎     | 13/30 [00:45<01:01,  3.60s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.60s/it]
 50%|█████     | 15/30 [00:53<00:54,  3.60s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.61s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.61s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.62s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.62s/it]
 67%|██████▋   | 20/30 [01:11<00:36,  3.62s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.62s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.62s/it]
 77%|███████▋  | 23/30 [01:22<00:25,  3.63s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.63s/it]
 83%|████████▎ | 25/30 [01:29<00:18,  3.62s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.62s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.62s/it]
 93%|█████████▎| 28/30 [01:40<00:07,  3.62s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.62s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.79s/it]
 20%|██        | 6/30 [00:22<01:32,  3.84s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.86s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.89s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.89s/it]
 37%|███▋      | 11/30 [00:41<01:14,  3.90s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:20<00:35,  3.90s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|███████���█▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.91s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.01s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.50s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.81s/it]
 20%|██        | 6/30 [00:22<01:32,  3.86s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.88s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.91s/it]
 30%|███       | 9/30 [00:34<01:21,  3.90s/it]
 33%|███▎      | 10/30 [00:38<01:18,  3.90s/it]
 37%|███▋      | 11/30 [00:42<01:14,  3.90s/it]
 40%|████      | 12/30 [00:45<01:10,  3.90s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.91s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.91s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.90s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.90s/it]
 63%|██████▎   | 19/30 [01:13<00:42,  3.90s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.90s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.91s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.90s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.90s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.90s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.90s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.90s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.90s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.90s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.90s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:27,  3.02s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.51s/it]
 10%|█         | 3/30 [00:10<01:39,  3.67s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.82s/it]
 20%|██        | 6/30 [00:22<01:32,  3.85s/it]
 23%|██▎       | 7/30 [00:26<01:29,  3.87s/it]
 27%|██▋       | 8/30 [00:30<01:25,  3.88s/it]
 30%|███       | 9/30 [00:34<01:21,  3.88s/it]
 33%|███▎      | 10/30 [00:38<01:17,  3.88s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.89s/it]
 40%|████      | 12/30 [00:45<01:10,  3.89s/it]
 43%|████▎     | 13/30 [00:49<01:06,  3.90s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.90s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.90s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.90s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.91s/it]
 60%|██████    | 18/30 [01:09<00:46,  3.91s/it]
 63%|██████▎   | 19/30 [01:13<00:43,  3.92s/it]
 67%|██████▋   | 20/30 [01:17<00:39,  3.92s/it]
 70%|███████   | 21/30 [01:21<00:35,  3.92s/it]
 73%|███████▎  | 22/30 [01:25<00:31,  3.93s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.93s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.93s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.93s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.93s/it]
 90%|█████████ | 27/30 [01:44<00:11,  3.93s/it]
 93%|█████████▎| 28/30 [01:48<00:07,  3.93s/it]
 97%|█████████▋| 29/30 [01:52<00:03,  3.93s/it]
+
+
+Saving full-length videos (301 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000007/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000007/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000007/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000008.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real compact apartment kitchen with smooth eggshell white walls and a light beige-yellow LVP floor; a single stainless steel fridge with a stainless steel handle is positioned against the wall, no other appliances visible, gentle natural light from off-camera. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 288 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000008/ext1/depth/original_depth.mp4
+  ext1 mask: 288 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000008/ext1/mask/rgb_robot.mp4
+  wrist depth: 288 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000008/wrist/depth/original_depth.mp4
+  wrist mask: 288 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000008/wrist/mask/rgb_robot.mp4
+Total source frames: 288
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:19,  2.76s/it]
  7%|▋         | 2/30 [00:06<01:29,  3.19s/it]
 10%|█         | 3/30 [00:09<01:30,  3.34s/it]
 13%|█▎        | 4/30 [00:13<01:28,  3.40s/it]
 17%|█▋        | 5/30 [00:16<01:25,  3.44s/it]
 20%|██        | 6/30 [00:20<01:23,  3.46s/it]
 23%|██▎       | 7/30 [00:23<01:19,  3.48s/it]
 27%|██▋       | 8/30 [00:27<01:16,  3.50s/it]
 30%|███       | 9/30 [00:30<01:13,  3.50s/it]
 33%|███▎      | 10/30 [00:34<01:10,  3.51s/it]
 37%|███▋      | 11/30 [00:37<01:06,  3.51s/it]
 40%|████      | 12/30 [00:41<01:03,  3.51s/it]
 43%|████▎     | 13/30 [00:44<00:59,  3.51s/it]
 47%|████▋     | 14/30 [00:48<00:56,  3.52s/it]
 50%|█████     | 15/30 [00:51<00:52,  3.52s/it]
 53%|█████▎    | 16/30 [00:55<00:49,  3.52s/it]
 57%|█████▋    | 17/30 [00:58<00:45,  3.52s/it]
 60%|██████    | 18/30 [01:02<00:42,  3.52s/it]
 63%|██████▎   | 19/30 [01:06<00:38,  3.52s/it]
 67%|██████▋   | 20/30 [01:09<00:35,  3.53s/it]
 70%|███████   | 21/30 [01:13<00:31,  3.53s/it]
 73%|███████▎  | 22/30 [01:16<00:28,  3.53s/it]
 77%|███████▋  | 23/30 [01:20<00:24,  3.53s/it]
 80%|████████  | 24/30 [01:23<00:21,  3.54s/it]
 83%|████████▎ | 25/30 [01:27<00:17,  3.54s/it]
 87%|████████▋ | 26/30 [01:30<00:14,  3.54s/it]
 90%|█████████ | 27/30 [01:34<00:10,  3.53s/it]
 93%|█████████▎| 28/30 [01:37<00:07,  3.53s/it]
 97%|█████████▋| 29/30 [01:41<00:03,  3.53s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.60s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:33,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.75s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:23,  3.78s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:16,  3.81s/it]
 37%|███▋      | 11/30 [00:40<01:12,  3.80s/it]
 40%|████      | 12/30 [00:44<01:08,  3.80s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.80s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.80s/it]
 50%|█████     | 15/30 [00:56<00:57,  3.80s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.80s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.80s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.80s/it]
 67%|██████▋   | 20/30 [01:15<00:38,  3.80s/it]
 70%|███████   | 21/30 [01:19<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.80s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:38<00:15,  3.81s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|████████��▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:26,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:36,  3.45s/it]
 10%|█         | 3/30 [00:10<01:37,  3.60s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.76s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.79s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.79s/it]
 37%|███▋      | 11/30 [00:40<01:12,  3.79s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.80s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.80s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.80s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.80s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.80s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.80s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:25,  2.97s/it]
  7%|▋         | 2/30 [00:06<01:37,  3.46s/it]
 10%|█         | 3/30 [00:10<01:37,  3.61s/it]
 13%|█▎        | 4/30 [00:14<01:35,  3.68s/it]
 17%|█▋        | 5/30 [00:18<01:32,  3.72s/it]
 20%|██        | 6/30 [00:21<01:29,  3.74s/it]
 23%|██▎       | 7/30 [00:25<01:26,  3.75s/it]
 27%|██▋       | 8/30 [00:29<01:22,  3.77s/it]
 30%|███       | 9/30 [00:33<01:19,  3.78s/it]
 33%|███▎      | 10/30 [00:37<01:15,  3.78s/it]
 37%|███▋      | 11/30 [00:40<01:11,  3.78s/it]
 40%|████      | 12/30 [00:44<01:08,  3.79s/it]
 43%|████▎     | 13/30 [00:48<01:04,  3.79s/it]
 47%|████▋     | 14/30 [00:52<01:00,  3.79s/it]
 50%|█████     | 15/30 [00:56<00:56,  3.79s/it]
 53%|█████▎    | 16/30 [00:59<00:53,  3.79s/it]
 57%|█████▋    | 17/30 [01:03<00:49,  3.79s/it]
 60%|██████    | 18/30 [01:07<00:45,  3.79s/it]
 63%|██████▎   | 19/30 [01:11<00:41,  3.79s/it]
 67%|██████▋   | 20/30 [01:15<00:37,  3.80s/it]
 70%|███████   | 21/30 [01:18<00:34,  3.80s/it]
 73%|███████▎  | 22/30 [01:22<00:30,  3.80s/it]
 77%|███████▋  | 23/30 [01:26<00:26,  3.80s/it]
 80%|████████  | 24/30 [01:30<00:22,  3.80s/it]
 83%|████████▎ | 25/30 [01:34<00:18,  3.80s/it]
 87%|████████▋ | 26/30 [01:37<00:15,  3.80s/it]
 90%|█████████ | 27/30 [01:41<00:11,  3.80s/it]
 93%|█████████▎| 28/30 [01:45<00:07,  3.79s/it]
 97%|█████████▋| 29/30 [01:49<00:03,  3.79s/it]
+
+
+Saving full-length videos (288 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000008/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000008/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000008/comparison.mp4
+
+Done!

fridge_m/chunk_008/_logs/b4_ep00000009.log

@@ -0,0 +1,134 @@
+Views: ['ext1', 'wrist']
+Prompt: Video of a real small contemporary kitchen with pale dove gray walls and a white matte tile floor; a tall stainless steel single-door fridge with a stainless steel pull handle sits centered against the wall, empty surroundings, soft even fill light. A Franka Panda robot arm is mounted on a black mobile platform in front of a stainless steel single door refrigerator with a stainless steel handle. The robot has a white-and-black two-finger parallel gripper. The fridge body is bright brushed stainless steel; the door has a flat slab-handle finished in matched stainless steel. Lighting is soft, even, and daylight-balanced; no harsh shadows or color casts on the metal. Photorealistic, sharp focus, no motion blur, no text or watermarks.
+Loading source videos...
+  ext1 depth: 299 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000009/ext1/depth/original_depth.mp4
+  ext1 mask: 299 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000009/ext1/mask/rgb_robot.mp4
+  wrist depth: 299 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000009/wrist/depth/original_depth.mp4
+  wrist mask: 299 frames from /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_depth/b4/h5_v4/episode_00000009/wrist/mask/rgb_robot.mp4
+Total source frames: 299
+Clip length: 81, overlap: 1
+Number of clips: 4
+Loading pipeline...
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors"
+Loaded model: {
+    "model_name": "wan_video_dit",
+    "model_class": "diffsynth.models.wan_video_dit.WanModel",
+    "extra_kwargs": {
+        "has_image_input": true,
+        "patch_size": [
+            1,
+            2,
+            2
+        ],
+        "in_dim": 48,
+        "dim": 1536,
+        "ffn_dim": 8960,
+        "freq_dim": 256,
+        "text_dim": 4096,
+        "out_dim": 16,
+        "num_heads": 12,
+        "num_layers": 30,
+        "eps": 1e-06,
+        "has_ref_conv": true
+    }
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_text_encoder",
+    "model_class": "diffsynth.models.wan_video_text_encoder.WanTextEncoder",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-DualControl-3View-PrevFrame_minimal_lora/step-4740.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+============================================================
+Generating clip 1/4
+  No previous frame (first clip)
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:02<01:21,  2.79s/it]
  7%|▋         | 2/30 [00:06<01:30,  3.24s/it]
 10%|█         | 3/30 [00:09<01:31,  3.39s/it]
 13%|█▎        | 4/30 [00:13<01:29,  3.45s/it]
 17%|█▋        | 5/30 [00:17<01:27,  3.49s/it]
 20%|██        | 6/30 [00:20<01:24,  3.52s/it]
 23%|██▎       | 7/30 [00:24<01:21,  3.53s/it]
 27%|██▋       | 8/30 [00:27<01:18,  3.55s/it]
 30%|███       | 9/30 [00:31<01:14,  3.56s/it]
 33%|███▎      | 10/30 [00:34<01:11,  3.56s/it]
 37%|███▋      | 11/30 [00:38<01:07,  3.57s/it]
 40%|████      | 12/30 [00:42<01:04,  3.57s/it]
 43%|████▎     | 13/30 [00:45<01:00,  3.58s/it]
 47%|████▋     | 14/30 [00:49<00:57,  3.58s/it]
 50%|█████     | 15/30 [00:52<00:53,  3.59s/it]
 53%|█████▎    | 16/30 [00:56<00:50,  3.60s/it]
 57%|█████▋    | 17/30 [01:00<00:46,  3.60s/it]
 60%|██████    | 18/30 [01:03<00:43,  3.59s/it]
 63%|██████▎   | 19/30 [01:07<00:39,  3.60s/it]
 67%|██████▋   | 20/30 [01:10<00:35,  3.60s/it]
 70%|███████   | 21/30 [01:14<00:32,  3.60s/it]
 73%|███████▎  | 22/30 [01:18<00:28,  3.60s/it]
 77%|███████▋  | 23/30 [01:21<00:25,  3.60s/it]
 80%|████████  | 24/30 [01:25<00:21,  3.60s/it]
 83%|████████▎ | 25/30 [01:28<00:17,  3.60s/it]
 87%|████████▋ | 26/30 [01:32<00:14,  3.60s/it]
 90%|█████████ | 27/30 [01:36<00:10,  3.60s/it]
 93%|█████████▎| 28/30 [01:39<00:07,  3.60s/it]
 97%|█████████▋| 29/30 [01:43<00:03,  3.60s/it]
+
+
+============================================================
+Generating clip 2/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.83s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.84s/it]
 30%|███       | 9/30 [00:33<01:20,  3.85s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.86s/it]
 40%|████      | 12/30 [00:45<01:09,  3.86s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.86s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.87s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.87s/it]
 57%|█████▋    | 17/30 [01:04<00:50,  3.87s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.87s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.87s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.87s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.87s/it]
 73%|███████▎  | 22/30 [01:24<00:30,  3.87s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:35<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 3/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.03s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.52s/it]
 10%|█         | 3/30 [00:10<01:39,  3.68s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.75s/it]
 17%|█▋        | 5/30 [00:18<01:34,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.82s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.87s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.87s/it]
 47%|████▋     | 14/30 [00:53<01:01,  3.87s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.89s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.89s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.89s/it]
 87%|████████▋ | 26/30 [01:40<00:15,  3.89s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.89s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+============================================================
+Generating clip 4/4
+  Using previous frame conditioning
+============================================================
+
+
+
+
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:03<01:28,  3.04s/it]
  7%|▋         | 2/30 [00:06<01:38,  3.53s/it]
 10%|█         | 3/30 [00:10<01:39,  3.69s/it]
 13%|█▎        | 4/30 [00:14<01:37,  3.76s/it]
 17%|█▋        | 5/30 [00:18<01:35,  3.80s/it]
 20%|██        | 6/30 [00:22<01:31,  3.83s/it]
 23%|██▎       | 7/30 [00:26<01:28,  3.84s/it]
 27%|██▋       | 8/30 [00:30<01:24,  3.85s/it]
 30%|███       | 9/30 [00:34<01:21,  3.86s/it]
 33%|███▎      | 10/30 [00:37<01:17,  3.86s/it]
 37%|███▋      | 11/30 [00:41<01:13,  3.87s/it]
 40%|████      | 12/30 [00:45<01:09,  3.87s/it]
 43%|████▎     | 13/30 [00:49<01:05,  3.88s/it]
 47%|████▋     | 14/30 [00:53<01:02,  3.88s/it]
 50%|█████     | 15/30 [00:57<00:58,  3.88s/it]
 53%|█████▎    | 16/30 [01:01<00:54,  3.88s/it]
 57%|█████▋    | 17/30 [01:05<00:50,  3.88s/it]
 60%|██████    | 18/30 [01:08<00:46,  3.88s/it]
 63%|██████▎   | 19/30 [01:12<00:42,  3.88s/it]
 67%|██████▋   | 20/30 [01:16<00:38,  3.88s/it]
 70%|███████   | 21/30 [01:20<00:34,  3.88s/it]
 73%|███████▎  | 22/30 [01:24<00:31,  3.88s/it]
 77%|███████▋  | 23/30 [01:28<00:27,  3.88s/it]
 80%|████████  | 24/30 [01:32<00:23,  3.88s/it]
 83%|████████▎ | 25/30 [01:36<00:19,  3.88s/it]
 87%|████████▋ | 26/30 [01:39<00:15,  3.88s/it]
 90%|█████████ | 27/30 [01:43<00:11,  3.88s/it]
 93%|█████████▎| 28/30 [01:47<00:07,  3.88s/it]
 97%|█████████▋| 29/30 [01:51<00:03,  3.88s/it]
+
+
+Saving full-length videos (299 frames)...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  ext1: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000009/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  wrist: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000009/generated_wrist.mp4
+Saving comparison video...
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  Comparison: /scr/ravenh/fridge_m/_workdir_gen/chunk_008/_inf_tmp/b4_ep00000009/comparison.mp4
+
+Done!

fridge_m/chunk_008/generated_episode_00000000_exo_camera_1_batch_1_of_4_cam_rand.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b3a7ef25c7765357eab5417c2f6b2f024de8af0c760b62aed6954682dbe6825
+size 238733

fridge_m/chunk_008/generated_episode_00000000_exo_camera_1_batch_2_of_4_cam_rand.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6de63f05fb58b50624bd390657a2625c0bdbf7059eb3e399c0616bff20562e0c
+size 277745

fridge_m/chunk_008/generated_episode_00000000_exo_camera_1_batch_3_of_4_cam_rand.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35a351c11028d201c46b366b0cf87016fda7de89da2253681a6e575fe977a4ad
+size 226142

fridge_m/chunk_008/generated_episode_00000000_exo_camera_1_batch_4_of_4_cam_rand.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b355f0ec95909722693a42477c87f677d4121c3163dfdd1f0e3d953b64e1c61
+size 236637