Commit History
Yesterday
-
Add MiniMax-H3-Image INT8 ConvRot checkpoints and VAE, update notice
Verified signaturefe905e1 -
Add MiniMax-H3-Image BF16 checkpoints (full and pruned) with license and notice
Verified signatured2fa3bd
2 days ago
-
Add minimax_h3_dmad_4step_full_critic_lora_bf16.safetensors (DMAD 4step full_critic student, ZhengmingYu/DMAD, MiniMax H3 Community License)
Verified signature4e6c06f -
Add minimax_h3_dmad_8step_full_critic_lora_bf16.safetensors (DMAD 8step full_critic student, ZhengmingYu/DMAD, MiniMax H3 Community License)
Verified signatureb5c0e05 -
Move the Viggle Image Swap LoRA to Qwen_image_2
Verified signature008096b -
Add Viggle Image Swap v0.1 LoRA for Qwen Image 2.1
Verified signatureff66d35 -
Open-H3 Regenerate 2K IC LoRA (BF16), HyperFlow 8-step LoRA for pruned H3, LMS Sharp v1.0, notices
Verified signature039c30b
3 days ago
-
pDMD 2-NFE LoRA (unmodified upstream file) with its Apache 2.0 licence and notice
Verified signature3d3c6fb -
Open-H3 Regenerate 2K LoRA in BF16 (was FP32), identical outputs
Verified signature1e90646 -
HyperFlow H3 Flash: low-rank down factors quantized to INT8 ConvRot (smaller files, faster steps); NOTICE updated
Verified signatured1c34b0 -
Add HyperFlow H3 Flash FL2VA and Ref2VA (INT8 ConvRot, repacked for WanGP) with LICENSE and NOTICE
Verified signature9b5fc50 -
Add Singularity V1.3 Ref2VA pruned INT8 ConvRot
Verified signatureWarmBloodAban Singularity V1.3 Ref2VA pruned INT8 (Apache-2.0) with its AdaLN LoRA map and the ConvRot marker in the header. Tensor data unchanged.0962673
7 days ago
-
Add the Viggle Remake swap and edit LoRAs
Verified signatureUnmodified LoRA weights of Viggle/Viggle-Remake at 21665fa7a1e038e0a5312fb15f37b18d0f06f117, renamed, with its LICENSE and NOTICE.1393e9b -
Remove the merged P-H3-Image transformers: the model runs on FL2VA Pruned plus its LoRA
Verified signature936ea57
8 days ago
-
Add loras/p-h3-image_lora_bf16.safetensors
Verified signatureaa89a38 -
Add P-H3-Image-pruned_rank8_bf16.safetensors
Verified signaturedb28236 -
Add P-H3-Image-pruned_rank8_int8_convrot.safetensors
Verified signatured87d3d8 -
Add P-H3-Image-Qwen3-VL-4B/P-H3-Image-Qwen3-VL-4B_bf16.safetensors
Verified signatureac22ab8 -
Add P-H3-Image-Qwen3-VL-4B/P-H3-Image-Qwen3-VL-4B_int8_convrot.safetensors
Verified signature579959e -
Add p_h3_image/p_h3_image_vae_fp16.safetensors
Verified signaturecffdf96 -
Add p_h3_image/p_h3_image_vae_int8_convrot.safetensors
Verified signaturef4a90fc -
Add P-H3-Image-Qwen3-VL-4B/tokenizer_config.json
Verified signaturef5615b9 -
Add P-H3-Image-Qwen3-VL-4B/tokenizer.json
Verified signature7f7fcdc -
Add P-H3-Image-Qwen3-VL-4B/config.json
Verified signaturea1f9101 -
Add p_h3_image/NOTICE
Verified signature63ea914 -
Add p_h3_image/LICENSE-APACHE
Verified signatured687396 -
Add p_h3_image/LICENSE
Verified signature12692f6
11 days ago
-
Add SeedVR2 7B INT8 ConvRot
Verified signature6820fad
13 days ago
-
Add Singularity V2 INT8 ConvRot checkpoints with embedded AdaLN maps
Verified signatureFull and pruned Singularity V2 INT8 ConvRot checkpoints from https://huggingface.co/WarmBloodAban/Minimax-h3_Singularity at revision 1259cf16cd0b655ba691a32fb46a409142762972. Each safetensors header embeds the validated Singularity V2 LoRA/AdaLN conversion map in wangp_h3_lora_affine metadata. Original tensor payloads are unchanged.d0d3fde -
Add Pika H3 Flex FL2VA and Ref2VA BF16 adapters
Verified signatureAdd the released six-step rank-128 adapters in BF16, matching upstream inference precision. Preserve model license and attribution, source hashes and conversion provenance. Requires the matching Pika re-noise sampler.2375fd1
14 days ago
-
Add Open-H3 Regenerate 2K refiner LoRA
Verified signature5905b20
17 days ago
-
Add WanGP PTD RefAsymOPD adapters and output heads
Verified signaturePublish eight scheduled RefAsymOPD adapters and the prepared output heads at the paths used by the full and pruned WanGP defaults. Include source attribution, license, and validation status.de1028e
18 days ago
-
Add latest H3 LightX2V Turbo LoRAs with explicit alpha tensors
Verified signatureAdd FL2V 4-step v1.3, FL2V 8-step v1.1, Ref2V 4-step v1.0 and Ref2V 8-step v1.1. Preserve original BF16 weights and add per-module alpha=8 tensors from the upstream metadata.8b3f78b
21 days ago
-
Add FastH3-Omni-8-Step-V1-pruned_rank8_bf16.safetensors
Verified signatureb575601 -
Add FastH3-Omni-8-Step-V1-pruned_rank8_int8_convrot.safetensors
Verified signatured0bc6d8 -
Add FastH3-Omni-NVFP4-QAD-pruned_rank8_nvfp4.safetensors
Verified signature007f3fe
27 days ago
-
Add TaoMate FL2VA 2026-10-15 and LightVAE Lite BF16/INT8
Verified signature630975d
30 days ago
-
Remove DMAD adapter and notice
Verified signature62e5db0 -
Add MiniMax H3 DMAD full-critic 4-step LoRA
Verified signatured785fd3
-
Add H3 Veda T2VA and R2VA predictors in BF16 and INT8 ConvRot
Verified signaturePrepared WanGP predictor modules for full and pruned H3 variants. BF16 weights preserve upstream dequantization; INT8 files use the shared ConvRot format. Source revisions and tile plans are retained in checkpoint metadata.49fe56a -
Add ConvRot file-level format metadata to prepared checkpoints
Verified signatureAdd quantization_format=int8_convrot and convrot_format=comfy_quant_v1. Preserve every tensor payload, dtype, shape and filename. Verified identical quantization detection, maps and runtime tensors; detection and conversion each perform zero per-layer comfy_quant JSON reads with the header markers.98b5288 -
Consolidate main history while preserving all current files
8d90934