Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-v4.6/checkpoint-13312/ema.safetensors +3 -0
- checkpoints-v4.6/checkpoint-13312/eval_state.json +3 -0
- checkpoints-v4.6/checkpoint-13312/model.safetensors +3 -0
- checkpoints-v4.6/checkpoint-13312/optimizer.pt +3 -0
- checkpoints-v4.6/checkpoint-13312/rng_state.pth +3 -0
- checkpoints-v4.6/checkpoint-13312/scaler.pt +3 -0
- checkpoints-v4.6/checkpoint-13312/scheduler.pt +3 -0
- checkpoints-v4.6/checkpoint-13312/trainer_state.json +801 -0
- checkpoints-v4.6/checkpoint-13312/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -52,3 +52,4 @@ checkpoints-v5.9/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs
|
|
| 52 |
checkpoints-v5.10/checkpoint-7168/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 53 |
checkpoints-v4.5/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 54 |
checkpoints-v5.11/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 52 |
checkpoints-v5.10/checkpoint-7168/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 53 |
checkpoints-v4.5/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 54 |
checkpoints-v5.11/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 55 |
+
checkpoints-v4.6/checkpoint-13312/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-v4.6/checkpoint-13312/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:410a4caf7dc8c831f6f1c1477872f6b73da8daa8d98c1c4bfae14bf53ffcd2c4
|
| 3 |
+
size 38935592
|
checkpoints-v4.6/checkpoint-13312/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a4e1429d83533955dcc8c481dc16057b82eab1b76885f4e56bc479796557f271
|
| 3 |
+
size 43383753
|
checkpoints-v4.6/checkpoint-13312/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d4e1fb35adc5b66c532bd54189eabd5c673b00e23c03af16bf1edfdee192238
|
| 3 |
+
size 38935624
|
checkpoints-v4.6/checkpoint-13312/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e7fd128e28cfae34f0d8050491d298d7b9f9b403b0c904f2c39e315c96c7de1
|
| 3 |
+
size 77992907
|
checkpoints-v4.6/checkpoint-13312/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1501339ed9a444ab026787513d9fc3ce539db5aee2e8e12315639e60161a5ee0
|
| 3 |
+
size 14645
|
checkpoints-v4.6/checkpoint-13312/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a34fec6a0d847723344253412c25b018523e427d3c1952286ab62a2afd2b427
|
| 3 |
+
size 1383
|
checkpoints-v4.6/checkpoint-13312/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f5fdf3e9e8835f981e5fcf3f2ce5d6a8ac2e3b6cfb6f151337a5e9701411146
|
| 3 |
+
size 1465
|
checkpoints-v4.6/checkpoint-13312/trainer_state.json
ADDED
|
@@ -0,0 +1,801 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.6148445799270241,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 13312,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011823934229365849,
|
| 14 |
+
"grad_norm": 2.166717529296875,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 6.863234996795654,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.9976925849914551,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 1.0293713808059692,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.9951309561729431,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 0.5387133359909058,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 1.1936287879943848,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 0.4395764172077179,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_bleu": 0.630805532752935,
|
| 43 |
+
"eval_ce_loss": 1.5585694417833738,
|
| 44 |
+
"eval_cos_loss": 0.12051772464015713,
|
| 45 |
+
"eval_loss": 0.41065708449169924,
|
| 46 |
+
"eval_mse_loss": 0.22680107500727317,
|
| 47 |
+
"eval_rec_loss": 0.015947288232876763,
|
| 48 |
+
"flow/cos_sim": 0.8794823028997744,
|
| 49 |
+
"flow/improvement_ratio": 0.9763019112147153,
|
| 50 |
+
"flow/mag_ratio_mean": 0.8738831932141901,
|
| 51 |
+
"flow/mag_ratio_std": 0.17382950611310463,
|
| 52 |
+
"step": 1024
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.047295736917463395,
|
| 56 |
+
"eval_bleu": 0.630805532752935,
|
| 57 |
+
"eval_ce_loss": 1.5585694417833738,
|
| 58 |
+
"eval_cos_loss": 0.12051772464015713,
|
| 59 |
+
"eval_loss": 0.41065708449169924,
|
| 60 |
+
"eval_mse_loss": 0.22680107500727317,
|
| 61 |
+
"eval_rec_loss": 0.015947288232876763,
|
| 62 |
+
"eval_runtime": 149.5036,
|
| 63 |
+
"eval_samples_per_second": 187.24,
|
| 64 |
+
"eval_steps_per_second": 2.93,
|
| 65 |
+
"flow/cos_sim": 0.8794823028997744,
|
| 66 |
+
"flow/improvement_ratio": 0.9763019112147153,
|
| 67 |
+
"flow/mag_ratio_mean": 0.8738831932141901,
|
| 68 |
+
"flow/mag_ratio_std": 0.17382950611310463,
|
| 69 |
+
"step": 1024
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.05911967114682925,
|
| 73 |
+
"grad_norm": 1.413312315940857,
|
| 74 |
+
"learning_rate": 0.0009967551747861387,
|
| 75 |
+
"loss": 0.40079358220100403,
|
| 76 |
+
"step": 1280
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.0709436053761951,
|
| 80 |
+
"grad_norm": 0.4399195909500122,
|
| 81 |
+
"learning_rate": 0.000994232528651847,
|
| 82 |
+
"loss": 0.38004180788993835,
|
| 83 |
+
"step": 1536
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.08276753960556095,
|
| 87 |
+
"grad_norm": 1.061546802520752,
|
| 88 |
+
"learning_rate": 0.0009909945800260092,
|
| 89 |
+
"loss": 0.3571281135082245,
|
| 90 |
+
"step": 1792
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.09459147383492679,
|
| 94 |
+
"grad_norm": 0.4095996916294098,
|
| 95 |
+
"learning_rate": 0.0009870460151900522,
|
| 96 |
+
"loss": 0.34413549304008484,
|
| 97 |
+
"step": 2048
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.09459147383492679,
|
| 101 |
+
"eval_bleu": 0.6370628963183373,
|
| 102 |
+
"eval_ce_loss": 1.4389138337411838,
|
| 103 |
+
"eval_cos_loss": 0.09142084970866164,
|
| 104 |
+
"eval_loss": 0.33612148960431415,
|
| 105 |
+
"eval_mse_loss": 0.17856517835566987,
|
| 106 |
+
"eval_rec_loss": 0.004522840211489004,
|
| 107 |
+
"flow/cos_sim": 0.9085791546460156,
|
| 108 |
+
"flow/improvement_ratio": 0.9776561622205935,
|
| 109 |
+
"flow/mag_ratio_mean": 0.9092166116248527,
|
| 110 |
+
"flow/mag_ratio_std": 0.14505240416417928,
|
| 111 |
+
"step": 2048
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.09459147383492679,
|
| 115 |
+
"eval_bleu": 0.6370628963183373,
|
| 116 |
+
"eval_ce_loss": 1.4389138337411838,
|
| 117 |
+
"eval_cos_loss": 0.09142084970866164,
|
| 118 |
+
"eval_loss": 0.33612148960431415,
|
| 119 |
+
"eval_mse_loss": 0.17856517835566987,
|
| 120 |
+
"eval_rec_loss": 0.004522840211489004,
|
| 121 |
+
"eval_runtime": 144.0023,
|
| 122 |
+
"eval_samples_per_second": 194.393,
|
| 123 |
+
"eval_steps_per_second": 3.042,
|
| 124 |
+
"flow/cos_sim": 0.9085791546460156,
|
| 125 |
+
"flow/improvement_ratio": 0.9776561622205935,
|
| 126 |
+
"flow/mag_ratio_mean": 0.9092166116248527,
|
| 127 |
+
"flow/mag_ratio_std": 0.14505240416417928,
|
| 128 |
+
"step": 2048
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.10641540806429264,
|
| 132 |
+
"grad_norm": 1.3680005073547363,
|
| 133 |
+
"learning_rate": 0.0009823925488998885,
|
| 134 |
+
"loss": 0.3346008360385895,
|
| 135 |
+
"step": 2304
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1182393422936585,
|
| 139 |
+
"grad_norm": 0.9834803938865662,
|
| 140 |
+
"learning_rate": 0.0009770409161149525,
|
| 141 |
+
"loss": 0.32358497381210327,
|
| 142 |
+
"step": 2560
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.13006327652302435,
|
| 146 |
+
"grad_norm": 0.7012990117073059,
|
| 147 |
+
"learning_rate": 0.0009709988622506973,
|
| 148 |
+
"loss": 0.31395941972732544,
|
| 149 |
+
"step": 2816
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.1418872107523902,
|
| 153 |
+
"grad_norm": 0.4390225112438202,
|
| 154 |
+
"learning_rate": 0.000964275131968659,
|
| 155 |
+
"loss": 0.30854108929634094,
|
| 156 |
+
"step": 3072
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1418872107523902,
|
| 160 |
+
"eval_bleu": 0.6707740050213097,
|
| 161 |
+
"eval_ce_loss": 1.2450400457534616,
|
| 162 |
+
"eval_cos_loss": 0.08531694868622033,
|
| 163 |
+
"eval_loss": 0.3036313052523082,
|
| 164 |
+
"eval_mse_loss": 0.16824311817481638,
|
| 165 |
+
"eval_rec_loss": 0.0023524861730428916,
|
| 166 |
+
"flow/cos_sim": 0.9146830740312463,
|
| 167 |
+
"flow/improvement_ratio": 0.9760218352487643,
|
| 168 |
+
"flow/mag_ratio_mean": 0.9162033348595171,
|
| 169 |
+
"flow/mag_ratio_std": 0.12317379108166586,
|
| 170 |
+
"step": 3072
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.1418872107523902,
|
| 174 |
+
"eval_bleu": 0.6707740050213097,
|
| 175 |
+
"eval_ce_loss": 1.2450400457534616,
|
| 176 |
+
"eval_cos_loss": 0.08531694868622033,
|
| 177 |
+
"eval_loss": 0.3036313052523082,
|
| 178 |
+
"eval_mse_loss": 0.16824311817481638,
|
| 179 |
+
"eval_rec_loss": 0.0023524861730428916,
|
| 180 |
+
"eval_runtime": 145.6816,
|
| 181 |
+
"eval_samples_per_second": 192.152,
|
| 182 |
+
"eval_steps_per_second": 3.007,
|
| 183 |
+
"flow/cos_sim": 0.9146830740312463,
|
| 184 |
+
"flow/improvement_ratio": 0.9760218352487643,
|
| 185 |
+
"flow/mag_ratio_mean": 0.9162033348595171,
|
| 186 |
+
"flow/mag_ratio_std": 0.12317379108166586,
|
| 187 |
+
"step": 3072
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.15371114498175603,
|
| 191 |
+
"grad_norm": 0.6193675398826599,
|
| 192 |
+
"learning_rate": 0.0009568794565203123,
|
| 193 |
+
"loss": 0.30133724212646484,
|
| 194 |
+
"step": 3328
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.1655350792111219,
|
| 198 |
+
"grad_norm": 0.5207021236419678,
|
| 199 |
+
"learning_rate": 0.0009488225396630347,
|
| 200 |
+
"loss": 0.29369795322418213,
|
| 201 |
+
"step": 3584
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.17735901344048774,
|
| 205 |
+
"grad_norm": 0.5521863102912903,
|
| 206 |
+
"learning_rate": 0.0009401160421685646,
|
| 207 |
+
"loss": 0.2897779941558838,
|
| 208 |
+
"step": 3840
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"grad_norm": 0.4839329421520233,
|
| 213 |
+
"learning_rate": 0.0009307725649463714,
|
| 214 |
+
"loss": 0.2864342927932739,
|
| 215 |
+
"step": 4096
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.18918294766985358,
|
| 219 |
+
"eval_bleu": 0.6704121655421459,
|
| 220 |
+
"eval_ce_loss": 1.2046921730858007,
|
| 221 |
+
"eval_cos_loss": 0.07965970783575213,
|
| 222 |
+
"eval_loss": 0.2842630836857508,
|
| 223 |
+
"eval_mse_loss": 0.1541883000534147,
|
| 224 |
+
"eval_rec_loss": 0.0016395924527162992,
|
| 225 |
+
"flow/cos_sim": 0.9203403159363629,
|
| 226 |
+
"flow/improvement_ratio": 0.9744544083669305,
|
| 227 |
+
"flow/mag_ratio_mean": 0.9212693351588838,
|
| 228 |
+
"flow/mag_ratio_std": 0.10148854444832563,
|
| 229 |
+
"step": 4096
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.18918294766985358,
|
| 233 |
+
"eval_bleu": 0.6704121655421459,
|
| 234 |
+
"eval_ce_loss": 1.2046921730858007,
|
| 235 |
+
"eval_cos_loss": 0.07965970783575213,
|
| 236 |
+
"eval_loss": 0.2842630836857508,
|
| 237 |
+
"eval_mse_loss": 0.1541883000534147,
|
| 238 |
+
"eval_rec_loss": 0.0016395924527162992,
|
| 239 |
+
"eval_runtime": 144.2359,
|
| 240 |
+
"eval_samples_per_second": 194.078,
|
| 241 |
+
"eval_steps_per_second": 3.037,
|
| 242 |
+
"flow/cos_sim": 0.9203403159363629,
|
| 243 |
+
"flow/improvement_ratio": 0.9744544083669305,
|
| 244 |
+
"flow/mag_ratio_mean": 0.9212693351588838,
|
| 245 |
+
"flow/mag_ratio_std": 0.10148854444832563,
|
| 246 |
+
"step": 4096
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.20100688189921945,
|
| 250 |
+
"grad_norm": 0.9023747444152832,
|
| 251 |
+
"learning_rate": 0.0009208056308063659,
|
| 252 |
+
"loss": 0.30037108063697815,
|
| 253 |
+
"step": 4352
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.2128308161285853,
|
| 257 |
+
"grad_norm": 0.8910952806472778,
|
| 258 |
+
"learning_rate": 0.0009102296648873445,
|
| 259 |
+
"loss": 0.28084370493888855,
|
| 260 |
+
"step": 4608
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.22465475035795113,
|
| 264 |
+
"grad_norm": 0.43350881338119507,
|
| 265 |
+
"learning_rate": 0.0008990599737794927,
|
| 266 |
+
"loss": 0.276326060295105,
|
| 267 |
+
"step": 4864
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.236478684587317,
|
| 271 |
+
"grad_norm": 0.4241744577884674,
|
| 272 |
+
"learning_rate": 0.0008873127233711644,
|
| 273 |
+
"loss": 0.2720130980014801,
|
| 274 |
+
"step": 5120
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.236478684587317,
|
| 278 |
+
"eval_bleu": 0.6884395396484843,
|
| 279 |
+
"eval_ce_loss": 1.1380460211947627,
|
| 280 |
+
"eval_cos_loss": 0.0743916420410621,
|
| 281 |
+
"eval_loss": 0.26738106996909666,
|
| 282 |
+
"eval_mse_loss": 0.14492133019925796,
|
| 283 |
+
"eval_rec_loss": 0.0012159729793058606,
|
| 284 |
+
"flow/cos_sim": 0.9256083738041795,
|
| 285 |
+
"flow/improvement_ratio": 0.9722109245110865,
|
| 286 |
+
"flow/mag_ratio_mean": 0.9361066815515632,
|
| 287 |
+
"flow/mag_ratio_std": 0.08903367672812994,
|
| 288 |
+
"step": 5120
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.236478684587317,
|
| 292 |
+
"eval_bleu": 0.6884395396484843,
|
| 293 |
+
"eval_ce_loss": 1.1380460211947627,
|
| 294 |
+
"eval_cos_loss": 0.0743916420410621,
|
| 295 |
+
"eval_loss": 0.26738106996909666,
|
| 296 |
+
"eval_mse_loss": 0.14492133019925796,
|
| 297 |
+
"eval_rec_loss": 0.0012159729793058606,
|
| 298 |
+
"eval_runtime": 145.1812,
|
| 299 |
+
"eval_samples_per_second": 192.814,
|
| 300 |
+
"eval_steps_per_second": 3.017,
|
| 301 |
+
"flow/cos_sim": 0.9256083738041795,
|
| 302 |
+
"flow/improvement_ratio": 0.9722109245110865,
|
| 303 |
+
"flow/mag_ratio_mean": 0.9361066815515632,
|
| 304 |
+
"flow/mag_ratio_std": 0.08903367672812994,
|
| 305 |
+
"step": 5120
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.24830261881668284,
|
| 309 |
+
"grad_norm": 0.7429539561271667,
|
| 310 |
+
"learning_rate": 0.0008750049154520011,
|
| 311 |
+
"loss": 0.2687661945819855,
|
| 312 |
+
"step": 5376
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.2601265530460487,
|
| 316 |
+
"grad_norm": 0.4944668114185333,
|
| 317 |
+
"learning_rate": 0.0008621543631062487,
|
| 318 |
+
"loss": 0.26752030849456787,
|
| 319 |
+
"step": 5632
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.27195048727541454,
|
| 323 |
+
"grad_norm": 0.4112600088119507,
|
| 324 |
+
"learning_rate": 0.0008487796649318904,
|
| 325 |
+
"loss": 0.26332396268844604,
|
| 326 |
+
"step": 5888
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.2837744215047804,
|
| 330 |
+
"grad_norm": 0.5803443789482117,
|
| 331 |
+
"learning_rate": 0.0008349001781229053,
|
| 332 |
+
"loss": 0.2650764286518097,
|
| 333 |
+
"step": 6144
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.2837744215047804,
|
| 337 |
+
"eval_bleu": 0.6901447099765771,
|
| 338 |
+
"eval_ce_loss": 1.1182073426845411,
|
| 339 |
+
"eval_cos_loss": 0.07293294577701995,
|
| 340 |
+
"eval_loss": 0.2625204250216484,
|
| 341 |
+
"eval_mse_loss": 0.14227469792785166,
|
| 342 |
+
"eval_rec_loss": 0.0011316965778382733,
|
| 343 |
+
"flow/cos_sim": 0.9270670652117359,
|
| 344 |
+
"flow/improvement_ratio": 0.972748759402532,
|
| 345 |
+
"flow/mag_ratio_mean": 0.9329337213681713,
|
| 346 |
+
"flow/mag_ratio_std": 0.08477971859310315,
|
| 347 |
+
"step": 6144
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.2837744215047804,
|
| 351 |
+
"eval_bleu": 0.6901447099765771,
|
| 352 |
+
"eval_ce_loss": 1.1182073426845411,
|
| 353 |
+
"eval_cos_loss": 0.07293294577701995,
|
| 354 |
+
"eval_loss": 0.2625204250216484,
|
| 355 |
+
"eval_mse_loss": 0.14227469792785166,
|
| 356 |
+
"eval_rec_loss": 0.0011316965778382733,
|
| 357 |
+
"eval_runtime": 143.6154,
|
| 358 |
+
"eval_samples_per_second": 194.916,
|
| 359 |
+
"eval_steps_per_second": 3.05,
|
| 360 |
+
"flow/cos_sim": 0.9270670652117359,
|
| 361 |
+
"flow/improvement_ratio": 0.972748759402532,
|
| 362 |
+
"flow/mag_ratio_mean": 0.9329337213681713,
|
| 363 |
+
"flow/mag_ratio_std": 0.08477971859310315,
|
| 364 |
+
"step": 6144
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.2955983557341462,
|
| 368 |
+
"grad_norm": 0.5521038770675659,
|
| 369 |
+
"learning_rate": 0.0008205359904536107,
|
| 370 |
+
"loss": 0.2589625120162964,
|
| 371 |
+
"step": 6400
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.30742228996351206,
|
| 375 |
+
"grad_norm": 0.5204539895057678,
|
| 376 |
+
"learning_rate": 0.0008057078912056363,
|
| 377 |
+
"loss": 0.2580261528491974,
|
| 378 |
+
"step": 6656
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.3192462241928779,
|
| 382 |
+
"grad_norm": 0.5325313806533813,
|
| 383 |
+
"learning_rate": 0.0007904373410796086,
|
| 384 |
+
"loss": 0.25551143288612366,
|
| 385 |
+
"step": 6912
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.3310701584222438,
|
| 389 |
+
"grad_norm": 0.3639472424983978,
|
| 390 |
+
"learning_rate": 0.0007747464411350876,
|
| 391 |
+
"loss": 0.25460946559906006,
|
| 392 |
+
"step": 7168
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.3310701584222438,
|
| 396 |
+
"eval_bleu": 0.7071091482680488,
|
| 397 |
+
"eval_ce_loss": 1.0378137462215336,
|
| 398 |
+
"eval_cos_loss": 0.07100786366521224,
|
| 399 |
+
"eval_loss": 0.2512795734283042,
|
| 400 |
+
"eval_mse_loss": 0.13953843853993503,
|
| 401 |
+
"eval_rec_loss": 0.0008589719193517103,
|
| 402 |
+
"flow/cos_sim": 0.9289921618759904,
|
| 403 |
+
"flow/improvement_ratio": 0.9735576084487514,
|
| 404 |
+
"flow/mag_ratio_mean": 0.9378764381691745,
|
| 405 |
+
"flow/mag_ratio_std": 0.08363016963549401,
|
| 406 |
+
"step": 7168
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.3310701584222438,
|
| 410 |
+
"eval_bleu": 0.7071091482680488,
|
| 411 |
+
"eval_ce_loss": 1.0378137462215336,
|
| 412 |
+
"eval_cos_loss": 0.07100786366521224,
|
| 413 |
+
"eval_loss": 0.2512795734283042,
|
| 414 |
+
"eval_mse_loss": 0.13953843853993503,
|
| 415 |
+
"eval_rec_loss": 0.0008589719193517103,
|
| 416 |
+
"eval_runtime": 145.8938,
|
| 417 |
+
"eval_samples_per_second": 191.872,
|
| 418 |
+
"eval_steps_per_second": 3.002,
|
| 419 |
+
"flow/cos_sim": 0.9289921618759904,
|
| 420 |
+
"flow/improvement_ratio": 0.9735576084487514,
|
| 421 |
+
"flow/mag_ratio_mean": 0.9378764381691745,
|
| 422 |
+
"flow/mag_ratio_std": 0.08363016963549401,
|
| 423 |
+
"step": 7168
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.34289409265160964,
|
| 427 |
+
"grad_norm": 0.26728010177612305,
|
| 428 |
+
"learning_rate": 0.000758657900803716,
|
| 429 |
+
"loss": 0.25558242201805115,
|
| 430 |
+
"step": 7424
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3547180268809755,
|
| 434 |
+
"grad_norm": 0.3367791175842285,
|
| 435 |
+
"learning_rate": 0.000742195005021869,
|
| 436 |
+
"loss": 0.2547113597393036,
|
| 437 |
+
"step": 7680
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3665419611103413,
|
| 441 |
+
"grad_norm": 0.6501409411430359,
|
| 442 |
+
"learning_rate": 0.0007253815805303786,
|
| 443 |
+
"loss": 0.2516462504863739,
|
| 444 |
+
"step": 7936
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.37836589533970716,
|
| 448 |
+
"grad_norm": 0.41470277309417725,
|
| 449 |
+
"learning_rate": 0.0007082419613901028,
|
| 450 |
+
"loss": 0.25756070017814636,
|
| 451 |
+
"step": 8192
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.37836589533970716,
|
| 455 |
+
"eval_bleu": 0.7183146382916741,
|
| 456 |
+
"eval_ce_loss": 1.0048845314408001,
|
| 457 |
+
"eval_cos_loss": 0.07200061048478841,
|
| 458 |
+
"eval_loss": 0.2508213480245577,
|
| 459 |
+
"eval_mse_loss": 0.14229333698409333,
|
| 460 |
+
"eval_rec_loss": 0.0008394962100133504,
|
| 461 |
+
"flow/cos_sim": 0.9279994152180137,
|
| 462 |
+
"flow/improvement_ratio": 0.9741122624645494,
|
| 463 |
+
"flow/mag_ratio_mean": 0.9306073022759669,
|
| 464 |
+
"flow/mag_ratio_std": 0.08248136017352478,
|
| 465 |
+
"step": 8192
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.37836589533970716,
|
| 469 |
+
"eval_bleu": 0.7183146382916741,
|
| 470 |
+
"eval_ce_loss": 1.0048845314408001,
|
| 471 |
+
"eval_cos_loss": 0.07200061048478841,
|
| 472 |
+
"eval_loss": 0.2508213480245577,
|
| 473 |
+
"eval_mse_loss": 0.14229333698409333,
|
| 474 |
+
"eval_rec_loss": 0.0008394962100133504,
|
| 475 |
+
"eval_runtime": 144.5073,
|
| 476 |
+
"eval_samples_per_second": 193.713,
|
| 477 |
+
"eval_steps_per_second": 3.031,
|
| 478 |
+
"flow/cos_sim": 0.9279994152180137,
|
| 479 |
+
"flow/improvement_ratio": 0.9741122624645494,
|
| 480 |
+
"flow/mag_ratio_mean": 0.9306073022759669,
|
| 481 |
+
"flow/mag_ratio_std": 0.08248136017352478,
|
| 482 |
+
"step": 8192
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.390189829569073,
|
| 486 |
+
"grad_norm": 0.5503519773483276,
|
| 487 |
+
"learning_rate": 0.0006908009537632514,
|
| 488 |
+
"loss": 0.24980439245700836,
|
| 489 |
+
"step": 8448
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.4020137637984389,
|
| 493 |
+
"grad_norm": 0.5455402135848999,
|
| 494 |
+
"learning_rate": 0.0006730838000114403,
|
| 495 |
+
"loss": 0.25091099739074707,
|
| 496 |
+
"step": 8704
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.41383769802780473,
|
| 500 |
+
"grad_norm": 0.29903262853622437,
|
| 501 |
+
"learning_rate": 0.0006551161421624341,
|
| 502 |
+
"loss": 0.24810341000556946,
|
| 503 |
+
"step": 8960
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.4256616322571706,
|
| 507 |
+
"grad_norm": 0.5439685583114624,
|
| 508 |
+
"learning_rate": 0.0006369239847984517,
|
| 509 |
+
"loss": 0.24486252665519714,
|
| 510 |
+
"step": 9216
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 0.4256616322571706,
|
| 514 |
+
"eval_bleu": 0.7059557795131848,
|
| 515 |
+
"eval_ce_loss": 1.0441748459045201,
|
| 516 |
+
"eval_cos_loss": 0.06742798249539175,
|
| 517 |
+
"eval_loss": 0.24551006137780404,
|
| 518 |
+
"eval_mse_loss": 0.13372647651730607,
|
| 519 |
+
"eval_rec_loss": 0.0006233008009167367,
|
| 520 |
+
"flow/cos_sim": 0.9325720338516583,
|
| 521 |
+
"flow/improvement_ratio": 0.9731077969618583,
|
| 522 |
+
"flow/mag_ratio_mean": 0.934824386688128,
|
| 523 |
+
"flow/mag_ratio_std": 0.079195149838244,
|
| 524 |
+
"step": 9216
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.4256616322571706,
|
| 528 |
+
"eval_bleu": 0.7059557795131848,
|
| 529 |
+
"eval_ce_loss": 1.0441748459045201,
|
| 530 |
+
"eval_cos_loss": 0.06742798249539175,
|
| 531 |
+
"eval_loss": 0.24551006137780404,
|
| 532 |
+
"eval_mse_loss": 0.13372647651730607,
|
| 533 |
+
"eval_rec_loss": 0.0006233008009167367,
|
| 534 |
+
"eval_runtime": 143.7536,
|
| 535 |
+
"eval_samples_per_second": 194.729,
|
| 536 |
+
"eval_steps_per_second": 3.047,
|
| 537 |
+
"flow/cos_sim": 0.9325720338516583,
|
| 538 |
+
"flow/improvement_ratio": 0.9731077969618583,
|
| 539 |
+
"flow/mag_ratio_mean": 0.934824386688128,
|
| 540 |
+
"flow/mag_ratio_std": 0.079195149838244,
|
| 541 |
+
"step": 9216
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.4374855664865364,
|
| 545 |
+
"grad_norm": 0.6247985363006592,
|
| 546 |
+
"learning_rate": 0.0006185336574197479,
|
| 547 |
+
"loss": 0.24821333587169647,
|
| 548 |
+
"step": 9472
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.44930950071590225,
|
| 552 |
+
"grad_norm": 0.5992924571037292,
|
| 553 |
+
"learning_rate": 0.0005999717763379407,
|
| 554 |
+
"loss": 0.2463277131319046,
|
| 555 |
+
"step": 9728
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.4611334349452681,
|
| 559 |
+
"grad_norm": 0.5051174759864807,
|
| 560 |
+
"learning_rate": 0.0005812652061542363,
|
| 561 |
+
"loss": 0.24300755560398102,
|
| 562 |
+
"step": 9984
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.472957369174634,
|
| 566 |
+
"grad_norm": 0.40211889147758484,
|
| 567 |
+
"learning_rate": 0.0005624410208783071,
|
| 568 |
+
"loss": 0.24105887115001678,
|
| 569 |
+
"step": 10240
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.472957369174634,
|
| 573 |
+
"eval_bleu": 0.7120039792514781,
|
| 574 |
+
"eval_ce_loss": 1.005759668472695,
|
| 575 |
+
"eval_cos_loss": 0.06631913622118295,
|
| 576 |
+
"eval_loss": 0.2399323729619588,
|
| 577 |
+
"eval_mse_loss": 0.1321649712820848,
|
| 578 |
+
"eval_rec_loss": 0.0005595201312656808,
|
| 579 |
+
"flow/cos_sim": 0.9336808790627136,
|
| 580 |
+
"flow/improvement_ratio": 0.9762771404497156,
|
| 581 |
+
"flow/mag_ratio_mean": 0.926664409572131,
|
| 582 |
+
"flow/mag_ratio_std": 0.0785302072763443,
|
| 583 |
+
"step": 10240
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.472957369174634,
|
| 587 |
+
"eval_bleu": 0.7120039792514781,
|
| 588 |
+
"eval_ce_loss": 1.005759668472695,
|
| 589 |
+
"eval_cos_loss": 0.06631913622118295,
|
| 590 |
+
"eval_loss": 0.2399323729619588,
|
| 591 |
+
"eval_mse_loss": 0.1321649712820848,
|
| 592 |
+
"eval_rec_loss": 0.0005595201312656808,
|
| 593 |
+
"eval_runtime": 145.1926,
|
| 594 |
+
"eval_samples_per_second": 192.799,
|
| 595 |
+
"eval_steps_per_second": 3.017,
|
| 596 |
+
"flow/cos_sim": 0.9336808790627136,
|
| 597 |
+
"flow/improvement_ratio": 0.9762771404497156,
|
| 598 |
+
"flow/mag_ratio_mean": 0.926664409572131,
|
| 599 |
+
"flow/mag_ratio_std": 0.0785302072763443,
|
| 600 |
+
"step": 10240
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.48478130340399983,
|
| 604 |
+
"grad_norm": 0.5163204073905945,
|
| 605 |
+
"learning_rate": 0.0005435264647440881,
|
| 606 |
+
"loss": 0.2431734949350357,
|
| 607 |
+
"step": 10496
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.49660523763336567,
|
| 611 |
+
"grad_norm": 0.3941446840763092,
|
| 612 |
+
"learning_rate": 0.000524548912779213,
|
| 613 |
+
"loss": 0.23893995583057404,
|
| 614 |
+
"step": 10752
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5084291718627315,
|
| 618 |
+
"grad_norm": 0.600309431552887,
|
| 619 |
+
"learning_rate": 0.0005055358311851499,
|
| 620 |
+
"loss": 0.24141335487365723,
|
| 621 |
+
"step": 11008
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5202531060920974,
|
| 625 |
+
"grad_norm": 0.2635146975517273,
|
| 626 |
+
"learning_rate": 0.0004865147375853812,
|
| 627 |
+
"loss": 0.23994745314121246,
|
| 628 |
+
"step": 11264
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 0.5202531060920974,
|
| 632 |
+
"eval_bleu": 0.721584420569106,
|
| 633 |
+
"eval_ce_loss": 0.9708276380142665,
|
| 634 |
+
"eval_cos_loss": 0.06598166635761794,
|
| 635 |
+
"eval_loss": 0.236370724664159,
|
| 636 |
+
"eval_mse_loss": 0.13219797124737473,
|
| 637 |
+
"eval_rec_loss": 0.0004918215776982103,
|
| 638 |
+
"flow/cos_sim": 0.9340183556624199,
|
| 639 |
+
"flow/improvement_ratio": 0.9756744954411842,
|
| 640 |
+
"flow/mag_ratio_mean": 0.9309794351662675,
|
| 641 |
+
"flow/mag_ratio_std": 0.07776250285372886,
|
| 642 |
+
"step": 11264
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.5202531060920974,
|
| 646 |
+
"eval_bleu": 0.721584420569106,
|
| 647 |
+
"eval_ce_loss": 0.9708276380142665,
|
| 648 |
+
"eval_cos_loss": 0.06598166635761794,
|
| 649 |
+
"eval_loss": 0.236370724664159,
|
| 650 |
+
"eval_mse_loss": 0.13219797124737473,
|
| 651 |
+
"eval_rec_loss": 0.0004918215776982103,
|
| 652 |
+
"eval_runtime": 145.3935,
|
| 653 |
+
"eval_samples_per_second": 192.533,
|
| 654 |
+
"eval_steps_per_second": 3.013,
|
| 655 |
+
"flow/cos_sim": 0.9340183556624199,
|
| 656 |
+
"flow/improvement_ratio": 0.9756744954411842,
|
| 657 |
+
"flow/mag_ratio_mean": 0.9309794351662675,
|
| 658 |
+
"flow/mag_ratio_std": 0.07776250285372886,
|
| 659 |
+
"step": 11264
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.5320770403214632,
|
| 663 |
+
"grad_norm": 0.7631673216819763,
|
| 664 |
+
"learning_rate": 0.0004675131611991607,
|
| 665 |
+
"loss": 0.2390274852514267,
|
| 666 |
+
"step": 11520
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.5439009745508291,
|
| 670 |
+
"grad_norm": 0.3735812306404114,
|
| 671 |
+
"learning_rate": 0.0004485586029984899,
|
| 672 |
+
"loss": 0.23720039427280426,
|
| 673 |
+
"step": 11776
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.5557249087801949,
|
| 677 |
+
"grad_norm": 0.31787776947021484,
|
| 678 |
+
"learning_rate": 0.00042967849590597266,
|
| 679 |
+
"loss": 0.2358924150466919,
|
| 680 |
+
"step": 12032
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.5675488430095608,
|
| 684 |
+
"grad_norm": 0.4151494801044464,
|
| 685 |
+
"learning_rate": 0.0004109001650911621,
|
| 686 |
+
"loss": 0.23592068254947662,
|
| 687 |
+
"step": 12288
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.5675488430095608,
|
| 691 |
+
"eval_bleu": 0.7193234778931977,
|
| 692 |
+
"eval_ce_loss": 0.9832807419370843,
|
| 693 |
+
"eval_cos_loss": 0.06677046464968762,
|
| 694 |
+
"eval_loss": 0.2394936037118032,
|
| 695 |
+
"eval_mse_loss": 0.1340124544147487,
|
| 696 |
+
"eval_rec_loss": 0.00047602696994042604,
|
| 697 |
+
"flow/cos_sim": 0.933229555278064,
|
| 698 |
+
"flow/improvement_ratio": 0.9741054156871691,
|
| 699 |
+
"flow/mag_ratio_mean": 0.9458617875565133,
|
| 700 |
+
"flow/mag_ratio_std": 0.07636498908201854,
|
| 701 |
+
"step": 12288
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.5675488430095608,
|
| 705 |
+
"eval_bleu": 0.7193234778931977,
|
| 706 |
+
"eval_ce_loss": 0.9832807419370843,
|
| 707 |
+
"eval_cos_loss": 0.06677046464968762,
|
| 708 |
+
"eval_loss": 0.2394936037118032,
|
| 709 |
+
"eval_mse_loss": 0.1340124544147487,
|
| 710 |
+
"eval_rec_loss": 0.00047602696994042604,
|
| 711 |
+
"eval_runtime": 146.0267,
|
| 712 |
+
"eval_samples_per_second": 191.698,
|
| 713 |
+
"eval_steps_per_second": 2.999,
|
| 714 |
+
"flow/cos_sim": 0.933229555278064,
|
| 715 |
+
"flow/improvement_ratio": 0.9741054156871691,
|
| 716 |
+
"flow/mag_ratio_mean": 0.9458617875565133,
|
| 717 |
+
"flow/mag_ratio_std": 0.07636498908201854,
|
| 718 |
+
"step": 12288
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.5793727772389267,
|
| 722 |
+
"grad_norm": 0.5829018950462341,
|
| 723 |
+
"learning_rate": 0.0003922507884228551,
|
| 724 |
+
"loss": 0.23718097805976868,
|
| 725 |
+
"step": 12544
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.5911967114682924,
|
| 729 |
+
"grad_norm": 0.24561040103435516,
|
| 730 |
+
"learning_rate": 0.00037375735713457723,
|
| 731 |
+
"loss": 0.23286746442317963,
|
| 732 |
+
"step": 12800
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.6030206456976583,
|
| 736 |
+
"grad_norm": 0.2852168381214142,
|
| 737 |
+
"learning_rate": 0.00035544663676018276,
|
| 738 |
+
"loss": 0.23593547940254211,
|
| 739 |
+
"step": 13056
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.6148445799270241,
|
| 743 |
+
"grad_norm": 0.387464314699173,
|
| 744 |
+
"learning_rate": 0.00033734512839611255,
|
| 745 |
+
"loss": 0.23200258612632751,
|
| 746 |
+
"step": 13312
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.6148445799270241,
|
| 750 |
+
"eval_bleu": 0.7243775501478836,
|
| 751 |
+
"eval_ce_loss": 0.9625263297802782,
|
| 752 |
+
"eval_cos_loss": 0.06494614632468518,
|
| 753 |
+
"eval_loss": 0.2341489744581044,
|
| 754 |
+
"eval_mse_loss": 0.13099264763465754,
|
| 755 |
+
"eval_rec_loss": 0.00040907774940659357,
|
| 756 |
+
"flow/cos_sim": 0.9350538734159514,
|
| 757 |
+
"flow/improvement_ratio": 0.9740996707792151,
|
| 758 |
+
"flow/mag_ratio_mean": 0.9335339481699957,
|
| 759 |
+
"flow/mag_ratio_std": 0.07727514897987722,
|
| 760 |
+
"step": 13312
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.6148445799270241,
|
| 764 |
+
"eval_bleu": 0.7243775501478836,
|
| 765 |
+
"eval_ce_loss": 0.9625263297802782,
|
| 766 |
+
"eval_cos_loss": 0.06494614632468518,
|
| 767 |
+
"eval_loss": 0.2341489744581044,
|
| 768 |
+
"eval_mse_loss": 0.13099264763465754,
|
| 769 |
+
"eval_rec_loss": 0.00040907774940659357,
|
| 770 |
+
"eval_runtime": 144.1869,
|
| 771 |
+
"eval_samples_per_second": 194.144,
|
| 772 |
+
"eval_steps_per_second": 3.038,
|
| 773 |
+
"flow/cos_sim": 0.9350538734159514,
|
| 774 |
+
"flow/improvement_ratio": 0.9740996707792151,
|
| 775 |
+
"flow/mag_ratio_mean": 0.9335339481699957,
|
| 776 |
+
"flow/mag_ratio_std": 0.07727514897987722,
|
| 777 |
+
"step": 13312
|
| 778 |
+
}
|
| 779 |
+
],
|
| 780 |
+
"logging_steps": 256,
|
| 781 |
+
"max_steps": 21651,
|
| 782 |
+
"num_input_tokens_seen": 0,
|
| 783 |
+
"num_train_epochs": 1,
|
| 784 |
+
"save_steps": 1024,
|
| 785 |
+
"stateful_callbacks": {
|
| 786 |
+
"TrainerControl": {
|
| 787 |
+
"args": {
|
| 788 |
+
"should_epoch_stop": false,
|
| 789 |
+
"should_evaluate": false,
|
| 790 |
+
"should_log": false,
|
| 791 |
+
"should_save": true,
|
| 792 |
+
"should_training_stop": false
|
| 793 |
+
},
|
| 794 |
+
"attributes": {}
|
| 795 |
+
}
|
| 796 |
+
},
|
| 797 |
+
"total_flos": 0.0,
|
| 798 |
+
"train_batch_size": 64,
|
| 799 |
+
"trial_name": null,
|
| 800 |
+
"trial_params": null
|
| 801 |
+
}
|
checkpoints-v4.6/checkpoint-13312/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee9cb7ebeac20164f12c8a5bf77bbe01722a7871151494aba288a393bf7de081
|
| 3 |
+
size 5137
|