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
|
@@ -61,3 +61,4 @@ checkpoints-v5.12-b/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=l
|
|
| 61 |
checkpoints-v5.12-d/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 62 |
checkpoints-v5.13/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 63 |
checkpoints-v5.13-b/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 61 |
checkpoints-v5.12-d/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 62 |
checkpoints-v5.13/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 63 |
checkpoints-v5.13-b/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 64 |
+
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:e1bda22a3c6aa3475bc245d613fd0811bce6d20b3b46f8ead16a8e11a899d835
|
| 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 sha256:8ff8d8722fde014283b11a352f5964755f7cf2ab00488648b395320fe35956b5
|
| 3 |
+
size 43356375
|
checkpoints-v4.6+/checkpoint-13312/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e00d4750932024e45dca2db12d387840ad224f2748c233de646662bf5f0d78fb
|
| 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 sha256:3621eeff3751ef43a4e51f0a3a3febae23d0b7418521dd37af4fee981e9a80b7
|
| 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 sha256:39cc7c96286260fa904379127512cbdc6d21a90eef555b0c583a9e629d4b6ada
|
| 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": 1.9140732288360596,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 6.9310712814331055,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.966057538986206,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 1.017655372619629,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.6036155223846436,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 0.5299859046936035,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.6692569851875305,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 0.43251049518585205,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_bleu": 0.6267380536600159,
|
| 43 |
+
"eval_ce_loss": 1.563563620935292,
|
| 44 |
+
"eval_cos_loss": 0.11390362637432198,
|
| 45 |
+
"eval_loss": 0.40128198402113024,
|
| 46 |
+
"eval_mse_loss": 0.2176408769061032,
|
| 47 |
+
"eval_rec_loss": 0.015894381031734215,
|
| 48 |
+
"flow/cos_sim": 0.8860963983078526,
|
| 49 |
+
"flow/improvement_ratio": 0.9765386980146034,
|
| 50 |
+
"flow/mag_ratio_mean": 0.8838755380889597,
|
| 51 |
+
"flow/mag_ratio_std": 0.1646749794006892,
|
| 52 |
+
"step": 1024
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.047295736917463395,
|
| 56 |
+
"eval_bleu": 0.6267380536600159,
|
| 57 |
+
"eval_ce_loss": 1.563563620935292,
|
| 58 |
+
"eval_cos_loss": 0.11390362637432198,
|
| 59 |
+
"eval_loss": 0.40128198402113024,
|
| 60 |
+
"eval_mse_loss": 0.2176408769061032,
|
| 61 |
+
"eval_rec_loss": 0.015894381031734215,
|
| 62 |
+
"eval_runtime": 145.0392,
|
| 63 |
+
"eval_samples_per_second": 193.003,
|
| 64 |
+
"eval_steps_per_second": 3.02,
|
| 65 |
+
"flow/cos_sim": 0.8860963983078526,
|
| 66 |
+
"flow/improvement_ratio": 0.9765386980146034,
|
| 67 |
+
"flow/mag_ratio_mean": 0.8838755380889597,
|
| 68 |
+
"flow/mag_ratio_std": 0.1646749794006892,
|
| 69 |
+
"step": 1024
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.05911967114682925,
|
| 73 |
+
"grad_norm": 0.46750983595848083,
|
| 74 |
+
"learning_rate": 0.0009967551747861387,
|
| 75 |
+
"loss": 0.3926387429237366,
|
| 76 |
+
"step": 1280
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.0709436053761951,
|
| 80 |
+
"grad_norm": 0.5673767328262329,
|
| 81 |
+
"learning_rate": 0.000994232528651847,
|
| 82 |
+
"loss": 0.3741423189640045,
|
| 83 |
+
"step": 1536
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.08276753960556095,
|
| 87 |
+
"grad_norm": 0.5806387066841125,
|
| 88 |
+
"learning_rate": 0.0009909945800260092,
|
| 89 |
+
"loss": 0.35307449102401733,
|
| 90 |
+
"step": 1792
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.09459147383492679,
|
| 94 |
+
"grad_norm": 0.6838217973709106,
|
| 95 |
+
"learning_rate": 0.0009870460151900522,
|
| 96 |
+
"loss": 0.3399304151535034,
|
| 97 |
+
"step": 2048
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.09459147383492679,
|
| 101 |
+
"eval_bleu": 0.6478349738418474,
|
| 102 |
+
"eval_ce_loss": 1.3434572242982856,
|
| 103 |
+
"eval_cos_loss": 0.09738318450306649,
|
| 104 |
+
"eval_loss": 0.3373560005943525,
|
| 105 |
+
"eval_mse_loss": 0.18800275102597938,
|
| 106 |
+
"eval_rec_loss": 0.005269206022433341,
|
| 107 |
+
"flow/cos_sim": 0.9026168323542973,
|
| 108 |
+
"flow/improvement_ratio": 0.9768092384621433,
|
| 109 |
+
"flow/mag_ratio_mean": 0.9078535956062682,
|
| 110 |
+
"flow/mag_ratio_std": 0.12002966954419601,
|
| 111 |
+
"step": 2048
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.09459147383492679,
|
| 115 |
+
"eval_bleu": 0.6478349738418474,
|
| 116 |
+
"eval_ce_loss": 1.3434572242982856,
|
| 117 |
+
"eval_cos_loss": 0.09738318450306649,
|
| 118 |
+
"eval_loss": 0.3373560005943525,
|
| 119 |
+
"eval_mse_loss": 0.18800275102597938,
|
| 120 |
+
"eval_rec_loss": 0.005269206022433341,
|
| 121 |
+
"eval_runtime": 143.5996,
|
| 122 |
+
"eval_samples_per_second": 194.938,
|
| 123 |
+
"eval_steps_per_second": 3.05,
|
| 124 |
+
"flow/cos_sim": 0.9026168323542973,
|
| 125 |
+
"flow/improvement_ratio": 0.9768092384621433,
|
| 126 |
+
"flow/mag_ratio_mean": 0.9078535956062682,
|
| 127 |
+
"flow/mag_ratio_std": 0.12002966954419601,
|
| 128 |
+
"step": 2048
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.10641540806429264,
|
| 132 |
+
"grad_norm": 0.6462650299072266,
|
| 133 |
+
"learning_rate": 0.0009823925488998885,
|
| 134 |
+
"loss": 0.3380688428878784,
|
| 135 |
+
"step": 2304
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1182393422936585,
|
| 139 |
+
"grad_norm": 0.6545562744140625,
|
| 140 |
+
"learning_rate": 0.0009770409161149525,
|
| 141 |
+
"loss": 0.3214532732963562,
|
| 142 |
+
"step": 2560
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.13006327652302435,
|
| 146 |
+
"grad_norm": 0.6667947173118591,
|
| 147 |
+
"learning_rate": 0.0009709988622506973,
|
| 148 |
+
"loss": 0.31182003021240234,
|
| 149 |
+
"step": 2816
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.1418872107523902,
|
| 153 |
+
"grad_norm": 0.600439727306366,
|
| 154 |
+
"learning_rate": 0.000964275131968659,
|
| 155 |
+
"loss": 0.3082171380519867,
|
| 156 |
+
"step": 3072
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1418872107523902,
|
| 160 |
+
"eval_bleu": 0.6794402302458825,
|
| 161 |
+
"eval_ce_loss": 1.18399681810919,
|
| 162 |
+
"eval_cos_loss": 0.08983433964453875,
|
| 163 |
+
"eval_loss": 0.3044066195414491,
|
| 164 |
+
"eval_mse_loss": 0.17457320809908652,
|
| 165 |
+
"eval_rec_loss": 0.0024502936500277726,
|
| 166 |
+
"flow/cos_sim": 0.9101656942062726,
|
| 167 |
+
"flow/improvement_ratio": 0.9734727941691603,
|
| 168 |
+
"flow/mag_ratio_mean": 0.9100713347463303,
|
| 169 |
+
"flow/mag_ratio_std": 0.09228039255629392,
|
| 170 |
+
"step": 3072
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.1418872107523902,
|
| 174 |
+
"eval_bleu": 0.6794402302458825,
|
| 175 |
+
"eval_ce_loss": 1.18399681810919,
|
| 176 |
+
"eval_cos_loss": 0.08983433964453875,
|
| 177 |
+
"eval_loss": 0.3044066195414491,
|
| 178 |
+
"eval_mse_loss": 0.17457320809908652,
|
| 179 |
+
"eval_rec_loss": 0.0024502936500277726,
|
| 180 |
+
"eval_runtime": 142.616,
|
| 181 |
+
"eval_samples_per_second": 196.282,
|
| 182 |
+
"eval_steps_per_second": 3.071,
|
| 183 |
+
"flow/cos_sim": 0.9101656942062726,
|
| 184 |
+
"flow/improvement_ratio": 0.9734727941691603,
|
| 185 |
+
"flow/mag_ratio_mean": 0.9100713347463303,
|
| 186 |
+
"flow/mag_ratio_std": 0.09228039255629392,
|
| 187 |
+
"step": 3072
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.15371114498175603,
|
| 191 |
+
"grad_norm": 0.5188813209533691,
|
| 192 |
+
"learning_rate": 0.0009568794565203123,
|
| 193 |
+
"loss": 0.3006798326969147,
|
| 194 |
+
"step": 3328
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.1655350792111219,
|
| 198 |
+
"grad_norm": 0.3702296316623688,
|
| 199 |
+
"learning_rate": 0.0009488225396630347,
|
| 200 |
+
"loss": 0.29647037386894226,
|
| 201 |
+
"step": 3584
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.17735901344048774,
|
| 205 |
+
"grad_norm": 0.49310481548309326,
|
| 206 |
+
"learning_rate": 0.0009401160421685646,
|
| 207 |
+
"loss": 0.3130263388156891,
|
| 208 |
+
"step": 3840
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"grad_norm": 0.5644649267196655,
|
| 213 |
+
"learning_rate": 0.0009307725649463714,
|
| 214 |
+
"loss": 0.2926834523677826,
|
| 215 |
+
"step": 4096
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.18918294766985358,
|
| 219 |
+
"eval_bleu": 0.6839715035857402,
|
| 220 |
+
"eval_ce_loss": 1.1360810732460458,
|
| 221 |
+
"eval_cos_loss": 0.08503438343536364,
|
| 222 |
+
"eval_loss": 0.2889118402788084,
|
| 223 |
+
"eval_mse_loss": 0.16517186124030858,
|
| 224 |
+
"eval_rec_loss": 0.0016284317239192974,
|
| 225 |
+
"flow/cos_sim": 0.9149656362457362,
|
| 226 |
+
"flow/improvement_ratio": 0.9750959768415042,
|
| 227 |
+
"flow/mag_ratio_mean": 0.9128544130553938,
|
| 228 |
+
"flow/mag_ratio_std": 0.09354445787325297,
|
| 229 |
+
"step": 4096
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.18918294766985358,
|
| 233 |
+
"eval_bleu": 0.6839715035857402,
|
| 234 |
+
"eval_ce_loss": 1.1360810732460458,
|
| 235 |
+
"eval_cos_loss": 0.08503438343536364,
|
| 236 |
+
"eval_loss": 0.2889118402788084,
|
| 237 |
+
"eval_mse_loss": 0.16517186124030858,
|
| 238 |
+
"eval_rec_loss": 0.0016284317239192974,
|
| 239 |
+
"eval_runtime": 143.3522,
|
| 240 |
+
"eval_samples_per_second": 195.274,
|
| 241 |
+
"eval_steps_per_second": 3.055,
|
| 242 |
+
"flow/cos_sim": 0.9149656362457362,
|
| 243 |
+
"flow/improvement_ratio": 0.9750959768415042,
|
| 244 |
+
"flow/mag_ratio_mean": 0.9128544130553938,
|
| 245 |
+
"flow/mag_ratio_std": 0.09354445787325297,
|
| 246 |
+
"step": 4096
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.20100688189921945,
|
| 250 |
+
"grad_norm": 0.3648587763309479,
|
| 251 |
+
"learning_rate": 0.0009208056308063659,
|
| 252 |
+
"loss": 0.28915518522262573,
|
| 253 |
+
"step": 4352
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.2128308161285853,
|
| 257 |
+
"grad_norm": 0.5004331469535828,
|
| 258 |
+
"learning_rate": 0.0009102296648873445,
|
| 259 |
+
"loss": 0.28424444794654846,
|
| 260 |
+
"step": 4608
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.22465475035795113,
|
| 264 |
+
"grad_norm": 0.27222171425819397,
|
| 265 |
+
"learning_rate": 0.0008990599737794927,
|
| 266 |
+
"loss": 0.2838270366191864,
|
| 267 |
+
"step": 4864
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.236478684587317,
|
| 271 |
+
"grad_norm": 0.39166298508644104,
|
| 272 |
+
"learning_rate": 0.0008873127233711644,
|
| 273 |
+
"loss": 0.2802566885948181,
|
| 274 |
+
"step": 5120
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.236478684587317,
|
| 278 |
+
"eval_bleu": 0.6966429437264656,
|
| 279 |
+
"eval_ce_loss": 1.0860335096784923,
|
| 280 |
+
"eval_cos_loss": 0.08069509963534738,
|
| 281 |
+
"eval_loss": 0.27587767567944854,
|
| 282 |
+
"eval_mse_loss": 0.15802378180228413,
|
| 283 |
+
"eval_rec_loss": 0.0011810324893420196,
|
| 284 |
+
"flow/cos_sim": 0.9193049165756191,
|
| 285 |
+
"flow/improvement_ratio": 0.9735453040360316,
|
| 286 |
+
"flow/mag_ratio_mean": 0.9165048223652251,
|
| 287 |
+
"flow/mag_ratio_std": 0.08227625193252955,
|
| 288 |
+
"step": 5120
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.236478684587317,
|
| 292 |
+
"eval_bleu": 0.6966429437264656,
|
| 293 |
+
"eval_ce_loss": 1.0860335096784923,
|
| 294 |
+
"eval_cos_loss": 0.08069509963534738,
|
| 295 |
+
"eval_loss": 0.27587767567944854,
|
| 296 |
+
"eval_mse_loss": 0.15802378180228413,
|
| 297 |
+
"eval_rec_loss": 0.0011810324893420196,
|
| 298 |
+
"eval_runtime": 144.8851,
|
| 299 |
+
"eval_samples_per_second": 193.208,
|
| 300 |
+
"eval_steps_per_second": 3.023,
|
| 301 |
+
"flow/cos_sim": 0.9193049165756191,
|
| 302 |
+
"flow/improvement_ratio": 0.9735453040360316,
|
| 303 |
+
"flow/mag_ratio_mean": 0.9165048223652251,
|
| 304 |
+
"flow/mag_ratio_std": 0.08227625193252955,
|
| 305 |
+
"step": 5120
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.24830261881668284,
|
| 309 |
+
"grad_norm": 0.4552461802959442,
|
| 310 |
+
"learning_rate": 0.0008750049154520011,
|
| 311 |
+
"loss": 0.27747780084609985,
|
| 312 |
+
"step": 5376
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.2601265530460487,
|
| 316 |
+
"grad_norm": 0.3430848717689514,
|
| 317 |
+
"learning_rate": 0.0008621543631062487,
|
| 318 |
+
"loss": 0.2891384959220886,
|
| 319 |
+
"step": 5632
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.27195048727541454,
|
| 323 |
+
"grad_norm": 0.4337022304534912,
|
| 324 |
+
"learning_rate": 0.0008487796649318904,
|
| 325 |
+
"loss": 0.274080365896225,
|
| 326 |
+
"step": 5888
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.2837744215047804,
|
| 330 |
+
"grad_norm": 0.40342026948928833,
|
| 331 |
+
"learning_rate": 0.0008349001781229053,
|
| 332 |
+
"loss": 0.2737179100513458,
|
| 333 |
+
"step": 6144
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.2837744215047804,
|
| 337 |
+
"eval_bleu": 0.7005510372485733,
|
| 338 |
+
"eval_ce_loss": 1.067978764981984,
|
| 339 |
+
"eval_cos_loss": 0.0791517910994079,
|
| 340 |
+
"eval_loss": 0.2710464134744313,
|
| 341 |
+
"eval_mse_loss": 0.15530247285485813,
|
| 342 |
+
"eval_rec_loss": 0.00103088276002016,
|
| 343 |
+
"flow/cos_sim": 0.9208482295682986,
|
| 344 |
+
"flow/improvement_ratio": 0.9731569932476026,
|
| 345 |
+
"flow/mag_ratio_mean": 0.9284612283314744,
|
| 346 |
+
"flow/mag_ratio_std": 0.07813496222574962,
|
| 347 |
+
"step": 6144
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.2837744215047804,
|
| 351 |
+
"eval_bleu": 0.7005510372485733,
|
| 352 |
+
"eval_ce_loss": 1.067978764981984,
|
| 353 |
+
"eval_cos_loss": 0.0791517910994079,
|
| 354 |
+
"eval_loss": 0.2710464134744313,
|
| 355 |
+
"eval_mse_loss": 0.15530247285485813,
|
| 356 |
+
"eval_rec_loss": 0.00103088276002016,
|
| 357 |
+
"eval_runtime": 142.9519,
|
| 358 |
+
"eval_samples_per_second": 195.821,
|
| 359 |
+
"eval_steps_per_second": 3.064,
|
| 360 |
+
"flow/cos_sim": 0.9208482295682986,
|
| 361 |
+
"flow/improvement_ratio": 0.9731569932476026,
|
| 362 |
+
"flow/mag_ratio_mean": 0.9284612283314744,
|
| 363 |
+
"flow/mag_ratio_std": 0.07813496222574962,
|
| 364 |
+
"step": 6144
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.2955983557341462,
|
| 368 |
+
"grad_norm": 0.36669906973838806,
|
| 369 |
+
"learning_rate": 0.0008205359904536107,
|
| 370 |
+
"loss": 0.2686825394630432,
|
| 371 |
+
"step": 6400
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.30742228996351206,
|
| 375 |
+
"grad_norm": 0.2844366729259491,
|
| 376 |
+
"learning_rate": 0.0008057078912056363,
|
| 377 |
+
"loss": 0.26841098070144653,
|
| 378 |
+
"step": 6656
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.3192462241928779,
|
| 382 |
+
"grad_norm": 0.341339647769928,
|
| 383 |
+
"learning_rate": 0.0007904373410796086,
|
| 384 |
+
"loss": 0.2661970555782318,
|
| 385 |
+
"step": 6912
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.3310701584222438,
|
| 389 |
+
"grad_norm": 0.35041218996047974,
|
| 390 |
+
"learning_rate": 0.0007747464411350876,
|
| 391 |
+
"loss": 0.26556745171546936,
|
| 392 |
+
"step": 7168
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.3310701584222438,
|
| 396 |
+
"eval_bleu": 0.710653856755971,
|
| 397 |
+
"eval_ce_loss": 1.0037036097213015,
|
| 398 |
+
"eval_cos_loss": 0.0774884378733156,
|
| 399 |
+
"eval_loss": 0.2617515635817018,
|
| 400 |
+
"eval_mse_loss": 0.15282062886784611,
|
| 401 |
+
"eval_rec_loss": 0.0008117282769037305,
|
| 402 |
+
"flow/cos_sim": 0.9225115811443765,
|
| 403 |
+
"flow/improvement_ratio": 0.9752856880834658,
|
| 404 |
+
"flow/mag_ratio_mean": 0.9207794453999768,
|
| 405 |
+
"flow/mag_ratio_std": 0.08132727078447059,
|
| 406 |
+
"step": 7168
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.3310701584222438,
|
| 410 |
+
"eval_bleu": 0.710653856755971,
|
| 411 |
+
"eval_ce_loss": 1.0037036097213015,
|
| 412 |
+
"eval_cos_loss": 0.0774884378733156,
|
| 413 |
+
"eval_loss": 0.2617515635817018,
|
| 414 |
+
"eval_mse_loss": 0.15282062886784611,
|
| 415 |
+
"eval_rec_loss": 0.0008117282769037305,
|
| 416 |
+
"eval_runtime": 142.6116,
|
| 417 |
+
"eval_samples_per_second": 196.288,
|
| 418 |
+
"eval_steps_per_second": 3.071,
|
| 419 |
+
"flow/cos_sim": 0.9225115811443765,
|
| 420 |
+
"flow/improvement_ratio": 0.9752856880834658,
|
| 421 |
+
"flow/mag_ratio_mean": 0.9207794453999768,
|
| 422 |
+
"flow/mag_ratio_std": 0.08132727078447059,
|
| 423 |
+
"step": 7168
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.34289409265160964,
|
| 427 |
+
"grad_norm": 0.328535258769989,
|
| 428 |
+
"learning_rate": 0.000758657900803716,
|
| 429 |
+
"loss": 0.27869582176208496,
|
| 430 |
+
"step": 7424
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3547180268809755,
|
| 434 |
+
"grad_norm": 0.27767524123191833,
|
| 435 |
+
"learning_rate": 0.000742195005021869,
|
| 436 |
+
"loss": 0.26837044954299927,
|
| 437 |
+
"step": 7680
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3665419611103413,
|
| 441 |
+
"grad_norm": 0.40343427658081055,
|
| 442 |
+
"learning_rate": 0.0007253815805303786,
|
| 443 |
+
"loss": 0.26287224888801575,
|
| 444 |
+
"step": 7936
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.37836589533970716,
|
| 448 |
+
"grad_norm": 0.3494727611541748,
|
| 449 |
+
"learning_rate": 0.0007082419613901028,
|
| 450 |
+
"loss": 0.2608710527420044,
|
| 451 |
+
"step": 8192
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.37836589533970716,
|
| 455 |
+
"eval_bleu": 0.7258561394806508,
|
| 456 |
+
"eval_ce_loss": 0.954990051584701,
|
| 457 |
+
"eval_cos_loss": 0.07798272708099183,
|
| 458 |
+
"eval_loss": 0.2584831827239359,
|
| 459 |
+
"eval_mse_loss": 0.15443320324023566,
|
| 460 |
+
"eval_rec_loss": 0.0007526992644982298,
|
| 461 |
+
"flow/cos_sim": 0.9220172950666244,
|
| 462 |
+
"flow/improvement_ratio": 0.9747015749482804,
|
| 463 |
+
"flow/mag_ratio_mean": 0.9233354187175019,
|
| 464 |
+
"flow/mag_ratio_std": 0.07788482179108276,
|
| 465 |
+
"step": 8192
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.37836589533970716,
|
| 469 |
+
"eval_bleu": 0.7258561394806508,
|
| 470 |
+
"eval_ce_loss": 0.954990051584701,
|
| 471 |
+
"eval_cos_loss": 0.07798272708099183,
|
| 472 |
+
"eval_loss": 0.2584831827239359,
|
| 473 |
+
"eval_mse_loss": 0.15443320324023566,
|
| 474 |
+
"eval_rec_loss": 0.0007526992644982298,
|
| 475 |
+
"eval_runtime": 144.1923,
|
| 476 |
+
"eval_samples_per_second": 194.137,
|
| 477 |
+
"eval_steps_per_second": 3.038,
|
| 478 |
+
"flow/cos_sim": 0.9220172950666244,
|
| 479 |
+
"flow/improvement_ratio": 0.9747015749482804,
|
| 480 |
+
"flow/mag_ratio_mean": 0.9233354187175019,
|
| 481 |
+
"flow/mag_ratio_std": 0.07788482179108276,
|
| 482 |
+
"step": 8192
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.390189829569073,
|
| 486 |
+
"grad_norm": 0.4279082417488098,
|
| 487 |
+
"learning_rate": 0.0006908009537632514,
|
| 488 |
+
"loss": 0.2599097788333893,
|
| 489 |
+
"step": 8448
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.4020137637984389,
|
| 493 |
+
"grad_norm": 0.43606042861938477,
|
| 494 |
+
"learning_rate": 0.0006730838000114403,
|
| 495 |
+
"loss": 0.26111704111099243,
|
| 496 |
+
"step": 8704
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.41383769802780473,
|
| 500 |
+
"grad_norm": 0.21560809016227722,
|
| 501 |
+
"learning_rate": 0.0006551161421624341,
|
| 502 |
+
"loss": 0.268413245677948,
|
| 503 |
+
"step": 8960
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.4256616322571706,
|
| 507 |
+
"grad_norm": 0.5756855607032776,
|
| 508 |
+
"learning_rate": 0.0006369239847984517,
|
| 509 |
+
"loss": 0.2562841475009918,
|
| 510 |
+
"step": 9216
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 0.4256616322571706,
|
| 514 |
+
"eval_bleu": 0.7095258647762687,
|
| 515 |
+
"eval_ce_loss": 1.0158189947474492,
|
| 516 |
+
"eval_cos_loss": 0.0742807937743457,
|
| 517 |
+
"eval_loss": 0.256784121705789,
|
| 518 |
+
"eval_mse_loss": 0.14713338680871546,
|
| 519 |
+
"eval_rec_loss": 0.0006407552403591884,
|
| 520 |
+
"flow/cos_sim": 0.9257192251072627,
|
| 521 |
+
"flow/improvement_ratio": 0.9743275895510635,
|
| 522 |
+
"flow/mag_ratio_mean": 0.9202466026016566,
|
| 523 |
+
"flow/mag_ratio_std": 0.0764387545588354,
|
| 524 |
+
"step": 9216
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.4256616322571706,
|
| 528 |
+
"eval_bleu": 0.7095258647762687,
|
| 529 |
+
"eval_ce_loss": 1.0158189947474492,
|
| 530 |
+
"eval_cos_loss": 0.0742807937743457,
|
| 531 |
+
"eval_loss": 0.256784121705789,
|
| 532 |
+
"eval_mse_loss": 0.14713338680871546,
|
| 533 |
+
"eval_rec_loss": 0.0006407552403591884,
|
| 534 |
+
"eval_runtime": 145.076,
|
| 535 |
+
"eval_samples_per_second": 192.954,
|
| 536 |
+
"eval_steps_per_second": 3.019,
|
| 537 |
+
"flow/cos_sim": 0.9257192251072627,
|
| 538 |
+
"flow/improvement_ratio": 0.9743275895510635,
|
| 539 |
+
"flow/mag_ratio_mean": 0.9202466026016566,
|
| 540 |
+
"flow/mag_ratio_std": 0.0764387545588354,
|
| 541 |
+
"step": 9216
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.4374855664865364,
|
| 545 |
+
"grad_norm": 0.376579225063324,
|
| 546 |
+
"learning_rate": 0.0006185336574197479,
|
| 547 |
+
"loss": 0.25477761030197144,
|
| 548 |
+
"step": 9472
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.44930950071590225,
|
| 552 |
+
"grad_norm": 0.4893060624599457,
|
| 553 |
+
"learning_rate": 0.0005999717763379407,
|
| 554 |
+
"loss": 0.2554967701435089,
|
| 555 |
+
"step": 9728
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.4611334349452681,
|
| 559 |
+
"grad_norm": 0.5277368426322937,
|
| 560 |
+
"learning_rate": 0.0005812652061542363,
|
| 561 |
+
"loss": 0.2537921667098999,
|
| 562 |
+
"step": 9984
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.472957369174634,
|
| 566 |
+
"grad_norm": 0.3493780791759491,
|
| 567 |
+
"learning_rate": 0.0005624410208783071,
|
| 568 |
+
"loss": 0.2518165707588196,
|
| 569 |
+
"step": 10240
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.472957369174634,
|
| 573 |
+
"eval_bleu": 0.7175600139987302,
|
| 574 |
+
"eval_ce_loss": 0.9684542848231041,
|
| 575 |
+
"eval_cos_loss": 0.07330035061902923,
|
| 576 |
+
"eval_loss": 0.25065779049782994,
|
| 577 |
+
"eval_mse_loss": 0.1458930650292194,
|
| 578 |
+
"eval_rec_loss": 0.000589260474122746,
|
| 579 |
+
"flow/cos_sim": 0.9266996725236989,
|
| 580 |
+
"flow/improvement_ratio": 0.9765301047394809,
|
| 581 |
+
"flow/mag_ratio_mean": 0.9194769399351181,
|
| 582 |
+
"flow/mag_ratio_std": 0.07280441308946914,
|
| 583 |
+
"step": 10240
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.472957369174634,
|
| 587 |
+
"eval_bleu": 0.7175600139987302,
|
| 588 |
+
"eval_ce_loss": 0.9684542848231041,
|
| 589 |
+
"eval_cos_loss": 0.07330035061902923,
|
| 590 |
+
"eval_loss": 0.25065779049782994,
|
| 591 |
+
"eval_mse_loss": 0.1458930650292194,
|
| 592 |
+
"eval_rec_loss": 0.000589260474122746,
|
| 593 |
+
"eval_runtime": 144.907,
|
| 594 |
+
"eval_samples_per_second": 193.179,
|
| 595 |
+
"eval_steps_per_second": 3.023,
|
| 596 |
+
"flow/cos_sim": 0.9266996725236989,
|
| 597 |
+
"flow/improvement_ratio": 0.9765301047394809,
|
| 598 |
+
"flow/mag_ratio_mean": 0.9194769399351181,
|
| 599 |
+
"flow/mag_ratio_std": 0.07280441308946914,
|
| 600 |
+
"step": 10240
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.48478130340399983,
|
| 604 |
+
"grad_norm": 0.5341387391090393,
|
| 605 |
+
"learning_rate": 0.0005435264647440881,
|
| 606 |
+
"loss": 0.2537766695022583,
|
| 607 |
+
"step": 10496
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.49660523763336567,
|
| 611 |
+
"grad_norm": 0.37015679478645325,
|
| 612 |
+
"learning_rate": 0.000524548912779213,
|
| 613 |
+
"loss": 0.2499093860387802,
|
| 614 |
+
"step": 10752
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5084291718627315,
|
| 618 |
+
"grad_norm": 0.6591759920120239,
|
| 619 |
+
"learning_rate": 0.0005055358311851499,
|
| 620 |
+
"loss": 0.25186970829963684,
|
| 621 |
+
"step": 11008
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5202531060920974,
|
| 625 |
+
"grad_norm": 0.31277748942375183,
|
| 626 |
+
"learning_rate": 0.0004865147375853812,
|
| 627 |
+
"loss": 0.25062766671180725,
|
| 628 |
+
"step": 11264
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 0.5202531060920974,
|
| 632 |
+
"eval_bleu": 0.7260803484693739,
|
| 633 |
+
"eval_ce_loss": 0.9400546782092961,
|
| 634 |
+
"eval_cos_loss": 0.07318596699228298,
|
| 635 |
+
"eval_loss": 0.2483225218169221,
|
| 636 |
+
"eval_mse_loss": 0.14649179712210073,
|
| 637 |
+
"eval_rec_loss": 0.0005066586955975142,
|
| 638 |
+
"flow/cos_sim": 0.9268140568308634,
|
| 639 |
+
"flow/improvement_ratio": 0.9762515691045213,
|
| 640 |
+
"flow/mag_ratio_mean": 0.9157658136326429,
|
| 641 |
+
"flow/mag_ratio_std": 0.07639177793373256,
|
| 642 |
+
"step": 11264
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.5202531060920974,
|
| 646 |
+
"eval_bleu": 0.7260803484693739,
|
| 647 |
+
"eval_ce_loss": 0.9400546782092961,
|
| 648 |
+
"eval_cos_loss": 0.07318596699228298,
|
| 649 |
+
"eval_loss": 0.2483225218169221,
|
| 650 |
+
"eval_mse_loss": 0.14649179712210073,
|
| 651 |
+
"eval_rec_loss": 0.0005066586955975142,
|
| 652 |
+
"eval_runtime": 143.4087,
|
| 653 |
+
"eval_samples_per_second": 195.197,
|
| 654 |
+
"eval_steps_per_second": 3.054,
|
| 655 |
+
"flow/cos_sim": 0.9268140568308634,
|
| 656 |
+
"flow/improvement_ratio": 0.9762515691045213,
|
| 657 |
+
"flow/mag_ratio_mean": 0.9157658136326429,
|
| 658 |
+
"flow/mag_ratio_std": 0.07639177793373256,
|
| 659 |
+
"step": 11264
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.5320770403214632,
|
| 663 |
+
"grad_norm": 0.7305059432983398,
|
| 664 |
+
"learning_rate": 0.0004675131611991607,
|
| 665 |
+
"loss": 0.2499980330467224,
|
| 666 |
+
"step": 11520
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.5439009745508291,
|
| 670 |
+
"grad_norm": 0.2771352231502533,
|
| 671 |
+
"learning_rate": 0.0004485586029984899,
|
| 672 |
+
"loss": 0.24800576269626617,
|
| 673 |
+
"step": 11776
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.5557249087801949,
|
| 677 |
+
"grad_norm": 0.2686781883239746,
|
| 678 |
+
"learning_rate": 0.00042967849590597266,
|
| 679 |
+
"loss": 0.24677802622318268,
|
| 680 |
+
"step": 12032
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.5675488430095608,
|
| 684 |
+
"grad_norm": 0.37717506289482117,
|
| 685 |
+
"learning_rate": 0.0004109001650911621,
|
| 686 |
+
"loss": 0.24666211009025574,
|
| 687 |
+
"step": 12288
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.5675488430095608,
|
| 691 |
+
"eval_bleu": 0.724567852062454,
|
| 692 |
+
"eval_ce_loss": 0.9509265820185343,
|
| 693 |
+
"eval_cos_loss": 0.07338333341663014,
|
| 694 |
+
"eval_loss": 0.24947158049911125,
|
| 695 |
+
"eval_mse_loss": 0.1465663725518745,
|
| 696 |
+
"eval_rec_loss": 0.00047421499268288965,
|
| 697 |
+
"flow/cos_sim": 0.9266166863920482,
|
| 698 |
+
"flow/improvement_ratio": 0.9748387396607769,
|
| 699 |
+
"flow/mag_ratio_mean": 0.9328112143631939,
|
| 700 |
+
"flow/mag_ratio_std": 0.07196474395113993,
|
| 701 |
+
"step": 12288
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.5675488430095608,
|
| 705 |
+
"eval_bleu": 0.724567852062454,
|
| 706 |
+
"eval_ce_loss": 0.9509265820185343,
|
| 707 |
+
"eval_cos_loss": 0.07338333341663014,
|
| 708 |
+
"eval_loss": 0.24947158049911125,
|
| 709 |
+
"eval_mse_loss": 0.1465663725518745,
|
| 710 |
+
"eval_rec_loss": 0.00047421499268288965,
|
| 711 |
+
"eval_runtime": 144.4458,
|
| 712 |
+
"eval_samples_per_second": 193.796,
|
| 713 |
+
"eval_steps_per_second": 3.032,
|
| 714 |
+
"flow/cos_sim": 0.9266166863920482,
|
| 715 |
+
"flow/improvement_ratio": 0.9748387396607769,
|
| 716 |
+
"flow/mag_ratio_mean": 0.9328112143631939,
|
| 717 |
+
"flow/mag_ratio_std": 0.07196474395113993,
|
| 718 |
+
"step": 12288
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.5793727772389267,
|
| 722 |
+
"grad_norm": 0.7573108673095703,
|
| 723 |
+
"learning_rate": 0.0003922507884228551,
|
| 724 |
+
"loss": 0.24700023233890533,
|
| 725 |
+
"step": 12544
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.5911967114682924,
|
| 729 |
+
"grad_norm": 0.2122378647327423,
|
| 730 |
+
"learning_rate": 0.00037375735713457723,
|
| 731 |
+
"loss": 0.2433316856622696,
|
| 732 |
+
"step": 12800
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.6030206456976583,
|
| 736 |
+
"grad_norm": 0.22209292650222778,
|
| 737 |
+
"learning_rate": 0.00035544663676018276,
|
| 738 |
+
"loss": 0.246763676404953,
|
| 739 |
+
"step": 13056
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.6148445799270241,
|
| 743 |
+
"grad_norm": 0.309989333152771,
|
| 744 |
+
"learning_rate": 0.00033734512839611255,
|
| 745 |
+
"loss": 0.24262374639511108,
|
| 746 |
+
"step": 13312
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.6148445799270241,
|
| 750 |
+
"eval_bleu": 0.7300156788258321,
|
| 751 |
+
"eval_ce_loss": 0.933146977737614,
|
| 752 |
+
"eval_cos_loss": 0.07192521648626077,
|
| 753 |
+
"eval_loss": 0.24540512995224564,
|
| 754 |
+
"eval_mse_loss": 0.14446957482526837,
|
| 755 |
+
"eval_rec_loss": 0.00042833433703537476,
|
| 756 |
+
"flow/cos_sim": 0.9280747922315989,
|
| 757 |
+
"flow/improvement_ratio": 0.9743193072271129,
|
| 758 |
+
"flow/mag_ratio_mean": 0.9250430587219866,
|
| 759 |
+
"flow/mag_ratio_std": 0.07064938510355612,
|
| 760 |
+
"step": 13312
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.6148445799270241,
|
| 764 |
+
"eval_bleu": 0.7300156788258321,
|
| 765 |
+
"eval_ce_loss": 0.933146977737614,
|
| 766 |
+
"eval_cos_loss": 0.07192521648626077,
|
| 767 |
+
"eval_loss": 0.24540512995224564,
|
| 768 |
+
"eval_mse_loss": 0.14446957482526837,
|
| 769 |
+
"eval_rec_loss": 0.00042833433703537476,
|
| 770 |
+
"eval_runtime": 144.2347,
|
| 771 |
+
"eval_samples_per_second": 194.079,
|
| 772 |
+
"eval_steps_per_second": 3.037,
|
| 773 |
+
"flow/cos_sim": 0.9280747922315989,
|
| 774 |
+
"flow/improvement_ratio": 0.9743193072271129,
|
| 775 |
+
"flow/mag_ratio_mean": 0.9250430587219866,
|
| 776 |
+
"flow/mag_ratio_std": 0.07064938510355612,
|
| 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 sha256:61a16bb839f687414b8e48611327c4b9cfddeefe38c031ca70808f9a97c476b7
|
| 3 |
+
size 5137
|