Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-v5.14/checkpoint-21504/ema.safetensors +3 -0
- checkpoints-v5.14/checkpoint-21504/eval_state.json +3 -0
- checkpoints-v5.14/checkpoint-21504/model.safetensors +3 -0
- checkpoints-v5.14/checkpoint-21504/optimizer.pt +3 -0
- checkpoints-v5.14/checkpoint-21504/rng_state.pth +3 -0
- checkpoints-v5.14/checkpoint-21504/scaler.pt +3 -0
- checkpoints-v5.14/checkpoint-21504/scheduler.pt +3 -0
- checkpoints-v5.14/checkpoint-21504/trainer_state.json +790 -0
- checkpoints-v5.14/checkpoint-21504/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -66,3 +66,4 @@ checkpoints-v4.6+/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs
|
|
| 66 |
checkpoints-v5.13-c/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 67 |
checkpoints-v5.14/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 68 |
checkpoints-v4.6++/checkpoint-2048/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 66 |
checkpoints-v5.13-c/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 67 |
checkpoints-v5.14/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 68 |
checkpoints-v4.6++/checkpoint-2048/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 69 |
+
checkpoints-v5.14/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-v5.14/checkpoint-21504/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ae6169501d9e43630e70c2ca690583767a7ffa4120a40fba3848fbb0022f20e
|
| 3 |
+
size 54599376
|
checkpoints-v5.14/checkpoint-21504/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a4dbacc25b040e2f24970f53d48e8753df25d77f0818ee6859dc0c7f486554d
|
| 3 |
+
size 60888526
|
checkpoints-v5.14/checkpoint-21504/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:738e9c1c18a532ce6e34eaca576c03a23af7a618e77758472821130385173a99
|
| 3 |
+
size 54599408
|
checkpoints-v5.14/checkpoint-21504/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a48896df390473207a5ece14cbdfc8ae5473663bdb84cba5fcbc291fff022ca
|
| 3 |
+
size 76550347
|
checkpoints-v5.14/checkpoint-21504/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:68be200dd5afac5ffeee9329157efed8e99b5252978576a76158c0c50e8caf83
|
| 3 |
+
size 14645
|
checkpoints-v5.14/checkpoint-21504/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:992372cb435a366b554522120ba7a2848629239d69ea81fd324957d1bc803d95
|
| 3 |
+
size 1383
|
checkpoints-v5.14/checkpoint-21504/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dfc6cbde811b5a96de4547d7d760e5ca70342e4231e1fe6ca2d690d1aebbaf29
|
| 3 |
+
size 1465
|
checkpoints-v5.14/checkpoint-21504/trainer_state.json
ADDED
|
@@ -0,0 +1,790 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9267567392850217,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 21504,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.04413127329928675,
|
| 14 |
+
"grad_norm": 2.3938751220703125,
|
| 15 |
+
"learning_rate": 9.990234375e-05,
|
| 16 |
+
"loss": 7.639987945556641,
|
| 17 |
+
"step": 1024
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.04413127329928675,
|
| 21 |
+
"eval_bleu": 0.3665367955679541,
|
| 22 |
+
"eval_ce_clean_loss": 2.1982549197638215,
|
| 23 |
+
"eval_ce_pred_loss": 4.012731915597977,
|
| 24 |
+
"eval_flow_mse_loss": 0.6777123596622492,
|
| 25 |
+
"eval_loss": 3.277240484253938,
|
| 26 |
+
"flow/cos_sim": 0.5957727816059137,
|
| 27 |
+
"flow/improvement_ratio": 0.9601070273405453,
|
| 28 |
+
"flow/mag_ratio_mean": 0.5324122500317946,
|
| 29 |
+
"flow/mag_ratio_std": 0.14340894446888966,
|
| 30 |
+
"step": 1024
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04413127329928675,
|
| 34 |
+
"eval_bleu": 0.3665367955679541,
|
| 35 |
+
"eval_ce_clean_loss": 2.1982549197638215,
|
| 36 |
+
"eval_ce_pred_loss": 4.012731915597977,
|
| 37 |
+
"eval_flow_mse_loss": 0.6777123596622492,
|
| 38 |
+
"eval_loss": 3.277240484253938,
|
| 39 |
+
"eval_runtime": 210.3423,
|
| 40 |
+
"eval_samples_per_second": 142.625,
|
| 41 |
+
"eval_steps_per_second": 2.23,
|
| 42 |
+
"flow/cos_sim": 0.5957727816059137,
|
| 43 |
+
"flow/improvement_ratio": 0.9601070273405453,
|
| 44 |
+
"flow/mag_ratio_mean": 0.5324122500317946,
|
| 45 |
+
"flow/mag_ratio_std": 0.14340894446888966,
|
| 46 |
+
"step": 1024
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.0882625465985735,
|
| 50 |
+
"grad_norm": 2.127551794052124,
|
| 51 |
+
"learning_rate": 9.9476028157316e-05,
|
| 52 |
+
"loss": 1.9875907897949219,
|
| 53 |
+
"step": 2048
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.0882625465985735,
|
| 57 |
+
"eval_bleu": 0.47239301802941436,
|
| 58 |
+
"eval_ce_clean_loss": 0.4944258190548496,
|
| 59 |
+
"eval_ce_pred_loss": 2.776211817635656,
|
| 60 |
+
"eval_flow_mse_loss": 0.4178365491854865,
|
| 61 |
+
"eval_loss": 1.1898835559389485,
|
| 62 |
+
"flow/cos_sim": 0.7644255845023117,
|
| 63 |
+
"flow/improvement_ratio": 0.9770590040221143,
|
| 64 |
+
"flow/mag_ratio_mean": 0.7315180154243258,
|
| 65 |
+
"flow/mag_ratio_std": 0.20312415069735634,
|
| 66 |
+
"step": 2048
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.0882625465985735,
|
| 70 |
+
"eval_bleu": 0.47239301802941436,
|
| 71 |
+
"eval_ce_clean_loss": 0.4944258190548496,
|
| 72 |
+
"eval_ce_pred_loss": 2.776211817635656,
|
| 73 |
+
"eval_flow_mse_loss": 0.4178365491854865,
|
| 74 |
+
"eval_loss": 1.1898835559389485,
|
| 75 |
+
"eval_runtime": 205.6011,
|
| 76 |
+
"eval_samples_per_second": 145.914,
|
| 77 |
+
"eval_steps_per_second": 2.281,
|
| 78 |
+
"flow/cos_sim": 0.7644255845023117,
|
| 79 |
+
"flow/improvement_ratio": 0.9770590040221143,
|
| 80 |
+
"flow/mag_ratio_mean": 0.7315180154243258,
|
| 81 |
+
"flow/mag_ratio_std": 0.20312415069735634,
|
| 82 |
+
"step": 2048
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.13239381989786023,
|
| 86 |
+
"grad_norm": 0.5170137286186218,
|
| 87 |
+
"learning_rate": 9.791307026072513e-05,
|
| 88 |
+
"loss": 0.8957924842834473,
|
| 89 |
+
"step": 3072
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.13239381989786023,
|
| 93 |
+
"eval_bleu": 0.5217005787375273,
|
| 94 |
+
"eval_ce_clean_loss": 0.12680107509212962,
|
| 95 |
+
"eval_ce_pred_loss": 2.1341091328338266,
|
| 96 |
+
"eval_flow_mse_loss": 0.3295880077617255,
|
| 97 |
+
"eval_loss": 0.669800000404244,
|
| 98 |
+
"flow/cos_sim": 0.8211929453715586,
|
| 99 |
+
"flow/improvement_ratio": 0.9835899890358768,
|
| 100 |
+
"flow/mag_ratio_mean": 0.7927031459838851,
|
| 101 |
+
"flow/mag_ratio_std": 0.19854979916040832,
|
| 102 |
+
"step": 3072
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.13239381989786023,
|
| 106 |
+
"eval_bleu": 0.5217005787375273,
|
| 107 |
+
"eval_ce_clean_loss": 0.12680107509212962,
|
| 108 |
+
"eval_ce_pred_loss": 2.1341091328338266,
|
| 109 |
+
"eval_flow_mse_loss": 0.3295880077617255,
|
| 110 |
+
"eval_loss": 0.669800000404244,
|
| 111 |
+
"eval_runtime": 207.9355,
|
| 112 |
+
"eval_samples_per_second": 144.276,
|
| 113 |
+
"eval_steps_per_second": 2.256,
|
| 114 |
+
"flow/cos_sim": 0.8211929453715586,
|
| 115 |
+
"flow/improvement_ratio": 0.9835899890358768,
|
| 116 |
+
"flow/mag_ratio_mean": 0.7927031459838851,
|
| 117 |
+
"flow/mag_ratio_std": 0.19854979916040832,
|
| 118 |
+
"step": 3072
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.176525093197147,
|
| 122 |
+
"grad_norm": 0.926017701625824,
|
| 123 |
+
"learning_rate": 9.53439476074686e-05,
|
| 124 |
+
"loss": 0.6218850016593933,
|
| 125 |
+
"step": 4096
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.176525093197147,
|
| 129 |
+
"eval_bleu": 0.543675748301631,
|
| 130 |
+
"eval_ce_clean_loss": 0.05681392484938285,
|
| 131 |
+
"eval_ce_pred_loss": 1.8018918128918484,
|
| 132 |
+
"eval_flow_mse_loss": 0.3036336704040133,
|
| 133 |
+
"eval_loss": 0.5406367787038848,
|
| 134 |
+
"flow/cos_sim": 0.8355540915338724,
|
| 135 |
+
"flow/improvement_ratio": 0.9839625680116194,
|
| 136 |
+
"flow/mag_ratio_mean": 0.8140800381774334,
|
| 137 |
+
"flow/mag_ratio_std": 0.20049285866431335,
|
| 138 |
+
"step": 4096
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.176525093197147,
|
| 142 |
+
"eval_bleu": 0.543675748301631,
|
| 143 |
+
"eval_ce_clean_loss": 0.05681392484938285,
|
| 144 |
+
"eval_ce_pred_loss": 1.8018918128918484,
|
| 145 |
+
"eval_flow_mse_loss": 0.3036336704040133,
|
| 146 |
+
"eval_loss": 0.5406367787038848,
|
| 147 |
+
"eval_runtime": 208.0044,
|
| 148 |
+
"eval_samples_per_second": 144.228,
|
| 149 |
+
"eval_steps_per_second": 2.255,
|
| 150 |
+
"flow/cos_sim": 0.8355540915338724,
|
| 151 |
+
"flow/improvement_ratio": 0.9839625680116194,
|
| 152 |
+
"flow/mag_ratio_mean": 0.8140800381774334,
|
| 153 |
+
"flow/mag_ratio_std": 0.20049285866431335,
|
| 154 |
+
"step": 4096
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.22065636649643372,
|
| 158 |
+
"grad_norm": 0.8565777540206909,
|
| 159 |
+
"learning_rate": 9.182261125213742e-05,
|
| 160 |
+
"loss": 0.533421516418457,
|
| 161 |
+
"step": 5120
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.22065636649643372,
|
| 165 |
+
"eval_bleu": 0.5544799404246258,
|
| 166 |
+
"eval_ce_clean_loss": 0.03229137011237744,
|
| 167 |
+
"eval_ce_pred_loss": 1.5962520919120642,
|
| 168 |
+
"eval_flow_mse_loss": 0.2900454838202198,
|
| 169 |
+
"eval_loss": 0.481962066596505,
|
| 170 |
+
"flow/cos_sim": 0.8430635610114791,
|
| 171 |
+
"flow/improvement_ratio": 0.9848958956661509,
|
| 172 |
+
"flow/mag_ratio_mean": 0.8249702423111971,
|
| 173 |
+
"flow/mag_ratio_std": 0.18978321196427986,
|
| 174 |
+
"step": 5120
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.22065636649643372,
|
| 178 |
+
"eval_bleu": 0.5544799404246258,
|
| 179 |
+
"eval_ce_clean_loss": 0.03229137011237744,
|
| 180 |
+
"eval_ce_pred_loss": 1.5962520919120642,
|
| 181 |
+
"eval_flow_mse_loss": 0.2900454838202198,
|
| 182 |
+
"eval_loss": 0.481962066596505,
|
| 183 |
+
"eval_runtime": 209.2596,
|
| 184 |
+
"eval_samples_per_second": 143.363,
|
| 185 |
+
"eval_steps_per_second": 2.241,
|
| 186 |
+
"flow/cos_sim": 0.8430635610114791,
|
| 187 |
+
"flow/improvement_ratio": 0.9848958956661509,
|
| 188 |
+
"flow/mag_ratio_mean": 0.8249702423111971,
|
| 189 |
+
"flow/mag_ratio_std": 0.18978321196427986,
|
| 190 |
+
"step": 5120
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.26478763979572045,
|
| 194 |
+
"grad_norm": 0.459722101688385,
|
| 195 |
+
"learning_rate": 8.742770483354739e-05,
|
| 196 |
+
"loss": 0.4872491657733917,
|
| 197 |
+
"step": 6144
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.26478763979572045,
|
| 201 |
+
"eval_bleu": 0.5527687698410321,
|
| 202 |
+
"eval_ce_clean_loss": 0.020704815237681622,
|
| 203 |
+
"eval_ce_pred_loss": 1.5302676085724252,
|
| 204 |
+
"eval_flow_mse_loss": 0.27337305222365904,
|
| 205 |
+
"eval_loss": 0.4471046316471181,
|
| 206 |
+
"flow/cos_sim": 0.8521691379008263,
|
| 207 |
+
"flow/improvement_ratio": 0.9867461412700255,
|
| 208 |
+
"flow/mag_ratio_mean": 0.8359249935729671,
|
| 209 |
+
"flow/mag_ratio_std": 0.1747948888903742,
|
| 210 |
+
"step": 6144
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.26478763979572045,
|
| 214 |
+
"eval_bleu": 0.5527687698410321,
|
| 215 |
+
"eval_ce_clean_loss": 0.020704815237681622,
|
| 216 |
+
"eval_ce_pred_loss": 1.5302676085724252,
|
| 217 |
+
"eval_flow_mse_loss": 0.27337305222365904,
|
| 218 |
+
"eval_loss": 0.4471046316471181,
|
| 219 |
+
"eval_runtime": 209.107,
|
| 220 |
+
"eval_samples_per_second": 143.467,
|
| 221 |
+
"eval_steps_per_second": 2.243,
|
| 222 |
+
"flow/cos_sim": 0.8521691379008263,
|
| 223 |
+
"flow/improvement_ratio": 0.9867461412700255,
|
| 224 |
+
"flow/mag_ratio_mean": 0.8359249935729671,
|
| 225 |
+
"flow/mag_ratio_std": 0.1747948888903742,
|
| 226 |
+
"step": 6144
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.30891891309500724,
|
| 230 |
+
"grad_norm": 0.5934994220733643,
|
| 231 |
+
"learning_rate": 8.224294338515429e-05,
|
| 232 |
+
"loss": 0.45292773842811584,
|
| 233 |
+
"step": 7168
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.30891891309500724,
|
| 237 |
+
"eval_bleu": 0.5556870699184919,
|
| 238 |
+
"eval_ce_clean_loss": 0.014376114403554944,
|
| 239 |
+
"eval_ce_pred_loss": 1.4703641462681898,
|
| 240 |
+
"eval_flow_mse_loss": 0.25588471872974305,
|
| 241 |
+
"eval_loss": 0.4172972510618442,
|
| 242 |
+
"flow/cos_sim": 0.8633063356759452,
|
| 243 |
+
"flow/improvement_ratio": 0.9885822310884879,
|
| 244 |
+
"flow/mag_ratio_mean": 0.8422111430401995,
|
| 245 |
+
"flow/mag_ratio_std": 0.16398038163876483,
|
| 246 |
+
"step": 7168
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.30891891309500724,
|
| 250 |
+
"eval_bleu": 0.5556870699184919,
|
| 251 |
+
"eval_ce_clean_loss": 0.014376114403554944,
|
| 252 |
+
"eval_ce_pred_loss": 1.4703641462681898,
|
| 253 |
+
"eval_flow_mse_loss": 0.25588471872974305,
|
| 254 |
+
"eval_loss": 0.4172972510618442,
|
| 255 |
+
"eval_runtime": 209.704,
|
| 256 |
+
"eval_samples_per_second": 143.059,
|
| 257 |
+
"eval_steps_per_second": 2.236,
|
| 258 |
+
"flow/cos_sim": 0.8633063356759452,
|
| 259 |
+
"flow/improvement_ratio": 0.9885822310884879,
|
| 260 |
+
"flow/mag_ratio_mean": 0.8422111430401995,
|
| 261 |
+
"flow/mag_ratio_std": 0.16398038163876483,
|
| 262 |
+
"step": 7168
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.353050186394294,
|
| 266 |
+
"grad_norm": 0.3148420453071594,
|
| 267 |
+
"learning_rate": 7.638710244802891e-05,
|
| 268 |
+
"loss": 0.4292431175708771,
|
| 269 |
+
"step": 8192
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.353050186394294,
|
| 273 |
+
"eval_bleu": 0.5566824370986413,
|
| 274 |
+
"eval_ce_clean_loss": 0.010665512040082707,
|
| 275 |
+
"eval_ce_pred_loss": 1.437801378749327,
|
| 276 |
+
"eval_flow_mse_loss": 0.24223278170582582,
|
| 277 |
+
"eval_loss": 0.39667843399779884,
|
| 278 |
+
"flow/cos_sim": 0.8718424904575226,
|
| 279 |
+
"flow/improvement_ratio": 0.9907243636879586,
|
| 280 |
+
"flow/mag_ratio_mean": 0.8495659919689967,
|
| 281 |
+
"flow/mag_ratio_std": 0.1551145709780996,
|
| 282 |
+
"step": 8192
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.353050186394294,
|
| 286 |
+
"eval_bleu": 0.5566824370986413,
|
| 287 |
+
"eval_ce_clean_loss": 0.010665512040082707,
|
| 288 |
+
"eval_ce_pred_loss": 1.437801378749327,
|
| 289 |
+
"eval_flow_mse_loss": 0.24223278170582582,
|
| 290 |
+
"eval_loss": 0.39667843399779884,
|
| 291 |
+
"eval_runtime": 206.6493,
|
| 292 |
+
"eval_samples_per_second": 145.173,
|
| 293 |
+
"eval_steps_per_second": 2.27,
|
| 294 |
+
"flow/cos_sim": 0.8718424904575226,
|
| 295 |
+
"flow/improvement_ratio": 0.9907243636879586,
|
| 296 |
+
"flow/mag_ratio_mean": 0.8495659919689967,
|
| 297 |
+
"flow/mag_ratio_std": 0.1551145709780996,
|
| 298 |
+
"step": 8192
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.3971814596935807,
|
| 302 |
+
"grad_norm": 0.5599232912063599,
|
| 303 |
+
"learning_rate": 6.997172522088177e-05,
|
| 304 |
+
"loss": 0.41261157393455505,
|
| 305 |
+
"step": 9216
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.3971814596935807,
|
| 309 |
+
"eval_bleu": 0.5663410724389681,
|
| 310 |
+
"eval_ce_clean_loss": 0.008277329295051537,
|
| 311 |
+
"eval_ce_pred_loss": 1.3597534239165057,
|
| 312 |
+
"eval_flow_mse_loss": 0.2412289542747713,
|
| 313 |
+
"eval_loss": 0.38548162909967304,
|
| 314 |
+
"flow/cos_sim": 0.8725989290646144,
|
| 315 |
+
"flow/improvement_ratio": 0.9909563410256718,
|
| 316 |
+
"flow/mag_ratio_mean": 0.8561376066350225,
|
| 317 |
+
"flow/mag_ratio_std": 0.15248549610439902,
|
| 318 |
+
"step": 9216
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.3971814596935807,
|
| 322 |
+
"eval_bleu": 0.5663410724389681,
|
| 323 |
+
"eval_ce_clean_loss": 0.008277329295051537,
|
| 324 |
+
"eval_ce_pred_loss": 1.3597534239165057,
|
| 325 |
+
"eval_flow_mse_loss": 0.2412289542747713,
|
| 326 |
+
"eval_loss": 0.38548162909967304,
|
| 327 |
+
"eval_runtime": 208.3126,
|
| 328 |
+
"eval_samples_per_second": 144.014,
|
| 329 |
+
"eval_steps_per_second": 2.251,
|
| 330 |
+
"flow/cos_sim": 0.8725989290646144,
|
| 331 |
+
"flow/improvement_ratio": 0.9909563410256718,
|
| 332 |
+
"flow/mag_ratio_mean": 0.8561376066350225,
|
| 333 |
+
"flow/mag_ratio_std": 0.15248549610439902,
|
| 334 |
+
"step": 9216
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.44131273299286744,
|
| 338 |
+
"grad_norm": 0.5740413069725037,
|
| 339 |
+
"learning_rate": 6.314377890922702e-05,
|
| 340 |
+
"loss": 0.4009917378425598,
|
| 341 |
+
"step": 10240
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 0.44131273299286744,
|
| 345 |
+
"eval_bleu": 0.5765275121149777,
|
| 346 |
+
"eval_ce_clean_loss": 0.0067116839581353865,
|
| 347 |
+
"eval_ce_pred_loss": 1.3255969242754775,
|
| 348 |
+
"eval_flow_mse_loss": 0.23518445449216024,
|
| 349 |
+
"eval_loss": 0.37445583431197127,
|
| 350 |
+
"flow/cos_sim": 0.8770843943807362,
|
| 351 |
+
"flow/improvement_ratio": 0.9924178827546044,
|
| 352 |
+
"flow/mag_ratio_mean": 0.855633486181434,
|
| 353 |
+
"flow/mag_ratio_std": 0.14714058946127068,
|
| 354 |
+
"step": 10240
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.44131273299286744,
|
| 358 |
+
"eval_bleu": 0.5765275121149777,
|
| 359 |
+
"eval_ce_clean_loss": 0.0067116839581353865,
|
| 360 |
+
"eval_ce_pred_loss": 1.3255969242754775,
|
| 361 |
+
"eval_flow_mse_loss": 0.23518445449216024,
|
| 362 |
+
"eval_loss": 0.37445583431197127,
|
| 363 |
+
"eval_runtime": 208.5305,
|
| 364 |
+
"eval_samples_per_second": 143.864,
|
| 365 |
+
"eval_steps_per_second": 2.249,
|
| 366 |
+
"flow/cos_sim": 0.8770843943807362,
|
| 367 |
+
"flow/improvement_ratio": 0.9924178827546044,
|
| 368 |
+
"flow/mag_ratio_mean": 0.855633486181434,
|
| 369 |
+
"flow/mag_ratio_std": 0.14714058946127068,
|
| 370 |
+
"step": 10240
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.4854440062921542,
|
| 374 |
+
"grad_norm": 0.4819375276565552,
|
| 375 |
+
"learning_rate": 5.604035379537632e-05,
|
| 376 |
+
"loss": 0.3924267590045929,
|
| 377 |
+
"step": 11264
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.4854440062921542,
|
| 381 |
+
"eval_bleu": 0.5897455473229686,
|
| 382 |
+
"eval_ce_clean_loss": 0.00563325951405084,
|
| 383 |
+
"eval_ce_pred_loss": 1.2748802623260758,
|
| 384 |
+
"eval_flow_mse_loss": 0.23001743392395313,
|
| 385 |
+
"eval_loss": 0.3631387221406518,
|
| 386 |
+
"flow/cos_sim": 0.8802281503738371,
|
| 387 |
+
"flow/improvement_ratio": 0.9929161580132523,
|
| 388 |
+
"flow/mag_ratio_mean": 0.8601305431394435,
|
| 389 |
+
"flow/mag_ratio_std": 0.13978371800962033,
|
| 390 |
+
"step": 11264
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.4854440062921542,
|
| 394 |
+
"eval_bleu": 0.5897455473229686,
|
| 395 |
+
"eval_ce_clean_loss": 0.00563325951405084,
|
| 396 |
+
"eval_ce_pred_loss": 1.2748802623260758,
|
| 397 |
+
"eval_flow_mse_loss": 0.23001743392395313,
|
| 398 |
+
"eval_loss": 0.3631387221406518,
|
| 399 |
+
"eval_runtime": 210.9194,
|
| 400 |
+
"eval_samples_per_second": 142.234,
|
| 401 |
+
"eval_steps_per_second": 2.224,
|
| 402 |
+
"flow/cos_sim": 0.8802281503738371,
|
| 403 |
+
"flow/improvement_ratio": 0.9929161580132523,
|
| 404 |
+
"flow/mag_ratio_mean": 0.8601305431394435,
|
| 405 |
+
"flow/mag_ratio_std": 0.13978371800962033,
|
| 406 |
+
"step": 11264
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.5295752795914409,
|
| 410 |
+
"grad_norm": 0.7128424048423767,
|
| 411 |
+
"learning_rate": 4.880324964674035e-05,
|
| 412 |
+
"loss": 0.38400018215179443,
|
| 413 |
+
"step": 12288
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.5295752795914409,
|
| 417 |
+
"eval_bleu": 0.5985666163591713,
|
| 418 |
+
"eval_ce_clean_loss": 0.004892549006147251,
|
| 419 |
+
"eval_ce_pred_loss": 1.2226355028813327,
|
| 420 |
+
"eval_flow_mse_loss": 0.22944619641629363,
|
| 421 |
+
"eval_loss": 0.3566022975993817,
|
| 422 |
+
"flow/cos_sim": 0.8810523618767256,
|
| 423 |
+
"flow/improvement_ratio": 0.993417012030636,
|
| 424 |
+
"flow/mag_ratio_mean": 0.8616911000026061,
|
| 425 |
+
"flow/mag_ratio_std": 0.13965364706033329,
|
| 426 |
+
"step": 12288
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.5295752795914409,
|
| 430 |
+
"eval_bleu": 0.5985666163591713,
|
| 431 |
+
"eval_ce_clean_loss": 0.004892549006147251,
|
| 432 |
+
"eval_ce_pred_loss": 1.2226355028813327,
|
| 433 |
+
"eval_flow_mse_loss": 0.22944619641629363,
|
| 434 |
+
"eval_loss": 0.3566022975993817,
|
| 435 |
+
"eval_runtime": 206.3231,
|
| 436 |
+
"eval_samples_per_second": 145.403,
|
| 437 |
+
"eval_steps_per_second": 2.273,
|
| 438 |
+
"flow/cos_sim": 0.8810523618767256,
|
| 439 |
+
"flow/improvement_ratio": 0.993417012030636,
|
| 440 |
+
"flow/mag_ratio_mean": 0.8616911000026061,
|
| 441 |
+
"flow/mag_ratio_std": 0.13965364706033329,
|
| 442 |
+
"step": 12288
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.5737065528907277,
|
| 446 |
+
"grad_norm": 3.4912893772125244,
|
| 447 |
+
"learning_rate": 4.159127701277096e-05,
|
| 448 |
+
"loss": 0.3788807988166809,
|
| 449 |
+
"step": 13312
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.5737065528907277,
|
| 453 |
+
"eval_bleu": 0.5998961073680158,
|
| 454 |
+
"eval_ce_clean_loss": 0.00434517013621447,
|
| 455 |
+
"eval_ce_pred_loss": 1.250772520804456,
|
| 456 |
+
"eval_flow_mse_loss": 0.22432834110153255,
|
| 457 |
+
"eval_loss": 0.35375076519655013,
|
| 458 |
+
"flow/cos_sim": 0.8840986825764052,
|
| 459 |
+
"flow/improvement_ratio": 0.9934716673293856,
|
| 460 |
+
"flow/mag_ratio_mean": 0.8680964381710045,
|
| 461 |
+
"flow/mag_ratio_std": 0.13179212478178143,
|
| 462 |
+
"step": 13312
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.5737065528907277,
|
| 466 |
+
"eval_bleu": 0.5998961073680158,
|
| 467 |
+
"eval_ce_clean_loss": 0.00434517013621447,
|
| 468 |
+
"eval_ce_pred_loss": 1.250772520804456,
|
| 469 |
+
"eval_flow_mse_loss": 0.22432834110153255,
|
| 470 |
+
"eval_loss": 0.35375076519655013,
|
| 471 |
+
"eval_runtime": 208.4206,
|
| 472 |
+
"eval_samples_per_second": 143.94,
|
| 473 |
+
"eval_steps_per_second": 2.25,
|
| 474 |
+
"flow/cos_sim": 0.8840986825764052,
|
| 475 |
+
"flow/improvement_ratio": 0.9934716673293856,
|
| 476 |
+
"flow/mag_ratio_mean": 0.8680964381710045,
|
| 477 |
+
"flow/mag_ratio_std": 0.13179212478178143,
|
| 478 |
+
"step": 13312
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6178378261900145,
|
| 482 |
+
"grad_norm": 0.6786638498306274,
|
| 483 |
+
"learning_rate": 3.456935793454373e-05,
|
| 484 |
+
"loss": 0.3726285398006439,
|
| 485 |
+
"step": 14336
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6178378261900145,
|
| 489 |
+
"eval_bleu": 0.601626045312055,
|
| 490 |
+
"eval_ce_clean_loss": 0.0038832511889613243,
|
| 491 |
+
"eval_ce_pred_loss": 1.2371323627195379,
|
| 492 |
+
"eval_flow_mse_loss": 0.22229076604217862,
|
| 493 |
+
"eval_loss": 0.34988725547597355,
|
| 494 |
+
"flow/cos_sim": 0.8853790543989332,
|
| 495 |
+
"flow/improvement_ratio": 0.9933634653274439,
|
| 496 |
+
"flow/mag_ratio_mean": 0.8658980943246691,
|
| 497 |
+
"flow/mag_ratio_std": 0.13267156143368944,
|
| 498 |
+
"step": 14336
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.6178378261900145,
|
| 502 |
+
"eval_bleu": 0.601626045312055,
|
| 503 |
+
"eval_ce_clean_loss": 0.0038832511889613243,
|
| 504 |
+
"eval_ce_pred_loss": 1.2371323627195379,
|
| 505 |
+
"eval_flow_mse_loss": 0.22229076604217862,
|
| 506 |
+
"eval_loss": 0.34988725547597355,
|
| 507 |
+
"eval_runtime": 207.0167,
|
| 508 |
+
"eval_samples_per_second": 144.916,
|
| 509 |
+
"eval_steps_per_second": 2.266,
|
| 510 |
+
"flow/cos_sim": 0.8853790543989332,
|
| 511 |
+
"flow/improvement_ratio": 0.9933634653274439,
|
| 512 |
+
"flow/mag_ratio_mean": 0.8658980943246691,
|
| 513 |
+
"flow/mag_ratio_std": 0.13267156143368944,
|
| 514 |
+
"step": 14336
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6619690994893012,
|
| 518 |
+
"grad_norm": 0.642800509929657,
|
| 519 |
+
"learning_rate": 2.78638664786528e-05,
|
| 520 |
+
"loss": 0.3696843087673187,
|
| 521 |
+
"step": 15360
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6619690994893012,
|
| 525 |
+
"eval_bleu": 0.6059630768563422,
|
| 526 |
+
"eval_ce_clean_loss": 0.0036327009496868834,
|
| 527 |
+
"eval_ce_pred_loss": 1.2279239570154057,
|
| 528 |
+
"eval_flow_mse_loss": 0.22302042141652056,
|
| 529 |
+
"eval_loss": 0.3494455208783465,
|
| 530 |
+
"flow/cos_sim": 0.8849434552670542,
|
| 531 |
+
"flow/improvement_ratio": 0.9929700529397424,
|
| 532 |
+
"flow/mag_ratio_mean": 0.8687539150211603,
|
| 533 |
+
"flow/mag_ratio_std": 0.13310390089684204,
|
| 534 |
+
"step": 15360
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.6619690994893012,
|
| 538 |
+
"eval_bleu": 0.6059630768563422,
|
| 539 |
+
"eval_ce_clean_loss": 0.0036327009496868834,
|
| 540 |
+
"eval_ce_pred_loss": 1.2279239570154057,
|
| 541 |
+
"eval_flow_mse_loss": 0.22302042141652056,
|
| 542 |
+
"eval_loss": 0.3494455208783465,
|
| 543 |
+
"eval_runtime": 208.9975,
|
| 544 |
+
"eval_samples_per_second": 143.542,
|
| 545 |
+
"eval_steps_per_second": 2.244,
|
| 546 |
+
"flow/cos_sim": 0.8849434552670542,
|
| 547 |
+
"flow/improvement_ratio": 0.9929700529397424,
|
| 548 |
+
"flow/mag_ratio_mean": 0.8687539150211603,
|
| 549 |
+
"flow/mag_ratio_std": 0.13310390089684204,
|
| 550 |
+
"step": 15360
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.706100372788588,
|
| 554 |
+
"grad_norm": 0.6865013837814331,
|
| 555 |
+
"learning_rate": 2.1616492832807178e-05,
|
| 556 |
+
"loss": 0.36596235632896423,
|
| 557 |
+
"step": 16384
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.706100372788588,
|
| 561 |
+
"eval_bleu": 0.6108133508544432,
|
| 562 |
+
"eval_ce_clean_loss": 0.0034194857946712215,
|
| 563 |
+
"eval_ce_pred_loss": 1.2112553603867733,
|
| 564 |
+
"eval_flow_mse_loss": 0.2210535094745632,
|
| 565 |
+
"eval_loss": 0.34559853403553015,
|
| 566 |
+
"flow/cos_sim": 0.8862113247293908,
|
| 567 |
+
"flow/improvement_ratio": 0.9932185677068828,
|
| 568 |
+
"flow/mag_ratio_mean": 0.8704400952182599,
|
| 569 |
+
"flow/mag_ratio_std": 0.1318186233038587,
|
| 570 |
+
"step": 16384
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.706100372788588,
|
| 574 |
+
"eval_bleu": 0.6108133508544432,
|
| 575 |
+
"eval_ce_clean_loss": 0.0034194857946712215,
|
| 576 |
+
"eval_ce_pred_loss": 1.2112553603867733,
|
| 577 |
+
"eval_flow_mse_loss": 0.2210535094745632,
|
| 578 |
+
"eval_loss": 0.34559853403553015,
|
| 579 |
+
"eval_runtime": 209.915,
|
| 580 |
+
"eval_samples_per_second": 142.915,
|
| 581 |
+
"eval_steps_per_second": 2.234,
|
| 582 |
+
"flow/cos_sim": 0.8862113247293908,
|
| 583 |
+
"flow/improvement_ratio": 0.9932185677068828,
|
| 584 |
+
"flow/mag_ratio_mean": 0.8704400952182599,
|
| 585 |
+
"flow/mag_ratio_std": 0.1318186233038587,
|
| 586 |
+
"step": 16384
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.7502316460878746,
|
| 590 |
+
"grad_norm": 0.4638829529285431,
|
| 591 |
+
"learning_rate": 1.5965167080089493e-05,
|
| 592 |
+
"loss": 0.3644106388092041,
|
| 593 |
+
"step": 17408
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.7502316460878746,
|
| 597 |
+
"eval_bleu": 0.6105079675429955,
|
| 598 |
+
"eval_ce_clean_loss": 0.00328515888986561,
|
| 599 |
+
"eval_ce_pred_loss": 1.21342312215742,
|
| 600 |
+
"eval_flow_mse_loss": 0.21814484179401195,
|
| 601 |
+
"eval_loss": 0.3427723162591076,
|
| 602 |
+
"flow/cos_sim": 0.8882733249206787,
|
| 603 |
+
"flow/improvement_ratio": 0.9942307769616784,
|
| 604 |
+
"flow/mag_ratio_mean": 0.8709119436329108,
|
| 605 |
+
"flow/mag_ratio_std": 0.12955253779379797,
|
| 606 |
+
"step": 17408
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.7502316460878746,
|
| 610 |
+
"eval_bleu": 0.6105079675429955,
|
| 611 |
+
"eval_ce_clean_loss": 0.00328515888986561,
|
| 612 |
+
"eval_ce_pred_loss": 1.21342312215742,
|
| 613 |
+
"eval_flow_mse_loss": 0.21814484179401195,
|
| 614 |
+
"eval_loss": 0.3427723162591076,
|
| 615 |
+
"eval_runtime": 208.6305,
|
| 616 |
+
"eval_samples_per_second": 143.795,
|
| 617 |
+
"eval_steps_per_second": 2.248,
|
| 618 |
+
"flow/cos_sim": 0.8882733249206787,
|
| 619 |
+
"flow/improvement_ratio": 0.9942307769616784,
|
| 620 |
+
"flow/mag_ratio_mean": 0.8709119436329108,
|
| 621 |
+
"flow/mag_ratio_std": 0.12955253779379797,
|
| 622 |
+
"step": 17408
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 0.7943629193871614,
|
| 626 |
+
"grad_norm": 0.5375549793243408,
|
| 627 |
+
"learning_rate": 1.1028565914977467e-05,
|
| 628 |
+
"loss": 0.36267560720443726,
|
| 629 |
+
"step": 18432
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.7943629193871614,
|
| 633 |
+
"eval_bleu": 0.6095697083754615,
|
| 634 |
+
"eval_ce_clean_loss": 0.0031694269122015725,
|
| 635 |
+
"eval_ce_pred_loss": 1.21946408042013,
|
| 636 |
+
"eval_flow_mse_loss": 0.21630297963426057,
|
| 637 |
+
"eval_loss": 0.3414188172898567,
|
| 638 |
+
"flow/cos_sim": 0.8893322530331642,
|
| 639 |
+
"flow/improvement_ratio": 0.9944999824200612,
|
| 640 |
+
"flow/mag_ratio_mean": 0.8717981498124503,
|
| 641 |
+
"flow/mag_ratio_std": 0.12810804766378422,
|
| 642 |
+
"step": 18432
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.7943629193871614,
|
| 646 |
+
"eval_bleu": 0.6095697083754615,
|
| 647 |
+
"eval_ce_clean_loss": 0.0031694269122015725,
|
| 648 |
+
"eval_ce_pred_loss": 1.21946408042013,
|
| 649 |
+
"eval_flow_mse_loss": 0.21630297963426057,
|
| 650 |
+
"eval_loss": 0.3414188172898567,
|
| 651 |
+
"eval_runtime": 206.2022,
|
| 652 |
+
"eval_samples_per_second": 145.488,
|
| 653 |
+
"eval_steps_per_second": 2.274,
|
| 654 |
+
"flow/cos_sim": 0.8893322530331642,
|
| 655 |
+
"flow/improvement_ratio": 0.9944999824200612,
|
| 656 |
+
"flow/mag_ratio_mean": 0.8717981498124503,
|
| 657 |
+
"flow/mag_ratio_std": 0.12810804766378422,
|
| 658 |
+
"step": 18432
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 0.8384941926864482,
|
| 662 |
+
"grad_norm": 0.34996357560157776,
|
| 663 |
+
"learning_rate": 6.910356960585646e-06,
|
| 664 |
+
"loss": 0.3601815104484558,
|
| 665 |
+
"step": 19456
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.8384941926864482,
|
| 669 |
+
"eval_bleu": 0.613705610293561,
|
| 670 |
+
"eval_ce_clean_loss": 0.0031111957139220397,
|
| 671 |
+
"eval_ce_pred_loss": 1.1972390986454766,
|
| 672 |
+
"eval_flow_mse_loss": 0.21534570565482955,
|
| 673 |
+
"eval_loss": 0.33818081364448643,
|
| 674 |
+
"flow/cos_sim": 0.8900876406159228,
|
| 675 |
+
"flow/improvement_ratio": 0.995116201926396,
|
| 676 |
+
"flow/mag_ratio_mean": 0.8732720582978304,
|
| 677 |
+
"flow/mag_ratio_std": 0.1259334259577143,
|
| 678 |
+
"step": 19456
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.8384941926864482,
|
| 682 |
+
"eval_bleu": 0.613705610293561,
|
| 683 |
+
"eval_ce_clean_loss": 0.0031111957139220397,
|
| 684 |
+
"eval_ce_pred_loss": 1.1972390986454766,
|
| 685 |
+
"eval_flow_mse_loss": 0.21534570565482955,
|
| 686 |
+
"eval_loss": 0.33818081364448643,
|
| 687 |
+
"eval_runtime": 206.4468,
|
| 688 |
+
"eval_samples_per_second": 145.316,
|
| 689 |
+
"eval_steps_per_second": 2.272,
|
| 690 |
+
"flow/cos_sim": 0.8900876406159228,
|
| 691 |
+
"flow/improvement_ratio": 0.995116201926396,
|
| 692 |
+
"flow/mag_ratio_mean": 0.8732720582978304,
|
| 693 |
+
"flow/mag_ratio_std": 0.1259334259577143,
|
| 694 |
+
"step": 19456
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 0.8826254659857349,
|
| 698 |
+
"grad_norm": 0.5631929039955139,
|
| 699 |
+
"learning_rate": 3.6996948354814763e-06,
|
| 700 |
+
"loss": 0.3600355088710785,
|
| 701 |
+
"step": 20480
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.8826254659857349,
|
| 705 |
+
"eval_bleu": 0.6092457491598383,
|
| 706 |
+
"eval_ce_clean_loss": 0.003062377157653354,
|
| 707 |
+
"eval_ce_pred_loss": 1.2164991043015583,
|
| 708 |
+
"eval_flow_mse_loss": 0.2156423478365453,
|
| 709 |
+
"eval_loss": 0.3403546378683688,
|
| 710 |
+
"flow/cos_sim": 0.8900357947420718,
|
| 711 |
+
"flow/improvement_ratio": 0.9949319583774884,
|
| 712 |
+
"flow/mag_ratio_mean": 0.8731580517693623,
|
| 713 |
+
"flow/mag_ratio_std": 0.12572082460943315,
|
| 714 |
+
"step": 20480
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.8826254659857349,
|
| 718 |
+
"eval_bleu": 0.6092457491598383,
|
| 719 |
+
"eval_ce_clean_loss": 0.003062377157653354,
|
| 720 |
+
"eval_ce_pred_loss": 1.2164991043015583,
|
| 721 |
+
"eval_flow_mse_loss": 0.2156423478365453,
|
| 722 |
+
"eval_loss": 0.3403546378683688,
|
| 723 |
+
"eval_runtime": 207.4975,
|
| 724 |
+
"eval_samples_per_second": 144.58,
|
| 725 |
+
"eval_steps_per_second": 2.26,
|
| 726 |
+
"flow/cos_sim": 0.8900357947420718,
|
| 727 |
+
"flow/improvement_ratio": 0.9949319583774884,
|
| 728 |
+
"flow/mag_ratio_mean": 0.8731580517693623,
|
| 729 |
+
"flow/mag_ratio_std": 0.12572082460943315,
|
| 730 |
+
"step": 20480
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.9267567392850217,
|
| 734 |
+
"grad_norm": 0.35356131196022034,
|
| 735 |
+
"learning_rate": 1.4577368571635218e-06,
|
| 736 |
+
"loss": 0.3606625497341156,
|
| 737 |
+
"step": 21504
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.9267567392850217,
|
| 741 |
+
"eval_bleu": 0.6115870524577083,
|
| 742 |
+
"eval_ce_clean_loss": 0.003047047608739524,
|
| 743 |
+
"eval_ce_pred_loss": 1.207809279658901,
|
| 744 |
+
"eval_flow_mse_loss": 0.2171662690669997,
|
| 745 |
+
"eval_loss": 0.3409942464152379,
|
| 746 |
+
"flow/cos_sim": 0.8890357672024384,
|
| 747 |
+
"flow/improvement_ratio": 0.9942183488213432,
|
| 748 |
+
"flow/mag_ratio_mean": 0.8708964179573792,
|
| 749 |
+
"flow/mag_ratio_std": 0.1279422494648362,
|
| 750 |
+
"step": 21504
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 0.9267567392850217,
|
| 754 |
+
"eval_bleu": 0.6115870524577083,
|
| 755 |
+
"eval_ce_clean_loss": 0.003047047608739524,
|
| 756 |
+
"eval_ce_pred_loss": 1.207809279658901,
|
| 757 |
+
"eval_flow_mse_loss": 0.2171662690669997,
|
| 758 |
+
"eval_loss": 0.3409942464152379,
|
| 759 |
+
"eval_runtime": 207.2504,
|
| 760 |
+
"eval_samples_per_second": 144.752,
|
| 761 |
+
"eval_steps_per_second": 2.263,
|
| 762 |
+
"flow/cos_sim": 0.8890357672024384,
|
| 763 |
+
"flow/improvement_ratio": 0.9942183488213432,
|
| 764 |
+
"flow/mag_ratio_mean": 0.8708964179573792,
|
| 765 |
+
"flow/mag_ratio_std": 0.1279422494648362,
|
| 766 |
+
"step": 21504
|
| 767 |
+
}
|
| 768 |
+
],
|
| 769 |
+
"logging_steps": 1024,
|
| 770 |
+
"max_steps": 23204,
|
| 771 |
+
"num_input_tokens_seen": 0,
|
| 772 |
+
"num_train_epochs": 1,
|
| 773 |
+
"save_steps": 1024,
|
| 774 |
+
"stateful_callbacks": {
|
| 775 |
+
"TrainerControl": {
|
| 776 |
+
"args": {
|
| 777 |
+
"should_epoch_stop": false,
|
| 778 |
+
"should_evaluate": false,
|
| 779 |
+
"should_log": false,
|
| 780 |
+
"should_save": true,
|
| 781 |
+
"should_training_stop": false
|
| 782 |
+
},
|
| 783 |
+
"attributes": {}
|
| 784 |
+
}
|
| 785 |
+
},
|
| 786 |
+
"total_flos": 0.0,
|
| 787 |
+
"train_batch_size": 64,
|
| 788 |
+
"trial_name": null,
|
| 789 |
+
"trial_params": null
|
| 790 |
+
}
|
checkpoints-v5.14/checkpoint-21504/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8469bbc81a2ba0be2f5b44007faafd15c75615abe30f4f4e56171816d31caa5b
|
| 3 |
+
size 5137
|