Upload folder using huggingface_hub
Browse files- checkpoints-v3-clean/checkpoint-12288/eval_state.json +0 -0
- checkpoints-v3-clean/checkpoint-12288/model.safetensors +3 -0
- checkpoints-v3-clean/checkpoint-12288/optimizer.pt +3 -0
- checkpoints-v3-clean/checkpoint-12288/rng_state.pth +3 -0
- checkpoints-v3-clean/checkpoint-12288/scaler.pt +3 -0
- checkpoints-v3-clean/checkpoint-12288/scheduler.pt +3 -0
- checkpoints-v3-clean/checkpoint-12288/trainer_state.json +958 -0
- checkpoints-v3-clean/checkpoint-12288/training_args.bin +3 -0
checkpoints-v3-clean/checkpoint-12288/eval_state.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoints-v3-clean/checkpoint-12288/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c4b584386b0bd04528c3fc3e91b16a5c4a87b30a81701770a4f202d345d3a250
|
| 3 |
+
size 37402680
|
checkpoints-v3-clean/checkpoint-12288/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b63e8f01490e3eb48a0cd5949ce60eeb8ae80703d620e20e4b4dedb7bf4f1c13
|
| 3 |
+
size 512267
|
checkpoints-v3-clean/checkpoint-12288/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f36e6c28b77555b6da6de84681647b558ac8ebc553a1b458e45112e416a213c
|
| 3 |
+
size 14645
|
checkpoints-v3-clean/checkpoint-12288/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d02347266bda28d7d0c3f532e0872dd78fc1e3d8144e77461b904455eb33a758
|
| 3 |
+
size 1383
|
checkpoints-v3-clean/checkpoint-12288/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fc29e591a5629b3478acd7a9ce88545c1f2b469012e3700ebb5314fc3e2dad70
|
| 3 |
+
size 1465
|
checkpoints-v3-clean/checkpoint-12288/trainer_state.json
ADDED
|
@@ -0,0 +1,958 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.127721938695964,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 12288,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.002660873722832583,
|
| 14 |
+
"grad_norm": 1.2785860300064087,
|
| 15 |
+
"learning_rate": 4.416349151368202e-06,
|
| 16 |
+
"loss": 10.8939,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.005321747445665166,
|
| 21 |
+
"grad_norm": 0.9874694347381592,
|
| 22 |
+
"learning_rate": 8.85001731901628e-06,
|
| 23 |
+
"loss": 10.1502,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.00798262116849775,
|
| 28 |
+
"grad_norm": 0.9198242425918579,
|
| 29 |
+
"learning_rate": 1.3283685486664357e-05,
|
| 30 |
+
"loss": 9.0734,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.010643494891330332,
|
| 35 |
+
"grad_norm": 0.868468701839447,
|
| 36 |
+
"learning_rate": 1.7717353654312436e-05,
|
| 37 |
+
"loss": 8.0763,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.010643494891330332,
|
| 42 |
+
"eval_acr_loss": 0.010333856916986406,
|
| 43 |
+
"eval_across_var": 0.013189065648475662,
|
| 44 |
+
"eval_bleu": 0.09159994619921694,
|
| 45 |
+
"eval_ce_loss": 7.027534931898117,
|
| 46 |
+
"eval_cos_loss": 0.9640954360365868,
|
| 47 |
+
"eval_cov": 0.06866264343261719,
|
| 48 |
+
"eval_cov_loss": 0.0075032426248071715,
|
| 49 |
+
"eval_glb_loss": 0.39646806567907333,
|
| 50 |
+
"eval_global_kurtosis": 3.131724640727043,
|
| 51 |
+
"eval_global_mean": 0.0048119425773620605,
|
| 52 |
+
"eval_global_var": 0.27025604248046875,
|
| 53 |
+
"eval_krt_loss": 0.017541410372359678,
|
| 54 |
+
"eval_loss": 7.459874868392944,
|
| 55 |
+
"eval_mean_loss": 2.4568188223383913e-05,
|
| 56 |
+
"eval_mse_loss": 1.9595371261239052,
|
| 57 |
+
"eval_per_loss": 0.34346967935562134,
|
| 58 |
+
"eval_per_var": 0.26409912109375,
|
| 59 |
+
"eval_within_var": 0.25767749128863215,
|
| 60 |
+
"eval_wth_loss": 0.41259176656603813,
|
| 61 |
+
"step": 1024
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.010643494891330332,
|
| 65 |
+
"eval_acr_loss": 0.010333856916986406,
|
| 66 |
+
"eval_across_var": 0.013189065648475662,
|
| 67 |
+
"eval_bleu": 0.09159994619921694,
|
| 68 |
+
"eval_ce_loss": 7.027534931898117,
|
| 69 |
+
"eval_cos_loss": 0.9640954360365868,
|
| 70 |
+
"eval_cov": 0.06866264343261719,
|
| 71 |
+
"eval_cov_loss": 0.0075032426248071715,
|
| 72 |
+
"eval_glb_loss": 0.39646806567907333,
|
| 73 |
+
"eval_global_kurtosis": 3.131724640727043,
|
| 74 |
+
"eval_global_mean": 0.0048119425773620605,
|
| 75 |
+
"eval_global_var": 0.27025604248046875,
|
| 76 |
+
"eval_krt_loss": 0.017541410372359678,
|
| 77 |
+
"eval_loss": 7.459874868392944,
|
| 78 |
+
"eval_mean_loss": 2.4568188223383913e-05,
|
| 79 |
+
"eval_mse_loss": 1.9595371261239052,
|
| 80 |
+
"eval_per_loss": 0.34346967935562134,
|
| 81 |
+
"eval_per_var": 0.26409912109375,
|
| 82 |
+
"eval_runtime": 10.9247,
|
| 83 |
+
"eval_samples_per_second": 183.071,
|
| 84 |
+
"eval_steps_per_second": 2.929,
|
| 85 |
+
"eval_within_var": 0.25767749128863215,
|
| 86 |
+
"eval_wth_loss": 0.41259176656603813,
|
| 87 |
+
"step": 1024
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.013304368614162915,
|
| 91 |
+
"grad_norm": 0.8875442147254944,
|
| 92 |
+
"learning_rate": 2.2151021821960514e-05,
|
| 93 |
+
"loss": 7.1132,
|
| 94 |
+
"step": 1280
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.0159652423369955,
|
| 98 |
+
"grad_norm": 0.8143702149391174,
|
| 99 |
+
"learning_rate": 2.6584689989608592e-05,
|
| 100 |
+
"loss": 6.1485,
|
| 101 |
+
"step": 1536
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.01862611605982808,
|
| 105 |
+
"grad_norm": 0.7273866534233093,
|
| 106 |
+
"learning_rate": 3.1018358157256674e-05,
|
| 107 |
+
"loss": 5.2325,
|
| 108 |
+
"step": 1792
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.021286989782660665,
|
| 112 |
+
"grad_norm": 0.5480166077613831,
|
| 113 |
+
"learning_rate": 3.5452026324904745e-05,
|
| 114 |
+
"loss": 4.3952,
|
| 115 |
+
"step": 2048
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.021286989782660665,
|
| 119 |
+
"eval_acr_loss": 0.011928890817216597,
|
| 120 |
+
"eval_across_var": 0.012925449758768082,
|
| 121 |
+
"eval_bleu": 0.3377202654954096,
|
| 122 |
+
"eval_ce_loss": 3.4600580409169197,
|
| 123 |
+
"eval_cos_loss": 0.9338241945952177,
|
| 124 |
+
"eval_cov": 0.07202339172363281,
|
| 125 |
+
"eval_cov_loss": 0.008123685591272078,
|
| 126 |
+
"eval_glb_loss": 0.3427571505308151,
|
| 127 |
+
"eval_global_kurtosis": 3.049249067902565,
|
| 128 |
+
"eval_global_mean": 0.0031435638666152954,
|
| 129 |
+
"eval_global_var": 0.314483642578125,
|
| 130 |
+
"eval_krt_loss": 0.002645308370119892,
|
| 131 |
+
"eval_loss": 3.876119486987591,
|
| 132 |
+
"eval_mean_loss": 1.1241194734168403e-05,
|
| 133 |
+
"eval_mse_loss": 1.949895329773426,
|
| 134 |
+
"eval_per_loss": 0.2958330735564232,
|
| 135 |
+
"eval_per_var": 0.306182861328125,
|
| 136 |
+
"eval_within_var": 0.3034430192783475,
|
| 137 |
+
"eval_wth_loss": 0.35589773673564196,
|
| 138 |
+
"step": 2048
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.021286989782660665,
|
| 142 |
+
"eval_acr_loss": 0.011928890817216597,
|
| 143 |
+
"eval_across_var": 0.012925449758768082,
|
| 144 |
+
"eval_bleu": 0.3377202654954096,
|
| 145 |
+
"eval_ce_loss": 3.4600580409169197,
|
| 146 |
+
"eval_cos_loss": 0.9338241945952177,
|
| 147 |
+
"eval_cov": 0.07202339172363281,
|
| 148 |
+
"eval_cov_loss": 0.008123685591272078,
|
| 149 |
+
"eval_glb_loss": 0.3427571505308151,
|
| 150 |
+
"eval_global_kurtosis": 3.049249067902565,
|
| 151 |
+
"eval_global_mean": 0.0031435638666152954,
|
| 152 |
+
"eval_global_var": 0.314483642578125,
|
| 153 |
+
"eval_krt_loss": 0.002645308370119892,
|
| 154 |
+
"eval_loss": 3.876119486987591,
|
| 155 |
+
"eval_mean_loss": 1.1241194734168403e-05,
|
| 156 |
+
"eval_mse_loss": 1.949895329773426,
|
| 157 |
+
"eval_per_loss": 0.2958330735564232,
|
| 158 |
+
"eval_per_var": 0.306182861328125,
|
| 159 |
+
"eval_runtime": 9.7829,
|
| 160 |
+
"eval_samples_per_second": 204.439,
|
| 161 |
+
"eval_steps_per_second": 3.271,
|
| 162 |
+
"eval_within_var": 0.3034430192783475,
|
| 163 |
+
"eval_wth_loss": 0.35589773673564196,
|
| 164 |
+
"step": 2048
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.023947863505493248,
|
| 168 |
+
"grad_norm": 0.4677615165710449,
|
| 169 |
+
"learning_rate": 3.988569449255283e-05,
|
| 170 |
+
"loss": 3.673,
|
| 171 |
+
"step": 2304
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.02660873722832583,
|
| 175 |
+
"grad_norm": 0.3998287618160248,
|
| 176 |
+
"learning_rate": 4.43193626602009e-05,
|
| 177 |
+
"loss": 3.091,
|
| 178 |
+
"step": 2560
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.029269610951158414,
|
| 182 |
+
"grad_norm": 0.34429243206977844,
|
| 183 |
+
"learning_rate": 4.875303082784898e-05,
|
| 184 |
+
"loss": 2.6144,
|
| 185 |
+
"step": 2816
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.031930484673991,
|
| 189 |
+
"grad_norm": 0.2926134467124939,
|
| 190 |
+
"learning_rate": 4.999952040357126e-05,
|
| 191 |
+
"loss": 2.2102,
|
| 192 |
+
"step": 3072
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.031930484673991,
|
| 196 |
+
"eval_acr_loss": 0.011340929486323148,
|
| 197 |
+
"eval_across_var": 0.02669821889139712,
|
| 198 |
+
"eval_bleu": 0.5843127273325036,
|
| 199 |
+
"eval_ce_loss": 1.5527473986148834,
|
| 200 |
+
"eval_cos_loss": 0.8454732391983271,
|
| 201 |
+
"eval_cov": 0.08180618286132812,
|
| 202 |
+
"eval_cov_loss": 0.010497079812921584,
|
| 203 |
+
"eval_glb_loss": 0.08526049554347992,
|
| 204 |
+
"eval_global_kurtosis": 3.0069182589650154,
|
| 205 |
+
"eval_global_mean": 0.0094224214553833,
|
| 206 |
+
"eval_global_var": 0.607940673828125,
|
| 207 |
+
"eval_krt_loss": 0.00011411440974029574,
|
| 208 |
+
"eval_loss": 1.9118921793997288,
|
| 209 |
+
"eval_mean_loss": 9.215879936164129e-05,
|
| 210 |
+
"eval_mse_loss": 1.832219734787941,
|
| 211 |
+
"eval_per_loss": 0.06583906384184957,
|
| 212 |
+
"eval_per_var": 0.593536376953125,
|
| 213 |
+
"eval_within_var": 0.5868291966617107,
|
| 214 |
+
"eval_wth_loss": 0.09813770232722163,
|
| 215 |
+
"step": 3072
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.031930484673991,
|
| 219 |
+
"eval_acr_loss": 0.011340929486323148,
|
| 220 |
+
"eval_across_var": 0.02669821889139712,
|
| 221 |
+
"eval_bleu": 0.5843127273325036,
|
| 222 |
+
"eval_ce_loss": 1.5527473986148834,
|
| 223 |
+
"eval_cos_loss": 0.8454732391983271,
|
| 224 |
+
"eval_cov": 0.08180618286132812,
|
| 225 |
+
"eval_cov_loss": 0.010497079812921584,
|
| 226 |
+
"eval_glb_loss": 0.08526049554347992,
|
| 227 |
+
"eval_global_kurtosis": 3.0069182589650154,
|
| 228 |
+
"eval_global_mean": 0.0094224214553833,
|
| 229 |
+
"eval_global_var": 0.607940673828125,
|
| 230 |
+
"eval_krt_loss": 0.00011411440974029574,
|
| 231 |
+
"eval_loss": 1.9118921793997288,
|
| 232 |
+
"eval_mean_loss": 9.215879936164129e-05,
|
| 233 |
+
"eval_mse_loss": 1.832219734787941,
|
| 234 |
+
"eval_per_loss": 0.06583906384184957,
|
| 235 |
+
"eval_per_var": 0.593536376953125,
|
| 236 |
+
"eval_runtime": 10.2115,
|
| 237 |
+
"eval_samples_per_second": 195.858,
|
| 238 |
+
"eval_steps_per_second": 3.134,
|
| 239 |
+
"eval_within_var": 0.5868291966617107,
|
| 240 |
+
"eval_wth_loss": 0.09813770232722163,
|
| 241 |
+
"step": 3072
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.034591358396823584,
|
| 245 |
+
"grad_norm": 0.26758256554603577,
|
| 246 |
+
"learning_rate": 4.9997257547617106e-05,
|
| 247 |
+
"loss": 1.8977,
|
| 248 |
+
"step": 3328
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.03725223211965616,
|
| 252 |
+
"grad_norm": 0.24744224548339844,
|
| 253 |
+
"learning_rate": 4.999313816462926e-05,
|
| 254 |
+
"loss": 1.6476,
|
| 255 |
+
"step": 3584
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.03991310584248875,
|
| 259 |
+
"grad_norm": 0.22395408153533936,
|
| 260 |
+
"learning_rate": 4.9987162560551115e-05,
|
| 261 |
+
"loss": 1.4364,
|
| 262 |
+
"step": 3840
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.04257397956532133,
|
| 266 |
+
"grad_norm": 0.20503944158554077,
|
| 267 |
+
"learning_rate": 4.997933117918617e-05,
|
| 268 |
+
"loss": 1.281,
|
| 269 |
+
"step": 4096
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.04257397956532133,
|
| 273 |
+
"eval_acr_loss": 0.010712883464293554,
|
| 274 |
+
"eval_across_var": 0.04058358340989798,
|
| 275 |
+
"eval_bleu": 0.74504639855184,
|
| 276 |
+
"eval_ce_loss": 0.8122093826532364,
|
| 277 |
+
"eval_cos_loss": 0.7478988897055387,
|
| 278 |
+
"eval_cov": 0.07811164855957031,
|
| 279 |
+
"eval_cov_loss": 0.009650954219978303,
|
| 280 |
+
"eval_glb_loss": 0.0014296220615506172,
|
| 281 |
+
"eval_global_kurtosis": 3.0172539353370667,
|
| 282 |
+
"eval_global_mean": 0.011150479316711426,
|
| 283 |
+
"eval_global_var": 0.86297607421875,
|
| 284 |
+
"eval_krt_loss": 0.0004142036146248529,
|
| 285 |
+
"eval_loss": 1.1253399811685085,
|
| 286 |
+
"eval_mean_loss": 0.00012954238400197937,
|
| 287 |
+
"eval_mse_loss": 1.6814398430287838,
|
| 288 |
+
"eval_per_loss": 9.741889432035578e-05,
|
| 289 |
+
"eval_per_var": 0.843017578125,
|
| 290 |
+
"eval_within_var": 0.8369333390146494,
|
| 291 |
+
"eval_wth_loss": 0.004123660655750427,
|
| 292 |
+
"step": 4096
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.04257397956532133,
|
| 296 |
+
"eval_acr_loss": 0.010712883464293554,
|
| 297 |
+
"eval_across_var": 0.04058358340989798,
|
| 298 |
+
"eval_bleu": 0.74504639855184,
|
| 299 |
+
"eval_ce_loss": 0.8122093826532364,
|
| 300 |
+
"eval_cos_loss": 0.7478988897055387,
|
| 301 |
+
"eval_cov": 0.07811164855957031,
|
| 302 |
+
"eval_cov_loss": 0.009650954219978303,
|
| 303 |
+
"eval_glb_loss": 0.0014296220615506172,
|
| 304 |
+
"eval_global_kurtosis": 3.0172539353370667,
|
| 305 |
+
"eval_global_mean": 0.011150479316711426,
|
| 306 |
+
"eval_global_var": 0.86297607421875,
|
| 307 |
+
"eval_krt_loss": 0.0004142036146248529,
|
| 308 |
+
"eval_loss": 1.1253399811685085,
|
| 309 |
+
"eval_mean_loss": 0.00012954238400197937,
|
| 310 |
+
"eval_mse_loss": 1.6814398430287838,
|
| 311 |
+
"eval_per_loss": 9.741889432035578e-05,
|
| 312 |
+
"eval_per_var": 0.843017578125,
|
| 313 |
+
"eval_runtime": 10.2891,
|
| 314 |
+
"eval_samples_per_second": 194.381,
|
| 315 |
+
"eval_steps_per_second": 3.11,
|
| 316 |
+
"eval_within_var": 0.8369333390146494,
|
| 317 |
+
"eval_wth_loss": 0.004123660655750427,
|
| 318 |
+
"step": 4096
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.045234853288153916,
|
| 322 |
+
"grad_norm": 0.19751055538654327,
|
| 323 |
+
"learning_rate": 4.99696446021651e-05,
|
| 324 |
+
"loss": 1.1395,
|
| 325 |
+
"step": 4352
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.047895727010986495,
|
| 329 |
+
"grad_norm": 0.18066278100013733,
|
| 330 |
+
"learning_rate": 4.995810354890248e-05,
|
| 331 |
+
"loss": 1.023,
|
| 332 |
+
"step": 4608
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.05055660073381908,
|
| 336 |
+
"grad_norm": 0.17854027450084686,
|
| 337 |
+
"learning_rate": 4.994470887654346e-05,
|
| 338 |
+
"loss": 0.9304,
|
| 339 |
+
"step": 4864
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.05321747445665166,
|
| 343 |
+
"grad_norm": 0.1603478193283081,
|
| 344 |
+
"learning_rate": 4.9929461579899996e-05,
|
| 345 |
+
"loss": 0.8495,
|
| 346 |
+
"step": 5120
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.05321747445665166,
|
| 350 |
+
"eval_acr_loss": 0.010327979252906516,
|
| 351 |
+
"eval_across_var": 0.04717686545336619,
|
| 352 |
+
"eval_bleu": 0.8372942377615584,
|
| 353 |
+
"eval_ce_loss": 0.4861879423260689,
|
| 354 |
+
"eval_cos_loss": 0.6623753104358912,
|
| 355 |
+
"eval_cov": 0.07553482055664062,
|
| 356 |
+
"eval_cov_loss": 0.009084547113161534,
|
| 357 |
+
"eval_glb_loss": 0.0,
|
| 358 |
+
"eval_global_kurtosis": 3.0330343022942543,
|
| 359 |
+
"eval_global_mean": 0.008684873580932617,
|
| 360 |
+
"eval_global_var": 0.9628448486328125,
|
| 361 |
+
"eval_krt_loss": 0.0012540334691948374,
|
| 362 |
+
"eval_loss": 0.765046201646328,
|
| 363 |
+
"eval_mean_loss": 8.126164811983472e-05,
|
| 364 |
+
"eval_mse_loss": 1.542122010141611,
|
| 365 |
+
"eval_per_loss": 0.0,
|
| 366 |
+
"eval_per_var": 0.9408111572265625,
|
| 367 |
+
"eval_within_var": 0.9378277324140072,
|
| 368 |
+
"eval_wth_loss": 2.4662538677944212e-08,
|
| 369 |
+
"step": 5120
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.05321747445665166,
|
| 373 |
+
"eval_acr_loss": 0.010327979252906516,
|
| 374 |
+
"eval_across_var": 0.04717686545336619,
|
| 375 |
+
"eval_bleu": 0.8372942377615584,
|
| 376 |
+
"eval_ce_loss": 0.4861879423260689,
|
| 377 |
+
"eval_cos_loss": 0.6623753104358912,
|
| 378 |
+
"eval_cov": 0.07553482055664062,
|
| 379 |
+
"eval_cov_loss": 0.009084547113161534,
|
| 380 |
+
"eval_glb_loss": 0.0,
|
| 381 |
+
"eval_global_kurtosis": 3.0330343022942543,
|
| 382 |
+
"eval_global_mean": 0.008684873580932617,
|
| 383 |
+
"eval_global_var": 0.9628448486328125,
|
| 384 |
+
"eval_krt_loss": 0.0012540334691948374,
|
| 385 |
+
"eval_loss": 0.765046201646328,
|
| 386 |
+
"eval_mean_loss": 8.126164811983472e-05,
|
| 387 |
+
"eval_mse_loss": 1.542122010141611,
|
| 388 |
+
"eval_per_loss": 0.0,
|
| 389 |
+
"eval_per_var": 0.9408111572265625,
|
| 390 |
+
"eval_runtime": 9.9429,
|
| 391 |
+
"eval_samples_per_second": 201.148,
|
| 392 |
+
"eval_steps_per_second": 3.218,
|
| 393 |
+
"eval_within_var": 0.9378277324140072,
|
| 394 |
+
"eval_wth_loss": 2.4662538677944212e-08,
|
| 395 |
+
"step": 5120
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.05587834817948425,
|
| 399 |
+
"grad_norm": 0.14991497993469238,
|
| 400 |
+
"learning_rate": 4.9912362791377065e-05,
|
| 401 |
+
"loss": 0.7781,
|
| 402 |
+
"step": 5376
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.05853922190231683,
|
| 406 |
+
"grad_norm": 0.14166131615638733,
|
| 407 |
+
"learning_rate": 4.989341378088849e-05,
|
| 408 |
+
"loss": 0.7173,
|
| 409 |
+
"step": 5632
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.061200095625149414,
|
| 413 |
+
"grad_norm": 0.15388458967208862,
|
| 414 |
+
"learning_rate": 4.987261595576267e-05,
|
| 415 |
+
"loss": 0.6644,
|
| 416 |
+
"step": 5888
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.063860969347982,
|
| 420 |
+
"grad_norm": 0.1334715038537979,
|
| 421 |
+
"learning_rate": 4.984997086063803e-05,
|
| 422 |
+
"loss": 0.6207,
|
| 423 |
+
"step": 6144
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.063860969347982,
|
| 427 |
+
"eval_acr_loss": 0.0101167297543725,
|
| 428 |
+
"eval_across_var": 0.054399247106630355,
|
| 429 |
+
"eval_bleu": 0.8895029308800785,
|
| 430 |
+
"eval_ce_loss": 0.3160733813419938,
|
| 431 |
+
"eval_cos_loss": 0.5900800842791796,
|
| 432 |
+
"eval_cov": 0.07538223266601562,
|
| 433 |
+
"eval_cov_loss": 0.009119293099502102,
|
| 434 |
+
"eval_glb_loss": 6.775409133297217e-06,
|
| 435 |
+
"eval_global_kurtosis": 3.032428242266178,
|
| 436 |
+
"eval_global_mean": 0.009008824825286865,
|
| 437 |
+
"eval_global_var": 1.085845947265625,
|
| 438 |
+
"eval_krt_loss": 0.0013425453408899557,
|
| 439 |
+
"eval_loss": 0.5660126637667418,
|
| 440 |
+
"eval_mean_loss": 8.861126445935952e-05,
|
| 441 |
+
"eval_mse_loss": 1.4217685647308826,
|
| 442 |
+
"eval_per_loss": 0.0,
|
| 443 |
+
"eval_per_var": 1.0615234375,
|
| 444 |
+
"eval_within_var": 1.0598372742533684,
|
| 445 |
+
"eval_wth_loss": 0.0,
|
| 446 |
+
"step": 6144
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.063860969347982,
|
| 450 |
+
"eval_acr_loss": 0.0101167297543725,
|
| 451 |
+
"eval_across_var": 0.054399247106630355,
|
| 452 |
+
"eval_bleu": 0.8895029308800785,
|
| 453 |
+
"eval_ce_loss": 0.3160733813419938,
|
| 454 |
+
"eval_cos_loss": 0.5900800842791796,
|
| 455 |
+
"eval_cov": 0.07538223266601562,
|
| 456 |
+
"eval_cov_loss": 0.009119293099502102,
|
| 457 |
+
"eval_glb_loss": 6.775409133297217e-06,
|
| 458 |
+
"eval_global_kurtosis": 3.032428242266178,
|
| 459 |
+
"eval_global_mean": 0.009008824825286865,
|
| 460 |
+
"eval_global_var": 1.085845947265625,
|
| 461 |
+
"eval_krt_loss": 0.0013425453408899557,
|
| 462 |
+
"eval_loss": 0.5660126637667418,
|
| 463 |
+
"eval_mean_loss": 8.861126445935952e-05,
|
| 464 |
+
"eval_mse_loss": 1.4217685647308826,
|
| 465 |
+
"eval_per_loss": 0.0,
|
| 466 |
+
"eval_per_var": 1.0615234375,
|
| 467 |
+
"eval_runtime": 10.0342,
|
| 468 |
+
"eval_samples_per_second": 199.319,
|
| 469 |
+
"eval_steps_per_second": 3.189,
|
| 470 |
+
"eval_within_var": 1.0598372742533684,
|
| 471 |
+
"eval_wth_loss": 0.0,
|
| 472 |
+
"step": 6144
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.06652184307081459,
|
| 476 |
+
"grad_norm": 0.13046643137931824,
|
| 477 |
+
"learning_rate": 4.982548017734831e-05,
|
| 478 |
+
"loss": 0.5778,
|
| 479 |
+
"step": 6400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.06918271679364717,
|
| 483 |
+
"grad_norm": 0.1357714831829071,
|
| 484 |
+
"learning_rate": 4.9799145724797655e-05,
|
| 485 |
+
"loss": 0.5412,
|
| 486 |
+
"step": 6656
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.07184359051647975,
|
| 490 |
+
"grad_norm": 0.12301656603813171,
|
| 491 |
+
"learning_rate": 4.977096945882557e-05,
|
| 492 |
+
"loss": 0.5079,
|
| 493 |
+
"step": 6912
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.07450446423931233,
|
| 497 |
+
"grad_norm": 0.12161236256361008,
|
| 498 |
+
"learning_rate": 4.974095347206157e-05,
|
| 499 |
+
"loss": 0.4804,
|
| 500 |
+
"step": 7168
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.07450446423931233,
|
| 504 |
+
"eval_acr_loss": 0.00983446673490107,
|
| 505 |
+
"eval_across_var": 0.05932739016134292,
|
| 506 |
+
"eval_bleu": 0.9201149551889397,
|
| 507 |
+
"eval_ce_loss": 0.2183996201492846,
|
| 508 |
+
"eval_cos_loss": 0.5297921635210514,
|
| 509 |
+
"eval_cov": 0.074615478515625,
|
| 510 |
+
"eval_cov_loss": 0.008976185490610078,
|
| 511 |
+
"eval_glb_loss": 0.0027152854308951646,
|
| 512 |
+
"eval_global_kurtosis": 3.0397911220788956,
|
| 513 |
+
"eval_global_mean": 0.00715222954750061,
|
| 514 |
+
"eval_global_var": 1.149993896484375,
|
| 515 |
+
"eval_krt_loss": 0.0019899249493846582,
|
| 516 |
+
"eval_loss": 0.44435011968016624,
|
| 517 |
+
"eval_mean_loss": 5.936167204367848e-05,
|
| 518 |
+
"eval_mse_loss": 1.3212404884397984,
|
| 519 |
+
"eval_per_loss": 1.6360521115643678e-06,
|
| 520 |
+
"eval_per_var": 1.12481689453125,
|
| 521 |
+
"eval_within_var": 1.1261104978621006,
|
| 522 |
+
"eval_wth_loss": 0.0010669257317950098,
|
| 523 |
+
"step": 7168
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.07450446423931233,
|
| 527 |
+
"eval_acr_loss": 0.00983446673490107,
|
| 528 |
+
"eval_across_var": 0.05932739016134292,
|
| 529 |
+
"eval_bleu": 0.9201149551889397,
|
| 530 |
+
"eval_ce_loss": 0.2183996201492846,
|
| 531 |
+
"eval_cos_loss": 0.5297921635210514,
|
| 532 |
+
"eval_cov": 0.074615478515625,
|
| 533 |
+
"eval_cov_loss": 0.008976185490610078,
|
| 534 |
+
"eval_glb_loss": 0.0027152854308951646,
|
| 535 |
+
"eval_global_kurtosis": 3.0397911220788956,
|
| 536 |
+
"eval_global_mean": 0.00715222954750061,
|
| 537 |
+
"eval_global_var": 1.149993896484375,
|
| 538 |
+
"eval_krt_loss": 0.0019899249493846582,
|
| 539 |
+
"eval_loss": 0.44435011968016624,
|
| 540 |
+
"eval_mean_loss": 5.936167204367848e-05,
|
| 541 |
+
"eval_mse_loss": 1.3212404884397984,
|
| 542 |
+
"eval_per_loss": 1.6360521115643678e-06,
|
| 543 |
+
"eval_per_var": 1.12481689453125,
|
| 544 |
+
"eval_runtime": 9.7644,
|
| 545 |
+
"eval_samples_per_second": 204.825,
|
| 546 |
+
"eval_steps_per_second": 3.277,
|
| 547 |
+
"eval_within_var": 1.1261104978621006,
|
| 548 |
+
"eval_wth_loss": 0.0010669257317950098,
|
| 549 |
+
"step": 7168
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.07716533796214492,
|
| 553 |
+
"grad_norm": 0.11638697236776352,
|
| 554 |
+
"learning_rate": 4.970909999376984e-05,
|
| 555 |
+
"loss": 0.4535,
|
| 556 |
+
"step": 7424
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.0798262116849775,
|
| 560 |
+
"grad_norm": 0.10645346343517303,
|
| 561 |
+
"learning_rate": 4.967541138968363e-05,
|
| 562 |
+
"loss": 0.4306,
|
| 563 |
+
"step": 7680
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.08248708540781008,
|
| 567 |
+
"grad_norm": 0.1007547378540039,
|
| 568 |
+
"learning_rate": 4.96398901618296e-05,
|
| 569 |
+
"loss": 0.4089,
|
| 570 |
+
"step": 7936
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.08514795913064266,
|
| 574 |
+
"grad_norm": 0.10934588313102722,
|
| 575 |
+
"learning_rate": 4.960253894834191e-05,
|
| 576 |
+
"loss": 0.3899,
|
| 577 |
+
"step": 8192
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.08514795913064266,
|
| 581 |
+
"eval_acr_loss": 0.009613318397896364,
|
| 582 |
+
"eval_across_var": 0.06101540511008352,
|
| 583 |
+
"eval_bleu": 0.9421627142951156,
|
| 584 |
+
"eval_ce_loss": 0.1593283866532147,
|
| 585 |
+
"eval_cos_loss": 0.48021325934678316,
|
| 586 |
+
"eval_cov": 0.07448005676269531,
|
| 587 |
+
"eval_cov_loss": 0.008969669608632103,
|
| 588 |
+
"eval_glb_loss": 0.0033801320823840797,
|
| 589 |
+
"eval_global_kurtosis": 3.043630987405777,
|
| 590 |
+
"eval_global_mean": 0.006347730755805969,
|
| 591 |
+
"eval_global_var": 1.156005859375,
|
| 592 |
+
"eval_krt_loss": 0.002347310634149835,
|
| 593 |
+
"eval_loss": 0.36554089840501547,
|
| 594 |
+
"eval_mean_loss": 4.930327887819885e-05,
|
| 595 |
+
"eval_mse_loss": 1.2402296103537083,
|
| 596 |
+
"eval_per_loss": 1.130322118569893e-05,
|
| 597 |
+
"eval_per_var": 1.13140869140625,
|
| 598 |
+
"eval_within_var": 1.1352268271148205,
|
| 599 |
+
"eval_wth_loss": 0.00167783152524488,
|
| 600 |
+
"step": 8192
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.08514795913064266,
|
| 604 |
+
"eval_acr_loss": 0.009613318397896364,
|
| 605 |
+
"eval_across_var": 0.06101540511008352,
|
| 606 |
+
"eval_bleu": 0.9421627142951156,
|
| 607 |
+
"eval_ce_loss": 0.1593283866532147,
|
| 608 |
+
"eval_cos_loss": 0.48021325934678316,
|
| 609 |
+
"eval_cov": 0.07448005676269531,
|
| 610 |
+
"eval_cov_loss": 0.008969669608632103,
|
| 611 |
+
"eval_glb_loss": 0.0033801320823840797,
|
| 612 |
+
"eval_global_kurtosis": 3.043630987405777,
|
| 613 |
+
"eval_global_mean": 0.006347730755805969,
|
| 614 |
+
"eval_global_var": 1.156005859375,
|
| 615 |
+
"eval_krt_loss": 0.002347310634149835,
|
| 616 |
+
"eval_loss": 0.36554089840501547,
|
| 617 |
+
"eval_mean_loss": 4.930327887819885e-05,
|
| 618 |
+
"eval_mse_loss": 1.2402296103537083,
|
| 619 |
+
"eval_per_loss": 1.130322118569893e-05,
|
| 620 |
+
"eval_per_var": 1.13140869140625,
|
| 621 |
+
"eval_runtime": 9.7714,
|
| 622 |
+
"eval_samples_per_second": 204.679,
|
| 623 |
+
"eval_steps_per_second": 3.275,
|
| 624 |
+
"eval_within_var": 1.1352268271148205,
|
| 625 |
+
"eval_wth_loss": 0.00167783152524488,
|
| 626 |
+
"step": 8192
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.08780883285347525,
|
| 630 |
+
"grad_norm": 0.0978470966219902,
|
| 631 |
+
"learning_rate": 4.95633605232664e-05,
|
| 632 |
+
"loss": 0.3731,
|
| 633 |
+
"step": 8448
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.09046970657630783,
|
| 637 |
+
"grad_norm": 0.10450848191976547,
|
| 638 |
+
"learning_rate": 4.952235779635446e-05,
|
| 639 |
+
"loss": 0.3564,
|
| 640 |
+
"step": 8704
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.09313058029914041,
|
| 644 |
+
"grad_norm": 0.09945756942033768,
|
| 645 |
+
"learning_rate": 4.947953381284698e-05,
|
| 646 |
+
"loss": 0.3427,
|
| 647 |
+
"step": 8960
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.09579145402197299,
|
| 651 |
+
"grad_norm": 0.10317794233560562,
|
| 652 |
+
"learning_rate": 4.943506966910453e-05,
|
| 653 |
+
"loss": 0.3282,
|
| 654 |
+
"step": 9216
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.09579145402197299,
|
| 658 |
+
"eval_acr_loss": 0.009452910453546792,
|
| 659 |
+
"eval_across_var": 0.06200917391106486,
|
| 660 |
+
"eval_bleu": 0.9571379147452593,
|
| 661 |
+
"eval_ce_loss": 0.11980621726252139,
|
| 662 |
+
"eval_cos_loss": 0.43968157004565,
|
| 663 |
+
"eval_cov": 0.07451438903808594,
|
| 664 |
+
"eval_cov_loss": 0.009015302639454603,
|
| 665 |
+
"eval_glb_loss": 0.0033379476226400584,
|
| 666 |
+
"eval_global_kurtosis": 3.0460192561149597,
|
| 667 |
+
"eval_global_mean": 0.003891175612807274,
|
| 668 |
+
"eval_global_var": 1.155426025390625,
|
| 669 |
+
"eval_krt_loss": 0.0026789802294615583,
|
| 670 |
+
"eval_loss": 0.30993283446878195,
|
| 671 |
+
"eval_mean_loss": 2.458498882218496e-05,
|
| 672 |
+
"eval_mse_loss": 1.1764822266995907,
|
| 673 |
+
"eval_per_loss": 1.538718709070963e-05,
|
| 674 |
+
"eval_per_var": 1.13104248046875,
|
| 675 |
+
"eval_within_var": 1.1370676457881927,
|
| 676 |
+
"eval_wth_loss": 0.0018457178829738297,
|
| 677 |
+
"step": 9216
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.09579145402197299,
|
| 681 |
+
"eval_acr_loss": 0.009452910453546792,
|
| 682 |
+
"eval_across_var": 0.06200917391106486,
|
| 683 |
+
"eval_bleu": 0.9571379147452593,
|
| 684 |
+
"eval_ce_loss": 0.11980621726252139,
|
| 685 |
+
"eval_cos_loss": 0.43968157004565,
|
| 686 |
+
"eval_cov": 0.07451438903808594,
|
| 687 |
+
"eval_cov_loss": 0.009015302639454603,
|
| 688 |
+
"eval_glb_loss": 0.0033379476226400584,
|
| 689 |
+
"eval_global_kurtosis": 3.0460192561149597,
|
| 690 |
+
"eval_global_mean": 0.003891175612807274,
|
| 691 |
+
"eval_global_var": 1.155426025390625,
|
| 692 |
+
"eval_krt_loss": 0.0026789802294615583,
|
| 693 |
+
"eval_loss": 0.30993283446878195,
|
| 694 |
+
"eval_mean_loss": 2.458498882218496e-05,
|
| 695 |
+
"eval_mse_loss": 1.1764822266995907,
|
| 696 |
+
"eval_per_loss": 1.538718709070963e-05,
|
| 697 |
+
"eval_per_var": 1.13104248046875,
|
| 698 |
+
"eval_runtime": 10.3535,
|
| 699 |
+
"eval_samples_per_second": 193.172,
|
| 700 |
+
"eval_steps_per_second": 3.091,
|
| 701 |
+
"eval_within_var": 1.1370676457881927,
|
| 702 |
+
"eval_wth_loss": 0.0018457178829738297,
|
| 703 |
+
"step": 9216
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.09845232774480558,
|
| 707 |
+
"grad_norm": 0.09872151166200638,
|
| 708 |
+
"learning_rate": 4.938861993122928e-05,
|
| 709 |
+
"loss": 0.3186,
|
| 710 |
+
"step": 9472
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.10111320146763816,
|
| 714 |
+
"grad_norm": 0.09426742047071457,
|
| 715 |
+
"learning_rate": 4.934035886936657e-05,
|
| 716 |
+
"loss": 0.3055,
|
| 717 |
+
"step": 9728
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.10377407519047074,
|
| 721 |
+
"grad_norm": 0.09218986332416534,
|
| 722 |
+
"learning_rate": 4.929029006782824e-05,
|
| 723 |
+
"loss": 0.294,
|
| 724 |
+
"step": 9984
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.10643494891330332,
|
| 728 |
+
"grad_norm": 0.10592538118362427,
|
| 729 |
+
"learning_rate": 4.923841724518553e-05,
|
| 730 |
+
"loss": 0.2864,
|
| 731 |
+
"step": 10240
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.10643494891330332,
|
| 735 |
+
"eval_acr_loss": 0.009291668538935483,
|
| 736 |
+
"eval_across_var": 0.06304166710469872,
|
| 737 |
+
"eval_bleu": 0.9663901561046487,
|
| 738 |
+
"eval_ce_loss": 0.09433894976973534,
|
| 739 |
+
"eval_cos_loss": 0.40719034615904093,
|
| 740 |
+
"eval_cov": 0.07495689392089844,
|
| 741 |
+
"eval_cov_loss": 0.009098658832954243,
|
| 742 |
+
"eval_glb_loss": 0.003349686012370512,
|
| 743 |
+
"eval_global_kurtosis": 3.040327101945877,
|
| 744 |
+
"eval_global_mean": 0.004347227513790131,
|
| 745 |
+
"eval_global_var": 1.15533447265625,
|
| 746 |
+
"eval_krt_loss": 0.0023327972218396553,
|
| 747 |
+
"eval_loss": 0.27165302727371454,
|
| 748 |
+
"eval_mean_loss": 2.88587275694141e-05,
|
| 749 |
+
"eval_mse_loss": 1.1290423460304737,
|
| 750 |
+
"eval_per_loss": 2.1555981511767186e-05,
|
| 751 |
+
"eval_per_var": 1.13153076171875,
|
| 752 |
+
"eval_within_var": 1.1392224319279194,
|
| 753 |
+
"eval_wth_loss": 0.0020331931153663163,
|
| 754 |
+
"step": 10240
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.10643494891330332,
|
| 758 |
+
"eval_acr_loss": 0.009291668538935483,
|
| 759 |
+
"eval_across_var": 0.06304166710469872,
|
| 760 |
+
"eval_bleu": 0.9663901561046487,
|
| 761 |
+
"eval_ce_loss": 0.09433894976973534,
|
| 762 |
+
"eval_cos_loss": 0.40719034615904093,
|
| 763 |
+
"eval_cov": 0.07495689392089844,
|
| 764 |
+
"eval_cov_loss": 0.009098658832954243,
|
| 765 |
+
"eval_glb_loss": 0.003349686012370512,
|
| 766 |
+
"eval_global_kurtosis": 3.040327101945877,
|
| 767 |
+
"eval_global_mean": 0.004347227513790131,
|
| 768 |
+
"eval_global_var": 1.15533447265625,
|
| 769 |
+
"eval_krt_loss": 0.0023327972218396553,
|
| 770 |
+
"eval_loss": 0.27165302727371454,
|
| 771 |
+
"eval_mean_loss": 2.88587275694141e-05,
|
| 772 |
+
"eval_mse_loss": 1.1290423460304737,
|
| 773 |
+
"eval_per_loss": 2.1555981511767186e-05,
|
| 774 |
+
"eval_per_var": 1.13153076171875,
|
| 775 |
+
"eval_runtime": 9.8187,
|
| 776 |
+
"eval_samples_per_second": 203.694,
|
| 777 |
+
"eval_steps_per_second": 3.259,
|
| 778 |
+
"eval_within_var": 1.1392224319279194,
|
| 779 |
+
"eval_wth_loss": 0.0020331931153663163,
|
| 780 |
+
"step": 10240
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.10909582263613592,
|
| 784 |
+
"grad_norm": 0.09304367750883102,
|
| 785 |
+
"learning_rate": 4.918474425399298e-05,
|
| 786 |
+
"loss": 0.2784,
|
| 787 |
+
"step": 10496
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.1117566963589685,
|
| 791 |
+
"grad_norm": 0.10622046142816544,
|
| 792 |
+
"learning_rate": 4.912927508050224e-05,
|
| 793 |
+
"loss": 0.2678,
|
| 794 |
+
"step": 10752
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.11441757008180108,
|
| 798 |
+
"grad_norm": 0.08184197545051575,
|
| 799 |
+
"learning_rate": 4.907201384436601e-05,
|
| 800 |
+
"loss": 0.2623,
|
| 801 |
+
"step": 11008
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.11707844380463366,
|
| 805 |
+
"grad_norm": 0.09012167900800705,
|
| 806 |
+
"learning_rate": 4.901296479833215e-05,
|
| 807 |
+
"loss": 0.255,
|
| 808 |
+
"step": 11264
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.11707844380463366,
|
| 812 |
+
"eval_acr_loss": 0.009119439113419503,
|
| 813 |
+
"eval_across_var": 0.06416496774181724,
|
| 814 |
+
"eval_bleu": 0.9719281384812661,
|
| 815 |
+
"eval_ce_loss": 0.07750852114986628,
|
| 816 |
+
"eval_cos_loss": 0.3806372629478574,
|
| 817 |
+
"eval_cov": 0.0753173828125,
|
| 818 |
+
"eval_cov_loss": 0.009160888177575544,
|
| 819 |
+
"eval_glb_loss": 0.003381317452294752,
|
| 820 |
+
"eval_global_kurtosis": 3.0355821326375008,
|
| 821 |
+
"eval_global_mean": 0.0027519892901182175,
|
| 822 |
+
"eval_global_var": 1.155487060546875,
|
| 823 |
+
"eval_krt_loss": 0.001984861595602183,
|
| 824 |
+
"eval_loss": 0.2444434049539268,
|
| 825 |
+
"eval_mean_loss": 1.832497667531019e-05,
|
| 826 |
+
"eval_mse_loss": 1.0934999212622643,
|
| 827 |
+
"eval_per_loss": 2.6998875114259135e-05,
|
| 828 |
+
"eval_per_var": 1.132049560546875,
|
| 829 |
+
"eval_within_var": 1.1408722698688507,
|
| 830 |
+
"eval_wth_loss": 0.0021871452599953045,
|
| 831 |
+
"step": 11264
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.11707844380463366,
|
| 835 |
+
"eval_acr_loss": 0.009119439113419503,
|
| 836 |
+
"eval_across_var": 0.06416496774181724,
|
| 837 |
+
"eval_bleu": 0.9719281384812661,
|
| 838 |
+
"eval_ce_loss": 0.07750852114986628,
|
| 839 |
+
"eval_cos_loss": 0.3806372629478574,
|
| 840 |
+
"eval_cov": 0.0753173828125,
|
| 841 |
+
"eval_cov_loss": 0.009160888177575544,
|
| 842 |
+
"eval_glb_loss": 0.003381317452294752,
|
| 843 |
+
"eval_global_kurtosis": 3.0355821326375008,
|
| 844 |
+
"eval_global_mean": 0.0027519892901182175,
|
| 845 |
+
"eval_global_var": 1.155487060546875,
|
| 846 |
+
"eval_krt_loss": 0.001984861595602183,
|
| 847 |
+
"eval_loss": 0.2444434049539268,
|
| 848 |
+
"eval_mean_loss": 1.832497667531019e-05,
|
| 849 |
+
"eval_mse_loss": 1.0934999212622643,
|
| 850 |
+
"eval_per_loss": 2.6998875114259135e-05,
|
| 851 |
+
"eval_per_var": 1.132049560546875,
|
| 852 |
+
"eval_runtime": 10.5086,
|
| 853 |
+
"eval_samples_per_second": 190.321,
|
| 854 |
+
"eval_steps_per_second": 3.045,
|
| 855 |
+
"eval_within_var": 1.1408722698688507,
|
| 856 |
+
"eval_wth_loss": 0.0021871452599953045,
|
| 857 |
+
"step": 11264
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.11973931752746625,
|
| 861 |
+
"grad_norm": 0.08071508258581161,
|
| 862 |
+
"learning_rate": 4.8952132327927744e-05,
|
| 863 |
+
"loss": 0.2487,
|
| 864 |
+
"step": 11520
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.12240019125029883,
|
| 868 |
+
"grad_norm": 0.09782595932483673,
|
| 869 |
+
"learning_rate": 4.888952095113344e-05,
|
| 870 |
+
"loss": 0.2424,
|
| 871 |
+
"step": 11776
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.1250610649731314,
|
| 875 |
+
"grad_norm": 0.08052252233028412,
|
| 876 |
+
"learning_rate": 4.8825135318047885e-05,
|
| 877 |
+
"loss": 0.2379,
|
| 878 |
+
"step": 12032
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.127721938695964,
|
| 882 |
+
"grad_norm": 0.08125212043523788,
|
| 883 |
+
"learning_rate": 4.875898021054239e-05,
|
| 884 |
+
"loss": 0.2333,
|
| 885 |
+
"step": 12288
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.127721938695964,
|
| 889 |
+
"eval_acr_loss": 0.008987350542156491,
|
| 890 |
+
"eval_across_var": 0.06508267612662166,
|
| 891 |
+
"eval_bleu": 0.9753456951784488,
|
| 892 |
+
"eval_ce_loss": 0.06519961968297139,
|
| 893 |
+
"eval_cos_loss": 0.3581163277849555,
|
| 894 |
+
"eval_cov": 0.07555389404296875,
|
| 895 |
+
"eval_cov_loss": 0.009179135347949341,
|
| 896 |
+
"eval_glb_loss": 0.0035121891705784947,
|
| 897 |
+
"eval_global_kurtosis": 3.0320599749684334,
|
| 898 |
+
"eval_global_mean": 0.002966819331049919,
|
| 899 |
+
"eval_global_var": 1.156524658203125,
|
| 900 |
+
"eval_krt_loss": 0.0019287865870936116,
|
| 901 |
+
"eval_loss": 0.2234154366888106,
|
| 902 |
+
"eval_mean_loss": 1.9722260523491464e-05,
|
| 903 |
+
"eval_mse_loss": 1.0655292384326458,
|
| 904 |
+
"eval_per_loss": 3.665547956188675e-05,
|
| 905 |
+
"eval_per_var": 1.133544921875,
|
| 906 |
+
"eval_within_var": 1.1434408612549305,
|
| 907 |
+
"eval_wth_loss": 0.0024261503635898407,
|
| 908 |
+
"step": 12288
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.127721938695964,
|
| 912 |
+
"eval_acr_loss": 0.008987350542156491,
|
| 913 |
+
"eval_across_var": 0.06508267612662166,
|
| 914 |
+
"eval_bleu": 0.9753456951784488,
|
| 915 |
+
"eval_ce_loss": 0.06519961968297139,
|
| 916 |
+
"eval_cos_loss": 0.3581163277849555,
|
| 917 |
+
"eval_cov": 0.07555389404296875,
|
| 918 |
+
"eval_cov_loss": 0.009179135347949341,
|
| 919 |
+
"eval_glb_loss": 0.0035121891705784947,
|
| 920 |
+
"eval_global_kurtosis": 3.0320599749684334,
|
| 921 |
+
"eval_global_mean": 0.002966819331049919,
|
| 922 |
+
"eval_global_var": 1.156524658203125,
|
| 923 |
+
"eval_krt_loss": 0.0019287865870936116,
|
| 924 |
+
"eval_loss": 0.2234154366888106,
|
| 925 |
+
"eval_mean_loss": 1.9722260523491464e-05,
|
| 926 |
+
"eval_mse_loss": 1.0655292384326458,
|
| 927 |
+
"eval_per_loss": 3.665547956188675e-05,
|
| 928 |
+
"eval_per_var": 1.133544921875,
|
| 929 |
+
"eval_runtime": 9.9204,
|
| 930 |
+
"eval_samples_per_second": 201.605,
|
| 931 |
+
"eval_steps_per_second": 3.226,
|
| 932 |
+
"eval_within_var": 1.1434408612549305,
|
| 933 |
+
"eval_wth_loss": 0.0024261503635898407,
|
| 934 |
+
"step": 12288
|
| 935 |
+
}
|
| 936 |
+
],
|
| 937 |
+
"logging_steps": 256,
|
| 938 |
+
"max_steps": 96209,
|
| 939 |
+
"num_input_tokens_seen": 0,
|
| 940 |
+
"num_train_epochs": 1,
|
| 941 |
+
"save_steps": 1024,
|
| 942 |
+
"stateful_callbacks": {
|
| 943 |
+
"TrainerControl": {
|
| 944 |
+
"args": {
|
| 945 |
+
"should_epoch_stop": false,
|
| 946 |
+
"should_evaluate": false,
|
| 947 |
+
"should_log": false,
|
| 948 |
+
"should_save": true,
|
| 949 |
+
"should_training_stop": false
|
| 950 |
+
},
|
| 951 |
+
"attributes": {}
|
| 952 |
+
}
|
| 953 |
+
},
|
| 954 |
+
"total_flos": 0.0,
|
| 955 |
+
"train_batch_size": 64,
|
| 956 |
+
"trial_name": null,
|
| 957 |
+
"trial_params": null
|
| 958 |
+
}
|
checkpoints-v3-clean/checkpoint-12288/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8f3ab7adcc6155e451263bcf71495874027ca0e4b88c0533266ab8b55dd4176
|
| 3 |
+
size 5777
|