End of training
Browse files- all_results.json +8 -0
- test_results.json +8 -0
- trainer_state.json +1038 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.005917159763314,
|
| 3 |
+
"eval_accuracy": 0.890927624872579,
|
| 4 |
+
"eval_loss": 0.4576520621776581,
|
| 5 |
+
"eval_runtime": 1124.6542,
|
| 6 |
+
"eval_samples_per_second": 0.872,
|
| 7 |
+
"eval_steps_per_second": 0.437
|
| 8 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.005917159763314,
|
| 3 |
+
"eval_accuracy": 0.890927624872579,
|
| 4 |
+
"eval_loss": 0.4576520621776581,
|
| 5 |
+
"eval_runtime": 1124.6542,
|
| 6 |
+
"eval_samples_per_second": 0.872,
|
| 7 |
+
"eval_steps_per_second": 0.437
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1038 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8978675645342312,
|
| 3 |
+
"best_model_checkpoint": "videomae-base-finetuned-dd\\checkpoint-1344",
|
| 4 |
+
"epoch": 4.005917159763314,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1352,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0073964497041420114,
|
| 13 |
+
"grad_norm": 22.897619247436523,
|
| 14 |
+
"learning_rate": 3.6764705882352942e-06,
|
| 15 |
+
"loss": 0.7648,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.014792899408284023,
|
| 20 |
+
"grad_norm": 6.239708423614502,
|
| 21 |
+
"learning_rate": 7.3529411764705884e-06,
|
| 22 |
+
"loss": 0.6619,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.022189349112426034,
|
| 27 |
+
"grad_norm": 11.775946617126465,
|
| 28 |
+
"learning_rate": 1.1029411764705883e-05,
|
| 29 |
+
"loss": 0.6844,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.029585798816568046,
|
| 34 |
+
"grad_norm": 12.948445320129395,
|
| 35 |
+
"learning_rate": 1.4705882352941177e-05,
|
| 36 |
+
"loss": 0.6158,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03698224852071006,
|
| 41 |
+
"grad_norm": 18.087614059448242,
|
| 42 |
+
"learning_rate": 1.8382352941176472e-05,
|
| 43 |
+
"loss": 0.6569,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04437869822485207,
|
| 48 |
+
"grad_norm": 10.075726509094238,
|
| 49 |
+
"learning_rate": 2.2058823529411766e-05,
|
| 50 |
+
"loss": 0.6572,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.051775147928994084,
|
| 55 |
+
"grad_norm": 4.17144250869751,
|
| 56 |
+
"learning_rate": 2.5735294117647057e-05,
|
| 57 |
+
"loss": 0.3788,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.05917159763313609,
|
| 62 |
+
"grad_norm": 24.00593376159668,
|
| 63 |
+
"learning_rate": 2.9411764705882354e-05,
|
| 64 |
+
"loss": 0.5752,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06656804733727811,
|
| 69 |
+
"grad_norm": 0.1779569834470749,
|
| 70 |
+
"learning_rate": 3.308823529411765e-05,
|
| 71 |
+
"loss": 0.488,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.07396449704142012,
|
| 76 |
+
"grad_norm": 2.1989564895629883,
|
| 77 |
+
"learning_rate": 3.6764705882352945e-05,
|
| 78 |
+
"loss": 1.0225,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08136094674556213,
|
| 83 |
+
"grad_norm": 16.50670623779297,
|
| 84 |
+
"learning_rate": 4.044117647058824e-05,
|
| 85 |
+
"loss": 1.0776,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08875739644970414,
|
| 90 |
+
"grad_norm": 36.278076171875,
|
| 91 |
+
"learning_rate": 4.411764705882353e-05,
|
| 92 |
+
"loss": 1.1623,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09615384615384616,
|
| 97 |
+
"grad_norm": 8.829483032226562,
|
| 98 |
+
"learning_rate": 4.7794117647058826e-05,
|
| 99 |
+
"loss": 0.9484,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.10355029585798817,
|
| 104 |
+
"grad_norm": 0.47137251496315,
|
| 105 |
+
"learning_rate": 4.983552631578948e-05,
|
| 106 |
+
"loss": 0.2775,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.11094674556213018,
|
| 111 |
+
"grad_norm": 1.3937591314315796,
|
| 112 |
+
"learning_rate": 4.942434210526316e-05,
|
| 113 |
+
"loss": 0.959,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.11834319526627218,
|
| 118 |
+
"grad_norm": 14.168859481811523,
|
| 119 |
+
"learning_rate": 4.901315789473684e-05,
|
| 120 |
+
"loss": 1.2776,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1257396449704142,
|
| 125 |
+
"grad_norm": 209.46127319335938,
|
| 126 |
+
"learning_rate": 4.860197368421053e-05,
|
| 127 |
+
"loss": 1.1028,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.13313609467455623,
|
| 132 |
+
"grad_norm": 0.22284063696861267,
|
| 133 |
+
"learning_rate": 4.819078947368421e-05,
|
| 134 |
+
"loss": 0.2779,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.14053254437869822,
|
| 139 |
+
"grad_norm": 0.16761070489883423,
|
| 140 |
+
"learning_rate": 4.7779605263157896e-05,
|
| 141 |
+
"loss": 1.05,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.14792899408284024,
|
| 146 |
+
"grad_norm": 33.303768157958984,
|
| 147 |
+
"learning_rate": 4.736842105263158e-05,
|
| 148 |
+
"loss": 0.5783,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.15532544378698224,
|
| 153 |
+
"grad_norm": 1.759582281112671,
|
| 154 |
+
"learning_rate": 4.6957236842105265e-05,
|
| 155 |
+
"loss": 0.5052,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.16272189349112426,
|
| 160 |
+
"grad_norm": 0.16995219886302948,
|
| 161 |
+
"learning_rate": 4.654605263157895e-05,
|
| 162 |
+
"loss": 0.8945,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.17011834319526628,
|
| 167 |
+
"grad_norm": 5.4278740882873535,
|
| 168 |
+
"learning_rate": 4.6134868421052635e-05,
|
| 169 |
+
"loss": 1.1972,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.17751479289940827,
|
| 174 |
+
"grad_norm": 5.3494038581848145,
|
| 175 |
+
"learning_rate": 4.572368421052632e-05,
|
| 176 |
+
"loss": 0.5657,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.1849112426035503,
|
| 181 |
+
"grad_norm": 38.29714584350586,
|
| 182 |
+
"learning_rate": 4.5312500000000004e-05,
|
| 183 |
+
"loss": 0.6919,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.19230769230769232,
|
| 188 |
+
"grad_norm": 31.136438369750977,
|
| 189 |
+
"learning_rate": 4.490131578947369e-05,
|
| 190 |
+
"loss": 0.4502,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.1997041420118343,
|
| 195 |
+
"grad_norm": 16.622379302978516,
|
| 196 |
+
"learning_rate": 4.449013157894737e-05,
|
| 197 |
+
"loss": 0.4002,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.20710059171597633,
|
| 202 |
+
"grad_norm": 0.7369871735572815,
|
| 203 |
+
"learning_rate": 4.407894736842105e-05,
|
| 204 |
+
"loss": 0.7867,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.21449704142011836,
|
| 209 |
+
"grad_norm": 1.566986083984375,
|
| 210 |
+
"learning_rate": 4.3667763157894735e-05,
|
| 211 |
+
"loss": 0.9427,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.22189349112426035,
|
| 216 |
+
"grad_norm": 20.51339340209961,
|
| 217 |
+
"learning_rate": 4.3256578947368426e-05,
|
| 218 |
+
"loss": 1.3318,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.22928994082840237,
|
| 223 |
+
"grad_norm": 16.819499969482422,
|
| 224 |
+
"learning_rate": 4.284539473684211e-05,
|
| 225 |
+
"loss": 0.7456,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.23668639053254437,
|
| 230 |
+
"grad_norm": 0.06786404550075531,
|
| 231 |
+
"learning_rate": 4.2434210526315796e-05,
|
| 232 |
+
"loss": 0.0428,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2440828402366864,
|
| 237 |
+
"grad_norm": 18.684722900390625,
|
| 238 |
+
"learning_rate": 4.202302631578947e-05,
|
| 239 |
+
"loss": 0.5777,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2485207100591716,
|
| 244 |
+
"eval_accuracy": 0.8372615039281706,
|
| 245 |
+
"eval_loss": 0.8213610649108887,
|
| 246 |
+
"eval_runtime": 1015.4502,
|
| 247 |
+
"eval_samples_per_second": 0.877,
|
| 248 |
+
"eval_steps_per_second": 0.439,
|
| 249 |
+
"step": 336
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.0029585798816567,
|
| 253 |
+
"grad_norm": 0.05207992345094681,
|
| 254 |
+
"learning_rate": 4.161184210526316e-05,
|
| 255 |
+
"loss": 0.3122,
|
| 256 |
+
"step": 340
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.0103550295857988,
|
| 260 |
+
"grad_norm": 0.08761809021234512,
|
| 261 |
+
"learning_rate": 4.120065789473684e-05,
|
| 262 |
+
"loss": 0.8974,
|
| 263 |
+
"step": 350
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.017751479289941,
|
| 267 |
+
"grad_norm": 2.162889003753662,
|
| 268 |
+
"learning_rate": 4.078947368421053e-05,
|
| 269 |
+
"loss": 0.5468,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.0251479289940828,
|
| 274 |
+
"grad_norm": 0.3578147888183594,
|
| 275 |
+
"learning_rate": 4.037828947368421e-05,
|
| 276 |
+
"loss": 0.4065,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.032544378698225,
|
| 281 |
+
"grad_norm": 0.1941654235124588,
|
| 282 |
+
"learning_rate": 3.9967105263157896e-05,
|
| 283 |
+
"loss": 0.5691,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.0399408284023668,
|
| 288 |
+
"grad_norm": 0.4009058475494385,
|
| 289 |
+
"learning_rate": 3.955592105263158e-05,
|
| 290 |
+
"loss": 1.0974,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.047337278106509,
|
| 295 |
+
"grad_norm": 1.8374804258346558,
|
| 296 |
+
"learning_rate": 3.9144736842105265e-05,
|
| 297 |
+
"loss": 0.3989,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.054733727810651,
|
| 302 |
+
"grad_norm": 0.1476195603609085,
|
| 303 |
+
"learning_rate": 3.873355263157895e-05,
|
| 304 |
+
"loss": 0.9985,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.0621301775147929,
|
| 309 |
+
"grad_norm": 0.23249651491641998,
|
| 310 |
+
"learning_rate": 3.8322368421052634e-05,
|
| 311 |
+
"loss": 0.0116,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.069526627218935,
|
| 316 |
+
"grad_norm": 120.95174407958984,
|
| 317 |
+
"learning_rate": 3.791118421052632e-05,
|
| 318 |
+
"loss": 0.2631,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.0769230769230769,
|
| 323 |
+
"grad_norm": 0.18491333723068237,
|
| 324 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 325 |
+
"loss": 0.6313,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.084319526627219,
|
| 330 |
+
"grad_norm": 139.33804321289062,
|
| 331 |
+
"learning_rate": 3.708881578947369e-05,
|
| 332 |
+
"loss": 0.8515,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.0917159763313609,
|
| 337 |
+
"grad_norm": 0.22990567982196808,
|
| 338 |
+
"learning_rate": 3.6677631578947366e-05,
|
| 339 |
+
"loss": 0.7119,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.099112426035503,
|
| 344 |
+
"grad_norm": 1.84480881690979,
|
| 345 |
+
"learning_rate": 3.626644736842105e-05,
|
| 346 |
+
"loss": 0.2762,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.106508875739645,
|
| 351 |
+
"grad_norm": 12.12067699432373,
|
| 352 |
+
"learning_rate": 3.5855263157894735e-05,
|
| 353 |
+
"loss": 0.3234,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.113905325443787,
|
| 358 |
+
"grad_norm": 25.058439254760742,
|
| 359 |
+
"learning_rate": 3.5444078947368426e-05,
|
| 360 |
+
"loss": 0.869,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.121301775147929,
|
| 365 |
+
"grad_norm": 20.12900161743164,
|
| 366 |
+
"learning_rate": 3.503289473684211e-05,
|
| 367 |
+
"loss": 1.085,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.128698224852071,
|
| 372 |
+
"grad_norm": 12.086252212524414,
|
| 373 |
+
"learning_rate": 3.4621710526315795e-05,
|
| 374 |
+
"loss": 0.7208,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.136094674556213,
|
| 379 |
+
"grad_norm": 25.66025161743164,
|
| 380 |
+
"learning_rate": 3.421052631578947e-05,
|
| 381 |
+
"loss": 0.6842,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.143491124260355,
|
| 386 |
+
"grad_norm": 15.72556209564209,
|
| 387 |
+
"learning_rate": 3.379934210526316e-05,
|
| 388 |
+
"loss": 0.7922,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.150887573964497,
|
| 393 |
+
"grad_norm": 0.22508259117603302,
|
| 394 |
+
"learning_rate": 3.338815789473684e-05,
|
| 395 |
+
"loss": 0.4978,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.1582840236686391,
|
| 400 |
+
"grad_norm": 0.3129032254219055,
|
| 401 |
+
"learning_rate": 3.297697368421053e-05,
|
| 402 |
+
"loss": 0.9733,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.165680473372781,
|
| 407 |
+
"grad_norm": 2.5810675621032715,
|
| 408 |
+
"learning_rate": 3.256578947368421e-05,
|
| 409 |
+
"loss": 1.0189,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.1730769230769231,
|
| 414 |
+
"grad_norm": 0.23227569460868835,
|
| 415 |
+
"learning_rate": 3.2154605263157896e-05,
|
| 416 |
+
"loss": 0.1902,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.180473372781065,
|
| 421 |
+
"grad_norm": 41.80963134765625,
|
| 422 |
+
"learning_rate": 3.174342105263158e-05,
|
| 423 |
+
"loss": 0.676,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.1878698224852071,
|
| 428 |
+
"grad_norm": 0.23230111598968506,
|
| 429 |
+
"learning_rate": 3.1332236842105265e-05,
|
| 430 |
+
"loss": 0.7534,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.195266272189349,
|
| 435 |
+
"grad_norm": 0.22635290026664734,
|
| 436 |
+
"learning_rate": 3.092105263157895e-05,
|
| 437 |
+
"loss": 0.4239,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.202662721893491,
|
| 442 |
+
"grad_norm": 1.2736923694610596,
|
| 443 |
+
"learning_rate": 3.0509868421052634e-05,
|
| 444 |
+
"loss": 0.1946,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.2100591715976332,
|
| 449 |
+
"grad_norm": 138.90115356445312,
|
| 450 |
+
"learning_rate": 3.009868421052632e-05,
|
| 451 |
+
"loss": 0.3972,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.217455621301775,
|
| 456 |
+
"grad_norm": 0.8171183466911316,
|
| 457 |
+
"learning_rate": 2.96875e-05,
|
| 458 |
+
"loss": 0.3475,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.2248520710059172,
|
| 463 |
+
"grad_norm": 0.14913396537303925,
|
| 464 |
+
"learning_rate": 2.9276315789473684e-05,
|
| 465 |
+
"loss": 0.0097,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.232248520710059,
|
| 470 |
+
"grad_norm": 0.16601596772670746,
|
| 471 |
+
"learning_rate": 2.886513157894737e-05,
|
| 472 |
+
"loss": 0.7906,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.2396449704142012,
|
| 477 |
+
"grad_norm": 0.0358225516974926,
|
| 478 |
+
"learning_rate": 2.8453947368421054e-05,
|
| 479 |
+
"loss": 1.0754,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.2470414201183433,
|
| 484 |
+
"grad_norm": 18.65558433532715,
|
| 485 |
+
"learning_rate": 2.8042763157894735e-05,
|
| 486 |
+
"loss": 1.144,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.2492603550295858,
|
| 491 |
+
"eval_accuracy": 0.8754208754208754,
|
| 492 |
+
"eval_loss": 0.4453337490558624,
|
| 493 |
+
"eval_runtime": 1018.2198,
|
| 494 |
+
"eval_samples_per_second": 0.875,
|
| 495 |
+
"eval_steps_per_second": 0.438,
|
| 496 |
+
"step": 673
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.0051775147928996,
|
| 500 |
+
"grad_norm": 0.26922106742858887,
|
| 501 |
+
"learning_rate": 2.7631578947368426e-05,
|
| 502 |
+
"loss": 0.0545,
|
| 503 |
+
"step": 680
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.0125739644970415,
|
| 507 |
+
"grad_norm": 0.08718305081129074,
|
| 508 |
+
"learning_rate": 2.7220394736842107e-05,
|
| 509 |
+
"loss": 0.2609,
|
| 510 |
+
"step": 690
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.0199704142011834,
|
| 514 |
+
"grad_norm": 0.062128640711307526,
|
| 515 |
+
"learning_rate": 2.6809210526315792e-05,
|
| 516 |
+
"loss": 0.205,
|
| 517 |
+
"step": 700
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.0273668639053253,
|
| 521 |
+
"grad_norm": 0.03102479875087738,
|
| 522 |
+
"learning_rate": 2.6398026315789476e-05,
|
| 523 |
+
"loss": 0.2414,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.0347633136094676,
|
| 528 |
+
"grad_norm": 0.1195179671049118,
|
| 529 |
+
"learning_rate": 2.598684210526316e-05,
|
| 530 |
+
"loss": 0.8337,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.0421597633136095,
|
| 535 |
+
"grad_norm": 0.18848393857479095,
|
| 536 |
+
"learning_rate": 2.5575657894736842e-05,
|
| 537 |
+
"loss": 0.938,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.0495562130177514,
|
| 542 |
+
"grad_norm": 0.30771324038505554,
|
| 543 |
+
"learning_rate": 2.5164473684210527e-05,
|
| 544 |
+
"loss": 0.5181,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.0569526627218937,
|
| 549 |
+
"grad_norm": 45.85493469238281,
|
| 550 |
+
"learning_rate": 2.4753289473684215e-05,
|
| 551 |
+
"loss": 0.9593,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.0643491124260356,
|
| 556 |
+
"grad_norm": 0.8830978870391846,
|
| 557 |
+
"learning_rate": 2.4342105263157896e-05,
|
| 558 |
+
"loss": 0.0696,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.0717455621301775,
|
| 563 |
+
"grad_norm": 0.23656447231769562,
|
| 564 |
+
"learning_rate": 2.393092105263158e-05,
|
| 565 |
+
"loss": 0.3063,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.0791420118343193,
|
| 570 |
+
"grad_norm": 14.187068939208984,
|
| 571 |
+
"learning_rate": 2.3519736842105265e-05,
|
| 572 |
+
"loss": 1.1477,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.0865384615384617,
|
| 577 |
+
"grad_norm": 20.795085906982422,
|
| 578 |
+
"learning_rate": 2.3108552631578946e-05,
|
| 579 |
+
"loss": 0.6215,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.0939349112426036,
|
| 584 |
+
"grad_norm": 0.27157312631607056,
|
| 585 |
+
"learning_rate": 2.2697368421052634e-05,
|
| 586 |
+
"loss": 0.271,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.1013313609467454,
|
| 591 |
+
"grad_norm": 26.25171661376953,
|
| 592 |
+
"learning_rate": 2.228618421052632e-05,
|
| 593 |
+
"loss": 0.9012,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.1087278106508878,
|
| 598 |
+
"grad_norm": 0.18786455690860748,
|
| 599 |
+
"learning_rate": 2.1875e-05,
|
| 600 |
+
"loss": 0.1285,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.1161242603550297,
|
| 605 |
+
"grad_norm": 0.09338750690221786,
|
| 606 |
+
"learning_rate": 2.1463815789473684e-05,
|
| 607 |
+
"loss": 0.6236,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.1235207100591715,
|
| 612 |
+
"grad_norm": 0.15658584237098694,
|
| 613 |
+
"learning_rate": 2.105263157894737e-05,
|
| 614 |
+
"loss": 0.2534,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.1309171597633134,
|
| 619 |
+
"grad_norm": 0.03897108510136604,
|
| 620 |
+
"learning_rate": 2.0641447368421053e-05,
|
| 621 |
+
"loss": 0.0317,
|
| 622 |
+
"step": 850
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.1383136094674557,
|
| 626 |
+
"grad_norm": 0.17228557169437408,
|
| 627 |
+
"learning_rate": 2.0230263157894738e-05,
|
| 628 |
+
"loss": 0.5894,
|
| 629 |
+
"step": 860
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.1457100591715976,
|
| 633 |
+
"grad_norm": 37.38359069824219,
|
| 634 |
+
"learning_rate": 1.9819078947368423e-05,
|
| 635 |
+
"loss": 0.3589,
|
| 636 |
+
"step": 870
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.1531065088757395,
|
| 640 |
+
"grad_norm": 14.766074180603027,
|
| 641 |
+
"learning_rate": 1.9407894736842107e-05,
|
| 642 |
+
"loss": 0.2452,
|
| 643 |
+
"step": 880
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.160502958579882,
|
| 647 |
+
"grad_norm": 0.09724285453557968,
|
| 648 |
+
"learning_rate": 1.8996710526315788e-05,
|
| 649 |
+
"loss": 0.4429,
|
| 650 |
+
"step": 890
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.1678994082840237,
|
| 654 |
+
"grad_norm": 0.20985926687717438,
|
| 655 |
+
"learning_rate": 1.8585526315789476e-05,
|
| 656 |
+
"loss": 0.3037,
|
| 657 |
+
"step": 900
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.1752958579881656,
|
| 661 |
+
"grad_norm": 17.582971572875977,
|
| 662 |
+
"learning_rate": 1.8174342105263157e-05,
|
| 663 |
+
"loss": 0.8678,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.1826923076923075,
|
| 668 |
+
"grad_norm": 0.10595466196537018,
|
| 669 |
+
"learning_rate": 1.7763157894736842e-05,
|
| 670 |
+
"loss": 0.5378,
|
| 671 |
+
"step": 920
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.19008875739645,
|
| 675 |
+
"grad_norm": 0.08267045021057129,
|
| 676 |
+
"learning_rate": 1.7351973684210527e-05,
|
| 677 |
+
"loss": 0.3613,
|
| 678 |
+
"step": 930
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.1974852071005917,
|
| 682 |
+
"grad_norm": 0.33162182569503784,
|
| 683 |
+
"learning_rate": 1.694078947368421e-05,
|
| 684 |
+
"loss": 0.4509,
|
| 685 |
+
"step": 940
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 2.2048816568047336,
|
| 689 |
+
"grad_norm": 0.070041224360466,
|
| 690 |
+
"learning_rate": 1.6529605263157896e-05,
|
| 691 |
+
"loss": 0.2242,
|
| 692 |
+
"step": 950
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.212278106508876,
|
| 696 |
+
"grad_norm": 0.07664915174245834,
|
| 697 |
+
"learning_rate": 1.611842105263158e-05,
|
| 698 |
+
"loss": 0.0046,
|
| 699 |
+
"step": 960
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.219674556213018,
|
| 703 |
+
"grad_norm": 84.51457977294922,
|
| 704 |
+
"learning_rate": 1.5707236842105265e-05,
|
| 705 |
+
"loss": 0.4331,
|
| 706 |
+
"step": 970
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.2270710059171597,
|
| 710 |
+
"grad_norm": 0.13595078885555267,
|
| 711 |
+
"learning_rate": 1.5296052631578946e-05,
|
| 712 |
+
"loss": 0.2439,
|
| 713 |
+
"step": 980
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.234467455621302,
|
| 717 |
+
"grad_norm": 107.51094818115234,
|
| 718 |
+
"learning_rate": 1.4884868421052634e-05,
|
| 719 |
+
"loss": 0.1195,
|
| 720 |
+
"step": 990
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.241863905325444,
|
| 724 |
+
"grad_norm": 6.981244087219238,
|
| 725 |
+
"learning_rate": 1.4473684210526317e-05,
|
| 726 |
+
"loss": 0.2644,
|
| 727 |
+
"step": 1000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 2.2492603550295858,
|
| 731 |
+
"grad_norm": 0.01494936365634203,
|
| 732 |
+
"learning_rate": 1.4062500000000001e-05,
|
| 733 |
+
"loss": 0.2783,
|
| 734 |
+
"step": 1010
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.2492603550295858,
|
| 738 |
+
"eval_accuracy": 0.8170594837261503,
|
| 739 |
+
"eval_loss": 0.8930483460426331,
|
| 740 |
+
"eval_runtime": 977.9049,
|
| 741 |
+
"eval_samples_per_second": 0.911,
|
| 742 |
+
"eval_steps_per_second": 0.456,
|
| 743 |
+
"step": 1010
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.007396449704142,
|
| 747 |
+
"grad_norm": 2.669276714324951,
|
| 748 |
+
"learning_rate": 1.3651315789473684e-05,
|
| 749 |
+
"loss": 0.6067,
|
| 750 |
+
"step": 1020
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.014792899408284,
|
| 754 |
+
"grad_norm": 0.0429365374147892,
|
| 755 |
+
"learning_rate": 1.3240131578947369e-05,
|
| 756 |
+
"loss": 0.2312,
|
| 757 |
+
"step": 1030
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.022189349112426,
|
| 761 |
+
"grad_norm": 0.0598057359457016,
|
| 762 |
+
"learning_rate": 1.2828947368421055e-05,
|
| 763 |
+
"loss": 0.0344,
|
| 764 |
+
"step": 1040
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.029585798816568,
|
| 768 |
+
"grad_norm": 0.0926346629858017,
|
| 769 |
+
"learning_rate": 1.2417763157894738e-05,
|
| 770 |
+
"loss": 0.9614,
|
| 771 |
+
"step": 1050
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.03698224852071,
|
| 775 |
+
"grad_norm": 0.01585511490702629,
|
| 776 |
+
"learning_rate": 1.200657894736842e-05,
|
| 777 |
+
"loss": 0.7316,
|
| 778 |
+
"step": 1060
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.044378698224852,
|
| 782 |
+
"grad_norm": 0.10066540539264679,
|
| 783 |
+
"learning_rate": 1.1595394736842107e-05,
|
| 784 |
+
"loss": 0.0041,
|
| 785 |
+
"step": 1070
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.051775147928994,
|
| 789 |
+
"grad_norm": 0.06809567660093307,
|
| 790 |
+
"learning_rate": 1.118421052631579e-05,
|
| 791 |
+
"loss": 0.0092,
|
| 792 |
+
"step": 1080
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.059171597633136,
|
| 796 |
+
"grad_norm": 0.14008688926696777,
|
| 797 |
+
"learning_rate": 1.0773026315789474e-05,
|
| 798 |
+
"loss": 1.0007,
|
| 799 |
+
"step": 1090
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.0665680473372783,
|
| 803 |
+
"grad_norm": 0.030784226953983307,
|
| 804 |
+
"learning_rate": 1.0361842105263159e-05,
|
| 805 |
+
"loss": 0.4198,
|
| 806 |
+
"step": 1100
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.07396449704142,
|
| 810 |
+
"grad_norm": 0.07336018979549408,
|
| 811 |
+
"learning_rate": 9.950657894736842e-06,
|
| 812 |
+
"loss": 0.2221,
|
| 813 |
+
"step": 1110
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.081360946745562,
|
| 817 |
+
"grad_norm": 0.06217151880264282,
|
| 818 |
+
"learning_rate": 9.539473684210528e-06,
|
| 819 |
+
"loss": 0.2991,
|
| 820 |
+
"step": 1120
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.088757396449704,
|
| 824 |
+
"grad_norm": 0.07419371604919434,
|
| 825 |
+
"learning_rate": 9.128289473684211e-06,
|
| 826 |
+
"loss": 0.6161,
|
| 827 |
+
"step": 1130
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.0961538461538463,
|
| 831 |
+
"grad_norm": 0.09845346957445145,
|
| 832 |
+
"learning_rate": 8.717105263157894e-06,
|
| 833 |
+
"loss": 0.204,
|
| 834 |
+
"step": 1140
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.103550295857988,
|
| 838 |
+
"grad_norm": 0.020927123725414276,
|
| 839 |
+
"learning_rate": 8.30592105263158e-06,
|
| 840 |
+
"loss": 0.5403,
|
| 841 |
+
"step": 1150
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.11094674556213,
|
| 845 |
+
"grad_norm": 0.07396041601896286,
|
| 846 |
+
"learning_rate": 7.894736842105263e-06,
|
| 847 |
+
"loss": 0.4637,
|
| 848 |
+
"step": 1160
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.1183431952662723,
|
| 852 |
+
"grad_norm": 0.11606968194246292,
|
| 853 |
+
"learning_rate": 7.483552631578948e-06,
|
| 854 |
+
"loss": 1.3375,
|
| 855 |
+
"step": 1170
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.1257396449704142,
|
| 859 |
+
"grad_norm": 0.08246757835149765,
|
| 860 |
+
"learning_rate": 7.072368421052632e-06,
|
| 861 |
+
"loss": 0.4418,
|
| 862 |
+
"step": 1180
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.133136094674556,
|
| 866 |
+
"grad_norm": 0.17207257449626923,
|
| 867 |
+
"learning_rate": 6.661184210526317e-06,
|
| 868 |
+
"loss": 0.2426,
|
| 869 |
+
"step": 1190
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.140532544378698,
|
| 873 |
+
"grad_norm": 0.49036145210266113,
|
| 874 |
+
"learning_rate": 6.25e-06,
|
| 875 |
+
"loss": 0.2545,
|
| 876 |
+
"step": 1200
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.1479289940828403,
|
| 880 |
+
"grad_norm": 68.61939239501953,
|
| 881 |
+
"learning_rate": 5.838815789473685e-06,
|
| 882 |
+
"loss": 0.5084,
|
| 883 |
+
"step": 1210
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 3.155325443786982,
|
| 887 |
+
"grad_norm": 0.08831863105297089,
|
| 888 |
+
"learning_rate": 5.4276315789473686e-06,
|
| 889 |
+
"loss": 0.0049,
|
| 890 |
+
"step": 1220
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.162721893491124,
|
| 894 |
+
"grad_norm": 0.10537251830101013,
|
| 895 |
+
"learning_rate": 5.016447368421053e-06,
|
| 896 |
+
"loss": 0.1406,
|
| 897 |
+
"step": 1230
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.1701183431952664,
|
| 901 |
+
"grad_norm": 17.6816349029541,
|
| 902 |
+
"learning_rate": 4.605263157894737e-06,
|
| 903 |
+
"loss": 0.486,
|
| 904 |
+
"step": 1240
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 3.1775147928994083,
|
| 908 |
+
"grad_norm": 0.461302250623703,
|
| 909 |
+
"learning_rate": 4.194078947368421e-06,
|
| 910 |
+
"loss": 0.0042,
|
| 911 |
+
"step": 1250
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.18491124260355,
|
| 915 |
+
"grad_norm": 0.04792853444814682,
|
| 916 |
+
"learning_rate": 3.7828947368421055e-06,
|
| 917 |
+
"loss": 0.719,
|
| 918 |
+
"step": 1260
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 3.1923076923076925,
|
| 922 |
+
"grad_norm": 0.058049630373716354,
|
| 923 |
+
"learning_rate": 3.3717105263157897e-06,
|
| 924 |
+
"loss": 0.2919,
|
| 925 |
+
"step": 1270
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 3.1997041420118344,
|
| 929 |
+
"grad_norm": 13.063404083251953,
|
| 930 |
+
"learning_rate": 2.960526315789474e-06,
|
| 931 |
+
"loss": 0.2856,
|
| 932 |
+
"step": 1280
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 3.2071005917159763,
|
| 936 |
+
"grad_norm": 0.4628206193447113,
|
| 937 |
+
"learning_rate": 2.549342105263158e-06,
|
| 938 |
+
"loss": 0.0601,
|
| 939 |
+
"step": 1290
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 3.214497041420118,
|
| 943 |
+
"grad_norm": 0.045592982321977615,
|
| 944 |
+
"learning_rate": 2.138157894736842e-06,
|
| 945 |
+
"loss": 0.53,
|
| 946 |
+
"step": 1300
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 3.2218934911242605,
|
| 950 |
+
"grad_norm": 0.04378387704491615,
|
| 951 |
+
"learning_rate": 1.7269736842105266e-06,
|
| 952 |
+
"loss": 0.0033,
|
| 953 |
+
"step": 1310
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.2292899408284024,
|
| 957 |
+
"grad_norm": 0.052677396684885025,
|
| 958 |
+
"learning_rate": 1.3157894736842106e-06,
|
| 959 |
+
"loss": 0.003,
|
| 960 |
+
"step": 1320
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 3.2366863905325443,
|
| 964 |
+
"grad_norm": 0.07824493199586868,
|
| 965 |
+
"learning_rate": 9.046052631578948e-07,
|
| 966 |
+
"loss": 0.3084,
|
| 967 |
+
"step": 1330
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 3.2440828402366866,
|
| 971 |
+
"grad_norm": 0.05082060024142265,
|
| 972 |
+
"learning_rate": 4.934210526315789e-07,
|
| 973 |
+
"loss": 0.0037,
|
| 974 |
+
"step": 1340
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 3.247041420118343,
|
| 978 |
+
"eval_accuracy": 0.8978675645342312,
|
| 979 |
+
"eval_loss": 0.40774551033973694,
|
| 980 |
+
"eval_runtime": 1030.7199,
|
| 981 |
+
"eval_samples_per_second": 0.864,
|
| 982 |
+
"eval_steps_per_second": 0.433,
|
| 983 |
+
"step": 1344
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 4.004437869822485,
|
| 987 |
+
"grad_norm": 0.045703090727329254,
|
| 988 |
+
"learning_rate": 8.223684210526316e-08,
|
| 989 |
+
"loss": 0.2113,
|
| 990 |
+
"step": 1350
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 4.005917159763314,
|
| 994 |
+
"eval_accuracy": 0.8978675645342312,
|
| 995 |
+
"eval_loss": 0.40793702006340027,
|
| 996 |
+
"eval_runtime": 1018.6267,
|
| 997 |
+
"eval_samples_per_second": 0.875,
|
| 998 |
+
"eval_steps_per_second": 0.438,
|
| 999 |
+
"step": 1352
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.005917159763314,
|
| 1003 |
+
"step": 1352,
|
| 1004 |
+
"total_flos": 3.3668665207526523e+18,
|
| 1005 |
+
"train_loss": 0.5293050300165978,
|
| 1006 |
+
"train_runtime": 11070.5286,
|
| 1007 |
+
"train_samples_per_second": 0.244,
|
| 1008 |
+
"train_steps_per_second": 0.122
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 4.005917159763314,
|
| 1012 |
+
"eval_accuracy": 0.890927624872579,
|
| 1013 |
+
"eval_loss": 0.4576520323753357,
|
| 1014 |
+
"eval_runtime": 1130.3017,
|
| 1015 |
+
"eval_samples_per_second": 0.868,
|
| 1016 |
+
"eval_steps_per_second": 0.434,
|
| 1017 |
+
"step": 1352
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.005917159763314,
|
| 1021 |
+
"eval_accuracy": 0.890927624872579,
|
| 1022 |
+
"eval_loss": 0.4576520621776581,
|
| 1023 |
+
"eval_runtime": 1124.6542,
|
| 1024 |
+
"eval_samples_per_second": 0.872,
|
| 1025 |
+
"eval_steps_per_second": 0.437,
|
| 1026 |
+
"step": 1352
|
| 1027 |
+
}
|
| 1028 |
+
],
|
| 1029 |
+
"logging_steps": 10,
|
| 1030 |
+
"max_steps": 1352,
|
| 1031 |
+
"num_input_tokens_seen": 0,
|
| 1032 |
+
"num_train_epochs": 9223372036854775807,
|
| 1033 |
+
"save_steps": 500,
|
| 1034 |
+
"total_flos": 3.3668665207526523e+18,
|
| 1035 |
+
"train_batch_size": 2,
|
| 1036 |
+
"trial_name": null,
|
| 1037 |
+
"trial_params": null
|
| 1038 |
+
}
|