Training in progress, step 500
Browse files- all_results.json +17 -0
- eval_results.json +12 -0
- model.safetensors +1 -1
- train_results.json +8 -0
- trainer_state.json +861 -0
- training_args.bin +1 -1
all_results.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.4192811608208402,
|
| 4 |
+
"eval_bleu": 0.13449590326973765,
|
| 5 |
+
"eval_loss": 3.201490640640259,
|
| 6 |
+
"eval_perplexity": 24.569126652935235,
|
| 7 |
+
"eval_runtime": 12.2552,
|
| 8 |
+
"eval_samples": 1141,
|
| 9 |
+
"eval_samples_per_second": 93.103,
|
| 10 |
+
"eval_steps_per_second": 1.469,
|
| 11 |
+
"perplexity": 24.569126652935235,
|
| 12 |
+
"total_flos": 1.0587061010143642e+18,
|
| 13 |
+
"train_loss": 3.7750147520893753,
|
| 14 |
+
"train_runtime": 15697.9119,
|
| 15 |
+
"train_samples_per_second": 36.309,
|
| 16 |
+
"train_steps_per_second": 0.568
|
| 17 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.4192811608208402,
|
| 4 |
+
"eval_bleu": 0.13449590326973765,
|
| 5 |
+
"eval_loss": 3.201490640640259,
|
| 6 |
+
"eval_perplexity": 24.569126652935235,
|
| 7 |
+
"eval_runtime": 12.2552,
|
| 8 |
+
"eval_samples": 1141,
|
| 9 |
+
"eval_samples_per_second": 93.103,
|
| 10 |
+
"eval_steps_per_second": 1.469,
|
| 11 |
+
"perplexity": 24.569126652935235
|
| 12 |
+
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1419374664
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b765c548960fd047ec41d12f45252cf489516aa303338551af02eb00be09aac
|
| 3 |
size 1419374664
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 1.0587061010143642e+18,
|
| 4 |
+
"train_loss": 3.7750147520893753,
|
| 5 |
+
"train_runtime": 15697.9119,
|
| 6 |
+
"train_samples_per_second": 36.309,
|
| 7 |
+
"train_steps_per_second": 0.568
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,861 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.2064151763916016,
|
| 3 |
+
"best_model_checkpoint": "./output/models/rotating-head-gp-gpt2-medium-wikitext/checkpoint-8500",
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 8910,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05611672278338945,
|
| 13 |
+
"grad_norm": 1.7030696868896484,
|
| 14 |
+
"learning_rate": 1.1223344556677892e-05,
|
| 15 |
+
"loss": 8.993,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.1122334455667789,
|
| 20 |
+
"grad_norm": 1.4508291482925415,
|
| 21 |
+
"learning_rate": 2.2446689113355783e-05,
|
| 22 |
+
"loss": 7.3838,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.16835016835016836,
|
| 27 |
+
"grad_norm": 1.6375716924667358,
|
| 28 |
+
"learning_rate": 3.3670033670033675e-05,
|
| 29 |
+
"loss": 6.5795,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.2244668911335578,
|
| 34 |
+
"grad_norm": 1.417483925819397,
|
| 35 |
+
"learning_rate": 4.4893378226711566e-05,
|
| 36 |
+
"loss": 6.1961,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.28058361391694725,
|
| 41 |
+
"grad_norm": 1.7931348085403442,
|
| 42 |
+
"learning_rate": 5.611672278338945e-05,
|
| 43 |
+
"loss": 5.9051,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.28058361391694725,
|
| 48 |
+
"eval_accuracy": 0.2238085814179224,
|
| 49 |
+
"eval_bleu": 0.04813423949717817,
|
| 50 |
+
"eval_loss": 5.7417893409729,
|
| 51 |
+
"eval_perplexity": 311.6215095481783,
|
| 52 |
+
"eval_runtime": 12.102,
|
| 53 |
+
"eval_samples_per_second": 94.282,
|
| 54 |
+
"eval_steps_per_second": 1.487,
|
| 55 |
+
"step": 500
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.3367003367003367,
|
| 59 |
+
"grad_norm": 1.502821445465088,
|
| 60 |
+
"learning_rate": 6.734006734006735e-05,
|
| 61 |
+
"loss": 5.6571,
|
| 62 |
+
"step": 600
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.39281705948372614,
|
| 66 |
+
"grad_norm": 1.267421007156372,
|
| 67 |
+
"learning_rate": 7.856341189674523e-05,
|
| 68 |
+
"loss": 5.4352,
|
| 69 |
+
"step": 700
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.4489337822671156,
|
| 73 |
+
"grad_norm": 1.4694616794586182,
|
| 74 |
+
"learning_rate": 8.978675645342313e-05,
|
| 75 |
+
"loss": 5.2521,
|
| 76 |
+
"step": 800
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.5050505050505051,
|
| 80 |
+
"grad_norm": 1.2315590381622314,
|
| 81 |
+
"learning_rate": 9.988776655443322e-05,
|
| 82 |
+
"loss": 5.0471,
|
| 83 |
+
"step": 900
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.5611672278338945,
|
| 87 |
+
"grad_norm": 0.9069780707359314,
|
| 88 |
+
"learning_rate": 9.864072827035791e-05,
|
| 89 |
+
"loss": 4.8617,
|
| 90 |
+
"step": 1000
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.5611672278338945,
|
| 94 |
+
"eval_accuracy": 0.28116767459732034,
|
| 95 |
+
"eval_bleu": 0.07213900102284179,
|
| 96 |
+
"eval_loss": 4.740506649017334,
|
| 97 |
+
"eval_perplexity": 114.4921943457123,
|
| 98 |
+
"eval_runtime": 12.1272,
|
| 99 |
+
"eval_samples_per_second": 94.086,
|
| 100 |
+
"eval_steps_per_second": 1.484,
|
| 101 |
+
"step": 1000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.6172839506172839,
|
| 105 |
+
"grad_norm": 0.9436570405960083,
|
| 106 |
+
"learning_rate": 9.73936899862826e-05,
|
| 107 |
+
"loss": 4.7137,
|
| 108 |
+
"step": 1100
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.6734006734006734,
|
| 112 |
+
"grad_norm": 0.8758283853530884,
|
| 113 |
+
"learning_rate": 9.614665170220725e-05,
|
| 114 |
+
"loss": 4.5873,
|
| 115 |
+
"step": 1200
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.7295173961840629,
|
| 119 |
+
"grad_norm": 0.8786563277244568,
|
| 120 |
+
"learning_rate": 9.489961341813194e-05,
|
| 121 |
+
"loss": 4.4798,
|
| 122 |
+
"step": 1300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.7856341189674523,
|
| 126 |
+
"grad_norm": 0.9080842733383179,
|
| 127 |
+
"learning_rate": 9.365257513405662e-05,
|
| 128 |
+
"loss": 4.3915,
|
| 129 |
+
"step": 1400
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.8417508417508418,
|
| 133 |
+
"grad_norm": 0.9573113322257996,
|
| 134 |
+
"learning_rate": 9.24055368499813e-05,
|
| 135 |
+
"loss": 4.2992,
|
| 136 |
+
"step": 1500
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.8417508417508418,
|
| 140 |
+
"eval_accuracy": 0.317916663453968,
|
| 141 |
+
"eval_bleu": 0.08316362600465955,
|
| 142 |
+
"eval_loss": 4.227725982666016,
|
| 143 |
+
"eval_perplexity": 68.56114553586224,
|
| 144 |
+
"eval_runtime": 12.2167,
|
| 145 |
+
"eval_samples_per_second": 93.397,
|
| 146 |
+
"eval_steps_per_second": 1.473,
|
| 147 |
+
"step": 1500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.8978675645342312,
|
| 151 |
+
"grad_norm": 1.070070505142212,
|
| 152 |
+
"learning_rate": 9.115849856590598e-05,
|
| 153 |
+
"loss": 4.2381,
|
| 154 |
+
"step": 1600
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.9539842873176206,
|
| 158 |
+
"grad_norm": 1.0821105241775513,
|
| 159 |
+
"learning_rate": 8.991146028183066e-05,
|
| 160 |
+
"loss": 4.1477,
|
| 161 |
+
"step": 1700
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.0101010101010102,
|
| 165 |
+
"grad_norm": 1.0546807050704956,
|
| 166 |
+
"learning_rate": 8.866442199775533e-05,
|
| 167 |
+
"loss": 4.0864,
|
| 168 |
+
"step": 1800
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.0662177328843996,
|
| 172 |
+
"grad_norm": 1.0770821571350098,
|
| 173 |
+
"learning_rate": 8.741738371368002e-05,
|
| 174 |
+
"loss": 4.0053,
|
| 175 |
+
"step": 1900
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.122334455667789,
|
| 179 |
+
"grad_norm": 0.9692613482475281,
|
| 180 |
+
"learning_rate": 8.617034542960469e-05,
|
| 181 |
+
"loss": 3.9585,
|
| 182 |
+
"step": 2000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.122334455667789,
|
| 186 |
+
"eval_accuracy": 0.34663219226844794,
|
| 187 |
+
"eval_bleu": 0.09053001364011155,
|
| 188 |
+
"eval_loss": 3.9252164363861084,
|
| 189 |
+
"eval_perplexity": 50.66404241449891,
|
| 190 |
+
"eval_runtime": 12.3364,
|
| 191 |
+
"eval_samples_per_second": 92.491,
|
| 192 |
+
"eval_steps_per_second": 1.459,
|
| 193 |
+
"step": 2000
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.1784511784511784,
|
| 197 |
+
"grad_norm": 0.9078754782676697,
|
| 198 |
+
"learning_rate": 8.492330714552937e-05,
|
| 199 |
+
"loss": 3.9194,
|
| 200 |
+
"step": 2100
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.2345679012345678,
|
| 204 |
+
"grad_norm": 1.207867980003357,
|
| 205 |
+
"learning_rate": 8.367626886145406e-05,
|
| 206 |
+
"loss": 3.8743,
|
| 207 |
+
"step": 2200
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.2906846240179575,
|
| 211 |
+
"grad_norm": 1.2701904773712158,
|
| 212 |
+
"learning_rate": 8.242923057737873e-05,
|
| 213 |
+
"loss": 3.8596,
|
| 214 |
+
"step": 2300
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.3468013468013469,
|
| 218 |
+
"grad_norm": 1.4610146284103394,
|
| 219 |
+
"learning_rate": 8.11821922933034e-05,
|
| 220 |
+
"loss": 3.821,
|
| 221 |
+
"step": 2400
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.4029180695847363,
|
| 225 |
+
"grad_norm": 2.153532028198242,
|
| 226 |
+
"learning_rate": 7.993515400922809e-05,
|
| 227 |
+
"loss": 3.7838,
|
| 228 |
+
"step": 2500
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.4029180695847363,
|
| 232 |
+
"eval_accuracy": 0.36260658663191814,
|
| 233 |
+
"eval_bleu": 0.09858369505966838,
|
| 234 |
+
"eval_loss": 3.7564430236816406,
|
| 235 |
+
"eval_perplexity": 42.7959308162683,
|
| 236 |
+
"eval_runtime": 12.2808,
|
| 237 |
+
"eval_samples_per_second": 92.909,
|
| 238 |
+
"eval_steps_per_second": 1.466,
|
| 239 |
+
"step": 2500
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.4590347923681257,
|
| 243 |
+
"grad_norm": 1.3955634832382202,
|
| 244 |
+
"learning_rate": 7.868811572515277e-05,
|
| 245 |
+
"loss": 3.7668,
|
| 246 |
+
"step": 2600
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.5151515151515151,
|
| 250 |
+
"grad_norm": 1.3527532815933228,
|
| 251 |
+
"learning_rate": 7.744107744107744e-05,
|
| 252 |
+
"loss": 3.7424,
|
| 253 |
+
"step": 2700
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.5712682379349046,
|
| 257 |
+
"grad_norm": 0.9464800357818604,
|
| 258 |
+
"learning_rate": 7.619403915700213e-05,
|
| 259 |
+
"loss": 3.708,
|
| 260 |
+
"step": 2800
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.627384960718294,
|
| 264 |
+
"grad_norm": 1.9855220317840576,
|
| 265 |
+
"learning_rate": 7.49470008729268e-05,
|
| 266 |
+
"loss": 3.6981,
|
| 267 |
+
"step": 2900
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.6835016835016834,
|
| 271 |
+
"grad_norm": 2.4134933948516846,
|
| 272 |
+
"learning_rate": 7.369996258885148e-05,
|
| 273 |
+
"loss": 3.6863,
|
| 274 |
+
"step": 3000
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 1.6835016835016834,
|
| 278 |
+
"eval_accuracy": 0.3738518885956052,
|
| 279 |
+
"eval_bleu": 0.10686414409488354,
|
| 280 |
+
"eval_loss": 3.6387925148010254,
|
| 281 |
+
"eval_perplexity": 38.04586915451721,
|
| 282 |
+
"eval_runtime": 12.2104,
|
| 283 |
+
"eval_samples_per_second": 93.445,
|
| 284 |
+
"eval_steps_per_second": 1.474,
|
| 285 |
+
"step": 3000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.7396184062850728,
|
| 289 |
+
"grad_norm": 2.7664296627044678,
|
| 290 |
+
"learning_rate": 7.245292430477615e-05,
|
| 291 |
+
"loss": 3.6503,
|
| 292 |
+
"step": 3100
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.7957351290684624,
|
| 296 |
+
"grad_norm": 2.4914608001708984,
|
| 297 |
+
"learning_rate": 7.120588602070084e-05,
|
| 298 |
+
"loss": 3.6382,
|
| 299 |
+
"step": 3200
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.8518518518518519,
|
| 303 |
+
"grad_norm": 2.2810556888580322,
|
| 304 |
+
"learning_rate": 6.995884773662552e-05,
|
| 305 |
+
"loss": 3.6126,
|
| 306 |
+
"step": 3300
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 1.9079685746352413,
|
| 310 |
+
"grad_norm": 1.139969825744629,
|
| 311 |
+
"learning_rate": 6.871180945255021e-05,
|
| 312 |
+
"loss": 3.6042,
|
| 313 |
+
"step": 3400
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 1.964085297418631,
|
| 317 |
+
"grad_norm": 1.3077082633972168,
|
| 318 |
+
"learning_rate": 6.746477116847487e-05,
|
| 319 |
+
"loss": 3.5869,
|
| 320 |
+
"step": 3500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.964085297418631,
|
| 324 |
+
"eval_accuracy": 0.3825578735533218,
|
| 325 |
+
"eval_bleu": 0.11004806187463387,
|
| 326 |
+
"eval_loss": 3.5517890453338623,
|
| 327 |
+
"eval_perplexity": 34.875655837249155,
|
| 328 |
+
"eval_runtime": 12.1547,
|
| 329 |
+
"eval_samples_per_second": 93.873,
|
| 330 |
+
"eval_steps_per_second": 1.481,
|
| 331 |
+
"step": 3500
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 2.0202020202020203,
|
| 335 |
+
"grad_norm": 2.2133948802948,
|
| 336 |
+
"learning_rate": 6.621773288439955e-05,
|
| 337 |
+
"loss": 3.5521,
|
| 338 |
+
"step": 3600
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.0763187429854097,
|
| 342 |
+
"grad_norm": 1.2750334739685059,
|
| 343 |
+
"learning_rate": 6.497069460032424e-05,
|
| 344 |
+
"loss": 3.4978,
|
| 345 |
+
"step": 3700
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.132435465768799,
|
| 349 |
+
"grad_norm": 1.4778505563735962,
|
| 350 |
+
"learning_rate": 6.372365631624892e-05,
|
| 351 |
+
"loss": 3.4785,
|
| 352 |
+
"step": 3800
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.1885521885521886,
|
| 356 |
+
"grad_norm": 1.5821393728256226,
|
| 357 |
+
"learning_rate": 6.247661803217359e-05,
|
| 358 |
+
"loss": 3.4804,
|
| 359 |
+
"step": 3900
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 2.244668911335578,
|
| 363 |
+
"grad_norm": 1.2070300579071045,
|
| 364 |
+
"learning_rate": 6.122957974809826e-05,
|
| 365 |
+
"loss": 3.4733,
|
| 366 |
+
"step": 4000
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 2.244668911335578,
|
| 370 |
+
"eval_accuracy": 0.38860631419507335,
|
| 371 |
+
"eval_bleu": 0.11590933164079384,
|
| 372 |
+
"eval_loss": 3.4845945835113525,
|
| 373 |
+
"eval_perplexity": 32.60920412059944,
|
| 374 |
+
"eval_runtime": 12.166,
|
| 375 |
+
"eval_samples_per_second": 93.786,
|
| 376 |
+
"eval_steps_per_second": 1.48,
|
| 377 |
+
"step": 4000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.3007856341189674,
|
| 381 |
+
"grad_norm": 1.0372384786605835,
|
| 382 |
+
"learning_rate": 5.998254146402295e-05,
|
| 383 |
+
"loss": 3.4604,
|
| 384 |
+
"step": 4100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.356902356902357,
|
| 388 |
+
"grad_norm": 1.3492435216903687,
|
| 389 |
+
"learning_rate": 5.8735503179947625e-05,
|
| 390 |
+
"loss": 3.4395,
|
| 391 |
+
"step": 4200
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.4130190796857462,
|
| 395 |
+
"grad_norm": 1.0056568384170532,
|
| 396 |
+
"learning_rate": 5.748846489587231e-05,
|
| 397 |
+
"loss": 3.4359,
|
| 398 |
+
"step": 4300
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.4691358024691357,
|
| 402 |
+
"grad_norm": 2.51788067817688,
|
| 403 |
+
"learning_rate": 5.624142661179699e-05,
|
| 404 |
+
"loss": 3.4198,
|
| 405 |
+
"step": 4400
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.525252525252525,
|
| 409 |
+
"grad_norm": 2.3806328773498535,
|
| 410 |
+
"learning_rate": 5.4994388327721666e-05,
|
| 411 |
+
"loss": 3.4122,
|
| 412 |
+
"step": 4500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 2.525252525252525,
|
| 416 |
+
"eval_accuracy": 0.39410559754909646,
|
| 417 |
+
"eval_bleu": 0.12123148173651561,
|
| 418 |
+
"eval_loss": 3.430687665939331,
|
| 419 |
+
"eval_perplexity": 30.897882867432642,
|
| 420 |
+
"eval_runtime": 12.1642,
|
| 421 |
+
"eval_samples_per_second": 93.8,
|
| 422 |
+
"eval_steps_per_second": 1.48,
|
| 423 |
+
"step": 4500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.581369248035915,
|
| 427 |
+
"grad_norm": 1.4060360193252563,
|
| 428 |
+
"learning_rate": 5.374735004364634e-05,
|
| 429 |
+
"loss": 3.4137,
|
| 430 |
+
"step": 4600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 2.637485970819304,
|
| 434 |
+
"grad_norm": 1.47615647315979,
|
| 435 |
+
"learning_rate": 5.250031175957102e-05,
|
| 436 |
+
"loss": 3.3987,
|
| 437 |
+
"step": 4700
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.6936026936026938,
|
| 441 |
+
"grad_norm": 3.111424446105957,
|
| 442 |
+
"learning_rate": 5.12532734754957e-05,
|
| 443 |
+
"loss": 3.3934,
|
| 444 |
+
"step": 4800
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.749719416386083,
|
| 448 |
+
"grad_norm": 1.385186791419983,
|
| 449 |
+
"learning_rate": 5.000623519142038e-05,
|
| 450 |
+
"loss": 3.3842,
|
| 451 |
+
"step": 4900
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.8058361391694726,
|
| 455 |
+
"grad_norm": 1.4257937669754028,
|
| 456 |
+
"learning_rate": 4.8759196907345056e-05,
|
| 457 |
+
"loss": 3.3791,
|
| 458 |
+
"step": 5000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.8058361391694726,
|
| 462 |
+
"eval_accuracy": 0.39906000721357937,
|
| 463 |
+
"eval_bleu": 0.12230826630862061,
|
| 464 |
+
"eval_loss": 3.380352735519409,
|
| 465 |
+
"eval_perplexity": 29.38113305489852,
|
| 466 |
+
"eval_runtime": 12.1813,
|
| 467 |
+
"eval_samples_per_second": 93.668,
|
| 468 |
+
"eval_steps_per_second": 1.478,
|
| 469 |
+
"step": 5000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 2.861952861952862,
|
| 473 |
+
"grad_norm": 2.4099814891815186,
|
| 474 |
+
"learning_rate": 4.751215862326974e-05,
|
| 475 |
+
"loss": 3.3747,
|
| 476 |
+
"step": 5100
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.9180695847362514,
|
| 480 |
+
"grad_norm": 2.5423948764801025,
|
| 481 |
+
"learning_rate": 4.626512033919442e-05,
|
| 482 |
+
"loss": 3.3545,
|
| 483 |
+
"step": 5200
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 2.974186307519641,
|
| 487 |
+
"grad_norm": 2.864335775375366,
|
| 488 |
+
"learning_rate": 4.5018082055119096e-05,
|
| 489 |
+
"loss": 3.3586,
|
| 490 |
+
"step": 5300
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 3.0303030303030303,
|
| 494 |
+
"grad_norm": 2.2575385570526123,
|
| 495 |
+
"learning_rate": 4.3771043771043774e-05,
|
| 496 |
+
"loss": 3.3028,
|
| 497 |
+
"step": 5400
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 3.0864197530864197,
|
| 501 |
+
"grad_norm": 2.1973252296447754,
|
| 502 |
+
"learning_rate": 4.252400548696845e-05,
|
| 503 |
+
"loss": 3.2616,
|
| 504 |
+
"step": 5500
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 3.0864197530864197,
|
| 508 |
+
"eval_accuracy": 0.40264623561674817,
|
| 509 |
+
"eval_bleu": 0.12218084899348862,
|
| 510 |
+
"eval_loss": 3.3446905612945557,
|
| 511 |
+
"eval_perplexity": 28.35180116448892,
|
| 512 |
+
"eval_runtime": 12.1937,
|
| 513 |
+
"eval_samples_per_second": 93.573,
|
| 514 |
+
"eval_steps_per_second": 1.476,
|
| 515 |
+
"step": 5500
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 3.142536475869809,
|
| 519 |
+
"grad_norm": 1.4363641738891602,
|
| 520 |
+
"learning_rate": 4.127696720289313e-05,
|
| 521 |
+
"loss": 3.2647,
|
| 522 |
+
"step": 5600
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 3.1986531986531985,
|
| 526 |
+
"grad_norm": 1.4158735275268555,
|
| 527 |
+
"learning_rate": 4.002992891881781e-05,
|
| 528 |
+
"loss": 3.2614,
|
| 529 |
+
"step": 5700
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 3.254769921436588,
|
| 533 |
+
"grad_norm": 1.5934867858886719,
|
| 534 |
+
"learning_rate": 3.8782890634742486e-05,
|
| 535 |
+
"loss": 3.2573,
|
| 536 |
+
"step": 5800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 3.3108866442199774,
|
| 540 |
+
"grad_norm": 1.3518158197402954,
|
| 541 |
+
"learning_rate": 3.7535852350667164e-05,
|
| 542 |
+
"loss": 3.2584,
|
| 543 |
+
"step": 5900
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 3.3670033670033668,
|
| 547 |
+
"grad_norm": 1.7532979249954224,
|
| 548 |
+
"learning_rate": 3.628881406659185e-05,
|
| 549 |
+
"loss": 3.2499,
|
| 550 |
+
"step": 6000
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 3.3670033670033668,
|
| 554 |
+
"eval_accuracy": 0.4066985195027942,
|
| 555 |
+
"eval_bleu": 0.1260881414008664,
|
| 556 |
+
"eval_loss": 3.3095922470092773,
|
| 557 |
+
"eval_perplexity": 27.373961381341072,
|
| 558 |
+
"eval_runtime": 12.162,
|
| 559 |
+
"eval_samples_per_second": 93.817,
|
| 560 |
+
"eval_steps_per_second": 1.48,
|
| 561 |
+
"step": 6000
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 3.4231200897867566,
|
| 565 |
+
"grad_norm": 1.3179600238800049,
|
| 566 |
+
"learning_rate": 3.504177578251652e-05,
|
| 567 |
+
"loss": 3.2306,
|
| 568 |
+
"step": 6100
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 3.479236812570146,
|
| 572 |
+
"grad_norm": 3.5212039947509766,
|
| 573 |
+
"learning_rate": 3.3794737498441205e-05,
|
| 574 |
+
"loss": 3.2421,
|
| 575 |
+
"step": 6200
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 3.5353535353535355,
|
| 579 |
+
"grad_norm": 1.3254704475402832,
|
| 580 |
+
"learning_rate": 3.254769921436588e-05,
|
| 581 |
+
"loss": 3.2305,
|
| 582 |
+
"step": 6300
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 3.591470258136925,
|
| 586 |
+
"grad_norm": 3.1436195373535156,
|
| 587 |
+
"learning_rate": 3.130066093029056e-05,
|
| 588 |
+
"loss": 3.2314,
|
| 589 |
+
"step": 6400
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 3.6475869809203143,
|
| 593 |
+
"grad_norm": 2.763392210006714,
|
| 594 |
+
"learning_rate": 3.0053622646215242e-05,
|
| 595 |
+
"loss": 3.2277,
|
| 596 |
+
"step": 6500
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 3.6475869809203143,
|
| 600 |
+
"eval_accuracy": 0.41002601857539517,
|
| 601 |
+
"eval_bleu": 0.12992587011026865,
|
| 602 |
+
"eval_loss": 3.2811858654022217,
|
| 603 |
+
"eval_perplexity": 26.607306683090435,
|
| 604 |
+
"eval_runtime": 12.2242,
|
| 605 |
+
"eval_samples_per_second": 93.34,
|
| 606 |
+
"eval_steps_per_second": 1.472,
|
| 607 |
+
"step": 6500
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 3.7037037037037037,
|
| 611 |
+
"grad_norm": 1.2504295110702515,
|
| 612 |
+
"learning_rate": 2.880658436213992e-05,
|
| 613 |
+
"loss": 3.219,
|
| 614 |
+
"step": 6600
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 3.759820426487093,
|
| 618 |
+
"grad_norm": 1.316945195198059,
|
| 619 |
+
"learning_rate": 2.7559546078064598e-05,
|
| 620 |
+
"loss": 3.2109,
|
| 621 |
+
"step": 6700
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 3.8159371492704826,
|
| 625 |
+
"grad_norm": 1.9210931062698364,
|
| 626 |
+
"learning_rate": 2.6312507793989276e-05,
|
| 627 |
+
"loss": 3.1992,
|
| 628 |
+
"step": 6800
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 3.872053872053872,
|
| 632 |
+
"grad_norm": 1.6544625759124756,
|
| 633 |
+
"learning_rate": 2.5065469509913957e-05,
|
| 634 |
+
"loss": 3.1897,
|
| 635 |
+
"step": 6900
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 3.9281705948372614,
|
| 639 |
+
"grad_norm": 1.2679674625396729,
|
| 640 |
+
"learning_rate": 2.3818431225838632e-05,
|
| 641 |
+
"loss": 3.1992,
|
| 642 |
+
"step": 7000
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 3.9281705948372614,
|
| 646 |
+
"eval_accuracy": 0.4128223514726582,
|
| 647 |
+
"eval_bleu": 0.1304892476483967,
|
| 648 |
+
"eval_loss": 3.252329111099243,
|
| 649 |
+
"eval_perplexity": 25.850478491671673,
|
| 650 |
+
"eval_runtime": 12.1672,
|
| 651 |
+
"eval_samples_per_second": 93.777,
|
| 652 |
+
"eval_steps_per_second": 1.479,
|
| 653 |
+
"step": 7000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.984287317620651,
|
| 657 |
+
"grad_norm": 1.1226177215576172,
|
| 658 |
+
"learning_rate": 2.2571392941763313e-05,
|
| 659 |
+
"loss": 3.1958,
|
| 660 |
+
"step": 7100
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 4.040404040404041,
|
| 664 |
+
"grad_norm": 1.0558810234069824,
|
| 665 |
+
"learning_rate": 2.132435465768799e-05,
|
| 666 |
+
"loss": 3.13,
|
| 667 |
+
"step": 7200
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 4.09652076318743,
|
| 671 |
+
"grad_norm": 1.1988404989242554,
|
| 672 |
+
"learning_rate": 2.007731637361267e-05,
|
| 673 |
+
"loss": 3.1249,
|
| 674 |
+
"step": 7300
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 4.1526374859708195,
|
| 678 |
+
"grad_norm": 1.914546251296997,
|
| 679 |
+
"learning_rate": 1.883027808953735e-05,
|
| 680 |
+
"loss": 3.1276,
|
| 681 |
+
"step": 7400
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 4.2087542087542085,
|
| 685 |
+
"grad_norm": 1.7904256582260132,
|
| 686 |
+
"learning_rate": 1.758323980546203e-05,
|
| 687 |
+
"loss": 3.13,
|
| 688 |
+
"step": 7500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 4.2087542087542085,
|
| 692 |
+
"eval_accuracy": 0.4154396299656541,
|
| 693 |
+
"eval_bleu": 0.13260343650795992,
|
| 694 |
+
"eval_loss": 3.233205795288086,
|
| 695 |
+
"eval_perplexity": 25.360828416165713,
|
| 696 |
+
"eval_runtime": 12.2696,
|
| 697 |
+
"eval_samples_per_second": 92.994,
|
| 698 |
+
"eval_steps_per_second": 1.467,
|
| 699 |
+
"step": 7500
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 4.264870931537598,
|
| 703 |
+
"grad_norm": 2.1336634159088135,
|
| 704 |
+
"learning_rate": 1.6336201521386706e-05,
|
| 705 |
+
"loss": 3.1166,
|
| 706 |
+
"step": 7600
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 4.320987654320987,
|
| 710 |
+
"grad_norm": 1.4960366487503052,
|
| 711 |
+
"learning_rate": 1.5089163237311384e-05,
|
| 712 |
+
"loss": 3.1033,
|
| 713 |
+
"step": 7700
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 4.377104377104377,
|
| 717 |
+
"grad_norm": 1.4756426811218262,
|
| 718 |
+
"learning_rate": 1.3842124953236066e-05,
|
| 719 |
+
"loss": 3.1145,
|
| 720 |
+
"step": 7800
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 4.433221099887767,
|
| 724 |
+
"grad_norm": 1.741774082183838,
|
| 725 |
+
"learning_rate": 1.2595086669160744e-05,
|
| 726 |
+
"loss": 3.1115,
|
| 727 |
+
"step": 7900
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 4.489337822671156,
|
| 731 |
+
"grad_norm": 1.0106322765350342,
|
| 732 |
+
"learning_rate": 1.1348048385085423e-05,
|
| 733 |
+
"loss": 3.0915,
|
| 734 |
+
"step": 8000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 4.489337822671156,
|
| 738 |
+
"eval_accuracy": 0.4168352262553727,
|
| 739 |
+
"eval_bleu": 0.1317196572953092,
|
| 740 |
+
"eval_loss": 3.220038414001465,
|
| 741 |
+
"eval_perplexity": 25.029081630049546,
|
| 742 |
+
"eval_runtime": 12.1779,
|
| 743 |
+
"eval_samples_per_second": 93.694,
|
| 744 |
+
"eval_steps_per_second": 1.478,
|
| 745 |
+
"step": 8000
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 4.545454545454545,
|
| 749 |
+
"grad_norm": 1.2306262254714966,
|
| 750 |
+
"learning_rate": 1.0101010101010101e-05,
|
| 751 |
+
"loss": 3.1086,
|
| 752 |
+
"step": 8100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.601571268237935,
|
| 756 |
+
"grad_norm": 1.141204833984375,
|
| 757 |
+
"learning_rate": 8.853971816934781e-06,
|
| 758 |
+
"loss": 3.0995,
|
| 759 |
+
"step": 8200
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.657687991021325,
|
| 763 |
+
"grad_norm": 2.441620349884033,
|
| 764 |
+
"learning_rate": 7.606933532859459e-06,
|
| 765 |
+
"loss": 3.106,
|
| 766 |
+
"step": 8300
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 4.713804713804714,
|
| 770 |
+
"grad_norm": 0.9955108761787415,
|
| 771 |
+
"learning_rate": 6.359895248784138e-06,
|
| 772 |
+
"loss": 3.0982,
|
| 773 |
+
"step": 8400
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 4.7699214365881035,
|
| 777 |
+
"grad_norm": 1.119258165359497,
|
| 778 |
+
"learning_rate": 5.112856964708817e-06,
|
| 779 |
+
"loss": 3.1011,
|
| 780 |
+
"step": 8500
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 4.7699214365881035,
|
| 784 |
+
"eval_accuracy": 0.4185786507179739,
|
| 785 |
+
"eval_bleu": 0.13349497596917598,
|
| 786 |
+
"eval_loss": 3.2064151763916016,
|
| 787 |
+
"eval_perplexity": 24.690416598736768,
|
| 788 |
+
"eval_runtime": 12.2406,
|
| 789 |
+
"eval_samples_per_second": 93.214,
|
| 790 |
+
"eval_steps_per_second": 1.471,
|
| 791 |
+
"step": 8500
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 4.8260381593714925,
|
| 795 |
+
"grad_norm": 1.438043236732483,
|
| 796 |
+
"learning_rate": 3.865818680633495e-06,
|
| 797 |
+
"loss": 3.0915,
|
| 798 |
+
"step": 8600
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 4.882154882154882,
|
| 802 |
+
"grad_norm": 1.4048082828521729,
|
| 803 |
+
"learning_rate": 2.6187803965581742e-06,
|
| 804 |
+
"loss": 3.1015,
|
| 805 |
+
"step": 8700
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 4.938271604938271,
|
| 809 |
+
"grad_norm": 1.0656932592391968,
|
| 810 |
+
"learning_rate": 1.3717421124828533e-06,
|
| 811 |
+
"loss": 3.0977,
|
| 812 |
+
"step": 8800
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 4.994388327721661,
|
| 816 |
+
"grad_norm": 0.9609319567680359,
|
| 817 |
+
"learning_rate": 1.2470382840753213e-07,
|
| 818 |
+
"loss": 3.0882,
|
| 819 |
+
"step": 8900
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 5.0,
|
| 823 |
+
"step": 8910,
|
| 824 |
+
"total_flos": 1.0587061010143642e+18,
|
| 825 |
+
"train_loss": 3.7750147520893753,
|
| 826 |
+
"train_runtime": 15697.9119,
|
| 827 |
+
"train_samples_per_second": 36.309,
|
| 828 |
+
"train_steps_per_second": 0.568
|
| 829 |
+
}
|
| 830 |
+
],
|
| 831 |
+
"logging_steps": 100,
|
| 832 |
+
"max_steps": 8910,
|
| 833 |
+
"num_input_tokens_seen": 0,
|
| 834 |
+
"num_train_epochs": 5,
|
| 835 |
+
"save_steps": 500,
|
| 836 |
+
"stateful_callbacks": {
|
| 837 |
+
"EarlyStoppingCallback": {
|
| 838 |
+
"args": {
|
| 839 |
+
"early_stopping_patience": 2,
|
| 840 |
+
"early_stopping_threshold": 0.0
|
| 841 |
+
},
|
| 842 |
+
"attributes": {
|
| 843 |
+
"early_stopping_patience_counter": 0
|
| 844 |
+
}
|
| 845 |
+
},
|
| 846 |
+
"TrainerControl": {
|
| 847 |
+
"args": {
|
| 848 |
+
"should_epoch_stop": false,
|
| 849 |
+
"should_evaluate": false,
|
| 850 |
+
"should_log": false,
|
| 851 |
+
"should_save": true,
|
| 852 |
+
"should_training_stop": true
|
| 853 |
+
},
|
| 854 |
+
"attributes": {}
|
| 855 |
+
}
|
| 856 |
+
},
|
| 857 |
+
"total_flos": 1.0587061010143642e+18,
|
| 858 |
+
"train_batch_size": 64,
|
| 859 |
+
"trial_name": null,
|
| 860 |
+
"trial_params": null
|
| 861 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5560
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9698132b7711dafc8a37071c574a7e35f5549f06a7d2a7c7a9c45ef5912876ca
|
| 3 |
size 5560
|