End of training
Browse files- README.md +14 -2
- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +2442 -0
README.md
CHANGED
|
@@ -1,11 +1,23 @@
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
|
|
|
|
|
|
| 4 |
metrics:
|
| 5 |
- accuracy
|
| 6 |
model-index:
|
| 7 |
- name: smolm-autoreg-bpe-counterfactual-babylm-indef_articles_with_pl_nouns-removal-3e-4
|
| 8 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -13,7 +25,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# smolm-autoreg-bpe-counterfactual-babylm-indef_articles_with_pl_nouns-removal-3e-4
|
| 15 |
|
| 16 |
-
This model was trained from scratch on
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
- Loss: 3.4054
|
| 19 |
- Accuracy: 0.4099
|
|
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
| 4 |
+
datasets:
|
| 5 |
+
- kanishka/counterfactual_babylm_aann_indef_articles_with_pl_nouns_removal
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
| 8 |
model-index:
|
| 9 |
- name: smolm-autoreg-bpe-counterfactual-babylm-indef_articles_with_pl_nouns-removal-3e-4
|
| 10 |
+
results:
|
| 11 |
+
- task:
|
| 12 |
+
name: Causal Language Modeling
|
| 13 |
+
type: text-generation
|
| 14 |
+
dataset:
|
| 15 |
+
name: kanishka/counterfactual_babylm_aann_indef_articles_with_pl_nouns_removal
|
| 16 |
+
type: kanishka/counterfactual_babylm_aann_indef_articles_with_pl_nouns_removal
|
| 17 |
+
metrics:
|
| 18 |
+
- name: Accuracy
|
| 19 |
+
type: accuracy
|
| 20 |
+
value: 0.40992788926628976
|
| 21 |
---
|
| 22 |
|
| 23 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 25 |
|
| 26 |
# smolm-autoreg-bpe-counterfactual-babylm-indef_articles_with_pl_nouns-removal-3e-4
|
| 27 |
|
| 28 |
+
This model was trained from scratch on the kanishka/counterfactual_babylm_aann_indef_articles_with_pl_nouns_removal dataset.
|
| 29 |
It achieves the following results on the evaluation set:
|
| 30 |
- Loss: 3.4054
|
| 31 |
- Accuracy: 0.4099
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.40992788926628976,
|
| 4 |
+
"eval_loss": 3.4053595066070557,
|
| 5 |
+
"eval_runtime": 148.7271,
|
| 6 |
+
"eval_samples": 57920,
|
| 7 |
+
"eval_samples_per_second": 389.438,
|
| 8 |
+
"eval_steps_per_second": 6.085,
|
| 9 |
+
"perplexity": 30.12512395849166,
|
| 10 |
+
"train_loss": 3.0558329318429918,
|
| 11 |
+
"train_runtime": 48026.3336,
|
| 12 |
+
"train_samples": 595216,
|
| 13 |
+
"train_samples_per_second": 247.871,
|
| 14 |
+
"train_steps_per_second": 7.746
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.40992788926628976,
|
| 4 |
+
"eval_loss": 3.4053595066070557,
|
| 5 |
+
"eval_runtime": 148.7271,
|
| 6 |
+
"eval_samples": 57920,
|
| 7 |
+
"eval_samples_per_second": 389.438,
|
| 8 |
+
"eval_steps_per_second": 6.085,
|
| 9 |
+
"perplexity": 30.12512395849166
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"train_loss": 3.0558329318429918,
|
| 4 |
+
"train_runtime": 48026.3336,
|
| 5 |
+
"train_samples": 595216,
|
| 6 |
+
"train_samples_per_second": 247.871,
|
| 7 |
+
"train_steps_per_second": 7.746
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2442 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 372020,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05,
|
| 13 |
+
"learning_rate": 9.375e-06,
|
| 14 |
+
"loss": 6.8474,
|
| 15 |
+
"step": 1000
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.11,
|
| 19 |
+
"learning_rate": 1.875e-05,
|
| 20 |
+
"loss": 5.3572,
|
| 21 |
+
"step": 2000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.16,
|
| 25 |
+
"learning_rate": 2.8125e-05,
|
| 26 |
+
"loss": 5.0383,
|
| 27 |
+
"step": 3000
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.22,
|
| 31 |
+
"learning_rate": 3.75e-05,
|
| 32 |
+
"loss": 4.8113,
|
| 33 |
+
"step": 4000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.27,
|
| 37 |
+
"learning_rate": 4.6874999999999994e-05,
|
| 38 |
+
"loss": 4.6399,
|
| 39 |
+
"step": 5000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.32,
|
| 43 |
+
"learning_rate": 5.625e-05,
|
| 44 |
+
"loss": 4.5038,
|
| 45 |
+
"step": 6000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.38,
|
| 49 |
+
"learning_rate": 6.5625e-05,
|
| 50 |
+
"loss": 4.3831,
|
| 51 |
+
"step": 7000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.43,
|
| 55 |
+
"learning_rate": 7.5e-05,
|
| 56 |
+
"loss": 4.2865,
|
| 57 |
+
"step": 8000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.48,
|
| 61 |
+
"learning_rate": 8.437499999999999e-05,
|
| 62 |
+
"loss": 4.1961,
|
| 63 |
+
"step": 9000
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.54,
|
| 67 |
+
"learning_rate": 9.374999999999999e-05,
|
| 68 |
+
"loss": 4.1295,
|
| 69 |
+
"step": 10000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.59,
|
| 73 |
+
"learning_rate": 0.000103115625,
|
| 74 |
+
"loss": 4.057,
|
| 75 |
+
"step": 11000
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.65,
|
| 79 |
+
"learning_rate": 0.000112490625,
|
| 80 |
+
"loss": 3.9959,
|
| 81 |
+
"step": 12000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.7,
|
| 85 |
+
"learning_rate": 0.000121865625,
|
| 86 |
+
"loss": 3.9361,
|
| 87 |
+
"step": 13000
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.75,
|
| 91 |
+
"learning_rate": 0.000131221875,
|
| 92 |
+
"loss": 3.8845,
|
| 93 |
+
"step": 14000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.81,
|
| 97 |
+
"learning_rate": 0.000140596875,
|
| 98 |
+
"loss": 3.844,
|
| 99 |
+
"step": 15000
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.86,
|
| 103 |
+
"learning_rate": 0.0001499625,
|
| 104 |
+
"loss": 3.8116,
|
| 105 |
+
"step": 16000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.91,
|
| 109 |
+
"learning_rate": 0.000159328125,
|
| 110 |
+
"loss": 3.7724,
|
| 111 |
+
"step": 17000
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.97,
|
| 115 |
+
"learning_rate": 0.000168703125,
|
| 116 |
+
"loss": 3.7446,
|
| 117 |
+
"step": 18000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 1.0,
|
| 121 |
+
"eval_accuracy": 0.3461573860094159,
|
| 122 |
+
"eval_loss": 3.9129602909088135,
|
| 123 |
+
"eval_runtime": 148.9938,
|
| 124 |
+
"eval_samples_per_second": 388.741,
|
| 125 |
+
"eval_steps_per_second": 6.074,
|
| 126 |
+
"step": 18601
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.02,
|
| 130 |
+
"learning_rate": 0.00017806875,
|
| 131 |
+
"loss": 3.7109,
|
| 132 |
+
"step": 19000
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 1.08,
|
| 136 |
+
"learning_rate": 0.00018744374999999999,
|
| 137 |
+
"loss": 3.6733,
|
| 138 |
+
"step": 20000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 1.13,
|
| 142 |
+
"learning_rate": 0.00019680937499999996,
|
| 143 |
+
"loss": 3.649,
|
| 144 |
+
"step": 21000
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.18,
|
| 148 |
+
"learning_rate": 0.00020618437499999995,
|
| 149 |
+
"loss": 3.6311,
|
| 150 |
+
"step": 22000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.24,
|
| 154 |
+
"learning_rate": 0.00021555937499999998,
|
| 155 |
+
"loss": 3.6133,
|
| 156 |
+
"step": 23000
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.29,
|
| 160 |
+
"learning_rate": 0.00022492499999999998,
|
| 161 |
+
"loss": 3.6013,
|
| 162 |
+
"step": 24000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.34,
|
| 166 |
+
"learning_rate": 0.00023429999999999998,
|
| 167 |
+
"loss": 3.5853,
|
| 168 |
+
"step": 25000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.4,
|
| 172 |
+
"learning_rate": 0.00024366562499999997,
|
| 173 |
+
"loss": 3.5709,
|
| 174 |
+
"step": 26000
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.45,
|
| 178 |
+
"learning_rate": 0.000253040625,
|
| 179 |
+
"loss": 3.566,
|
| 180 |
+
"step": 27000
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 1.51,
|
| 184 |
+
"learning_rate": 0.000262415625,
|
| 185 |
+
"loss": 3.5454,
|
| 186 |
+
"step": 28000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.56,
|
| 190 |
+
"learning_rate": 0.000271790625,
|
| 191 |
+
"loss": 3.5315,
|
| 192 |
+
"step": 29000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.61,
|
| 196 |
+
"learning_rate": 0.00028115624999999994,
|
| 197 |
+
"loss": 3.5155,
|
| 198 |
+
"step": 30000
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.67,
|
| 202 |
+
"learning_rate": 0.00029053124999999994,
|
| 203 |
+
"loss": 3.5075,
|
| 204 |
+
"step": 31000
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.72,
|
| 208 |
+
"learning_rate": 0.00029990624999999993,
|
| 209 |
+
"loss": 3.499,
|
| 210 |
+
"step": 32000
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.77,
|
| 214 |
+
"learning_rate": 0.00029912740427033704,
|
| 215 |
+
"loss": 3.4882,
|
| 216 |
+
"step": 33000
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.83,
|
| 220 |
+
"learning_rate": 0.00029824510322922176,
|
| 221 |
+
"loss": 3.4677,
|
| 222 |
+
"step": 34000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 1.88,
|
| 226 |
+
"learning_rate": 0.00029736280218810654,
|
| 227 |
+
"loss": 3.4639,
|
| 228 |
+
"step": 35000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.94,
|
| 232 |
+
"learning_rate": 0.0002964805011469913,
|
| 233 |
+
"loss": 3.4498,
|
| 234 |
+
"step": 36000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.99,
|
| 238 |
+
"learning_rate": 0.0002955990824069172,
|
| 239 |
+
"loss": 3.4295,
|
| 240 |
+
"step": 37000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.0,
|
| 244 |
+
"eval_accuracy": 0.37515525238085,
|
| 245 |
+
"eval_loss": 3.6268346309661865,
|
| 246 |
+
"eval_runtime": 150.4919,
|
| 247 |
+
"eval_samples_per_second": 384.871,
|
| 248 |
+
"eval_steps_per_second": 6.014,
|
| 249 |
+
"step": 37202
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.04,
|
| 253 |
+
"learning_rate": 0.00029471678136580196,
|
| 254 |
+
"loss": 3.3882,
|
| 255 |
+
"step": 38000
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 2.1,
|
| 259 |
+
"learning_rate": 0.0002938353626257279,
|
| 260 |
+
"loss": 3.3783,
|
| 261 |
+
"step": 39000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 2.15,
|
| 265 |
+
"learning_rate": 0.00029295306158461267,
|
| 266 |
+
"loss": 3.3682,
|
| 267 |
+
"step": 40000
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 2.2,
|
| 271 |
+
"learning_rate": 0.00029207076054349744,
|
| 272 |
+
"loss": 3.3669,
|
| 273 |
+
"step": 41000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 2.26,
|
| 277 |
+
"learning_rate": 0.0002911893418034233,
|
| 278 |
+
"loss": 3.3568,
|
| 279 |
+
"step": 42000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 2.31,
|
| 283 |
+
"learning_rate": 0.0002903070407623081,
|
| 284 |
+
"loss": 3.3567,
|
| 285 |
+
"step": 43000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 2.37,
|
| 289 |
+
"learning_rate": 0.00028942562202223396,
|
| 290 |
+
"loss": 3.3465,
|
| 291 |
+
"step": 44000
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.42,
|
| 295 |
+
"learning_rate": 0.00028854332098111874,
|
| 296 |
+
"loss": 3.3394,
|
| 297 |
+
"step": 45000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.47,
|
| 301 |
+
"learning_rate": 0.0002876610199400035,
|
| 302 |
+
"loss": 3.3411,
|
| 303 |
+
"step": 46000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.53,
|
| 307 |
+
"learning_rate": 0.0002867787188988883,
|
| 308 |
+
"loss": 3.3295,
|
| 309 |
+
"step": 47000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.58,
|
| 313 |
+
"learning_rate": 0.00028589730015881416,
|
| 314 |
+
"loss": 3.3291,
|
| 315 |
+
"step": 48000
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 2.63,
|
| 319 |
+
"learning_rate": 0.00028501588141874003,
|
| 320 |
+
"loss": 3.3172,
|
| 321 |
+
"step": 49000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.69,
|
| 325 |
+
"learning_rate": 0.0002841335803776248,
|
| 326 |
+
"loss": 3.3157,
|
| 327 |
+
"step": 50000
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 2.74,
|
| 331 |
+
"learning_rate": 0.00028325216163755073,
|
| 332 |
+
"loss": 3.3098,
|
| 333 |
+
"step": 51000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.8,
|
| 337 |
+
"learning_rate": 0.0002823698605964355,
|
| 338 |
+
"loss": 3.3071,
|
| 339 |
+
"step": 52000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.85,
|
| 343 |
+
"learning_rate": 0.00028148755955532023,
|
| 344 |
+
"loss": 3.2994,
|
| 345 |
+
"step": 53000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.9,
|
| 349 |
+
"learning_rate": 0.00028060614081524616,
|
| 350 |
+
"loss": 3.2919,
|
| 351 |
+
"step": 54000
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 2.96,
|
| 355 |
+
"learning_rate": 0.0002797238397741309,
|
| 356 |
+
"loss": 3.2878,
|
| 357 |
+
"step": 55000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 3.0,
|
| 361 |
+
"eval_accuracy": 0.38779592675795943,
|
| 362 |
+
"eval_loss": 3.5073165893554688,
|
| 363 |
+
"eval_runtime": 150.2329,
|
| 364 |
+
"eval_samples_per_second": 385.535,
|
| 365 |
+
"eval_steps_per_second": 6.024,
|
| 366 |
+
"step": 55803
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 3.01,
|
| 370 |
+
"learning_rate": 0.00027884153873301566,
|
| 371 |
+
"loss": 3.278,
|
| 372 |
+
"step": 56000
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.06,
|
| 376 |
+
"learning_rate": 0.0002779610022939827,
|
| 377 |
+
"loss": 3.2281,
|
| 378 |
+
"step": 57000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 3.12,
|
| 382 |
+
"learning_rate": 0.00027707870125286745,
|
| 383 |
+
"loss": 3.2237,
|
| 384 |
+
"step": 58000
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 3.17,
|
| 388 |
+
"learning_rate": 0.0002761972825127933,
|
| 389 |
+
"loss": 3.2292,
|
| 390 |
+
"step": 59000
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 3.23,
|
| 394 |
+
"learning_rate": 0.0002753149814716781,
|
| 395 |
+
"loss": 3.2293,
|
| 396 |
+
"step": 60000
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 3.28,
|
| 400 |
+
"learning_rate": 0.0002744326804305629,
|
| 401 |
+
"loss": 3.2232,
|
| 402 |
+
"step": 61000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.33,
|
| 406 |
+
"learning_rate": 0.00027355037938944765,
|
| 407 |
+
"loss": 3.2279,
|
| 408 |
+
"step": 62000
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 3.39,
|
| 412 |
+
"learning_rate": 0.00027266807834833243,
|
| 413 |
+
"loss": 3.2248,
|
| 414 |
+
"step": 63000
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.44,
|
| 418 |
+
"learning_rate": 0.0002717857773072172,
|
| 419 |
+
"loss": 3.2182,
|
| 420 |
+
"step": 64000
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 3.49,
|
| 424 |
+
"learning_rate": 0.00027090347626610193,
|
| 425 |
+
"loss": 3.2216,
|
| 426 |
+
"step": 65000
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 3.55,
|
| 430 |
+
"learning_rate": 0.00027002205752602786,
|
| 431 |
+
"loss": 3.2212,
|
| 432 |
+
"step": 66000
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 3.6,
|
| 436 |
+
"learning_rate": 0.0002691406387859537,
|
| 437 |
+
"loss": 3.2125,
|
| 438 |
+
"step": 67000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 3.66,
|
| 442 |
+
"learning_rate": 0.0002682583377448385,
|
| 443 |
+
"loss": 3.2149,
|
| 444 |
+
"step": 68000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.71,
|
| 448 |
+
"learning_rate": 0.0002673760367037233,
|
| 449 |
+
"loss": 3.2145,
|
| 450 |
+
"step": 69000
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 3.76,
|
| 454 |
+
"learning_rate": 0.00026649373566260806,
|
| 455 |
+
"loss": 3.2114,
|
| 456 |
+
"step": 70000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.82,
|
| 460 |
+
"learning_rate": 0.00026561231692253393,
|
| 461 |
+
"loss": 3.2074,
|
| 462 |
+
"step": 71000
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 3.87,
|
| 466 |
+
"learning_rate": 0.0002647300158814187,
|
| 467 |
+
"loss": 3.2033,
|
| 468 |
+
"step": 72000
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 3.92,
|
| 472 |
+
"learning_rate": 0.00026384859714134463,
|
| 473 |
+
"loss": 3.2111,
|
| 474 |
+
"step": 73000
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 3.98,
|
| 478 |
+
"learning_rate": 0.00026296629610022935,
|
| 479 |
+
"loss": 3.2031,
|
| 480 |
+
"step": 74000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 4.0,
|
| 484 |
+
"eval_accuracy": 0.39482574918846874,
|
| 485 |
+
"eval_loss": 3.46301007270813,
|
| 486 |
+
"eval_runtime": 149.3064,
|
| 487 |
+
"eval_samples_per_second": 387.927,
|
| 488 |
+
"eval_steps_per_second": 6.061,
|
| 489 |
+
"step": 74404
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 4.03,
|
| 493 |
+
"learning_rate": 0.00026208399505911413,
|
| 494 |
+
"loss": 3.16,
|
| 495 |
+
"step": 75000
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 4.09,
|
| 499 |
+
"learning_rate": 0.0002612016940179989,
|
| 500 |
+
"loss": 3.1399,
|
| 501 |
+
"step": 76000
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 4.14,
|
| 505 |
+
"learning_rate": 0.00026032027527792483,
|
| 506 |
+
"loss": 3.1365,
|
| 507 |
+
"step": 77000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 4.19,
|
| 511 |
+
"learning_rate": 0.0002594379742368096,
|
| 512 |
+
"loss": 3.1442,
|
| 513 |
+
"step": 78000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 4.25,
|
| 517 |
+
"learning_rate": 0.0002585565554967355,
|
| 518 |
+
"loss": 3.1495,
|
| 519 |
+
"step": 79000
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 4.3,
|
| 523 |
+
"learning_rate": 0.00025767425445562026,
|
| 524 |
+
"loss": 3.1482,
|
| 525 |
+
"step": 80000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 4.35,
|
| 529 |
+
"learning_rate": 0.000256791953414505,
|
| 530 |
+
"loss": 3.1491,
|
| 531 |
+
"step": 81000
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 4.41,
|
| 535 |
+
"learning_rate": 0.0002559105346744309,
|
| 536 |
+
"loss": 3.1457,
|
| 537 |
+
"step": 82000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.46,
|
| 541 |
+
"learning_rate": 0.0002550291159343568,
|
| 542 |
+
"loss": 3.1516,
|
| 543 |
+
"step": 83000
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 4.52,
|
| 547 |
+
"learning_rate": 0.00025414681489324155,
|
| 548 |
+
"loss": 3.1469,
|
| 549 |
+
"step": 84000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 4.57,
|
| 553 |
+
"learning_rate": 0.00025326451385212633,
|
| 554 |
+
"loss": 3.1491,
|
| 555 |
+
"step": 85000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 4.62,
|
| 559 |
+
"learning_rate": 0.0002523822128110111,
|
| 560 |
+
"loss": 3.1484,
|
| 561 |
+
"step": 86000
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 4.68,
|
| 565 |
+
"learning_rate": 0.0002514999117698959,
|
| 566 |
+
"loss": 3.1452,
|
| 567 |
+
"step": 87000
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 4.73,
|
| 571 |
+
"learning_rate": 0.00025061849302982175,
|
| 572 |
+
"loss": 3.1454,
|
| 573 |
+
"step": 88000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 4.78,
|
| 577 |
+
"learning_rate": 0.00024973619198870653,
|
| 578 |
+
"loss": 3.1461,
|
| 579 |
+
"step": 89000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 4.84,
|
| 583 |
+
"learning_rate": 0.0002488547732486324,
|
| 584 |
+
"loss": 3.1483,
|
| 585 |
+
"step": 90000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 4.89,
|
| 589 |
+
"learning_rate": 0.0002479724722075172,
|
| 590 |
+
"loss": 3.1441,
|
| 591 |
+
"step": 91000
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.95,
|
| 595 |
+
"learning_rate": 0.00024709017116640196,
|
| 596 |
+
"loss": 3.1448,
|
| 597 |
+
"step": 92000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.0,
|
| 601 |
+
"learning_rate": 0.00024620787012528673,
|
| 602 |
+
"loss": 3.1443,
|
| 603 |
+
"step": 93000
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 5.0,
|
| 607 |
+
"eval_accuracy": 0.39941869370337724,
|
| 608 |
+
"eval_loss": 3.407740354537964,
|
| 609 |
+
"eval_runtime": 149.7411,
|
| 610 |
+
"eval_samples_per_second": 386.801,
|
| 611 |
+
"eval_steps_per_second": 6.044,
|
| 612 |
+
"step": 93005
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 5.05,
|
| 616 |
+
"learning_rate": 0.0002453264513852126,
|
| 617 |
+
"loss": 3.0734,
|
| 618 |
+
"step": 94000
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.11,
|
| 622 |
+
"learning_rate": 0.0002444441503440974,
|
| 623 |
+
"loss": 3.0793,
|
| 624 |
+
"step": 95000
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 5.16,
|
| 628 |
+
"learning_rate": 0.00024356273160402328,
|
| 629 |
+
"loss": 3.0807,
|
| 630 |
+
"step": 96000
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 5.21,
|
| 634 |
+
"learning_rate": 0.00024268043056290805,
|
| 635 |
+
"loss": 3.0863,
|
| 636 |
+
"step": 97000
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 5.27,
|
| 640 |
+
"learning_rate": 0.0002417981295217928,
|
| 641 |
+
"loss": 3.0939,
|
| 642 |
+
"step": 98000
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 5.32,
|
| 646 |
+
"learning_rate": 0.0002409167107817187,
|
| 647 |
+
"loss": 3.0949,
|
| 648 |
+
"step": 99000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 5.38,
|
| 652 |
+
"learning_rate": 0.00024003440974060345,
|
| 653 |
+
"loss": 3.0931,
|
| 654 |
+
"step": 100000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 5.43,
|
| 658 |
+
"learning_rate": 0.00023915299100052938,
|
| 659 |
+
"loss": 3.0941,
|
| 660 |
+
"step": 101000
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 5.48,
|
| 664 |
+
"learning_rate": 0.00023827068995941413,
|
| 665 |
+
"loss": 3.0931,
|
| 666 |
+
"step": 102000
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 5.54,
|
| 670 |
+
"learning_rate": 0.0002373883889182989,
|
| 671 |
+
"loss": 3.094,
|
| 672 |
+
"step": 103000
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 5.59,
|
| 676 |
+
"learning_rate": 0.00023650697017822477,
|
| 677 |
+
"loss": 3.0968,
|
| 678 |
+
"step": 104000
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 5.64,
|
| 682 |
+
"learning_rate": 0.00023562466913710955,
|
| 683 |
+
"loss": 3.098,
|
| 684 |
+
"step": 105000
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 5.7,
|
| 688 |
+
"learning_rate": 0.00023474325039703542,
|
| 689 |
+
"loss": 3.0946,
|
| 690 |
+
"step": 106000
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 5.75,
|
| 694 |
+
"learning_rate": 0.00023386094935592023,
|
| 695 |
+
"loss": 3.0951,
|
| 696 |
+
"step": 107000
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 5.81,
|
| 700 |
+
"learning_rate": 0.00023297953061584612,
|
| 701 |
+
"loss": 3.0984,
|
| 702 |
+
"step": 108000
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 5.86,
|
| 706 |
+
"learning_rate": 0.00023209722957473087,
|
| 707 |
+
"loss": 3.094,
|
| 708 |
+
"step": 109000
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 5.91,
|
| 712 |
+
"learning_rate": 0.00023121492853361565,
|
| 713 |
+
"loss": 3.099,
|
| 714 |
+
"step": 110000
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 5.97,
|
| 718 |
+
"learning_rate": 0.00023033350979354152,
|
| 719 |
+
"loss": 3.0973,
|
| 720 |
+
"step": 111000
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 6.0,
|
| 724 |
+
"eval_accuracy": 0.402754033471634,
|
| 725 |
+
"eval_loss": 3.372433662414551,
|
| 726 |
+
"eval_runtime": 149.8825,
|
| 727 |
+
"eval_samples_per_second": 386.436,
|
| 728 |
+
"eval_steps_per_second": 6.038,
|
| 729 |
+
"step": 111606
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 6.02,
|
| 733 |
+
"learning_rate": 0.0002294512087524263,
|
| 734 |
+
"loss": 3.0661,
|
| 735 |
+
"step": 112000
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 6.07,
|
| 739 |
+
"learning_rate": 0.0002285697900123522,
|
| 740 |
+
"loss": 3.0288,
|
| 741 |
+
"step": 113000
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 6.13,
|
| 745 |
+
"learning_rate": 0.00022768748897123697,
|
| 746 |
+
"loss": 3.0356,
|
| 747 |
+
"step": 114000
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 6.18,
|
| 751 |
+
"learning_rate": 0.00022680518793012175,
|
| 752 |
+
"loss": 3.0368,
|
| 753 |
+
"step": 115000
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 6.24,
|
| 757 |
+
"learning_rate": 0.0002259228868890065,
|
| 758 |
+
"loss": 3.0423,
|
| 759 |
+
"step": 116000
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 6.29,
|
| 763 |
+
"learning_rate": 0.0002250414681489324,
|
| 764 |
+
"loss": 3.0489,
|
| 765 |
+
"step": 117000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 6.34,
|
| 769 |
+
"learning_rate": 0.00022415916710781717,
|
| 770 |
+
"loss": 3.0487,
|
| 771 |
+
"step": 118000
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 6.4,
|
| 775 |
+
"learning_rate": 0.00022327774836774307,
|
| 776 |
+
"loss": 3.0484,
|
| 777 |
+
"step": 119000
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 6.45,
|
| 781 |
+
"learning_rate": 0.00022239544732662785,
|
| 782 |
+
"loss": 3.0518,
|
| 783 |
+
"step": 120000
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 6.51,
|
| 787 |
+
"learning_rate": 0.0002215131462855126,
|
| 788 |
+
"loss": 3.0504,
|
| 789 |
+
"step": 121000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 6.56,
|
| 793 |
+
"learning_rate": 0.00022063084524439738,
|
| 794 |
+
"loss": 3.0549,
|
| 795 |
+
"step": 122000
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 6.61,
|
| 799 |
+
"learning_rate": 0.00021974942650432325,
|
| 800 |
+
"loss": 3.0526,
|
| 801 |
+
"step": 123000
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 6.67,
|
| 805 |
+
"learning_rate": 0.00021886889006529023,
|
| 806 |
+
"loss": 3.0552,
|
| 807 |
+
"step": 124000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 6.72,
|
| 811 |
+
"learning_rate": 0.00021798658902417504,
|
| 812 |
+
"loss": 3.0549,
|
| 813 |
+
"step": 125000
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 6.77,
|
| 817 |
+
"learning_rate": 0.00021710428798305982,
|
| 818 |
+
"loss": 3.0557,
|
| 819 |
+
"step": 126000
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 6.83,
|
| 823 |
+
"learning_rate": 0.00021622198694194457,
|
| 824 |
+
"loss": 3.0567,
|
| 825 |
+
"step": 127000
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 6.88,
|
| 829 |
+
"learning_rate": 0.00021534056820187046,
|
| 830 |
+
"loss": 3.0581,
|
| 831 |
+
"step": 128000
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 6.94,
|
| 835 |
+
"learning_rate": 0.00021445826716075521,
|
| 836 |
+
"loss": 3.0577,
|
| 837 |
+
"step": 129000
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 6.99,
|
| 841 |
+
"learning_rate": 0.0002135768484206811,
|
| 842 |
+
"loss": 3.0617,
|
| 843 |
+
"step": 130000
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 7.0,
|
| 847 |
+
"eval_accuracy": 0.40619524851129696,
|
| 848 |
+
"eval_loss": 3.3562216758728027,
|
| 849 |
+
"eval_runtime": 149.8137,
|
| 850 |
+
"eval_samples_per_second": 386.613,
|
| 851 |
+
"eval_steps_per_second": 6.041,
|
| 852 |
+
"step": 130207
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 7.04,
|
| 856 |
+
"learning_rate": 0.00021269454737956592,
|
| 857 |
+
"loss": 2.9986,
|
| 858 |
+
"step": 131000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 7.1,
|
| 862 |
+
"learning_rate": 0.00021181224633845067,
|
| 863 |
+
"loss": 2.9938,
|
| 864 |
+
"step": 132000
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 7.15,
|
| 868 |
+
"learning_rate": 0.00021092994529733544,
|
| 869 |
+
"loss": 2.997,
|
| 870 |
+
"step": 133000
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 7.2,
|
| 874 |
+
"learning_rate": 0.00021004764425622022,
|
| 875 |
+
"loss": 3.0043,
|
| 876 |
+
"step": 134000
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 7.26,
|
| 880 |
+
"learning_rate": 0.0002091662255161461,
|
| 881 |
+
"loss": 3.0065,
|
| 882 |
+
"step": 135000
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 7.31,
|
| 886 |
+
"learning_rate": 0.00020828392447503087,
|
| 887 |
+
"loss": 3.0121,
|
| 888 |
+
"step": 136000
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 7.37,
|
| 892 |
+
"learning_rate": 0.00020740162343391562,
|
| 893 |
+
"loss": 3.011,
|
| 894 |
+
"step": 137000
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 7.42,
|
| 898 |
+
"learning_rate": 0.00020652020469384152,
|
| 899 |
+
"loss": 3.0123,
|
| 900 |
+
"step": 138000
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 7.47,
|
| 904 |
+
"learning_rate": 0.00020563790365272627,
|
| 905 |
+
"loss": 3.014,
|
| 906 |
+
"step": 139000
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 7.53,
|
| 910 |
+
"learning_rate": 0.0002047564849126522,
|
| 911 |
+
"loss": 3.0197,
|
| 912 |
+
"step": 140000
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 7.58,
|
| 916 |
+
"learning_rate": 0.00020387418387153697,
|
| 917 |
+
"loss": 3.019,
|
| 918 |
+
"step": 141000
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 7.63,
|
| 922 |
+
"learning_rate": 0.00020299188283042172,
|
| 923 |
+
"loss": 3.0203,
|
| 924 |
+
"step": 142000
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 7.69,
|
| 928 |
+
"learning_rate": 0.00020211046409034761,
|
| 929 |
+
"loss": 3.0176,
|
| 930 |
+
"step": 143000
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 7.74,
|
| 934 |
+
"learning_rate": 0.00020122816304923236,
|
| 935 |
+
"loss": 3.0197,
|
| 936 |
+
"step": 144000
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 7.8,
|
| 940 |
+
"learning_rate": 0.0002003467443091583,
|
| 941 |
+
"loss": 3.021,
|
| 942 |
+
"step": 145000
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 7.85,
|
| 946 |
+
"learning_rate": 0.00019946444326804304,
|
| 947 |
+
"loss": 3.028,
|
| 948 |
+
"step": 146000
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 7.9,
|
| 952 |
+
"learning_rate": 0.00019858302452796894,
|
| 953 |
+
"loss": 3.0219,
|
| 954 |
+
"step": 147000
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 7.96,
|
| 958 |
+
"learning_rate": 0.00019770072348685369,
|
| 959 |
+
"loss": 3.0252,
|
| 960 |
+
"step": 148000
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 8.0,
|
| 964 |
+
"eval_accuracy": 0.40585283337274547,
|
| 965 |
+
"eval_loss": 3.364830255508423,
|
| 966 |
+
"eval_runtime": 150.706,
|
| 967 |
+
"eval_samples_per_second": 384.325,
|
| 968 |
+
"eval_steps_per_second": 6.005,
|
| 969 |
+
"step": 148808
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 8.01,
|
| 973 |
+
"learning_rate": 0.00019681842244573846,
|
| 974 |
+
"loss": 3.0107,
|
| 975 |
+
"step": 149000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 8.06,
|
| 979 |
+
"learning_rate": 0.00019593700370566433,
|
| 980 |
+
"loss": 2.9562,
|
| 981 |
+
"step": 150000
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 8.12,
|
| 985 |
+
"learning_rate": 0.0001950547026645491,
|
| 986 |
+
"loss": 2.9643,
|
| 987 |
+
"step": 151000
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 8.17,
|
| 991 |
+
"learning_rate": 0.000194173283924475,
|
| 992 |
+
"loss": 2.9706,
|
| 993 |
+
"step": 152000
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 8.23,
|
| 997 |
+
"learning_rate": 0.00019329098288335979,
|
| 998 |
+
"loss": 2.974,
|
| 999 |
+
"step": 153000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 8.28,
|
| 1003 |
+
"learning_rate": 0.00019240868184224456,
|
| 1004 |
+
"loss": 2.9751,
|
| 1005 |
+
"step": 154000
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 8.33,
|
| 1009 |
+
"learning_rate": 0.00019152638080112934,
|
| 1010 |
+
"loss": 2.9802,
|
| 1011 |
+
"step": 155000
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 8.39,
|
| 1015 |
+
"learning_rate": 0.0001906449620610552,
|
| 1016 |
+
"loss": 2.9811,
|
| 1017 |
+
"step": 156000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 8.44,
|
| 1021 |
+
"learning_rate": 0.00018976354332098108,
|
| 1022 |
+
"loss": 2.9821,
|
| 1023 |
+
"step": 157000
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 8.49,
|
| 1027 |
+
"learning_rate": 0.000188882124580907,
|
| 1028 |
+
"loss": 2.9879,
|
| 1029 |
+
"step": 158000
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 8.55,
|
| 1033 |
+
"learning_rate": 0.00018799982353979175,
|
| 1034 |
+
"loss": 2.9908,
|
| 1035 |
+
"step": 159000
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 8.6,
|
| 1039 |
+
"learning_rate": 0.00018711840479971765,
|
| 1040 |
+
"loss": 2.9866,
|
| 1041 |
+
"step": 160000
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 8.66,
|
| 1045 |
+
"learning_rate": 0.0001862361037586024,
|
| 1046 |
+
"loss": 2.9891,
|
| 1047 |
+
"step": 161000
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 8.71,
|
| 1051 |
+
"learning_rate": 0.00018535380271748718,
|
| 1052 |
+
"loss": 2.9896,
|
| 1053 |
+
"step": 162000
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 8.76,
|
| 1057 |
+
"learning_rate": 0.00018447150167637196,
|
| 1058 |
+
"loss": 2.9915,
|
| 1059 |
+
"step": 163000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 8.82,
|
| 1063 |
+
"learning_rate": 0.00018358920063525676,
|
| 1064 |
+
"loss": 2.9924,
|
| 1065 |
+
"step": 164000
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 8.87,
|
| 1069 |
+
"learning_rate": 0.0001827068995941415,
|
| 1070 |
+
"loss": 2.9922,
|
| 1071 |
+
"step": 165000
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 8.92,
|
| 1075 |
+
"learning_rate": 0.0001818254808540674,
|
| 1076 |
+
"loss": 2.9917,
|
| 1077 |
+
"step": 166000
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 8.98,
|
| 1081 |
+
"learning_rate": 0.00018094317981295216,
|
| 1082 |
+
"loss": 2.994,
|
| 1083 |
+
"step": 167000
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 9.0,
|
| 1087 |
+
"eval_accuracy": 0.407066838467657,
|
| 1088 |
+
"eval_loss": 3.3581929206848145,
|
| 1089 |
+
"eval_runtime": 150.1138,
|
| 1090 |
+
"eval_samples_per_second": 385.841,
|
| 1091 |
+
"eval_steps_per_second": 6.029,
|
| 1092 |
+
"step": 167409
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 9.03,
|
| 1096 |
+
"learning_rate": 0.00018006087877183694,
|
| 1097 |
+
"loss": 2.9573,
|
| 1098 |
+
"step": 168000
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 9.09,
|
| 1102 |
+
"learning_rate": 0.0001791785777307217,
|
| 1103 |
+
"loss": 2.9322,
|
| 1104 |
+
"step": 169000
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 9.14,
|
| 1108 |
+
"learning_rate": 0.00017829715899064758,
|
| 1109 |
+
"loss": 2.9376,
|
| 1110 |
+
"step": 170000
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 9.19,
|
| 1114 |
+
"learning_rate": 0.0001774148579495324,
|
| 1115 |
+
"loss": 2.9392,
|
| 1116 |
+
"step": 171000
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 9.25,
|
| 1120 |
+
"learning_rate": 0.0001765325569084171,
|
| 1121 |
+
"loss": 2.9452,
|
| 1122 |
+
"step": 172000
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 9.3,
|
| 1126 |
+
"learning_rate": 0.00017565025586730192,
|
| 1127 |
+
"loss": 2.9497,
|
| 1128 |
+
"step": 173000
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 9.35,
|
| 1132 |
+
"learning_rate": 0.00017476883712722778,
|
| 1133 |
+
"loss": 2.9484,
|
| 1134 |
+
"step": 174000
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 9.41,
|
| 1138 |
+
"learning_rate": 0.00017388653608611256,
|
| 1139 |
+
"loss": 2.9534,
|
| 1140 |
+
"step": 175000
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 9.46,
|
| 1144 |
+
"learning_rate": 0.00017300423504499734,
|
| 1145 |
+
"loss": 2.9566,
|
| 1146 |
+
"step": 176000
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 9.52,
|
| 1150 |
+
"learning_rate": 0.00017212193400388212,
|
| 1151 |
+
"loss": 2.9573,
|
| 1152 |
+
"step": 177000
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 9.57,
|
| 1156 |
+
"learning_rate": 0.00017124139756484913,
|
| 1157 |
+
"loss": 2.9621,
|
| 1158 |
+
"step": 178000
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 9.62,
|
| 1162 |
+
"learning_rate": 0.00017035909652373388,
|
| 1163 |
+
"loss": 2.9643,
|
| 1164 |
+
"step": 179000
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 9.68,
|
| 1168 |
+
"learning_rate": 0.00016947679548261866,
|
| 1169 |
+
"loss": 2.9616,
|
| 1170 |
+
"step": 180000
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 9.73,
|
| 1174 |
+
"learning_rate": 0.00016859537674254453,
|
| 1175 |
+
"loss": 2.9635,
|
| 1176 |
+
"step": 181000
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 9.78,
|
| 1180 |
+
"learning_rate": 0.0001677130757014293,
|
| 1181 |
+
"loss": 2.9642,
|
| 1182 |
+
"step": 182000
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 9.84,
|
| 1186 |
+
"learning_rate": 0.00016683077466031409,
|
| 1187 |
+
"loss": 2.9676,
|
| 1188 |
+
"step": 183000
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 9.89,
|
| 1192 |
+
"learning_rate": 0.00016594847361919884,
|
| 1193 |
+
"loss": 2.9648,
|
| 1194 |
+
"step": 184000
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 9.95,
|
| 1198 |
+
"learning_rate": 0.00016506881948120697,
|
| 1199 |
+
"loss": 2.9712,
|
| 1200 |
+
"step": 185000
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 10.0,
|
| 1204 |
+
"learning_rate": 0.00016418651844009175,
|
| 1205 |
+
"loss": 2.9693,
|
| 1206 |
+
"step": 186000
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 10.0,
|
| 1210 |
+
"eval_accuracy": 0.407509351419911,
|
| 1211 |
+
"eval_loss": 3.36879825592041,
|
| 1212 |
+
"eval_runtime": 150.0406,
|
| 1213 |
+
"eval_samples_per_second": 386.029,
|
| 1214 |
+
"eval_steps_per_second": 6.032,
|
| 1215 |
+
"step": 186010
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 10.05,
|
| 1219 |
+
"learning_rate": 0.0001633042173989765,
|
| 1220 |
+
"loss": 2.9024,
|
| 1221 |
+
"step": 187000
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 10.11,
|
| 1225 |
+
"learning_rate": 0.00016242191635786128,
|
| 1226 |
+
"loss": 2.9062,
|
| 1227 |
+
"step": 188000
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 10.16,
|
| 1231 |
+
"learning_rate": 0.00016153961531674605,
|
| 1232 |
+
"loss": 2.9137,
|
| 1233 |
+
"step": 189000
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 10.21,
|
| 1237 |
+
"learning_rate": 0.00016065731427563083,
|
| 1238 |
+
"loss": 2.9198,
|
| 1239 |
+
"step": 190000
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 10.27,
|
| 1243 |
+
"learning_rate": 0.00015977589553555673,
|
| 1244 |
+
"loss": 2.9199,
|
| 1245 |
+
"step": 191000
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 10.32,
|
| 1249 |
+
"learning_rate": 0.0001588935944944415,
|
| 1250 |
+
"loss": 2.9239,
|
| 1251 |
+
"step": 192000
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 10.38,
|
| 1255 |
+
"learning_rate": 0.00015801129345332626,
|
| 1256 |
+
"loss": 2.9313,
|
| 1257 |
+
"step": 193000
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 10.43,
|
| 1261 |
+
"learning_rate": 0.00015712899241221103,
|
| 1262 |
+
"loss": 2.9295,
|
| 1263 |
+
"step": 194000
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 10.48,
|
| 1267 |
+
"learning_rate": 0.0001562475736721369,
|
| 1268 |
+
"loss": 2.9337,
|
| 1269 |
+
"step": 195000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 10.54,
|
| 1273 |
+
"learning_rate": 0.0001553661549320628,
|
| 1274 |
+
"loss": 2.9304,
|
| 1275 |
+
"step": 196000
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 10.59,
|
| 1279 |
+
"learning_rate": 0.00015448385389094755,
|
| 1280 |
+
"loss": 2.9378,
|
| 1281 |
+
"step": 197000
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 10.64,
|
| 1285 |
+
"learning_rate": 0.00015360155284983236,
|
| 1286 |
+
"loss": 2.9367,
|
| 1287 |
+
"step": 198000
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 10.7,
|
| 1291 |
+
"learning_rate": 0.00015272013410975825,
|
| 1292 |
+
"loss": 2.9404,
|
| 1293 |
+
"step": 199000
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 10.75,
|
| 1297 |
+
"learning_rate": 0.000151837833068643,
|
| 1298 |
+
"loss": 2.9415,
|
| 1299 |
+
"step": 200000
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 10.81,
|
| 1303 |
+
"learning_rate": 0.00015095553202752778,
|
| 1304 |
+
"loss": 2.9428,
|
| 1305 |
+
"step": 201000
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 10.86,
|
| 1309 |
+
"learning_rate": 0.00015007411328745365,
|
| 1310 |
+
"loss": 2.9405,
|
| 1311 |
+
"step": 202000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 10.91,
|
| 1315 |
+
"learning_rate": 0.00014919181224633843,
|
| 1316 |
+
"loss": 2.943,
|
| 1317 |
+
"step": 203000
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 10.97,
|
| 1321 |
+
"learning_rate": 0.00014831039350626432,
|
| 1322 |
+
"loss": 2.9383,
|
| 1323 |
+
"step": 204000
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 11.0,
|
| 1327 |
+
"eval_accuracy": 0.4091652648494099,
|
| 1328 |
+
"eval_loss": 3.3513429164886475,
|
| 1329 |
+
"eval_runtime": 149.8996,
|
| 1330 |
+
"eval_samples_per_second": 386.392,
|
| 1331 |
+
"eval_steps_per_second": 6.037,
|
| 1332 |
+
"step": 204611
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 11.02,
|
| 1336 |
+
"learning_rate": 0.0001474280924651491,
|
| 1337 |
+
"loss": 2.9199,
|
| 1338 |
+
"step": 205000
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 11.07,
|
| 1342 |
+
"learning_rate": 0.00014654579142403385,
|
| 1343 |
+
"loss": 2.8839,
|
| 1344 |
+
"step": 206000
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 11.13,
|
| 1348 |
+
"learning_rate": 0.00014566349038291863,
|
| 1349 |
+
"loss": 2.8913,
|
| 1350 |
+
"step": 207000
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 11.18,
|
| 1354 |
+
"learning_rate": 0.00014478207164284453,
|
| 1355 |
+
"loss": 2.8901,
|
| 1356 |
+
"step": 208000
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 11.24,
|
| 1360 |
+
"learning_rate": 0.0001439006529027704,
|
| 1361 |
+
"loss": 2.8959,
|
| 1362 |
+
"step": 209000
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 11.29,
|
| 1366 |
+
"learning_rate": 0.0001430183518616552,
|
| 1367 |
+
"loss": 2.8992,
|
| 1368 |
+
"step": 210000
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 11.34,
|
| 1372 |
+
"learning_rate": 0.00014213605082053995,
|
| 1373 |
+
"loss": 2.9022,
|
| 1374 |
+
"step": 211000
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 11.4,
|
| 1378 |
+
"learning_rate": 0.00014125374977942473,
|
| 1379 |
+
"loss": 2.9044,
|
| 1380 |
+
"step": 212000
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 11.45,
|
| 1384 |
+
"learning_rate": 0.0001403714487383095,
|
| 1385 |
+
"loss": 2.9076,
|
| 1386 |
+
"step": 213000
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 11.5,
|
| 1390 |
+
"learning_rate": 0.0001394900299982354,
|
| 1391 |
+
"loss": 2.9106,
|
| 1392 |
+
"step": 214000
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 11.56,
|
| 1396 |
+
"learning_rate": 0.00013860772895712015,
|
| 1397 |
+
"loss": 2.9124,
|
| 1398 |
+
"step": 215000
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 11.61,
|
| 1402 |
+
"learning_rate": 0.00013772631021704605,
|
| 1403 |
+
"loss": 2.9112,
|
| 1404 |
+
"step": 216000
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 11.67,
|
| 1408 |
+
"learning_rate": 0.00013684489147697192,
|
| 1409 |
+
"loss": 2.9151,
|
| 1410 |
+
"step": 217000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 11.72,
|
| 1414 |
+
"learning_rate": 0.0001359625904358567,
|
| 1415 |
+
"loss": 2.916,
|
| 1416 |
+
"step": 218000
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 11.77,
|
| 1420 |
+
"learning_rate": 0.0001350811716957826,
|
| 1421 |
+
"loss": 2.9181,
|
| 1422 |
+
"step": 219000
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 11.83,
|
| 1426 |
+
"learning_rate": 0.00013419887065466737,
|
| 1427 |
+
"loss": 2.9202,
|
| 1428 |
+
"step": 220000
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 11.88,
|
| 1432 |
+
"learning_rate": 0.00013331656961355212,
|
| 1433 |
+
"loss": 2.9172,
|
| 1434 |
+
"step": 221000
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 11.93,
|
| 1438 |
+
"learning_rate": 0.0001324342685724369,
|
| 1439 |
+
"loss": 2.9191,
|
| 1440 |
+
"step": 222000
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 11.99,
|
| 1444 |
+
"learning_rate": 0.00013155196753132168,
|
| 1445 |
+
"loss": 2.9188,
|
| 1446 |
+
"step": 223000
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 12.0,
|
| 1450 |
+
"eval_accuracy": 0.4086410613324161,
|
| 1451 |
+
"eval_loss": 3.3659284114837646,
|
| 1452 |
+
"eval_runtime": 149.7389,
|
| 1453 |
+
"eval_samples_per_second": 386.807,
|
| 1454 |
+
"eval_steps_per_second": 6.044,
|
| 1455 |
+
"step": 223212
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 12.04,
|
| 1459 |
+
"learning_rate": 0.00013067054879124757,
|
| 1460 |
+
"loss": 2.8731,
|
| 1461 |
+
"step": 224000
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 12.1,
|
| 1465 |
+
"learning_rate": 0.00012978913005117344,
|
| 1466 |
+
"loss": 2.8633,
|
| 1467 |
+
"step": 225000
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 12.15,
|
| 1471 |
+
"learning_rate": 0.00012890682901005822,
|
| 1472 |
+
"loss": 2.8684,
|
| 1473 |
+
"step": 226000
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 12.2,
|
| 1477 |
+
"learning_rate": 0.000128024527968943,
|
| 1478 |
+
"loss": 2.8681,
|
| 1479 |
+
"step": 227000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 12.26,
|
| 1483 |
+
"learning_rate": 0.00012714222692782778,
|
| 1484 |
+
"loss": 2.8766,
|
| 1485 |
+
"step": 228000
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 12.31,
|
| 1489 |
+
"learning_rate": 0.00012625992588671253,
|
| 1490 |
+
"loss": 2.8833,
|
| 1491 |
+
"step": 229000
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 12.36,
|
| 1495 |
+
"learning_rate": 0.0001253776248455973,
|
| 1496 |
+
"loss": 2.8843,
|
| 1497 |
+
"step": 230000
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 12.42,
|
| 1501 |
+
"learning_rate": 0.0001244962061055232,
|
| 1502 |
+
"loss": 2.8805,
|
| 1503 |
+
"step": 231000
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 12.47,
|
| 1507 |
+
"learning_rate": 0.00012361478736544907,
|
| 1508 |
+
"loss": 2.8888,
|
| 1509 |
+
"step": 232000
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 12.53,
|
| 1513 |
+
"learning_rate": 0.00012273248632433385,
|
| 1514 |
+
"loss": 2.8911,
|
| 1515 |
+
"step": 233000
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 12.58,
|
| 1519 |
+
"learning_rate": 0.00012185018528321863,
|
| 1520 |
+
"loss": 2.8895,
|
| 1521 |
+
"step": 234000
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 12.63,
|
| 1525 |
+
"learning_rate": 0.00012096876654314452,
|
| 1526 |
+
"loss": 2.8906,
|
| 1527 |
+
"step": 235000
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 12.69,
|
| 1531 |
+
"learning_rate": 0.00012008646550202929,
|
| 1532 |
+
"loss": 2.8921,
|
| 1533 |
+
"step": 236000
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 12.74,
|
| 1537 |
+
"learning_rate": 0.00011920504676195517,
|
| 1538 |
+
"loss": 2.8938,
|
| 1539 |
+
"step": 237000
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 12.8,
|
| 1543 |
+
"learning_rate": 0.00011832274572083993,
|
| 1544 |
+
"loss": 2.894,
|
| 1545 |
+
"step": 238000
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 12.85,
|
| 1549 |
+
"learning_rate": 0.00011744044467972471,
|
| 1550 |
+
"loss": 2.8977,
|
| 1551 |
+
"step": 239000
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 12.9,
|
| 1555 |
+
"learning_rate": 0.00011655814363860949,
|
| 1556 |
+
"loss": 2.8978,
|
| 1557 |
+
"step": 240000
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 12.96,
|
| 1561 |
+
"learning_rate": 0.00011567672489853537,
|
| 1562 |
+
"loss": 2.8978,
|
| 1563 |
+
"step": 241000
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 13.0,
|
| 1567 |
+
"eval_accuracy": 0.4096840939871445,
|
| 1568 |
+
"eval_loss": 3.3580610752105713,
|
| 1569 |
+
"eval_runtime": 149.9409,
|
| 1570 |
+
"eval_samples_per_second": 386.285,
|
| 1571 |
+
"eval_steps_per_second": 6.036,
|
| 1572 |
+
"step": 241813
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 13.01,
|
| 1576 |
+
"learning_rate": 0.00011479442385742015,
|
| 1577 |
+
"loss": 2.8892,
|
| 1578 |
+
"step": 242000
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 13.06,
|
| 1582 |
+
"learning_rate": 0.00011391212281630491,
|
| 1583 |
+
"loss": 2.839,
|
| 1584 |
+
"step": 243000
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 13.12,
|
| 1588 |
+
"learning_rate": 0.0001130307040762308,
|
| 1589 |
+
"loss": 2.8503,
|
| 1590 |
+
"step": 244000
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 13.17,
|
| 1594 |
+
"learning_rate": 0.00011214840303511556,
|
| 1595 |
+
"loss": 2.8516,
|
| 1596 |
+
"step": 245000
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 13.23,
|
| 1600 |
+
"learning_rate": 0.00011126610199400035,
|
| 1601 |
+
"loss": 2.8554,
|
| 1602 |
+
"step": 246000
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 13.28,
|
| 1606 |
+
"learning_rate": 0.00011038468325392623,
|
| 1607 |
+
"loss": 2.8568,
|
| 1608 |
+
"step": 247000
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 13.33,
|
| 1612 |
+
"learning_rate": 0.000109502382212811,
|
| 1613 |
+
"loss": 2.862,
|
| 1614 |
+
"step": 248000
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 13.39,
|
| 1618 |
+
"learning_rate": 0.00010862008117169576,
|
| 1619 |
+
"loss": 2.8615,
|
| 1620 |
+
"step": 249000
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 13.44,
|
| 1624 |
+
"learning_rate": 0.00010773778013058055,
|
| 1625 |
+
"loss": 2.8704,
|
| 1626 |
+
"step": 250000
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 13.49,
|
| 1630 |
+
"learning_rate": 0.00010685636139050644,
|
| 1631 |
+
"loss": 2.8627,
|
| 1632 |
+
"step": 251000
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 13.55,
|
| 1636 |
+
"learning_rate": 0.00010597494265043232,
|
| 1637 |
+
"loss": 2.8684,
|
| 1638 |
+
"step": 252000
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 13.6,
|
| 1642 |
+
"learning_rate": 0.0001050926416093171,
|
| 1643 |
+
"loss": 2.8714,
|
| 1644 |
+
"step": 253000
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 13.66,
|
| 1648 |
+
"learning_rate": 0.00010421034056820186,
|
| 1649 |
+
"loss": 2.8749,
|
| 1650 |
+
"step": 254000
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 13.71,
|
| 1654 |
+
"learning_rate": 0.00010332892182812776,
|
| 1655 |
+
"loss": 2.87,
|
| 1656 |
+
"step": 255000
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 13.76,
|
| 1660 |
+
"learning_rate": 0.00010244662078701252,
|
| 1661 |
+
"loss": 2.8721,
|
| 1662 |
+
"step": 256000
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 13.82,
|
| 1666 |
+
"learning_rate": 0.0001015643197458973,
|
| 1667 |
+
"loss": 2.8775,
|
| 1668 |
+
"step": 257000
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 13.87,
|
| 1672 |
+
"learning_rate": 0.00010068201870478206,
|
| 1673 |
+
"loss": 2.8779,
|
| 1674 |
+
"step": 258000
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 13.92,
|
| 1678 |
+
"learning_rate": 9.980148226574907e-05,
|
| 1679 |
+
"loss": 2.8696,
|
| 1680 |
+
"step": 259000
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 13.98,
|
| 1684 |
+
"learning_rate": 9.891918122463383e-05,
|
| 1685 |
+
"loss": 2.8784,
|
| 1686 |
+
"step": 260000
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 14.0,
|
| 1690 |
+
"eval_accuracy": 0.4103098060923963,
|
| 1691 |
+
"eval_loss": 3.3657193183898926,
|
| 1692 |
+
"eval_runtime": 154.6068,
|
| 1693 |
+
"eval_samples_per_second": 374.628,
|
| 1694 |
+
"eval_steps_per_second": 5.854,
|
| 1695 |
+
"step": 260414
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 14.03,
|
| 1699 |
+
"learning_rate": 9.803776248455973e-05,
|
| 1700 |
+
"loss": 2.8434,
|
| 1701 |
+
"step": 261000
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 14.09,
|
| 1705 |
+
"learning_rate": 9.715546144344449e-05,
|
| 1706 |
+
"loss": 2.824,
|
| 1707 |
+
"step": 262000
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 14.14,
|
| 1711 |
+
"learning_rate": 9.627404270337037e-05,
|
| 1712 |
+
"loss": 2.8286,
|
| 1713 |
+
"step": 263000
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 14.19,
|
| 1717 |
+
"learning_rate": 9.539174166225517e-05,
|
| 1718 |
+
"loss": 2.8351,
|
| 1719 |
+
"step": 264000
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 14.25,
|
| 1723 |
+
"learning_rate": 9.450944062113993e-05,
|
| 1724 |
+
"loss": 2.8403,
|
| 1725 |
+
"step": 265000
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 14.3,
|
| 1729 |
+
"learning_rate": 9.362713958002469e-05,
|
| 1730 |
+
"loss": 2.8419,
|
| 1731 |
+
"step": 266000
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 14.35,
|
| 1735 |
+
"learning_rate": 9.274483853890947e-05,
|
| 1736 |
+
"loss": 2.8428,
|
| 1737 |
+
"step": 267000
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 14.41,
|
| 1741 |
+
"learning_rate": 9.186341979883535e-05,
|
| 1742 |
+
"loss": 2.8455,
|
| 1743 |
+
"step": 268000
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 14.46,
|
| 1747 |
+
"learning_rate": 9.098200105876124e-05,
|
| 1748 |
+
"loss": 2.8471,
|
| 1749 |
+
"step": 269000
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 14.52,
|
| 1753 |
+
"learning_rate": 9.009970001764601e-05,
|
| 1754 |
+
"loss": 2.8474,
|
| 1755 |
+
"step": 270000
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 14.57,
|
| 1759 |
+
"learning_rate": 8.92182812775719e-05,
|
| 1760 |
+
"loss": 2.8482,
|
| 1761 |
+
"step": 271000
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 14.62,
|
| 1765 |
+
"learning_rate": 8.833598023645668e-05,
|
| 1766 |
+
"loss": 2.8493,
|
| 1767 |
+
"step": 272000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 14.68,
|
| 1771 |
+
"learning_rate": 8.745367919534144e-05,
|
| 1772 |
+
"loss": 2.8536,
|
| 1773 |
+
"step": 273000
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 14.73,
|
| 1777 |
+
"learning_rate": 8.65713781542262e-05,
|
| 1778 |
+
"loss": 2.853,
|
| 1779 |
+
"step": 274000
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 14.78,
|
| 1783 |
+
"learning_rate": 8.5689077113111e-05,
|
| 1784 |
+
"loss": 2.8526,
|
| 1785 |
+
"step": 275000
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 14.84,
|
| 1789 |
+
"learning_rate": 8.4808540674078e-05,
|
| 1790 |
+
"loss": 2.8543,
|
| 1791 |
+
"step": 276000
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 14.89,
|
| 1795 |
+
"learning_rate": 8.392623963296276e-05,
|
| 1796 |
+
"loss": 2.8549,
|
| 1797 |
+
"step": 277000
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 14.95,
|
| 1801 |
+
"learning_rate": 8.304393859184754e-05,
|
| 1802 |
+
"loss": 2.8575,
|
| 1803 |
+
"step": 278000
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 15.0,
|
| 1807 |
+
"learning_rate": 8.216251985177341e-05,
|
| 1808 |
+
"loss": 2.8592,
|
| 1809 |
+
"step": 279000
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 15.0,
|
| 1813 |
+
"eval_accuracy": 0.4102248237203603,
|
| 1814 |
+
"eval_loss": 3.3692827224731445,
|
| 1815 |
+
"eval_runtime": 155.1467,
|
| 1816 |
+
"eval_samples_per_second": 373.324,
|
| 1817 |
+
"eval_steps_per_second": 5.833,
|
| 1818 |
+
"step": 279015
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 15.05,
|
| 1822 |
+
"learning_rate": 8.12802188106582e-05,
|
| 1823 |
+
"loss": 2.806,
|
| 1824 |
+
"step": 280000
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 15.11,
|
| 1828 |
+
"learning_rate": 8.039880007058407e-05,
|
| 1829 |
+
"loss": 2.8122,
|
| 1830 |
+
"step": 281000
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 15.16,
|
| 1834 |
+
"learning_rate": 7.951649902946885e-05,
|
| 1835 |
+
"loss": 2.8176,
|
| 1836 |
+
"step": 282000
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 15.21,
|
| 1840 |
+
"learning_rate": 7.863419798835361e-05,
|
| 1841 |
+
"loss": 2.8156,
|
| 1842 |
+
"step": 283000
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 15.27,
|
| 1846 |
+
"learning_rate": 7.775277924827951e-05,
|
| 1847 |
+
"loss": 2.8164,
|
| 1848 |
+
"step": 284000
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 15.32,
|
| 1852 |
+
"learning_rate": 7.687047820716427e-05,
|
| 1853 |
+
"loss": 2.8215,
|
| 1854 |
+
"step": 285000
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 15.38,
|
| 1858 |
+
"learning_rate": 7.598905946709017e-05,
|
| 1859 |
+
"loss": 2.8236,
|
| 1860 |
+
"step": 286000
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 15.43,
|
| 1864 |
+
"learning_rate": 7.510675842597493e-05,
|
| 1865 |
+
"loss": 2.8233,
|
| 1866 |
+
"step": 287000
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 15.48,
|
| 1870 |
+
"learning_rate": 7.422533968590081e-05,
|
| 1871 |
+
"loss": 2.8257,
|
| 1872 |
+
"step": 288000
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 15.54,
|
| 1876 |
+
"learning_rate": 7.334303864478559e-05,
|
| 1877 |
+
"loss": 2.8313,
|
| 1878 |
+
"step": 289000
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 15.59,
|
| 1882 |
+
"learning_rate": 7.246161990471149e-05,
|
| 1883 |
+
"loss": 2.8306,
|
| 1884 |
+
"step": 290000
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 15.64,
|
| 1888 |
+
"learning_rate": 7.157931886359625e-05,
|
| 1889 |
+
"loss": 2.8365,
|
| 1890 |
+
"step": 291000
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 15.7,
|
| 1894 |
+
"learning_rate": 7.069701782248102e-05,
|
| 1895 |
+
"loss": 2.8369,
|
| 1896 |
+
"step": 292000
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 15.75,
|
| 1900 |
+
"learning_rate": 6.981559908240691e-05,
|
| 1901 |
+
"loss": 2.8358,
|
| 1902 |
+
"step": 293000
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 15.81,
|
| 1906 |
+
"learning_rate": 6.893329804129168e-05,
|
| 1907 |
+
"loss": 2.8321,
|
| 1908 |
+
"step": 294000
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 15.86,
|
| 1912 |
+
"learning_rate": 6.805187930121757e-05,
|
| 1913 |
+
"loss": 2.8366,
|
| 1914 |
+
"step": 295000
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 15.91,
|
| 1918 |
+
"learning_rate": 6.716957826010234e-05,
|
| 1919 |
+
"loss": 2.8378,
|
| 1920 |
+
"step": 296000
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 15.97,
|
| 1924 |
+
"learning_rate": 6.62872772189871e-05,
|
| 1925 |
+
"loss": 2.8415,
|
| 1926 |
+
"step": 297000
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 16.0,
|
| 1930 |
+
"eval_accuracy": 0.40921652299159444,
|
| 1931 |
+
"eval_loss": 3.386690616607666,
|
| 1932 |
+
"eval_runtime": 150.4052,
|
| 1933 |
+
"eval_samples_per_second": 385.093,
|
| 1934 |
+
"eval_steps_per_second": 6.017,
|
| 1935 |
+
"step": 297616
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 16.02,
|
| 1939 |
+
"learning_rate": 6.540497617787188e-05,
|
| 1940 |
+
"loss": 2.8232,
|
| 1941 |
+
"step": 298000
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 16.07,
|
| 1945 |
+
"learning_rate": 6.452267513675666e-05,
|
| 1946 |
+
"loss": 2.7984,
|
| 1947 |
+
"step": 299000
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 16.13,
|
| 1951 |
+
"learning_rate": 6.364213869772366e-05,
|
| 1952 |
+
"loss": 2.795,
|
| 1953 |
+
"step": 300000
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 16.18,
|
| 1957 |
+
"learning_rate": 6.275983765660842e-05,
|
| 1958 |
+
"loss": 2.7994,
|
| 1959 |
+
"step": 301000
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 16.24,
|
| 1963 |
+
"learning_rate": 6.187841891653431e-05,
|
| 1964 |
+
"loss": 2.8011,
|
| 1965 |
+
"step": 302000
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 16.29,
|
| 1969 |
+
"learning_rate": 6.0996117875419085e-05,
|
| 1970 |
+
"loss": 2.8058,
|
| 1971 |
+
"step": 303000
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 16.34,
|
| 1975 |
+
"learning_rate": 6.011469913534497e-05,
|
| 1976 |
+
"loss": 2.8093,
|
| 1977 |
+
"step": 304000
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 16.4,
|
| 1981 |
+
"learning_rate": 5.9232398094229746e-05,
|
| 1982 |
+
"loss": 2.8096,
|
| 1983 |
+
"step": 305000
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 16.45,
|
| 1987 |
+
"learning_rate": 5.8350979354155636e-05,
|
| 1988 |
+
"loss": 2.8079,
|
| 1989 |
+
"step": 306000
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 16.5,
|
| 1993 |
+
"learning_rate": 5.74686783130404e-05,
|
| 1994 |
+
"loss": 2.8131,
|
| 1995 |
+
"step": 307000
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 16.56,
|
| 1999 |
+
"learning_rate": 5.658637727192517e-05,
|
| 2000 |
+
"loss": 2.8123,
|
| 2001 |
+
"step": 308000
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 16.61,
|
| 2005 |
+
"learning_rate": 5.570495853185107e-05,
|
| 2006 |
+
"loss": 2.8134,
|
| 2007 |
+
"step": 309000
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 16.67,
|
| 2011 |
+
"learning_rate": 5.482265749073583e-05,
|
| 2012 |
+
"loss": 2.817,
|
| 2013 |
+
"step": 310000
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 16.72,
|
| 2017 |
+
"learning_rate": 5.39403564496206e-05,
|
| 2018 |
+
"loss": 2.8147,
|
| 2019 |
+
"step": 311000
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 16.77,
|
| 2023 |
+
"learning_rate": 5.305805540850538e-05,
|
| 2024 |
+
"loss": 2.8145,
|
| 2025 |
+
"step": 312000
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 16.83,
|
| 2029 |
+
"learning_rate": 5.217575436739015e-05,
|
| 2030 |
+
"loss": 2.8123,
|
| 2031 |
+
"step": 313000
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 16.88,
|
| 2035 |
+
"learning_rate": 5.129521792835715e-05,
|
| 2036 |
+
"loss": 2.8173,
|
| 2037 |
+
"step": 314000
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 16.93,
|
| 2041 |
+
"learning_rate": 5.041291688724192e-05,
|
| 2042 |
+
"loss": 2.8142,
|
| 2043 |
+
"step": 315000
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 16.99,
|
| 2047 |
+
"learning_rate": 4.95306158461267e-05,
|
| 2048 |
+
"loss": 2.8198,
|
| 2049 |
+
"step": 316000
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 17.0,
|
| 2053 |
+
"eval_accuracy": 0.41009368886643593,
|
| 2054 |
+
"eval_loss": 3.3789703845977783,
|
| 2055 |
+
"eval_runtime": 149.2821,
|
| 2056 |
+
"eval_samples_per_second": 387.99,
|
| 2057 |
+
"eval_steps_per_second": 6.062,
|
| 2058 |
+
"step": 316217
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 17.04,
|
| 2062 |
+
"learning_rate": 4.8649197106052584e-05,
|
| 2063 |
+
"loss": 2.7896,
|
| 2064 |
+
"step": 317000
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 17.1,
|
| 2068 |
+
"learning_rate": 4.7766896064937355e-05,
|
| 2069 |
+
"loss": 2.7827,
|
| 2070 |
+
"step": 318000
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 17.15,
|
| 2074 |
+
"learning_rate": 4.6884595023822125e-05,
|
| 2075 |
+
"loss": 2.7812,
|
| 2076 |
+
"step": 319000
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 17.2,
|
| 2080 |
+
"learning_rate": 4.600317628374801e-05,
|
| 2081 |
+
"loss": 2.7871,
|
| 2082 |
+
"step": 320000
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 17.26,
|
| 2086 |
+
"learning_rate": 4.512175754367389e-05,
|
| 2087 |
+
"loss": 2.7895,
|
| 2088 |
+
"step": 321000
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 17.31,
|
| 2092 |
+
"learning_rate": 4.423945650255867e-05,
|
| 2093 |
+
"loss": 2.7907,
|
| 2094 |
+
"step": 322000
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 17.36,
|
| 2098 |
+
"learning_rate": 4.335715546144344e-05,
|
| 2099 |
+
"loss": 2.788,
|
| 2100 |
+
"step": 323000
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 17.42,
|
| 2104 |
+
"learning_rate": 4.247485442032821e-05,
|
| 2105 |
+
"loss": 2.7862,
|
| 2106 |
+
"step": 324000
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 17.47,
|
| 2110 |
+
"learning_rate": 4.159255337921299e-05,
|
| 2111 |
+
"loss": 2.7948,
|
| 2112 |
+
"step": 325000
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 17.53,
|
| 2116 |
+
"learning_rate": 4.071113463913887e-05,
|
| 2117 |
+
"loss": 2.7939,
|
| 2118 |
+
"step": 326000
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 17.58,
|
| 2122 |
+
"learning_rate": 3.982883359802364e-05,
|
| 2123 |
+
"loss": 2.7953,
|
| 2124 |
+
"step": 327000
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 17.63,
|
| 2128 |
+
"learning_rate": 3.894653255690842e-05,
|
| 2129 |
+
"loss": 2.7983,
|
| 2130 |
+
"step": 328000
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 17.69,
|
| 2134 |
+
"learning_rate": 3.80651138168343e-05,
|
| 2135 |
+
"loss": 2.7977,
|
| 2136 |
+
"step": 329000
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 17.74,
|
| 2140 |
+
"learning_rate": 3.7182812775719074e-05,
|
| 2141 |
+
"loss": 2.7977,
|
| 2142 |
+
"step": 330000
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 17.79,
|
| 2146 |
+
"learning_rate": 3.6300511734603844e-05,
|
| 2147 |
+
"loss": 2.7985,
|
| 2148 |
+
"step": 331000
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 17.85,
|
| 2152 |
+
"learning_rate": 3.541909299452973e-05,
|
| 2153 |
+
"loss": 2.7956,
|
| 2154 |
+
"step": 332000
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 17.9,
|
| 2158 |
+
"learning_rate": 3.4536791953414505e-05,
|
| 2159 |
+
"loss": 2.8001,
|
| 2160 |
+
"step": 333000
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 17.96,
|
| 2164 |
+
"learning_rate": 3.3654490912299276e-05,
|
| 2165 |
+
"loss": 2.8013,
|
| 2166 |
+
"step": 334000
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 18.0,
|
| 2170 |
+
"eval_accuracy": 0.40988771578132727,
|
| 2171 |
+
"eval_loss": 3.392399311065674,
|
| 2172 |
+
"eval_runtime": 149.3924,
|
| 2173 |
+
"eval_samples_per_second": 387.704,
|
| 2174 |
+
"eval_steps_per_second": 6.058,
|
| 2175 |
+
"step": 334818
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 18.01,
|
| 2179 |
+
"learning_rate": 3.2772189871184046e-05,
|
| 2180 |
+
"loss": 2.795,
|
| 2181 |
+
"step": 335000
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 18.06,
|
| 2185 |
+
"learning_rate": 3.189077113110993e-05,
|
| 2186 |
+
"loss": 2.7694,
|
| 2187 |
+
"step": 336000
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 18.12,
|
| 2191 |
+
"learning_rate": 3.101023469207693e-05,
|
| 2192 |
+
"loss": 2.7677,
|
| 2193 |
+
"step": 337000
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 18.17,
|
| 2197 |
+
"learning_rate": 3.0127933650961703e-05,
|
| 2198 |
+
"loss": 2.7737,
|
| 2199 |
+
"step": 338000
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 18.22,
|
| 2203 |
+
"learning_rate": 2.9245632609846477e-05,
|
| 2204 |
+
"loss": 2.772,
|
| 2205 |
+
"step": 339000
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 18.28,
|
| 2209 |
+
"learning_rate": 2.836333156873125e-05,
|
| 2210 |
+
"loss": 2.7774,
|
| 2211 |
+
"step": 340000
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 18.33,
|
| 2215 |
+
"learning_rate": 2.748103052761602e-05,
|
| 2216 |
+
"loss": 2.7737,
|
| 2217 |
+
"step": 341000
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 18.39,
|
| 2221 |
+
"learning_rate": 2.6599611787541905e-05,
|
| 2222 |
+
"loss": 2.7789,
|
| 2223 |
+
"step": 342000
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 18.44,
|
| 2227 |
+
"learning_rate": 2.571731074642668e-05,
|
| 2228 |
+
"loss": 2.779,
|
| 2229 |
+
"step": 343000
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 18.49,
|
| 2233 |
+
"learning_rate": 2.483500970531145e-05,
|
| 2234 |
+
"loss": 2.7798,
|
| 2235 |
+
"step": 344000
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 18.55,
|
| 2239 |
+
"learning_rate": 2.3953590965237337e-05,
|
| 2240 |
+
"loss": 2.7792,
|
| 2241 |
+
"step": 345000
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 18.6,
|
| 2245 |
+
"learning_rate": 2.307128992412211e-05,
|
| 2246 |
+
"loss": 2.7788,
|
| 2247 |
+
"step": 346000
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 18.65,
|
| 2251 |
+
"learning_rate": 2.2189871184047998e-05,
|
| 2252 |
+
"loss": 2.7776,
|
| 2253 |
+
"step": 347000
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 18.71,
|
| 2257 |
+
"learning_rate": 2.1307570142932765e-05,
|
| 2258 |
+
"loss": 2.779,
|
| 2259 |
+
"step": 348000
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 18.76,
|
| 2263 |
+
"learning_rate": 2.042526910181754e-05,
|
| 2264 |
+
"loss": 2.7764,
|
| 2265 |
+
"step": 349000
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 18.82,
|
| 2269 |
+
"learning_rate": 1.9542968060702313e-05,
|
| 2270 |
+
"loss": 2.7811,
|
| 2271 |
+
"step": 350000
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 18.87,
|
| 2275 |
+
"learning_rate": 1.8661549320628196e-05,
|
| 2276 |
+
"loss": 2.7809,
|
| 2277 |
+
"step": 351000
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 18.92,
|
| 2281 |
+
"learning_rate": 1.7780130580554083e-05,
|
| 2282 |
+
"loss": 2.7794,
|
| 2283 |
+
"step": 352000
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 18.98,
|
| 2287 |
+
"learning_rate": 1.6897829539438854e-05,
|
| 2288 |
+
"loss": 2.7836,
|
| 2289 |
+
"step": 353000
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 19.0,
|
| 2293 |
+
"eval_accuracy": 0.4097668594277361,
|
| 2294 |
+
"eval_loss": 3.401383399963379,
|
| 2295 |
+
"eval_runtime": 149.562,
|
| 2296 |
+
"eval_samples_per_second": 387.264,
|
| 2297 |
+
"eval_steps_per_second": 6.051,
|
| 2298 |
+
"step": 353419
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 19.03,
|
| 2302 |
+
"learning_rate": 1.6017293100405857e-05,
|
| 2303 |
+
"loss": 2.7676,
|
| 2304 |
+
"step": 354000
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 19.08,
|
| 2308 |
+
"learning_rate": 1.5134992059290627e-05,
|
| 2309 |
+
"loss": 2.7617,
|
| 2310 |
+
"step": 355000
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 19.14,
|
| 2314 |
+
"learning_rate": 1.4252691018175401e-05,
|
| 2315 |
+
"loss": 2.7597,
|
| 2316 |
+
"step": 356000
|
| 2317 |
+
},
|
| 2318 |
+
{
|
| 2319 |
+
"epoch": 19.19,
|
| 2320 |
+
"learning_rate": 1.3371272278101286e-05,
|
| 2321 |
+
"loss": 2.7629,
|
| 2322 |
+
"step": 357000
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 19.25,
|
| 2326 |
+
"learning_rate": 1.2488971236986057e-05,
|
| 2327 |
+
"loss": 2.7652,
|
| 2328 |
+
"step": 358000
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 19.3,
|
| 2332 |
+
"learning_rate": 1.1606670195870831e-05,
|
| 2333 |
+
"loss": 2.7624,
|
| 2334 |
+
"step": 359000
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 19.35,
|
| 2338 |
+
"learning_rate": 1.0725251455796716e-05,
|
| 2339 |
+
"loss": 2.7646,
|
| 2340 |
+
"step": 360000
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 19.41,
|
| 2344 |
+
"learning_rate": 9.84295041468149e-06,
|
| 2345 |
+
"loss": 2.7645,
|
| 2346 |
+
"step": 361000
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 19.46,
|
| 2350 |
+
"learning_rate": 8.96064937356626e-06,
|
| 2351 |
+
"loss": 2.7625,
|
| 2352 |
+
"step": 362000
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 19.52,
|
| 2356 |
+
"learning_rate": 8.078348332451031e-06,
|
| 2357 |
+
"loss": 2.7621,
|
| 2358 |
+
"step": 363000
|
| 2359 |
+
},
|
| 2360 |
+
{
|
| 2361 |
+
"epoch": 19.57,
|
| 2362 |
+
"learning_rate": 7.196047291335803e-06,
|
| 2363 |
+
"loss": 2.7639,
|
| 2364 |
+
"step": 364000
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 19.62,
|
| 2368 |
+
"learning_rate": 6.3146285512616895e-06,
|
| 2369 |
+
"loss": 2.7626,
|
| 2370 |
+
"step": 365000
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 19.68,
|
| 2374 |
+
"learning_rate": 5.432327510146462e-06,
|
| 2375 |
+
"loss": 2.7637,
|
| 2376 |
+
"step": 366000
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 19.73,
|
| 2380 |
+
"learning_rate": 4.550026469031233e-06,
|
| 2381 |
+
"loss": 2.7605,
|
| 2382 |
+
"step": 367000
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 19.78,
|
| 2386 |
+
"learning_rate": 3.66860772895712e-06,
|
| 2387 |
+
"loss": 2.7683,
|
| 2388 |
+
"step": 368000
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 19.84,
|
| 2392 |
+
"learning_rate": 2.7863066878418916e-06,
|
| 2393 |
+
"loss": 2.7608,
|
| 2394 |
+
"step": 369000
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 19.89,
|
| 2398 |
+
"learning_rate": 1.904005646726663e-06,
|
| 2399 |
+
"loss": 2.765,
|
| 2400 |
+
"step": 370000
|
| 2401 |
+
},
|
| 2402 |
+
{
|
| 2403 |
+
"epoch": 19.95,
|
| 2404 |
+
"learning_rate": 1.0225869066525498e-06,
|
| 2405 |
+
"loss": 2.7635,
|
| 2406 |
+
"step": 371000
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 20.0,
|
| 2410 |
+
"learning_rate": 1.4028586553732134e-07,
|
| 2411 |
+
"loss": 2.7626,
|
| 2412 |
+
"step": 372000
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 20.0,
|
| 2416 |
+
"eval_accuracy": 0.40992788926628976,
|
| 2417 |
+
"eval_loss": 3.4053595066070557,
|
| 2418 |
+
"eval_runtime": 149.4807,
|
| 2419 |
+
"eval_samples_per_second": 387.475,
|
| 2420 |
+
"eval_steps_per_second": 6.054,
|
| 2421 |
+
"step": 372020
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 20.0,
|
| 2425 |
+
"step": 372020,
|
| 2426 |
+
"total_flos": 1.56740238729216e+18,
|
| 2427 |
+
"train_loss": 3.0558329318429918,
|
| 2428 |
+
"train_runtime": 48026.3336,
|
| 2429 |
+
"train_samples_per_second": 247.871,
|
| 2430 |
+
"train_steps_per_second": 7.746
|
| 2431 |
+
}
|
| 2432 |
+
],
|
| 2433 |
+
"logging_steps": 1000,
|
| 2434 |
+
"max_steps": 372020,
|
| 2435 |
+
"num_input_tokens_seen": 0,
|
| 2436 |
+
"num_train_epochs": 20,
|
| 2437 |
+
"save_steps": 5000,
|
| 2438 |
+
"total_flos": 1.56740238729216e+18,
|
| 2439 |
+
"train_batch_size": 32,
|
| 2440 |
+
"trial_name": null,
|
| 2441 |
+
"trial_params": null
|
| 2442 |
+
}
|