End of training
Browse files- README.md +16 -4
- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +2662 -0
README.md
CHANGED
|
@@ -1,11 +1,23 @@
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
|
|
|
|
|
|
| 4 |
metrics:
|
| 5 |
- accuracy
|
| 6 |
model-index:
|
| 7 |
- name: bert_12_layer_model_v4_complete_training_48
|
| 8 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -13,10 +25,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# bert_12_layer_model_v4_complete_training_48
|
| 15 |
|
| 16 |
-
This model is a fine-tuned version of [](https://huggingface.co/) on
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 4.
|
| 19 |
-
- Accuracy: 0.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
| 4 |
+
datasets:
|
| 5 |
+
- gokuls/wiki_book_corpus_complete_processed_bert_dataset
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
| 8 |
model-index:
|
| 9 |
- name: bert_12_layer_model_v4_complete_training_48
|
| 10 |
+
results:
|
| 11 |
+
- task:
|
| 12 |
+
name: Masked Language Modeling
|
| 13 |
+
type: fill-mask
|
| 14 |
+
dataset:
|
| 15 |
+
name: gokuls/wiki_book_corpus_complete_processed_bert_dataset
|
| 16 |
+
type: gokuls/wiki_book_corpus_complete_processed_bert_dataset
|
| 17 |
+
metrics:
|
| 18 |
+
- name: Accuracy
|
| 19 |
+
type: accuracy
|
| 20 |
+
value: 0.2879346070609049
|
| 21 |
---
|
| 22 |
|
| 23 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 25 |
|
| 26 |
# bert_12_layer_model_v4_complete_training_48
|
| 27 |
|
| 28 |
+
This model is a fine-tuned version of [](https://huggingface.co/) on the gokuls/wiki_book_corpus_complete_processed_bert_dataset dataset.
|
| 29 |
It achieves the following results on the evaluation set:
|
| 30 |
+
- Loss: 4.7042
|
| 31 |
+
- Accuracy: 0.2879
|
| 32 |
|
| 33 |
## Model description
|
| 34 |
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.68,
|
| 3 |
+
"eval_accuracy": 0.2879346070609049,
|
| 4 |
+
"eval_loss": 4.704182147979736,
|
| 5 |
+
"eval_runtime": 1786.8613,
|
| 6 |
+
"eval_samples": 308356,
|
| 7 |
+
"eval_samples_per_second": 172.569,
|
| 8 |
+
"eval_steps_per_second": 3.596,
|
| 9 |
+
"perplexity": 110.4079506469723,
|
| 10 |
+
"train_loss": 5.814965769895094,
|
| 11 |
+
"train_runtime": 198000.2186,
|
| 12 |
+
"train_samples": 5858758,
|
| 13 |
+
"train_samples_per_second": 147.948,
|
| 14 |
+
"train_steps_per_second": 3.082
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.68,
|
| 3 |
+
"eval_accuracy": 0.2879346070609049,
|
| 4 |
+
"eval_loss": 4.704182147979736,
|
| 5 |
+
"eval_runtime": 1786.8613,
|
| 6 |
+
"eval_samples": 308356,
|
| 7 |
+
"eval_samples_per_second": 172.569,
|
| 8 |
+
"eval_steps_per_second": 3.596,
|
| 9 |
+
"perplexity": 110.4079506469723
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.68,
|
| 3 |
+
"train_loss": 5.814965769895094,
|
| 4 |
+
"train_runtime": 198000.2186,
|
| 5 |
+
"train_samples": 5858758,
|
| 6 |
+
"train_samples_per_second": 147.948,
|
| 7 |
+
"train_steps_per_second": 3.082
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2662 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 4.705814361572266,
|
| 3 |
+
"best_model_checkpoint": "bert_12_layer_model_v4_complete_training_48/checkpoint-200000",
|
| 4 |
+
"epoch": 1.6781612020514838,
|
| 5 |
+
"eval_steps": 10000,
|
| 6 |
+
"global_step": 204833,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 5.000000000000001e-07,
|
| 14 |
+
"loss": 10.0556,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 20 |
+
"loss": 9.3294,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 1.5e-06,
|
| 26 |
+
"loss": 8.9103,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.02,
|
| 31 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 32 |
+
"loss": 8.4426,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.02,
|
| 37 |
+
"learning_rate": 2.5e-06,
|
| 38 |
+
"loss": 7.9629,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.02,
|
| 43 |
+
"learning_rate": 3e-06,
|
| 44 |
+
"loss": 7.5474,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03,
|
| 49 |
+
"learning_rate": 3.5e-06,
|
| 50 |
+
"loss": 7.2526,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03,
|
| 55 |
+
"learning_rate": 4.000000000000001e-06,
|
| 56 |
+
"loss": 7.0825,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.04,
|
| 61 |
+
"learning_rate": 4.5e-06,
|
| 62 |
+
"loss": 6.9837,
|
| 63 |
+
"step": 4500
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.04,
|
| 67 |
+
"learning_rate": 5e-06,
|
| 68 |
+
"loss": 6.9078,
|
| 69 |
+
"step": 5000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.05,
|
| 73 |
+
"learning_rate": 5.500000000000001e-06,
|
| 74 |
+
"loss": 6.8501,
|
| 75 |
+
"step": 5500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.05,
|
| 79 |
+
"learning_rate": 6e-06,
|
| 80 |
+
"loss": 6.8013,
|
| 81 |
+
"step": 6000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.05,
|
| 85 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 86 |
+
"loss": 6.763,
|
| 87 |
+
"step": 6500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.06,
|
| 91 |
+
"learning_rate": 7e-06,
|
| 92 |
+
"loss": 6.7338,
|
| 93 |
+
"step": 7000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.06,
|
| 97 |
+
"learning_rate": 7.500000000000001e-06,
|
| 98 |
+
"loss": 6.7016,
|
| 99 |
+
"step": 7500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.07,
|
| 103 |
+
"learning_rate": 8.000000000000001e-06,
|
| 104 |
+
"loss": 6.668,
|
| 105 |
+
"step": 8000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.07,
|
| 109 |
+
"learning_rate": 8.5e-06,
|
| 110 |
+
"loss": 6.6369,
|
| 111 |
+
"step": 8500
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.07,
|
| 115 |
+
"learning_rate": 9e-06,
|
| 116 |
+
"loss": 6.6162,
|
| 117 |
+
"step": 9000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.08,
|
| 121 |
+
"learning_rate": 9.5e-06,
|
| 122 |
+
"loss": 6.5918,
|
| 123 |
+
"step": 9500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.08,
|
| 127 |
+
"learning_rate": 1e-05,
|
| 128 |
+
"loss": 6.5774,
|
| 129 |
+
"step": 10000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.08,
|
| 133 |
+
"eval_accuracy": 0.12525079805443665,
|
| 134 |
+
"eval_loss": 6.53986120223999,
|
| 135 |
+
"eval_runtime": 1806.8145,
|
| 136 |
+
"eval_samples_per_second": 170.663,
|
| 137 |
+
"eval_steps_per_second": 3.556,
|
| 138 |
+
"step": 10000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.09,
|
| 142 |
+
"learning_rate": 9.991670692498626e-06,
|
| 143 |
+
"loss": 6.5584,
|
| 144 |
+
"step": 10500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.09,
|
| 148 |
+
"learning_rate": 9.983341384997253e-06,
|
| 149 |
+
"loss": 6.5297,
|
| 150 |
+
"step": 11000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.09,
|
| 154 |
+
"learning_rate": 9.975012077495878e-06,
|
| 155 |
+
"loss": 6.5098,
|
| 156 |
+
"step": 11500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1,
|
| 160 |
+
"learning_rate": 9.966682769994504e-06,
|
| 161 |
+
"loss": 6.4989,
|
| 162 |
+
"step": 12000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.1,
|
| 166 |
+
"learning_rate": 9.95835346249313e-06,
|
| 167 |
+
"loss": 6.4844,
|
| 168 |
+
"step": 12500
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.11,
|
| 172 |
+
"learning_rate": 9.950024154991756e-06,
|
| 173 |
+
"loss": 6.464,
|
| 174 |
+
"step": 13000
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.11,
|
| 178 |
+
"learning_rate": 9.941694847490381e-06,
|
| 179 |
+
"loss": 6.4558,
|
| 180 |
+
"step": 13500
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.11,
|
| 184 |
+
"learning_rate": 9.933365539989006e-06,
|
| 185 |
+
"loss": 6.449,
|
| 186 |
+
"step": 14000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.12,
|
| 190 |
+
"learning_rate": 9.925036232487631e-06,
|
| 191 |
+
"loss": 6.4263,
|
| 192 |
+
"step": 14500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.12,
|
| 196 |
+
"learning_rate": 9.916706924986258e-06,
|
| 197 |
+
"loss": 6.4189,
|
| 198 |
+
"step": 15000
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.13,
|
| 202 |
+
"learning_rate": 9.908377617484883e-06,
|
| 203 |
+
"loss": 6.4102,
|
| 204 |
+
"step": 15500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.13,
|
| 208 |
+
"learning_rate": 9.90004830998351e-06,
|
| 209 |
+
"loss": 6.4047,
|
| 210 |
+
"step": 16000
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.14,
|
| 214 |
+
"learning_rate": 9.891719002482134e-06,
|
| 215 |
+
"loss": 6.3937,
|
| 216 |
+
"step": 16500
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.14,
|
| 220 |
+
"learning_rate": 9.883389694980761e-06,
|
| 221 |
+
"loss": 6.3809,
|
| 222 |
+
"step": 17000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.14,
|
| 226 |
+
"learning_rate": 9.875060387479386e-06,
|
| 227 |
+
"loss": 6.3775,
|
| 228 |
+
"step": 17500
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.15,
|
| 232 |
+
"learning_rate": 9.866731079978011e-06,
|
| 233 |
+
"loss": 6.3645,
|
| 234 |
+
"step": 18000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.15,
|
| 238 |
+
"learning_rate": 9.858401772476636e-06,
|
| 239 |
+
"loss": 6.3562,
|
| 240 |
+
"step": 18500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.16,
|
| 244 |
+
"learning_rate": 9.850072464975263e-06,
|
| 245 |
+
"loss": 6.3436,
|
| 246 |
+
"step": 19000
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.16,
|
| 250 |
+
"learning_rate": 9.84174315747389e-06,
|
| 251 |
+
"loss": 6.3436,
|
| 252 |
+
"step": 19500
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.16,
|
| 256 |
+
"learning_rate": 9.833413849972514e-06,
|
| 257 |
+
"loss": 6.3254,
|
| 258 |
+
"step": 20000
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.16,
|
| 262 |
+
"eval_accuracy": 0.13876505547385287,
|
| 263 |
+
"eval_loss": 6.310290813446045,
|
| 264 |
+
"eval_runtime": 1824.2928,
|
| 265 |
+
"eval_samples_per_second": 169.028,
|
| 266 |
+
"eval_steps_per_second": 3.522,
|
| 267 |
+
"step": 20000
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.17,
|
| 271 |
+
"learning_rate": 9.82508454247114e-06,
|
| 272 |
+
"loss": 6.3235,
|
| 273 |
+
"step": 20500
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.17,
|
| 277 |
+
"learning_rate": 9.816755234969766e-06,
|
| 278 |
+
"loss": 6.3226,
|
| 279 |
+
"step": 21000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.18,
|
| 283 |
+
"learning_rate": 9.808425927468391e-06,
|
| 284 |
+
"loss": 6.3075,
|
| 285 |
+
"step": 21500
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.18,
|
| 289 |
+
"learning_rate": 9.800096619967016e-06,
|
| 290 |
+
"loss": 6.3025,
|
| 291 |
+
"step": 22000
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.18,
|
| 295 |
+
"learning_rate": 9.791767312465643e-06,
|
| 296 |
+
"loss": 6.2989,
|
| 297 |
+
"step": 22500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.19,
|
| 301 |
+
"learning_rate": 9.783438004964268e-06,
|
| 302 |
+
"loss": 6.299,
|
| 303 |
+
"step": 23000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.19,
|
| 307 |
+
"learning_rate": 9.775108697462895e-06,
|
| 308 |
+
"loss": 6.2843,
|
| 309 |
+
"step": 23500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.2,
|
| 313 |
+
"learning_rate": 9.76677938996152e-06,
|
| 314 |
+
"loss": 6.285,
|
| 315 |
+
"step": 24000
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.2,
|
| 319 |
+
"learning_rate": 9.758450082460145e-06,
|
| 320 |
+
"loss": 6.2705,
|
| 321 |
+
"step": 24500
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.2,
|
| 325 |
+
"learning_rate": 9.750120774958771e-06,
|
| 326 |
+
"loss": 6.2822,
|
| 327 |
+
"step": 25000
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.21,
|
| 331 |
+
"learning_rate": 9.741791467457396e-06,
|
| 332 |
+
"loss": 6.2683,
|
| 333 |
+
"step": 25500
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.21,
|
| 337 |
+
"learning_rate": 9.733462159956021e-06,
|
| 338 |
+
"loss": 6.2659,
|
| 339 |
+
"step": 26000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.22,
|
| 343 |
+
"learning_rate": 9.725132852454648e-06,
|
| 344 |
+
"loss": 6.2596,
|
| 345 |
+
"step": 26500
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.22,
|
| 349 |
+
"learning_rate": 9.716803544953275e-06,
|
| 350 |
+
"loss": 6.2482,
|
| 351 |
+
"step": 27000
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.23,
|
| 355 |
+
"learning_rate": 9.7084742374519e-06,
|
| 356 |
+
"loss": 6.2425,
|
| 357 |
+
"step": 27500
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.23,
|
| 361 |
+
"learning_rate": 9.700144929950525e-06,
|
| 362 |
+
"loss": 6.243,
|
| 363 |
+
"step": 28000
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.23,
|
| 367 |
+
"learning_rate": 9.69181562244915e-06,
|
| 368 |
+
"loss": 6.244,
|
| 369 |
+
"step": 28500
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.24,
|
| 373 |
+
"learning_rate": 9.683486314947776e-06,
|
| 374 |
+
"loss": 6.2352,
|
| 375 |
+
"step": 29000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.24,
|
| 379 |
+
"learning_rate": 9.675157007446401e-06,
|
| 380 |
+
"loss": 6.2333,
|
| 381 |
+
"step": 29500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.25,
|
| 385 |
+
"learning_rate": 9.666827699945028e-06,
|
| 386 |
+
"loss": 6.2278,
|
| 387 |
+
"step": 30000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.25,
|
| 391 |
+
"eval_accuracy": 0.14427574743432298,
|
| 392 |
+
"eval_loss": 6.211411476135254,
|
| 393 |
+
"eval_runtime": 1813.005,
|
| 394 |
+
"eval_samples_per_second": 170.08,
|
| 395 |
+
"eval_steps_per_second": 3.544,
|
| 396 |
+
"step": 30000
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.25,
|
| 400 |
+
"learning_rate": 9.658498392443653e-06,
|
| 401 |
+
"loss": 6.2199,
|
| 402 |
+
"step": 30500
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.25,
|
| 406 |
+
"learning_rate": 9.65016908494228e-06,
|
| 407 |
+
"loss": 6.2198,
|
| 408 |
+
"step": 31000
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.26,
|
| 412 |
+
"learning_rate": 9.641839777440905e-06,
|
| 413 |
+
"loss": 6.2182,
|
| 414 |
+
"step": 31500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.26,
|
| 418 |
+
"learning_rate": 9.63351046993953e-06,
|
| 419 |
+
"loss": 6.2189,
|
| 420 |
+
"step": 32000
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.27,
|
| 424 |
+
"learning_rate": 9.625181162438155e-06,
|
| 425 |
+
"loss": 6.211,
|
| 426 |
+
"step": 32500
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.27,
|
| 430 |
+
"learning_rate": 9.616851854936781e-06,
|
| 431 |
+
"loss": 6.2024,
|
| 432 |
+
"step": 33000
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.27,
|
| 436 |
+
"learning_rate": 9.608522547435406e-06,
|
| 437 |
+
"loss": 6.1984,
|
| 438 |
+
"step": 33500
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.28,
|
| 442 |
+
"learning_rate": 9.600193239934033e-06,
|
| 443 |
+
"loss": 6.2002,
|
| 444 |
+
"step": 34000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.28,
|
| 448 |
+
"learning_rate": 9.591863932432658e-06,
|
| 449 |
+
"loss": 6.195,
|
| 450 |
+
"step": 34500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.29,
|
| 454 |
+
"learning_rate": 9.583534624931285e-06,
|
| 455 |
+
"loss": 6.1938,
|
| 456 |
+
"step": 35000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.29,
|
| 460 |
+
"learning_rate": 9.57520531742991e-06,
|
| 461 |
+
"loss": 6.2003,
|
| 462 |
+
"step": 35500
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.29,
|
| 466 |
+
"learning_rate": 9.566876009928535e-06,
|
| 467 |
+
"loss": 6.1859,
|
| 468 |
+
"step": 36000
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 0.3,
|
| 472 |
+
"learning_rate": 9.55854670242716e-06,
|
| 473 |
+
"loss": 6.1807,
|
| 474 |
+
"step": 36500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.3,
|
| 478 |
+
"learning_rate": 9.550217394925787e-06,
|
| 479 |
+
"loss": 6.18,
|
| 480 |
+
"step": 37000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.31,
|
| 484 |
+
"learning_rate": 9.541888087424413e-06,
|
| 485 |
+
"loss": 6.1864,
|
| 486 |
+
"step": 37500
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.31,
|
| 490 |
+
"learning_rate": 9.533558779923038e-06,
|
| 491 |
+
"loss": 6.169,
|
| 492 |
+
"step": 38000
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.32,
|
| 496 |
+
"learning_rate": 9.525229472421663e-06,
|
| 497 |
+
"loss": 6.1768,
|
| 498 |
+
"step": 38500
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.32,
|
| 502 |
+
"learning_rate": 9.51690016492029e-06,
|
| 503 |
+
"loss": 6.1708,
|
| 504 |
+
"step": 39000
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.32,
|
| 508 |
+
"learning_rate": 9.508570857418915e-06,
|
| 509 |
+
"loss": 6.1663,
|
| 510 |
+
"step": 39500
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 0.33,
|
| 514 |
+
"learning_rate": 9.50024154991754e-06,
|
| 515 |
+
"loss": 6.1712,
|
| 516 |
+
"step": 40000
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.33,
|
| 520 |
+
"eval_accuracy": 0.1475426777304807,
|
| 521 |
+
"eval_loss": 6.149131774902344,
|
| 522 |
+
"eval_runtime": 1803.2109,
|
| 523 |
+
"eval_samples_per_second": 171.004,
|
| 524 |
+
"eval_steps_per_second": 3.563,
|
| 525 |
+
"step": 40000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.33,
|
| 529 |
+
"learning_rate": 9.491912242416167e-06,
|
| 530 |
+
"loss": 6.1603,
|
| 531 |
+
"step": 40500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.34,
|
| 535 |
+
"learning_rate": 9.483582934914792e-06,
|
| 536 |
+
"loss": 6.1614,
|
| 537 |
+
"step": 41000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.34,
|
| 541 |
+
"learning_rate": 9.475253627413418e-06,
|
| 542 |
+
"loss": 6.1602,
|
| 543 |
+
"step": 41500
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.34,
|
| 547 |
+
"learning_rate": 9.466924319912043e-06,
|
| 548 |
+
"loss": 6.1587,
|
| 549 |
+
"step": 42000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.35,
|
| 553 |
+
"learning_rate": 9.458595012410668e-06,
|
| 554 |
+
"loss": 6.1588,
|
| 555 |
+
"step": 42500
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.35,
|
| 559 |
+
"learning_rate": 9.450265704909295e-06,
|
| 560 |
+
"loss": 6.1506,
|
| 561 |
+
"step": 43000
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.36,
|
| 565 |
+
"learning_rate": 9.44193639740792e-06,
|
| 566 |
+
"loss": 6.1512,
|
| 567 |
+
"step": 43500
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.36,
|
| 571 |
+
"learning_rate": 9.433607089906545e-06,
|
| 572 |
+
"loss": 6.1491,
|
| 573 |
+
"step": 44000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.36,
|
| 577 |
+
"learning_rate": 9.425277782405172e-06,
|
| 578 |
+
"loss": 6.1434,
|
| 579 |
+
"step": 44500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.37,
|
| 583 |
+
"learning_rate": 9.416948474903798e-06,
|
| 584 |
+
"loss": 6.1409,
|
| 585 |
+
"step": 45000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.37,
|
| 589 |
+
"learning_rate": 9.408619167402423e-06,
|
| 590 |
+
"loss": 6.1363,
|
| 591 |
+
"step": 45500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.38,
|
| 595 |
+
"learning_rate": 9.400289859901048e-06,
|
| 596 |
+
"loss": 6.1389,
|
| 597 |
+
"step": 46000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.38,
|
| 601 |
+
"learning_rate": 9.391960552399673e-06,
|
| 602 |
+
"loss": 6.1324,
|
| 603 |
+
"step": 46500
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.39,
|
| 607 |
+
"learning_rate": 9.3836312448983e-06,
|
| 608 |
+
"loss": 6.1361,
|
| 609 |
+
"step": 47000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.39,
|
| 613 |
+
"learning_rate": 9.375301937396925e-06,
|
| 614 |
+
"loss": 6.1389,
|
| 615 |
+
"step": 47500
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.39,
|
| 619 |
+
"learning_rate": 9.366972629895552e-06,
|
| 620 |
+
"loss": 6.1341,
|
| 621 |
+
"step": 48000
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.4,
|
| 625 |
+
"learning_rate": 9.358643322394177e-06,
|
| 626 |
+
"loss": 6.1251,
|
| 627 |
+
"step": 48500
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.4,
|
| 631 |
+
"learning_rate": 9.350314014892804e-06,
|
| 632 |
+
"loss": 6.1259,
|
| 633 |
+
"step": 49000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.41,
|
| 637 |
+
"learning_rate": 9.341984707391429e-06,
|
| 638 |
+
"loss": 6.1298,
|
| 639 |
+
"step": 49500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.41,
|
| 643 |
+
"learning_rate": 9.333655399890054e-06,
|
| 644 |
+
"loss": 6.12,
|
| 645 |
+
"step": 50000
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.41,
|
| 649 |
+
"eval_accuracy": 0.14918066299504143,
|
| 650 |
+
"eval_loss": 6.108561038970947,
|
| 651 |
+
"eval_runtime": 1794.239,
|
| 652 |
+
"eval_samples_per_second": 171.859,
|
| 653 |
+
"eval_steps_per_second": 3.581,
|
| 654 |
+
"step": 50000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.41,
|
| 658 |
+
"learning_rate": 9.325326092388679e-06,
|
| 659 |
+
"loss": 6.1252,
|
| 660 |
+
"step": 50500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.42,
|
| 664 |
+
"learning_rate": 9.316996784887305e-06,
|
| 665 |
+
"loss": 6.1187,
|
| 666 |
+
"step": 51000
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.42,
|
| 670 |
+
"learning_rate": 9.30866747738593e-06,
|
| 671 |
+
"loss": 6.1217,
|
| 672 |
+
"step": 51500
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.43,
|
| 676 |
+
"learning_rate": 9.300338169884557e-06,
|
| 677 |
+
"loss": 6.1137,
|
| 678 |
+
"step": 52000
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.43,
|
| 682 |
+
"learning_rate": 9.292008862383182e-06,
|
| 683 |
+
"loss": 6.113,
|
| 684 |
+
"step": 52500
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.43,
|
| 688 |
+
"learning_rate": 9.283679554881809e-06,
|
| 689 |
+
"loss": 6.1066,
|
| 690 |
+
"step": 53000
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.44,
|
| 694 |
+
"learning_rate": 9.275350247380434e-06,
|
| 695 |
+
"loss": 6.1104,
|
| 696 |
+
"step": 53500
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.44,
|
| 700 |
+
"learning_rate": 9.267020939879059e-06,
|
| 701 |
+
"loss": 6.1091,
|
| 702 |
+
"step": 54000
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.45,
|
| 706 |
+
"learning_rate": 9.258691632377685e-06,
|
| 707 |
+
"loss": 6.0986,
|
| 708 |
+
"step": 54500
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.45,
|
| 712 |
+
"learning_rate": 9.25036232487631e-06,
|
| 713 |
+
"loss": 6.1056,
|
| 714 |
+
"step": 55000
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.45,
|
| 718 |
+
"learning_rate": 9.242033017374937e-06,
|
| 719 |
+
"loss": 6.1055,
|
| 720 |
+
"step": 55500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.46,
|
| 724 |
+
"learning_rate": 9.233703709873562e-06,
|
| 725 |
+
"loss": 6.1067,
|
| 726 |
+
"step": 56000
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.46,
|
| 730 |
+
"learning_rate": 9.225374402372189e-06,
|
| 731 |
+
"loss": 6.1028,
|
| 732 |
+
"step": 56500
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.47,
|
| 736 |
+
"learning_rate": 9.217045094870814e-06,
|
| 737 |
+
"loss": 6.1013,
|
| 738 |
+
"step": 57000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.47,
|
| 742 |
+
"learning_rate": 9.208715787369439e-06,
|
| 743 |
+
"loss": 6.1004,
|
| 744 |
+
"step": 57500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.48,
|
| 748 |
+
"learning_rate": 9.200386479868064e-06,
|
| 749 |
+
"loss": 6.0907,
|
| 750 |
+
"step": 58000
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 0.48,
|
| 754 |
+
"learning_rate": 9.19205717236669e-06,
|
| 755 |
+
"loss": 6.1033,
|
| 756 |
+
"step": 58500
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 0.48,
|
| 760 |
+
"learning_rate": 9.183727864865315e-06,
|
| 761 |
+
"loss": 6.0882,
|
| 762 |
+
"step": 59000
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.49,
|
| 766 |
+
"learning_rate": 9.175398557363942e-06,
|
| 767 |
+
"loss": 6.0939,
|
| 768 |
+
"step": 59500
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.49,
|
| 772 |
+
"learning_rate": 9.167069249862567e-06,
|
| 773 |
+
"loss": 6.0914,
|
| 774 |
+
"step": 60000
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.49,
|
| 778 |
+
"eval_accuracy": 0.14997233538934182,
|
| 779 |
+
"eval_loss": 6.0781354904174805,
|
| 780 |
+
"eval_runtime": 1803.6606,
|
| 781 |
+
"eval_samples_per_second": 170.961,
|
| 782 |
+
"eval_steps_per_second": 3.562,
|
| 783 |
+
"step": 60000
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.5,
|
| 787 |
+
"learning_rate": 9.158739942361194e-06,
|
| 788 |
+
"loss": 6.0949,
|
| 789 |
+
"step": 60500
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.5,
|
| 793 |
+
"learning_rate": 9.150410634859819e-06,
|
| 794 |
+
"loss": 6.0895,
|
| 795 |
+
"step": 61000
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.5,
|
| 799 |
+
"learning_rate": 9.142081327358444e-06,
|
| 800 |
+
"loss": 6.0897,
|
| 801 |
+
"step": 61500
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.51,
|
| 805 |
+
"learning_rate": 9.133752019857069e-06,
|
| 806 |
+
"loss": 6.0892,
|
| 807 |
+
"step": 62000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.51,
|
| 811 |
+
"learning_rate": 9.125422712355696e-06,
|
| 812 |
+
"loss": 6.0818,
|
| 813 |
+
"step": 62500
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.52,
|
| 817 |
+
"learning_rate": 9.11709340485432e-06,
|
| 818 |
+
"loss": 6.0875,
|
| 819 |
+
"step": 63000
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.52,
|
| 823 |
+
"learning_rate": 9.108764097352947e-06,
|
| 824 |
+
"loss": 6.0838,
|
| 825 |
+
"step": 63500
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.52,
|
| 829 |
+
"learning_rate": 9.100434789851572e-06,
|
| 830 |
+
"loss": 6.0708,
|
| 831 |
+
"step": 64000
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.53,
|
| 835 |
+
"learning_rate": 9.092105482350199e-06,
|
| 836 |
+
"loss": 6.0753,
|
| 837 |
+
"step": 64500
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.53,
|
| 841 |
+
"learning_rate": 9.083776174848824e-06,
|
| 842 |
+
"loss": 6.0786,
|
| 843 |
+
"step": 65000
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.54,
|
| 847 |
+
"learning_rate": 9.075446867347449e-06,
|
| 848 |
+
"loss": 6.0781,
|
| 849 |
+
"step": 65500
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.54,
|
| 853 |
+
"learning_rate": 9.067117559846076e-06,
|
| 854 |
+
"loss": 6.0831,
|
| 855 |
+
"step": 66000
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.54,
|
| 859 |
+
"learning_rate": 9.0587882523447e-06,
|
| 860 |
+
"loss": 6.0692,
|
| 861 |
+
"step": 66500
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.55,
|
| 865 |
+
"learning_rate": 9.050458944843327e-06,
|
| 866 |
+
"loss": 6.0783,
|
| 867 |
+
"step": 67000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.55,
|
| 871 |
+
"learning_rate": 9.042129637341952e-06,
|
| 872 |
+
"loss": 6.0708,
|
| 873 |
+
"step": 67500
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.56,
|
| 877 |
+
"learning_rate": 9.033800329840577e-06,
|
| 878 |
+
"loss": 6.0648,
|
| 879 |
+
"step": 68000
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.56,
|
| 883 |
+
"learning_rate": 9.025471022339204e-06,
|
| 884 |
+
"loss": 6.0696,
|
| 885 |
+
"step": 68500
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.57,
|
| 889 |
+
"learning_rate": 9.017141714837829e-06,
|
| 890 |
+
"loss": 6.0674,
|
| 891 |
+
"step": 69000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.57,
|
| 895 |
+
"learning_rate": 9.008812407336454e-06,
|
| 896 |
+
"loss": 6.0652,
|
| 897 |
+
"step": 69500
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.57,
|
| 901 |
+
"learning_rate": 9.00048309983508e-06,
|
| 902 |
+
"loss": 6.0676,
|
| 903 |
+
"step": 70000
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.57,
|
| 907 |
+
"eval_accuracy": 0.15047221025546462,
|
| 908 |
+
"eval_loss": 6.054013252258301,
|
| 909 |
+
"eval_runtime": 1810.7781,
|
| 910 |
+
"eval_samples_per_second": 170.289,
|
| 911 |
+
"eval_steps_per_second": 3.548,
|
| 912 |
+
"step": 70000
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.58,
|
| 916 |
+
"learning_rate": 8.992153792333706e-06,
|
| 917 |
+
"loss": 6.0658,
|
| 918 |
+
"step": 70500
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.58,
|
| 922 |
+
"learning_rate": 8.983824484832332e-06,
|
| 923 |
+
"loss": 6.0653,
|
| 924 |
+
"step": 71000
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.59,
|
| 928 |
+
"learning_rate": 8.975495177330957e-06,
|
| 929 |
+
"loss": 6.0616,
|
| 930 |
+
"step": 71500
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.59,
|
| 934 |
+
"learning_rate": 8.967165869829582e-06,
|
| 935 |
+
"loss": 6.0673,
|
| 936 |
+
"step": 72000
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.59,
|
| 940 |
+
"learning_rate": 8.958836562328209e-06,
|
| 941 |
+
"loss": 6.0589,
|
| 942 |
+
"step": 72500
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 0.6,
|
| 946 |
+
"learning_rate": 8.950507254826834e-06,
|
| 947 |
+
"loss": 6.0565,
|
| 948 |
+
"step": 73000
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.6,
|
| 952 |
+
"learning_rate": 8.942177947325459e-06,
|
| 953 |
+
"loss": 6.0558,
|
| 954 |
+
"step": 73500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.61,
|
| 958 |
+
"learning_rate": 8.933848639824086e-06,
|
| 959 |
+
"loss": 6.055,
|
| 960 |
+
"step": 74000
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.61,
|
| 964 |
+
"learning_rate": 8.925519332322713e-06,
|
| 965 |
+
"loss": 6.0538,
|
| 966 |
+
"step": 74500
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.61,
|
| 970 |
+
"learning_rate": 8.917190024821338e-06,
|
| 971 |
+
"loss": 6.0593,
|
| 972 |
+
"step": 75000
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 0.62,
|
| 976 |
+
"learning_rate": 8.908860717319963e-06,
|
| 977 |
+
"loss": 6.0565,
|
| 978 |
+
"step": 75500
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.62,
|
| 982 |
+
"learning_rate": 8.900531409818588e-06,
|
| 983 |
+
"loss": 6.0579,
|
| 984 |
+
"step": 76000
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.63,
|
| 988 |
+
"learning_rate": 8.892202102317214e-06,
|
| 989 |
+
"loss": 6.0533,
|
| 990 |
+
"step": 76500
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.63,
|
| 994 |
+
"learning_rate": 8.88387279481584e-06,
|
| 995 |
+
"loss": 6.0536,
|
| 996 |
+
"step": 77000
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.63,
|
| 1000 |
+
"learning_rate": 8.875543487314466e-06,
|
| 1001 |
+
"loss": 6.0477,
|
| 1002 |
+
"step": 77500
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 0.64,
|
| 1006 |
+
"learning_rate": 8.867214179813091e-06,
|
| 1007 |
+
"loss": 6.0453,
|
| 1008 |
+
"step": 78000
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.64,
|
| 1012 |
+
"learning_rate": 8.858884872311718e-06,
|
| 1013 |
+
"loss": 6.0512,
|
| 1014 |
+
"step": 78500
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 0.65,
|
| 1018 |
+
"learning_rate": 8.850555564810343e-06,
|
| 1019 |
+
"loss": 6.0429,
|
| 1020 |
+
"step": 79000
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.65,
|
| 1024 |
+
"learning_rate": 8.842226257308968e-06,
|
| 1025 |
+
"loss": 6.0466,
|
| 1026 |
+
"step": 79500
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.66,
|
| 1030 |
+
"learning_rate": 8.833896949807593e-06,
|
| 1031 |
+
"loss": 6.0492,
|
| 1032 |
+
"step": 80000
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.66,
|
| 1036 |
+
"eval_accuracy": 0.15117882498687546,
|
| 1037 |
+
"eval_loss": 6.034516334533691,
|
| 1038 |
+
"eval_runtime": 1810.1741,
|
| 1039 |
+
"eval_samples_per_second": 170.346,
|
| 1040 |
+
"eval_steps_per_second": 3.549,
|
| 1041 |
+
"step": 80000
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.66,
|
| 1045 |
+
"learning_rate": 8.82556764230622e-06,
|
| 1046 |
+
"loss": 6.0465,
|
| 1047 |
+
"step": 80500
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.66,
|
| 1051 |
+
"learning_rate": 8.817238334804844e-06,
|
| 1052 |
+
"loss": 6.0447,
|
| 1053 |
+
"step": 81000
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.67,
|
| 1057 |
+
"learning_rate": 8.808909027303471e-06,
|
| 1058 |
+
"loss": 6.0407,
|
| 1059 |
+
"step": 81500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.67,
|
| 1063 |
+
"learning_rate": 8.800579719802096e-06,
|
| 1064 |
+
"loss": 6.0431,
|
| 1065 |
+
"step": 82000
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.68,
|
| 1069 |
+
"learning_rate": 8.792250412300723e-06,
|
| 1070 |
+
"loss": 6.0431,
|
| 1071 |
+
"step": 82500
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.68,
|
| 1075 |
+
"learning_rate": 8.783921104799348e-06,
|
| 1076 |
+
"loss": 6.037,
|
| 1077 |
+
"step": 83000
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.68,
|
| 1081 |
+
"learning_rate": 8.775591797297973e-06,
|
| 1082 |
+
"loss": 6.0392,
|
| 1083 |
+
"step": 83500
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.69,
|
| 1087 |
+
"learning_rate": 8.767262489796598e-06,
|
| 1088 |
+
"loss": 6.0371,
|
| 1089 |
+
"step": 84000
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.69,
|
| 1093 |
+
"learning_rate": 8.758933182295224e-06,
|
| 1094 |
+
"loss": 6.0282,
|
| 1095 |
+
"step": 84500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.7,
|
| 1099 |
+
"learning_rate": 8.750603874793851e-06,
|
| 1100 |
+
"loss": 6.0355,
|
| 1101 |
+
"step": 85000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.7,
|
| 1105 |
+
"learning_rate": 8.742274567292476e-06,
|
| 1106 |
+
"loss": 6.031,
|
| 1107 |
+
"step": 85500
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.7,
|
| 1111 |
+
"learning_rate": 8.733945259791101e-06,
|
| 1112 |
+
"loss": 6.0366,
|
| 1113 |
+
"step": 86000
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 0.71,
|
| 1117 |
+
"learning_rate": 8.725615952289728e-06,
|
| 1118 |
+
"loss": 6.0268,
|
| 1119 |
+
"step": 86500
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.71,
|
| 1123 |
+
"learning_rate": 8.717286644788353e-06,
|
| 1124 |
+
"loss": 6.0266,
|
| 1125 |
+
"step": 87000
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.72,
|
| 1129 |
+
"learning_rate": 8.708957337286978e-06,
|
| 1130 |
+
"loss": 6.0357,
|
| 1131 |
+
"step": 87500
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.72,
|
| 1135 |
+
"learning_rate": 8.700628029785605e-06,
|
| 1136 |
+
"loss": 6.0274,
|
| 1137 |
+
"step": 88000
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.73,
|
| 1141 |
+
"learning_rate": 8.69229872228423e-06,
|
| 1142 |
+
"loss": 6.0286,
|
| 1143 |
+
"step": 88500
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.73,
|
| 1147 |
+
"learning_rate": 8.683969414782856e-06,
|
| 1148 |
+
"loss": 6.0298,
|
| 1149 |
+
"step": 89000
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.73,
|
| 1153 |
+
"learning_rate": 8.675640107281481e-06,
|
| 1154 |
+
"loss": 6.02,
|
| 1155 |
+
"step": 89500
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 0.74,
|
| 1159 |
+
"learning_rate": 8.667310799780106e-06,
|
| 1160 |
+
"loss": 6.028,
|
| 1161 |
+
"step": 90000
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.74,
|
| 1165 |
+
"eval_accuracy": 0.15156892809200398,
|
| 1166 |
+
"eval_loss": 6.0156707763671875,
|
| 1167 |
+
"eval_runtime": 1828.1199,
|
| 1168 |
+
"eval_samples_per_second": 168.674,
|
| 1169 |
+
"eval_steps_per_second": 3.515,
|
| 1170 |
+
"step": 90000
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 0.74,
|
| 1174 |
+
"learning_rate": 8.658981492278733e-06,
|
| 1175 |
+
"loss": 6.0266,
|
| 1176 |
+
"step": 90500
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 0.75,
|
| 1180 |
+
"learning_rate": 8.650652184777358e-06,
|
| 1181 |
+
"loss": 6.0212,
|
| 1182 |
+
"step": 91000
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.75,
|
| 1186 |
+
"learning_rate": 8.642322877275983e-06,
|
| 1187 |
+
"loss": 6.02,
|
| 1188 |
+
"step": 91500
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.75,
|
| 1192 |
+
"learning_rate": 8.63399356977461e-06,
|
| 1193 |
+
"loss": 6.0243,
|
| 1194 |
+
"step": 92000
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.76,
|
| 1198 |
+
"learning_rate": 8.625664262273236e-06,
|
| 1199 |
+
"loss": 6.0218,
|
| 1200 |
+
"step": 92500
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.76,
|
| 1204 |
+
"learning_rate": 8.617334954771861e-06,
|
| 1205 |
+
"loss": 6.0209,
|
| 1206 |
+
"step": 93000
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.77,
|
| 1210 |
+
"learning_rate": 8.609005647270486e-06,
|
| 1211 |
+
"loss": 6.0217,
|
| 1212 |
+
"step": 93500
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 0.77,
|
| 1216 |
+
"learning_rate": 8.600676339769111e-06,
|
| 1217 |
+
"loss": 6.0186,
|
| 1218 |
+
"step": 94000
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 0.77,
|
| 1222 |
+
"learning_rate": 8.592347032267738e-06,
|
| 1223 |
+
"loss": 6.0168,
|
| 1224 |
+
"step": 94500
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.78,
|
| 1228 |
+
"learning_rate": 8.584017724766363e-06,
|
| 1229 |
+
"loss": 6.0078,
|
| 1230 |
+
"step": 95000
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 0.78,
|
| 1234 |
+
"learning_rate": 8.57568841726499e-06,
|
| 1235 |
+
"loss": 6.0245,
|
| 1236 |
+
"step": 95500
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 0.79,
|
| 1240 |
+
"learning_rate": 8.567359109763615e-06,
|
| 1241 |
+
"loss": 6.0165,
|
| 1242 |
+
"step": 96000
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.79,
|
| 1246 |
+
"learning_rate": 8.559029802262241e-06,
|
| 1247 |
+
"loss": 6.0122,
|
| 1248 |
+
"step": 96500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.79,
|
| 1252 |
+
"learning_rate": 8.550700494760866e-06,
|
| 1253 |
+
"loss": 6.0171,
|
| 1254 |
+
"step": 97000
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 0.8,
|
| 1258 |
+
"learning_rate": 8.542371187259491e-06,
|
| 1259 |
+
"loss": 6.0088,
|
| 1260 |
+
"step": 97500
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 0.8,
|
| 1264 |
+
"learning_rate": 8.534041879758118e-06,
|
| 1265 |
+
"loss": 6.0039,
|
| 1266 |
+
"step": 98000
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.81,
|
| 1270 |
+
"learning_rate": 8.525712572256743e-06,
|
| 1271 |
+
"loss": 6.0001,
|
| 1272 |
+
"step": 98500
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.81,
|
| 1276 |
+
"learning_rate": 8.517383264755368e-06,
|
| 1277 |
+
"loss": 5.9801,
|
| 1278 |
+
"step": 99000
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.82,
|
| 1282 |
+
"learning_rate": 8.509053957253995e-06,
|
| 1283 |
+
"loss": 5.9577,
|
| 1284 |
+
"step": 99500
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.82,
|
| 1288 |
+
"learning_rate": 8.500724649752622e-06,
|
| 1289 |
+
"loss": 5.9337,
|
| 1290 |
+
"step": 100000
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.82,
|
| 1294 |
+
"eval_accuracy": 0.15331601926979785,
|
| 1295 |
+
"eval_loss": 5.898829936981201,
|
| 1296 |
+
"eval_runtime": 1812.7063,
|
| 1297 |
+
"eval_samples_per_second": 170.108,
|
| 1298 |
+
"eval_steps_per_second": 3.544,
|
| 1299 |
+
"step": 100000
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.82,
|
| 1303 |
+
"learning_rate": 8.492395342251247e-06,
|
| 1304 |
+
"loss": 5.907,
|
| 1305 |
+
"step": 100500
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.83,
|
| 1309 |
+
"learning_rate": 8.484066034749872e-06,
|
| 1310 |
+
"loss": 5.886,
|
| 1311 |
+
"step": 101000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.83,
|
| 1315 |
+
"learning_rate": 8.475736727248497e-06,
|
| 1316 |
+
"loss": 5.8732,
|
| 1317 |
+
"step": 101500
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 0.84,
|
| 1321 |
+
"learning_rate": 8.467407419747123e-06,
|
| 1322 |
+
"loss": 5.8616,
|
| 1323 |
+
"step": 102000
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 0.84,
|
| 1327 |
+
"learning_rate": 8.459078112245748e-06,
|
| 1328 |
+
"loss": 5.8567,
|
| 1329 |
+
"step": 102500
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.84,
|
| 1333 |
+
"learning_rate": 8.450748804744375e-06,
|
| 1334 |
+
"loss": 5.8453,
|
| 1335 |
+
"step": 103000
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.85,
|
| 1339 |
+
"learning_rate": 8.442419497243e-06,
|
| 1340 |
+
"loss": 5.8401,
|
| 1341 |
+
"step": 103500
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.85,
|
| 1345 |
+
"learning_rate": 8.434090189741627e-06,
|
| 1346 |
+
"loss": 5.8362,
|
| 1347 |
+
"step": 104000
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.86,
|
| 1351 |
+
"learning_rate": 8.425760882240252e-06,
|
| 1352 |
+
"loss": 5.8284,
|
| 1353 |
+
"step": 104500
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.86,
|
| 1357 |
+
"learning_rate": 8.417431574738877e-06,
|
| 1358 |
+
"loss": 5.8164,
|
| 1359 |
+
"step": 105000
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.86,
|
| 1363 |
+
"learning_rate": 8.409102267237502e-06,
|
| 1364 |
+
"loss": 5.8103,
|
| 1365 |
+
"step": 105500
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 0.87,
|
| 1369 |
+
"learning_rate": 8.400772959736128e-06,
|
| 1370 |
+
"loss": 5.8114,
|
| 1371 |
+
"step": 106000
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.87,
|
| 1375 |
+
"learning_rate": 8.392443652234753e-06,
|
| 1376 |
+
"loss": 5.8097,
|
| 1377 |
+
"step": 106500
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.88,
|
| 1381 |
+
"learning_rate": 8.38411434473338e-06,
|
| 1382 |
+
"loss": 5.8062,
|
| 1383 |
+
"step": 107000
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 0.88,
|
| 1387 |
+
"learning_rate": 8.375785037232005e-06,
|
| 1388 |
+
"loss": 5.7963,
|
| 1389 |
+
"step": 107500
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.88,
|
| 1393 |
+
"learning_rate": 8.367455729730632e-06,
|
| 1394 |
+
"loss": 5.7905,
|
| 1395 |
+
"step": 108000
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.89,
|
| 1399 |
+
"learning_rate": 8.359126422229257e-06,
|
| 1400 |
+
"loss": 5.7889,
|
| 1401 |
+
"step": 108500
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.89,
|
| 1405 |
+
"learning_rate": 8.350797114727882e-06,
|
| 1406 |
+
"loss": 5.7848,
|
| 1407 |
+
"step": 109000
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 0.9,
|
| 1411 |
+
"learning_rate": 8.342467807226507e-06,
|
| 1412 |
+
"loss": 5.7796,
|
| 1413 |
+
"step": 109500
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.9,
|
| 1417 |
+
"learning_rate": 8.334138499725133e-06,
|
| 1418 |
+
"loss": 5.7697,
|
| 1419 |
+
"step": 110000
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.9,
|
| 1423 |
+
"eval_accuracy": 0.1653736661861396,
|
| 1424 |
+
"eval_loss": 5.740232467651367,
|
| 1425 |
+
"eval_runtime": 1802.5446,
|
| 1426 |
+
"eval_samples_per_second": 171.067,
|
| 1427 |
+
"eval_steps_per_second": 3.564,
|
| 1428 |
+
"step": 110000
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.91,
|
| 1432 |
+
"learning_rate": 8.32580919222376e-06,
|
| 1433 |
+
"loss": 5.7694,
|
| 1434 |
+
"step": 110500
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.91,
|
| 1438 |
+
"learning_rate": 8.317479884722385e-06,
|
| 1439 |
+
"loss": 5.7703,
|
| 1440 |
+
"step": 111000
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.91,
|
| 1444 |
+
"learning_rate": 8.30915057722101e-06,
|
| 1445 |
+
"loss": 5.7666,
|
| 1446 |
+
"step": 111500
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 0.92,
|
| 1450 |
+
"learning_rate": 8.300821269719637e-06,
|
| 1451 |
+
"loss": 5.7614,
|
| 1452 |
+
"step": 112000
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.92,
|
| 1456 |
+
"learning_rate": 8.292491962218262e-06,
|
| 1457 |
+
"loss": 5.7536,
|
| 1458 |
+
"step": 112500
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.93,
|
| 1462 |
+
"learning_rate": 8.284162654716887e-06,
|
| 1463 |
+
"loss": 5.7505,
|
| 1464 |
+
"step": 113000
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 0.93,
|
| 1468 |
+
"learning_rate": 8.275833347215514e-06,
|
| 1469 |
+
"loss": 5.7471,
|
| 1470 |
+
"step": 113500
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.93,
|
| 1474 |
+
"learning_rate": 8.267504039714139e-06,
|
| 1475 |
+
"loss": 5.7439,
|
| 1476 |
+
"step": 114000
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.94,
|
| 1480 |
+
"learning_rate": 8.259174732212765e-06,
|
| 1481 |
+
"loss": 5.7347,
|
| 1482 |
+
"step": 114500
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.94,
|
| 1486 |
+
"learning_rate": 8.25084542471139e-06,
|
| 1487 |
+
"loss": 5.7354,
|
| 1488 |
+
"step": 115000
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.95,
|
| 1492 |
+
"learning_rate": 8.242516117210015e-06,
|
| 1493 |
+
"loss": 5.7304,
|
| 1494 |
+
"step": 115500
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.95,
|
| 1498 |
+
"learning_rate": 8.234186809708642e-06,
|
| 1499 |
+
"loss": 5.728,
|
| 1500 |
+
"step": 116000
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.95,
|
| 1504 |
+
"learning_rate": 8.225857502207267e-06,
|
| 1505 |
+
"loss": 5.7234,
|
| 1506 |
+
"step": 116500
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.96,
|
| 1510 |
+
"learning_rate": 8.217528194705892e-06,
|
| 1511 |
+
"loss": 5.7152,
|
| 1512 |
+
"step": 117000
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 0.96,
|
| 1516 |
+
"learning_rate": 8.209198887204519e-06,
|
| 1517 |
+
"loss": 5.7106,
|
| 1518 |
+
"step": 117500
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.97,
|
| 1522 |
+
"learning_rate": 8.200869579703145e-06,
|
| 1523 |
+
"loss": 5.7065,
|
| 1524 |
+
"step": 118000
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 0.97,
|
| 1528 |
+
"learning_rate": 8.19254027220177e-06,
|
| 1529 |
+
"loss": 5.7019,
|
| 1530 |
+
"step": 118500
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.97,
|
| 1534 |
+
"learning_rate": 8.184210964700395e-06,
|
| 1535 |
+
"loss": 5.7067,
|
| 1536 |
+
"step": 119000
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.98,
|
| 1540 |
+
"learning_rate": 8.17588165719902e-06,
|
| 1541 |
+
"loss": 5.6976,
|
| 1542 |
+
"step": 119500
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.98,
|
| 1546 |
+
"learning_rate": 8.167552349697647e-06,
|
| 1547 |
+
"loss": 5.6918,
|
| 1548 |
+
"step": 120000
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 0.98,
|
| 1552 |
+
"eval_accuracy": 0.1776737422744013,
|
| 1553 |
+
"eval_loss": 5.638702392578125,
|
| 1554 |
+
"eval_runtime": 1780.2997,
|
| 1555 |
+
"eval_samples_per_second": 173.205,
|
| 1556 |
+
"eval_steps_per_second": 3.609,
|
| 1557 |
+
"step": 120000
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.99,
|
| 1561 |
+
"learning_rate": 8.159223042196272e-06,
|
| 1562 |
+
"loss": 5.6797,
|
| 1563 |
+
"step": 120500
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.99,
|
| 1567 |
+
"learning_rate": 8.150893734694899e-06,
|
| 1568 |
+
"loss": 5.6854,
|
| 1569 |
+
"step": 121000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 1.0,
|
| 1573 |
+
"learning_rate": 8.142564427193524e-06,
|
| 1574 |
+
"loss": 5.6752,
|
| 1575 |
+
"step": 121500
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 1.0,
|
| 1579 |
+
"learning_rate": 8.13423511969215e-06,
|
| 1580 |
+
"loss": 5.6689,
|
| 1581 |
+
"step": 122000
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 1.0,
|
| 1585 |
+
"learning_rate": 8.125905812190775e-06,
|
| 1586 |
+
"loss": 5.6641,
|
| 1587 |
+
"step": 122500
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 1.01,
|
| 1591 |
+
"learning_rate": 8.1175765046894e-06,
|
| 1592 |
+
"loss": 5.6585,
|
| 1593 |
+
"step": 123000
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 1.01,
|
| 1597 |
+
"learning_rate": 8.109247197188025e-06,
|
| 1598 |
+
"loss": 5.6506,
|
| 1599 |
+
"step": 123500
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.02,
|
| 1603 |
+
"learning_rate": 8.100917889686652e-06,
|
| 1604 |
+
"loss": 5.6594,
|
| 1605 |
+
"step": 124000
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.02,
|
| 1609 |
+
"learning_rate": 8.092588582185277e-06,
|
| 1610 |
+
"loss": 5.6498,
|
| 1611 |
+
"step": 124500
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 1.02,
|
| 1615 |
+
"learning_rate": 8.084259274683904e-06,
|
| 1616 |
+
"loss": 5.6498,
|
| 1617 |
+
"step": 125000
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 1.03,
|
| 1621 |
+
"learning_rate": 8.075929967182529e-06,
|
| 1622 |
+
"loss": 5.6415,
|
| 1623 |
+
"step": 125500
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 1.03,
|
| 1627 |
+
"learning_rate": 8.067600659681156e-06,
|
| 1628 |
+
"loss": 5.6332,
|
| 1629 |
+
"step": 126000
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.04,
|
| 1633 |
+
"learning_rate": 8.05927135217978e-06,
|
| 1634 |
+
"loss": 5.6328,
|
| 1635 |
+
"step": 126500
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 1.04,
|
| 1639 |
+
"learning_rate": 8.050942044678406e-06,
|
| 1640 |
+
"loss": 5.6157,
|
| 1641 |
+
"step": 127000
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.04,
|
| 1645 |
+
"learning_rate": 8.04261273717703e-06,
|
| 1646 |
+
"loss": 5.6211,
|
| 1647 |
+
"step": 127500
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.05,
|
| 1651 |
+
"learning_rate": 8.034283429675657e-06,
|
| 1652 |
+
"loss": 5.6171,
|
| 1653 |
+
"step": 128000
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 1.05,
|
| 1657 |
+
"learning_rate": 8.025954122174284e-06,
|
| 1658 |
+
"loss": 5.6092,
|
| 1659 |
+
"step": 128500
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 1.06,
|
| 1663 |
+
"learning_rate": 8.017624814672909e-06,
|
| 1664 |
+
"loss": 5.6144,
|
| 1665 |
+
"step": 129000
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 1.06,
|
| 1669 |
+
"learning_rate": 8.009295507171534e-06,
|
| 1670 |
+
"loss": 5.6023,
|
| 1671 |
+
"step": 129500
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 1.07,
|
| 1675 |
+
"learning_rate": 8.00096619967016e-06,
|
| 1676 |
+
"loss": 5.6026,
|
| 1677 |
+
"step": 130000
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 1.07,
|
| 1681 |
+
"eval_accuracy": 0.19102818031266838,
|
| 1682 |
+
"eval_loss": 5.534846305847168,
|
| 1683 |
+
"eval_runtime": 1782.1016,
|
| 1684 |
+
"eval_samples_per_second": 173.029,
|
| 1685 |
+
"eval_steps_per_second": 3.605,
|
| 1686 |
+
"step": 130000
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 1.07,
|
| 1690 |
+
"learning_rate": 7.992636892168786e-06,
|
| 1691 |
+
"loss": 5.5977,
|
| 1692 |
+
"step": 130500
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 1.07,
|
| 1696 |
+
"learning_rate": 7.98430758466741e-06,
|
| 1697 |
+
"loss": 5.5854,
|
| 1698 |
+
"step": 131000
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 1.08,
|
| 1702 |
+
"learning_rate": 7.975978277166037e-06,
|
| 1703 |
+
"loss": 5.5849,
|
| 1704 |
+
"step": 131500
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.08,
|
| 1708 |
+
"learning_rate": 7.967648969664662e-06,
|
| 1709 |
+
"loss": 5.5764,
|
| 1710 |
+
"step": 132000
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.09,
|
| 1714 |
+
"learning_rate": 7.959319662163289e-06,
|
| 1715 |
+
"loss": 5.5828,
|
| 1716 |
+
"step": 132500
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 1.09,
|
| 1720 |
+
"learning_rate": 7.950990354661914e-06,
|
| 1721 |
+
"loss": 5.577,
|
| 1722 |
+
"step": 133000
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 1.09,
|
| 1726 |
+
"learning_rate": 7.942661047160539e-06,
|
| 1727 |
+
"loss": 5.5639,
|
| 1728 |
+
"step": 133500
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 1.1,
|
| 1732 |
+
"learning_rate": 7.934331739659166e-06,
|
| 1733 |
+
"loss": 5.5682,
|
| 1734 |
+
"step": 134000
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 1.1,
|
| 1738 |
+
"learning_rate": 7.92600243215779e-06,
|
| 1739 |
+
"loss": 5.5651,
|
| 1740 |
+
"step": 134500
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 1.11,
|
| 1744 |
+
"learning_rate": 7.917673124656416e-06,
|
| 1745 |
+
"loss": 5.5511,
|
| 1746 |
+
"step": 135000
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.11,
|
| 1750 |
+
"learning_rate": 7.909343817155042e-06,
|
| 1751 |
+
"loss": 5.5536,
|
| 1752 |
+
"step": 135500
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.11,
|
| 1756 |
+
"learning_rate": 7.901014509653669e-06,
|
| 1757 |
+
"loss": 5.5439,
|
| 1758 |
+
"step": 136000
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 1.12,
|
| 1762 |
+
"learning_rate": 7.892685202152294e-06,
|
| 1763 |
+
"loss": 5.5373,
|
| 1764 |
+
"step": 136500
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 1.12,
|
| 1768 |
+
"learning_rate": 7.884355894650919e-06,
|
| 1769 |
+
"loss": 5.5369,
|
| 1770 |
+
"step": 137000
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 1.13,
|
| 1774 |
+
"learning_rate": 7.876026587149544e-06,
|
| 1775 |
+
"loss": 5.5323,
|
| 1776 |
+
"step": 137500
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 1.13,
|
| 1780 |
+
"learning_rate": 7.86769727964817e-06,
|
| 1781 |
+
"loss": 5.5285,
|
| 1782 |
+
"step": 138000
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 1.13,
|
| 1786 |
+
"learning_rate": 7.859367972146796e-06,
|
| 1787 |
+
"loss": 5.526,
|
| 1788 |
+
"step": 138500
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.14,
|
| 1792 |
+
"learning_rate": 7.851038664645423e-06,
|
| 1793 |
+
"loss": 5.5149,
|
| 1794 |
+
"step": 139000
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.14,
|
| 1798 |
+
"learning_rate": 7.842709357144048e-06,
|
| 1799 |
+
"loss": 5.5194,
|
| 1800 |
+
"step": 139500
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 1.15,
|
| 1804 |
+
"learning_rate": 7.834380049642674e-06,
|
| 1805 |
+
"loss": 5.5066,
|
| 1806 |
+
"step": 140000
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 1.15,
|
| 1810 |
+
"eval_accuracy": 0.20347976075310614,
|
| 1811 |
+
"eval_loss": 5.432919979095459,
|
| 1812 |
+
"eval_runtime": 1769.3726,
|
| 1813 |
+
"eval_samples_per_second": 174.274,
|
| 1814 |
+
"eval_steps_per_second": 3.631,
|
| 1815 |
+
"step": 140000
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.15,
|
| 1819 |
+
"learning_rate": 7.8260507421413e-06,
|
| 1820 |
+
"loss": 5.501,
|
| 1821 |
+
"step": 140500
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 1.16,
|
| 1825 |
+
"learning_rate": 7.817721434639924e-06,
|
| 1826 |
+
"loss": 5.5057,
|
| 1827 |
+
"step": 141000
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 1.16,
|
| 1831 |
+
"learning_rate": 7.809392127138551e-06,
|
| 1832 |
+
"loss": 5.4965,
|
| 1833 |
+
"step": 141500
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 1.16,
|
| 1837 |
+
"learning_rate": 7.801062819637176e-06,
|
| 1838 |
+
"loss": 5.4912,
|
| 1839 |
+
"step": 142000
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 1.17,
|
| 1843 |
+
"learning_rate": 7.792733512135801e-06,
|
| 1844 |
+
"loss": 5.4938,
|
| 1845 |
+
"step": 142500
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 1.17,
|
| 1849 |
+
"learning_rate": 7.784404204634428e-06,
|
| 1850 |
+
"loss": 5.4841,
|
| 1851 |
+
"step": 143000
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.18,
|
| 1855 |
+
"learning_rate": 7.776074897133054e-06,
|
| 1856 |
+
"loss": 5.4816,
|
| 1857 |
+
"step": 143500
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.18,
|
| 1861 |
+
"learning_rate": 7.76774558963168e-06,
|
| 1862 |
+
"loss": 5.4753,
|
| 1863 |
+
"step": 144000
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 1.18,
|
| 1867 |
+
"learning_rate": 7.759416282130304e-06,
|
| 1868 |
+
"loss": 5.4732,
|
| 1869 |
+
"step": 144500
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 1.19,
|
| 1873 |
+
"learning_rate": 7.75108697462893e-06,
|
| 1874 |
+
"loss": 5.4726,
|
| 1875 |
+
"step": 145000
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 1.19,
|
| 1879 |
+
"learning_rate": 7.742757667127556e-06,
|
| 1880 |
+
"loss": 5.4659,
|
| 1881 |
+
"step": 145500
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 1.2,
|
| 1885 |
+
"learning_rate": 7.734428359626181e-06,
|
| 1886 |
+
"loss": 5.4565,
|
| 1887 |
+
"step": 146000
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 1.2,
|
| 1891 |
+
"learning_rate": 7.726099052124808e-06,
|
| 1892 |
+
"loss": 5.4571,
|
| 1893 |
+
"step": 146500
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.2,
|
| 1897 |
+
"learning_rate": 7.717769744623433e-06,
|
| 1898 |
+
"loss": 5.4561,
|
| 1899 |
+
"step": 147000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.21,
|
| 1903 |
+
"learning_rate": 7.70944043712206e-06,
|
| 1904 |
+
"loss": 5.4514,
|
| 1905 |
+
"step": 147500
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 1.21,
|
| 1909 |
+
"learning_rate": 7.701111129620684e-06,
|
| 1910 |
+
"loss": 5.4428,
|
| 1911 |
+
"step": 148000
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 1.22,
|
| 1915 |
+
"learning_rate": 7.69278182211931e-06,
|
| 1916 |
+
"loss": 5.4455,
|
| 1917 |
+
"step": 148500
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 1.22,
|
| 1921 |
+
"learning_rate": 7.684452514617934e-06,
|
| 1922 |
+
"loss": 5.4338,
|
| 1923 |
+
"step": 149000
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 1.22,
|
| 1927 |
+
"learning_rate": 7.676123207116561e-06,
|
| 1928 |
+
"loss": 5.4338,
|
| 1929 |
+
"step": 149500
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 1.23,
|
| 1933 |
+
"learning_rate": 7.667793899615186e-06,
|
| 1934 |
+
"loss": 5.4294,
|
| 1935 |
+
"step": 150000
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.23,
|
| 1939 |
+
"eval_accuracy": 0.2143715688125589,
|
| 1940 |
+
"eval_loss": 5.332610607147217,
|
| 1941 |
+
"eval_runtime": 1768.6133,
|
| 1942 |
+
"eval_samples_per_second": 174.349,
|
| 1943 |
+
"eval_steps_per_second": 3.633,
|
| 1944 |
+
"step": 150000
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 1.23,
|
| 1948 |
+
"learning_rate": 7.659464592113813e-06,
|
| 1949 |
+
"loss": 5.4237,
|
| 1950 |
+
"step": 150500
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 1.24,
|
| 1954 |
+
"learning_rate": 7.651135284612438e-06,
|
| 1955 |
+
"loss": 5.4188,
|
| 1956 |
+
"step": 151000
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.24,
|
| 1960 |
+
"learning_rate": 7.642805977111065e-06,
|
| 1961 |
+
"loss": 5.4166,
|
| 1962 |
+
"step": 151500
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.25,
|
| 1966 |
+
"learning_rate": 7.63447666960969e-06,
|
| 1967 |
+
"loss": 5.4135,
|
| 1968 |
+
"step": 152000
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 1.25,
|
| 1972 |
+
"learning_rate": 7.6261473621083145e-06,
|
| 1973 |
+
"loss": 5.4115,
|
| 1974 |
+
"step": 152500
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 1.25,
|
| 1978 |
+
"learning_rate": 7.61781805460694e-06,
|
| 1979 |
+
"loss": 5.4036,
|
| 1980 |
+
"step": 153000
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 1.26,
|
| 1984 |
+
"learning_rate": 7.609488747105567e-06,
|
| 1985 |
+
"loss": 5.4039,
|
| 1986 |
+
"step": 153500
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 1.26,
|
| 1990 |
+
"learning_rate": 7.601159439604192e-06,
|
| 1991 |
+
"loss": 5.395,
|
| 1992 |
+
"step": 154000
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 1.27,
|
| 1996 |
+
"learning_rate": 7.592830132102817e-06,
|
| 1997 |
+
"loss": 5.3897,
|
| 1998 |
+
"step": 154500
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.27,
|
| 2002 |
+
"learning_rate": 7.584500824601443e-06,
|
| 2003 |
+
"loss": 5.3822,
|
| 2004 |
+
"step": 155000
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.27,
|
| 2008 |
+
"learning_rate": 7.57617151710007e-06,
|
| 2009 |
+
"loss": 5.379,
|
| 2010 |
+
"step": 155500
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 1.28,
|
| 2014 |
+
"learning_rate": 7.567842209598695e-06,
|
| 2015 |
+
"loss": 5.3777,
|
| 2016 |
+
"step": 156000
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 1.28,
|
| 2020 |
+
"learning_rate": 7.5595129020973205e-06,
|
| 2021 |
+
"loss": 5.3705,
|
| 2022 |
+
"step": 156500
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 1.29,
|
| 2026 |
+
"learning_rate": 7.5511835945959455e-06,
|
| 2027 |
+
"loss": 5.3615,
|
| 2028 |
+
"step": 157000
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 1.29,
|
| 2032 |
+
"learning_rate": 7.542854287094572e-06,
|
| 2033 |
+
"loss": 5.3632,
|
| 2034 |
+
"step": 157500
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 1.29,
|
| 2038 |
+
"learning_rate": 7.534524979593197e-06,
|
| 2039 |
+
"loss": 5.3605,
|
| 2040 |
+
"step": 158000
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.3,
|
| 2044 |
+
"learning_rate": 7.526195672091823e-06,
|
| 2045 |
+
"loss": 5.3553,
|
| 2046 |
+
"step": 158500
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.3,
|
| 2050 |
+
"learning_rate": 7.517866364590448e-06,
|
| 2051 |
+
"loss": 5.3458,
|
| 2052 |
+
"step": 159000
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 1.31,
|
| 2056 |
+
"learning_rate": 7.509537057089075e-06,
|
| 2057 |
+
"loss": 5.3459,
|
| 2058 |
+
"step": 159500
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 1.31,
|
| 2062 |
+
"learning_rate": 7.5012077495877e-06,
|
| 2063 |
+
"loss": 5.3402,
|
| 2064 |
+
"step": 160000
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 1.31,
|
| 2068 |
+
"eval_accuracy": 0.2270330166918787,
|
| 2069 |
+
"eval_loss": 5.230414867401123,
|
| 2070 |
+
"eval_runtime": 1766.226,
|
| 2071 |
+
"eval_samples_per_second": 174.585,
|
| 2072 |
+
"eval_steps_per_second": 3.638,
|
| 2073 |
+
"step": 160000
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 1.31,
|
| 2077 |
+
"learning_rate": 7.492878442086326e-06,
|
| 2078 |
+
"loss": 5.3345,
|
| 2079 |
+
"step": 160500
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 1.32,
|
| 2083 |
+
"learning_rate": 7.484549134584951e-06,
|
| 2084 |
+
"loss": 5.3311,
|
| 2085 |
+
"step": 161000
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 1.32,
|
| 2089 |
+
"learning_rate": 7.476219827083577e-06,
|
| 2090 |
+
"loss": 5.3308,
|
| 2091 |
+
"step": 161500
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 1.33,
|
| 2095 |
+
"learning_rate": 7.467890519582202e-06,
|
| 2096 |
+
"loss": 5.3175,
|
| 2097 |
+
"step": 162000
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 1.33,
|
| 2101 |
+
"learning_rate": 7.459561212080828e-06,
|
| 2102 |
+
"loss": 5.3256,
|
| 2103 |
+
"step": 162500
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.34,
|
| 2107 |
+
"learning_rate": 7.451231904579453e-06,
|
| 2108 |
+
"loss": 5.3168,
|
| 2109 |
+
"step": 163000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.34,
|
| 2113 |
+
"learning_rate": 7.44290259707808e-06,
|
| 2114 |
+
"loss": 5.3122,
|
| 2115 |
+
"step": 163500
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 1.34,
|
| 2119 |
+
"learning_rate": 7.434573289576706e-06,
|
| 2120 |
+
"loss": 5.3019,
|
| 2121 |
+
"step": 164000
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 1.35,
|
| 2125 |
+
"learning_rate": 7.426243982075331e-06,
|
| 2126 |
+
"loss": 5.2948,
|
| 2127 |
+
"step": 164500
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 1.35,
|
| 2131 |
+
"learning_rate": 7.417914674573956e-06,
|
| 2132 |
+
"loss": 5.2968,
|
| 2133 |
+
"step": 165000
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 1.36,
|
| 2137 |
+
"learning_rate": 7.409585367072582e-06,
|
| 2138 |
+
"loss": 5.2876,
|
| 2139 |
+
"step": 165500
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 1.36,
|
| 2143 |
+
"learning_rate": 7.401256059571208e-06,
|
| 2144 |
+
"loss": 5.2824,
|
| 2145 |
+
"step": 166000
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.36,
|
| 2149 |
+
"learning_rate": 7.392926752069833e-06,
|
| 2150 |
+
"loss": 5.2717,
|
| 2151 |
+
"step": 166500
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.37,
|
| 2155 |
+
"learning_rate": 7.384597444568459e-06,
|
| 2156 |
+
"loss": 5.2779,
|
| 2157 |
+
"step": 167000
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 1.37,
|
| 2161 |
+
"learning_rate": 7.376268137067085e-06,
|
| 2162 |
+
"loss": 5.2697,
|
| 2163 |
+
"step": 167500
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 1.38,
|
| 2167 |
+
"learning_rate": 7.367938829565711e-06,
|
| 2168 |
+
"loss": 5.2618,
|
| 2169 |
+
"step": 168000
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 1.38,
|
| 2173 |
+
"learning_rate": 7.359609522064336e-06,
|
| 2174 |
+
"loss": 5.2582,
|
| 2175 |
+
"step": 168500
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 1.38,
|
| 2179 |
+
"learning_rate": 7.351280214562962e-06,
|
| 2180 |
+
"loss": 5.253,
|
| 2181 |
+
"step": 169000
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 1.39,
|
| 2185 |
+
"learning_rate": 7.3429509070615875e-06,
|
| 2186 |
+
"loss": 5.2486,
|
| 2187 |
+
"step": 169500
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.39,
|
| 2191 |
+
"learning_rate": 7.334621599560213e-06,
|
| 2192 |
+
"loss": 5.2397,
|
| 2193 |
+
"step": 170000
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.39,
|
| 2197 |
+
"eval_accuracy": 0.2406091159386724,
|
| 2198 |
+
"eval_loss": 5.117036819458008,
|
| 2199 |
+
"eval_runtime": 1771.1496,
|
| 2200 |
+
"eval_samples_per_second": 174.099,
|
| 2201 |
+
"eval_steps_per_second": 3.628,
|
| 2202 |
+
"step": 170000
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 1.4,
|
| 2206 |
+
"learning_rate": 7.326292292058838e-06,
|
| 2207 |
+
"loss": 5.2303,
|
| 2208 |
+
"step": 170500
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 1.4,
|
| 2212 |
+
"learning_rate": 7.317962984557464e-06,
|
| 2213 |
+
"loss": 5.2302,
|
| 2214 |
+
"step": 171000
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.41,
|
| 2218 |
+
"learning_rate": 7.309633677056091e-06,
|
| 2219 |
+
"loss": 5.2246,
|
| 2220 |
+
"step": 171500
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 1.41,
|
| 2224 |
+
"learning_rate": 7.301304369554716e-06,
|
| 2225 |
+
"loss": 5.2181,
|
| 2226 |
+
"step": 172000
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 1.41,
|
| 2230 |
+
"learning_rate": 7.292975062053341e-06,
|
| 2231 |
+
"loss": 5.2146,
|
| 2232 |
+
"step": 172500
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 1.42,
|
| 2236 |
+
"learning_rate": 7.284645754551967e-06,
|
| 2237 |
+
"loss": 5.2077,
|
| 2238 |
+
"step": 173000
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 1.42,
|
| 2242 |
+
"learning_rate": 7.2763164470505934e-06,
|
| 2243 |
+
"loss": 5.2077,
|
| 2244 |
+
"step": 173500
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 1.43,
|
| 2248 |
+
"learning_rate": 7.2679871395492184e-06,
|
| 2249 |
+
"loss": 5.2002,
|
| 2250 |
+
"step": 174000
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 1.43,
|
| 2254 |
+
"learning_rate": 7.259657832047844e-06,
|
| 2255 |
+
"loss": 5.1931,
|
| 2256 |
+
"step": 174500
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.43,
|
| 2260 |
+
"learning_rate": 7.251328524546469e-06,
|
| 2261 |
+
"loss": 5.188,
|
| 2262 |
+
"step": 175000
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 1.44,
|
| 2266 |
+
"learning_rate": 7.242999217045096e-06,
|
| 2267 |
+
"loss": 5.1806,
|
| 2268 |
+
"step": 175500
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 1.44,
|
| 2272 |
+
"learning_rate": 7.234669909543721e-06,
|
| 2273 |
+
"loss": 5.1811,
|
| 2274 |
+
"step": 176000
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 1.45,
|
| 2278 |
+
"learning_rate": 7.226340602042347e-06,
|
| 2279 |
+
"loss": 5.17,
|
| 2280 |
+
"step": 176500
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 1.45,
|
| 2284 |
+
"learning_rate": 7.218011294540972e-06,
|
| 2285 |
+
"loss": 5.1635,
|
| 2286 |
+
"step": 177000
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 1.45,
|
| 2290 |
+
"learning_rate": 7.2096819870395985e-06,
|
| 2291 |
+
"loss": 5.157,
|
| 2292 |
+
"step": 177500
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 1.46,
|
| 2296 |
+
"learning_rate": 7.2013526795382235e-06,
|
| 2297 |
+
"loss": 5.1505,
|
| 2298 |
+
"step": 178000
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 1.46,
|
| 2302 |
+
"learning_rate": 7.193023372036849e-06,
|
| 2303 |
+
"loss": 5.1446,
|
| 2304 |
+
"step": 178500
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 1.47,
|
| 2308 |
+
"learning_rate": 7.184694064535474e-06,
|
| 2309 |
+
"loss": 5.1397,
|
| 2310 |
+
"step": 179000
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 1.47,
|
| 2314 |
+
"learning_rate": 7.176364757034101e-06,
|
| 2315 |
+
"loss": 5.1393,
|
| 2316 |
+
"step": 179500
|
| 2317 |
+
},
|
| 2318 |
+
{
|
| 2319 |
+
"epoch": 1.47,
|
| 2320 |
+
"learning_rate": 7.168035449532726e-06,
|
| 2321 |
+
"loss": 5.1356,
|
| 2322 |
+
"step": 180000
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 1.47,
|
| 2326 |
+
"eval_accuracy": 0.2563928525528852,
|
| 2327 |
+
"eval_loss": 4.97933292388916,
|
| 2328 |
+
"eval_runtime": 1777.4907,
|
| 2329 |
+
"eval_samples_per_second": 173.478,
|
| 2330 |
+
"eval_steps_per_second": 3.615,
|
| 2331 |
+
"step": 180000
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 1.48,
|
| 2335 |
+
"learning_rate": 7.159706142031352e-06,
|
| 2336 |
+
"loss": 5.12,
|
| 2337 |
+
"step": 180500
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 1.48,
|
| 2341 |
+
"learning_rate": 7.151376834529977e-06,
|
| 2342 |
+
"loss": 5.1107,
|
| 2343 |
+
"step": 181000
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 1.49,
|
| 2347 |
+
"learning_rate": 7.143047527028604e-06,
|
| 2348 |
+
"loss": 5.1081,
|
| 2349 |
+
"step": 181500
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 1.49,
|
| 2353 |
+
"learning_rate": 7.1347182195272295e-06,
|
| 2354 |
+
"loss": 5.108,
|
| 2355 |
+
"step": 182000
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 1.5,
|
| 2359 |
+
"learning_rate": 7.1263889120258545e-06,
|
| 2360 |
+
"loss": 5.099,
|
| 2361 |
+
"step": 182500
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 1.5,
|
| 2365 |
+
"learning_rate": 7.118059604524481e-06,
|
| 2366 |
+
"loss": 5.0923,
|
| 2367 |
+
"step": 183000
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 1.5,
|
| 2371 |
+
"learning_rate": 7.109730297023106e-06,
|
| 2372 |
+
"loss": 5.088,
|
| 2373 |
+
"step": 183500
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 1.51,
|
| 2377 |
+
"learning_rate": 7.101400989521732e-06,
|
| 2378 |
+
"loss": 5.0822,
|
| 2379 |
+
"step": 184000
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 1.51,
|
| 2383 |
+
"learning_rate": 7.093071682020357e-06,
|
| 2384 |
+
"loss": 5.0788,
|
| 2385 |
+
"step": 184500
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 1.52,
|
| 2389 |
+
"learning_rate": 7.084742374518984e-06,
|
| 2390 |
+
"loss": 5.0631,
|
| 2391 |
+
"step": 185000
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 1.52,
|
| 2395 |
+
"learning_rate": 7.076413067017609e-06,
|
| 2396 |
+
"loss": 5.0649,
|
| 2397 |
+
"step": 185500
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 1.52,
|
| 2401 |
+
"learning_rate": 7.068083759516235e-06,
|
| 2402 |
+
"loss": 5.0524,
|
| 2403 |
+
"step": 186000
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 1.53,
|
| 2407 |
+
"learning_rate": 7.05975445201486e-06,
|
| 2408 |
+
"loss": 5.0413,
|
| 2409 |
+
"step": 186500
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 1.53,
|
| 2413 |
+
"learning_rate": 7.051425144513486e-06,
|
| 2414 |
+
"loss": 5.0458,
|
| 2415 |
+
"step": 187000
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 1.54,
|
| 2419 |
+
"learning_rate": 7.043095837012111e-06,
|
| 2420 |
+
"loss": 5.0394,
|
| 2421 |
+
"step": 187500
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 1.54,
|
| 2425 |
+
"learning_rate": 7.034766529510737e-06,
|
| 2426 |
+
"loss": 5.0297,
|
| 2427 |
+
"step": 188000
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 1.54,
|
| 2431 |
+
"learning_rate": 7.026437222009362e-06,
|
| 2432 |
+
"loss": 5.0247,
|
| 2433 |
+
"step": 188500
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 1.55,
|
| 2437 |
+
"learning_rate": 7.018107914507989e-06,
|
| 2438 |
+
"loss": 5.0194,
|
| 2439 |
+
"step": 189000
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 1.55,
|
| 2443 |
+
"learning_rate": 7.009778607006615e-06,
|
| 2444 |
+
"loss": 5.0154,
|
| 2445 |
+
"step": 189500
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 1.56,
|
| 2449 |
+
"learning_rate": 7.00144929950524e-06,
|
| 2450 |
+
"loss": 5.0099,
|
| 2451 |
+
"step": 190000
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 1.56,
|
| 2455 |
+
"eval_accuracy": 0.2730389650205668,
|
| 2456 |
+
"eval_loss": 4.837229251861572,
|
| 2457 |
+
"eval_runtime": 1776.3101,
|
| 2458 |
+
"eval_samples_per_second": 173.594,
|
| 2459 |
+
"eval_steps_per_second": 3.617,
|
| 2460 |
+
"step": 190000
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 1.56,
|
| 2464 |
+
"learning_rate": 6.993119992003865e-06,
|
| 2465 |
+
"loss": 5.0008,
|
| 2466 |
+
"step": 190500
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 1.56,
|
| 2470 |
+
"learning_rate": 6.984790684502491e-06,
|
| 2471 |
+
"loss": 4.9927,
|
| 2472 |
+
"step": 191000
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 1.57,
|
| 2476 |
+
"learning_rate": 6.976461377001117e-06,
|
| 2477 |
+
"loss": 4.983,
|
| 2478 |
+
"step": 191500
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 1.57,
|
| 2482 |
+
"learning_rate": 6.968132069499742e-06,
|
| 2483 |
+
"loss": 4.9762,
|
| 2484 |
+
"step": 192000
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 1.58,
|
| 2488 |
+
"learning_rate": 6.959802761998368e-06,
|
| 2489 |
+
"loss": 4.9734,
|
| 2490 |
+
"step": 192500
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 1.58,
|
| 2494 |
+
"learning_rate": 6.951473454496994e-06,
|
| 2495 |
+
"loss": 4.9717,
|
| 2496 |
+
"step": 193000
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 1.59,
|
| 2500 |
+
"learning_rate": 6.94314414699562e-06,
|
| 2501 |
+
"loss": 4.9597,
|
| 2502 |
+
"step": 193500
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 1.59,
|
| 2506 |
+
"learning_rate": 6.934814839494245e-06,
|
| 2507 |
+
"loss": 4.9551,
|
| 2508 |
+
"step": 194000
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 1.59,
|
| 2512 |
+
"learning_rate": 6.926485531992871e-06,
|
| 2513 |
+
"loss": 4.9559,
|
| 2514 |
+
"step": 194500
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 1.6,
|
| 2518 |
+
"learning_rate": 6.9181562244914965e-06,
|
| 2519 |
+
"loss": 4.9434,
|
| 2520 |
+
"step": 195000
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 1.6,
|
| 2524 |
+
"learning_rate": 6.909826916990122e-06,
|
| 2525 |
+
"loss": 4.9297,
|
| 2526 |
+
"step": 195500
|
| 2527 |
+
},
|
| 2528 |
+
{
|
| 2529 |
+
"epoch": 1.61,
|
| 2530 |
+
"learning_rate": 6.901497609488747e-06,
|
| 2531 |
+
"loss": 4.9285,
|
| 2532 |
+
"step": 196000
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 1.61,
|
| 2536 |
+
"learning_rate": 6.893168301987373e-06,
|
| 2537 |
+
"loss": 4.9264,
|
| 2538 |
+
"step": 196500
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 1.61,
|
| 2542 |
+
"learning_rate": 6.884838994485999e-06,
|
| 2543 |
+
"loss": 4.9244,
|
| 2544 |
+
"step": 197000
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 1.62,
|
| 2548 |
+
"learning_rate": 6.876509686984625e-06,
|
| 2549 |
+
"loss": 4.9106,
|
| 2550 |
+
"step": 197500
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 1.62,
|
| 2554 |
+
"learning_rate": 6.86818037948325e-06,
|
| 2555 |
+
"loss": 4.9086,
|
| 2556 |
+
"step": 198000
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 1.63,
|
| 2560 |
+
"learning_rate": 6.859851071981876e-06,
|
| 2561 |
+
"loss": 4.9002,
|
| 2562 |
+
"step": 198500
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 1.63,
|
| 2566 |
+
"learning_rate": 6.8515217644805024e-06,
|
| 2567 |
+
"loss": 4.896,
|
| 2568 |
+
"step": 199000
|
| 2569 |
+
},
|
| 2570 |
+
{
|
| 2571 |
+
"epoch": 1.63,
|
| 2572 |
+
"learning_rate": 6.8431924569791274e-06,
|
| 2573 |
+
"loss": 4.8849,
|
| 2574 |
+
"step": 199500
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 1.64,
|
| 2578 |
+
"learning_rate": 6.8348631494777525e-06,
|
| 2579 |
+
"loss": 4.885,
|
| 2580 |
+
"step": 200000
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 1.64,
|
| 2584 |
+
"eval_accuracy": 0.28782873953338056,
|
| 2585 |
+
"eval_loss": 4.705814361572266,
|
| 2586 |
+
"eval_runtime": 1797.5806,
|
| 2587 |
+
"eval_samples_per_second": 171.539,
|
| 2588 |
+
"eval_steps_per_second": 3.574,
|
| 2589 |
+
"step": 200000
|
| 2590 |
+
},
|
| 2591 |
+
{
|
| 2592 |
+
"epoch": 1.64,
|
| 2593 |
+
"learning_rate": 6.826533841976378e-06,
|
| 2594 |
+
"loss": 4.8816,
|
| 2595 |
+
"step": 200500
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 1.65,
|
| 2599 |
+
"learning_rate": 6.818204534475005e-06,
|
| 2600 |
+
"loss": 4.8729,
|
| 2601 |
+
"step": 201000
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 1.65,
|
| 2605 |
+
"learning_rate": 6.80987522697363e-06,
|
| 2606 |
+
"loss": 4.865,
|
| 2607 |
+
"step": 201500
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 1.65,
|
| 2611 |
+
"learning_rate": 6.801545919472256e-06,
|
| 2612 |
+
"loss": 4.8596,
|
| 2613 |
+
"step": 202000
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 1.66,
|
| 2617 |
+
"learning_rate": 6.793216611970881e-06,
|
| 2618 |
+
"loss": 4.8521,
|
| 2619 |
+
"step": 202500
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 1.66,
|
| 2623 |
+
"learning_rate": 6.7848873044695075e-06,
|
| 2624 |
+
"loss": 4.8476,
|
| 2625 |
+
"step": 203000
|
| 2626 |
+
},
|
| 2627 |
+
{
|
| 2628 |
+
"epoch": 1.67,
|
| 2629 |
+
"learning_rate": 6.7765579969681325e-06,
|
| 2630 |
+
"loss": 4.8398,
|
| 2631 |
+
"step": 203500
|
| 2632 |
+
},
|
| 2633 |
+
{
|
| 2634 |
+
"epoch": 1.67,
|
| 2635 |
+
"learning_rate": 6.768228689466758e-06,
|
| 2636 |
+
"loss": 4.835,
|
| 2637 |
+
"step": 204000
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 1.68,
|
| 2641 |
+
"learning_rate": 6.759899381965383e-06,
|
| 2642 |
+
"loss": 4.8266,
|
| 2643 |
+
"step": 204500
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 1.68,
|
| 2647 |
+
"step": 204833,
|
| 2648 |
+
"total_flos": 2.843961674711433e+18,
|
| 2649 |
+
"train_loss": 5.814965769895094,
|
| 2650 |
+
"train_runtime": 198000.2186,
|
| 2651 |
+
"train_samples_per_second": 147.948,
|
| 2652 |
+
"train_steps_per_second": 3.082
|
| 2653 |
+
}
|
| 2654 |
+
],
|
| 2655 |
+
"logging_steps": 500,
|
| 2656 |
+
"max_steps": 610290,
|
| 2657 |
+
"num_train_epochs": 5,
|
| 2658 |
+
"save_steps": 10000,
|
| 2659 |
+
"total_flos": 2.843961674711433e+18,
|
| 2660 |
+
"trial_name": null,
|
| 2661 |
+
"trial_params": null
|
| 2662 |
+
}
|