End of training
Browse files- README.md +2 -1
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +3179 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: llama3.1
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3.1-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: oh_scale_x2_compute_equal
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# oh_scale_x2_compute_equal
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.9233
|
| 21 |
|
|
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3.1-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: oh_scale_x2_compute_equal
|
|
|
|
| 16 |
|
| 17 |
# oh_scale_x2_compute_equal
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on the mlfoundations-dev/oh-dcft-v1.3_no-curation_gpt-4o-mini_scale_2x dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.9233
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.9963586709148835,
|
| 3 |
+
"eval_loss": 0.9232881665229797,
|
| 4 |
+
"eval_runtime": 379.7133,
|
| 5 |
+
"eval_samples_per_second": 38.979,
|
| 6 |
+
"eval_steps_per_second": 0.611,
|
| 7 |
+
"total_flos": 7356554202316800.0,
|
| 8 |
+
"train_loss": 0.6099752168050664,
|
| 9 |
+
"train_runtime": 144891.7905,
|
| 10 |
+
"train_samples_per_second": 15.527,
|
| 11 |
+
"train_steps_per_second": 0.03
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.9963586709148835,
|
| 3 |
+
"eval_loss": 0.9232881665229797,
|
| 4 |
+
"eval_runtime": 379.7133,
|
| 5 |
+
"eval_samples_per_second": 38.979,
|
| 6 |
+
"eval_steps_per_second": 0.611
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.9963586709148835,
|
| 3 |
+
"total_flos": 7356554202316800.0,
|
| 4 |
+
"train_loss": 0.6099752168050664,
|
| 5 |
+
"train_runtime": 144891.7905,
|
| 6 |
+
"train_samples_per_second": 15.527,
|
| 7 |
+
"train_steps_per_second": 0.03
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3179 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 7.9963586709148835,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4392,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.018206645425580335,
|
| 13 |
+
"grad_norm": 15.35500484249196,
|
| 14 |
+
"learning_rate": 5e-06,
|
| 15 |
+
"loss": 1.0839,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03641329085116067,
|
| 20 |
+
"grad_norm": 2.618926418245599,
|
| 21 |
+
"learning_rate": 5e-06,
|
| 22 |
+
"loss": 0.9627,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05461993627674101,
|
| 27 |
+
"grad_norm": 2.080636783209041,
|
| 28 |
+
"learning_rate": 5e-06,
|
| 29 |
+
"loss": 0.9308,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07282658170232134,
|
| 34 |
+
"grad_norm": 1.436134442652753,
|
| 35 |
+
"learning_rate": 5e-06,
|
| 36 |
+
"loss": 0.897,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09103322712790168,
|
| 41 |
+
"grad_norm": 1.581598449271548,
|
| 42 |
+
"learning_rate": 5e-06,
|
| 43 |
+
"loss": 0.8783,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10923987255348203,
|
| 48 |
+
"grad_norm": 1.3764881165967209,
|
| 49 |
+
"learning_rate": 5e-06,
|
| 50 |
+
"loss": 0.8594,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12744651797906237,
|
| 55 |
+
"grad_norm": 0.8508264971874785,
|
| 56 |
+
"learning_rate": 5e-06,
|
| 57 |
+
"loss": 0.8476,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14565316340464268,
|
| 62 |
+
"grad_norm": 0.9826078435096594,
|
| 63 |
+
"learning_rate": 5e-06,
|
| 64 |
+
"loss": 0.8351,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16385980883022302,
|
| 69 |
+
"grad_norm": 0.7948332220104163,
|
| 70 |
+
"learning_rate": 5e-06,
|
| 71 |
+
"loss": 0.8261,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18206645425580337,
|
| 76 |
+
"grad_norm": 0.7335767518254585,
|
| 77 |
+
"learning_rate": 5e-06,
|
| 78 |
+
"loss": 0.8185,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2002730996813837,
|
| 83 |
+
"grad_norm": 1.0534204936405038,
|
| 84 |
+
"learning_rate": 5e-06,
|
| 85 |
+
"loss": 0.8204,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21847974510696405,
|
| 90 |
+
"grad_norm": 0.847321277840478,
|
| 91 |
+
"learning_rate": 5e-06,
|
| 92 |
+
"loss": 0.8108,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23668639053254437,
|
| 97 |
+
"grad_norm": 0.6210895897293146,
|
| 98 |
+
"learning_rate": 5e-06,
|
| 99 |
+
"loss": 0.8084,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.25489303595812474,
|
| 104 |
+
"grad_norm": 0.5781940572538745,
|
| 105 |
+
"learning_rate": 5e-06,
|
| 106 |
+
"loss": 0.8103,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.27309968138370505,
|
| 111 |
+
"grad_norm": 0.9063984151725509,
|
| 112 |
+
"learning_rate": 5e-06,
|
| 113 |
+
"loss": 0.8077,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.29130632680928537,
|
| 118 |
+
"grad_norm": 0.6571941221537146,
|
| 119 |
+
"learning_rate": 5e-06,
|
| 120 |
+
"loss": 0.799,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.30951297223486574,
|
| 125 |
+
"grad_norm": 0.5952826448438556,
|
| 126 |
+
"learning_rate": 5e-06,
|
| 127 |
+
"loss": 0.7999,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.32771961766044605,
|
| 132 |
+
"grad_norm": 1.4601461457218337,
|
| 133 |
+
"learning_rate": 5e-06,
|
| 134 |
+
"loss": 0.8004,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3459262630860264,
|
| 139 |
+
"grad_norm": 1.3162465370008012,
|
| 140 |
+
"learning_rate": 5e-06,
|
| 141 |
+
"loss": 0.7973,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.36413290851160673,
|
| 146 |
+
"grad_norm": 1.2972378846813408,
|
| 147 |
+
"learning_rate": 5e-06,
|
| 148 |
+
"loss": 0.7959,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.38233955393718705,
|
| 153 |
+
"grad_norm": 0.9152791184028909,
|
| 154 |
+
"learning_rate": 5e-06,
|
| 155 |
+
"loss": 0.7892,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4005461993627674,
|
| 160 |
+
"grad_norm": 0.823555803312372,
|
| 161 |
+
"learning_rate": 5e-06,
|
| 162 |
+
"loss": 0.7932,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.41875284478834773,
|
| 167 |
+
"grad_norm": 0.7145315436546384,
|
| 168 |
+
"learning_rate": 5e-06,
|
| 169 |
+
"loss": 0.7848,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4369594902139281,
|
| 174 |
+
"grad_norm": 0.5776875953876701,
|
| 175 |
+
"learning_rate": 5e-06,
|
| 176 |
+
"loss": 0.791,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4551661356395084,
|
| 181 |
+
"grad_norm": 0.6846063811362885,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 0.7875,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.47337278106508873,
|
| 188 |
+
"grad_norm": 0.7232578149621245,
|
| 189 |
+
"learning_rate": 5e-06,
|
| 190 |
+
"loss": 0.7843,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4915794264906691,
|
| 195 |
+
"grad_norm": 0.6962364276167124,
|
| 196 |
+
"learning_rate": 5e-06,
|
| 197 |
+
"loss": 0.7839,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5097860719162495,
|
| 202 |
+
"grad_norm": 0.8881236192032435,
|
| 203 |
+
"learning_rate": 5e-06,
|
| 204 |
+
"loss": 0.779,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5279927173418297,
|
| 209 |
+
"grad_norm": 0.6192460064199938,
|
| 210 |
+
"learning_rate": 5e-06,
|
| 211 |
+
"loss": 0.7821,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5461993627674101,
|
| 216 |
+
"grad_norm": 0.7429540921229376,
|
| 217 |
+
"learning_rate": 5e-06,
|
| 218 |
+
"loss": 0.7771,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5644060081929905,
|
| 223 |
+
"grad_norm": 0.5678182946639223,
|
| 224 |
+
"learning_rate": 5e-06,
|
| 225 |
+
"loss": 0.7837,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5826126536185707,
|
| 230 |
+
"grad_norm": 0.7927883549153515,
|
| 231 |
+
"learning_rate": 5e-06,
|
| 232 |
+
"loss": 0.7795,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6008192990441511,
|
| 237 |
+
"grad_norm": 0.6630873675584549,
|
| 238 |
+
"learning_rate": 5e-06,
|
| 239 |
+
"loss": 0.7788,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6190259444697315,
|
| 244 |
+
"grad_norm": 0.8821281227865143,
|
| 245 |
+
"learning_rate": 5e-06,
|
| 246 |
+
"loss": 0.7817,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6372325898953118,
|
| 251 |
+
"grad_norm": 0.5862391322505137,
|
| 252 |
+
"learning_rate": 5e-06,
|
| 253 |
+
"loss": 0.7818,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6554392353208921,
|
| 258 |
+
"grad_norm": 0.7933124674797363,
|
| 259 |
+
"learning_rate": 5e-06,
|
| 260 |
+
"loss": 0.7714,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6736458807464725,
|
| 265 |
+
"grad_norm": 0.5912540694609474,
|
| 266 |
+
"learning_rate": 5e-06,
|
| 267 |
+
"loss": 0.7732,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6918525261720528,
|
| 272 |
+
"grad_norm": 0.5222298817170612,
|
| 273 |
+
"learning_rate": 5e-06,
|
| 274 |
+
"loss": 0.7697,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7100591715976331,
|
| 279 |
+
"grad_norm": 0.6154935564611252,
|
| 280 |
+
"learning_rate": 5e-06,
|
| 281 |
+
"loss": 0.7716,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7282658170232135,
|
| 286 |
+
"grad_norm": 0.7305900161607543,
|
| 287 |
+
"learning_rate": 5e-06,
|
| 288 |
+
"loss": 0.7708,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7464724624487938,
|
| 293 |
+
"grad_norm": 0.7798387910594561,
|
| 294 |
+
"learning_rate": 5e-06,
|
| 295 |
+
"loss": 0.7724,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7646791078743741,
|
| 300 |
+
"grad_norm": 1.100346410427369,
|
| 301 |
+
"learning_rate": 5e-06,
|
| 302 |
+
"loss": 0.7716,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7828857532999545,
|
| 307 |
+
"grad_norm": 0.657718484418182,
|
| 308 |
+
"learning_rate": 5e-06,
|
| 309 |
+
"loss": 0.7666,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.8010923987255348,
|
| 314 |
+
"grad_norm": 0.6103104885939142,
|
| 315 |
+
"learning_rate": 5e-06,
|
| 316 |
+
"loss": 0.765,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8192990441511152,
|
| 321 |
+
"grad_norm": 0.6884438533501283,
|
| 322 |
+
"learning_rate": 5e-06,
|
| 323 |
+
"loss": 0.7699,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8375056895766955,
|
| 328 |
+
"grad_norm": 0.6270481654771743,
|
| 329 |
+
"learning_rate": 5e-06,
|
| 330 |
+
"loss": 0.767,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8557123350022758,
|
| 335 |
+
"grad_norm": 0.5861612234613632,
|
| 336 |
+
"learning_rate": 5e-06,
|
| 337 |
+
"loss": 0.7645,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8739189804278562,
|
| 342 |
+
"grad_norm": 0.586876291200088,
|
| 343 |
+
"learning_rate": 5e-06,
|
| 344 |
+
"loss": 0.7686,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8921256258534365,
|
| 349 |
+
"grad_norm": 0.7466830111479166,
|
| 350 |
+
"learning_rate": 5e-06,
|
| 351 |
+
"loss": 0.7688,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9103322712790168,
|
| 356 |
+
"grad_norm": 0.6792188281050604,
|
| 357 |
+
"learning_rate": 5e-06,
|
| 358 |
+
"loss": 0.7685,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9285389167045972,
|
| 363 |
+
"grad_norm": 0.6846050616906328,
|
| 364 |
+
"learning_rate": 5e-06,
|
| 365 |
+
"loss": 0.7616,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9467455621301775,
|
| 370 |
+
"grad_norm": 0.7324962552104101,
|
| 371 |
+
"learning_rate": 5e-06,
|
| 372 |
+
"loss": 0.762,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9649522075557578,
|
| 377 |
+
"grad_norm": 0.6893904439545421,
|
| 378 |
+
"learning_rate": 5e-06,
|
| 379 |
+
"loss": 0.7664,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9831588529813382,
|
| 384 |
+
"grad_norm": 0.6529796330088168,
|
| 385 |
+
"learning_rate": 5e-06,
|
| 386 |
+
"loss": 0.7631,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.9995448338643604,
|
| 391 |
+
"eval_loss": 0.7638739347457886,
|
| 392 |
+
"eval_runtime": 379.9866,
|
| 393 |
+
"eval_samples_per_second": 38.951,
|
| 394 |
+
"eval_steps_per_second": 0.611,
|
| 395 |
+
"step": 549
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0013654984069185,
|
| 399 |
+
"grad_norm": 1.2061555179450747,
|
| 400 |
+
"learning_rate": 5e-06,
|
| 401 |
+
"loss": 0.812,
|
| 402 |
+
"step": 550
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.019572143832499,
|
| 406 |
+
"grad_norm": 0.8129440249983992,
|
| 407 |
+
"learning_rate": 5e-06,
|
| 408 |
+
"loss": 0.7222,
|
| 409 |
+
"step": 560
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0377787892580792,
|
| 413 |
+
"grad_norm": 0.6593013104808815,
|
| 414 |
+
"learning_rate": 5e-06,
|
| 415 |
+
"loss": 0.7205,
|
| 416 |
+
"step": 570
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.0559854346836595,
|
| 420 |
+
"grad_norm": 0.6107941470930411,
|
| 421 |
+
"learning_rate": 5e-06,
|
| 422 |
+
"loss": 0.7205,
|
| 423 |
+
"step": 580
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.07419208010924,
|
| 427 |
+
"grad_norm": 0.8276670919993867,
|
| 428 |
+
"learning_rate": 5e-06,
|
| 429 |
+
"loss": 0.7144,
|
| 430 |
+
"step": 590
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.0923987255348202,
|
| 434 |
+
"grad_norm": 0.6253185542660747,
|
| 435 |
+
"learning_rate": 5e-06,
|
| 436 |
+
"loss": 0.7136,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1106053709604005,
|
| 441 |
+
"grad_norm": 0.6108900775704884,
|
| 442 |
+
"learning_rate": 5e-06,
|
| 443 |
+
"loss": 0.7189,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.128812016385981,
|
| 448 |
+
"grad_norm": 0.8988785705551109,
|
| 449 |
+
"learning_rate": 5e-06,
|
| 450 |
+
"loss": 0.7122,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.1470186618115612,
|
| 455 |
+
"grad_norm": 0.616453203309773,
|
| 456 |
+
"learning_rate": 5e-06,
|
| 457 |
+
"loss": 0.7138,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.1652253072371415,
|
| 462 |
+
"grad_norm": 0.8816478964333492,
|
| 463 |
+
"learning_rate": 5e-06,
|
| 464 |
+
"loss": 0.7174,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.183431952662722,
|
| 469 |
+
"grad_norm": 0.6581015672880766,
|
| 470 |
+
"learning_rate": 5e-06,
|
| 471 |
+
"loss": 0.7183,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2016385980883022,
|
| 476 |
+
"grad_norm": 0.6253002121195789,
|
| 477 |
+
"learning_rate": 5e-06,
|
| 478 |
+
"loss": 0.7169,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2198452435138827,
|
| 483 |
+
"grad_norm": 0.7483871477416271,
|
| 484 |
+
"learning_rate": 5e-06,
|
| 485 |
+
"loss": 0.7136,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.238051888939463,
|
| 490 |
+
"grad_norm": 0.6941445361796307,
|
| 491 |
+
"learning_rate": 5e-06,
|
| 492 |
+
"loss": 0.7114,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.2562585343650432,
|
| 497 |
+
"grad_norm": 0.6143667364149711,
|
| 498 |
+
"learning_rate": 5e-06,
|
| 499 |
+
"loss": 0.7185,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.2744651797906235,
|
| 504 |
+
"grad_norm": 0.6592963882014911,
|
| 505 |
+
"learning_rate": 5e-06,
|
| 506 |
+
"loss": 0.7166,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.292671825216204,
|
| 511 |
+
"grad_norm": 0.6189597729803721,
|
| 512 |
+
"learning_rate": 5e-06,
|
| 513 |
+
"loss": 0.7158,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.3108784706417842,
|
| 518 |
+
"grad_norm": 0.6920980650430207,
|
| 519 |
+
"learning_rate": 5e-06,
|
| 520 |
+
"loss": 0.7144,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3290851160673647,
|
| 525 |
+
"grad_norm": 0.49530512173626284,
|
| 526 |
+
"learning_rate": 5e-06,
|
| 527 |
+
"loss": 0.718,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.347291761492945,
|
| 532 |
+
"grad_norm": 0.8064053486625922,
|
| 533 |
+
"learning_rate": 5e-06,
|
| 534 |
+
"loss": 0.7131,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.3654984069185252,
|
| 539 |
+
"grad_norm": 0.5821998941102782,
|
| 540 |
+
"learning_rate": 5e-06,
|
| 541 |
+
"loss": 0.7136,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.3837050523441057,
|
| 546 |
+
"grad_norm": 0.5735380427187717,
|
| 547 |
+
"learning_rate": 5e-06,
|
| 548 |
+
"loss": 0.7154,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.401911697769686,
|
| 553 |
+
"grad_norm": 0.5961452096313207,
|
| 554 |
+
"learning_rate": 5e-06,
|
| 555 |
+
"loss": 0.7148,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.4201183431952662,
|
| 560 |
+
"grad_norm": 0.6364200555389256,
|
| 561 |
+
"learning_rate": 5e-06,
|
| 562 |
+
"loss": 0.7152,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.4383249886208467,
|
| 567 |
+
"grad_norm": 0.618866609217087,
|
| 568 |
+
"learning_rate": 5e-06,
|
| 569 |
+
"loss": 0.7139,
|
| 570 |
+
"step": 790
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.456531634046427,
|
| 574 |
+
"grad_norm": 0.5153453443480923,
|
| 575 |
+
"learning_rate": 5e-06,
|
| 576 |
+
"loss": 0.7169,
|
| 577 |
+
"step": 800
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.4747382794720072,
|
| 581 |
+
"grad_norm": 0.5921562610346923,
|
| 582 |
+
"learning_rate": 5e-06,
|
| 583 |
+
"loss": 0.7157,
|
| 584 |
+
"step": 810
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.4929449248975877,
|
| 588 |
+
"grad_norm": 0.6864315768684789,
|
| 589 |
+
"learning_rate": 5e-06,
|
| 590 |
+
"loss": 0.7186,
|
| 591 |
+
"step": 820
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.511151570323168,
|
| 595 |
+
"grad_norm": 0.6718756873067311,
|
| 596 |
+
"learning_rate": 5e-06,
|
| 597 |
+
"loss": 0.713,
|
| 598 |
+
"step": 830
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.5293582157487484,
|
| 602 |
+
"grad_norm": 0.5292103286821733,
|
| 603 |
+
"learning_rate": 5e-06,
|
| 604 |
+
"loss": 0.7121,
|
| 605 |
+
"step": 840
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.5475648611743287,
|
| 609 |
+
"grad_norm": 0.5807337157950755,
|
| 610 |
+
"learning_rate": 5e-06,
|
| 611 |
+
"loss": 0.7161,
|
| 612 |
+
"step": 850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.565771506599909,
|
| 616 |
+
"grad_norm": 0.5725693174193143,
|
| 617 |
+
"learning_rate": 5e-06,
|
| 618 |
+
"loss": 0.7144,
|
| 619 |
+
"step": 860
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.5839781520254892,
|
| 623 |
+
"grad_norm": 0.6693290134206946,
|
| 624 |
+
"learning_rate": 5e-06,
|
| 625 |
+
"loss": 0.7137,
|
| 626 |
+
"step": 870
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.6021847974510697,
|
| 630 |
+
"grad_norm": 0.5960183936149324,
|
| 631 |
+
"learning_rate": 5e-06,
|
| 632 |
+
"loss": 0.718,
|
| 633 |
+
"step": 880
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.62039144287665,
|
| 637 |
+
"grad_norm": 0.6717348604009659,
|
| 638 |
+
"learning_rate": 5e-06,
|
| 639 |
+
"loss": 0.7107,
|
| 640 |
+
"step": 890
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.6385980883022304,
|
| 644 |
+
"grad_norm": 0.6147573815012755,
|
| 645 |
+
"learning_rate": 5e-06,
|
| 646 |
+
"loss": 0.7101,
|
| 647 |
+
"step": 900
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.6568047337278107,
|
| 651 |
+
"grad_norm": 0.5451936467703803,
|
| 652 |
+
"learning_rate": 5e-06,
|
| 653 |
+
"loss": 0.7128,
|
| 654 |
+
"step": 910
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.675011379153391,
|
| 658 |
+
"grad_norm": 0.7294513986677604,
|
| 659 |
+
"learning_rate": 5e-06,
|
| 660 |
+
"loss": 0.7104,
|
| 661 |
+
"step": 920
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.6932180245789712,
|
| 665 |
+
"grad_norm": 0.6756274493055726,
|
| 666 |
+
"learning_rate": 5e-06,
|
| 667 |
+
"loss": 0.7115,
|
| 668 |
+
"step": 930
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.7114246700045517,
|
| 672 |
+
"grad_norm": 0.6503407532587065,
|
| 673 |
+
"learning_rate": 5e-06,
|
| 674 |
+
"loss": 0.7105,
|
| 675 |
+
"step": 940
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.7296313154301322,
|
| 679 |
+
"grad_norm": 0.5369254794383076,
|
| 680 |
+
"learning_rate": 5e-06,
|
| 681 |
+
"loss": 0.7116,
|
| 682 |
+
"step": 950
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.7478379608557124,
|
| 686 |
+
"grad_norm": 0.6658976879506388,
|
| 687 |
+
"learning_rate": 5e-06,
|
| 688 |
+
"loss": 0.7108,
|
| 689 |
+
"step": 960
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.7660446062812927,
|
| 693 |
+
"grad_norm": 0.5443862825931207,
|
| 694 |
+
"learning_rate": 5e-06,
|
| 695 |
+
"loss": 0.7107,
|
| 696 |
+
"step": 970
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.784251251706873,
|
| 700 |
+
"grad_norm": 0.6571290419201206,
|
| 701 |
+
"learning_rate": 5e-06,
|
| 702 |
+
"loss": 0.7118,
|
| 703 |
+
"step": 980
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.8024578971324532,
|
| 707 |
+
"grad_norm": 0.543057907363159,
|
| 708 |
+
"learning_rate": 5e-06,
|
| 709 |
+
"loss": 0.7109,
|
| 710 |
+
"step": 990
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.8206645425580337,
|
| 714 |
+
"grad_norm": 0.5863888559339765,
|
| 715 |
+
"learning_rate": 5e-06,
|
| 716 |
+
"loss": 0.7128,
|
| 717 |
+
"step": 1000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.8388711879836142,
|
| 721 |
+
"grad_norm": 0.7512763814183313,
|
| 722 |
+
"learning_rate": 5e-06,
|
| 723 |
+
"loss": 0.7129,
|
| 724 |
+
"step": 1010
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.8570778334091944,
|
| 728 |
+
"grad_norm": 0.6875839517913319,
|
| 729 |
+
"learning_rate": 5e-06,
|
| 730 |
+
"loss": 0.7163,
|
| 731 |
+
"step": 1020
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.8752844788347747,
|
| 735 |
+
"grad_norm": 0.5392039305234562,
|
| 736 |
+
"learning_rate": 5e-06,
|
| 737 |
+
"loss": 0.712,
|
| 738 |
+
"step": 1030
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.893491124260355,
|
| 742 |
+
"grad_norm": 0.5715846033792595,
|
| 743 |
+
"learning_rate": 5e-06,
|
| 744 |
+
"loss": 0.7106,
|
| 745 |
+
"step": 1040
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.9116977696859354,
|
| 749 |
+
"grad_norm": 0.554967637717086,
|
| 750 |
+
"learning_rate": 5e-06,
|
| 751 |
+
"loss": 0.7135,
|
| 752 |
+
"step": 1050
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.9299044151115157,
|
| 756 |
+
"grad_norm": 0.6064174196129051,
|
| 757 |
+
"learning_rate": 5e-06,
|
| 758 |
+
"loss": 0.7113,
|
| 759 |
+
"step": 1060
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.9481110605370962,
|
| 763 |
+
"grad_norm": 0.57876695036443,
|
| 764 |
+
"learning_rate": 5e-06,
|
| 765 |
+
"loss": 0.7133,
|
| 766 |
+
"step": 1070
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.9663177059626764,
|
| 770 |
+
"grad_norm": 0.6168643227515411,
|
| 771 |
+
"learning_rate": 5e-06,
|
| 772 |
+
"loss": 0.7107,
|
| 773 |
+
"step": 1080
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.9845243513882567,
|
| 777 |
+
"grad_norm": 0.5918567965220687,
|
| 778 |
+
"learning_rate": 5e-06,
|
| 779 |
+
"loss": 0.7148,
|
| 780 |
+
"step": 1090
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.9990896677287209,
|
| 784 |
+
"eval_loss": 0.7494105696678162,
|
| 785 |
+
"eval_runtime": 371.303,
|
| 786 |
+
"eval_samples_per_second": 39.862,
|
| 787 |
+
"eval_steps_per_second": 0.625,
|
| 788 |
+
"step": 1098
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.002730996813837,
|
| 792 |
+
"grad_norm": 0.7112677344136977,
|
| 793 |
+
"learning_rate": 5e-06,
|
| 794 |
+
"loss": 0.7534,
|
| 795 |
+
"step": 1100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 2.020937642239417,
|
| 799 |
+
"grad_norm": 0.7657125714853397,
|
| 800 |
+
"learning_rate": 5e-06,
|
| 801 |
+
"loss": 0.6594,
|
| 802 |
+
"step": 1110
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.039144287664998,
|
| 806 |
+
"grad_norm": 0.7269679004826854,
|
| 807 |
+
"learning_rate": 5e-06,
|
| 808 |
+
"loss": 0.6665,
|
| 809 |
+
"step": 1120
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.057350933090578,
|
| 813 |
+
"grad_norm": 0.5908930298865153,
|
| 814 |
+
"learning_rate": 5e-06,
|
| 815 |
+
"loss": 0.6641,
|
| 816 |
+
"step": 1130
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 2.0755575785161584,
|
| 820 |
+
"grad_norm": 0.6685407405530368,
|
| 821 |
+
"learning_rate": 5e-06,
|
| 822 |
+
"loss": 0.6611,
|
| 823 |
+
"step": 1140
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 2.0937642239417387,
|
| 827 |
+
"grad_norm": 0.6835975141817704,
|
| 828 |
+
"learning_rate": 5e-06,
|
| 829 |
+
"loss": 0.6631,
|
| 830 |
+
"step": 1150
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.111970869367319,
|
| 834 |
+
"grad_norm": 0.6561709037357517,
|
| 835 |
+
"learning_rate": 5e-06,
|
| 836 |
+
"loss": 0.6631,
|
| 837 |
+
"step": 1160
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 2.1301775147928996,
|
| 841 |
+
"grad_norm": 0.9447854836936921,
|
| 842 |
+
"learning_rate": 5e-06,
|
| 843 |
+
"loss": 0.6641,
|
| 844 |
+
"step": 1170
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 2.14838416021848,
|
| 848 |
+
"grad_norm": 0.6554564933004847,
|
| 849 |
+
"learning_rate": 5e-06,
|
| 850 |
+
"loss": 0.6644,
|
| 851 |
+
"step": 1180
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 2.16659080564406,
|
| 855 |
+
"grad_norm": 1.0127997383891072,
|
| 856 |
+
"learning_rate": 5e-06,
|
| 857 |
+
"loss": 0.6711,
|
| 858 |
+
"step": 1190
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.1847974510696404,
|
| 862 |
+
"grad_norm": 0.7206567370516284,
|
| 863 |
+
"learning_rate": 5e-06,
|
| 864 |
+
"loss": 0.6668,
|
| 865 |
+
"step": 1200
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 2.2030040964952207,
|
| 869 |
+
"grad_norm": 0.655937928590562,
|
| 870 |
+
"learning_rate": 5e-06,
|
| 871 |
+
"loss": 0.6717,
|
| 872 |
+
"step": 1210
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.221210741920801,
|
| 876 |
+
"grad_norm": 0.8360483477674778,
|
| 877 |
+
"learning_rate": 5e-06,
|
| 878 |
+
"loss": 0.6689,
|
| 879 |
+
"step": 1220
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.2394173873463816,
|
| 883 |
+
"grad_norm": 0.6092424114015346,
|
| 884 |
+
"learning_rate": 5e-06,
|
| 885 |
+
"loss": 0.6628,
|
| 886 |
+
"step": 1230
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 2.257624032771962,
|
| 890 |
+
"grad_norm": 0.6406882124202334,
|
| 891 |
+
"learning_rate": 5e-06,
|
| 892 |
+
"loss": 0.6651,
|
| 893 |
+
"step": 1240
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 2.275830678197542,
|
| 897 |
+
"grad_norm": 0.7614678401383996,
|
| 898 |
+
"learning_rate": 5e-06,
|
| 899 |
+
"loss": 0.6677,
|
| 900 |
+
"step": 1250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 2.2940373236231224,
|
| 904 |
+
"grad_norm": 0.7284934278440872,
|
| 905 |
+
"learning_rate": 5e-06,
|
| 906 |
+
"loss": 0.6649,
|
| 907 |
+
"step": 1260
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 2.3122439690487027,
|
| 911 |
+
"grad_norm": 0.6736018159497253,
|
| 912 |
+
"learning_rate": 5e-06,
|
| 913 |
+
"loss": 0.6661,
|
| 914 |
+
"step": 1270
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.330450614474283,
|
| 918 |
+
"grad_norm": 0.682991318579073,
|
| 919 |
+
"learning_rate": 5e-06,
|
| 920 |
+
"loss": 0.6706,
|
| 921 |
+
"step": 1280
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 2.3486572598998636,
|
| 925 |
+
"grad_norm": 0.6067823760231827,
|
| 926 |
+
"learning_rate": 5e-06,
|
| 927 |
+
"loss": 0.6642,
|
| 928 |
+
"step": 1290
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 2.366863905325444,
|
| 932 |
+
"grad_norm": 0.6016632433423775,
|
| 933 |
+
"learning_rate": 5e-06,
|
| 934 |
+
"loss": 0.6675,
|
| 935 |
+
"step": 1300
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 2.385070550751024,
|
| 939 |
+
"grad_norm": 0.6784572170158043,
|
| 940 |
+
"learning_rate": 5e-06,
|
| 941 |
+
"loss": 0.6658,
|
| 942 |
+
"step": 1310
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 2.4032771961766044,
|
| 946 |
+
"grad_norm": 0.6804813510854688,
|
| 947 |
+
"learning_rate": 5e-06,
|
| 948 |
+
"loss": 0.6693,
|
| 949 |
+
"step": 1320
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 2.4214838416021847,
|
| 953 |
+
"grad_norm": 0.6667510974726526,
|
| 954 |
+
"learning_rate": 5e-06,
|
| 955 |
+
"loss": 0.6667,
|
| 956 |
+
"step": 1330
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.4396904870277654,
|
| 960 |
+
"grad_norm": 0.5712701093393895,
|
| 961 |
+
"learning_rate": 5e-06,
|
| 962 |
+
"loss": 0.6693,
|
| 963 |
+
"step": 1340
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 2.4578971324533456,
|
| 967 |
+
"grad_norm": 0.5495878729178928,
|
| 968 |
+
"learning_rate": 5e-06,
|
| 969 |
+
"loss": 0.6691,
|
| 970 |
+
"step": 1350
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 2.476103777878926,
|
| 974 |
+
"grad_norm": 0.6863248077276337,
|
| 975 |
+
"learning_rate": 5e-06,
|
| 976 |
+
"loss": 0.6697,
|
| 977 |
+
"step": 1360
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 2.494310423304506,
|
| 981 |
+
"grad_norm": 0.6281725692693549,
|
| 982 |
+
"learning_rate": 5e-06,
|
| 983 |
+
"loss": 0.6659,
|
| 984 |
+
"step": 1370
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 2.5125170687300864,
|
| 988 |
+
"grad_norm": 0.584114706811986,
|
| 989 |
+
"learning_rate": 5e-06,
|
| 990 |
+
"loss": 0.6689,
|
| 991 |
+
"step": 1380
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 2.5307237141556667,
|
| 995 |
+
"grad_norm": 0.8069687501187531,
|
| 996 |
+
"learning_rate": 5e-06,
|
| 997 |
+
"loss": 0.6713,
|
| 998 |
+
"step": 1390
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.548930359581247,
|
| 1002 |
+
"grad_norm": 0.5837408246005996,
|
| 1003 |
+
"learning_rate": 5e-06,
|
| 1004 |
+
"loss": 0.6702,
|
| 1005 |
+
"step": 1400
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 2.5671370050068276,
|
| 1009 |
+
"grad_norm": 0.6602631718487662,
|
| 1010 |
+
"learning_rate": 5e-06,
|
| 1011 |
+
"loss": 0.6729,
|
| 1012 |
+
"step": 1410
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 2.585343650432408,
|
| 1016 |
+
"grad_norm": 0.5751520967539265,
|
| 1017 |
+
"learning_rate": 5e-06,
|
| 1018 |
+
"loss": 0.6727,
|
| 1019 |
+
"step": 1420
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 2.603550295857988,
|
| 1023 |
+
"grad_norm": 0.5753704561331634,
|
| 1024 |
+
"learning_rate": 5e-06,
|
| 1025 |
+
"loss": 0.6638,
|
| 1026 |
+
"step": 1430
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 2.6217569412835684,
|
| 1030 |
+
"grad_norm": 0.5947582650600889,
|
| 1031 |
+
"learning_rate": 5e-06,
|
| 1032 |
+
"loss": 0.6687,
|
| 1033 |
+
"step": 1440
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.6399635867091487,
|
| 1037 |
+
"grad_norm": 0.6421455226025955,
|
| 1038 |
+
"learning_rate": 5e-06,
|
| 1039 |
+
"loss": 0.6673,
|
| 1040 |
+
"step": 1450
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.6581702321347294,
|
| 1044 |
+
"grad_norm": 0.625971391993565,
|
| 1045 |
+
"learning_rate": 5e-06,
|
| 1046 |
+
"loss": 0.6723,
|
| 1047 |
+
"step": 1460
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 2.6763768775603096,
|
| 1051 |
+
"grad_norm": 0.6230286833589651,
|
| 1052 |
+
"learning_rate": 5e-06,
|
| 1053 |
+
"loss": 0.6697,
|
| 1054 |
+
"step": 1470
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.69458352298589,
|
| 1058 |
+
"grad_norm": 0.6704340829719386,
|
| 1059 |
+
"learning_rate": 5e-06,
|
| 1060 |
+
"loss": 0.6726,
|
| 1061 |
+
"step": 1480
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 2.71279016841147,
|
| 1065 |
+
"grad_norm": 0.7115047592117129,
|
| 1066 |
+
"learning_rate": 5e-06,
|
| 1067 |
+
"loss": 0.6697,
|
| 1068 |
+
"step": 1490
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.7309968138370504,
|
| 1072 |
+
"grad_norm": 0.5348550201712026,
|
| 1073 |
+
"learning_rate": 5e-06,
|
| 1074 |
+
"loss": 0.6713,
|
| 1075 |
+
"step": 1500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 2.749203459262631,
|
| 1079 |
+
"grad_norm": 0.6018887550964896,
|
| 1080 |
+
"learning_rate": 5e-06,
|
| 1081 |
+
"loss": 0.6713,
|
| 1082 |
+
"step": 1510
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.7674101046882114,
|
| 1086 |
+
"grad_norm": 0.5829651029728877,
|
| 1087 |
+
"learning_rate": 5e-06,
|
| 1088 |
+
"loss": 0.6711,
|
| 1089 |
+
"step": 1520
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 2.7856167501137916,
|
| 1093 |
+
"grad_norm": 0.7257452058424095,
|
| 1094 |
+
"learning_rate": 5e-06,
|
| 1095 |
+
"loss": 0.6679,
|
| 1096 |
+
"step": 1530
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 2.803823395539372,
|
| 1100 |
+
"grad_norm": 0.7324462033376344,
|
| 1101 |
+
"learning_rate": 5e-06,
|
| 1102 |
+
"loss": 0.6705,
|
| 1103 |
+
"step": 1540
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 2.822030040964952,
|
| 1107 |
+
"grad_norm": 0.6518477868742856,
|
| 1108 |
+
"learning_rate": 5e-06,
|
| 1109 |
+
"loss": 0.6714,
|
| 1110 |
+
"step": 1550
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.8402366863905324,
|
| 1114 |
+
"grad_norm": 0.5915504419734868,
|
| 1115 |
+
"learning_rate": 5e-06,
|
| 1116 |
+
"loss": 0.6678,
|
| 1117 |
+
"step": 1560
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 2.8584433318161127,
|
| 1121 |
+
"grad_norm": 0.5596845389686367,
|
| 1122 |
+
"learning_rate": 5e-06,
|
| 1123 |
+
"loss": 0.6738,
|
| 1124 |
+
"step": 1570
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.8766499772416934,
|
| 1128 |
+
"grad_norm": 0.6211271364874157,
|
| 1129 |
+
"learning_rate": 5e-06,
|
| 1130 |
+
"loss": 0.6682,
|
| 1131 |
+
"step": 1580
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 2.8948566226672736,
|
| 1135 |
+
"grad_norm": 0.6029083750419233,
|
| 1136 |
+
"learning_rate": 5e-06,
|
| 1137 |
+
"loss": 0.6689,
|
| 1138 |
+
"step": 1590
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 2.913063268092854,
|
| 1142 |
+
"grad_norm": 0.6997154043188799,
|
| 1143 |
+
"learning_rate": 5e-06,
|
| 1144 |
+
"loss": 0.6654,
|
| 1145 |
+
"step": 1600
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 2.931269913518434,
|
| 1149 |
+
"grad_norm": 0.6561810265940334,
|
| 1150 |
+
"learning_rate": 5e-06,
|
| 1151 |
+
"loss": 0.6648,
|
| 1152 |
+
"step": 1610
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.9494765589440144,
|
| 1156 |
+
"grad_norm": 0.5949195869956836,
|
| 1157 |
+
"learning_rate": 5e-06,
|
| 1158 |
+
"loss": 0.6674,
|
| 1159 |
+
"step": 1620
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 2.967683204369595,
|
| 1163 |
+
"grad_norm": 0.6572460988263176,
|
| 1164 |
+
"learning_rate": 5e-06,
|
| 1165 |
+
"loss": 0.6748,
|
| 1166 |
+
"step": 1630
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.9858898497951754,
|
| 1170 |
+
"grad_norm": 0.6466132865038215,
|
| 1171 |
+
"learning_rate": 5e-06,
|
| 1172 |
+
"loss": 0.6707,
|
| 1173 |
+
"step": 1640
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 2.9986345015930813,
|
| 1177 |
+
"eval_loss": 0.7486923933029175,
|
| 1178 |
+
"eval_runtime": 371.6394,
|
| 1179 |
+
"eval_samples_per_second": 39.826,
|
| 1180 |
+
"eval_steps_per_second": 0.624,
|
| 1181 |
+
"step": 1647
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 3.0040964952207556,
|
| 1185 |
+
"grad_norm": 0.9066558799698319,
|
| 1186 |
+
"learning_rate": 5e-06,
|
| 1187 |
+
"loss": 0.7123,
|
| 1188 |
+
"step": 1650
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 3.022303140646336,
|
| 1192 |
+
"grad_norm": 0.6985831692717679,
|
| 1193 |
+
"learning_rate": 5e-06,
|
| 1194 |
+
"loss": 0.6192,
|
| 1195 |
+
"step": 1660
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 3.040509786071916,
|
| 1199 |
+
"grad_norm": 0.7297757031418619,
|
| 1200 |
+
"learning_rate": 5e-06,
|
| 1201 |
+
"loss": 0.6192,
|
| 1202 |
+
"step": 1670
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 3.0587164314974964,
|
| 1206 |
+
"grad_norm": 0.7384505423128972,
|
| 1207 |
+
"learning_rate": 5e-06,
|
| 1208 |
+
"loss": 0.6218,
|
| 1209 |
+
"step": 1680
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 3.076923076923077,
|
| 1213 |
+
"grad_norm": 0.7273583338941372,
|
| 1214 |
+
"learning_rate": 5e-06,
|
| 1215 |
+
"loss": 0.6211,
|
| 1216 |
+
"step": 1690
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 3.0951297223486574,
|
| 1220 |
+
"grad_norm": 0.7471474278740236,
|
| 1221 |
+
"learning_rate": 5e-06,
|
| 1222 |
+
"loss": 0.621,
|
| 1223 |
+
"step": 1700
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 3.1133363677742376,
|
| 1227 |
+
"grad_norm": 0.6193936409655777,
|
| 1228 |
+
"learning_rate": 5e-06,
|
| 1229 |
+
"loss": 0.6201,
|
| 1230 |
+
"step": 1710
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 3.131543013199818,
|
| 1234 |
+
"grad_norm": 0.6710258448973723,
|
| 1235 |
+
"learning_rate": 5e-06,
|
| 1236 |
+
"loss": 0.6224,
|
| 1237 |
+
"step": 1720
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 3.149749658625398,
|
| 1241 |
+
"grad_norm": 0.7409460472770332,
|
| 1242 |
+
"learning_rate": 5e-06,
|
| 1243 |
+
"loss": 0.625,
|
| 1244 |
+
"step": 1730
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 3.1679563040509784,
|
| 1248 |
+
"grad_norm": 0.709501536492518,
|
| 1249 |
+
"learning_rate": 5e-06,
|
| 1250 |
+
"loss": 0.6204,
|
| 1251 |
+
"step": 1740
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 3.186162949476559,
|
| 1255 |
+
"grad_norm": 0.7649819943694469,
|
| 1256 |
+
"learning_rate": 5e-06,
|
| 1257 |
+
"loss": 0.6163,
|
| 1258 |
+
"step": 1750
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 3.2043695949021394,
|
| 1262 |
+
"grad_norm": 0.7173580779194709,
|
| 1263 |
+
"learning_rate": 5e-06,
|
| 1264 |
+
"loss": 0.6234,
|
| 1265 |
+
"step": 1760
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 3.2225762403277196,
|
| 1269 |
+
"grad_norm": 0.6457493603144049,
|
| 1270 |
+
"learning_rate": 5e-06,
|
| 1271 |
+
"loss": 0.6213,
|
| 1272 |
+
"step": 1770
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 3.2407828857533,
|
| 1276 |
+
"grad_norm": 0.6074440281419403,
|
| 1277 |
+
"learning_rate": 5e-06,
|
| 1278 |
+
"loss": 0.6229,
|
| 1279 |
+
"step": 1780
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 3.25898953117888,
|
| 1283 |
+
"grad_norm": 0.7025871444184026,
|
| 1284 |
+
"learning_rate": 5e-06,
|
| 1285 |
+
"loss": 0.621,
|
| 1286 |
+
"step": 1790
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.277196176604461,
|
| 1290 |
+
"grad_norm": 0.7659855454008233,
|
| 1291 |
+
"learning_rate": 5e-06,
|
| 1292 |
+
"loss": 0.6244,
|
| 1293 |
+
"step": 1800
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 3.295402822030041,
|
| 1297 |
+
"grad_norm": 0.6532860177479579,
|
| 1298 |
+
"learning_rate": 5e-06,
|
| 1299 |
+
"loss": 0.6229,
|
| 1300 |
+
"step": 1810
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 3.3136094674556213,
|
| 1304 |
+
"grad_norm": 0.5432923070386573,
|
| 1305 |
+
"learning_rate": 5e-06,
|
| 1306 |
+
"loss": 0.6226,
|
| 1307 |
+
"step": 1820
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 3.3318161128812016,
|
| 1311 |
+
"grad_norm": 0.6927506071213173,
|
| 1312 |
+
"learning_rate": 5e-06,
|
| 1313 |
+
"loss": 0.6241,
|
| 1314 |
+
"step": 1830
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 3.350022758306782,
|
| 1318 |
+
"grad_norm": 0.6319912049544428,
|
| 1319 |
+
"learning_rate": 5e-06,
|
| 1320 |
+
"loss": 0.6234,
|
| 1321 |
+
"step": 1840
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 3.368229403732362,
|
| 1325 |
+
"grad_norm": 0.6730493262615627,
|
| 1326 |
+
"learning_rate": 5e-06,
|
| 1327 |
+
"loss": 0.6239,
|
| 1328 |
+
"step": 1850
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 3.386436049157943,
|
| 1332 |
+
"grad_norm": 0.624938299600225,
|
| 1333 |
+
"learning_rate": 5e-06,
|
| 1334 |
+
"loss": 0.6225,
|
| 1335 |
+
"step": 1860
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 3.404642694583523,
|
| 1339 |
+
"grad_norm": 0.623354310700811,
|
| 1340 |
+
"learning_rate": 5e-06,
|
| 1341 |
+
"loss": 0.6255,
|
| 1342 |
+
"step": 1870
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 3.4228493400091033,
|
| 1346 |
+
"grad_norm": 0.6329249624990303,
|
| 1347 |
+
"learning_rate": 5e-06,
|
| 1348 |
+
"loss": 0.6253,
|
| 1349 |
+
"step": 1880
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 3.4410559854346836,
|
| 1353 |
+
"grad_norm": 0.691830816858052,
|
| 1354 |
+
"learning_rate": 5e-06,
|
| 1355 |
+
"loss": 0.6313,
|
| 1356 |
+
"step": 1890
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 3.459262630860264,
|
| 1360 |
+
"grad_norm": 0.6032142068639882,
|
| 1361 |
+
"learning_rate": 5e-06,
|
| 1362 |
+
"loss": 0.6279,
|
| 1363 |
+
"step": 1900
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.477469276285844,
|
| 1367 |
+
"grad_norm": 0.6620277244742816,
|
| 1368 |
+
"learning_rate": 5e-06,
|
| 1369 |
+
"loss": 0.6279,
|
| 1370 |
+
"step": 1910
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 3.495675921711425,
|
| 1374 |
+
"grad_norm": 0.6334218656408422,
|
| 1375 |
+
"learning_rate": 5e-06,
|
| 1376 |
+
"loss": 0.6299,
|
| 1377 |
+
"step": 1920
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 3.513882567137005,
|
| 1381 |
+
"grad_norm": 0.7375169169002789,
|
| 1382 |
+
"learning_rate": 5e-06,
|
| 1383 |
+
"loss": 0.6281,
|
| 1384 |
+
"step": 1930
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 3.5320892125625853,
|
| 1388 |
+
"grad_norm": 0.6448512236031347,
|
| 1389 |
+
"learning_rate": 5e-06,
|
| 1390 |
+
"loss": 0.6275,
|
| 1391 |
+
"step": 1940
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 3.5502958579881656,
|
| 1395 |
+
"grad_norm": 0.5665149421787481,
|
| 1396 |
+
"learning_rate": 5e-06,
|
| 1397 |
+
"loss": 0.6299,
|
| 1398 |
+
"step": 1950
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 3.568502503413746,
|
| 1402 |
+
"grad_norm": 0.5999159680891181,
|
| 1403 |
+
"learning_rate": 5e-06,
|
| 1404 |
+
"loss": 0.6286,
|
| 1405 |
+
"step": 1960
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 3.5867091488393266,
|
| 1409 |
+
"grad_norm": 0.6281107227517486,
|
| 1410 |
+
"learning_rate": 5e-06,
|
| 1411 |
+
"loss": 0.6248,
|
| 1412 |
+
"step": 1970
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 3.604915794264907,
|
| 1416 |
+
"grad_norm": 0.6353135942879086,
|
| 1417 |
+
"learning_rate": 5e-06,
|
| 1418 |
+
"loss": 0.6281,
|
| 1419 |
+
"step": 1980
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 3.623122439690487,
|
| 1423 |
+
"grad_norm": 0.6193631021536377,
|
| 1424 |
+
"learning_rate": 5e-06,
|
| 1425 |
+
"loss": 0.6278,
|
| 1426 |
+
"step": 1990
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 3.6413290851160673,
|
| 1430 |
+
"grad_norm": 0.5918570154763652,
|
| 1431 |
+
"learning_rate": 5e-06,
|
| 1432 |
+
"loss": 0.6296,
|
| 1433 |
+
"step": 2000
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 3.6595357305416476,
|
| 1437 |
+
"grad_norm": 0.6010286514232586,
|
| 1438 |
+
"learning_rate": 5e-06,
|
| 1439 |
+
"loss": 0.6312,
|
| 1440 |
+
"step": 2010
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 3.6777423759672283,
|
| 1444 |
+
"grad_norm": 0.556597304136066,
|
| 1445 |
+
"learning_rate": 5e-06,
|
| 1446 |
+
"loss": 0.6282,
|
| 1447 |
+
"step": 2020
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 3.695949021392808,
|
| 1451 |
+
"grad_norm": 0.7271994139802581,
|
| 1452 |
+
"learning_rate": 5e-06,
|
| 1453 |
+
"loss": 0.6282,
|
| 1454 |
+
"step": 2030
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 3.714155666818389,
|
| 1458 |
+
"grad_norm": 0.6169419004473696,
|
| 1459 |
+
"learning_rate": 5e-06,
|
| 1460 |
+
"loss": 0.633,
|
| 1461 |
+
"step": 2040
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 3.732362312243969,
|
| 1465 |
+
"grad_norm": 0.6025566845682745,
|
| 1466 |
+
"learning_rate": 5e-06,
|
| 1467 |
+
"loss": 0.6299,
|
| 1468 |
+
"step": 2050
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 3.7505689576695493,
|
| 1472 |
+
"grad_norm": 0.5736503199358731,
|
| 1473 |
+
"learning_rate": 5e-06,
|
| 1474 |
+
"loss": 0.6307,
|
| 1475 |
+
"step": 2060
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 3.7687756030951296,
|
| 1479 |
+
"grad_norm": 0.6112656378611477,
|
| 1480 |
+
"learning_rate": 5e-06,
|
| 1481 |
+
"loss": 0.6314,
|
| 1482 |
+
"step": 2070
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 3.78698224852071,
|
| 1486 |
+
"grad_norm": 0.683407900900997,
|
| 1487 |
+
"learning_rate": 5e-06,
|
| 1488 |
+
"loss": 0.6292,
|
| 1489 |
+
"step": 2080
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 3.8051888939462906,
|
| 1493 |
+
"grad_norm": 0.5823312600994301,
|
| 1494 |
+
"learning_rate": 5e-06,
|
| 1495 |
+
"loss": 0.6302,
|
| 1496 |
+
"step": 2090
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 3.823395539371871,
|
| 1500 |
+
"grad_norm": 0.7722774266919119,
|
| 1501 |
+
"learning_rate": 5e-06,
|
| 1502 |
+
"loss": 0.6329,
|
| 1503 |
+
"step": 2100
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 3.841602184797451,
|
| 1507 |
+
"grad_norm": 0.5962312514475476,
|
| 1508 |
+
"learning_rate": 5e-06,
|
| 1509 |
+
"loss": 0.6336,
|
| 1510 |
+
"step": 2110
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 3.8598088302230313,
|
| 1514 |
+
"grad_norm": 0.5704378657470786,
|
| 1515 |
+
"learning_rate": 5e-06,
|
| 1516 |
+
"loss": 0.6343,
|
| 1517 |
+
"step": 2120
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 3.8780154756486116,
|
| 1521 |
+
"grad_norm": 0.6221717567681257,
|
| 1522 |
+
"learning_rate": 5e-06,
|
| 1523 |
+
"loss": 0.6323,
|
| 1524 |
+
"step": 2130
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 3.8962221210741923,
|
| 1528 |
+
"grad_norm": 0.8715901410739767,
|
| 1529 |
+
"learning_rate": 5e-06,
|
| 1530 |
+
"loss": 0.6289,
|
| 1531 |
+
"step": 2140
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 3.9144287664997726,
|
| 1535 |
+
"grad_norm": 0.7264743862880063,
|
| 1536 |
+
"learning_rate": 5e-06,
|
| 1537 |
+
"loss": 0.6319,
|
| 1538 |
+
"step": 2150
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 3.932635411925353,
|
| 1542 |
+
"grad_norm": 0.5646894741460851,
|
| 1543 |
+
"learning_rate": 5e-06,
|
| 1544 |
+
"loss": 0.6326,
|
| 1545 |
+
"step": 2160
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 3.950842057350933,
|
| 1549 |
+
"grad_norm": 0.6303177348076553,
|
| 1550 |
+
"learning_rate": 5e-06,
|
| 1551 |
+
"loss": 0.6347,
|
| 1552 |
+
"step": 2170
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 3.9690487027765133,
|
| 1556 |
+
"grad_norm": 0.5882842156900018,
|
| 1557 |
+
"learning_rate": 5e-06,
|
| 1558 |
+
"loss": 0.6255,
|
| 1559 |
+
"step": 2180
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 3.987255348202094,
|
| 1563 |
+
"grad_norm": 0.7637770706292029,
|
| 1564 |
+
"learning_rate": 5e-06,
|
| 1565 |
+
"loss": 0.6324,
|
| 1566 |
+
"step": 2190
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 4.0,
|
| 1570 |
+
"eval_loss": 0.7583181262016296,
|
| 1571 |
+
"eval_runtime": 379.7995,
|
| 1572 |
+
"eval_samples_per_second": 38.971,
|
| 1573 |
+
"eval_steps_per_second": 0.611,
|
| 1574 |
+
"step": 2197
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 4.005461993627674,
|
| 1578 |
+
"grad_norm": 1.3302589404721668,
|
| 1579 |
+
"learning_rate": 5e-06,
|
| 1580 |
+
"loss": 0.6642,
|
| 1581 |
+
"step": 2200
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 4.023668639053255,
|
| 1585 |
+
"grad_norm": 1.0479601737495694,
|
| 1586 |
+
"learning_rate": 5e-06,
|
| 1587 |
+
"loss": 0.5747,
|
| 1588 |
+
"step": 2210
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 4.041875284478834,
|
| 1592 |
+
"grad_norm": 0.8236224363789262,
|
| 1593 |
+
"learning_rate": 5e-06,
|
| 1594 |
+
"loss": 0.5739,
|
| 1595 |
+
"step": 2220
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 4.060081929904415,
|
| 1599 |
+
"grad_norm": 0.6514312376533014,
|
| 1600 |
+
"learning_rate": 5e-06,
|
| 1601 |
+
"loss": 0.5717,
|
| 1602 |
+
"step": 2230
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 4.078288575329996,
|
| 1606 |
+
"grad_norm": 0.6470033460056318,
|
| 1607 |
+
"learning_rate": 5e-06,
|
| 1608 |
+
"loss": 0.577,
|
| 1609 |
+
"step": 2240
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 4.096495220755576,
|
| 1613 |
+
"grad_norm": 0.6842183198102436,
|
| 1614 |
+
"learning_rate": 5e-06,
|
| 1615 |
+
"loss": 0.5772,
|
| 1616 |
+
"step": 2250
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.114701866181156,
|
| 1620 |
+
"grad_norm": 0.7773595484089195,
|
| 1621 |
+
"learning_rate": 5e-06,
|
| 1622 |
+
"loss": 0.5787,
|
| 1623 |
+
"step": 2260
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 4.132908511606736,
|
| 1627 |
+
"grad_norm": 0.8512328139098053,
|
| 1628 |
+
"learning_rate": 5e-06,
|
| 1629 |
+
"loss": 0.5782,
|
| 1630 |
+
"step": 2270
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 4.151115157032317,
|
| 1634 |
+
"grad_norm": 0.7556248928216761,
|
| 1635 |
+
"learning_rate": 5e-06,
|
| 1636 |
+
"loss": 0.5794,
|
| 1637 |
+
"step": 2280
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 4.1693218024578975,
|
| 1641 |
+
"grad_norm": 0.649345136898707,
|
| 1642 |
+
"learning_rate": 5e-06,
|
| 1643 |
+
"loss": 0.5806,
|
| 1644 |
+
"step": 2290
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 4.187528447883477,
|
| 1648 |
+
"grad_norm": 0.7218760078584716,
|
| 1649 |
+
"learning_rate": 5e-06,
|
| 1650 |
+
"loss": 0.5797,
|
| 1651 |
+
"step": 2300
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 4.205735093309058,
|
| 1655 |
+
"grad_norm": 0.6931763868077586,
|
| 1656 |
+
"learning_rate": 5e-06,
|
| 1657 |
+
"loss": 0.5767,
|
| 1658 |
+
"step": 2310
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 4.223941738734638,
|
| 1662 |
+
"grad_norm": 0.764188364334872,
|
| 1663 |
+
"learning_rate": 5e-06,
|
| 1664 |
+
"loss": 0.5778,
|
| 1665 |
+
"step": 2320
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 4.2421483841602186,
|
| 1669 |
+
"grad_norm": 0.7472509039059302,
|
| 1670 |
+
"learning_rate": 5e-06,
|
| 1671 |
+
"loss": 0.5786,
|
| 1672 |
+
"step": 2330
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 4.260355029585799,
|
| 1676 |
+
"grad_norm": 0.7681897599903893,
|
| 1677 |
+
"learning_rate": 5e-06,
|
| 1678 |
+
"loss": 0.5824,
|
| 1679 |
+
"step": 2340
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 4.278561675011379,
|
| 1683 |
+
"grad_norm": 0.6853235525321896,
|
| 1684 |
+
"learning_rate": 5e-06,
|
| 1685 |
+
"loss": 0.578,
|
| 1686 |
+
"step": 2350
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 4.29676832043696,
|
| 1690 |
+
"grad_norm": 0.6371627957254833,
|
| 1691 |
+
"learning_rate": 5e-06,
|
| 1692 |
+
"loss": 0.5767,
|
| 1693 |
+
"step": 2360
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 4.31497496586254,
|
| 1697 |
+
"grad_norm": 0.62518793171934,
|
| 1698 |
+
"learning_rate": 5e-06,
|
| 1699 |
+
"loss": 0.5828,
|
| 1700 |
+
"step": 2370
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 4.33318161128812,
|
| 1704 |
+
"grad_norm": 0.6835225300328861,
|
| 1705 |
+
"learning_rate": 5e-06,
|
| 1706 |
+
"loss": 0.5845,
|
| 1707 |
+
"step": 2380
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 4.3513882567137,
|
| 1711 |
+
"grad_norm": 0.6635739669645049,
|
| 1712 |
+
"learning_rate": 5e-06,
|
| 1713 |
+
"loss": 0.5831,
|
| 1714 |
+
"step": 2390
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 4.369594902139281,
|
| 1718 |
+
"grad_norm": 0.6905352305966937,
|
| 1719 |
+
"learning_rate": 5e-06,
|
| 1720 |
+
"loss": 0.5838,
|
| 1721 |
+
"step": 2400
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 4.3878015475648615,
|
| 1725 |
+
"grad_norm": 0.675578669245968,
|
| 1726 |
+
"learning_rate": 5e-06,
|
| 1727 |
+
"loss": 0.5838,
|
| 1728 |
+
"step": 2410
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 4.406008192990441,
|
| 1732 |
+
"grad_norm": 0.6472341173499976,
|
| 1733 |
+
"learning_rate": 5e-06,
|
| 1734 |
+
"loss": 0.5873,
|
| 1735 |
+
"step": 2420
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 4.424214838416022,
|
| 1739 |
+
"grad_norm": 0.6312793186052579,
|
| 1740 |
+
"learning_rate": 5e-06,
|
| 1741 |
+
"loss": 0.5862,
|
| 1742 |
+
"step": 2430
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 4.442421483841602,
|
| 1746 |
+
"grad_norm": 0.6451118193568525,
|
| 1747 |
+
"learning_rate": 5e-06,
|
| 1748 |
+
"loss": 0.5825,
|
| 1749 |
+
"step": 2440
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 4.4606281292671826,
|
| 1753 |
+
"grad_norm": 0.6570443368465694,
|
| 1754 |
+
"learning_rate": 5e-06,
|
| 1755 |
+
"loss": 0.5854,
|
| 1756 |
+
"step": 2450
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 4.478834774692763,
|
| 1760 |
+
"grad_norm": 0.658937290588797,
|
| 1761 |
+
"learning_rate": 5e-06,
|
| 1762 |
+
"loss": 0.5866,
|
| 1763 |
+
"step": 2460
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 4.497041420118343,
|
| 1767 |
+
"grad_norm": 0.7077581907222236,
|
| 1768 |
+
"learning_rate": 5e-06,
|
| 1769 |
+
"loss": 0.5869,
|
| 1770 |
+
"step": 2470
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 4.515248065543924,
|
| 1774 |
+
"grad_norm": 0.6792893518112693,
|
| 1775 |
+
"learning_rate": 5e-06,
|
| 1776 |
+
"loss": 0.5873,
|
| 1777 |
+
"step": 2480
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 4.533454710969504,
|
| 1781 |
+
"grad_norm": 0.7399002369000202,
|
| 1782 |
+
"learning_rate": 5e-06,
|
| 1783 |
+
"loss": 0.5832,
|
| 1784 |
+
"step": 2490
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 4.551661356395084,
|
| 1788 |
+
"grad_norm": 0.7316128957109573,
|
| 1789 |
+
"learning_rate": 5e-06,
|
| 1790 |
+
"loss": 0.5878,
|
| 1791 |
+
"step": 2500
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 4.569868001820664,
|
| 1795 |
+
"grad_norm": 0.709723130607844,
|
| 1796 |
+
"learning_rate": 5e-06,
|
| 1797 |
+
"loss": 0.5844,
|
| 1798 |
+
"step": 2510
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 4.588074647246245,
|
| 1802 |
+
"grad_norm": 0.6462822194584583,
|
| 1803 |
+
"learning_rate": 5e-06,
|
| 1804 |
+
"loss": 0.5883,
|
| 1805 |
+
"step": 2520
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 4.6062812926718255,
|
| 1809 |
+
"grad_norm": 0.7572060951703973,
|
| 1810 |
+
"learning_rate": 5e-06,
|
| 1811 |
+
"loss": 0.5914,
|
| 1812 |
+
"step": 2530
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 4.624487938097405,
|
| 1816 |
+
"grad_norm": 0.7698033683902142,
|
| 1817 |
+
"learning_rate": 5e-06,
|
| 1818 |
+
"loss": 0.5882,
|
| 1819 |
+
"step": 2540
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 4.642694583522986,
|
| 1823 |
+
"grad_norm": 0.7913214081907007,
|
| 1824 |
+
"learning_rate": 5e-06,
|
| 1825 |
+
"loss": 0.5865,
|
| 1826 |
+
"step": 2550
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 4.660901228948566,
|
| 1830 |
+
"grad_norm": 0.652908829617188,
|
| 1831 |
+
"learning_rate": 5e-06,
|
| 1832 |
+
"loss": 0.5885,
|
| 1833 |
+
"step": 2560
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 4.6791078743741465,
|
| 1837 |
+
"grad_norm": 0.7196290964249576,
|
| 1838 |
+
"learning_rate": 5e-06,
|
| 1839 |
+
"loss": 0.585,
|
| 1840 |
+
"step": 2570
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 4.697314519799727,
|
| 1844 |
+
"grad_norm": 0.6963685189277059,
|
| 1845 |
+
"learning_rate": 5e-06,
|
| 1846 |
+
"loss": 0.5874,
|
| 1847 |
+
"step": 2580
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 4.715521165225307,
|
| 1851 |
+
"grad_norm": 0.6213064117674314,
|
| 1852 |
+
"learning_rate": 5e-06,
|
| 1853 |
+
"loss": 0.5886,
|
| 1854 |
+
"step": 2590
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 4.733727810650888,
|
| 1858 |
+
"grad_norm": 0.6931094000594287,
|
| 1859 |
+
"learning_rate": 5e-06,
|
| 1860 |
+
"loss": 0.5931,
|
| 1861 |
+
"step": 2600
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 4.751934456076468,
|
| 1865 |
+
"grad_norm": 0.6421262946164188,
|
| 1866 |
+
"learning_rate": 5e-06,
|
| 1867 |
+
"loss": 0.5877,
|
| 1868 |
+
"step": 2610
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 4.770141101502048,
|
| 1872 |
+
"grad_norm": 0.7592504356723478,
|
| 1873 |
+
"learning_rate": 5e-06,
|
| 1874 |
+
"loss": 0.5901,
|
| 1875 |
+
"step": 2620
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 4.788347746927629,
|
| 1879 |
+
"grad_norm": 0.5755276719477025,
|
| 1880 |
+
"learning_rate": 5e-06,
|
| 1881 |
+
"loss": 0.5877,
|
| 1882 |
+
"step": 2630
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 4.806554392353209,
|
| 1886 |
+
"grad_norm": 0.7579206553363019,
|
| 1887 |
+
"learning_rate": 5e-06,
|
| 1888 |
+
"loss": 0.593,
|
| 1889 |
+
"step": 2640
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 4.8247610377787895,
|
| 1893 |
+
"grad_norm": 0.7227783750503406,
|
| 1894 |
+
"learning_rate": 5e-06,
|
| 1895 |
+
"loss": 0.592,
|
| 1896 |
+
"step": 2650
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 4.842967683204369,
|
| 1900 |
+
"grad_norm": 0.6143720314418375,
|
| 1901 |
+
"learning_rate": 5e-06,
|
| 1902 |
+
"loss": 0.5923,
|
| 1903 |
+
"step": 2660
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 4.86117432862995,
|
| 1907 |
+
"grad_norm": 0.684711113595212,
|
| 1908 |
+
"learning_rate": 5e-06,
|
| 1909 |
+
"loss": 0.5906,
|
| 1910 |
+
"step": 2670
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 4.879380974055531,
|
| 1914 |
+
"grad_norm": 0.6041365723044803,
|
| 1915 |
+
"learning_rate": 5e-06,
|
| 1916 |
+
"loss": 0.5943,
|
| 1917 |
+
"step": 2680
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 4.8975876194811105,
|
| 1921 |
+
"grad_norm": 0.6690754978083606,
|
| 1922 |
+
"learning_rate": 5e-06,
|
| 1923 |
+
"loss": 0.5909,
|
| 1924 |
+
"step": 2690
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 4.915794264906691,
|
| 1928 |
+
"grad_norm": 0.6408949623705275,
|
| 1929 |
+
"learning_rate": 5e-06,
|
| 1930 |
+
"loss": 0.5938,
|
| 1931 |
+
"step": 2700
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 4.934000910332271,
|
| 1935 |
+
"grad_norm": 0.6310380336524642,
|
| 1936 |
+
"learning_rate": 5e-06,
|
| 1937 |
+
"loss": 0.5896,
|
| 1938 |
+
"step": 2710
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 4.952207555757852,
|
| 1942 |
+
"grad_norm": 0.7658719466797836,
|
| 1943 |
+
"learning_rate": 5e-06,
|
| 1944 |
+
"loss": 0.5925,
|
| 1945 |
+
"step": 2720
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 4.970414201183432,
|
| 1949 |
+
"grad_norm": 0.6764346949045434,
|
| 1950 |
+
"learning_rate": 5e-06,
|
| 1951 |
+
"loss": 0.587,
|
| 1952 |
+
"step": 2730
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 4.988620846609012,
|
| 1956 |
+
"grad_norm": 0.6578929361436391,
|
| 1957 |
+
"learning_rate": 5e-06,
|
| 1958 |
+
"loss": 0.5947,
|
| 1959 |
+
"step": 2740
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 4.999544833864361,
|
| 1963 |
+
"eval_loss": 0.7756755352020264,
|
| 1964 |
+
"eval_runtime": 379.7908,
|
| 1965 |
+
"eval_samples_per_second": 38.971,
|
| 1966 |
+
"eval_steps_per_second": 0.611,
|
| 1967 |
+
"step": 2746
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 5.006827492034593,
|
| 1971 |
+
"grad_norm": 1.6971405469705954,
|
| 1972 |
+
"learning_rate": 5e-06,
|
| 1973 |
+
"loss": 0.6175,
|
| 1974 |
+
"step": 2750
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 5.025034137460173,
|
| 1978 |
+
"grad_norm": 1.1249548156787617,
|
| 1979 |
+
"learning_rate": 5e-06,
|
| 1980 |
+
"loss": 0.5284,
|
| 1981 |
+
"step": 2760
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 5.0432407828857535,
|
| 1985 |
+
"grad_norm": 0.8687787258109687,
|
| 1986 |
+
"learning_rate": 5e-06,
|
| 1987 |
+
"loss": 0.5308,
|
| 1988 |
+
"step": 2770
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 5.061447428311333,
|
| 1992 |
+
"grad_norm": 0.8717275105088685,
|
| 1993 |
+
"learning_rate": 5e-06,
|
| 1994 |
+
"loss": 0.5312,
|
| 1995 |
+
"step": 2780
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 5.079654073736914,
|
| 1999 |
+
"grad_norm": 0.8465418731492865,
|
| 2000 |
+
"learning_rate": 5e-06,
|
| 2001 |
+
"loss": 0.5304,
|
| 2002 |
+
"step": 2790
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 5.097860719162495,
|
| 2006 |
+
"grad_norm": 0.7886283884740087,
|
| 2007 |
+
"learning_rate": 5e-06,
|
| 2008 |
+
"loss": 0.5279,
|
| 2009 |
+
"step": 2800
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 5.1160673645880745,
|
| 2013 |
+
"grad_norm": 0.8398942783728177,
|
| 2014 |
+
"learning_rate": 5e-06,
|
| 2015 |
+
"loss": 0.5363,
|
| 2016 |
+
"step": 2810
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 5.134274010013655,
|
| 2020 |
+
"grad_norm": 0.769923514886696,
|
| 2021 |
+
"learning_rate": 5e-06,
|
| 2022 |
+
"loss": 0.5338,
|
| 2023 |
+
"step": 2820
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 5.152480655439235,
|
| 2027 |
+
"grad_norm": 0.7868390424124061,
|
| 2028 |
+
"learning_rate": 5e-06,
|
| 2029 |
+
"loss": 0.5321,
|
| 2030 |
+
"step": 2830
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 5.170687300864816,
|
| 2034 |
+
"grad_norm": 0.7950518134725912,
|
| 2035 |
+
"learning_rate": 5e-06,
|
| 2036 |
+
"loss": 0.534,
|
| 2037 |
+
"step": 2840
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 5.188893946290396,
|
| 2041 |
+
"grad_norm": 0.9279901061572416,
|
| 2042 |
+
"learning_rate": 5e-06,
|
| 2043 |
+
"loss": 0.5363,
|
| 2044 |
+
"step": 2850
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 5.207100591715976,
|
| 2048 |
+
"grad_norm": 0.7594304483347795,
|
| 2049 |
+
"learning_rate": 5e-06,
|
| 2050 |
+
"loss": 0.534,
|
| 2051 |
+
"step": 2860
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 5.225307237141557,
|
| 2055 |
+
"grad_norm": 0.8252531851052891,
|
| 2056 |
+
"learning_rate": 5e-06,
|
| 2057 |
+
"loss": 0.5349,
|
| 2058 |
+
"step": 2870
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 5.243513882567137,
|
| 2062 |
+
"grad_norm": 0.7626524916862092,
|
| 2063 |
+
"learning_rate": 5e-06,
|
| 2064 |
+
"loss": 0.534,
|
| 2065 |
+
"step": 2880
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 5.2617205279927175,
|
| 2069 |
+
"grad_norm": 0.78634458156834,
|
| 2070 |
+
"learning_rate": 5e-06,
|
| 2071 |
+
"loss": 0.5368,
|
| 2072 |
+
"step": 2890
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 5.279927173418297,
|
| 2076 |
+
"grad_norm": 0.7800860821607554,
|
| 2077 |
+
"learning_rate": 5e-06,
|
| 2078 |
+
"loss": 0.5348,
|
| 2079 |
+
"step": 2900
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 5.298133818843878,
|
| 2083 |
+
"grad_norm": 0.7198214581579875,
|
| 2084 |
+
"learning_rate": 5e-06,
|
| 2085 |
+
"loss": 0.539,
|
| 2086 |
+
"step": 2910
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 5.316340464269459,
|
| 2090 |
+
"grad_norm": 0.8018455337446776,
|
| 2091 |
+
"learning_rate": 5e-06,
|
| 2092 |
+
"loss": 0.5354,
|
| 2093 |
+
"step": 2920
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 5.3345471096950385,
|
| 2097 |
+
"grad_norm": 0.6375711217756587,
|
| 2098 |
+
"learning_rate": 5e-06,
|
| 2099 |
+
"loss": 0.538,
|
| 2100 |
+
"step": 2930
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 5.352753755120619,
|
| 2104 |
+
"grad_norm": 0.7615279567325839,
|
| 2105 |
+
"learning_rate": 5e-06,
|
| 2106 |
+
"loss": 0.5388,
|
| 2107 |
+
"step": 2940
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 5.370960400546199,
|
| 2111 |
+
"grad_norm": 0.8634503306148038,
|
| 2112 |
+
"learning_rate": 5e-06,
|
| 2113 |
+
"loss": 0.5387,
|
| 2114 |
+
"step": 2950
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 5.38916704597178,
|
| 2118 |
+
"grad_norm": 0.728328161040519,
|
| 2119 |
+
"learning_rate": 5e-06,
|
| 2120 |
+
"loss": 0.5392,
|
| 2121 |
+
"step": 2960
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 5.4073736913973605,
|
| 2125 |
+
"grad_norm": 0.6714611489114068,
|
| 2126 |
+
"learning_rate": 5e-06,
|
| 2127 |
+
"loss": 0.5393,
|
| 2128 |
+
"step": 2970
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 5.42558033682294,
|
| 2132 |
+
"grad_norm": 0.864160433406934,
|
| 2133 |
+
"learning_rate": 5e-06,
|
| 2134 |
+
"loss": 0.5406,
|
| 2135 |
+
"step": 2980
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 5.443786982248521,
|
| 2139 |
+
"grad_norm": 0.7597928888037461,
|
| 2140 |
+
"learning_rate": 5e-06,
|
| 2141 |
+
"loss": 0.5368,
|
| 2142 |
+
"step": 2990
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 5.461993627674101,
|
| 2146 |
+
"grad_norm": 0.7636646866418205,
|
| 2147 |
+
"learning_rate": 5e-06,
|
| 2148 |
+
"loss": 0.5403,
|
| 2149 |
+
"step": 3000
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 5.4802002730996815,
|
| 2153 |
+
"grad_norm": 0.7478936763176718,
|
| 2154 |
+
"learning_rate": 5e-06,
|
| 2155 |
+
"loss": 0.5411,
|
| 2156 |
+
"step": 3010
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 5.498406918525261,
|
| 2160 |
+
"grad_norm": 0.7764354445181351,
|
| 2161 |
+
"learning_rate": 5e-06,
|
| 2162 |
+
"loss": 0.5411,
|
| 2163 |
+
"step": 3020
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 5.516613563950842,
|
| 2167 |
+
"grad_norm": 0.8232735346509414,
|
| 2168 |
+
"learning_rate": 5e-06,
|
| 2169 |
+
"loss": 0.5433,
|
| 2170 |
+
"step": 3030
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 5.534820209376423,
|
| 2174 |
+
"grad_norm": 0.7742332026679888,
|
| 2175 |
+
"learning_rate": 5e-06,
|
| 2176 |
+
"loss": 0.5459,
|
| 2177 |
+
"step": 3040
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 5.5530268548020025,
|
| 2181 |
+
"grad_norm": 0.8806327502433587,
|
| 2182 |
+
"learning_rate": 5e-06,
|
| 2183 |
+
"loss": 0.5456,
|
| 2184 |
+
"step": 3050
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 5.571233500227583,
|
| 2188 |
+
"grad_norm": 0.8857449291804894,
|
| 2189 |
+
"learning_rate": 5e-06,
|
| 2190 |
+
"loss": 0.5434,
|
| 2191 |
+
"step": 3060
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 5.589440145653163,
|
| 2195 |
+
"grad_norm": 0.9161045429573774,
|
| 2196 |
+
"learning_rate": 5e-06,
|
| 2197 |
+
"loss": 0.5381,
|
| 2198 |
+
"step": 3070
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 5.607646791078744,
|
| 2202 |
+
"grad_norm": 0.7948317610189037,
|
| 2203 |
+
"learning_rate": 5e-06,
|
| 2204 |
+
"loss": 0.5424,
|
| 2205 |
+
"step": 3080
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 5.6258534365043245,
|
| 2209 |
+
"grad_norm": 0.7867485103330927,
|
| 2210 |
+
"learning_rate": 5e-06,
|
| 2211 |
+
"loss": 0.5434,
|
| 2212 |
+
"step": 3090
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 5.644060081929904,
|
| 2216 |
+
"grad_norm": 0.7569175444017539,
|
| 2217 |
+
"learning_rate": 5e-06,
|
| 2218 |
+
"loss": 0.5456,
|
| 2219 |
+
"step": 3100
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 5.662266727355485,
|
| 2223 |
+
"grad_norm": 0.7425402095430522,
|
| 2224 |
+
"learning_rate": 5e-06,
|
| 2225 |
+
"loss": 0.5437,
|
| 2226 |
+
"step": 3110
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 5.680473372781065,
|
| 2230 |
+
"grad_norm": 0.8566468508292944,
|
| 2231 |
+
"learning_rate": 5e-06,
|
| 2232 |
+
"loss": 0.543,
|
| 2233 |
+
"step": 3120
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 5.6986800182066455,
|
| 2237 |
+
"grad_norm": 0.7291267040452126,
|
| 2238 |
+
"learning_rate": 5e-06,
|
| 2239 |
+
"loss": 0.5456,
|
| 2240 |
+
"step": 3130
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 5.716886663632225,
|
| 2244 |
+
"grad_norm": 0.9339173337499906,
|
| 2245 |
+
"learning_rate": 5e-06,
|
| 2246 |
+
"loss": 0.5488,
|
| 2247 |
+
"step": 3140
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 5.735093309057806,
|
| 2251 |
+
"grad_norm": 0.668254133197116,
|
| 2252 |
+
"learning_rate": 5e-06,
|
| 2253 |
+
"loss": 0.5461,
|
| 2254 |
+
"step": 3150
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 5.753299954483387,
|
| 2258 |
+
"grad_norm": 0.733779005468029,
|
| 2259 |
+
"learning_rate": 5e-06,
|
| 2260 |
+
"loss": 0.548,
|
| 2261 |
+
"step": 3160
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 5.7715065999089665,
|
| 2265 |
+
"grad_norm": 0.6817898908805843,
|
| 2266 |
+
"learning_rate": 5e-06,
|
| 2267 |
+
"loss": 0.5471,
|
| 2268 |
+
"step": 3170
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 5.789713245334547,
|
| 2272 |
+
"grad_norm": 0.7061838303703719,
|
| 2273 |
+
"learning_rate": 5e-06,
|
| 2274 |
+
"loss": 0.5435,
|
| 2275 |
+
"step": 3180
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 5.807919890760127,
|
| 2279 |
+
"grad_norm": 0.6788561279223211,
|
| 2280 |
+
"learning_rate": 5e-06,
|
| 2281 |
+
"loss": 0.5485,
|
| 2282 |
+
"step": 3190
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 5.826126536185708,
|
| 2286 |
+
"grad_norm": 0.743542468839561,
|
| 2287 |
+
"learning_rate": 5e-06,
|
| 2288 |
+
"loss": 0.5442,
|
| 2289 |
+
"step": 3200
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 5.8443331816112885,
|
| 2293 |
+
"grad_norm": 0.7126719114866883,
|
| 2294 |
+
"learning_rate": 5e-06,
|
| 2295 |
+
"loss": 0.545,
|
| 2296 |
+
"step": 3210
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 5.862539827036868,
|
| 2300 |
+
"grad_norm": 0.7700757305885421,
|
| 2301 |
+
"learning_rate": 5e-06,
|
| 2302 |
+
"loss": 0.5476,
|
| 2303 |
+
"step": 3220
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 5.880746472462449,
|
| 2307 |
+
"grad_norm": 0.8580615971421329,
|
| 2308 |
+
"learning_rate": 5e-06,
|
| 2309 |
+
"loss": 0.5496,
|
| 2310 |
+
"step": 3230
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 5.898953117888029,
|
| 2314 |
+
"grad_norm": 0.6651944277970823,
|
| 2315 |
+
"learning_rate": 5e-06,
|
| 2316 |
+
"loss": 0.5467,
|
| 2317 |
+
"step": 3240
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 5.9171597633136095,
|
| 2321 |
+
"grad_norm": 0.7351660383841082,
|
| 2322 |
+
"learning_rate": 5e-06,
|
| 2323 |
+
"loss": 0.5498,
|
| 2324 |
+
"step": 3250
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 5.93536640873919,
|
| 2328 |
+
"grad_norm": 0.7432057163838226,
|
| 2329 |
+
"learning_rate": 5e-06,
|
| 2330 |
+
"loss": 0.5457,
|
| 2331 |
+
"step": 3260
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 5.95357305416477,
|
| 2335 |
+
"grad_norm": 0.7880258422896124,
|
| 2336 |
+
"learning_rate": 5e-06,
|
| 2337 |
+
"loss": 0.5484,
|
| 2338 |
+
"step": 3270
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 5.971779699590351,
|
| 2342 |
+
"grad_norm": 0.7829267364761494,
|
| 2343 |
+
"learning_rate": 5e-06,
|
| 2344 |
+
"loss": 0.5522,
|
| 2345 |
+
"step": 3280
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 5.9899863450159305,
|
| 2349 |
+
"grad_norm": 0.755975814943294,
|
| 2350 |
+
"learning_rate": 5e-06,
|
| 2351 |
+
"loss": 0.5504,
|
| 2352 |
+
"step": 3290
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 5.999089667728721,
|
| 2356 |
+
"eval_loss": 0.8121561408042908,
|
| 2357 |
+
"eval_runtime": 385.6914,
|
| 2358 |
+
"eval_samples_per_second": 38.375,
|
| 2359 |
+
"eval_steps_per_second": 0.602,
|
| 2360 |
+
"step": 3295
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 6.008192990441511,
|
| 2364 |
+
"grad_norm": 1.229518253610835,
|
| 2365 |
+
"learning_rate": 5e-06,
|
| 2366 |
+
"loss": 0.5616,
|
| 2367 |
+
"step": 3300
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 6.026399635867092,
|
| 2371 |
+
"grad_norm": 0.9690039493999913,
|
| 2372 |
+
"learning_rate": 5e-06,
|
| 2373 |
+
"loss": 0.4806,
|
| 2374 |
+
"step": 3310
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 6.044606281292672,
|
| 2378 |
+
"grad_norm": 0.975061524739041,
|
| 2379 |
+
"learning_rate": 5e-06,
|
| 2380 |
+
"loss": 0.4807,
|
| 2381 |
+
"step": 3320
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 6.0628129267182524,
|
| 2385 |
+
"grad_norm": 0.8693905009394387,
|
| 2386 |
+
"learning_rate": 5e-06,
|
| 2387 |
+
"loss": 0.48,
|
| 2388 |
+
"step": 3330
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 6.081019572143832,
|
| 2392 |
+
"grad_norm": 0.8533904059239436,
|
| 2393 |
+
"learning_rate": 5e-06,
|
| 2394 |
+
"loss": 0.4814,
|
| 2395 |
+
"step": 3340
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 6.099226217569413,
|
| 2399 |
+
"grad_norm": 0.9743084553071109,
|
| 2400 |
+
"learning_rate": 5e-06,
|
| 2401 |
+
"loss": 0.4832,
|
| 2402 |
+
"step": 3350
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 6.117432862994993,
|
| 2406 |
+
"grad_norm": 0.8428587686314428,
|
| 2407 |
+
"learning_rate": 5e-06,
|
| 2408 |
+
"loss": 0.4836,
|
| 2409 |
+
"step": 3360
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 6.1356395084205735,
|
| 2413 |
+
"grad_norm": 0.8485721574906323,
|
| 2414 |
+
"learning_rate": 5e-06,
|
| 2415 |
+
"loss": 0.4842,
|
| 2416 |
+
"step": 3370
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 6.153846153846154,
|
| 2420 |
+
"grad_norm": 0.9049634307737825,
|
| 2421 |
+
"learning_rate": 5e-06,
|
| 2422 |
+
"loss": 0.4871,
|
| 2423 |
+
"step": 3380
|
| 2424 |
+
},
|
| 2425 |
+
{
|
| 2426 |
+
"epoch": 6.172052799271734,
|
| 2427 |
+
"grad_norm": 0.8485250729351949,
|
| 2428 |
+
"learning_rate": 5e-06,
|
| 2429 |
+
"loss": 0.4891,
|
| 2430 |
+
"step": 3390
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 6.190259444697315,
|
| 2434 |
+
"grad_norm": 0.8677063446883995,
|
| 2435 |
+
"learning_rate": 5e-06,
|
| 2436 |
+
"loss": 0.4863,
|
| 2437 |
+
"step": 3400
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 6.2084660901228945,
|
| 2441 |
+
"grad_norm": 0.8067384460711394,
|
| 2442 |
+
"learning_rate": 5e-06,
|
| 2443 |
+
"loss": 0.4847,
|
| 2444 |
+
"step": 3410
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 6.226672735548475,
|
| 2448 |
+
"grad_norm": 0.8491856381367863,
|
| 2449 |
+
"learning_rate": 5e-06,
|
| 2450 |
+
"loss": 0.4872,
|
| 2451 |
+
"step": 3420
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 6.244879380974056,
|
| 2455 |
+
"grad_norm": 0.8434795569738153,
|
| 2456 |
+
"learning_rate": 5e-06,
|
| 2457 |
+
"loss": 0.4893,
|
| 2458 |
+
"step": 3430
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 6.263086026399636,
|
| 2462 |
+
"grad_norm": 0.877180400082389,
|
| 2463 |
+
"learning_rate": 5e-06,
|
| 2464 |
+
"loss": 0.488,
|
| 2465 |
+
"step": 3440
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 6.2812926718252164,
|
| 2469 |
+
"grad_norm": 0.7544909031927692,
|
| 2470 |
+
"learning_rate": 5e-06,
|
| 2471 |
+
"loss": 0.4926,
|
| 2472 |
+
"step": 3450
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 6.299499317250796,
|
| 2476 |
+
"grad_norm": 0.7786685673980743,
|
| 2477 |
+
"learning_rate": 5e-06,
|
| 2478 |
+
"loss": 0.4903,
|
| 2479 |
+
"step": 3460
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"epoch": 6.317705962676377,
|
| 2483 |
+
"grad_norm": 0.8813259287698524,
|
| 2484 |
+
"learning_rate": 5e-06,
|
| 2485 |
+
"loss": 0.4928,
|
| 2486 |
+
"step": 3470
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 6.335912608101957,
|
| 2490 |
+
"grad_norm": 0.8143534255862093,
|
| 2491 |
+
"learning_rate": 5e-06,
|
| 2492 |
+
"loss": 0.489,
|
| 2493 |
+
"step": 3480
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 6.3541192535275375,
|
| 2497 |
+
"grad_norm": 0.813093240140628,
|
| 2498 |
+
"learning_rate": 5e-06,
|
| 2499 |
+
"loss": 0.4895,
|
| 2500 |
+
"step": 3490
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 6.372325898953118,
|
| 2504 |
+
"grad_norm": 0.8234651442993571,
|
| 2505 |
+
"learning_rate": 5e-06,
|
| 2506 |
+
"loss": 0.4874,
|
| 2507 |
+
"step": 3500
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 6.390532544378698,
|
| 2511 |
+
"grad_norm": 0.7583974230123129,
|
| 2512 |
+
"learning_rate": 5e-06,
|
| 2513 |
+
"loss": 0.4894,
|
| 2514 |
+
"step": 3510
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 6.408739189804279,
|
| 2518 |
+
"grad_norm": 0.8045709791368885,
|
| 2519 |
+
"learning_rate": 5e-06,
|
| 2520 |
+
"loss": 0.4924,
|
| 2521 |
+
"step": 3520
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 6.4269458352298585,
|
| 2525 |
+
"grad_norm": 0.8021037971072071,
|
| 2526 |
+
"learning_rate": 5e-06,
|
| 2527 |
+
"loss": 0.4958,
|
| 2528 |
+
"step": 3530
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 6.445152480655439,
|
| 2532 |
+
"grad_norm": 0.8860947638978545,
|
| 2533 |
+
"learning_rate": 5e-06,
|
| 2534 |
+
"loss": 0.4936,
|
| 2535 |
+
"step": 3540
|
| 2536 |
+
},
|
| 2537 |
+
{
|
| 2538 |
+
"epoch": 6.46335912608102,
|
| 2539 |
+
"grad_norm": 0.778005059457562,
|
| 2540 |
+
"learning_rate": 5e-06,
|
| 2541 |
+
"loss": 0.4904,
|
| 2542 |
+
"step": 3550
|
| 2543 |
+
},
|
| 2544 |
+
{
|
| 2545 |
+
"epoch": 6.4815657715066,
|
| 2546 |
+
"grad_norm": 0.8879225211432363,
|
| 2547 |
+
"learning_rate": 5e-06,
|
| 2548 |
+
"loss": 0.4996,
|
| 2549 |
+
"step": 3560
|
| 2550 |
+
},
|
| 2551 |
+
{
|
| 2552 |
+
"epoch": 6.49977241693218,
|
| 2553 |
+
"grad_norm": 0.86115853799121,
|
| 2554 |
+
"learning_rate": 5e-06,
|
| 2555 |
+
"loss": 0.4958,
|
| 2556 |
+
"step": 3570
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 6.51797906235776,
|
| 2560 |
+
"grad_norm": 0.821923783763917,
|
| 2561 |
+
"learning_rate": 5e-06,
|
| 2562 |
+
"loss": 0.4998,
|
| 2563 |
+
"step": 3580
|
| 2564 |
+
},
|
| 2565 |
+
{
|
| 2566 |
+
"epoch": 6.536185707783341,
|
| 2567 |
+
"grad_norm": 0.7488573715925901,
|
| 2568 |
+
"learning_rate": 5e-06,
|
| 2569 |
+
"loss": 0.4942,
|
| 2570 |
+
"step": 3590
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"epoch": 6.554392353208922,
|
| 2574 |
+
"grad_norm": 0.766478109512254,
|
| 2575 |
+
"learning_rate": 5e-06,
|
| 2576 |
+
"loss": 0.4927,
|
| 2577 |
+
"step": 3600
|
| 2578 |
+
},
|
| 2579 |
+
{
|
| 2580 |
+
"epoch": 6.5725989986345015,
|
| 2581 |
+
"grad_norm": 0.9993393195281365,
|
| 2582 |
+
"learning_rate": 5e-06,
|
| 2583 |
+
"loss": 0.5009,
|
| 2584 |
+
"step": 3610
|
| 2585 |
+
},
|
| 2586 |
+
{
|
| 2587 |
+
"epoch": 6.590805644060082,
|
| 2588 |
+
"grad_norm": 1.0105814992493984,
|
| 2589 |
+
"learning_rate": 5e-06,
|
| 2590 |
+
"loss": 0.4902,
|
| 2591 |
+
"step": 3620
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 6.609012289485662,
|
| 2595 |
+
"grad_norm": 0.918270960348232,
|
| 2596 |
+
"learning_rate": 5e-06,
|
| 2597 |
+
"loss": 0.5004,
|
| 2598 |
+
"step": 3630
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 6.627218934911243,
|
| 2602 |
+
"grad_norm": 0.7543985269097522,
|
| 2603 |
+
"learning_rate": 5e-06,
|
| 2604 |
+
"loss": 0.4972,
|
| 2605 |
+
"step": 3640
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 6.645425580336823,
|
| 2609 |
+
"grad_norm": 0.9591362111868698,
|
| 2610 |
+
"learning_rate": 5e-06,
|
| 2611 |
+
"loss": 0.4978,
|
| 2612 |
+
"step": 3650
|
| 2613 |
+
},
|
| 2614 |
+
{
|
| 2615 |
+
"epoch": 6.663632225762403,
|
| 2616 |
+
"grad_norm": 0.8732535087129343,
|
| 2617 |
+
"learning_rate": 5e-06,
|
| 2618 |
+
"loss": 0.4962,
|
| 2619 |
+
"step": 3660
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 6.681838871187984,
|
| 2623 |
+
"grad_norm": 0.8907840024813226,
|
| 2624 |
+
"learning_rate": 5e-06,
|
| 2625 |
+
"loss": 0.4984,
|
| 2626 |
+
"step": 3670
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 6.700045516613564,
|
| 2630 |
+
"grad_norm": 0.8081950411988741,
|
| 2631 |
+
"learning_rate": 5e-06,
|
| 2632 |
+
"loss": 0.4995,
|
| 2633 |
+
"step": 3680
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 6.718252162039144,
|
| 2637 |
+
"grad_norm": 0.7957053619070132,
|
| 2638 |
+
"learning_rate": 5e-06,
|
| 2639 |
+
"loss": 0.5007,
|
| 2640 |
+
"step": 3690
|
| 2641 |
+
},
|
| 2642 |
+
{
|
| 2643 |
+
"epoch": 6.736458807464724,
|
| 2644 |
+
"grad_norm": 1.0396693696127743,
|
| 2645 |
+
"learning_rate": 5e-06,
|
| 2646 |
+
"loss": 0.5,
|
| 2647 |
+
"step": 3700
|
| 2648 |
+
},
|
| 2649 |
+
{
|
| 2650 |
+
"epoch": 6.754665452890305,
|
| 2651 |
+
"grad_norm": 0.8276731947592968,
|
| 2652 |
+
"learning_rate": 5e-06,
|
| 2653 |
+
"loss": 0.5003,
|
| 2654 |
+
"step": 3710
|
| 2655 |
+
},
|
| 2656 |
+
{
|
| 2657 |
+
"epoch": 6.772872098315886,
|
| 2658 |
+
"grad_norm": 0.740304498205183,
|
| 2659 |
+
"learning_rate": 5e-06,
|
| 2660 |
+
"loss": 0.498,
|
| 2661 |
+
"step": 3720
|
| 2662 |
+
},
|
| 2663 |
+
{
|
| 2664 |
+
"epoch": 6.7910787437414655,
|
| 2665 |
+
"grad_norm": 0.7463156387752775,
|
| 2666 |
+
"learning_rate": 5e-06,
|
| 2667 |
+
"loss": 0.4994,
|
| 2668 |
+
"step": 3730
|
| 2669 |
+
},
|
| 2670 |
+
{
|
| 2671 |
+
"epoch": 6.809285389167046,
|
| 2672 |
+
"grad_norm": 0.8016121539442179,
|
| 2673 |
+
"learning_rate": 5e-06,
|
| 2674 |
+
"loss": 0.4997,
|
| 2675 |
+
"step": 3740
|
| 2676 |
+
},
|
| 2677 |
+
{
|
| 2678 |
+
"epoch": 6.827492034592626,
|
| 2679 |
+
"grad_norm": 0.8418329142580222,
|
| 2680 |
+
"learning_rate": 5e-06,
|
| 2681 |
+
"loss": 0.4982,
|
| 2682 |
+
"step": 3750
|
| 2683 |
+
},
|
| 2684 |
+
{
|
| 2685 |
+
"epoch": 6.845698680018207,
|
| 2686 |
+
"grad_norm": 0.8007293534037732,
|
| 2687 |
+
"learning_rate": 5e-06,
|
| 2688 |
+
"loss": 0.5008,
|
| 2689 |
+
"step": 3760
|
| 2690 |
+
},
|
| 2691 |
+
{
|
| 2692 |
+
"epoch": 6.8639053254437865,
|
| 2693 |
+
"grad_norm": 0.7962992985732563,
|
| 2694 |
+
"learning_rate": 5e-06,
|
| 2695 |
+
"loss": 0.4983,
|
| 2696 |
+
"step": 3770
|
| 2697 |
+
},
|
| 2698 |
+
{
|
| 2699 |
+
"epoch": 6.882111970869367,
|
| 2700 |
+
"grad_norm": 0.8192800444422219,
|
| 2701 |
+
"learning_rate": 5e-06,
|
| 2702 |
+
"loss": 0.5012,
|
| 2703 |
+
"step": 3780
|
| 2704 |
+
},
|
| 2705 |
+
{
|
| 2706 |
+
"epoch": 6.900318616294948,
|
| 2707 |
+
"grad_norm": 0.7933819169487161,
|
| 2708 |
+
"learning_rate": 5e-06,
|
| 2709 |
+
"loss": 0.4993,
|
| 2710 |
+
"step": 3790
|
| 2711 |
+
},
|
| 2712 |
+
{
|
| 2713 |
+
"epoch": 6.918525261720528,
|
| 2714 |
+
"grad_norm": 0.7719557878224939,
|
| 2715 |
+
"learning_rate": 5e-06,
|
| 2716 |
+
"loss": 0.4969,
|
| 2717 |
+
"step": 3800
|
| 2718 |
+
},
|
| 2719 |
+
{
|
| 2720 |
+
"epoch": 6.936731907146108,
|
| 2721 |
+
"grad_norm": 0.7794726223761498,
|
| 2722 |
+
"learning_rate": 5e-06,
|
| 2723 |
+
"loss": 0.4997,
|
| 2724 |
+
"step": 3810
|
| 2725 |
+
},
|
| 2726 |
+
{
|
| 2727 |
+
"epoch": 6.954938552571688,
|
| 2728 |
+
"grad_norm": 0.7490795619365134,
|
| 2729 |
+
"learning_rate": 5e-06,
|
| 2730 |
+
"loss": 0.4995,
|
| 2731 |
+
"step": 3820
|
| 2732 |
+
},
|
| 2733 |
+
{
|
| 2734 |
+
"epoch": 6.973145197997269,
|
| 2735 |
+
"grad_norm": 0.8534935347128176,
|
| 2736 |
+
"learning_rate": 5e-06,
|
| 2737 |
+
"loss": 0.5045,
|
| 2738 |
+
"step": 3830
|
| 2739 |
+
},
|
| 2740 |
+
{
|
| 2741 |
+
"epoch": 6.99135184342285,
|
| 2742 |
+
"grad_norm": 0.8266159513619395,
|
| 2743 |
+
"learning_rate": 5e-06,
|
| 2744 |
+
"loss": 0.5002,
|
| 2745 |
+
"step": 3840
|
| 2746 |
+
},
|
| 2747 |
+
{
|
| 2748 |
+
"epoch": 6.998634501593082,
|
| 2749 |
+
"eval_loss": 0.8599892854690552,
|
| 2750 |
+
"eval_runtime": 382.211,
|
| 2751 |
+
"eval_samples_per_second": 38.725,
|
| 2752 |
+
"eval_steps_per_second": 0.607,
|
| 2753 |
+
"step": 3844
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 7.0095584888484295,
|
| 2757 |
+
"grad_norm": 1.4929469131394506,
|
| 2758 |
+
"learning_rate": 5e-06,
|
| 2759 |
+
"loss": 0.507,
|
| 2760 |
+
"step": 3850
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 7.02776513427401,
|
| 2764 |
+
"grad_norm": 0.9806727302100092,
|
| 2765 |
+
"learning_rate": 5e-06,
|
| 2766 |
+
"loss": 0.4258,
|
| 2767 |
+
"step": 3860
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 7.04597177969959,
|
| 2771 |
+
"grad_norm": 1.0186437284049983,
|
| 2772 |
+
"learning_rate": 5e-06,
|
| 2773 |
+
"loss": 0.4304,
|
| 2774 |
+
"step": 3870
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 7.064178425125171,
|
| 2778 |
+
"grad_norm": 0.8346902224337589,
|
| 2779 |
+
"learning_rate": 5e-06,
|
| 2780 |
+
"loss": 0.4315,
|
| 2781 |
+
"step": 3880
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 7.082385070550751,
|
| 2785 |
+
"grad_norm": 0.8895988414802425,
|
| 2786 |
+
"learning_rate": 5e-06,
|
| 2787 |
+
"loss": 0.4321,
|
| 2788 |
+
"step": 3890
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 7.100591715976331,
|
| 2792 |
+
"grad_norm": 0.8968273584319876,
|
| 2793 |
+
"learning_rate": 5e-06,
|
| 2794 |
+
"loss": 0.4316,
|
| 2795 |
+
"step": 3900
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 7.118798361401912,
|
| 2799 |
+
"grad_norm": 1.0430199362279293,
|
| 2800 |
+
"learning_rate": 5e-06,
|
| 2801 |
+
"loss": 0.4352,
|
| 2802 |
+
"step": 3910
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 7.137005006827492,
|
| 2806 |
+
"grad_norm": 0.9925445246056604,
|
| 2807 |
+
"learning_rate": 5e-06,
|
| 2808 |
+
"loss": 0.4335,
|
| 2809 |
+
"step": 3920
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 7.155211652253072,
|
| 2813 |
+
"grad_norm": 0.9236834017922617,
|
| 2814 |
+
"learning_rate": 5e-06,
|
| 2815 |
+
"loss": 0.4344,
|
| 2816 |
+
"step": 3930
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 7.173418297678653,
|
| 2820 |
+
"grad_norm": 0.9006887934639467,
|
| 2821 |
+
"learning_rate": 5e-06,
|
| 2822 |
+
"loss": 0.4354,
|
| 2823 |
+
"step": 3940
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 7.191624943104233,
|
| 2827 |
+
"grad_norm": 1.021746418559959,
|
| 2828 |
+
"learning_rate": 5e-06,
|
| 2829 |
+
"loss": 0.4337,
|
| 2830 |
+
"step": 3950
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 7.209831588529814,
|
| 2834 |
+
"grad_norm": 0.8820115554192361,
|
| 2835 |
+
"learning_rate": 5e-06,
|
| 2836 |
+
"loss": 0.4356,
|
| 2837 |
+
"step": 3960
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 7.2280382339553935,
|
| 2841 |
+
"grad_norm": 0.9450373403984087,
|
| 2842 |
+
"learning_rate": 5e-06,
|
| 2843 |
+
"loss": 0.4346,
|
| 2844 |
+
"step": 3970
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 7.246244879380974,
|
| 2848 |
+
"grad_norm": 0.9599720579833242,
|
| 2849 |
+
"learning_rate": 5e-06,
|
| 2850 |
+
"loss": 0.437,
|
| 2851 |
+
"step": 3980
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 7.264451524806554,
|
| 2855 |
+
"grad_norm": 0.944022553759221,
|
| 2856 |
+
"learning_rate": 5e-06,
|
| 2857 |
+
"loss": 0.4373,
|
| 2858 |
+
"step": 3990
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 7.282658170232135,
|
| 2862 |
+
"grad_norm": 1.1470515929421334,
|
| 2863 |
+
"learning_rate": 5e-06,
|
| 2864 |
+
"loss": 0.437,
|
| 2865 |
+
"step": 4000
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 7.300864815657715,
|
| 2869 |
+
"grad_norm": 0.9229924188228826,
|
| 2870 |
+
"learning_rate": 5e-06,
|
| 2871 |
+
"loss": 0.4412,
|
| 2872 |
+
"step": 4010
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 7.319071461083295,
|
| 2876 |
+
"grad_norm": 1.1004139942729736,
|
| 2877 |
+
"learning_rate": 5e-06,
|
| 2878 |
+
"loss": 0.4404,
|
| 2879 |
+
"step": 4020
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 7.337278106508876,
|
| 2883 |
+
"grad_norm": 0.9234895336641952,
|
| 2884 |
+
"learning_rate": 5e-06,
|
| 2885 |
+
"loss": 0.4413,
|
| 2886 |
+
"step": 4030
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 7.355484751934456,
|
| 2890 |
+
"grad_norm": 0.9191069378971881,
|
| 2891 |
+
"learning_rate": 5e-06,
|
| 2892 |
+
"loss": 0.4374,
|
| 2893 |
+
"step": 4040
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 7.373691397360036,
|
| 2897 |
+
"grad_norm": 0.9565597630772228,
|
| 2898 |
+
"learning_rate": 5e-06,
|
| 2899 |
+
"loss": 0.4427,
|
| 2900 |
+
"step": 4050
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 7.391898042785617,
|
| 2904 |
+
"grad_norm": 0.8991823052622302,
|
| 2905 |
+
"learning_rate": 5e-06,
|
| 2906 |
+
"loss": 0.4408,
|
| 2907 |
+
"step": 4060
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 7.410104688211197,
|
| 2911 |
+
"grad_norm": 0.9463545273318638,
|
| 2912 |
+
"learning_rate": 5e-06,
|
| 2913 |
+
"loss": 0.4428,
|
| 2914 |
+
"step": 4070
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 7.428311333636778,
|
| 2918 |
+
"grad_norm": 0.8995180480466028,
|
| 2919 |
+
"learning_rate": 5e-06,
|
| 2920 |
+
"loss": 0.439,
|
| 2921 |
+
"step": 4080
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 7.4465179790623575,
|
| 2925 |
+
"grad_norm": 0.8848211155155944,
|
| 2926 |
+
"learning_rate": 5e-06,
|
| 2927 |
+
"loss": 0.4408,
|
| 2928 |
+
"step": 4090
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 7.464724624487938,
|
| 2932 |
+
"grad_norm": 0.9344565804246009,
|
| 2933 |
+
"learning_rate": 5e-06,
|
| 2934 |
+
"loss": 0.4426,
|
| 2935 |
+
"step": 4100
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 7.482931269913518,
|
| 2939 |
+
"grad_norm": 0.9397047554338785,
|
| 2940 |
+
"learning_rate": 5e-06,
|
| 2941 |
+
"loss": 0.4422,
|
| 2942 |
+
"step": 4110
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 7.501137915339099,
|
| 2946 |
+
"grad_norm": 1.0544561679078621,
|
| 2947 |
+
"learning_rate": 5e-06,
|
| 2948 |
+
"loss": 0.4433,
|
| 2949 |
+
"step": 4120
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 7.519344560764679,
|
| 2953 |
+
"grad_norm": 1.018669421540764,
|
| 2954 |
+
"learning_rate": 5e-06,
|
| 2955 |
+
"loss": 0.4425,
|
| 2956 |
+
"step": 4130
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 7.537551206190259,
|
| 2960 |
+
"grad_norm": 0.884634669314439,
|
| 2961 |
+
"learning_rate": 5e-06,
|
| 2962 |
+
"loss": 0.4447,
|
| 2963 |
+
"step": 4140
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 7.55575785161584,
|
| 2967 |
+
"grad_norm": 0.8973368938963501,
|
| 2968 |
+
"learning_rate": 5e-06,
|
| 2969 |
+
"loss": 0.4414,
|
| 2970 |
+
"step": 4150
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 7.57396449704142,
|
| 2974 |
+
"grad_norm": 1.0298222416647689,
|
| 2975 |
+
"learning_rate": 5e-06,
|
| 2976 |
+
"loss": 0.4475,
|
| 2977 |
+
"step": 4160
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 7.592171142467,
|
| 2981 |
+
"grad_norm": 0.9090837637474475,
|
| 2982 |
+
"learning_rate": 5e-06,
|
| 2983 |
+
"loss": 0.4427,
|
| 2984 |
+
"step": 4170
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 7.610377787892581,
|
| 2988 |
+
"grad_norm": 0.9114871413052325,
|
| 2989 |
+
"learning_rate": 5e-06,
|
| 2990 |
+
"loss": 0.4464,
|
| 2991 |
+
"step": 4180
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 7.628584433318161,
|
| 2995 |
+
"grad_norm": 0.8765155235059887,
|
| 2996 |
+
"learning_rate": 5e-06,
|
| 2997 |
+
"loss": 0.4459,
|
| 2998 |
+
"step": 4190
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 7.646791078743742,
|
| 3002 |
+
"grad_norm": 0.8572465195574555,
|
| 3003 |
+
"learning_rate": 5e-06,
|
| 3004 |
+
"loss": 0.4466,
|
| 3005 |
+
"step": 4200
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 7.6649977241693215,
|
| 3009 |
+
"grad_norm": 0.8619953574827645,
|
| 3010 |
+
"learning_rate": 5e-06,
|
| 3011 |
+
"loss": 0.4449,
|
| 3012 |
+
"step": 4210
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 7.683204369594902,
|
| 3016 |
+
"grad_norm": 0.9047350844927557,
|
| 3017 |
+
"learning_rate": 5e-06,
|
| 3018 |
+
"loss": 0.4491,
|
| 3019 |
+
"step": 4220
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 7.701411015020483,
|
| 3023 |
+
"grad_norm": 0.8935472556665444,
|
| 3024 |
+
"learning_rate": 5e-06,
|
| 3025 |
+
"loss": 0.4483,
|
| 3026 |
+
"step": 4230
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 7.719617660446063,
|
| 3030 |
+
"grad_norm": 0.8653782388568667,
|
| 3031 |
+
"learning_rate": 5e-06,
|
| 3032 |
+
"loss": 0.4475,
|
| 3033 |
+
"step": 4240
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 7.737824305871643,
|
| 3037 |
+
"grad_norm": 0.881016800265456,
|
| 3038 |
+
"learning_rate": 5e-06,
|
| 3039 |
+
"loss": 0.451,
|
| 3040 |
+
"step": 4250
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 7.756030951297223,
|
| 3044 |
+
"grad_norm": 0.9020813961697224,
|
| 3045 |
+
"learning_rate": 5e-06,
|
| 3046 |
+
"loss": 0.448,
|
| 3047 |
+
"step": 4260
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 7.774237596722804,
|
| 3051 |
+
"grad_norm": 0.9716168118712624,
|
| 3052 |
+
"learning_rate": 5e-06,
|
| 3053 |
+
"loss": 0.4523,
|
| 3054 |
+
"step": 4270
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 7.792444242148385,
|
| 3058 |
+
"grad_norm": 0.9179534064504127,
|
| 3059 |
+
"learning_rate": 5e-06,
|
| 3060 |
+
"loss": 0.4515,
|
| 3061 |
+
"step": 4280
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 7.810650887573964,
|
| 3065 |
+
"grad_norm": 0.9614798912890371,
|
| 3066 |
+
"learning_rate": 5e-06,
|
| 3067 |
+
"loss": 0.448,
|
| 3068 |
+
"step": 4290
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 7.828857532999545,
|
| 3072 |
+
"grad_norm": 0.9044673066961919,
|
| 3073 |
+
"learning_rate": 5e-06,
|
| 3074 |
+
"loss": 0.452,
|
| 3075 |
+
"step": 4300
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 7.847064178425125,
|
| 3079 |
+
"grad_norm": 0.9051177289315938,
|
| 3080 |
+
"learning_rate": 5e-06,
|
| 3081 |
+
"loss": 0.4502,
|
| 3082 |
+
"step": 4310
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 7.865270823850706,
|
| 3086 |
+
"grad_norm": 0.9751136269675025,
|
| 3087 |
+
"learning_rate": 5e-06,
|
| 3088 |
+
"loss": 0.4508,
|
| 3089 |
+
"step": 4320
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 7.883477469276286,
|
| 3093 |
+
"grad_norm": 0.9479807771925007,
|
| 3094 |
+
"learning_rate": 5e-06,
|
| 3095 |
+
"loss": 0.4481,
|
| 3096 |
+
"step": 4330
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 7.901684114701866,
|
| 3100 |
+
"grad_norm": 1.0019388262723266,
|
| 3101 |
+
"learning_rate": 5e-06,
|
| 3102 |
+
"loss": 0.4508,
|
| 3103 |
+
"step": 4340
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 7.919890760127447,
|
| 3107 |
+
"grad_norm": 0.8485770165242006,
|
| 3108 |
+
"learning_rate": 5e-06,
|
| 3109 |
+
"loss": 0.4488,
|
| 3110 |
+
"step": 4350
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 7.938097405553027,
|
| 3114 |
+
"grad_norm": 0.9440844580758145,
|
| 3115 |
+
"learning_rate": 5e-06,
|
| 3116 |
+
"loss": 0.453,
|
| 3117 |
+
"step": 4360
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 7.956304050978607,
|
| 3121 |
+
"grad_norm": 0.9450193155443942,
|
| 3122 |
+
"learning_rate": 5e-06,
|
| 3123 |
+
"loss": 0.4519,
|
| 3124 |
+
"step": 4370
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 7.974510696404187,
|
| 3128 |
+
"grad_norm": 0.9124151727103043,
|
| 3129 |
+
"learning_rate": 5e-06,
|
| 3130 |
+
"loss": 0.4542,
|
| 3131 |
+
"step": 4380
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 7.992717341829768,
|
| 3135 |
+
"grad_norm": 0.9474262270163373,
|
| 3136 |
+
"learning_rate": 5e-06,
|
| 3137 |
+
"loss": 0.4505,
|
| 3138 |
+
"step": 4390
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 7.9963586709148835,
|
| 3142 |
+
"eval_loss": 0.9232881665229797,
|
| 3143 |
+
"eval_runtime": 369.711,
|
| 3144 |
+
"eval_samples_per_second": 40.034,
|
| 3145 |
+
"eval_steps_per_second": 0.628,
|
| 3146 |
+
"step": 4392
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 7.9963586709148835,
|
| 3150 |
+
"step": 4392,
|
| 3151 |
+
"total_flos": 7356554202316800.0,
|
| 3152 |
+
"train_loss": 0.6099752168050664,
|
| 3153 |
+
"train_runtime": 144891.7905,
|
| 3154 |
+
"train_samples_per_second": 15.527,
|
| 3155 |
+
"train_steps_per_second": 0.03
|
| 3156 |
+
}
|
| 3157 |
+
],
|
| 3158 |
+
"logging_steps": 10,
|
| 3159 |
+
"max_steps": 4392,
|
| 3160 |
+
"num_input_tokens_seen": 0,
|
| 3161 |
+
"num_train_epochs": 8,
|
| 3162 |
+
"save_steps": 500,
|
| 3163 |
+
"stateful_callbacks": {
|
| 3164 |
+
"TrainerControl": {
|
| 3165 |
+
"args": {
|
| 3166 |
+
"should_epoch_stop": false,
|
| 3167 |
+
"should_evaluate": false,
|
| 3168 |
+
"should_log": false,
|
| 3169 |
+
"should_save": true,
|
| 3170 |
+
"should_training_stop": true
|
| 3171 |
+
},
|
| 3172 |
+
"attributes": {}
|
| 3173 |
+
}
|
| 3174 |
+
},
|
| 3175 |
+
"total_flos": 7356554202316800.0,
|
| 3176 |
+
"train_batch_size": 8,
|
| 3177 |
+
"trial_name": null,
|
| 3178 |
+
"trial_params": null
|
| 3179 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|