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 +2860 -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_x4_compute_equal
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# oh_scale_x4_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: 0.7223
|
| 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_x4_compute_equal
|
|
|
|
| 16 |
|
| 17 |
# oh_scale_x4_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_4x dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.7223
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.99849454271735,
|
| 3 |
+
"eval_loss": 0.7222821116447449,
|
| 4 |
+
"eval_runtime": 690.268,
|
| 5 |
+
"eval_samples_per_second": 38.896,
|
| 6 |
+
"eval_steps_per_second": 0.608,
|
| 7 |
+
"total_flos": 6673139006177280.0,
|
| 8 |
+
"train_loss": 0.6817261522194468,
|
| 9 |
+
"train_runtime": 130344.1738,
|
| 10 |
+
"train_samples_per_second": 15.654,
|
| 11 |
+
"train_steps_per_second": 0.031
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.99849454271735,
|
| 3 |
+
"eval_loss": 0.7222821116447449,
|
| 4 |
+
"eval_runtime": 690.268,
|
| 5 |
+
"eval_samples_per_second": 38.896,
|
| 6 |
+
"eval_steps_per_second": 0.608
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.99849454271735,
|
| 3 |
+
"total_flos": 6673139006177280.0,
|
| 4 |
+
"train_loss": 0.6817261522194468,
|
| 5 |
+
"train_runtime": 130344.1738,
|
| 6 |
+
"train_samples_per_second": 15.654,
|
| 7 |
+
"train_steps_per_second": 0.031
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2860 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.99849454271735,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3984,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0100363818843307,
|
| 13 |
+
"grad_norm": 33.33474786117092,
|
| 14 |
+
"learning_rate": 5e-06,
|
| 15 |
+
"loss": 1.0592,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0200727637686614,
|
| 20 |
+
"grad_norm": 1.3624139034064995,
|
| 21 |
+
"learning_rate": 5e-06,
|
| 22 |
+
"loss": 0.9442,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.030109145652992095,
|
| 27 |
+
"grad_norm": 1.4119904374590468,
|
| 28 |
+
"learning_rate": 5e-06,
|
| 29 |
+
"loss": 0.8939,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0401455275373228,
|
| 34 |
+
"grad_norm": 1.528739282183849,
|
| 35 |
+
"learning_rate": 5e-06,
|
| 36 |
+
"loss": 0.8733,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.050181909421653494,
|
| 41 |
+
"grad_norm": 1.0731508202908417,
|
| 42 |
+
"learning_rate": 5e-06,
|
| 43 |
+
"loss": 0.8605,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06021829130598419,
|
| 48 |
+
"grad_norm": 0.9332984692607159,
|
| 49 |
+
"learning_rate": 5e-06,
|
| 50 |
+
"loss": 0.8442,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07025467319031489,
|
| 55 |
+
"grad_norm": 0.872377007810567,
|
| 56 |
+
"learning_rate": 5e-06,
|
| 57 |
+
"loss": 0.8326,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0802910550746456,
|
| 62 |
+
"grad_norm": 1.1369570969633511,
|
| 63 |
+
"learning_rate": 5e-06,
|
| 64 |
+
"loss": 0.8255,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09032743695897628,
|
| 69 |
+
"grad_norm": 1.2039587178247435,
|
| 70 |
+
"learning_rate": 5e-06,
|
| 71 |
+
"loss": 0.8131,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.10036381884330699,
|
| 76 |
+
"grad_norm": 0.7786189445202707,
|
| 77 |
+
"learning_rate": 5e-06,
|
| 78 |
+
"loss": 0.8089,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11040020072763769,
|
| 83 |
+
"grad_norm": 0.8167899418786717,
|
| 84 |
+
"learning_rate": 5e-06,
|
| 85 |
+
"loss": 0.8015,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.12043658261196838,
|
| 90 |
+
"grad_norm": 0.7700801877494686,
|
| 91 |
+
"learning_rate": 5e-06,
|
| 92 |
+
"loss": 0.7984,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13047296449629908,
|
| 97 |
+
"grad_norm": 0.7344316347793653,
|
| 98 |
+
"learning_rate": 5e-06,
|
| 99 |
+
"loss": 0.8001,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.14050934638062978,
|
| 104 |
+
"grad_norm": 0.8770050061566579,
|
| 105 |
+
"learning_rate": 5e-06,
|
| 106 |
+
"loss": 0.7932,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1505457282649605,
|
| 111 |
+
"grad_norm": 0.996787607804233,
|
| 112 |
+
"learning_rate": 5e-06,
|
| 113 |
+
"loss": 0.7958,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1605821101492912,
|
| 118 |
+
"grad_norm": 0.6341583113229519,
|
| 119 |
+
"learning_rate": 5e-06,
|
| 120 |
+
"loss": 0.7933,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.17061849203362187,
|
| 125 |
+
"grad_norm": 0.7057153333092503,
|
| 126 |
+
"learning_rate": 5e-06,
|
| 127 |
+
"loss": 0.7864,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.18065487391795257,
|
| 132 |
+
"grad_norm": 0.5875164626639859,
|
| 133 |
+
"learning_rate": 5e-06,
|
| 134 |
+
"loss": 0.7864,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.19069125580228327,
|
| 139 |
+
"grad_norm": 0.696758643929393,
|
| 140 |
+
"learning_rate": 5e-06,
|
| 141 |
+
"loss": 0.7851,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.20072763768661397,
|
| 146 |
+
"grad_norm": 0.6720306069577169,
|
| 147 |
+
"learning_rate": 5e-06,
|
| 148 |
+
"loss": 0.782,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.21076401957094468,
|
| 153 |
+
"grad_norm": 0.7205930987361605,
|
| 154 |
+
"learning_rate": 5e-06,
|
| 155 |
+
"loss": 0.7802,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.22080040145527538,
|
| 160 |
+
"grad_norm": 0.7263170071853633,
|
| 161 |
+
"learning_rate": 5e-06,
|
| 162 |
+
"loss": 0.7766,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.23083678333960608,
|
| 167 |
+
"grad_norm": 0.6222177030840054,
|
| 168 |
+
"learning_rate": 5e-06,
|
| 169 |
+
"loss": 0.7811,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.24087316522393676,
|
| 174 |
+
"grad_norm": 0.7933389948657629,
|
| 175 |
+
"learning_rate": 5e-06,
|
| 176 |
+
"loss": 0.7785,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.25090954710826746,
|
| 181 |
+
"grad_norm": 0.677033021877631,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 0.774,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.26094592899259816,
|
| 188 |
+
"grad_norm": 1.1592201237448847,
|
| 189 |
+
"learning_rate": 5e-06,
|
| 190 |
+
"loss": 0.7685,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.27098231087692887,
|
| 195 |
+
"grad_norm": 0.8219803131042634,
|
| 196 |
+
"learning_rate": 5e-06,
|
| 197 |
+
"loss": 0.7733,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.28101869276125957,
|
| 202 |
+
"grad_norm": 0.8862071608668877,
|
| 203 |
+
"learning_rate": 5e-06,
|
| 204 |
+
"loss": 0.7717,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2910550746455903,
|
| 209 |
+
"grad_norm": 0.7936049147378009,
|
| 210 |
+
"learning_rate": 5e-06,
|
| 211 |
+
"loss": 0.7732,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.301091456529921,
|
| 216 |
+
"grad_norm": 1.0061666072778264,
|
| 217 |
+
"learning_rate": 5e-06,
|
| 218 |
+
"loss": 0.7694,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3111278384142517,
|
| 223 |
+
"grad_norm": 0.8032266175827166,
|
| 224 |
+
"learning_rate": 5e-06,
|
| 225 |
+
"loss": 0.7667,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.3211642202985824,
|
| 230 |
+
"grad_norm": 0.6353992902115431,
|
| 231 |
+
"learning_rate": 5e-06,
|
| 232 |
+
"loss": 0.7651,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3312006021829131,
|
| 237 |
+
"grad_norm": 0.771841795504432,
|
| 238 |
+
"learning_rate": 5e-06,
|
| 239 |
+
"loss": 0.7628,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.34123698406724373,
|
| 244 |
+
"grad_norm": 0.6845658986967137,
|
| 245 |
+
"learning_rate": 5e-06,
|
| 246 |
+
"loss": 0.7717,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.35127336595157443,
|
| 251 |
+
"grad_norm": 0.5670233457983912,
|
| 252 |
+
"learning_rate": 5e-06,
|
| 253 |
+
"loss": 0.7634,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.36130974783590514,
|
| 258 |
+
"grad_norm": 0.6478342801580839,
|
| 259 |
+
"learning_rate": 5e-06,
|
| 260 |
+
"loss": 0.7629,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.37134612972023584,
|
| 265 |
+
"grad_norm": 0.6993852271462582,
|
| 266 |
+
"learning_rate": 5e-06,
|
| 267 |
+
"loss": 0.7633,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.38138251160456654,
|
| 272 |
+
"grad_norm": 0.8808430833699983,
|
| 273 |
+
"learning_rate": 5e-06,
|
| 274 |
+
"loss": 0.7689,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.39141889348889725,
|
| 279 |
+
"grad_norm": 0.8240632882958299,
|
| 280 |
+
"learning_rate": 5e-06,
|
| 281 |
+
"loss": 0.7595,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.40145527537322795,
|
| 286 |
+
"grad_norm": 0.6461166154976471,
|
| 287 |
+
"learning_rate": 5e-06,
|
| 288 |
+
"loss": 0.758,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.41149165725755865,
|
| 293 |
+
"grad_norm": 0.6015328676217312,
|
| 294 |
+
"learning_rate": 5e-06,
|
| 295 |
+
"loss": 0.7615,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.42152803914188935,
|
| 300 |
+
"grad_norm": 0.6844312177243449,
|
| 301 |
+
"learning_rate": 5e-06,
|
| 302 |
+
"loss": 0.7604,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.43156442102622006,
|
| 307 |
+
"grad_norm": 0.5903552392115103,
|
| 308 |
+
"learning_rate": 5e-06,
|
| 309 |
+
"loss": 0.762,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.44160080291055076,
|
| 314 |
+
"grad_norm": 0.6188002518762006,
|
| 315 |
+
"learning_rate": 5e-06,
|
| 316 |
+
"loss": 0.7608,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.45163718479488146,
|
| 321 |
+
"grad_norm": 0.5972767152570548,
|
| 322 |
+
"learning_rate": 5e-06,
|
| 323 |
+
"loss": 0.7498,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.46167356667921217,
|
| 328 |
+
"grad_norm": 0.8557407016909138,
|
| 329 |
+
"learning_rate": 5e-06,
|
| 330 |
+
"loss": 0.7557,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.47170994856354287,
|
| 335 |
+
"grad_norm": 0.5850889857419986,
|
| 336 |
+
"learning_rate": 5e-06,
|
| 337 |
+
"loss": 0.7506,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.4817463304478735,
|
| 342 |
+
"grad_norm": 0.6245054223045923,
|
| 343 |
+
"learning_rate": 5e-06,
|
| 344 |
+
"loss": 0.7577,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.4917827123322042,
|
| 349 |
+
"grad_norm": 0.8464030388596748,
|
| 350 |
+
"learning_rate": 5e-06,
|
| 351 |
+
"loss": 0.7524,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5018190942165349,
|
| 356 |
+
"grad_norm": 0.5975779064483824,
|
| 357 |
+
"learning_rate": 5e-06,
|
| 358 |
+
"loss": 0.7542,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5118554761008657,
|
| 363 |
+
"grad_norm": 0.5557828388284775,
|
| 364 |
+
"learning_rate": 5e-06,
|
| 365 |
+
"loss": 0.7562,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5218918579851963,
|
| 370 |
+
"grad_norm": 0.5041527087050589,
|
| 371 |
+
"learning_rate": 5e-06,
|
| 372 |
+
"loss": 0.7505,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.5319282398695271,
|
| 377 |
+
"grad_norm": 0.606175554740277,
|
| 378 |
+
"learning_rate": 5e-06,
|
| 379 |
+
"loss": 0.7488,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5419646217538577,
|
| 384 |
+
"grad_norm": 0.6932655836595772,
|
| 385 |
+
"learning_rate": 5e-06,
|
| 386 |
+
"loss": 0.7459,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5520010036381884,
|
| 391 |
+
"grad_norm": 0.6342619574494085,
|
| 392 |
+
"learning_rate": 5e-06,
|
| 393 |
+
"loss": 0.748,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5620373855225191,
|
| 398 |
+
"grad_norm": 0.7273685186875936,
|
| 399 |
+
"learning_rate": 5e-06,
|
| 400 |
+
"loss": 0.7514,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.5720737674068498,
|
| 405 |
+
"grad_norm": 0.6175109686722693,
|
| 406 |
+
"learning_rate": 5e-06,
|
| 407 |
+
"loss": 0.7487,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5821101492911805,
|
| 412 |
+
"grad_norm": 0.5906521692541239,
|
| 413 |
+
"learning_rate": 5e-06,
|
| 414 |
+
"loss": 0.7444,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5921465311755112,
|
| 419 |
+
"grad_norm": 0.5770083684995156,
|
| 420 |
+
"learning_rate": 5e-06,
|
| 421 |
+
"loss": 0.7463,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.602182913059842,
|
| 426 |
+
"grad_norm": 0.5611205193626524,
|
| 427 |
+
"learning_rate": 5e-06,
|
| 428 |
+
"loss": 0.7501,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6122192949441726,
|
| 433 |
+
"grad_norm": 0.5626260973998957,
|
| 434 |
+
"learning_rate": 5e-06,
|
| 435 |
+
"loss": 0.7476,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6222556768285034,
|
| 440 |
+
"grad_norm": 0.5359369641368554,
|
| 441 |
+
"learning_rate": 5e-06,
|
| 442 |
+
"loss": 0.7438,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.632292058712834,
|
| 447 |
+
"grad_norm": 0.5986655577379103,
|
| 448 |
+
"learning_rate": 5e-06,
|
| 449 |
+
"loss": 0.7486,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6423284405971648,
|
| 454 |
+
"grad_norm": 0.6386177918361722,
|
| 455 |
+
"learning_rate": 5e-06,
|
| 456 |
+
"loss": 0.7485,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6523648224814954,
|
| 461 |
+
"grad_norm": 0.5978315440058417,
|
| 462 |
+
"learning_rate": 5e-06,
|
| 463 |
+
"loss": 0.7433,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.6624012043658262,
|
| 468 |
+
"grad_norm": 0.6036098551107931,
|
| 469 |
+
"learning_rate": 5e-06,
|
| 470 |
+
"loss": 0.7455,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.6724375862501568,
|
| 475 |
+
"grad_norm": 0.5636202992429957,
|
| 476 |
+
"learning_rate": 5e-06,
|
| 477 |
+
"loss": 0.7466,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6824739681344875,
|
| 482 |
+
"grad_norm": 0.702719868075261,
|
| 483 |
+
"learning_rate": 5e-06,
|
| 484 |
+
"loss": 0.7423,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6925103500188182,
|
| 489 |
+
"grad_norm": 0.5833095780838476,
|
| 490 |
+
"learning_rate": 5e-06,
|
| 491 |
+
"loss": 0.7398,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.7025467319031489,
|
| 496 |
+
"grad_norm": 0.6308085334601381,
|
| 497 |
+
"learning_rate": 5e-06,
|
| 498 |
+
"loss": 0.7429,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7125831137874796,
|
| 503 |
+
"grad_norm": 0.5487784356318717,
|
| 504 |
+
"learning_rate": 5e-06,
|
| 505 |
+
"loss": 0.7411,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.7226194956718103,
|
| 510 |
+
"grad_norm": 0.5865151577104253,
|
| 511 |
+
"learning_rate": 5e-06,
|
| 512 |
+
"loss": 0.7428,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.732655877556141,
|
| 517 |
+
"grad_norm": 0.6872870216284236,
|
| 518 |
+
"learning_rate": 5e-06,
|
| 519 |
+
"loss": 0.7431,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.7426922594404717,
|
| 524 |
+
"grad_norm": 0.7183251325599549,
|
| 525 |
+
"learning_rate": 5e-06,
|
| 526 |
+
"loss": 0.7423,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.7527286413248024,
|
| 531 |
+
"grad_norm": 0.7454754048807214,
|
| 532 |
+
"learning_rate": 5e-06,
|
| 533 |
+
"loss": 0.7387,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.7627650232091331,
|
| 538 |
+
"grad_norm": 0.5531190685288089,
|
| 539 |
+
"learning_rate": 5e-06,
|
| 540 |
+
"loss": 0.743,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.7728014050934638,
|
| 545 |
+
"grad_norm": 0.5534601111291695,
|
| 546 |
+
"learning_rate": 5e-06,
|
| 547 |
+
"loss": 0.741,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.7828377869777945,
|
| 552 |
+
"grad_norm": 0.5724229278027059,
|
| 553 |
+
"learning_rate": 5e-06,
|
| 554 |
+
"loss": 0.742,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.7928741688621253,
|
| 559 |
+
"grad_norm": 0.5444775257450564,
|
| 560 |
+
"learning_rate": 5e-06,
|
| 561 |
+
"loss": 0.7375,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.8029105507464559,
|
| 566 |
+
"grad_norm": 0.5172792176589391,
|
| 567 |
+
"learning_rate": 5e-06,
|
| 568 |
+
"loss": 0.7387,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8129469326307867,
|
| 573 |
+
"grad_norm": 0.5788117817392925,
|
| 574 |
+
"learning_rate": 5e-06,
|
| 575 |
+
"loss": 0.7376,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8229833145151173,
|
| 580 |
+
"grad_norm": 0.5261510224866858,
|
| 581 |
+
"learning_rate": 5e-06,
|
| 582 |
+
"loss": 0.739,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.833019696399448,
|
| 587 |
+
"grad_norm": 0.5280046534368359,
|
| 588 |
+
"learning_rate": 5e-06,
|
| 589 |
+
"loss": 0.7359,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.8430560782837787,
|
| 594 |
+
"grad_norm": 0.5641435748973659,
|
| 595 |
+
"learning_rate": 5e-06,
|
| 596 |
+
"loss": 0.7349,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.8530924601681094,
|
| 601 |
+
"grad_norm": 0.557571684728077,
|
| 602 |
+
"learning_rate": 5e-06,
|
| 603 |
+
"loss": 0.7365,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8631288420524401,
|
| 608 |
+
"grad_norm": 0.6296144133240885,
|
| 609 |
+
"learning_rate": 5e-06,
|
| 610 |
+
"loss": 0.7373,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.8731652239367708,
|
| 615 |
+
"grad_norm": 0.5966607955551202,
|
| 616 |
+
"learning_rate": 5e-06,
|
| 617 |
+
"loss": 0.7353,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.8832016058211015,
|
| 622 |
+
"grad_norm": 0.5563914474354427,
|
| 623 |
+
"learning_rate": 5e-06,
|
| 624 |
+
"loss": 0.7353,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8932379877054322,
|
| 629 |
+
"grad_norm": 0.536620041860774,
|
| 630 |
+
"learning_rate": 5e-06,
|
| 631 |
+
"loss": 0.7338,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.9032743695897629,
|
| 636 |
+
"grad_norm": 0.5387289422962349,
|
| 637 |
+
"learning_rate": 5e-06,
|
| 638 |
+
"loss": 0.7372,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9133107514740936,
|
| 643 |
+
"grad_norm": 0.6013347978585226,
|
| 644 |
+
"learning_rate": 5e-06,
|
| 645 |
+
"loss": 0.7337,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9233471333584243,
|
| 650 |
+
"grad_norm": 0.583886043468759,
|
| 651 |
+
"learning_rate": 5e-06,
|
| 652 |
+
"loss": 0.7336,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.933383515242755,
|
| 657 |
+
"grad_norm": 0.5382696935521819,
|
| 658 |
+
"learning_rate": 5e-06,
|
| 659 |
+
"loss": 0.7368,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.9434198971270857,
|
| 664 |
+
"grad_norm": 0.5653997036590966,
|
| 665 |
+
"learning_rate": 5e-06,
|
| 666 |
+
"loss": 0.7345,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.9534562790114164,
|
| 671 |
+
"grad_norm": 0.6567632931315763,
|
| 672 |
+
"learning_rate": 5e-06,
|
| 673 |
+
"loss": 0.7304,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.963492660895747,
|
| 678 |
+
"grad_norm": 0.6418916205752738,
|
| 679 |
+
"learning_rate": 5e-06,
|
| 680 |
+
"loss": 0.7301,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.9735290427800778,
|
| 685 |
+
"grad_norm": 0.6530499607318279,
|
| 686 |
+
"learning_rate": 5e-06,
|
| 687 |
+
"loss": 0.7387,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.9835654246644084,
|
| 692 |
+
"grad_norm": 0.5551081098759132,
|
| 693 |
+
"learning_rate": 5e-06,
|
| 694 |
+
"loss": 0.7333,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.9936018065487392,
|
| 699 |
+
"grad_norm": 0.6077836691860066,
|
| 700 |
+
"learning_rate": 5e-06,
|
| 701 |
+
"loss": 0.7336,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.9996236356793377,
|
| 706 |
+
"eval_loss": 0.7306540608406067,
|
| 707 |
+
"eval_runtime": 710.0823,
|
| 708 |
+
"eval_samples_per_second": 37.811,
|
| 709 |
+
"eval_steps_per_second": 0.591,
|
| 710 |
+
"step": 996
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.0036381884330698,
|
| 714 |
+
"grad_norm": 0.7501884083987931,
|
| 715 |
+
"learning_rate": 5e-06,
|
| 716 |
+
"loss": 0.7631,
|
| 717 |
+
"step": 1000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.0136745703174006,
|
| 721 |
+
"grad_norm": 0.6377477928653016,
|
| 722 |
+
"learning_rate": 5e-06,
|
| 723 |
+
"loss": 0.6949,
|
| 724 |
+
"step": 1010
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.0237109522017314,
|
| 728 |
+
"grad_norm": 0.5049710429815633,
|
| 729 |
+
"learning_rate": 5e-06,
|
| 730 |
+
"loss": 0.6968,
|
| 731 |
+
"step": 1020
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.033747334086062,
|
| 735 |
+
"grad_norm": 0.6837967725702919,
|
| 736 |
+
"learning_rate": 5e-06,
|
| 737 |
+
"loss": 0.6905,
|
| 738 |
+
"step": 1030
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.0437837159703927,
|
| 742 |
+
"grad_norm": 0.5356223402685194,
|
| 743 |
+
"learning_rate": 5e-06,
|
| 744 |
+
"loss": 0.6922,
|
| 745 |
+
"step": 1040
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.0538200978547234,
|
| 749 |
+
"grad_norm": 0.8370676577254016,
|
| 750 |
+
"learning_rate": 5e-06,
|
| 751 |
+
"loss": 0.6901,
|
| 752 |
+
"step": 1050
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.0638564797390542,
|
| 756 |
+
"grad_norm": 0.6105636596448625,
|
| 757 |
+
"learning_rate": 5e-06,
|
| 758 |
+
"loss": 0.6922,
|
| 759 |
+
"step": 1060
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.0738928616233847,
|
| 763 |
+
"grad_norm": 0.5927877282574158,
|
| 764 |
+
"learning_rate": 5e-06,
|
| 765 |
+
"loss": 0.6979,
|
| 766 |
+
"step": 1070
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.0839292435077155,
|
| 770 |
+
"grad_norm": 0.540102358596148,
|
| 771 |
+
"learning_rate": 5e-06,
|
| 772 |
+
"loss": 0.6958,
|
| 773 |
+
"step": 1080
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.0939656253920462,
|
| 777 |
+
"grad_norm": 0.59126294445945,
|
| 778 |
+
"learning_rate": 5e-06,
|
| 779 |
+
"loss": 0.6931,
|
| 780 |
+
"step": 1090
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.1040020072763768,
|
| 784 |
+
"grad_norm": 0.6608825970633089,
|
| 785 |
+
"learning_rate": 5e-06,
|
| 786 |
+
"loss": 0.6937,
|
| 787 |
+
"step": 1100
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.1140383891607075,
|
| 791 |
+
"grad_norm": 0.5164144607050836,
|
| 792 |
+
"learning_rate": 5e-06,
|
| 793 |
+
"loss": 0.6917,
|
| 794 |
+
"step": 1110
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.1240747710450383,
|
| 798 |
+
"grad_norm": 0.5814527717684509,
|
| 799 |
+
"learning_rate": 5e-06,
|
| 800 |
+
"loss": 0.6972,
|
| 801 |
+
"step": 1120
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.134111152929369,
|
| 805 |
+
"grad_norm": 0.8029370829094067,
|
| 806 |
+
"learning_rate": 5e-06,
|
| 807 |
+
"loss": 0.6911,
|
| 808 |
+
"step": 1130
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.1441475348136996,
|
| 812 |
+
"grad_norm": 0.5560867302991094,
|
| 813 |
+
"learning_rate": 5e-06,
|
| 814 |
+
"loss": 0.6934,
|
| 815 |
+
"step": 1140
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.1541839166980303,
|
| 819 |
+
"grad_norm": 0.5562556075209857,
|
| 820 |
+
"learning_rate": 5e-06,
|
| 821 |
+
"loss": 0.6966,
|
| 822 |
+
"step": 1150
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.164220298582361,
|
| 826 |
+
"grad_norm": 0.5466240170135728,
|
| 827 |
+
"learning_rate": 5e-06,
|
| 828 |
+
"loss": 0.6905,
|
| 829 |
+
"step": 1160
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.1742566804666918,
|
| 833 |
+
"grad_norm": 0.6181691352555871,
|
| 834 |
+
"learning_rate": 5e-06,
|
| 835 |
+
"loss": 0.6949,
|
| 836 |
+
"step": 1170
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.1842930623510224,
|
| 840 |
+
"grad_norm": 0.5816875585540926,
|
| 841 |
+
"learning_rate": 5e-06,
|
| 842 |
+
"loss": 0.6982,
|
| 843 |
+
"step": 1180
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.1943294442353531,
|
| 847 |
+
"grad_norm": 0.6682656613454141,
|
| 848 |
+
"learning_rate": 5e-06,
|
| 849 |
+
"loss": 0.6975,
|
| 850 |
+
"step": 1190
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.204365826119684,
|
| 854 |
+
"grad_norm": 0.5227039044223272,
|
| 855 |
+
"learning_rate": 5e-06,
|
| 856 |
+
"loss": 0.6938,
|
| 857 |
+
"step": 1200
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.2144022080040147,
|
| 861 |
+
"grad_norm": 0.5934796942110975,
|
| 862 |
+
"learning_rate": 5e-06,
|
| 863 |
+
"loss": 0.6928,
|
| 864 |
+
"step": 1210
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.2244385898883452,
|
| 868 |
+
"grad_norm": 0.7797726771495517,
|
| 869 |
+
"learning_rate": 5e-06,
|
| 870 |
+
"loss": 0.6966,
|
| 871 |
+
"step": 1220
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.234474971772676,
|
| 875 |
+
"grad_norm": 0.529604849325101,
|
| 876 |
+
"learning_rate": 5e-06,
|
| 877 |
+
"loss": 0.6923,
|
| 878 |
+
"step": 1230
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.2445113536570067,
|
| 882 |
+
"grad_norm": 0.6023982359465729,
|
| 883 |
+
"learning_rate": 5e-06,
|
| 884 |
+
"loss": 0.6945,
|
| 885 |
+
"step": 1240
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.2545477355413372,
|
| 889 |
+
"grad_norm": 0.5560385701612578,
|
| 890 |
+
"learning_rate": 5e-06,
|
| 891 |
+
"loss": 0.6944,
|
| 892 |
+
"step": 1250
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.264584117425668,
|
| 896 |
+
"grad_norm": 0.5283876353062209,
|
| 897 |
+
"learning_rate": 5e-06,
|
| 898 |
+
"loss": 0.6959,
|
| 899 |
+
"step": 1260
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.2746204993099988,
|
| 903 |
+
"grad_norm": 0.673569525320333,
|
| 904 |
+
"learning_rate": 5e-06,
|
| 905 |
+
"loss": 0.6957,
|
| 906 |
+
"step": 1270
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.2846568811943295,
|
| 910 |
+
"grad_norm": 0.5956349315336121,
|
| 911 |
+
"learning_rate": 5e-06,
|
| 912 |
+
"loss": 0.6919,
|
| 913 |
+
"step": 1280
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.29469326307866,
|
| 917 |
+
"grad_norm": 0.5672304692349656,
|
| 918 |
+
"learning_rate": 5e-06,
|
| 919 |
+
"loss": 0.6886,
|
| 920 |
+
"step": 1290
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.3047296449629908,
|
| 924 |
+
"grad_norm": 0.5877274033488099,
|
| 925 |
+
"learning_rate": 5e-06,
|
| 926 |
+
"loss": 0.6962,
|
| 927 |
+
"step": 1300
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.3147660268473216,
|
| 931 |
+
"grad_norm": 0.6060436152781087,
|
| 932 |
+
"learning_rate": 5e-06,
|
| 933 |
+
"loss": 0.6972,
|
| 934 |
+
"step": 1310
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.3248024087316521,
|
| 938 |
+
"grad_norm": 0.566578140937884,
|
| 939 |
+
"learning_rate": 5e-06,
|
| 940 |
+
"loss": 0.6929,
|
| 941 |
+
"step": 1320
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.3348387906159829,
|
| 945 |
+
"grad_norm": 0.5294558357936279,
|
| 946 |
+
"learning_rate": 5e-06,
|
| 947 |
+
"loss": 0.6879,
|
| 948 |
+
"step": 1330
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.3448751725003136,
|
| 952 |
+
"grad_norm": 0.6237162250745834,
|
| 953 |
+
"learning_rate": 5e-06,
|
| 954 |
+
"loss": 0.6963,
|
| 955 |
+
"step": 1340
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.3549115543846444,
|
| 959 |
+
"grad_norm": 0.5430804692697989,
|
| 960 |
+
"learning_rate": 5e-06,
|
| 961 |
+
"loss": 0.6879,
|
| 962 |
+
"step": 1350
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.3649479362689751,
|
| 966 |
+
"grad_norm": 0.7283462410509066,
|
| 967 |
+
"learning_rate": 5e-06,
|
| 968 |
+
"loss": 0.6976,
|
| 969 |
+
"step": 1360
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.3749843181533057,
|
| 973 |
+
"grad_norm": 0.6917759789030041,
|
| 974 |
+
"learning_rate": 5e-06,
|
| 975 |
+
"loss": 0.6852,
|
| 976 |
+
"step": 1370
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.3850207000376364,
|
| 980 |
+
"grad_norm": 0.5567789537021975,
|
| 981 |
+
"learning_rate": 5e-06,
|
| 982 |
+
"loss": 0.6948,
|
| 983 |
+
"step": 1380
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.3950570819219672,
|
| 987 |
+
"grad_norm": 0.5454605064190993,
|
| 988 |
+
"learning_rate": 5e-06,
|
| 989 |
+
"loss": 0.6954,
|
| 990 |
+
"step": 1390
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.4050934638062977,
|
| 994 |
+
"grad_norm": 0.5900124519339845,
|
| 995 |
+
"learning_rate": 5e-06,
|
| 996 |
+
"loss": 0.6921,
|
| 997 |
+
"step": 1400
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.4151298456906285,
|
| 1001 |
+
"grad_norm": 0.5690895340205371,
|
| 1002 |
+
"learning_rate": 5e-06,
|
| 1003 |
+
"loss": 0.6907,
|
| 1004 |
+
"step": 1410
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.4251662275749593,
|
| 1008 |
+
"grad_norm": 0.5672995415143901,
|
| 1009 |
+
"learning_rate": 5e-06,
|
| 1010 |
+
"loss": 0.6924,
|
| 1011 |
+
"step": 1420
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.43520260945929,
|
| 1015 |
+
"grad_norm": 0.5311142120372314,
|
| 1016 |
+
"learning_rate": 5e-06,
|
| 1017 |
+
"loss": 0.696,
|
| 1018 |
+
"step": 1430
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.4452389913436205,
|
| 1022 |
+
"grad_norm": 0.5999039652092255,
|
| 1023 |
+
"learning_rate": 5e-06,
|
| 1024 |
+
"loss": 0.6902,
|
| 1025 |
+
"step": 1440
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.4552753732279513,
|
| 1029 |
+
"grad_norm": 0.5933508701370325,
|
| 1030 |
+
"learning_rate": 5e-06,
|
| 1031 |
+
"loss": 0.6921,
|
| 1032 |
+
"step": 1450
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.465311755112282,
|
| 1036 |
+
"grad_norm": 0.5165015340634761,
|
| 1037 |
+
"learning_rate": 5e-06,
|
| 1038 |
+
"loss": 0.6926,
|
| 1039 |
+
"step": 1460
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.4753481369966126,
|
| 1043 |
+
"grad_norm": 0.542825456215961,
|
| 1044 |
+
"learning_rate": 5e-06,
|
| 1045 |
+
"loss": 0.6924,
|
| 1046 |
+
"step": 1470
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.4853845188809434,
|
| 1050 |
+
"grad_norm": 0.5173942736182007,
|
| 1051 |
+
"learning_rate": 5e-06,
|
| 1052 |
+
"loss": 0.6889,
|
| 1053 |
+
"step": 1480
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.4954209007652741,
|
| 1057 |
+
"grad_norm": 0.5444379147874959,
|
| 1058 |
+
"learning_rate": 5e-06,
|
| 1059 |
+
"loss": 0.6895,
|
| 1060 |
+
"step": 1490
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.5054572826496049,
|
| 1064 |
+
"grad_norm": 0.5186245031758688,
|
| 1065 |
+
"learning_rate": 5e-06,
|
| 1066 |
+
"loss": 0.6955,
|
| 1067 |
+
"step": 1500
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.5154936645339356,
|
| 1071 |
+
"grad_norm": 0.5273924158023976,
|
| 1072 |
+
"learning_rate": 5e-06,
|
| 1073 |
+
"loss": 0.6971,
|
| 1074 |
+
"step": 1510
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.5255300464182662,
|
| 1078 |
+
"grad_norm": 0.5676022244683365,
|
| 1079 |
+
"learning_rate": 5e-06,
|
| 1080 |
+
"loss": 0.695,
|
| 1081 |
+
"step": 1520
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.535566428302597,
|
| 1085 |
+
"grad_norm": 0.6271761201828187,
|
| 1086 |
+
"learning_rate": 5e-06,
|
| 1087 |
+
"loss": 0.6917,
|
| 1088 |
+
"step": 1530
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.5456028101869275,
|
| 1092 |
+
"grad_norm": 0.6124159982732021,
|
| 1093 |
+
"learning_rate": 5e-06,
|
| 1094 |
+
"loss": 0.6911,
|
| 1095 |
+
"step": 1540
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.5556391920712582,
|
| 1099 |
+
"grad_norm": 0.6862261314401948,
|
| 1100 |
+
"learning_rate": 5e-06,
|
| 1101 |
+
"loss": 0.6864,
|
| 1102 |
+
"step": 1550
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.565675573955589,
|
| 1106 |
+
"grad_norm": 0.6275955880807454,
|
| 1107 |
+
"learning_rate": 5e-06,
|
| 1108 |
+
"loss": 0.6936,
|
| 1109 |
+
"step": 1560
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.5757119558399197,
|
| 1113 |
+
"grad_norm": 0.6289883161946416,
|
| 1114 |
+
"learning_rate": 5e-06,
|
| 1115 |
+
"loss": 0.6894,
|
| 1116 |
+
"step": 1570
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.5857483377242505,
|
| 1120 |
+
"grad_norm": 0.580105480957792,
|
| 1121 |
+
"learning_rate": 5e-06,
|
| 1122 |
+
"loss": 0.693,
|
| 1123 |
+
"step": 1580
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.5957847196085813,
|
| 1127 |
+
"grad_norm": 0.5560879464825351,
|
| 1128 |
+
"learning_rate": 5e-06,
|
| 1129 |
+
"loss": 0.692,
|
| 1130 |
+
"step": 1590
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.6058211014929118,
|
| 1134 |
+
"grad_norm": 0.58969515178897,
|
| 1135 |
+
"learning_rate": 5e-06,
|
| 1136 |
+
"loss": 0.6916,
|
| 1137 |
+
"step": 1600
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.6158574833772426,
|
| 1141 |
+
"grad_norm": 0.626694720233302,
|
| 1142 |
+
"learning_rate": 5e-06,
|
| 1143 |
+
"loss": 0.6867,
|
| 1144 |
+
"step": 1610
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.625893865261573,
|
| 1148 |
+
"grad_norm": 0.5549455637786119,
|
| 1149 |
+
"learning_rate": 5e-06,
|
| 1150 |
+
"loss": 0.6897,
|
| 1151 |
+
"step": 1620
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.6359302471459038,
|
| 1155 |
+
"grad_norm": 0.5938553058669086,
|
| 1156 |
+
"learning_rate": 5e-06,
|
| 1157 |
+
"loss": 0.6921,
|
| 1158 |
+
"step": 1630
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.6459666290302346,
|
| 1162 |
+
"grad_norm": 0.5753903253177306,
|
| 1163 |
+
"learning_rate": 5e-06,
|
| 1164 |
+
"loss": 0.6848,
|
| 1165 |
+
"step": 1640
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.6560030109145654,
|
| 1169 |
+
"grad_norm": 0.5191405989431156,
|
| 1170 |
+
"learning_rate": 5e-06,
|
| 1171 |
+
"loss": 0.69,
|
| 1172 |
+
"step": 1650
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.6660393927988961,
|
| 1176 |
+
"grad_norm": 0.5561359520494403,
|
| 1177 |
+
"learning_rate": 5e-06,
|
| 1178 |
+
"loss": 0.6986,
|
| 1179 |
+
"step": 1660
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.6760757746832267,
|
| 1183 |
+
"grad_norm": 0.6203674148098636,
|
| 1184 |
+
"learning_rate": 5e-06,
|
| 1185 |
+
"loss": 0.6908,
|
| 1186 |
+
"step": 1670
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.6861121565675574,
|
| 1190 |
+
"grad_norm": 0.6125267283499654,
|
| 1191 |
+
"learning_rate": 5e-06,
|
| 1192 |
+
"loss": 0.6914,
|
| 1193 |
+
"step": 1680
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.696148538451888,
|
| 1197 |
+
"grad_norm": 0.6670104437553405,
|
| 1198 |
+
"learning_rate": 5e-06,
|
| 1199 |
+
"loss": 0.6921,
|
| 1200 |
+
"step": 1690
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.7061849203362187,
|
| 1204 |
+
"grad_norm": 0.6164770721352887,
|
| 1205 |
+
"learning_rate": 5e-06,
|
| 1206 |
+
"loss": 0.6906,
|
| 1207 |
+
"step": 1700
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.7162213022205495,
|
| 1211 |
+
"grad_norm": 0.5419884156579401,
|
| 1212 |
+
"learning_rate": 5e-06,
|
| 1213 |
+
"loss": 0.6943,
|
| 1214 |
+
"step": 1710
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.7262576841048802,
|
| 1218 |
+
"grad_norm": 0.5735849533542166,
|
| 1219 |
+
"learning_rate": 5e-06,
|
| 1220 |
+
"loss": 0.69,
|
| 1221 |
+
"step": 1720
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.736294065989211,
|
| 1225 |
+
"grad_norm": 0.5297233454269522,
|
| 1226 |
+
"learning_rate": 5e-06,
|
| 1227 |
+
"loss": 0.6892,
|
| 1228 |
+
"step": 1730
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.7463304478735417,
|
| 1232 |
+
"grad_norm": 0.5393616139715296,
|
| 1233 |
+
"learning_rate": 5e-06,
|
| 1234 |
+
"loss": 0.6915,
|
| 1235 |
+
"step": 1740
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.7563668297578723,
|
| 1239 |
+
"grad_norm": 0.5643636572625912,
|
| 1240 |
+
"learning_rate": 5e-06,
|
| 1241 |
+
"loss": 0.6946,
|
| 1242 |
+
"step": 1750
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.7664032116422028,
|
| 1246 |
+
"grad_norm": 0.5884196950446513,
|
| 1247 |
+
"learning_rate": 5e-06,
|
| 1248 |
+
"loss": 0.6888,
|
| 1249 |
+
"step": 1760
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.7764395935265336,
|
| 1253 |
+
"grad_norm": 0.5236101003130557,
|
| 1254 |
+
"learning_rate": 5e-06,
|
| 1255 |
+
"loss": 0.687,
|
| 1256 |
+
"step": 1770
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.7864759754108643,
|
| 1260 |
+
"grad_norm": 0.5035132462272621,
|
| 1261 |
+
"learning_rate": 5e-06,
|
| 1262 |
+
"loss": 0.6888,
|
| 1263 |
+
"step": 1780
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.796512357295195,
|
| 1267 |
+
"grad_norm": 0.503023782309955,
|
| 1268 |
+
"learning_rate": 5e-06,
|
| 1269 |
+
"loss": 0.697,
|
| 1270 |
+
"step": 1790
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.8065487391795259,
|
| 1274 |
+
"grad_norm": 0.6043844268122752,
|
| 1275 |
+
"learning_rate": 5e-06,
|
| 1276 |
+
"loss": 0.6939,
|
| 1277 |
+
"step": 1800
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.8165851210638566,
|
| 1281 |
+
"grad_norm": 0.5232601740320375,
|
| 1282 |
+
"learning_rate": 5e-06,
|
| 1283 |
+
"loss": 0.6904,
|
| 1284 |
+
"step": 1810
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.8266215029481871,
|
| 1288 |
+
"grad_norm": 0.5370830872283877,
|
| 1289 |
+
"learning_rate": 5e-06,
|
| 1290 |
+
"loss": 0.6911,
|
| 1291 |
+
"step": 1820
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.836657884832518,
|
| 1295 |
+
"grad_norm": 0.6459700462195024,
|
| 1296 |
+
"learning_rate": 5e-06,
|
| 1297 |
+
"loss": 0.6882,
|
| 1298 |
+
"step": 1830
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.8466942667168484,
|
| 1302 |
+
"grad_norm": 0.582019267718641,
|
| 1303 |
+
"learning_rate": 5e-06,
|
| 1304 |
+
"loss": 0.6922,
|
| 1305 |
+
"step": 1840
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.8567306486011792,
|
| 1309 |
+
"grad_norm": 0.5870687122095531,
|
| 1310 |
+
"learning_rate": 5e-06,
|
| 1311 |
+
"loss": 0.6922,
|
| 1312 |
+
"step": 1850
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.86676703048551,
|
| 1316 |
+
"grad_norm": 0.5078692726109669,
|
| 1317 |
+
"learning_rate": 5e-06,
|
| 1318 |
+
"loss": 0.6885,
|
| 1319 |
+
"step": 1860
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.8768034123698407,
|
| 1323 |
+
"grad_norm": 0.5040649883447891,
|
| 1324 |
+
"learning_rate": 5e-06,
|
| 1325 |
+
"loss": 0.6881,
|
| 1326 |
+
"step": 1870
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.8868397942541715,
|
| 1330 |
+
"grad_norm": 0.4995959267347782,
|
| 1331 |
+
"learning_rate": 5e-06,
|
| 1332 |
+
"loss": 0.6862,
|
| 1333 |
+
"step": 1880
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.8968761761385022,
|
| 1337 |
+
"grad_norm": 0.5881241901658035,
|
| 1338 |
+
"learning_rate": 5e-06,
|
| 1339 |
+
"loss": 0.6917,
|
| 1340 |
+
"step": 1890
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.9069125580228328,
|
| 1344 |
+
"grad_norm": 0.5588343821948456,
|
| 1345 |
+
"learning_rate": 5e-06,
|
| 1346 |
+
"loss": 0.6888,
|
| 1347 |
+
"step": 1900
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.9169489399071633,
|
| 1351 |
+
"grad_norm": 0.4870751715972866,
|
| 1352 |
+
"learning_rate": 5e-06,
|
| 1353 |
+
"loss": 0.6872,
|
| 1354 |
+
"step": 1910
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.926985321791494,
|
| 1358 |
+
"grad_norm": 0.5735655591316463,
|
| 1359 |
+
"learning_rate": 5e-06,
|
| 1360 |
+
"loss": 0.6947,
|
| 1361 |
+
"step": 1920
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.9370217036758248,
|
| 1365 |
+
"grad_norm": 0.6085871436334919,
|
| 1366 |
+
"learning_rate": 5e-06,
|
| 1367 |
+
"loss": 0.6888,
|
| 1368 |
+
"step": 1930
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.9470580855601556,
|
| 1372 |
+
"grad_norm": 0.6006262589490265,
|
| 1373 |
+
"learning_rate": 5e-06,
|
| 1374 |
+
"loss": 0.6884,
|
| 1375 |
+
"step": 1940
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.9570944674444863,
|
| 1379 |
+
"grad_norm": 0.6341474482034111,
|
| 1380 |
+
"learning_rate": 5e-06,
|
| 1381 |
+
"loss": 0.6886,
|
| 1382 |
+
"step": 1950
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.967130849328817,
|
| 1386 |
+
"grad_norm": 0.5362807629078432,
|
| 1387 |
+
"learning_rate": 5e-06,
|
| 1388 |
+
"loss": 0.6873,
|
| 1389 |
+
"step": 1960
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.9771672312131476,
|
| 1393 |
+
"grad_norm": 0.4940952174490118,
|
| 1394 |
+
"learning_rate": 5e-06,
|
| 1395 |
+
"loss": 0.6895,
|
| 1396 |
+
"step": 1970
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.9872036130974784,
|
| 1400 |
+
"grad_norm": 0.5068555383392377,
|
| 1401 |
+
"learning_rate": 5e-06,
|
| 1402 |
+
"loss": 0.6886,
|
| 1403 |
+
"step": 1980
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.997239994981809,
|
| 1407 |
+
"grad_norm": 0.5115531434616939,
|
| 1408 |
+
"learning_rate": 5e-06,
|
| 1409 |
+
"loss": 0.6923,
|
| 1410 |
+
"step": 1990
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.999247271358675,
|
| 1414 |
+
"eval_loss": 0.7167317867279053,
|
| 1415 |
+
"eval_runtime": 681.5165,
|
| 1416 |
+
"eval_samples_per_second": 39.396,
|
| 1417 |
+
"eval_steps_per_second": 0.616,
|
| 1418 |
+
"step": 1992
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 2.0072763768661397,
|
| 1422 |
+
"grad_norm": 0.7464092668086445,
|
| 1423 |
+
"learning_rate": 5e-06,
|
| 1424 |
+
"loss": 0.7021,
|
| 1425 |
+
"step": 2000
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 2.0173127587504704,
|
| 1429 |
+
"grad_norm": 0.6285168143241044,
|
| 1430 |
+
"learning_rate": 5e-06,
|
| 1431 |
+
"loss": 0.6464,
|
| 1432 |
+
"step": 2010
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 2.027349140634801,
|
| 1436 |
+
"grad_norm": 0.6153973121048567,
|
| 1437 |
+
"learning_rate": 5e-06,
|
| 1438 |
+
"loss": 0.6468,
|
| 1439 |
+
"step": 2020
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 2.037385522519132,
|
| 1443 |
+
"grad_norm": 0.5873302210415512,
|
| 1444 |
+
"learning_rate": 5e-06,
|
| 1445 |
+
"loss": 0.6466,
|
| 1446 |
+
"step": 2030
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 2.0474219044034627,
|
| 1450 |
+
"grad_norm": 0.6320791509681155,
|
| 1451 |
+
"learning_rate": 5e-06,
|
| 1452 |
+
"loss": 0.645,
|
| 1453 |
+
"step": 2040
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 2.057458286287793,
|
| 1457 |
+
"grad_norm": 0.7492672540101699,
|
| 1458 |
+
"learning_rate": 5e-06,
|
| 1459 |
+
"loss": 0.6484,
|
| 1460 |
+
"step": 2050
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 2.067494668172124,
|
| 1464 |
+
"grad_norm": 0.6049252570937941,
|
| 1465 |
+
"learning_rate": 5e-06,
|
| 1466 |
+
"loss": 0.6479,
|
| 1467 |
+
"step": 2060
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 2.0775310500564546,
|
| 1471 |
+
"grad_norm": 0.5658472447863507,
|
| 1472 |
+
"learning_rate": 5e-06,
|
| 1473 |
+
"loss": 0.646,
|
| 1474 |
+
"step": 2070
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 2.0875674319407853,
|
| 1478 |
+
"grad_norm": 0.5516870136698201,
|
| 1479 |
+
"learning_rate": 5e-06,
|
| 1480 |
+
"loss": 0.6492,
|
| 1481 |
+
"step": 2080
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 2.097603813825116,
|
| 1485 |
+
"grad_norm": 0.7029586116565276,
|
| 1486 |
+
"learning_rate": 5e-06,
|
| 1487 |
+
"loss": 0.6538,
|
| 1488 |
+
"step": 2090
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 2.107640195709447,
|
| 1492 |
+
"grad_norm": 0.5501781567288918,
|
| 1493 |
+
"learning_rate": 5e-06,
|
| 1494 |
+
"loss": 0.6468,
|
| 1495 |
+
"step": 2100
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 2.1176765775937776,
|
| 1499 |
+
"grad_norm": 0.5732859435604044,
|
| 1500 |
+
"learning_rate": 5e-06,
|
| 1501 |
+
"loss": 0.6468,
|
| 1502 |
+
"step": 2110
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 2.1277129594781083,
|
| 1506 |
+
"grad_norm": 0.6097094231081762,
|
| 1507 |
+
"learning_rate": 5e-06,
|
| 1508 |
+
"loss": 0.645,
|
| 1509 |
+
"step": 2120
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 2.1377493413624387,
|
| 1513 |
+
"grad_norm": 0.6168934436640728,
|
| 1514 |
+
"learning_rate": 5e-06,
|
| 1515 |
+
"loss": 0.6495,
|
| 1516 |
+
"step": 2130
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 2.1477857232467694,
|
| 1520 |
+
"grad_norm": 0.5791030315614372,
|
| 1521 |
+
"learning_rate": 5e-06,
|
| 1522 |
+
"loss": 0.6494,
|
| 1523 |
+
"step": 2140
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 2.1578221051311,
|
| 1527 |
+
"grad_norm": 0.5248289813751731,
|
| 1528 |
+
"learning_rate": 5e-06,
|
| 1529 |
+
"loss": 0.6502,
|
| 1530 |
+
"step": 2150
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 2.167858487015431,
|
| 1534 |
+
"grad_norm": 0.5339345181224099,
|
| 1535 |
+
"learning_rate": 5e-06,
|
| 1536 |
+
"loss": 0.6469,
|
| 1537 |
+
"step": 2160
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 2.1778948688997617,
|
| 1541 |
+
"grad_norm": 0.5503349261990413,
|
| 1542 |
+
"learning_rate": 5e-06,
|
| 1543 |
+
"loss": 0.6445,
|
| 1544 |
+
"step": 2170
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 2.1879312507840925,
|
| 1548 |
+
"grad_norm": 0.5464800148102408,
|
| 1549 |
+
"learning_rate": 5e-06,
|
| 1550 |
+
"loss": 0.6506,
|
| 1551 |
+
"step": 2180
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 2.197967632668423,
|
| 1555 |
+
"grad_norm": 0.5247947975936182,
|
| 1556 |
+
"learning_rate": 5e-06,
|
| 1557 |
+
"loss": 0.6527,
|
| 1558 |
+
"step": 2190
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 2.2080040145527535,
|
| 1562 |
+
"grad_norm": 0.5606999830634875,
|
| 1563 |
+
"learning_rate": 5e-06,
|
| 1564 |
+
"loss": 0.6493,
|
| 1565 |
+
"step": 2200
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 2.2180403964370843,
|
| 1569 |
+
"grad_norm": 0.5331049457160831,
|
| 1570 |
+
"learning_rate": 5e-06,
|
| 1571 |
+
"loss": 0.6537,
|
| 1572 |
+
"step": 2210
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 2.228076778321415,
|
| 1576 |
+
"grad_norm": 0.5988937144831796,
|
| 1577 |
+
"learning_rate": 5e-06,
|
| 1578 |
+
"loss": 0.6491,
|
| 1579 |
+
"step": 2220
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 2.238113160205746,
|
| 1583 |
+
"grad_norm": 0.5493199529484449,
|
| 1584 |
+
"learning_rate": 5e-06,
|
| 1585 |
+
"loss": 0.6497,
|
| 1586 |
+
"step": 2230
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 2.2481495420900766,
|
| 1590 |
+
"grad_norm": 0.5391276137070077,
|
| 1591 |
+
"learning_rate": 5e-06,
|
| 1592 |
+
"loss": 0.6479,
|
| 1593 |
+
"step": 2240
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 2.2581859239744073,
|
| 1597 |
+
"grad_norm": 0.5468874755147779,
|
| 1598 |
+
"learning_rate": 5e-06,
|
| 1599 |
+
"loss": 0.6481,
|
| 1600 |
+
"step": 2250
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 2.268222305858738,
|
| 1604 |
+
"grad_norm": 0.6698860969319781,
|
| 1605 |
+
"learning_rate": 5e-06,
|
| 1606 |
+
"loss": 0.6496,
|
| 1607 |
+
"step": 2260
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 2.278258687743069,
|
| 1611 |
+
"grad_norm": 0.5534091734548499,
|
| 1612 |
+
"learning_rate": 5e-06,
|
| 1613 |
+
"loss": 0.6495,
|
| 1614 |
+
"step": 2270
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 2.288295069627399,
|
| 1618 |
+
"grad_norm": 0.5251666746216365,
|
| 1619 |
+
"learning_rate": 5e-06,
|
| 1620 |
+
"loss": 0.6488,
|
| 1621 |
+
"step": 2280
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 2.29833145151173,
|
| 1625 |
+
"grad_norm": 0.6041634029692523,
|
| 1626 |
+
"learning_rate": 5e-06,
|
| 1627 |
+
"loss": 0.6463,
|
| 1628 |
+
"step": 2290
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 2.3083678333960607,
|
| 1632 |
+
"grad_norm": 0.6700682402301694,
|
| 1633 |
+
"learning_rate": 5e-06,
|
| 1634 |
+
"loss": 0.6512,
|
| 1635 |
+
"step": 2300
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 2.3184042152803914,
|
| 1639 |
+
"grad_norm": 0.5507167873654453,
|
| 1640 |
+
"learning_rate": 5e-06,
|
| 1641 |
+
"loss": 0.6531,
|
| 1642 |
+
"step": 2310
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 2.328440597164722,
|
| 1646 |
+
"grad_norm": 0.7110576199813733,
|
| 1647 |
+
"learning_rate": 5e-06,
|
| 1648 |
+
"loss": 0.651,
|
| 1649 |
+
"step": 2320
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 2.338476979049053,
|
| 1653 |
+
"grad_norm": 0.547011003517586,
|
| 1654 |
+
"learning_rate": 5e-06,
|
| 1655 |
+
"loss": 0.6496,
|
| 1656 |
+
"step": 2330
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 2.3485133609333837,
|
| 1660 |
+
"grad_norm": 0.5277892887616297,
|
| 1661 |
+
"learning_rate": 5e-06,
|
| 1662 |
+
"loss": 0.6478,
|
| 1663 |
+
"step": 2340
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 2.358549742817714,
|
| 1667 |
+
"grad_norm": 0.5418890015733175,
|
| 1668 |
+
"learning_rate": 5e-06,
|
| 1669 |
+
"loss": 0.6474,
|
| 1670 |
+
"step": 2350
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 2.3685861247020448,
|
| 1674 |
+
"grad_norm": 0.5453644592494074,
|
| 1675 |
+
"learning_rate": 5e-06,
|
| 1676 |
+
"loss": 0.6513,
|
| 1677 |
+
"step": 2360
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 2.3786225065863755,
|
| 1681 |
+
"grad_norm": 0.5570648512157045,
|
| 1682 |
+
"learning_rate": 5e-06,
|
| 1683 |
+
"loss": 0.6549,
|
| 1684 |
+
"step": 2370
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 2.3886588884707063,
|
| 1688 |
+
"grad_norm": 0.6533123491533362,
|
| 1689 |
+
"learning_rate": 5e-06,
|
| 1690 |
+
"loss": 0.6517,
|
| 1691 |
+
"step": 2380
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 2.398695270355037,
|
| 1695 |
+
"grad_norm": 0.5514173498092635,
|
| 1696 |
+
"learning_rate": 5e-06,
|
| 1697 |
+
"loss": 0.6479,
|
| 1698 |
+
"step": 2390
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 2.408731652239368,
|
| 1702 |
+
"grad_norm": 0.6169109034186734,
|
| 1703 |
+
"learning_rate": 5e-06,
|
| 1704 |
+
"loss": 0.6497,
|
| 1705 |
+
"step": 2400
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 2.4187680341236986,
|
| 1709 |
+
"grad_norm": 0.6092701719038681,
|
| 1710 |
+
"learning_rate": 5e-06,
|
| 1711 |
+
"loss": 0.6523,
|
| 1712 |
+
"step": 2410
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 2.4288044160080293,
|
| 1716 |
+
"grad_norm": 0.5995748884635175,
|
| 1717 |
+
"learning_rate": 5e-06,
|
| 1718 |
+
"loss": 0.653,
|
| 1719 |
+
"step": 2420
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 2.4388407978923596,
|
| 1723 |
+
"grad_norm": 0.5733910522734972,
|
| 1724 |
+
"learning_rate": 5e-06,
|
| 1725 |
+
"loss": 0.6482,
|
| 1726 |
+
"step": 2430
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 2.4488771797766904,
|
| 1730 |
+
"grad_norm": 0.5191638101081872,
|
| 1731 |
+
"learning_rate": 5e-06,
|
| 1732 |
+
"loss": 0.656,
|
| 1733 |
+
"step": 2440
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 2.458913561661021,
|
| 1737 |
+
"grad_norm": 0.8798002303204917,
|
| 1738 |
+
"learning_rate": 5e-06,
|
| 1739 |
+
"loss": 0.6501,
|
| 1740 |
+
"step": 2450
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 2.468949943545352,
|
| 1744 |
+
"grad_norm": 0.5055430681257954,
|
| 1745 |
+
"learning_rate": 5e-06,
|
| 1746 |
+
"loss": 0.6526,
|
| 1747 |
+
"step": 2460
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 2.4789863254296827,
|
| 1751 |
+
"grad_norm": 0.5305005320670955,
|
| 1752 |
+
"learning_rate": 5e-06,
|
| 1753 |
+
"loss": 0.6579,
|
| 1754 |
+
"step": 2470
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 2.4890227073140134,
|
| 1758 |
+
"grad_norm": 0.6181982492315344,
|
| 1759 |
+
"learning_rate": 5e-06,
|
| 1760 |
+
"loss": 0.6473,
|
| 1761 |
+
"step": 2480
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 2.499059089198344,
|
| 1765 |
+
"grad_norm": 0.6215467401287779,
|
| 1766 |
+
"learning_rate": 5e-06,
|
| 1767 |
+
"loss": 0.6511,
|
| 1768 |
+
"step": 2490
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 2.5090954710826745,
|
| 1772 |
+
"grad_norm": 0.5729138634014542,
|
| 1773 |
+
"learning_rate": 5e-06,
|
| 1774 |
+
"loss": 0.6577,
|
| 1775 |
+
"step": 2500
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 2.5191318529670053,
|
| 1779 |
+
"grad_norm": 0.5393679833546277,
|
| 1780 |
+
"learning_rate": 5e-06,
|
| 1781 |
+
"loss": 0.6536,
|
| 1782 |
+
"step": 2510
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 2.529168234851336,
|
| 1786 |
+
"grad_norm": 0.534381658436043,
|
| 1787 |
+
"learning_rate": 5e-06,
|
| 1788 |
+
"loss": 0.6489,
|
| 1789 |
+
"step": 2520
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 2.5392046167356668,
|
| 1793 |
+
"grad_norm": 0.539600655245499,
|
| 1794 |
+
"learning_rate": 5e-06,
|
| 1795 |
+
"loss": 0.6495,
|
| 1796 |
+
"step": 2530
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 2.5492409986199975,
|
| 1800 |
+
"grad_norm": 0.6226376194292436,
|
| 1801 |
+
"learning_rate": 5e-06,
|
| 1802 |
+
"loss": 0.651,
|
| 1803 |
+
"step": 2540
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 2.5592773805043283,
|
| 1807 |
+
"grad_norm": 0.5865717492190782,
|
| 1808 |
+
"learning_rate": 5e-06,
|
| 1809 |
+
"loss": 0.6541,
|
| 1810 |
+
"step": 2550
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 2.569313762388659,
|
| 1814 |
+
"grad_norm": 0.6438993538240664,
|
| 1815 |
+
"learning_rate": 5e-06,
|
| 1816 |
+
"loss": 0.6507,
|
| 1817 |
+
"step": 2560
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 2.57935014427299,
|
| 1821 |
+
"grad_norm": 0.7518127666950809,
|
| 1822 |
+
"learning_rate": 5e-06,
|
| 1823 |
+
"loss": 0.6499,
|
| 1824 |
+
"step": 2570
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 2.58938652615732,
|
| 1828 |
+
"grad_norm": 0.586864178840436,
|
| 1829 |
+
"learning_rate": 5e-06,
|
| 1830 |
+
"loss": 0.6616,
|
| 1831 |
+
"step": 2580
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 2.599422908041651,
|
| 1835 |
+
"grad_norm": 0.5941413788184908,
|
| 1836 |
+
"learning_rate": 5e-06,
|
| 1837 |
+
"loss": 0.6486,
|
| 1838 |
+
"step": 2590
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 2.6094592899259816,
|
| 1842 |
+
"grad_norm": 0.5451550588012898,
|
| 1843 |
+
"learning_rate": 5e-06,
|
| 1844 |
+
"loss": 0.6544,
|
| 1845 |
+
"step": 2600
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 2.6194956718103124,
|
| 1849 |
+
"grad_norm": 0.5147403638954634,
|
| 1850 |
+
"learning_rate": 5e-06,
|
| 1851 |
+
"loss": 0.6539,
|
| 1852 |
+
"step": 2610
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 2.629532053694643,
|
| 1856 |
+
"grad_norm": 0.5467938780782444,
|
| 1857 |
+
"learning_rate": 5e-06,
|
| 1858 |
+
"loss": 0.6475,
|
| 1859 |
+
"step": 2620
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 2.639568435578974,
|
| 1863 |
+
"grad_norm": 0.5458940947855774,
|
| 1864 |
+
"learning_rate": 5e-06,
|
| 1865 |
+
"loss": 0.6501,
|
| 1866 |
+
"step": 2630
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 2.6496048174633042,
|
| 1870 |
+
"grad_norm": 0.600288340260203,
|
| 1871 |
+
"learning_rate": 5e-06,
|
| 1872 |
+
"loss": 0.6533,
|
| 1873 |
+
"step": 2640
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 2.659641199347635,
|
| 1877 |
+
"grad_norm": 0.5329802788249394,
|
| 1878 |
+
"learning_rate": 5e-06,
|
| 1879 |
+
"loss": 0.6539,
|
| 1880 |
+
"step": 2650
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 2.6696775812319657,
|
| 1884 |
+
"grad_norm": 0.5494428968278945,
|
| 1885 |
+
"learning_rate": 5e-06,
|
| 1886 |
+
"loss": 0.6485,
|
| 1887 |
+
"step": 2660
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 2.6797139631162965,
|
| 1891 |
+
"grad_norm": 0.5583817948296362,
|
| 1892 |
+
"learning_rate": 5e-06,
|
| 1893 |
+
"loss": 0.6522,
|
| 1894 |
+
"step": 2670
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 2.6897503450006273,
|
| 1898 |
+
"grad_norm": 0.5760005041925667,
|
| 1899 |
+
"learning_rate": 5e-06,
|
| 1900 |
+
"loss": 0.6513,
|
| 1901 |
+
"step": 2680
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 2.699786726884958,
|
| 1905 |
+
"grad_norm": 0.6094121074922789,
|
| 1906 |
+
"learning_rate": 5e-06,
|
| 1907 |
+
"loss": 0.6507,
|
| 1908 |
+
"step": 2690
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 2.709823108769289,
|
| 1912 |
+
"grad_norm": 0.6168874365624798,
|
| 1913 |
+
"learning_rate": 5e-06,
|
| 1914 |
+
"loss": 0.6532,
|
| 1915 |
+
"step": 2700
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 2.7198594906536195,
|
| 1919 |
+
"grad_norm": 0.5204111827775895,
|
| 1920 |
+
"learning_rate": 5e-06,
|
| 1921 |
+
"loss": 0.652,
|
| 1922 |
+
"step": 2710
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 2.7298958725379503,
|
| 1926 |
+
"grad_norm": 0.5414555391536807,
|
| 1927 |
+
"learning_rate": 5e-06,
|
| 1928 |
+
"loss": 0.6529,
|
| 1929 |
+
"step": 2720
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 2.7399322544222806,
|
| 1933 |
+
"grad_norm": 0.5343532711007503,
|
| 1934 |
+
"learning_rate": 5e-06,
|
| 1935 |
+
"loss": 0.6551,
|
| 1936 |
+
"step": 2730
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 2.7499686363066114,
|
| 1940 |
+
"grad_norm": 0.5730034177518993,
|
| 1941 |
+
"learning_rate": 5e-06,
|
| 1942 |
+
"loss": 0.6559,
|
| 1943 |
+
"step": 2740
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 2.760005018190942,
|
| 1947 |
+
"grad_norm": 0.5413799739509437,
|
| 1948 |
+
"learning_rate": 5e-06,
|
| 1949 |
+
"loss": 0.6536,
|
| 1950 |
+
"step": 2750
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 2.770041400075273,
|
| 1954 |
+
"grad_norm": 0.58185222058302,
|
| 1955 |
+
"learning_rate": 5e-06,
|
| 1956 |
+
"loss": 0.6607,
|
| 1957 |
+
"step": 2760
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 2.7800777819596036,
|
| 1961 |
+
"grad_norm": 0.5659358615987367,
|
| 1962 |
+
"learning_rate": 5e-06,
|
| 1963 |
+
"loss": 0.6532,
|
| 1964 |
+
"step": 2770
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 2.7901141638439344,
|
| 1968 |
+
"grad_norm": 0.5269963486852614,
|
| 1969 |
+
"learning_rate": 5e-06,
|
| 1970 |
+
"loss": 0.6516,
|
| 1971 |
+
"step": 2780
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 2.8001505457282647,
|
| 1975 |
+
"grad_norm": 0.6056482620803397,
|
| 1976 |
+
"learning_rate": 5e-06,
|
| 1977 |
+
"loss": 0.6552,
|
| 1978 |
+
"step": 2790
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 2.8101869276125955,
|
| 1982 |
+
"grad_norm": 1.0590493469402826,
|
| 1983 |
+
"learning_rate": 5e-06,
|
| 1984 |
+
"loss": 0.6556,
|
| 1985 |
+
"step": 2800
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 2.8202233094969262,
|
| 1989 |
+
"grad_norm": 0.8284099367079102,
|
| 1990 |
+
"learning_rate": 5e-06,
|
| 1991 |
+
"loss": 0.6549,
|
| 1992 |
+
"step": 2810
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 2.830259691381257,
|
| 1996 |
+
"grad_norm": 0.6501574839000807,
|
| 1997 |
+
"learning_rate": 5e-06,
|
| 1998 |
+
"loss": 0.6507,
|
| 1999 |
+
"step": 2820
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 2.8402960732655878,
|
| 2003 |
+
"grad_norm": 0.5410870388612636,
|
| 2004 |
+
"learning_rate": 5e-06,
|
| 2005 |
+
"loss": 0.6526,
|
| 2006 |
+
"step": 2830
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 2.8503324551499185,
|
| 2010 |
+
"grad_norm": 0.6510662758185398,
|
| 2011 |
+
"learning_rate": 5e-06,
|
| 2012 |
+
"loss": 0.6519,
|
| 2013 |
+
"step": 2840
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 2.8603688370342493,
|
| 2017 |
+
"grad_norm": 0.5238204299865007,
|
| 2018 |
+
"learning_rate": 5e-06,
|
| 2019 |
+
"loss": 0.6593,
|
| 2020 |
+
"step": 2850
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 2.87040521891858,
|
| 2024 |
+
"grad_norm": 0.5397683523247623,
|
| 2025 |
+
"learning_rate": 5e-06,
|
| 2026 |
+
"loss": 0.6526,
|
| 2027 |
+
"step": 2860
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 2.880441600802911,
|
| 2031 |
+
"grad_norm": 0.5558371306126749,
|
| 2032 |
+
"learning_rate": 5e-06,
|
| 2033 |
+
"loss": 0.6521,
|
| 2034 |
+
"step": 2870
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 2.890477982687241,
|
| 2038 |
+
"grad_norm": 0.5084123704706736,
|
| 2039 |
+
"learning_rate": 5e-06,
|
| 2040 |
+
"loss": 0.655,
|
| 2041 |
+
"step": 2880
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 2.900514364571572,
|
| 2045 |
+
"grad_norm": 0.5363806596935408,
|
| 2046 |
+
"learning_rate": 5e-06,
|
| 2047 |
+
"loss": 0.6523,
|
| 2048 |
+
"step": 2890
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 2.9105507464559026,
|
| 2052 |
+
"grad_norm": 0.526536425381128,
|
| 2053 |
+
"learning_rate": 5e-06,
|
| 2054 |
+
"loss": 0.6549,
|
| 2055 |
+
"step": 2900
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 2.9205871283402334,
|
| 2059 |
+
"grad_norm": 0.5720711163344511,
|
| 2060 |
+
"learning_rate": 5e-06,
|
| 2061 |
+
"loss": 0.6516,
|
| 2062 |
+
"step": 2910
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 2.930623510224564,
|
| 2066 |
+
"grad_norm": 0.5168344329750222,
|
| 2067 |
+
"learning_rate": 5e-06,
|
| 2068 |
+
"loss": 0.6505,
|
| 2069 |
+
"step": 2920
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 2.940659892108895,
|
| 2073 |
+
"grad_norm": 0.5068041805158231,
|
| 2074 |
+
"learning_rate": 5e-06,
|
| 2075 |
+
"loss": 0.6523,
|
| 2076 |
+
"step": 2930
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 2.950696273993225,
|
| 2080 |
+
"grad_norm": 0.5854150052744146,
|
| 2081 |
+
"learning_rate": 5e-06,
|
| 2082 |
+
"loss": 0.6562,
|
| 2083 |
+
"step": 2940
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 2.960732655877556,
|
| 2087 |
+
"grad_norm": 0.6196831772444017,
|
| 2088 |
+
"learning_rate": 5e-06,
|
| 2089 |
+
"loss": 0.6542,
|
| 2090 |
+
"step": 2950
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 2.9707690377618867,
|
| 2094 |
+
"grad_norm": 0.5314563776407271,
|
| 2095 |
+
"learning_rate": 5e-06,
|
| 2096 |
+
"loss": 0.6561,
|
| 2097 |
+
"step": 2960
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 2.9808054196462175,
|
| 2101 |
+
"grad_norm": 0.5039099479336567,
|
| 2102 |
+
"learning_rate": 5e-06,
|
| 2103 |
+
"loss": 0.6563,
|
| 2104 |
+
"step": 2970
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 2.9908418015305482,
|
| 2108 |
+
"grad_norm": 0.5313007683420622,
|
| 2109 |
+
"learning_rate": 5e-06,
|
| 2110 |
+
"loss": 0.6521,
|
| 2111 |
+
"step": 2980
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 2.999874545226446,
|
| 2115 |
+
"eval_loss": 0.7146658897399902,
|
| 2116 |
+
"eval_runtime": 674.3767,
|
| 2117 |
+
"eval_samples_per_second": 39.813,
|
| 2118 |
+
"eval_steps_per_second": 0.623,
|
| 2119 |
+
"step": 2989
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 3.000878183414879,
|
| 2123 |
+
"grad_norm": 1.159167907299322,
|
| 2124 |
+
"learning_rate": 5e-06,
|
| 2125 |
+
"loss": 0.6898,
|
| 2126 |
+
"step": 2990
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 3.0109145652992098,
|
| 2130 |
+
"grad_norm": 0.7818114221168871,
|
| 2131 |
+
"learning_rate": 5e-06,
|
| 2132 |
+
"loss": 0.6112,
|
| 2133 |
+
"step": 3000
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 3.0209509471835405,
|
| 2137 |
+
"grad_norm": 0.7106384774672879,
|
| 2138 |
+
"learning_rate": 5e-06,
|
| 2139 |
+
"loss": 0.6059,
|
| 2140 |
+
"step": 3010
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 3.030987329067871,
|
| 2144 |
+
"grad_norm": 0.640003891116449,
|
| 2145 |
+
"learning_rate": 5e-06,
|
| 2146 |
+
"loss": 0.6085,
|
| 2147 |
+
"step": 3020
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 3.0410237109522016,
|
| 2151 |
+
"grad_norm": 0.6066326657235696,
|
| 2152 |
+
"learning_rate": 5e-06,
|
| 2153 |
+
"loss": 0.6077,
|
| 2154 |
+
"step": 3030
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 3.0510600928365323,
|
| 2158 |
+
"grad_norm": 0.6071064742090702,
|
| 2159 |
+
"learning_rate": 5e-06,
|
| 2160 |
+
"loss": 0.6116,
|
| 2161 |
+
"step": 3040
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 3.061096474720863,
|
| 2165 |
+
"grad_norm": 0.693281501487692,
|
| 2166 |
+
"learning_rate": 5e-06,
|
| 2167 |
+
"loss": 0.6052,
|
| 2168 |
+
"step": 3050
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 3.071132856605194,
|
| 2172 |
+
"grad_norm": 0.7693025911684378,
|
| 2173 |
+
"learning_rate": 5e-06,
|
| 2174 |
+
"loss": 0.6095,
|
| 2175 |
+
"step": 3060
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 3.0811692384895246,
|
| 2179 |
+
"grad_norm": 0.6184427233437038,
|
| 2180 |
+
"learning_rate": 5e-06,
|
| 2181 |
+
"loss": 0.6069,
|
| 2182 |
+
"step": 3070
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 3.0912056203738554,
|
| 2186 |
+
"grad_norm": 0.634870226376911,
|
| 2187 |
+
"learning_rate": 5e-06,
|
| 2188 |
+
"loss": 0.6099,
|
| 2189 |
+
"step": 3080
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 3.101242002258186,
|
| 2193 |
+
"grad_norm": 0.5947920487741215,
|
| 2194 |
+
"learning_rate": 5e-06,
|
| 2195 |
+
"loss": 0.6092,
|
| 2196 |
+
"step": 3090
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 3.1112783841425165,
|
| 2200 |
+
"grad_norm": 0.5475506951891964,
|
| 2201 |
+
"learning_rate": 5e-06,
|
| 2202 |
+
"loss": 0.6066,
|
| 2203 |
+
"step": 3100
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 3.121314766026847,
|
| 2207 |
+
"grad_norm": 0.5786846254856872,
|
| 2208 |
+
"learning_rate": 5e-06,
|
| 2209 |
+
"loss": 0.6077,
|
| 2210 |
+
"step": 3110
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 3.131351147911178,
|
| 2214 |
+
"grad_norm": 0.5837921428316006,
|
| 2215 |
+
"learning_rate": 5e-06,
|
| 2216 |
+
"loss": 0.6125,
|
| 2217 |
+
"step": 3120
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 3.1413875297955087,
|
| 2221 |
+
"grad_norm": 0.6388660075166559,
|
| 2222 |
+
"learning_rate": 5e-06,
|
| 2223 |
+
"loss": 0.6073,
|
| 2224 |
+
"step": 3130
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 3.1514239116798395,
|
| 2228 |
+
"grad_norm": 0.6247319700614546,
|
| 2229 |
+
"learning_rate": 5e-06,
|
| 2230 |
+
"loss": 0.6129,
|
| 2231 |
+
"step": 3140
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 3.1614602935641702,
|
| 2235 |
+
"grad_norm": 0.7220969862146115,
|
| 2236 |
+
"learning_rate": 5e-06,
|
| 2237 |
+
"loss": 0.6096,
|
| 2238 |
+
"step": 3150
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 3.171496675448501,
|
| 2242 |
+
"grad_norm": 0.5966143252277392,
|
| 2243 |
+
"learning_rate": 5e-06,
|
| 2244 |
+
"loss": 0.6108,
|
| 2245 |
+
"step": 3160
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 3.1815330573328313,
|
| 2249 |
+
"grad_norm": 0.5806668148525886,
|
| 2250 |
+
"learning_rate": 5e-06,
|
| 2251 |
+
"loss": 0.611,
|
| 2252 |
+
"step": 3170
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 3.191569439217162,
|
| 2256 |
+
"grad_norm": 0.5847564401984537,
|
| 2257 |
+
"learning_rate": 5e-06,
|
| 2258 |
+
"loss": 0.6099,
|
| 2259 |
+
"step": 3180
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 3.201605821101493,
|
| 2263 |
+
"grad_norm": 0.5685073324759383,
|
| 2264 |
+
"learning_rate": 5e-06,
|
| 2265 |
+
"loss": 0.6162,
|
| 2266 |
+
"step": 3190
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 3.2116422029858236,
|
| 2270 |
+
"grad_norm": 0.5806892343391038,
|
| 2271 |
+
"learning_rate": 5e-06,
|
| 2272 |
+
"loss": 0.6099,
|
| 2273 |
+
"step": 3200
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 3.2216785848701543,
|
| 2277 |
+
"grad_norm": 0.5629335787755336,
|
| 2278 |
+
"learning_rate": 5e-06,
|
| 2279 |
+
"loss": 0.6122,
|
| 2280 |
+
"step": 3210
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 3.231714966754485,
|
| 2284 |
+
"grad_norm": 0.6104998235017857,
|
| 2285 |
+
"learning_rate": 5e-06,
|
| 2286 |
+
"loss": 0.6128,
|
| 2287 |
+
"step": 3220
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 3.241751348638816,
|
| 2291 |
+
"grad_norm": 0.670576007712542,
|
| 2292 |
+
"learning_rate": 5e-06,
|
| 2293 |
+
"loss": 0.6122,
|
| 2294 |
+
"step": 3230
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 3.251787730523146,
|
| 2298 |
+
"grad_norm": 0.6597487550561909,
|
| 2299 |
+
"learning_rate": 5e-06,
|
| 2300 |
+
"loss": 0.6135,
|
| 2301 |
+
"step": 3240
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 3.261824112407477,
|
| 2305 |
+
"grad_norm": 0.5645378989833628,
|
| 2306 |
+
"learning_rate": 5e-06,
|
| 2307 |
+
"loss": 0.6149,
|
| 2308 |
+
"step": 3250
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 3.2718604942918077,
|
| 2312 |
+
"grad_norm": 0.5939861646065504,
|
| 2313 |
+
"learning_rate": 5e-06,
|
| 2314 |
+
"loss": 0.6097,
|
| 2315 |
+
"step": 3260
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 3.2818968761761385,
|
| 2319 |
+
"grad_norm": 0.6160406690736504,
|
| 2320 |
+
"learning_rate": 5e-06,
|
| 2321 |
+
"loss": 0.6086,
|
| 2322 |
+
"step": 3270
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 3.291933258060469,
|
| 2326 |
+
"grad_norm": 0.5487825374465094,
|
| 2327 |
+
"learning_rate": 5e-06,
|
| 2328 |
+
"loss": 0.6144,
|
| 2329 |
+
"step": 3280
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 3.3019696399448,
|
| 2333 |
+
"grad_norm": 0.6520181865316601,
|
| 2334 |
+
"learning_rate": 5e-06,
|
| 2335 |
+
"loss": 0.6183,
|
| 2336 |
+
"step": 3290
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 3.3120060218291307,
|
| 2340 |
+
"grad_norm": 0.5977506986605584,
|
| 2341 |
+
"learning_rate": 5e-06,
|
| 2342 |
+
"loss": 0.6147,
|
| 2343 |
+
"step": 3300
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 3.3220424037134615,
|
| 2347 |
+
"grad_norm": 0.6484133892242163,
|
| 2348 |
+
"learning_rate": 5e-06,
|
| 2349 |
+
"loss": 0.6151,
|
| 2350 |
+
"step": 3310
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 3.332078785597792,
|
| 2354 |
+
"grad_norm": 0.5970543245993525,
|
| 2355 |
+
"learning_rate": 5e-06,
|
| 2356 |
+
"loss": 0.6108,
|
| 2357 |
+
"step": 3320
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 3.3421151674821226,
|
| 2361 |
+
"grad_norm": 0.6116862845869632,
|
| 2362 |
+
"learning_rate": 5e-06,
|
| 2363 |
+
"loss": 0.6166,
|
| 2364 |
+
"step": 3330
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 3.3521515493664533,
|
| 2368 |
+
"grad_norm": 0.5580458755366267,
|
| 2369 |
+
"learning_rate": 5e-06,
|
| 2370 |
+
"loss": 0.6177,
|
| 2371 |
+
"step": 3340
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 3.362187931250784,
|
| 2375 |
+
"grad_norm": 0.7040073547476862,
|
| 2376 |
+
"learning_rate": 5e-06,
|
| 2377 |
+
"loss": 0.6181,
|
| 2378 |
+
"step": 3350
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 3.372224313135115,
|
| 2382 |
+
"grad_norm": 0.652807816857214,
|
| 2383 |
+
"learning_rate": 5e-06,
|
| 2384 |
+
"loss": 0.6186,
|
| 2385 |
+
"step": 3360
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 3.3822606950194456,
|
| 2389 |
+
"grad_norm": 0.6106384378999347,
|
| 2390 |
+
"learning_rate": 5e-06,
|
| 2391 |
+
"loss": 0.6189,
|
| 2392 |
+
"step": 3370
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 3.3922970769037764,
|
| 2396 |
+
"grad_norm": 0.6920160744092827,
|
| 2397 |
+
"learning_rate": 5e-06,
|
| 2398 |
+
"loss": 0.617,
|
| 2399 |
+
"step": 3380
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 3.4023334587881067,
|
| 2403 |
+
"grad_norm": 0.5814853765567533,
|
| 2404 |
+
"learning_rate": 5e-06,
|
| 2405 |
+
"loss": 0.6118,
|
| 2406 |
+
"step": 3390
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 3.4123698406724374,
|
| 2410 |
+
"grad_norm": 0.5389887117603109,
|
| 2411 |
+
"learning_rate": 5e-06,
|
| 2412 |
+
"loss": 0.6126,
|
| 2413 |
+
"step": 3400
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 3.422406222556768,
|
| 2417 |
+
"grad_norm": 0.582316325799389,
|
| 2418 |
+
"learning_rate": 5e-06,
|
| 2419 |
+
"loss": 0.6143,
|
| 2420 |
+
"step": 3410
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 3.432442604441099,
|
| 2424 |
+
"grad_norm": 0.5612761289810537,
|
| 2425 |
+
"learning_rate": 5e-06,
|
| 2426 |
+
"loss": 0.616,
|
| 2427 |
+
"step": 3420
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 3.4424789863254297,
|
| 2431 |
+
"grad_norm": 0.5315307027152637,
|
| 2432 |
+
"learning_rate": 5e-06,
|
| 2433 |
+
"loss": 0.6126,
|
| 2434 |
+
"step": 3430
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 3.4525153682097605,
|
| 2438 |
+
"grad_norm": 0.5632095541748001,
|
| 2439 |
+
"learning_rate": 5e-06,
|
| 2440 |
+
"loss": 0.6139,
|
| 2441 |
+
"step": 3440
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 3.462551750094091,
|
| 2445 |
+
"grad_norm": 0.5777253428377956,
|
| 2446 |
+
"learning_rate": 5e-06,
|
| 2447 |
+
"loss": 0.6153,
|
| 2448 |
+
"step": 3450
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 3.472588131978422,
|
| 2452 |
+
"grad_norm": 0.5802066633079221,
|
| 2453 |
+
"learning_rate": 5e-06,
|
| 2454 |
+
"loss": 0.6183,
|
| 2455 |
+
"step": 3460
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 3.4826245138627523,
|
| 2459 |
+
"grad_norm": 0.6858510179050318,
|
| 2460 |
+
"learning_rate": 5e-06,
|
| 2461 |
+
"loss": 0.6105,
|
| 2462 |
+
"step": 3470
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 3.492660895747083,
|
| 2466 |
+
"grad_norm": 0.6150923461042579,
|
| 2467 |
+
"learning_rate": 5e-06,
|
| 2468 |
+
"loss": 0.6123,
|
| 2469 |
+
"step": 3480
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 3.502697277631414,
|
| 2473 |
+
"grad_norm": 0.6850358509214438,
|
| 2474 |
+
"learning_rate": 5e-06,
|
| 2475 |
+
"loss": 0.6176,
|
| 2476 |
+
"step": 3490
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 3.5127336595157446,
|
| 2480 |
+
"grad_norm": 0.6221194287714066,
|
| 2481 |
+
"learning_rate": 5e-06,
|
| 2482 |
+
"loss": 0.6135,
|
| 2483 |
+
"step": 3500
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 3.5227700414000753,
|
| 2487 |
+
"grad_norm": 0.6337555357747637,
|
| 2488 |
+
"learning_rate": 5e-06,
|
| 2489 |
+
"loss": 0.6176,
|
| 2490 |
+
"step": 3510
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 3.532806423284406,
|
| 2494 |
+
"grad_norm": 0.5696342404252304,
|
| 2495 |
+
"learning_rate": 5e-06,
|
| 2496 |
+
"loss": 0.6194,
|
| 2497 |
+
"step": 3520
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 3.5428428051687364,
|
| 2501 |
+
"grad_norm": 0.5192096724412292,
|
| 2502 |
+
"learning_rate": 5e-06,
|
| 2503 |
+
"loss": 0.6169,
|
| 2504 |
+
"step": 3530
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 3.552879187053067,
|
| 2508 |
+
"grad_norm": 0.6461636488212382,
|
| 2509 |
+
"learning_rate": 5e-06,
|
| 2510 |
+
"loss": 0.6194,
|
| 2511 |
+
"step": 3540
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 3.562915568937398,
|
| 2515 |
+
"grad_norm": 0.5204792269879596,
|
| 2516 |
+
"learning_rate": 5e-06,
|
| 2517 |
+
"loss": 0.6134,
|
| 2518 |
+
"step": 3550
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 3.5729519508217287,
|
| 2522 |
+
"grad_norm": 0.5799708864875738,
|
| 2523 |
+
"learning_rate": 5e-06,
|
| 2524 |
+
"loss": 0.6188,
|
| 2525 |
+
"step": 3560
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 3.5829883327060594,
|
| 2529 |
+
"grad_norm": 0.5463250823773549,
|
| 2530 |
+
"learning_rate": 5e-06,
|
| 2531 |
+
"loss": 0.6176,
|
| 2532 |
+
"step": 3570
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 3.59302471459039,
|
| 2536 |
+
"grad_norm": 0.6314712032266755,
|
| 2537 |
+
"learning_rate": 5e-06,
|
| 2538 |
+
"loss": 0.6207,
|
| 2539 |
+
"step": 3580
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 3.603061096474721,
|
| 2543 |
+
"grad_norm": 0.6479660409480549,
|
| 2544 |
+
"learning_rate": 5e-06,
|
| 2545 |
+
"loss": 0.6164,
|
| 2546 |
+
"step": 3590
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 3.6130974783590517,
|
| 2550 |
+
"grad_norm": 0.6300826657668005,
|
| 2551 |
+
"learning_rate": 5e-06,
|
| 2552 |
+
"loss": 0.6161,
|
| 2553 |
+
"step": 3600
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 3.6231338602433825,
|
| 2557 |
+
"grad_norm": 0.6595382686169196,
|
| 2558 |
+
"learning_rate": 5e-06,
|
| 2559 |
+
"loss": 0.6206,
|
| 2560 |
+
"step": 3610
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 3.6331702421277132,
|
| 2564 |
+
"grad_norm": 0.5652565303110992,
|
| 2565 |
+
"learning_rate": 5e-06,
|
| 2566 |
+
"loss": 0.6196,
|
| 2567 |
+
"step": 3620
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 3.6432066240120435,
|
| 2571 |
+
"grad_norm": 0.579203582642288,
|
| 2572 |
+
"learning_rate": 5e-06,
|
| 2573 |
+
"loss": 0.6204,
|
| 2574 |
+
"step": 3630
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 3.6532430058963743,
|
| 2578 |
+
"grad_norm": 0.5437849615714947,
|
| 2579 |
+
"learning_rate": 5e-06,
|
| 2580 |
+
"loss": 0.6183,
|
| 2581 |
+
"step": 3640
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 3.663279387780705,
|
| 2585 |
+
"grad_norm": 0.5787593937881179,
|
| 2586 |
+
"learning_rate": 5e-06,
|
| 2587 |
+
"loss": 0.6167,
|
| 2588 |
+
"step": 3650
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 3.673315769665036,
|
| 2592 |
+
"grad_norm": 0.5811897003512283,
|
| 2593 |
+
"learning_rate": 5e-06,
|
| 2594 |
+
"loss": 0.6152,
|
| 2595 |
+
"step": 3660
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 3.6833521515493666,
|
| 2599 |
+
"grad_norm": 0.5178334072146914,
|
| 2600 |
+
"learning_rate": 5e-06,
|
| 2601 |
+
"loss": 0.612,
|
| 2602 |
+
"step": 3670
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 3.693388533433697,
|
| 2606 |
+
"grad_norm": 0.6269969593175169,
|
| 2607 |
+
"learning_rate": 5e-06,
|
| 2608 |
+
"loss": 0.616,
|
| 2609 |
+
"step": 3680
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 3.7034249153180276,
|
| 2613 |
+
"grad_norm": 0.6398231596326978,
|
| 2614 |
+
"learning_rate": 5e-06,
|
| 2615 |
+
"loss": 0.62,
|
| 2616 |
+
"step": 3690
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 3.7134612972023584,
|
| 2620 |
+
"grad_norm": 0.5787553408157999,
|
| 2621 |
+
"learning_rate": 5e-06,
|
| 2622 |
+
"loss": 0.6199,
|
| 2623 |
+
"step": 3700
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 3.723497679086689,
|
| 2627 |
+
"grad_norm": 0.5592510334692263,
|
| 2628 |
+
"learning_rate": 5e-06,
|
| 2629 |
+
"loss": 0.6172,
|
| 2630 |
+
"step": 3710
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 3.73353406097102,
|
| 2634 |
+
"grad_norm": 0.5911520225126639,
|
| 2635 |
+
"learning_rate": 5e-06,
|
| 2636 |
+
"loss": 0.6206,
|
| 2637 |
+
"step": 3720
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 3.7435704428553507,
|
| 2641 |
+
"grad_norm": 0.5509577144961972,
|
| 2642 |
+
"learning_rate": 5e-06,
|
| 2643 |
+
"loss": 0.6162,
|
| 2644 |
+
"step": 3730
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 3.7536068247396814,
|
| 2648 |
+
"grad_norm": 0.6230171740270747,
|
| 2649 |
+
"learning_rate": 5e-06,
|
| 2650 |
+
"loss": 0.6211,
|
| 2651 |
+
"step": 3740
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 3.763643206624012,
|
| 2655 |
+
"grad_norm": 0.606064062911815,
|
| 2656 |
+
"learning_rate": 5e-06,
|
| 2657 |
+
"loss": 0.6169,
|
| 2658 |
+
"step": 3750
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 3.773679588508343,
|
| 2662 |
+
"grad_norm": 0.5251378729385787,
|
| 2663 |
+
"learning_rate": 5e-06,
|
| 2664 |
+
"loss": 0.6142,
|
| 2665 |
+
"step": 3760
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 3.7837159703926737,
|
| 2669 |
+
"grad_norm": 0.5313902122829314,
|
| 2670 |
+
"learning_rate": 5e-06,
|
| 2671 |
+
"loss": 0.6171,
|
| 2672 |
+
"step": 3770
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 3.793752352277004,
|
| 2676 |
+
"grad_norm": 0.5738855269066635,
|
| 2677 |
+
"learning_rate": 5e-06,
|
| 2678 |
+
"loss": 0.614,
|
| 2679 |
+
"step": 3780
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 3.803788734161335,
|
| 2683 |
+
"grad_norm": 0.6094017649137001,
|
| 2684 |
+
"learning_rate": 5e-06,
|
| 2685 |
+
"loss": 0.6184,
|
| 2686 |
+
"step": 3790
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 3.8138251160456655,
|
| 2690 |
+
"grad_norm": 0.577775104243207,
|
| 2691 |
+
"learning_rate": 5e-06,
|
| 2692 |
+
"loss": 0.6147,
|
| 2693 |
+
"step": 3800
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 3.8238614979299963,
|
| 2697 |
+
"grad_norm": 0.5896473350091869,
|
| 2698 |
+
"learning_rate": 5e-06,
|
| 2699 |
+
"loss": 0.6191,
|
| 2700 |
+
"step": 3810
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 3.833897879814327,
|
| 2704 |
+
"grad_norm": 0.6197381864983481,
|
| 2705 |
+
"learning_rate": 5e-06,
|
| 2706 |
+
"loss": 0.6204,
|
| 2707 |
+
"step": 3820
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 3.8439342616986574,
|
| 2711 |
+
"grad_norm": 0.6385235404417998,
|
| 2712 |
+
"learning_rate": 5e-06,
|
| 2713 |
+
"loss": 0.6184,
|
| 2714 |
+
"step": 3830
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 3.853970643582988,
|
| 2718 |
+
"grad_norm": 0.5605836986977404,
|
| 2719 |
+
"learning_rate": 5e-06,
|
| 2720 |
+
"loss": 0.6184,
|
| 2721 |
+
"step": 3840
|
| 2722 |
+
},
|
| 2723 |
+
{
|
| 2724 |
+
"epoch": 3.864007025467319,
|
| 2725 |
+
"grad_norm": 0.5314662217673221,
|
| 2726 |
+
"learning_rate": 5e-06,
|
| 2727 |
+
"loss": 0.6175,
|
| 2728 |
+
"step": 3850
|
| 2729 |
+
},
|
| 2730 |
+
{
|
| 2731 |
+
"epoch": 3.8740434073516496,
|
| 2732 |
+
"grad_norm": 0.5617746957894754,
|
| 2733 |
+
"learning_rate": 5e-06,
|
| 2734 |
+
"loss": 0.6168,
|
| 2735 |
+
"step": 3860
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 3.8840797892359804,
|
| 2739 |
+
"grad_norm": 0.6130682994037887,
|
| 2740 |
+
"learning_rate": 5e-06,
|
| 2741 |
+
"loss": 0.6213,
|
| 2742 |
+
"step": 3870
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 3.894116171120311,
|
| 2746 |
+
"grad_norm": 0.6330660373549564,
|
| 2747 |
+
"learning_rate": 5e-06,
|
| 2748 |
+
"loss": 0.6237,
|
| 2749 |
+
"step": 3880
|
| 2750 |
+
},
|
| 2751 |
+
{
|
| 2752 |
+
"epoch": 3.904152553004642,
|
| 2753 |
+
"grad_norm": 0.5757814490358608,
|
| 2754 |
+
"learning_rate": 5e-06,
|
| 2755 |
+
"loss": 0.6199,
|
| 2756 |
+
"step": 3890
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 3.9141889348889727,
|
| 2760 |
+
"grad_norm": 0.5686860241059948,
|
| 2761 |
+
"learning_rate": 5e-06,
|
| 2762 |
+
"loss": 0.6166,
|
| 2763 |
+
"step": 3900
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 3.9242253167733034,
|
| 2767 |
+
"grad_norm": 0.577591496190582,
|
| 2768 |
+
"learning_rate": 5e-06,
|
| 2769 |
+
"loss": 0.6194,
|
| 2770 |
+
"step": 3910
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 3.934261698657634,
|
| 2774 |
+
"grad_norm": 0.5589041470451204,
|
| 2775 |
+
"learning_rate": 5e-06,
|
| 2776 |
+
"loss": 0.619,
|
| 2777 |
+
"step": 3920
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 3.9442980805419645,
|
| 2781 |
+
"grad_norm": 0.6004802459840047,
|
| 2782 |
+
"learning_rate": 5e-06,
|
| 2783 |
+
"loss": 0.6148,
|
| 2784 |
+
"step": 3930
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 3.9543344624262953,
|
| 2788 |
+
"grad_norm": 0.6153349703188992,
|
| 2789 |
+
"learning_rate": 5e-06,
|
| 2790 |
+
"loss": 0.6166,
|
| 2791 |
+
"step": 3940
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"epoch": 3.964370844310626,
|
| 2795 |
+
"grad_norm": 0.5218178205884076,
|
| 2796 |
+
"learning_rate": 5e-06,
|
| 2797 |
+
"loss": 0.619,
|
| 2798 |
+
"step": 3950
|
| 2799 |
+
},
|
| 2800 |
+
{
|
| 2801 |
+
"epoch": 3.974407226194957,
|
| 2802 |
+
"grad_norm": 0.547168472398349,
|
| 2803 |
+
"learning_rate": 5e-06,
|
| 2804 |
+
"loss": 0.6194,
|
| 2805 |
+
"step": 3960
|
| 2806 |
+
},
|
| 2807 |
+
{
|
| 2808 |
+
"epoch": 3.9844436080792875,
|
| 2809 |
+
"grad_norm": 0.5359836200059497,
|
| 2810 |
+
"learning_rate": 5e-06,
|
| 2811 |
+
"loss": 0.6225,
|
| 2812 |
+
"step": 3970
|
| 2813 |
+
},
|
| 2814 |
+
{
|
| 2815 |
+
"epoch": 3.994479989963618,
|
| 2816 |
+
"grad_norm": 0.7180200697231374,
|
| 2817 |
+
"learning_rate": 5e-06,
|
| 2818 |
+
"loss": 0.6204,
|
| 2819 |
+
"step": 3980
|
| 2820 |
+
},
|
| 2821 |
+
{
|
| 2822 |
+
"epoch": 3.99849454271735,
|
| 2823 |
+
"eval_loss": 0.7222821116447449,
|
| 2824 |
+
"eval_runtime": 679.219,
|
| 2825 |
+
"eval_samples_per_second": 39.529,
|
| 2826 |
+
"eval_steps_per_second": 0.618,
|
| 2827 |
+
"step": 3984
|
| 2828 |
+
},
|
| 2829 |
+
{
|
| 2830 |
+
"epoch": 3.99849454271735,
|
| 2831 |
+
"step": 3984,
|
| 2832 |
+
"total_flos": 6673139006177280.0,
|
| 2833 |
+
"train_loss": 0.6817261522194468,
|
| 2834 |
+
"train_runtime": 130344.1738,
|
| 2835 |
+
"train_samples_per_second": 15.654,
|
| 2836 |
+
"train_steps_per_second": 0.031
|
| 2837 |
+
}
|
| 2838 |
+
],
|
| 2839 |
+
"logging_steps": 10,
|
| 2840 |
+
"max_steps": 3984,
|
| 2841 |
+
"num_input_tokens_seen": 0,
|
| 2842 |
+
"num_train_epochs": 4,
|
| 2843 |
+
"save_steps": 500,
|
| 2844 |
+
"stateful_callbacks": {
|
| 2845 |
+
"TrainerControl": {
|
| 2846 |
+
"args": {
|
| 2847 |
+
"should_epoch_stop": false,
|
| 2848 |
+
"should_evaluate": false,
|
| 2849 |
+
"should_log": false,
|
| 2850 |
+
"should_save": true,
|
| 2851 |
+
"should_training_stop": true
|
| 2852 |
+
},
|
| 2853 |
+
"attributes": {}
|
| 2854 |
+
}
|
| 2855 |
+
},
|
| 2856 |
+
"total_flos": 6673139006177280.0,
|
| 2857 |
+
"train_batch_size": 8,
|
| 2858 |
+
"trial_name": null,
|
| 2859 |
+
"trial_params": null
|
| 2860 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|