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 +2880 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: oh-dcft-v3.1-SN-405B-hacky-qwen
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# oh-dcft-v3.1-SN-405B-hacky-qwen
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.5737
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: oh-dcft-v3.1-SN-405B-hacky-qwen
|
|
|
|
| 16 |
|
| 17 |
# oh-dcft-v3.1-SN-405B-hacky-qwen
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) on the mlfoundations-dev/oh-dcft-v3.1-SN-405B-hacky dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.5737
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 0.5737335681915283,
|
| 4 |
+
"eval_runtime": 339.9139,
|
| 5 |
+
"eval_samples_per_second": 26.571,
|
| 6 |
+
"eval_steps_per_second": 0.418,
|
| 7 |
+
"total_flos": 2108813507297280.0,
|
| 8 |
+
"train_loss": 0.4856802497430432,
|
| 9 |
+
"train_runtime": 54595.2546,
|
| 10 |
+
"train_samples_per_second": 9.429,
|
| 11 |
+
"train_steps_per_second": 0.074
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 0.5737335681915283,
|
| 4 |
+
"eval_runtime": 339.9139,
|
| 5 |
+
"eval_samples_per_second": 26.571,
|
| 6 |
+
"eval_steps_per_second": 0.418
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 2108813507297280.0,
|
| 4 |
+
"train_loss": 0.4856802497430432,
|
| 5 |
+
"train_runtime": 54595.2546,
|
| 6 |
+
"train_samples_per_second": 9.429,
|
| 7 |
+
"train_steps_per_second": 0.074
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2880 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4023,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.007457121551081283,
|
| 13 |
+
"grad_norm": 1.4923322722027186,
|
| 14 |
+
"learning_rate": 5e-06,
|
| 15 |
+
"loss": 0.6314,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.014914243102162566,
|
| 20 |
+
"grad_norm": 1.4277498493155112,
|
| 21 |
+
"learning_rate": 5e-06,
|
| 22 |
+
"loss": 0.5957,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02237136465324385,
|
| 27 |
+
"grad_norm": 1.3160638879347244,
|
| 28 |
+
"learning_rate": 5e-06,
|
| 29 |
+
"loss": 0.5858,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.02982848620432513,
|
| 34 |
+
"grad_norm": 1.3286844636967108,
|
| 35 |
+
"learning_rate": 5e-06,
|
| 36 |
+
"loss": 0.6083,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.037285607755406416,
|
| 41 |
+
"grad_norm": 1.3457945081617781,
|
| 42 |
+
"learning_rate": 5e-06,
|
| 43 |
+
"loss": 0.5833,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0447427293064877,
|
| 48 |
+
"grad_norm": 1.1775640475079898,
|
| 49 |
+
"learning_rate": 5e-06,
|
| 50 |
+
"loss": 0.5615,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05219985085756898,
|
| 55 |
+
"grad_norm": 0.7851058774790824,
|
| 56 |
+
"learning_rate": 5e-06,
|
| 57 |
+
"loss": 0.5599,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.05965697240865026,
|
| 62 |
+
"grad_norm": 0.6748980341595908,
|
| 63 |
+
"learning_rate": 5e-06,
|
| 64 |
+
"loss": 0.5738,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06711409395973154,
|
| 69 |
+
"grad_norm": 0.6098300182337207,
|
| 70 |
+
"learning_rate": 5e-06,
|
| 71 |
+
"loss": 0.5513,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.07457121551081283,
|
| 76 |
+
"grad_norm": 0.6353294403272493,
|
| 77 |
+
"learning_rate": 5e-06,
|
| 78 |
+
"loss": 0.5723,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08202833706189411,
|
| 83 |
+
"grad_norm": 0.5674389209790128,
|
| 84 |
+
"learning_rate": 5e-06,
|
| 85 |
+
"loss": 0.5666,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0894854586129754,
|
| 90 |
+
"grad_norm": 0.6279447283125179,
|
| 91 |
+
"learning_rate": 5e-06,
|
| 92 |
+
"loss": 0.5475,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09694258016405667,
|
| 97 |
+
"grad_norm": 0.576962331333649,
|
| 98 |
+
"learning_rate": 5e-06,
|
| 99 |
+
"loss": 0.5532,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.10439970171513796,
|
| 104 |
+
"grad_norm": 0.5745640803688579,
|
| 105 |
+
"learning_rate": 5e-06,
|
| 106 |
+
"loss": 0.5561,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.11185682326621924,
|
| 111 |
+
"grad_norm": 0.568035600435591,
|
| 112 |
+
"learning_rate": 5e-06,
|
| 113 |
+
"loss": 0.5542,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.11931394481730052,
|
| 118 |
+
"grad_norm": 0.5903955926152153,
|
| 119 |
+
"learning_rate": 5e-06,
|
| 120 |
+
"loss": 0.55,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1267710663683818,
|
| 125 |
+
"grad_norm": 0.6272192755787677,
|
| 126 |
+
"learning_rate": 5e-06,
|
| 127 |
+
"loss": 0.5577,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1342281879194631,
|
| 132 |
+
"grad_norm": 0.621732319302288,
|
| 133 |
+
"learning_rate": 5e-06,
|
| 134 |
+
"loss": 0.5579,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.14168530947054436,
|
| 139 |
+
"grad_norm": 0.582261312726868,
|
| 140 |
+
"learning_rate": 5e-06,
|
| 141 |
+
"loss": 0.5691,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.14914243102162567,
|
| 146 |
+
"grad_norm": 0.6038562541450465,
|
| 147 |
+
"learning_rate": 5e-06,
|
| 148 |
+
"loss": 0.5441,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.15659955257270694,
|
| 153 |
+
"grad_norm": 0.6342330659293374,
|
| 154 |
+
"learning_rate": 5e-06,
|
| 155 |
+
"loss": 0.5438,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.16405667412378822,
|
| 160 |
+
"grad_norm": 0.5605346796698036,
|
| 161 |
+
"learning_rate": 5e-06,
|
| 162 |
+
"loss": 0.5552,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.1715137956748695,
|
| 167 |
+
"grad_norm": 0.5998445612806728,
|
| 168 |
+
"learning_rate": 5e-06,
|
| 169 |
+
"loss": 0.57,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.1789709172259508,
|
| 174 |
+
"grad_norm": 0.5714415956910529,
|
| 175 |
+
"learning_rate": 5e-06,
|
| 176 |
+
"loss": 0.5588,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.18642803877703207,
|
| 181 |
+
"grad_norm": 0.6368975283895788,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 0.5435,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.19388516032811334,
|
| 188 |
+
"grad_norm": 0.5912645424796484,
|
| 189 |
+
"learning_rate": 5e-06,
|
| 190 |
+
"loss": 0.5371,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.20134228187919462,
|
| 195 |
+
"grad_norm": 0.5671791863380606,
|
| 196 |
+
"learning_rate": 5e-06,
|
| 197 |
+
"loss": 0.554,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.20879940343027592,
|
| 202 |
+
"grad_norm": 0.5825115142497552,
|
| 203 |
+
"learning_rate": 5e-06,
|
| 204 |
+
"loss": 0.5648,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2162565249813572,
|
| 209 |
+
"grad_norm": 0.6148559353437262,
|
| 210 |
+
"learning_rate": 5e-06,
|
| 211 |
+
"loss": 0.5626,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.22371364653243847,
|
| 216 |
+
"grad_norm": 0.5992171256664753,
|
| 217 |
+
"learning_rate": 5e-06,
|
| 218 |
+
"loss": 0.5495,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.23117076808351977,
|
| 223 |
+
"grad_norm": 0.5756923092743548,
|
| 224 |
+
"learning_rate": 5e-06,
|
| 225 |
+
"loss": 0.5492,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.23862788963460105,
|
| 230 |
+
"grad_norm": 0.5571239833558982,
|
| 231 |
+
"learning_rate": 5e-06,
|
| 232 |
+
"loss": 0.5382,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.24608501118568232,
|
| 237 |
+
"grad_norm": 0.6142758618521919,
|
| 238 |
+
"learning_rate": 5e-06,
|
| 239 |
+
"loss": 0.5626,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2535421327367636,
|
| 244 |
+
"grad_norm": 0.6071098554908272,
|
| 245 |
+
"learning_rate": 5e-06,
|
| 246 |
+
"loss": 0.5537,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.2609992542878449,
|
| 251 |
+
"grad_norm": 0.5503873105742949,
|
| 252 |
+
"learning_rate": 5e-06,
|
| 253 |
+
"loss": 0.5577,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.2684563758389262,
|
| 258 |
+
"grad_norm": 0.5837181817071831,
|
| 259 |
+
"learning_rate": 5e-06,
|
| 260 |
+
"loss": 0.5514,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.2759134973900075,
|
| 265 |
+
"grad_norm": 0.6207096163948669,
|
| 266 |
+
"learning_rate": 5e-06,
|
| 267 |
+
"loss": 0.5415,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2833706189410887,
|
| 272 |
+
"grad_norm": 0.5430609166419872,
|
| 273 |
+
"learning_rate": 5e-06,
|
| 274 |
+
"loss": 0.5641,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.29082774049217003,
|
| 279 |
+
"grad_norm": 0.6540545927611947,
|
| 280 |
+
"learning_rate": 5e-06,
|
| 281 |
+
"loss": 0.5481,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.29828486204325133,
|
| 286 |
+
"grad_norm": 0.5461956945004067,
|
| 287 |
+
"learning_rate": 5e-06,
|
| 288 |
+
"loss": 0.5513,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.3057419835943326,
|
| 293 |
+
"grad_norm": 0.6695239493079291,
|
| 294 |
+
"learning_rate": 5e-06,
|
| 295 |
+
"loss": 0.5638,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.3131991051454139,
|
| 300 |
+
"grad_norm": 0.586611622356162,
|
| 301 |
+
"learning_rate": 5e-06,
|
| 302 |
+
"loss": 0.5492,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.32065622669649513,
|
| 307 |
+
"grad_norm": 0.610109513110288,
|
| 308 |
+
"learning_rate": 5e-06,
|
| 309 |
+
"loss": 0.5607,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.32811334824757643,
|
| 314 |
+
"grad_norm": 0.6127769350008854,
|
| 315 |
+
"learning_rate": 5e-06,
|
| 316 |
+
"loss": 0.5645,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.33557046979865773,
|
| 321 |
+
"grad_norm": 0.6680654140600404,
|
| 322 |
+
"learning_rate": 5e-06,
|
| 323 |
+
"loss": 0.5549,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.343027591349739,
|
| 328 |
+
"grad_norm": 0.5826310904520773,
|
| 329 |
+
"learning_rate": 5e-06,
|
| 330 |
+
"loss": 0.5548,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.3504847129008203,
|
| 335 |
+
"grad_norm": 0.6239330754324169,
|
| 336 |
+
"learning_rate": 5e-06,
|
| 337 |
+
"loss": 0.5583,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.3579418344519016,
|
| 342 |
+
"grad_norm": 0.5877553356456545,
|
| 343 |
+
"learning_rate": 5e-06,
|
| 344 |
+
"loss": 0.545,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.36539895600298283,
|
| 349 |
+
"grad_norm": 0.5829722537190941,
|
| 350 |
+
"learning_rate": 5e-06,
|
| 351 |
+
"loss": 0.5395,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.37285607755406414,
|
| 356 |
+
"grad_norm": 0.6621755183448562,
|
| 357 |
+
"learning_rate": 5e-06,
|
| 358 |
+
"loss": 0.5456,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.38031319910514544,
|
| 363 |
+
"grad_norm": 0.5591184552346828,
|
| 364 |
+
"learning_rate": 5e-06,
|
| 365 |
+
"loss": 0.5367,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.3877703206562267,
|
| 370 |
+
"grad_norm": 0.565634611705387,
|
| 371 |
+
"learning_rate": 5e-06,
|
| 372 |
+
"loss": 0.5716,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.395227442207308,
|
| 377 |
+
"grad_norm": 0.5782734602622104,
|
| 378 |
+
"learning_rate": 5e-06,
|
| 379 |
+
"loss": 0.5545,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.40268456375838924,
|
| 384 |
+
"grad_norm": 0.5695093375447697,
|
| 385 |
+
"learning_rate": 5e-06,
|
| 386 |
+
"loss": 0.5646,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.41014168530947054,
|
| 391 |
+
"grad_norm": 0.6396895709952651,
|
| 392 |
+
"learning_rate": 5e-06,
|
| 393 |
+
"loss": 0.559,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.41759880686055184,
|
| 398 |
+
"grad_norm": 0.5626775345590961,
|
| 399 |
+
"learning_rate": 5e-06,
|
| 400 |
+
"loss": 0.5398,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4250559284116331,
|
| 405 |
+
"grad_norm": 0.6561792759757238,
|
| 406 |
+
"learning_rate": 5e-06,
|
| 407 |
+
"loss": 0.5506,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.4325130499627144,
|
| 412 |
+
"grad_norm": 0.6503627380598381,
|
| 413 |
+
"learning_rate": 5e-06,
|
| 414 |
+
"loss": 0.5624,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.4399701715137957,
|
| 419 |
+
"grad_norm": 0.5712768208534026,
|
| 420 |
+
"learning_rate": 5e-06,
|
| 421 |
+
"loss": 0.5493,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.44742729306487694,
|
| 426 |
+
"grad_norm": 0.632567133823459,
|
| 427 |
+
"learning_rate": 5e-06,
|
| 428 |
+
"loss": 0.5495,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.45488441461595824,
|
| 433 |
+
"grad_norm": 0.5891856850275913,
|
| 434 |
+
"learning_rate": 5e-06,
|
| 435 |
+
"loss": 0.5422,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.46234153616703955,
|
| 440 |
+
"grad_norm": 0.5745761726721252,
|
| 441 |
+
"learning_rate": 5e-06,
|
| 442 |
+
"loss": 0.5575,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.4697986577181208,
|
| 447 |
+
"grad_norm": 0.6304716115534362,
|
| 448 |
+
"learning_rate": 5e-06,
|
| 449 |
+
"loss": 0.552,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.4772557792692021,
|
| 454 |
+
"grad_norm": 0.5804599760466073,
|
| 455 |
+
"learning_rate": 5e-06,
|
| 456 |
+
"loss": 0.5574,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.48471290082028334,
|
| 461 |
+
"grad_norm": 0.5920559180390651,
|
| 462 |
+
"learning_rate": 5e-06,
|
| 463 |
+
"loss": 0.5515,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.49217002237136465,
|
| 468 |
+
"grad_norm": 0.5791705205755153,
|
| 469 |
+
"learning_rate": 5e-06,
|
| 470 |
+
"loss": 0.5574,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.49962714392244595,
|
| 475 |
+
"grad_norm": 0.6201405707302831,
|
| 476 |
+
"learning_rate": 5e-06,
|
| 477 |
+
"loss": 0.5585,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5070842654735273,
|
| 482 |
+
"grad_norm": 0.5753619878991011,
|
| 483 |
+
"learning_rate": 5e-06,
|
| 484 |
+
"loss": 0.556,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.5145413870246085,
|
| 489 |
+
"grad_norm": 0.5594987960581536,
|
| 490 |
+
"learning_rate": 5e-06,
|
| 491 |
+
"loss": 0.5437,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.5219985085756897,
|
| 496 |
+
"grad_norm": 0.5703994092235947,
|
| 497 |
+
"learning_rate": 5e-06,
|
| 498 |
+
"loss": 0.5568,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.5294556301267711,
|
| 503 |
+
"grad_norm": 0.5834034437186236,
|
| 504 |
+
"learning_rate": 5e-06,
|
| 505 |
+
"loss": 0.5431,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.5369127516778524,
|
| 510 |
+
"grad_norm": 0.560106669559214,
|
| 511 |
+
"learning_rate": 5e-06,
|
| 512 |
+
"loss": 0.5552,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.5443698732289336,
|
| 517 |
+
"grad_norm": 0.5679786278378056,
|
| 518 |
+
"learning_rate": 5e-06,
|
| 519 |
+
"loss": 0.554,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.551826994780015,
|
| 524 |
+
"grad_norm": 0.6429988192429542,
|
| 525 |
+
"learning_rate": 5e-06,
|
| 526 |
+
"loss": 0.5495,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.5592841163310962,
|
| 531 |
+
"grad_norm": 0.6267938520430484,
|
| 532 |
+
"learning_rate": 5e-06,
|
| 533 |
+
"loss": 0.5561,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5667412378821775,
|
| 538 |
+
"grad_norm": 0.6138480606756376,
|
| 539 |
+
"learning_rate": 5e-06,
|
| 540 |
+
"loss": 0.5532,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.5741983594332588,
|
| 545 |
+
"grad_norm": 0.6398337341244397,
|
| 546 |
+
"learning_rate": 5e-06,
|
| 547 |
+
"loss": 0.5567,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5816554809843401,
|
| 552 |
+
"grad_norm": 0.6201612138500541,
|
| 553 |
+
"learning_rate": 5e-06,
|
| 554 |
+
"loss": 0.5411,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5891126025354213,
|
| 559 |
+
"grad_norm": 0.5777153613729241,
|
| 560 |
+
"learning_rate": 5e-06,
|
| 561 |
+
"loss": 0.5598,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.5965697240865027,
|
| 566 |
+
"grad_norm": 0.5831925884746647,
|
| 567 |
+
"learning_rate": 5e-06,
|
| 568 |
+
"loss": 0.5505,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.6040268456375839,
|
| 573 |
+
"grad_norm": 0.5811154005536924,
|
| 574 |
+
"learning_rate": 5e-06,
|
| 575 |
+
"loss": 0.5301,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.6114839671886652,
|
| 580 |
+
"grad_norm": 0.5578916945797876,
|
| 581 |
+
"learning_rate": 5e-06,
|
| 582 |
+
"loss": 0.5616,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.6189410887397464,
|
| 587 |
+
"grad_norm": 0.6251996052710133,
|
| 588 |
+
"learning_rate": 5e-06,
|
| 589 |
+
"loss": 0.5444,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.6263982102908278,
|
| 594 |
+
"grad_norm": 0.5440172219846166,
|
| 595 |
+
"learning_rate": 5e-06,
|
| 596 |
+
"loss": 0.5596,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.633855331841909,
|
| 601 |
+
"grad_norm": 0.5904453872308907,
|
| 602 |
+
"learning_rate": 5e-06,
|
| 603 |
+
"loss": 0.5469,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.6413124533929903,
|
| 608 |
+
"grad_norm": 0.5659960741097526,
|
| 609 |
+
"learning_rate": 5e-06,
|
| 610 |
+
"loss": 0.562,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.6487695749440716,
|
| 615 |
+
"grad_norm": 0.5836414884442678,
|
| 616 |
+
"learning_rate": 5e-06,
|
| 617 |
+
"loss": 0.5388,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.6562266964951529,
|
| 622 |
+
"grad_norm": 0.6185085505610105,
|
| 623 |
+
"learning_rate": 5e-06,
|
| 624 |
+
"loss": 0.5406,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.6636838180462341,
|
| 629 |
+
"grad_norm": 0.5986837831143031,
|
| 630 |
+
"learning_rate": 5e-06,
|
| 631 |
+
"loss": 0.5549,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.6711409395973155,
|
| 636 |
+
"grad_norm": 0.5921990630549645,
|
| 637 |
+
"learning_rate": 5e-06,
|
| 638 |
+
"loss": 0.5291,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.6785980611483967,
|
| 643 |
+
"grad_norm": 0.5802544521725337,
|
| 644 |
+
"learning_rate": 5e-06,
|
| 645 |
+
"loss": 0.5523,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.686055182699478,
|
| 650 |
+
"grad_norm": 0.5851196561131752,
|
| 651 |
+
"learning_rate": 5e-06,
|
| 652 |
+
"loss": 0.5448,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.6935123042505593,
|
| 657 |
+
"grad_norm": 0.583877657890954,
|
| 658 |
+
"learning_rate": 5e-06,
|
| 659 |
+
"loss": 0.5415,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.7009694258016406,
|
| 664 |
+
"grad_norm": 0.5598625156605288,
|
| 665 |
+
"learning_rate": 5e-06,
|
| 666 |
+
"loss": 0.5712,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.7084265473527218,
|
| 671 |
+
"grad_norm": 0.6934347616736215,
|
| 672 |
+
"learning_rate": 5e-06,
|
| 673 |
+
"loss": 0.5519,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.7158836689038032,
|
| 678 |
+
"grad_norm": 0.5977390121445122,
|
| 679 |
+
"learning_rate": 5e-06,
|
| 680 |
+
"loss": 0.5551,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.7233407904548844,
|
| 685 |
+
"grad_norm": 0.5908530451006745,
|
| 686 |
+
"learning_rate": 5e-06,
|
| 687 |
+
"loss": 0.5561,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.7307979120059657,
|
| 692 |
+
"grad_norm": 0.5657057055912034,
|
| 693 |
+
"learning_rate": 5e-06,
|
| 694 |
+
"loss": 0.553,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.738255033557047,
|
| 699 |
+
"grad_norm": 0.5866240008585512,
|
| 700 |
+
"learning_rate": 5e-06,
|
| 701 |
+
"loss": 0.5403,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.7457121551081283,
|
| 706 |
+
"grad_norm": 0.5925600343817397,
|
| 707 |
+
"learning_rate": 5e-06,
|
| 708 |
+
"loss": 0.548,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.7531692766592095,
|
| 713 |
+
"grad_norm": 0.5905026803607015,
|
| 714 |
+
"learning_rate": 5e-06,
|
| 715 |
+
"loss": 0.5466,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.7606263982102909,
|
| 720 |
+
"grad_norm": 0.5939700140149752,
|
| 721 |
+
"learning_rate": 5e-06,
|
| 722 |
+
"loss": 0.5519,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.7680835197613721,
|
| 727 |
+
"grad_norm": 0.6358723434503033,
|
| 728 |
+
"learning_rate": 5e-06,
|
| 729 |
+
"loss": 0.5465,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.7755406413124534,
|
| 734 |
+
"grad_norm": 0.5893886156719156,
|
| 735 |
+
"learning_rate": 5e-06,
|
| 736 |
+
"loss": 0.5478,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.7829977628635347,
|
| 741 |
+
"grad_norm": 0.6124289839560092,
|
| 742 |
+
"learning_rate": 5e-06,
|
| 743 |
+
"loss": 0.5578,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.790454884414616,
|
| 748 |
+
"grad_norm": 0.5876929465940476,
|
| 749 |
+
"learning_rate": 5e-06,
|
| 750 |
+
"loss": 0.5486,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.7979120059656972,
|
| 755 |
+
"grad_norm": 0.5765792583371309,
|
| 756 |
+
"learning_rate": 5e-06,
|
| 757 |
+
"loss": 0.5453,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.8053691275167785,
|
| 762 |
+
"grad_norm": 0.5763345735177611,
|
| 763 |
+
"learning_rate": 5e-06,
|
| 764 |
+
"loss": 0.54,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.8128262490678598,
|
| 769 |
+
"grad_norm": 0.6392240553326028,
|
| 770 |
+
"learning_rate": 5e-06,
|
| 771 |
+
"loss": 0.5536,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.8202833706189411,
|
| 776 |
+
"grad_norm": 0.5783160074020695,
|
| 777 |
+
"learning_rate": 5e-06,
|
| 778 |
+
"loss": 0.5645,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.8277404921700223,
|
| 783 |
+
"grad_norm": 0.580827750660942,
|
| 784 |
+
"learning_rate": 5e-06,
|
| 785 |
+
"loss": 0.5618,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.8351976137211037,
|
| 790 |
+
"grad_norm": 0.5805219027688908,
|
| 791 |
+
"learning_rate": 5e-06,
|
| 792 |
+
"loss": 0.5378,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.8426547352721849,
|
| 797 |
+
"grad_norm": 0.5631938645784198,
|
| 798 |
+
"learning_rate": 5e-06,
|
| 799 |
+
"loss": 0.552,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.8501118568232662,
|
| 804 |
+
"grad_norm": 0.5460444743065537,
|
| 805 |
+
"learning_rate": 5e-06,
|
| 806 |
+
"loss": 0.5474,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.8575689783743475,
|
| 811 |
+
"grad_norm": 0.619229281669363,
|
| 812 |
+
"learning_rate": 5e-06,
|
| 813 |
+
"loss": 0.5568,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.8650260999254288,
|
| 818 |
+
"grad_norm": 0.5892594363626477,
|
| 819 |
+
"learning_rate": 5e-06,
|
| 820 |
+
"loss": 0.5343,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.87248322147651,
|
| 825 |
+
"grad_norm": 0.6139047400145734,
|
| 826 |
+
"learning_rate": 5e-06,
|
| 827 |
+
"loss": 0.5428,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.8799403430275914,
|
| 832 |
+
"grad_norm": 0.5966906346273294,
|
| 833 |
+
"learning_rate": 5e-06,
|
| 834 |
+
"loss": 0.5356,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.8873974645786726,
|
| 839 |
+
"grad_norm": 0.6535776740165109,
|
| 840 |
+
"learning_rate": 5e-06,
|
| 841 |
+
"loss": 0.5499,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.8948545861297539,
|
| 846 |
+
"grad_norm": 0.5662037475274219,
|
| 847 |
+
"learning_rate": 5e-06,
|
| 848 |
+
"loss": 0.5288,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.9023117076808352,
|
| 853 |
+
"grad_norm": 0.5995000278562762,
|
| 854 |
+
"learning_rate": 5e-06,
|
| 855 |
+
"loss": 0.5563,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.9097688292319165,
|
| 860 |
+
"grad_norm": 0.6346956605590084,
|
| 861 |
+
"learning_rate": 5e-06,
|
| 862 |
+
"loss": 0.5392,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.9172259507829977,
|
| 867 |
+
"grad_norm": 0.5657793485267905,
|
| 868 |
+
"learning_rate": 5e-06,
|
| 869 |
+
"loss": 0.551,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.9246830723340791,
|
| 874 |
+
"grad_norm": 0.5681455949615055,
|
| 875 |
+
"learning_rate": 5e-06,
|
| 876 |
+
"loss": 0.5603,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.9321401938851603,
|
| 881 |
+
"grad_norm": 0.6009029524168309,
|
| 882 |
+
"learning_rate": 5e-06,
|
| 883 |
+
"loss": 0.5468,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.9395973154362416,
|
| 888 |
+
"grad_norm": 0.6145844854677124,
|
| 889 |
+
"learning_rate": 5e-06,
|
| 890 |
+
"loss": 0.5609,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.947054436987323,
|
| 895 |
+
"grad_norm": 0.5696932141639989,
|
| 896 |
+
"learning_rate": 5e-06,
|
| 897 |
+
"loss": 0.5592,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.9545115585384042,
|
| 902 |
+
"grad_norm": 0.5890062683374891,
|
| 903 |
+
"learning_rate": 5e-06,
|
| 904 |
+
"loss": 0.5554,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.9619686800894854,
|
| 909 |
+
"grad_norm": 0.5938117712349276,
|
| 910 |
+
"learning_rate": 5e-06,
|
| 911 |
+
"loss": 0.5646,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.9694258016405667,
|
| 916 |
+
"grad_norm": 0.6257865475961188,
|
| 917 |
+
"learning_rate": 5e-06,
|
| 918 |
+
"loss": 0.5559,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.976882923191648,
|
| 923 |
+
"grad_norm": 0.5860984667723387,
|
| 924 |
+
"learning_rate": 5e-06,
|
| 925 |
+
"loss": 0.5486,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.9843400447427293,
|
| 930 |
+
"grad_norm": 0.5609191548184437,
|
| 931 |
+
"learning_rate": 5e-06,
|
| 932 |
+
"loss": 0.5341,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.9917971662938105,
|
| 937 |
+
"grad_norm": 0.6025440403637561,
|
| 938 |
+
"learning_rate": 5e-06,
|
| 939 |
+
"loss": 0.5446,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.9992542878448919,
|
| 944 |
+
"grad_norm": 0.5572472436768862,
|
| 945 |
+
"learning_rate": 5e-06,
|
| 946 |
+
"loss": 0.5404,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.0,
|
| 951 |
+
"eval_loss": 0.5463822484016418,
|
| 952 |
+
"eval_runtime": 339.2131,
|
| 953 |
+
"eval_samples_per_second": 26.626,
|
| 954 |
+
"eval_steps_per_second": 0.419,
|
| 955 |
+
"step": 1341
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.0067114093959733,
|
| 959 |
+
"grad_norm": 0.6160017220578684,
|
| 960 |
+
"learning_rate": 5e-06,
|
| 961 |
+
"loss": 0.4824,
|
| 962 |
+
"step": 1350
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.0141685309470545,
|
| 966 |
+
"grad_norm": 0.5591512386853066,
|
| 967 |
+
"learning_rate": 5e-06,
|
| 968 |
+
"loss": 0.4797,
|
| 969 |
+
"step": 1360
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.0216256524981358,
|
| 973 |
+
"grad_norm": 0.6118294509759679,
|
| 974 |
+
"learning_rate": 5e-06,
|
| 975 |
+
"loss": 0.4686,
|
| 976 |
+
"step": 1370
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.029082774049217,
|
| 980 |
+
"grad_norm": 0.5583913619404066,
|
| 981 |
+
"learning_rate": 5e-06,
|
| 982 |
+
"loss": 0.469,
|
| 983 |
+
"step": 1380
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.0365398956002982,
|
| 987 |
+
"grad_norm": 0.5818799856636863,
|
| 988 |
+
"learning_rate": 5e-06,
|
| 989 |
+
"loss": 0.4903,
|
| 990 |
+
"step": 1390
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.0439970171513795,
|
| 994 |
+
"grad_norm": 0.6171364008097712,
|
| 995 |
+
"learning_rate": 5e-06,
|
| 996 |
+
"loss": 0.4956,
|
| 997 |
+
"step": 1400
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.0514541387024607,
|
| 1001 |
+
"grad_norm": 0.6061235511507647,
|
| 1002 |
+
"learning_rate": 5e-06,
|
| 1003 |
+
"loss": 0.4933,
|
| 1004 |
+
"step": 1410
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.0589112602535422,
|
| 1008 |
+
"grad_norm": 0.6107761071918465,
|
| 1009 |
+
"learning_rate": 5e-06,
|
| 1010 |
+
"loss": 0.486,
|
| 1011 |
+
"step": 1420
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.0663683818046235,
|
| 1015 |
+
"grad_norm": 0.5547904203912964,
|
| 1016 |
+
"learning_rate": 5e-06,
|
| 1017 |
+
"loss": 0.4968,
|
| 1018 |
+
"step": 1430
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.0738255033557047,
|
| 1022 |
+
"grad_norm": 0.5709552455416576,
|
| 1023 |
+
"learning_rate": 5e-06,
|
| 1024 |
+
"loss": 0.4902,
|
| 1025 |
+
"step": 1440
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.081282624906786,
|
| 1029 |
+
"grad_norm": 0.6263774656448823,
|
| 1030 |
+
"learning_rate": 5e-06,
|
| 1031 |
+
"loss": 0.4703,
|
| 1032 |
+
"step": 1450
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.0887397464578672,
|
| 1036 |
+
"grad_norm": 0.5856371980460074,
|
| 1037 |
+
"learning_rate": 5e-06,
|
| 1038 |
+
"loss": 0.4816,
|
| 1039 |
+
"step": 1460
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.0961968680089484,
|
| 1043 |
+
"grad_norm": 0.5914165561619208,
|
| 1044 |
+
"learning_rate": 5e-06,
|
| 1045 |
+
"loss": 0.479,
|
| 1046 |
+
"step": 1470
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.10365398956003,
|
| 1050 |
+
"grad_norm": 0.6094767293533612,
|
| 1051 |
+
"learning_rate": 5e-06,
|
| 1052 |
+
"loss": 0.4834,
|
| 1053 |
+
"step": 1480
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.1111111111111112,
|
| 1057 |
+
"grad_norm": 0.5628759979218878,
|
| 1058 |
+
"learning_rate": 5e-06,
|
| 1059 |
+
"loss": 0.4807,
|
| 1060 |
+
"step": 1490
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.1185682326621924,
|
| 1064 |
+
"grad_norm": 0.5604069640766214,
|
| 1065 |
+
"learning_rate": 5e-06,
|
| 1066 |
+
"loss": 0.4945,
|
| 1067 |
+
"step": 1500
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.1260253542132737,
|
| 1071 |
+
"grad_norm": 0.6056154435588564,
|
| 1072 |
+
"learning_rate": 5e-06,
|
| 1073 |
+
"loss": 0.4756,
|
| 1074 |
+
"step": 1510
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.133482475764355,
|
| 1078 |
+
"grad_norm": 0.5813579925633866,
|
| 1079 |
+
"learning_rate": 5e-06,
|
| 1080 |
+
"loss": 0.4753,
|
| 1081 |
+
"step": 1520
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.1409395973154361,
|
| 1085 |
+
"grad_norm": 0.5387240892913252,
|
| 1086 |
+
"learning_rate": 5e-06,
|
| 1087 |
+
"loss": 0.474,
|
| 1088 |
+
"step": 1530
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.1483967188665176,
|
| 1092 |
+
"grad_norm": 0.5334644195163835,
|
| 1093 |
+
"learning_rate": 5e-06,
|
| 1094 |
+
"loss": 0.4786,
|
| 1095 |
+
"step": 1540
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.1558538404175989,
|
| 1099 |
+
"grad_norm": 0.6302845458131542,
|
| 1100 |
+
"learning_rate": 5e-06,
|
| 1101 |
+
"loss": 0.4658,
|
| 1102 |
+
"step": 1550
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.1633109619686801,
|
| 1106 |
+
"grad_norm": 0.5458667914397936,
|
| 1107 |
+
"learning_rate": 5e-06,
|
| 1108 |
+
"loss": 0.4884,
|
| 1109 |
+
"step": 1560
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.1707680835197614,
|
| 1113 |
+
"grad_norm": 0.5839445403317228,
|
| 1114 |
+
"learning_rate": 5e-06,
|
| 1115 |
+
"loss": 0.4907,
|
| 1116 |
+
"step": 1570
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.1782252050708426,
|
| 1120 |
+
"grad_norm": 0.5603070510925319,
|
| 1121 |
+
"learning_rate": 5e-06,
|
| 1122 |
+
"loss": 0.4872,
|
| 1123 |
+
"step": 1580
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.1856823266219239,
|
| 1127 |
+
"grad_norm": 0.6298096241603139,
|
| 1128 |
+
"learning_rate": 5e-06,
|
| 1129 |
+
"loss": 0.4816,
|
| 1130 |
+
"step": 1590
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.1931394481730053,
|
| 1134 |
+
"grad_norm": 0.6031228261187916,
|
| 1135 |
+
"learning_rate": 5e-06,
|
| 1136 |
+
"loss": 0.4676,
|
| 1137 |
+
"step": 1600
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.2005965697240866,
|
| 1141 |
+
"grad_norm": 0.5890973152880289,
|
| 1142 |
+
"learning_rate": 5e-06,
|
| 1143 |
+
"loss": 0.4789,
|
| 1144 |
+
"step": 1610
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.2080536912751678,
|
| 1148 |
+
"grad_norm": 0.5915397514486846,
|
| 1149 |
+
"learning_rate": 5e-06,
|
| 1150 |
+
"loss": 0.4816,
|
| 1151 |
+
"step": 1620
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.215510812826249,
|
| 1155 |
+
"grad_norm": 0.5786222706794956,
|
| 1156 |
+
"learning_rate": 5e-06,
|
| 1157 |
+
"loss": 0.4852,
|
| 1158 |
+
"step": 1630
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.2229679343773303,
|
| 1162 |
+
"grad_norm": 0.5699021496499598,
|
| 1163 |
+
"learning_rate": 5e-06,
|
| 1164 |
+
"loss": 0.4986,
|
| 1165 |
+
"step": 1640
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.2304250559284116,
|
| 1169 |
+
"grad_norm": 0.5949853159150295,
|
| 1170 |
+
"learning_rate": 5e-06,
|
| 1171 |
+
"loss": 0.4811,
|
| 1172 |
+
"step": 1650
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.2378821774794928,
|
| 1176 |
+
"grad_norm": 0.5673027327267307,
|
| 1177 |
+
"learning_rate": 5e-06,
|
| 1178 |
+
"loss": 0.4886,
|
| 1179 |
+
"step": 1660
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.2453392990305743,
|
| 1183 |
+
"grad_norm": 0.560192582433626,
|
| 1184 |
+
"learning_rate": 5e-06,
|
| 1185 |
+
"loss": 0.4656,
|
| 1186 |
+
"step": 1670
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.2527964205816555,
|
| 1190 |
+
"grad_norm": 0.6108254414303492,
|
| 1191 |
+
"learning_rate": 5e-06,
|
| 1192 |
+
"loss": 0.4715,
|
| 1193 |
+
"step": 1680
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.2602535421327368,
|
| 1197 |
+
"grad_norm": 0.5590093744521378,
|
| 1198 |
+
"learning_rate": 5e-06,
|
| 1199 |
+
"loss": 0.471,
|
| 1200 |
+
"step": 1690
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.267710663683818,
|
| 1204 |
+
"grad_norm": 0.5914565694717694,
|
| 1205 |
+
"learning_rate": 5e-06,
|
| 1206 |
+
"loss": 0.4736,
|
| 1207 |
+
"step": 1700
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.2751677852348993,
|
| 1211 |
+
"grad_norm": 0.5674996558281405,
|
| 1212 |
+
"learning_rate": 5e-06,
|
| 1213 |
+
"loss": 0.4798,
|
| 1214 |
+
"step": 1710
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.2826249067859807,
|
| 1218 |
+
"grad_norm": 0.6498966266280284,
|
| 1219 |
+
"learning_rate": 5e-06,
|
| 1220 |
+
"loss": 0.4792,
|
| 1221 |
+
"step": 1720
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.290082028337062,
|
| 1225 |
+
"grad_norm": 0.5750798444633526,
|
| 1226 |
+
"learning_rate": 5e-06,
|
| 1227 |
+
"loss": 0.4843,
|
| 1228 |
+
"step": 1730
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.2975391498881432,
|
| 1232 |
+
"grad_norm": 0.5825772193243091,
|
| 1233 |
+
"learning_rate": 5e-06,
|
| 1234 |
+
"loss": 0.4869,
|
| 1235 |
+
"step": 1740
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.3049962714392245,
|
| 1239 |
+
"grad_norm": 0.613520592437996,
|
| 1240 |
+
"learning_rate": 5e-06,
|
| 1241 |
+
"loss": 0.4838,
|
| 1242 |
+
"step": 1750
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.3124533929903057,
|
| 1246 |
+
"grad_norm": 0.6147541770677961,
|
| 1247 |
+
"learning_rate": 5e-06,
|
| 1248 |
+
"loss": 0.4746,
|
| 1249 |
+
"step": 1760
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.319910514541387,
|
| 1253 |
+
"grad_norm": 0.6241611435750868,
|
| 1254 |
+
"learning_rate": 5e-06,
|
| 1255 |
+
"loss": 0.4931,
|
| 1256 |
+
"step": 1770
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.3273676360924682,
|
| 1260 |
+
"grad_norm": 0.5926826684472789,
|
| 1261 |
+
"learning_rate": 5e-06,
|
| 1262 |
+
"loss": 0.4871,
|
| 1263 |
+
"step": 1780
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.3348247576435495,
|
| 1267 |
+
"grad_norm": 0.6066670454162487,
|
| 1268 |
+
"learning_rate": 5e-06,
|
| 1269 |
+
"loss": 0.4841,
|
| 1270 |
+
"step": 1790
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.342281879194631,
|
| 1274 |
+
"grad_norm": 0.6069960632718007,
|
| 1275 |
+
"learning_rate": 5e-06,
|
| 1276 |
+
"loss": 0.4862,
|
| 1277 |
+
"step": 1800
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.3497390007457122,
|
| 1281 |
+
"grad_norm": 0.5479395919590173,
|
| 1282 |
+
"learning_rate": 5e-06,
|
| 1283 |
+
"loss": 0.4766,
|
| 1284 |
+
"step": 1810
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.3571961222967934,
|
| 1288 |
+
"grad_norm": 0.5639904817681118,
|
| 1289 |
+
"learning_rate": 5e-06,
|
| 1290 |
+
"loss": 0.4812,
|
| 1291 |
+
"step": 1820
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.3646532438478747,
|
| 1295 |
+
"grad_norm": 0.5858843238376426,
|
| 1296 |
+
"learning_rate": 5e-06,
|
| 1297 |
+
"loss": 0.4686,
|
| 1298 |
+
"step": 1830
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.372110365398956,
|
| 1302 |
+
"grad_norm": 0.5734924915807309,
|
| 1303 |
+
"learning_rate": 5e-06,
|
| 1304 |
+
"loss": 0.477,
|
| 1305 |
+
"step": 1840
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.3795674869500374,
|
| 1309 |
+
"grad_norm": 0.5902217700915133,
|
| 1310 |
+
"learning_rate": 5e-06,
|
| 1311 |
+
"loss": 0.4889,
|
| 1312 |
+
"step": 1850
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.3870246085011186,
|
| 1316 |
+
"grad_norm": 0.5854924372795176,
|
| 1317 |
+
"learning_rate": 5e-06,
|
| 1318 |
+
"loss": 0.4792,
|
| 1319 |
+
"step": 1860
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.3944817300521999,
|
| 1323 |
+
"grad_norm": 0.6350410086083018,
|
| 1324 |
+
"learning_rate": 5e-06,
|
| 1325 |
+
"loss": 0.4897,
|
| 1326 |
+
"step": 1870
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.4019388516032811,
|
| 1330 |
+
"grad_norm": 0.5606930586639369,
|
| 1331 |
+
"learning_rate": 5e-06,
|
| 1332 |
+
"loss": 0.496,
|
| 1333 |
+
"step": 1880
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.4093959731543624,
|
| 1337 |
+
"grad_norm": 0.5586778910160224,
|
| 1338 |
+
"learning_rate": 5e-06,
|
| 1339 |
+
"loss": 0.4671,
|
| 1340 |
+
"step": 1890
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.4168530947054436,
|
| 1344 |
+
"grad_norm": 0.5732737315369484,
|
| 1345 |
+
"learning_rate": 5e-06,
|
| 1346 |
+
"loss": 0.4786,
|
| 1347 |
+
"step": 1900
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.4243102162565249,
|
| 1351 |
+
"grad_norm": 4.027507437517355,
|
| 1352 |
+
"learning_rate": 5e-06,
|
| 1353 |
+
"loss": 0.4886,
|
| 1354 |
+
"step": 1910
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.4317673378076063,
|
| 1358 |
+
"grad_norm": 0.5991590420531947,
|
| 1359 |
+
"learning_rate": 5e-06,
|
| 1360 |
+
"loss": 0.497,
|
| 1361 |
+
"step": 1920
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.4392244593586876,
|
| 1365 |
+
"grad_norm": 0.5828725709335938,
|
| 1366 |
+
"learning_rate": 5e-06,
|
| 1367 |
+
"loss": 0.4899,
|
| 1368 |
+
"step": 1930
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.4466815809097688,
|
| 1372 |
+
"grad_norm": 0.5473354356011987,
|
| 1373 |
+
"learning_rate": 5e-06,
|
| 1374 |
+
"loss": 0.4775,
|
| 1375 |
+
"step": 1940
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.45413870246085,
|
| 1379 |
+
"grad_norm": 0.6210979610287483,
|
| 1380 |
+
"learning_rate": 5e-06,
|
| 1381 |
+
"loss": 0.4907,
|
| 1382 |
+
"step": 1950
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.4615958240119313,
|
| 1386 |
+
"grad_norm": 0.5638329844959816,
|
| 1387 |
+
"learning_rate": 5e-06,
|
| 1388 |
+
"loss": 0.4759,
|
| 1389 |
+
"step": 1960
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.4690529455630128,
|
| 1393 |
+
"grad_norm": 0.5723878428151293,
|
| 1394 |
+
"learning_rate": 5e-06,
|
| 1395 |
+
"loss": 0.4897,
|
| 1396 |
+
"step": 1970
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.476510067114094,
|
| 1400 |
+
"grad_norm": 0.5521756950405313,
|
| 1401 |
+
"learning_rate": 5e-06,
|
| 1402 |
+
"loss": 0.4746,
|
| 1403 |
+
"step": 1980
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.4839671886651753,
|
| 1407 |
+
"grad_norm": 0.6691296670250785,
|
| 1408 |
+
"learning_rate": 5e-06,
|
| 1409 |
+
"loss": 0.4885,
|
| 1410 |
+
"step": 1990
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.4914243102162565,
|
| 1414 |
+
"grad_norm": 0.6141910929035425,
|
| 1415 |
+
"learning_rate": 5e-06,
|
| 1416 |
+
"loss": 0.4884,
|
| 1417 |
+
"step": 2000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.4988814317673378,
|
| 1421 |
+
"grad_norm": 0.5400373664496432,
|
| 1422 |
+
"learning_rate": 5e-06,
|
| 1423 |
+
"loss": 0.4799,
|
| 1424 |
+
"step": 2010
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.506338553318419,
|
| 1428 |
+
"grad_norm": 0.5645402137800272,
|
| 1429 |
+
"learning_rate": 5e-06,
|
| 1430 |
+
"loss": 0.4832,
|
| 1431 |
+
"step": 2020
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.5137956748695003,
|
| 1435 |
+
"grad_norm": 0.59297682178836,
|
| 1436 |
+
"learning_rate": 5e-06,
|
| 1437 |
+
"loss": 0.4765,
|
| 1438 |
+
"step": 2030
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.5212527964205815,
|
| 1442 |
+
"grad_norm": 0.6099720079302091,
|
| 1443 |
+
"learning_rate": 5e-06,
|
| 1444 |
+
"loss": 0.4876,
|
| 1445 |
+
"step": 2040
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.5287099179716628,
|
| 1449 |
+
"grad_norm": 0.5797412589969116,
|
| 1450 |
+
"learning_rate": 5e-06,
|
| 1451 |
+
"loss": 0.4867,
|
| 1452 |
+
"step": 2050
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.5361670395227442,
|
| 1456 |
+
"grad_norm": 0.6160597774931952,
|
| 1457 |
+
"learning_rate": 5e-06,
|
| 1458 |
+
"loss": 0.4952,
|
| 1459 |
+
"step": 2060
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.5436241610738255,
|
| 1463 |
+
"grad_norm": 0.5775983688665581,
|
| 1464 |
+
"learning_rate": 5e-06,
|
| 1465 |
+
"loss": 0.476,
|
| 1466 |
+
"step": 2070
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.5510812826249067,
|
| 1470 |
+
"grad_norm": 0.5846473827271578,
|
| 1471 |
+
"learning_rate": 5e-06,
|
| 1472 |
+
"loss": 0.4937,
|
| 1473 |
+
"step": 2080
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.5585384041759882,
|
| 1477 |
+
"grad_norm": 0.5702728266559334,
|
| 1478 |
+
"learning_rate": 5e-06,
|
| 1479 |
+
"loss": 0.4974,
|
| 1480 |
+
"step": 2090
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.5659955257270695,
|
| 1484 |
+
"grad_norm": 0.5870580247571568,
|
| 1485 |
+
"learning_rate": 5e-06,
|
| 1486 |
+
"loss": 0.4877,
|
| 1487 |
+
"step": 2100
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.5734526472781507,
|
| 1491 |
+
"grad_norm": 0.5817243215012845,
|
| 1492 |
+
"learning_rate": 5e-06,
|
| 1493 |
+
"loss": 0.4873,
|
| 1494 |
+
"step": 2110
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.580909768829232,
|
| 1498 |
+
"grad_norm": 0.5315491273649968,
|
| 1499 |
+
"learning_rate": 5e-06,
|
| 1500 |
+
"loss": 0.4798,
|
| 1501 |
+
"step": 2120
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.5883668903803132,
|
| 1505 |
+
"grad_norm": 0.6200889922410328,
|
| 1506 |
+
"learning_rate": 5e-06,
|
| 1507 |
+
"loss": 0.4699,
|
| 1508 |
+
"step": 2130
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.5958240119313944,
|
| 1512 |
+
"grad_norm": 0.6029121739064351,
|
| 1513 |
+
"learning_rate": 5e-06,
|
| 1514 |
+
"loss": 0.4995,
|
| 1515 |
+
"step": 2140
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.6032811334824757,
|
| 1519 |
+
"grad_norm": 0.6028582417864272,
|
| 1520 |
+
"learning_rate": 5e-06,
|
| 1521 |
+
"loss": 0.4959,
|
| 1522 |
+
"step": 2150
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.610738255033557,
|
| 1526 |
+
"grad_norm": 0.5893432209254877,
|
| 1527 |
+
"learning_rate": 5e-06,
|
| 1528 |
+
"loss": 0.4783,
|
| 1529 |
+
"step": 2160
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.6181953765846382,
|
| 1533 |
+
"grad_norm": 0.5884082032907385,
|
| 1534 |
+
"learning_rate": 5e-06,
|
| 1535 |
+
"loss": 0.4823,
|
| 1536 |
+
"step": 2170
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.6256524981357197,
|
| 1540 |
+
"grad_norm": 0.5921720057393691,
|
| 1541 |
+
"learning_rate": 5e-06,
|
| 1542 |
+
"loss": 0.4761,
|
| 1543 |
+
"step": 2180
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.633109619686801,
|
| 1547 |
+
"grad_norm": 0.5776696223531537,
|
| 1548 |
+
"learning_rate": 5e-06,
|
| 1549 |
+
"loss": 0.4823,
|
| 1550 |
+
"step": 2190
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.6405667412378822,
|
| 1554 |
+
"grad_norm": 0.5645109392805745,
|
| 1555 |
+
"learning_rate": 5e-06,
|
| 1556 |
+
"loss": 0.4827,
|
| 1557 |
+
"step": 2200
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.6480238627889636,
|
| 1561 |
+
"grad_norm": 0.5532331562154945,
|
| 1562 |
+
"learning_rate": 5e-06,
|
| 1563 |
+
"loss": 0.4878,
|
| 1564 |
+
"step": 2210
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.6554809843400449,
|
| 1568 |
+
"grad_norm": 0.5877417623919632,
|
| 1569 |
+
"learning_rate": 5e-06,
|
| 1570 |
+
"loss": 0.4854,
|
| 1571 |
+
"step": 2220
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.6629381058911261,
|
| 1575 |
+
"grad_norm": 0.5528378617144938,
|
| 1576 |
+
"learning_rate": 5e-06,
|
| 1577 |
+
"loss": 0.4876,
|
| 1578 |
+
"step": 2230
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.6703952274422074,
|
| 1582 |
+
"grad_norm": 0.6104249956677287,
|
| 1583 |
+
"learning_rate": 5e-06,
|
| 1584 |
+
"loss": 0.4976,
|
| 1585 |
+
"step": 2240
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.6778523489932886,
|
| 1589 |
+
"grad_norm": 0.5676314169193764,
|
| 1590 |
+
"learning_rate": 5e-06,
|
| 1591 |
+
"loss": 0.473,
|
| 1592 |
+
"step": 2250
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.6853094705443699,
|
| 1596 |
+
"grad_norm": 0.5696949870525951,
|
| 1597 |
+
"learning_rate": 5e-06,
|
| 1598 |
+
"loss": 0.4866,
|
| 1599 |
+
"step": 2260
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.692766592095451,
|
| 1603 |
+
"grad_norm": 0.5702619592671349,
|
| 1604 |
+
"learning_rate": 5e-06,
|
| 1605 |
+
"loss": 0.4909,
|
| 1606 |
+
"step": 2270
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.7002237136465324,
|
| 1610 |
+
"grad_norm": 0.5973494243465649,
|
| 1611 |
+
"learning_rate": 5e-06,
|
| 1612 |
+
"loss": 0.4914,
|
| 1613 |
+
"step": 2280
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.7076808351976136,
|
| 1617 |
+
"grad_norm": 0.5545197259083948,
|
| 1618 |
+
"learning_rate": 5e-06,
|
| 1619 |
+
"loss": 0.4804,
|
| 1620 |
+
"step": 2290
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.7151379567486948,
|
| 1624 |
+
"grad_norm": 0.5905506685708721,
|
| 1625 |
+
"learning_rate": 5e-06,
|
| 1626 |
+
"loss": 0.4925,
|
| 1627 |
+
"step": 2300
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.7225950782997763,
|
| 1631 |
+
"grad_norm": 0.6320966702911732,
|
| 1632 |
+
"learning_rate": 5e-06,
|
| 1633 |
+
"loss": 0.4913,
|
| 1634 |
+
"step": 2310
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.7300521998508576,
|
| 1638 |
+
"grad_norm": 0.5417972662105043,
|
| 1639 |
+
"learning_rate": 5e-06,
|
| 1640 |
+
"loss": 0.4769,
|
| 1641 |
+
"step": 2320
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.7375093214019388,
|
| 1645 |
+
"grad_norm": 0.597965564481012,
|
| 1646 |
+
"learning_rate": 5e-06,
|
| 1647 |
+
"loss": 0.4898,
|
| 1648 |
+
"step": 2330
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.7449664429530203,
|
| 1652 |
+
"grad_norm": 0.6144634291593705,
|
| 1653 |
+
"learning_rate": 5e-06,
|
| 1654 |
+
"loss": 0.4826,
|
| 1655 |
+
"step": 2340
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.7524235645041015,
|
| 1659 |
+
"grad_norm": 0.5987390856089427,
|
| 1660 |
+
"learning_rate": 5e-06,
|
| 1661 |
+
"loss": 0.4777,
|
| 1662 |
+
"step": 2350
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.7598806860551828,
|
| 1666 |
+
"grad_norm": 0.5693070234792923,
|
| 1667 |
+
"learning_rate": 5e-06,
|
| 1668 |
+
"loss": 0.4934,
|
| 1669 |
+
"step": 2360
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.767337807606264,
|
| 1673 |
+
"grad_norm": 0.5879521460157956,
|
| 1674 |
+
"learning_rate": 5e-06,
|
| 1675 |
+
"loss": 0.5053,
|
| 1676 |
+
"step": 2370
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.7747949291573453,
|
| 1680 |
+
"grad_norm": 0.5680219111185565,
|
| 1681 |
+
"learning_rate": 5e-06,
|
| 1682 |
+
"loss": 0.4817,
|
| 1683 |
+
"step": 2380
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.7822520507084265,
|
| 1687 |
+
"grad_norm": 0.6501364801989891,
|
| 1688 |
+
"learning_rate": 5e-06,
|
| 1689 |
+
"loss": 0.4853,
|
| 1690 |
+
"step": 2390
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.7897091722595078,
|
| 1694 |
+
"grad_norm": 0.615030215815147,
|
| 1695 |
+
"learning_rate": 5e-06,
|
| 1696 |
+
"loss": 0.4791,
|
| 1697 |
+
"step": 2400
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.797166293810589,
|
| 1701 |
+
"grad_norm": 0.5785652647623134,
|
| 1702 |
+
"learning_rate": 5e-06,
|
| 1703 |
+
"loss": 0.4809,
|
| 1704 |
+
"step": 2410
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.8046234153616703,
|
| 1708 |
+
"grad_norm": 0.6044924431994139,
|
| 1709 |
+
"learning_rate": 5e-06,
|
| 1710 |
+
"loss": 0.4929,
|
| 1711 |
+
"step": 2420
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.8120805369127517,
|
| 1715 |
+
"grad_norm": 0.5426508657118521,
|
| 1716 |
+
"learning_rate": 5e-06,
|
| 1717 |
+
"loss": 0.4847,
|
| 1718 |
+
"step": 2430
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.819537658463833,
|
| 1722 |
+
"grad_norm": 0.6009223844637949,
|
| 1723 |
+
"learning_rate": 5e-06,
|
| 1724 |
+
"loss": 0.4906,
|
| 1725 |
+
"step": 2440
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.8269947800149142,
|
| 1729 |
+
"grad_norm": 0.5743711413707431,
|
| 1730 |
+
"learning_rate": 5e-06,
|
| 1731 |
+
"loss": 0.4975,
|
| 1732 |
+
"step": 2450
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.8344519015659957,
|
| 1736 |
+
"grad_norm": 0.5983412684564712,
|
| 1737 |
+
"learning_rate": 5e-06,
|
| 1738 |
+
"loss": 0.4814,
|
| 1739 |
+
"step": 2460
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.841909023117077,
|
| 1743 |
+
"grad_norm": 0.5560773314183736,
|
| 1744 |
+
"learning_rate": 5e-06,
|
| 1745 |
+
"loss": 0.4811,
|
| 1746 |
+
"step": 2470
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.8493661446681582,
|
| 1750 |
+
"grad_norm": 0.5992209963884128,
|
| 1751 |
+
"learning_rate": 5e-06,
|
| 1752 |
+
"loss": 0.496,
|
| 1753 |
+
"step": 2480
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.8568232662192394,
|
| 1757 |
+
"grad_norm": 0.5387716311327034,
|
| 1758 |
+
"learning_rate": 5e-06,
|
| 1759 |
+
"loss": 0.4804,
|
| 1760 |
+
"step": 2490
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.8642803877703207,
|
| 1764 |
+
"grad_norm": 0.5990433932479565,
|
| 1765 |
+
"learning_rate": 5e-06,
|
| 1766 |
+
"loss": 0.4863,
|
| 1767 |
+
"step": 2500
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.871737509321402,
|
| 1771 |
+
"grad_norm": 0.5341574484686165,
|
| 1772 |
+
"learning_rate": 5e-06,
|
| 1773 |
+
"loss": 0.4909,
|
| 1774 |
+
"step": 2510
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.8791946308724832,
|
| 1778 |
+
"grad_norm": 0.5857022850771958,
|
| 1779 |
+
"learning_rate": 5e-06,
|
| 1780 |
+
"loss": 0.4831,
|
| 1781 |
+
"step": 2520
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.8866517524235644,
|
| 1785 |
+
"grad_norm": 0.596806200047546,
|
| 1786 |
+
"learning_rate": 5e-06,
|
| 1787 |
+
"loss": 0.5042,
|
| 1788 |
+
"step": 2530
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.8941088739746457,
|
| 1792 |
+
"grad_norm": 0.5831993654340557,
|
| 1793 |
+
"learning_rate": 5e-06,
|
| 1794 |
+
"loss": 0.5004,
|
| 1795 |
+
"step": 2540
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.901565995525727,
|
| 1799 |
+
"grad_norm": 0.5256203832071005,
|
| 1800 |
+
"learning_rate": 5e-06,
|
| 1801 |
+
"loss": 0.4802,
|
| 1802 |
+
"step": 2550
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.9090231170768084,
|
| 1806 |
+
"grad_norm": 0.6256000165719325,
|
| 1807 |
+
"learning_rate": 5e-06,
|
| 1808 |
+
"loss": 0.49,
|
| 1809 |
+
"step": 2560
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.9164802386278896,
|
| 1813 |
+
"grad_norm": 0.6025652662604613,
|
| 1814 |
+
"learning_rate": 5e-06,
|
| 1815 |
+
"loss": 0.4961,
|
| 1816 |
+
"step": 2570
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 1.9239373601789709,
|
| 1820 |
+
"grad_norm": 0.5851336978699863,
|
| 1821 |
+
"learning_rate": 5e-06,
|
| 1822 |
+
"loss": 0.493,
|
| 1823 |
+
"step": 2580
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.9313944817300523,
|
| 1827 |
+
"grad_norm": 0.5841404426808496,
|
| 1828 |
+
"learning_rate": 5e-06,
|
| 1829 |
+
"loss": 0.4788,
|
| 1830 |
+
"step": 2590
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.9388516032811336,
|
| 1834 |
+
"grad_norm": 0.5553137000045634,
|
| 1835 |
+
"learning_rate": 5e-06,
|
| 1836 |
+
"loss": 0.4944,
|
| 1837 |
+
"step": 2600
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 1.9463087248322148,
|
| 1841 |
+
"grad_norm": 0.5954485746159562,
|
| 1842 |
+
"learning_rate": 5e-06,
|
| 1843 |
+
"loss": 0.4924,
|
| 1844 |
+
"step": 2610
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.953765846383296,
|
| 1848 |
+
"grad_norm": 0.573701644653239,
|
| 1849 |
+
"learning_rate": 5e-06,
|
| 1850 |
+
"loss": 0.4909,
|
| 1851 |
+
"step": 2620
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.9612229679343773,
|
| 1855 |
+
"grad_norm": 0.6460568307562583,
|
| 1856 |
+
"learning_rate": 5e-06,
|
| 1857 |
+
"loss": 0.4981,
|
| 1858 |
+
"step": 2630
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.9686800894854586,
|
| 1862 |
+
"grad_norm": 0.5881894639448034,
|
| 1863 |
+
"learning_rate": 5e-06,
|
| 1864 |
+
"loss": 0.498,
|
| 1865 |
+
"step": 2640
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 1.9761372110365398,
|
| 1869 |
+
"grad_norm": 0.5442840215629223,
|
| 1870 |
+
"learning_rate": 5e-06,
|
| 1871 |
+
"loss": 0.4793,
|
| 1872 |
+
"step": 2650
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.983594332587621,
|
| 1876 |
+
"grad_norm": 0.5772375310283397,
|
| 1877 |
+
"learning_rate": 5e-06,
|
| 1878 |
+
"loss": 0.4877,
|
| 1879 |
+
"step": 2660
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 1.9910514541387023,
|
| 1883 |
+
"grad_norm": 0.5624955248751579,
|
| 1884 |
+
"learning_rate": 5e-06,
|
| 1885 |
+
"loss": 0.4852,
|
| 1886 |
+
"step": 2670
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.9985085756897838,
|
| 1890 |
+
"grad_norm": 0.5618877188539807,
|
| 1891 |
+
"learning_rate": 5e-06,
|
| 1892 |
+
"loss": 0.4804,
|
| 1893 |
+
"step": 2680
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.0,
|
| 1897 |
+
"eval_loss": 0.5512435436248779,
|
| 1898 |
+
"eval_runtime": 342.0704,
|
| 1899 |
+
"eval_samples_per_second": 26.404,
|
| 1900 |
+
"eval_steps_per_second": 0.415,
|
| 1901 |
+
"step": 2682
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 2.005965697240865,
|
| 1905 |
+
"grad_norm": 0.6558184793140344,
|
| 1906 |
+
"learning_rate": 5e-06,
|
| 1907 |
+
"loss": 0.4215,
|
| 1908 |
+
"step": 2690
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 2.0134228187919465,
|
| 1912 |
+
"grad_norm": 0.6034125785194365,
|
| 1913 |
+
"learning_rate": 5e-06,
|
| 1914 |
+
"loss": 0.4033,
|
| 1915 |
+
"step": 2700
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 2.0208799403430278,
|
| 1919 |
+
"grad_norm": 0.5610165111292696,
|
| 1920 |
+
"learning_rate": 5e-06,
|
| 1921 |
+
"loss": 0.3907,
|
| 1922 |
+
"step": 2710
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 2.028337061894109,
|
| 1926 |
+
"grad_norm": 0.580517844768164,
|
| 1927 |
+
"learning_rate": 5e-06,
|
| 1928 |
+
"loss": 0.3977,
|
| 1929 |
+
"step": 2720
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 2.0357941834451903,
|
| 1933 |
+
"grad_norm": 0.5514277294221982,
|
| 1934 |
+
"learning_rate": 5e-06,
|
| 1935 |
+
"loss": 0.4112,
|
| 1936 |
+
"step": 2730
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 2.0432513049962715,
|
| 1940 |
+
"grad_norm": 0.60029875754166,
|
| 1941 |
+
"learning_rate": 5e-06,
|
| 1942 |
+
"loss": 0.4176,
|
| 1943 |
+
"step": 2740
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 2.0507084265473527,
|
| 1947 |
+
"grad_norm": 0.6089301446363884,
|
| 1948 |
+
"learning_rate": 5e-06,
|
| 1949 |
+
"loss": 0.4026,
|
| 1950 |
+
"step": 2750
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 2.058165548098434,
|
| 1954 |
+
"grad_norm": 0.5849276057096263,
|
| 1955 |
+
"learning_rate": 5e-06,
|
| 1956 |
+
"loss": 0.409,
|
| 1957 |
+
"step": 2760
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 2.0656226696495152,
|
| 1961 |
+
"grad_norm": 0.5950355197672823,
|
| 1962 |
+
"learning_rate": 5e-06,
|
| 1963 |
+
"loss": 0.4024,
|
| 1964 |
+
"step": 2770
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 2.0730797912005965,
|
| 1968 |
+
"grad_norm": 0.5889941012752544,
|
| 1969 |
+
"learning_rate": 5e-06,
|
| 1970 |
+
"loss": 0.4175,
|
| 1971 |
+
"step": 2780
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 2.0805369127516777,
|
| 1975 |
+
"grad_norm": 0.5611907053787162,
|
| 1976 |
+
"learning_rate": 5e-06,
|
| 1977 |
+
"loss": 0.403,
|
| 1978 |
+
"step": 2790
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 2.087994034302759,
|
| 1982 |
+
"grad_norm": 0.5934324272584153,
|
| 1983 |
+
"learning_rate": 5e-06,
|
| 1984 |
+
"loss": 0.413,
|
| 1985 |
+
"step": 2800
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 2.0954511558538402,
|
| 1989 |
+
"grad_norm": 0.5774168779978719,
|
| 1990 |
+
"learning_rate": 5e-06,
|
| 1991 |
+
"loss": 0.4159,
|
| 1992 |
+
"step": 2810
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 2.1029082774049215,
|
| 1996 |
+
"grad_norm": 0.564340240629499,
|
| 1997 |
+
"learning_rate": 5e-06,
|
| 1998 |
+
"loss": 0.4107,
|
| 1999 |
+
"step": 2820
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 2.110365398956003,
|
| 2003 |
+
"grad_norm": 0.6144699399372608,
|
| 2004 |
+
"learning_rate": 5e-06,
|
| 2005 |
+
"loss": 0.415,
|
| 2006 |
+
"step": 2830
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 2.1178225205070844,
|
| 2010 |
+
"grad_norm": 0.623817632576955,
|
| 2011 |
+
"learning_rate": 5e-06,
|
| 2012 |
+
"loss": 0.415,
|
| 2013 |
+
"step": 2840
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 2.1252796420581657,
|
| 2017 |
+
"grad_norm": 0.5835224903490361,
|
| 2018 |
+
"learning_rate": 5e-06,
|
| 2019 |
+
"loss": 0.4096,
|
| 2020 |
+
"step": 2850
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 2.132736763609247,
|
| 2024 |
+
"grad_norm": 0.6186331596636564,
|
| 2025 |
+
"learning_rate": 5e-06,
|
| 2026 |
+
"loss": 0.4015,
|
| 2027 |
+
"step": 2860
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 2.140193885160328,
|
| 2031 |
+
"grad_norm": 0.5790608361059295,
|
| 2032 |
+
"learning_rate": 5e-06,
|
| 2033 |
+
"loss": 0.4068,
|
| 2034 |
+
"step": 2870
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 2.1476510067114094,
|
| 2038 |
+
"grad_norm": 0.5869215595357605,
|
| 2039 |
+
"learning_rate": 5e-06,
|
| 2040 |
+
"loss": 0.4215,
|
| 2041 |
+
"step": 2880
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 2.1551081282624907,
|
| 2045 |
+
"grad_norm": 0.5936702509963696,
|
| 2046 |
+
"learning_rate": 5e-06,
|
| 2047 |
+
"loss": 0.4141,
|
| 2048 |
+
"step": 2890
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 2.162565249813572,
|
| 2052 |
+
"grad_norm": 0.6381922955234081,
|
| 2053 |
+
"learning_rate": 5e-06,
|
| 2054 |
+
"loss": 0.407,
|
| 2055 |
+
"step": 2900
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 2.170022371364653,
|
| 2059 |
+
"grad_norm": 0.5862737710378301,
|
| 2060 |
+
"learning_rate": 5e-06,
|
| 2061 |
+
"loss": 0.417,
|
| 2062 |
+
"step": 2910
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 2.1774794929157344,
|
| 2066 |
+
"grad_norm": 0.5822255490097745,
|
| 2067 |
+
"learning_rate": 5e-06,
|
| 2068 |
+
"loss": 0.4225,
|
| 2069 |
+
"step": 2920
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 2.1849366144668156,
|
| 2073 |
+
"grad_norm": 0.5764505027268744,
|
| 2074 |
+
"learning_rate": 5e-06,
|
| 2075 |
+
"loss": 0.4138,
|
| 2076 |
+
"step": 2930
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 2.192393736017897,
|
| 2080 |
+
"grad_norm": 0.6073760733719726,
|
| 2081 |
+
"learning_rate": 5e-06,
|
| 2082 |
+
"loss": 0.4112,
|
| 2083 |
+
"step": 2940
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 2.1998508575689786,
|
| 2087 |
+
"grad_norm": 0.6091808835643772,
|
| 2088 |
+
"learning_rate": 5e-06,
|
| 2089 |
+
"loss": 0.4208,
|
| 2090 |
+
"step": 2950
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 2.20730797912006,
|
| 2094 |
+
"grad_norm": 0.6239375451432535,
|
| 2095 |
+
"learning_rate": 5e-06,
|
| 2096 |
+
"loss": 0.4167,
|
| 2097 |
+
"step": 2960
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 2.214765100671141,
|
| 2101 |
+
"grad_norm": 0.6130674669978408,
|
| 2102 |
+
"learning_rate": 5e-06,
|
| 2103 |
+
"loss": 0.413,
|
| 2104 |
+
"step": 2970
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 2.2222222222222223,
|
| 2108 |
+
"grad_norm": 0.5668324865537339,
|
| 2109 |
+
"learning_rate": 5e-06,
|
| 2110 |
+
"loss": 0.4126,
|
| 2111 |
+
"step": 2980
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 2.2296793437733036,
|
| 2115 |
+
"grad_norm": 0.5765769707939168,
|
| 2116 |
+
"learning_rate": 5e-06,
|
| 2117 |
+
"loss": 0.4189,
|
| 2118 |
+
"step": 2990
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 2.237136465324385,
|
| 2122 |
+
"grad_norm": 0.6216905758163017,
|
| 2123 |
+
"learning_rate": 5e-06,
|
| 2124 |
+
"loss": 0.416,
|
| 2125 |
+
"step": 3000
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 2.244593586875466,
|
| 2129 |
+
"grad_norm": 0.642406145303932,
|
| 2130 |
+
"learning_rate": 5e-06,
|
| 2131 |
+
"loss": 0.4182,
|
| 2132 |
+
"step": 3010
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 2.2520507084265473,
|
| 2136 |
+
"grad_norm": 0.6153726361933832,
|
| 2137 |
+
"learning_rate": 5e-06,
|
| 2138 |
+
"loss": 0.4401,
|
| 2139 |
+
"step": 3020
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 2.2595078299776286,
|
| 2143 |
+
"grad_norm": 0.6162592369179899,
|
| 2144 |
+
"learning_rate": 5e-06,
|
| 2145 |
+
"loss": 0.4093,
|
| 2146 |
+
"step": 3030
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 2.26696495152871,
|
| 2150 |
+
"grad_norm": 0.6062140595912686,
|
| 2151 |
+
"learning_rate": 5e-06,
|
| 2152 |
+
"loss": 0.4201,
|
| 2153 |
+
"step": 3040
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 2.274422073079791,
|
| 2157 |
+
"grad_norm": 0.6489911104793248,
|
| 2158 |
+
"learning_rate": 5e-06,
|
| 2159 |
+
"loss": 0.4216,
|
| 2160 |
+
"step": 3050
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 2.2818791946308723,
|
| 2164 |
+
"grad_norm": 0.5973147232668122,
|
| 2165 |
+
"learning_rate": 5e-06,
|
| 2166 |
+
"loss": 0.4225,
|
| 2167 |
+
"step": 3060
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 2.289336316181954,
|
| 2171 |
+
"grad_norm": 0.6029950937606031,
|
| 2172 |
+
"learning_rate": 5e-06,
|
| 2173 |
+
"loss": 0.4187,
|
| 2174 |
+
"step": 3070
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 2.2967934377330352,
|
| 2178 |
+
"grad_norm": 0.6231584364415401,
|
| 2179 |
+
"learning_rate": 5e-06,
|
| 2180 |
+
"loss": 0.4238,
|
| 2181 |
+
"step": 3080
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 2.3042505592841165,
|
| 2185 |
+
"grad_norm": 0.5561184622741537,
|
| 2186 |
+
"learning_rate": 5e-06,
|
| 2187 |
+
"loss": 0.4219,
|
| 2188 |
+
"step": 3090
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 2.3117076808351977,
|
| 2192 |
+
"grad_norm": 0.5941537102506256,
|
| 2193 |
+
"learning_rate": 5e-06,
|
| 2194 |
+
"loss": 0.4246,
|
| 2195 |
+
"step": 3100
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 2.319164802386279,
|
| 2199 |
+
"grad_norm": 0.6410414670886565,
|
| 2200 |
+
"learning_rate": 5e-06,
|
| 2201 |
+
"loss": 0.4071,
|
| 2202 |
+
"step": 3110
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 2.3266219239373602,
|
| 2206 |
+
"grad_norm": 0.5732228808060483,
|
| 2207 |
+
"learning_rate": 5e-06,
|
| 2208 |
+
"loss": 0.4114,
|
| 2209 |
+
"step": 3120
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 2.3340790454884415,
|
| 2213 |
+
"grad_norm": 0.5736385479610295,
|
| 2214 |
+
"learning_rate": 5e-06,
|
| 2215 |
+
"loss": 0.4154,
|
| 2216 |
+
"step": 3130
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 2.3415361670395227,
|
| 2220 |
+
"grad_norm": 0.5563198902731237,
|
| 2221 |
+
"learning_rate": 5e-06,
|
| 2222 |
+
"loss": 0.4342,
|
| 2223 |
+
"step": 3140
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 2.348993288590604,
|
| 2227 |
+
"grad_norm": 0.6010188015263107,
|
| 2228 |
+
"learning_rate": 5e-06,
|
| 2229 |
+
"loss": 0.4115,
|
| 2230 |
+
"step": 3150
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 2.356450410141685,
|
| 2234 |
+
"grad_norm": 0.6377162414933872,
|
| 2235 |
+
"learning_rate": 5e-06,
|
| 2236 |
+
"loss": 0.4129,
|
| 2237 |
+
"step": 3160
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 2.3639075316927665,
|
| 2241 |
+
"grad_norm": 0.5983638229139512,
|
| 2242 |
+
"learning_rate": 5e-06,
|
| 2243 |
+
"loss": 0.4116,
|
| 2244 |
+
"step": 3170
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 2.3713646532438477,
|
| 2248 |
+
"grad_norm": 0.6406800867031649,
|
| 2249 |
+
"learning_rate": 5e-06,
|
| 2250 |
+
"loss": 0.4257,
|
| 2251 |
+
"step": 3180
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 2.378821774794929,
|
| 2255 |
+
"grad_norm": 0.6165353698446668,
|
| 2256 |
+
"learning_rate": 5e-06,
|
| 2257 |
+
"loss": 0.4332,
|
| 2258 |
+
"step": 3190
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 2.3862788963460106,
|
| 2262 |
+
"grad_norm": 0.5950518706856058,
|
| 2263 |
+
"learning_rate": 5e-06,
|
| 2264 |
+
"loss": 0.4166,
|
| 2265 |
+
"step": 3200
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 2.393736017897092,
|
| 2269 |
+
"grad_norm": 0.5601525861175598,
|
| 2270 |
+
"learning_rate": 5e-06,
|
| 2271 |
+
"loss": 0.4214,
|
| 2272 |
+
"step": 3210
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 2.401193139448173,
|
| 2276 |
+
"grad_norm": 0.6338595567107324,
|
| 2277 |
+
"learning_rate": 5e-06,
|
| 2278 |
+
"loss": 0.4175,
|
| 2279 |
+
"step": 3220
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 2.4086502609992544,
|
| 2283 |
+
"grad_norm": 0.613546741805029,
|
| 2284 |
+
"learning_rate": 5e-06,
|
| 2285 |
+
"loss": 0.4205,
|
| 2286 |
+
"step": 3230
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 2.4161073825503356,
|
| 2290 |
+
"grad_norm": 0.6203881329193865,
|
| 2291 |
+
"learning_rate": 5e-06,
|
| 2292 |
+
"loss": 0.4106,
|
| 2293 |
+
"step": 3240
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 2.423564504101417,
|
| 2297 |
+
"grad_norm": 0.587446398589364,
|
| 2298 |
+
"learning_rate": 5e-06,
|
| 2299 |
+
"loss": 0.435,
|
| 2300 |
+
"step": 3250
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 2.431021625652498,
|
| 2304 |
+
"grad_norm": 0.6180935141585874,
|
| 2305 |
+
"learning_rate": 5e-06,
|
| 2306 |
+
"loss": 0.4148,
|
| 2307 |
+
"step": 3260
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 2.4384787472035794,
|
| 2311 |
+
"grad_norm": 0.5966462669608655,
|
| 2312 |
+
"learning_rate": 5e-06,
|
| 2313 |
+
"loss": 0.4071,
|
| 2314 |
+
"step": 3270
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 2.4459358687546606,
|
| 2318 |
+
"grad_norm": 0.6318483897236675,
|
| 2319 |
+
"learning_rate": 5e-06,
|
| 2320 |
+
"loss": 0.4222,
|
| 2321 |
+
"step": 3280
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 2.453392990305742,
|
| 2325 |
+
"grad_norm": 0.6445808903137813,
|
| 2326 |
+
"learning_rate": 5e-06,
|
| 2327 |
+
"loss": 0.4152,
|
| 2328 |
+
"step": 3290
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 2.460850111856823,
|
| 2332 |
+
"grad_norm": 0.6047687718392507,
|
| 2333 |
+
"learning_rate": 5e-06,
|
| 2334 |
+
"loss": 0.4051,
|
| 2335 |
+
"step": 3300
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 2.4683072334079044,
|
| 2339 |
+
"grad_norm": 0.6113185157541442,
|
| 2340 |
+
"learning_rate": 5e-06,
|
| 2341 |
+
"loss": 0.4221,
|
| 2342 |
+
"step": 3310
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 2.4757643549589856,
|
| 2346 |
+
"grad_norm": 0.5912448445164308,
|
| 2347 |
+
"learning_rate": 5e-06,
|
| 2348 |
+
"loss": 0.4116,
|
| 2349 |
+
"step": 3320
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 2.4832214765100673,
|
| 2353 |
+
"grad_norm": 0.6710834935010201,
|
| 2354 |
+
"learning_rate": 5e-06,
|
| 2355 |
+
"loss": 0.4211,
|
| 2356 |
+
"step": 3330
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 2.4906785980611486,
|
| 2360 |
+
"grad_norm": 0.5855123869668963,
|
| 2361 |
+
"learning_rate": 5e-06,
|
| 2362 |
+
"loss": 0.4176,
|
| 2363 |
+
"step": 3340
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 2.49813571961223,
|
| 2367 |
+
"grad_norm": 0.5672278638197414,
|
| 2368 |
+
"learning_rate": 5e-06,
|
| 2369 |
+
"loss": 0.4283,
|
| 2370 |
+
"step": 3350
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 2.505592841163311,
|
| 2374 |
+
"grad_norm": 0.6623637796001947,
|
| 2375 |
+
"learning_rate": 5e-06,
|
| 2376 |
+
"loss": 0.4207,
|
| 2377 |
+
"step": 3360
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 2.5130499627143923,
|
| 2381 |
+
"grad_norm": 0.5429235298023167,
|
| 2382 |
+
"learning_rate": 5e-06,
|
| 2383 |
+
"loss": 0.4239,
|
| 2384 |
+
"step": 3370
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 2.5205070842654735,
|
| 2388 |
+
"grad_norm": 0.5553083911368453,
|
| 2389 |
+
"learning_rate": 5e-06,
|
| 2390 |
+
"loss": 0.4266,
|
| 2391 |
+
"step": 3380
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 2.527964205816555,
|
| 2395 |
+
"grad_norm": 0.5876357465241916,
|
| 2396 |
+
"learning_rate": 5e-06,
|
| 2397 |
+
"loss": 0.4262,
|
| 2398 |
+
"step": 3390
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 2.535421327367636,
|
| 2402 |
+
"grad_norm": 0.6222644485385007,
|
| 2403 |
+
"learning_rate": 5e-06,
|
| 2404 |
+
"loss": 0.4082,
|
| 2405 |
+
"step": 3400
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 2.5428784489187173,
|
| 2409 |
+
"grad_norm": 0.6327192739316518,
|
| 2410 |
+
"learning_rate": 5e-06,
|
| 2411 |
+
"loss": 0.4123,
|
| 2412 |
+
"step": 3410
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 2.5503355704697985,
|
| 2416 |
+
"grad_norm": 0.5833750402335252,
|
| 2417 |
+
"learning_rate": 5e-06,
|
| 2418 |
+
"loss": 0.4177,
|
| 2419 |
+
"step": 3420
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 2.5577926920208798,
|
| 2423 |
+
"grad_norm": 0.642842699957225,
|
| 2424 |
+
"learning_rate": 5e-06,
|
| 2425 |
+
"loss": 0.4238,
|
| 2426 |
+
"step": 3430
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 2.5652498135719615,
|
| 2430 |
+
"grad_norm": 0.6050955007783061,
|
| 2431 |
+
"learning_rate": 5e-06,
|
| 2432 |
+
"loss": 0.4333,
|
| 2433 |
+
"step": 3440
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 2.5727069351230423,
|
| 2437 |
+
"grad_norm": 0.616598274296138,
|
| 2438 |
+
"learning_rate": 5e-06,
|
| 2439 |
+
"loss": 0.4253,
|
| 2440 |
+
"step": 3450
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 2.580164056674124,
|
| 2444 |
+
"grad_norm": 0.6524589708431916,
|
| 2445 |
+
"learning_rate": 5e-06,
|
| 2446 |
+
"loss": 0.4249,
|
| 2447 |
+
"step": 3460
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 2.587621178225205,
|
| 2451 |
+
"grad_norm": 0.5751310148688968,
|
| 2452 |
+
"learning_rate": 5e-06,
|
| 2453 |
+
"loss": 0.4177,
|
| 2454 |
+
"step": 3470
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 2.5950782997762865,
|
| 2458 |
+
"grad_norm": 0.627879388176911,
|
| 2459 |
+
"learning_rate": 5e-06,
|
| 2460 |
+
"loss": 0.4161,
|
| 2461 |
+
"step": 3480
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 2.6025354213273677,
|
| 2465 |
+
"grad_norm": 0.6201429646996841,
|
| 2466 |
+
"learning_rate": 5e-06,
|
| 2467 |
+
"loss": 0.4163,
|
| 2468 |
+
"step": 3490
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 2.609992542878449,
|
| 2472 |
+
"grad_norm": 0.5820543874841265,
|
| 2473 |
+
"learning_rate": 5e-06,
|
| 2474 |
+
"loss": 0.4186,
|
| 2475 |
+
"step": 3500
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 2.61744966442953,
|
| 2479 |
+
"grad_norm": 0.6017708715046648,
|
| 2480 |
+
"learning_rate": 5e-06,
|
| 2481 |
+
"loss": 0.428,
|
| 2482 |
+
"step": 3510
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 2.6249067859806114,
|
| 2486 |
+
"grad_norm": 0.6272846852335328,
|
| 2487 |
+
"learning_rate": 5e-06,
|
| 2488 |
+
"loss": 0.4171,
|
| 2489 |
+
"step": 3520
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 2.6323639075316927,
|
| 2493 |
+
"grad_norm": 0.6171944158126793,
|
| 2494 |
+
"learning_rate": 5e-06,
|
| 2495 |
+
"loss": 0.4208,
|
| 2496 |
+
"step": 3530
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 2.639821029082774,
|
| 2500 |
+
"grad_norm": 0.5757983007676734,
|
| 2501 |
+
"learning_rate": 5e-06,
|
| 2502 |
+
"loss": 0.4232,
|
| 2503 |
+
"step": 3540
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 2.647278150633855,
|
| 2507 |
+
"grad_norm": 0.615459114965357,
|
| 2508 |
+
"learning_rate": 5e-06,
|
| 2509 |
+
"loss": 0.4145,
|
| 2510 |
+
"step": 3550
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 2.6547352721849364,
|
| 2514 |
+
"grad_norm": 0.642264448040409,
|
| 2515 |
+
"learning_rate": 5e-06,
|
| 2516 |
+
"loss": 0.428,
|
| 2517 |
+
"step": 3560
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 2.662192393736018,
|
| 2521 |
+
"grad_norm": 0.5989193246950945,
|
| 2522 |
+
"learning_rate": 5e-06,
|
| 2523 |
+
"loss": 0.4271,
|
| 2524 |
+
"step": 3570
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 2.669649515287099,
|
| 2528 |
+
"grad_norm": 0.5516635613380466,
|
| 2529 |
+
"learning_rate": 5e-06,
|
| 2530 |
+
"loss": 0.4153,
|
| 2531 |
+
"step": 3580
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 2.6771066368381806,
|
| 2535 |
+
"grad_norm": 0.6619116578052296,
|
| 2536 |
+
"learning_rate": 5e-06,
|
| 2537 |
+
"loss": 0.4157,
|
| 2538 |
+
"step": 3590
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 2.684563758389262,
|
| 2542 |
+
"grad_norm": 0.6371720904571588,
|
| 2543 |
+
"learning_rate": 5e-06,
|
| 2544 |
+
"loss": 0.427,
|
| 2545 |
+
"step": 3600
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 2.692020879940343,
|
| 2549 |
+
"grad_norm": 0.5757772064087978,
|
| 2550 |
+
"learning_rate": 5e-06,
|
| 2551 |
+
"loss": 0.4141,
|
| 2552 |
+
"step": 3610
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 2.6994780014914244,
|
| 2556 |
+
"grad_norm": 0.6037614796783571,
|
| 2557 |
+
"learning_rate": 5e-06,
|
| 2558 |
+
"loss": 0.4169,
|
| 2559 |
+
"step": 3620
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 2.7069351230425056,
|
| 2563 |
+
"grad_norm": 0.6349534794022433,
|
| 2564 |
+
"learning_rate": 5e-06,
|
| 2565 |
+
"loss": 0.4363,
|
| 2566 |
+
"step": 3630
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 2.714392244593587,
|
| 2570 |
+
"grad_norm": 0.5873004454383765,
|
| 2571 |
+
"learning_rate": 5e-06,
|
| 2572 |
+
"loss": 0.4247,
|
| 2573 |
+
"step": 3640
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 2.721849366144668,
|
| 2577 |
+
"grad_norm": 0.5840849381722206,
|
| 2578 |
+
"learning_rate": 5e-06,
|
| 2579 |
+
"loss": 0.4275,
|
| 2580 |
+
"step": 3650
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 2.7293064876957494,
|
| 2584 |
+
"grad_norm": 0.5723518390413971,
|
| 2585 |
+
"learning_rate": 5e-06,
|
| 2586 |
+
"loss": 0.4294,
|
| 2587 |
+
"step": 3660
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 2.7367636092468306,
|
| 2591 |
+
"grad_norm": 0.5792593078543581,
|
| 2592 |
+
"learning_rate": 5e-06,
|
| 2593 |
+
"loss": 0.4274,
|
| 2594 |
+
"step": 3670
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 2.744220730797912,
|
| 2598 |
+
"grad_norm": 0.5770536517441345,
|
| 2599 |
+
"learning_rate": 5e-06,
|
| 2600 |
+
"loss": 0.4184,
|
| 2601 |
+
"step": 3680
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 2.751677852348993,
|
| 2605 |
+
"grad_norm": 0.5803564642723817,
|
| 2606 |
+
"learning_rate": 5e-06,
|
| 2607 |
+
"loss": 0.4172,
|
| 2608 |
+
"step": 3690
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 2.759134973900075,
|
| 2612 |
+
"grad_norm": 0.610053209093683,
|
| 2613 |
+
"learning_rate": 5e-06,
|
| 2614 |
+
"loss": 0.4189,
|
| 2615 |
+
"step": 3700
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 2.7665920954511556,
|
| 2619 |
+
"grad_norm": 0.6101684327110989,
|
| 2620 |
+
"learning_rate": 5e-06,
|
| 2621 |
+
"loss": 0.4219,
|
| 2622 |
+
"step": 3710
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 2.7740492170022373,
|
| 2626 |
+
"grad_norm": 0.5904638045362951,
|
| 2627 |
+
"learning_rate": 5e-06,
|
| 2628 |
+
"loss": 0.4199,
|
| 2629 |
+
"step": 3720
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 2.7815063385533185,
|
| 2633 |
+
"grad_norm": 0.6123335251469906,
|
| 2634 |
+
"learning_rate": 5e-06,
|
| 2635 |
+
"loss": 0.417,
|
| 2636 |
+
"step": 3730
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 2.7889634601043998,
|
| 2640 |
+
"grad_norm": 0.6433911863159033,
|
| 2641 |
+
"learning_rate": 5e-06,
|
| 2642 |
+
"loss": 0.4247,
|
| 2643 |
+
"step": 3740
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 2.796420581655481,
|
| 2647 |
+
"grad_norm": 0.6031149493474106,
|
| 2648 |
+
"learning_rate": 5e-06,
|
| 2649 |
+
"loss": 0.4185,
|
| 2650 |
+
"step": 3750
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 2.8038777032065623,
|
| 2654 |
+
"grad_norm": 0.5895736114737712,
|
| 2655 |
+
"learning_rate": 5e-06,
|
| 2656 |
+
"loss": 0.4289,
|
| 2657 |
+
"step": 3760
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 2.8113348247576435,
|
| 2661 |
+
"grad_norm": 0.5882027658002406,
|
| 2662 |
+
"learning_rate": 5e-06,
|
| 2663 |
+
"loss": 0.4295,
|
| 2664 |
+
"step": 3770
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 2.8187919463087248,
|
| 2668 |
+
"grad_norm": 0.6697395749462909,
|
| 2669 |
+
"learning_rate": 5e-06,
|
| 2670 |
+
"loss": 0.4278,
|
| 2671 |
+
"step": 3780
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 2.826249067859806,
|
| 2675 |
+
"grad_norm": 0.61486633354577,
|
| 2676 |
+
"learning_rate": 5e-06,
|
| 2677 |
+
"loss": 0.4194,
|
| 2678 |
+
"step": 3790
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 2.8337061894108873,
|
| 2682 |
+
"grad_norm": 0.6062660800570471,
|
| 2683 |
+
"learning_rate": 5e-06,
|
| 2684 |
+
"loss": 0.4191,
|
| 2685 |
+
"step": 3800
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 2.841163310961969,
|
| 2689 |
+
"grad_norm": 0.5823531490175686,
|
| 2690 |
+
"learning_rate": 5e-06,
|
| 2691 |
+
"loss": 0.4249,
|
| 2692 |
+
"step": 3810
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 2.8486204325130497,
|
| 2696 |
+
"grad_norm": 0.6594942240638506,
|
| 2697 |
+
"learning_rate": 5e-06,
|
| 2698 |
+
"loss": 0.4141,
|
| 2699 |
+
"step": 3820
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 2.8560775540641314,
|
| 2703 |
+
"grad_norm": 0.5894447911469776,
|
| 2704 |
+
"learning_rate": 5e-06,
|
| 2705 |
+
"loss": 0.4226,
|
| 2706 |
+
"step": 3830
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 2.8635346756152127,
|
| 2710 |
+
"grad_norm": 0.6257621624046176,
|
| 2711 |
+
"learning_rate": 5e-06,
|
| 2712 |
+
"loss": 0.4237,
|
| 2713 |
+
"step": 3840
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 2.870991797166294,
|
| 2717 |
+
"grad_norm": 0.5507526853748216,
|
| 2718 |
+
"learning_rate": 5e-06,
|
| 2719 |
+
"loss": 0.4247,
|
| 2720 |
+
"step": 3850
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 2.878448918717375,
|
| 2724 |
+
"grad_norm": 0.555588132655637,
|
| 2725 |
+
"learning_rate": 5e-06,
|
| 2726 |
+
"loss": 0.4259,
|
| 2727 |
+
"step": 3860
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 2.8859060402684564,
|
| 2731 |
+
"grad_norm": 0.617068246539021,
|
| 2732 |
+
"learning_rate": 5e-06,
|
| 2733 |
+
"loss": 0.4307,
|
| 2734 |
+
"step": 3870
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 2.8933631618195377,
|
| 2738 |
+
"grad_norm": 0.5957184006665759,
|
| 2739 |
+
"learning_rate": 5e-06,
|
| 2740 |
+
"loss": 0.4302,
|
| 2741 |
+
"step": 3880
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 2.900820283370619,
|
| 2745 |
+
"grad_norm": 0.5821306913096179,
|
| 2746 |
+
"learning_rate": 5e-06,
|
| 2747 |
+
"loss": 0.4197,
|
| 2748 |
+
"step": 3890
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 2.9082774049217,
|
| 2752 |
+
"grad_norm": 0.5948347831059887,
|
| 2753 |
+
"learning_rate": 5e-06,
|
| 2754 |
+
"loss": 0.4196,
|
| 2755 |
+
"step": 3900
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 2.9157345264727814,
|
| 2759 |
+
"grad_norm": 0.5748015527924519,
|
| 2760 |
+
"learning_rate": 5e-06,
|
| 2761 |
+
"loss": 0.4234,
|
| 2762 |
+
"step": 3910
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 2.9231916480238627,
|
| 2766 |
+
"grad_norm": 0.573330338580251,
|
| 2767 |
+
"learning_rate": 5e-06,
|
| 2768 |
+
"loss": 0.4155,
|
| 2769 |
+
"step": 3920
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 2.930648769574944,
|
| 2773 |
+
"grad_norm": 0.5959038101452754,
|
| 2774 |
+
"learning_rate": 5e-06,
|
| 2775 |
+
"loss": 0.4245,
|
| 2776 |
+
"step": 3930
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 2.9381058911260256,
|
| 2780 |
+
"grad_norm": 0.6327154198497399,
|
| 2781 |
+
"learning_rate": 5e-06,
|
| 2782 |
+
"loss": 0.4141,
|
| 2783 |
+
"step": 3940
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 2.9455630126771064,
|
| 2787 |
+
"grad_norm": 0.5879319738064069,
|
| 2788 |
+
"learning_rate": 5e-06,
|
| 2789 |
+
"loss": 0.4111,
|
| 2790 |
+
"step": 3950
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 2.953020134228188,
|
| 2794 |
+
"grad_norm": 0.5850847757389654,
|
| 2795 |
+
"learning_rate": 5e-06,
|
| 2796 |
+
"loss": 0.413,
|
| 2797 |
+
"step": 3960
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 2.9604772557792693,
|
| 2801 |
+
"grad_norm": 0.6440418621171623,
|
| 2802 |
+
"learning_rate": 5e-06,
|
| 2803 |
+
"loss": 0.4254,
|
| 2804 |
+
"step": 3970
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 2.9679343773303506,
|
| 2808 |
+
"grad_norm": 0.6104743322838082,
|
| 2809 |
+
"learning_rate": 5e-06,
|
| 2810 |
+
"loss": 0.4116,
|
| 2811 |
+
"step": 3980
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 2.975391498881432,
|
| 2815 |
+
"grad_norm": 0.5820609145361088,
|
| 2816 |
+
"learning_rate": 5e-06,
|
| 2817 |
+
"loss": 0.4188,
|
| 2818 |
+
"step": 3990
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 2.982848620432513,
|
| 2822 |
+
"grad_norm": 0.6266879443777426,
|
| 2823 |
+
"learning_rate": 5e-06,
|
| 2824 |
+
"loss": 0.4393,
|
| 2825 |
+
"step": 4000
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 2.9903057419835943,
|
| 2829 |
+
"grad_norm": 0.5751703131935023,
|
| 2830 |
+
"learning_rate": 5e-06,
|
| 2831 |
+
"loss": 0.4298,
|
| 2832 |
+
"step": 4010
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 2.9977628635346756,
|
| 2836 |
+
"grad_norm": 0.6279401240117848,
|
| 2837 |
+
"learning_rate": 5e-06,
|
| 2838 |
+
"loss": 0.4292,
|
| 2839 |
+
"step": 4020
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 3.0,
|
| 2843 |
+
"eval_loss": 0.5737335681915283,
|
| 2844 |
+
"eval_runtime": 341.8895,
|
| 2845 |
+
"eval_samples_per_second": 26.418,
|
| 2846 |
+
"eval_steps_per_second": 0.415,
|
| 2847 |
+
"step": 4023
|
| 2848 |
+
},
|
| 2849 |
+
{
|
| 2850 |
+
"epoch": 3.0,
|
| 2851 |
+
"step": 4023,
|
| 2852 |
+
"total_flos": 2108813507297280.0,
|
| 2853 |
+
"train_loss": 0.4856802497430432,
|
| 2854 |
+
"train_runtime": 54595.2546,
|
| 2855 |
+
"train_samples_per_second": 9.429,
|
| 2856 |
+
"train_steps_per_second": 0.074
|
| 2857 |
+
}
|
| 2858 |
+
],
|
| 2859 |
+
"logging_steps": 10,
|
| 2860 |
+
"max_steps": 4023,
|
| 2861 |
+
"num_input_tokens_seen": 0,
|
| 2862 |
+
"num_train_epochs": 3,
|
| 2863 |
+
"save_steps": 500,
|
| 2864 |
+
"stateful_callbacks": {
|
| 2865 |
+
"TrainerControl": {
|
| 2866 |
+
"args": {
|
| 2867 |
+
"should_epoch_stop": false,
|
| 2868 |
+
"should_evaluate": false,
|
| 2869 |
+
"should_log": false,
|
| 2870 |
+
"should_save": true,
|
| 2871 |
+
"should_training_stop": true
|
| 2872 |
+
},
|
| 2873 |
+
"attributes": {}
|
| 2874 |
+
}
|
| 2875 |
+
},
|
| 2876 |
+
"total_flos": 2108813507297280.0,
|
| 2877 |
+
"train_batch_size": 8,
|
| 2878 |
+
"trial_name": null,
|
| 2879 |
+
"trial_params": null
|
| 2880 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|