End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1533 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: OHprompts_GPT4oresponses_30k
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# OHprompts_GPT4oresponses_30k
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: OHprompts_GPT4oresponses_30k
|
|
|
|
| 16 |
|
| 17 |
# OHprompts_GPT4oresponses_30k
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mli-lab/OHprompts_GPT4oresponses_30k dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.986046511627907,
|
| 3 |
+
"total_flos": 1.1419933308302131e+17,
|
| 4 |
+
"train_loss": 0.6310646277917943,
|
| 5 |
+
"train_runtime": 4071.0206,
|
| 6 |
+
"train_samples_per_second": 5.059,
|
| 7 |
+
"train_steps_per_second": 0.052
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.986046511627907,
|
| 3 |
+
"total_flos": 1.1419933308302131e+17,
|
| 4 |
+
"train_loss": 0.6310646277917943,
|
| 5 |
+
"train_runtime": 4071.0206,
|
| 6 |
+
"train_samples_per_second": 5.059,
|
| 7 |
+
"train_steps_per_second": 0.052
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1533 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.986046511627907,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 213,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.013953488372093023,
|
| 13 |
+
"grad_norm": 4.180223031536278,
|
| 14 |
+
"learning_rate": 4.5454545454545457e-07,
|
| 15 |
+
"loss": 0.9256,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.027906976744186046,
|
| 20 |
+
"grad_norm": 4.538483624724508,
|
| 21 |
+
"learning_rate": 9.090909090909091e-07,
|
| 22 |
+
"loss": 0.8742,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04186046511627907,
|
| 27 |
+
"grad_norm": 4.966665379123455,
|
| 28 |
+
"learning_rate": 1.3636363636363636e-06,
|
| 29 |
+
"loss": 0.9453,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.05581395348837209,
|
| 34 |
+
"grad_norm": 4.269989274936151,
|
| 35 |
+
"learning_rate": 1.8181818181818183e-06,
|
| 36 |
+
"loss": 0.868,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.06976744186046512,
|
| 41 |
+
"grad_norm": 4.185160235663758,
|
| 42 |
+
"learning_rate": 2.2727272727272728e-06,
|
| 43 |
+
"loss": 0.8915,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08372093023255814,
|
| 48 |
+
"grad_norm": 3.6421422943394184,
|
| 49 |
+
"learning_rate": 2.7272727272727272e-06,
|
| 50 |
+
"loss": 0.9411,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.09767441860465116,
|
| 55 |
+
"grad_norm": 3.044369769290876,
|
| 56 |
+
"learning_rate": 3.181818181818182e-06,
|
| 57 |
+
"loss": 0.8872,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.11162790697674418,
|
| 62 |
+
"grad_norm": 2.3026052106225516,
|
| 63 |
+
"learning_rate": 3.6363636363636366e-06,
|
| 64 |
+
"loss": 0.8061,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.12558139534883722,
|
| 69 |
+
"grad_norm": 2.36814680746031,
|
| 70 |
+
"learning_rate": 4.0909090909090915e-06,
|
| 71 |
+
"loss": 0.8383,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.13953488372093023,
|
| 76 |
+
"grad_norm": 1.8945442019043834,
|
| 77 |
+
"learning_rate": 4.5454545454545455e-06,
|
| 78 |
+
"loss": 0.7922,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.15348837209302327,
|
| 83 |
+
"grad_norm": 1.843510350188421,
|
| 84 |
+
"learning_rate": 5e-06,
|
| 85 |
+
"loss": 0.8003,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.16744186046511628,
|
| 90 |
+
"grad_norm": 1.4520922792930948,
|
| 91 |
+
"learning_rate": 5.4545454545454545e-06,
|
| 92 |
+
"loss": 0.7304,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1813953488372093,
|
| 97 |
+
"grad_norm": 1.385363919907977,
|
| 98 |
+
"learning_rate": 5.90909090909091e-06,
|
| 99 |
+
"loss": 0.7748,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.19534883720930232,
|
| 104 |
+
"grad_norm": 1.0509763410634785,
|
| 105 |
+
"learning_rate": 6.363636363636364e-06,
|
| 106 |
+
"loss": 0.7428,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.20930232558139536,
|
| 111 |
+
"grad_norm": 1.0039237549101678,
|
| 112 |
+
"learning_rate": 6.818181818181818e-06,
|
| 113 |
+
"loss": 0.7535,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.22325581395348837,
|
| 118 |
+
"grad_norm": 0.8932008204248008,
|
| 119 |
+
"learning_rate": 7.272727272727273e-06,
|
| 120 |
+
"loss": 0.7221,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2372093023255814,
|
| 125 |
+
"grad_norm": 1.0256660986105275,
|
| 126 |
+
"learning_rate": 7.727272727272727e-06,
|
| 127 |
+
"loss": 0.7461,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25116279069767444,
|
| 132 |
+
"grad_norm": 1.042115817076506,
|
| 133 |
+
"learning_rate": 8.181818181818183e-06,
|
| 134 |
+
"loss": 0.732,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2651162790697674,
|
| 139 |
+
"grad_norm": 0.8800070105054354,
|
| 140 |
+
"learning_rate": 8.636363636363637e-06,
|
| 141 |
+
"loss": 0.7798,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.27906976744186046,
|
| 146 |
+
"grad_norm": 0.9186101656788263,
|
| 147 |
+
"learning_rate": 9.090909090909091e-06,
|
| 148 |
+
"loss": 0.7589,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2930232558139535,
|
| 153 |
+
"grad_norm": 0.8981856877871544,
|
| 154 |
+
"learning_rate": 9.545454545454547e-06,
|
| 155 |
+
"loss": 0.761,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.30697674418604654,
|
| 160 |
+
"grad_norm": 0.9442233006086749,
|
| 161 |
+
"learning_rate": 1e-05,
|
| 162 |
+
"loss": 0.7809,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3209302325581395,
|
| 167 |
+
"grad_norm": 0.8564785752559885,
|
| 168 |
+
"learning_rate": 9.999323662872998e-06,
|
| 169 |
+
"loss": 0.7985,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.33488372093023255,
|
| 174 |
+
"grad_norm": 0.8572582575558093,
|
| 175 |
+
"learning_rate": 9.99729483446475e-06,
|
| 176 |
+
"loss": 0.7573,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3488372093023256,
|
| 181 |
+
"grad_norm": 0.8121984298438342,
|
| 182 |
+
"learning_rate": 9.993914063644053e-06,
|
| 183 |
+
"loss": 0.7502,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.3627906976744186,
|
| 188 |
+
"grad_norm": 0.9475921551407652,
|
| 189 |
+
"learning_rate": 9.989182265027232e-06,
|
| 190 |
+
"loss": 0.7526,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3767441860465116,
|
| 195 |
+
"grad_norm": 0.7787851193825684,
|
| 196 |
+
"learning_rate": 9.98310071873072e-06,
|
| 197 |
+
"loss": 0.7138,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.39069767441860465,
|
| 202 |
+
"grad_norm": 0.8448946493089139,
|
| 203 |
+
"learning_rate": 9.975671070024741e-06,
|
| 204 |
+
"loss": 0.7801,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.4046511627906977,
|
| 209 |
+
"grad_norm": 0.8745787585831062,
|
| 210 |
+
"learning_rate": 9.966895328888195e-06,
|
| 211 |
+
"loss": 0.7315,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.4186046511627907,
|
| 216 |
+
"grad_norm": 0.8154328510100463,
|
| 217 |
+
"learning_rate": 9.956775869464901e-06,
|
| 218 |
+
"loss": 0.782,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.4325581395348837,
|
| 223 |
+
"grad_norm": 0.8123407449935709,
|
| 224 |
+
"learning_rate": 9.945315429421307e-06,
|
| 225 |
+
"loss": 0.7038,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.44651162790697674,
|
| 230 |
+
"grad_norm": 0.8368248745687986,
|
| 231 |
+
"learning_rate": 9.932517109205849e-06,
|
| 232 |
+
"loss": 0.7554,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.4604651162790698,
|
| 237 |
+
"grad_norm": 0.8324720480538125,
|
| 238 |
+
"learning_rate": 9.918384371210178e-06,
|
| 239 |
+
"loss": 0.7279,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.4744186046511628,
|
| 244 |
+
"grad_norm": 0.7542843937958372,
|
| 245 |
+
"learning_rate": 9.902921038832456e-06,
|
| 246 |
+
"loss": 0.8063,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.4883720930232558,
|
| 251 |
+
"grad_norm": 0.7016852905312764,
|
| 252 |
+
"learning_rate": 9.886131295443003e-06,
|
| 253 |
+
"loss": 0.7298,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5023255813953489,
|
| 258 |
+
"grad_norm": 0.7473857827397101,
|
| 259 |
+
"learning_rate": 9.868019683252543e-06,
|
| 260 |
+
"loss": 0.7411,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.5162790697674419,
|
| 265 |
+
"grad_norm": 0.8528258992946952,
|
| 266 |
+
"learning_rate": 9.848591102083375e-06,
|
| 267 |
+
"loss": 0.7173,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.5302325581395348,
|
| 272 |
+
"grad_norm": 0.8237901171986899,
|
| 273 |
+
"learning_rate": 9.82785080804381e-06,
|
| 274 |
+
"loss": 0.7054,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.5441860465116279,
|
| 279 |
+
"grad_norm": 0.8417001155591349,
|
| 280 |
+
"learning_rate": 9.805804412106197e-06,
|
| 281 |
+
"loss": 0.7231,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.5581395348837209,
|
| 286 |
+
"grad_norm": 0.7688769030864758,
|
| 287 |
+
"learning_rate": 9.782457878588977e-06,
|
| 288 |
+
"loss": 0.7644,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.5720930232558139,
|
| 293 |
+
"grad_norm": 0.7926630801835173,
|
| 294 |
+
"learning_rate": 9.75781752354311e-06,
|
| 295 |
+
"loss": 0.8192,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.586046511627907,
|
| 300 |
+
"grad_norm": 0.7753664995622703,
|
| 301 |
+
"learning_rate": 9.731890013043367e-06,
|
| 302 |
+
"loss": 0.6907,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.6,
|
| 307 |
+
"grad_norm": 0.794269792132402,
|
| 308 |
+
"learning_rate": 9.704682361384941e-06,
|
| 309 |
+
"loss": 0.7647,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.6139534883720931,
|
| 314 |
+
"grad_norm": 0.7415044200607281,
|
| 315 |
+
"learning_rate": 9.676201929185809e-06,
|
| 316 |
+
"loss": 0.7076,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.627906976744186,
|
| 321 |
+
"grad_norm": 0.7385780791151161,
|
| 322 |
+
"learning_rate": 9.646456421395447e-06,
|
| 323 |
+
"loss": 0.7175,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.641860465116279,
|
| 328 |
+
"grad_norm": 0.7003336741400412,
|
| 329 |
+
"learning_rate": 9.615453885210368e-06,
|
| 330 |
+
"loss": 0.738,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.6558139534883721,
|
| 335 |
+
"grad_norm": 0.7724918826496796,
|
| 336 |
+
"learning_rate": 9.583202707897075e-06,
|
| 337 |
+
"loss": 0.7614,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.6697674418604651,
|
| 342 |
+
"grad_norm": 0.7620497454219544,
|
| 343 |
+
"learning_rate": 9.549711614523007e-06,
|
| 344 |
+
"loss": 0.734,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.6837209302325581,
|
| 349 |
+
"grad_norm": 0.7500001225387097,
|
| 350 |
+
"learning_rate": 9.514989665596114e-06,
|
| 351 |
+
"loss": 0.7485,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.6976744186046512,
|
| 356 |
+
"grad_norm": 0.7787354702551742,
|
| 357 |
+
"learning_rate": 9.479046254613673e-06,
|
| 358 |
+
"loss": 0.7746,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.7116279069767442,
|
| 363 |
+
"grad_norm": 0.6950977978192039,
|
| 364 |
+
"learning_rate": 9.441891105521005e-06,
|
| 365 |
+
"loss": 0.7041,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.7255813953488373,
|
| 370 |
+
"grad_norm": 0.7539402070090733,
|
| 371 |
+
"learning_rate": 9.40353427008083e-06,
|
| 372 |
+
"loss": 0.7673,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.7395348837209302,
|
| 377 |
+
"grad_norm": 0.7260757420379652,
|
| 378 |
+
"learning_rate": 9.3639861251539e-06,
|
| 379 |
+
"loss": 0.731,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.7534883720930232,
|
| 384 |
+
"grad_norm": 0.7310930851217866,
|
| 385 |
+
"learning_rate": 9.323257369891702e-06,
|
| 386 |
+
"loss": 0.7278,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.7674418604651163,
|
| 391 |
+
"grad_norm": 0.7202232307933061,
|
| 392 |
+
"learning_rate": 9.281359022841966e-06,
|
| 393 |
+
"loss": 0.7308,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.7813953488372093,
|
| 398 |
+
"grad_norm": 0.774608059002569,
|
| 399 |
+
"learning_rate": 9.238302418967757e-06,
|
| 400 |
+
"loss": 0.748,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.7953488372093023,
|
| 405 |
+
"grad_norm": 0.7155020729739833,
|
| 406 |
+
"learning_rate": 9.194099206580981e-06,
|
| 407 |
+
"loss": 0.7142,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.8093023255813954,
|
| 412 |
+
"grad_norm": 0.705738515219601,
|
| 413 |
+
"learning_rate": 9.14876134419111e-06,
|
| 414 |
+
"loss": 0.7387,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.8232558139534883,
|
| 419 |
+
"grad_norm": 0.726638534200724,
|
| 420 |
+
"learning_rate": 9.102301097269974e-06,
|
| 421 |
+
"loss": 0.6992,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.8372093023255814,
|
| 426 |
+
"grad_norm": 0.7630628036771007,
|
| 427 |
+
"learning_rate": 9.05473103493355e-06,
|
| 428 |
+
"loss": 0.7535,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.8511627906976744,
|
| 433 |
+
"grad_norm": 0.7219220998253707,
|
| 434 |
+
"learning_rate": 9.006064026541549e-06,
|
| 435 |
+
"loss": 0.7062,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.8651162790697674,
|
| 440 |
+
"grad_norm": 0.7504793745970053,
|
| 441 |
+
"learning_rate": 8.956313238215824e-06,
|
| 442 |
+
"loss": 0.7332,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.8790697674418605,
|
| 447 |
+
"grad_norm": 0.7478848773737441,
|
| 448 |
+
"learning_rate": 8.905492129278478e-06,
|
| 449 |
+
"loss": 0.744,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.8930232558139535,
|
| 454 |
+
"grad_norm": 0.7393594785059273,
|
| 455 |
+
"learning_rate": 8.85361444861063e-06,
|
| 456 |
+
"loss": 0.7747,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.9069767441860465,
|
| 461 |
+
"grad_norm": 0.7677319315899734,
|
| 462 |
+
"learning_rate": 8.800694230932885e-06,
|
| 463 |
+
"loss": 0.7763,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.9209302325581395,
|
| 468 |
+
"grad_norm": 0.6968681433979783,
|
| 469 |
+
"learning_rate": 8.74674579300843e-06,
|
| 470 |
+
"loss": 0.7227,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.9348837209302325,
|
| 475 |
+
"grad_norm": 0.7515716718812465,
|
| 476 |
+
"learning_rate": 8.691783729769874e-06,
|
| 477 |
+
"loss": 0.7672,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.9488372093023256,
|
| 482 |
+
"grad_norm": 0.7070436887476904,
|
| 483 |
+
"learning_rate": 8.635822910370793e-06,
|
| 484 |
+
"loss": 0.7327,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.9627906976744186,
|
| 489 |
+
"grad_norm": 0.7188471720101923,
|
| 490 |
+
"learning_rate": 8.578878474163115e-06,
|
| 491 |
+
"loss": 0.7698,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.9767441860465116,
|
| 496 |
+
"grad_norm": 0.7153844950556374,
|
| 497 |
+
"learning_rate": 8.520965826601394e-06,
|
| 498 |
+
"loss": 0.7235,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.9906976744186047,
|
| 503 |
+
"grad_norm": 0.6866954166606222,
|
| 504 |
+
"learning_rate": 8.462100635075097e-06,
|
| 505 |
+
"loss": 0.7419,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.0116279069767442,
|
| 510 |
+
"grad_norm": 0.7525581036379952,
|
| 511 |
+
"learning_rate": 8.40229882467003e-06,
|
| 512 |
+
"loss": 0.6483,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.0255813953488373,
|
| 517 |
+
"grad_norm": 0.7612245277051721,
|
| 518 |
+
"learning_rate": 8.341576573860049e-06,
|
| 519 |
+
"loss": 0.6092,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.0395348837209302,
|
| 524 |
+
"grad_norm": 0.7476520856673246,
|
| 525 |
+
"learning_rate": 8.279950310130218e-06,
|
| 526 |
+
"loss": 0.6285,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.0534883720930233,
|
| 531 |
+
"grad_norm": 0.7213048511300673,
|
| 532 |
+
"learning_rate": 8.2174367055326e-06,
|
| 533 |
+
"loss": 0.6253,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.0674418604651164,
|
| 538 |
+
"grad_norm": 0.7060965908686724,
|
| 539 |
+
"learning_rate": 8.154052672175888e-06,
|
| 540 |
+
"loss": 0.6487,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.0813953488372092,
|
| 545 |
+
"grad_norm": 0.7359501279300943,
|
| 546 |
+
"learning_rate": 8.08981535765009e-06,
|
| 547 |
+
"loss": 0.6488,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.0953488372093023,
|
| 552 |
+
"grad_norm": 0.7471600782679423,
|
| 553 |
+
"learning_rate": 8.024742140387506e-06,
|
| 554 |
+
"loss": 0.6245,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.1093023255813954,
|
| 559 |
+
"grad_norm": 1.3000015372048261,
|
| 560 |
+
"learning_rate": 7.95885062496126e-06,
|
| 561 |
+
"loss": 0.6122,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.1232558139534883,
|
| 566 |
+
"grad_norm": 0.6835960202006859,
|
| 567 |
+
"learning_rate": 7.892158637322647e-06,
|
| 568 |
+
"loss": 0.6513,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.1372093023255814,
|
| 573 |
+
"grad_norm": 0.7131678249087022,
|
| 574 |
+
"learning_rate": 7.824684219978591e-06,
|
| 575 |
+
"loss": 0.6119,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.1511627906976745,
|
| 580 |
+
"grad_norm": 0.7023779935518183,
|
| 581 |
+
"learning_rate": 7.756445627110523e-06,
|
| 582 |
+
"loss": 0.5517,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.1651162790697676,
|
| 587 |
+
"grad_norm": 0.7320694358005044,
|
| 588 |
+
"learning_rate": 7.68746131963598e-06,
|
| 589 |
+
"loss": 0.6272,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.1790697674418604,
|
| 594 |
+
"grad_norm": 0.6891185877199024,
|
| 595 |
+
"learning_rate": 7.6177499602143e-06,
|
| 596 |
+
"loss": 0.623,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.1930232558139535,
|
| 601 |
+
"grad_norm": 0.6894582716338596,
|
| 602 |
+
"learning_rate": 7.547330408197695e-06,
|
| 603 |
+
"loss": 0.5828,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.2069767441860466,
|
| 608 |
+
"grad_norm": 0.690393431880645,
|
| 609 |
+
"learning_rate": 7.476221714529167e-06,
|
| 610 |
+
"loss": 0.6147,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.2209302325581395,
|
| 615 |
+
"grad_norm": 0.6701171413014578,
|
| 616 |
+
"learning_rate": 7.404443116588548e-06,
|
| 617 |
+
"loss": 0.6194,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.2348837209302326,
|
| 622 |
+
"grad_norm": 0.7111277733561969,
|
| 623 |
+
"learning_rate": 7.332014032988123e-06,
|
| 624 |
+
"loss": 0.5956,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.2488372093023257,
|
| 629 |
+
"grad_norm": 0.6544773295433196,
|
| 630 |
+
"learning_rate": 7.2589540583192165e-06,
|
| 631 |
+
"loss": 0.5981,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.2627906976744185,
|
| 636 |
+
"grad_norm": 0.6514675966676534,
|
| 637 |
+
"learning_rate": 7.185282957851175e-06,
|
| 638 |
+
"loss": 0.5996,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.2767441860465116,
|
| 643 |
+
"grad_norm": 0.6938285046387257,
|
| 644 |
+
"learning_rate": 7.111020662184174e-06,
|
| 645 |
+
"loss": 0.5923,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.2906976744186047,
|
| 650 |
+
"grad_norm": 0.6870448835381137,
|
| 651 |
+
"learning_rate": 7.036187261857289e-06,
|
| 652 |
+
"loss": 0.6124,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.3046511627906976,
|
| 657 |
+
"grad_norm": 0.680814272690559,
|
| 658 |
+
"learning_rate": 6.960803001913315e-06,
|
| 659 |
+
"loss": 0.5983,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.3186046511627907,
|
| 664 |
+
"grad_norm": 0.6748565705881445,
|
| 665 |
+
"learning_rate": 6.884888276421766e-06,
|
| 666 |
+
"loss": 0.611,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.3325581395348838,
|
| 671 |
+
"grad_norm": 0.6676791066748962,
|
| 672 |
+
"learning_rate": 6.8084636229615786e-06,
|
| 673 |
+
"loss": 0.5924,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.3465116279069766,
|
| 678 |
+
"grad_norm": 0.6901277126709034,
|
| 679 |
+
"learning_rate": 6.731549717064975e-06,
|
| 680 |
+
"loss": 0.6097,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.3604651162790697,
|
| 685 |
+
"grad_norm": 0.6992819720098142,
|
| 686 |
+
"learning_rate": 6.654167366624009e-06,
|
| 687 |
+
"loss": 0.672,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.3744186046511628,
|
| 692 |
+
"grad_norm": 0.7824503034259802,
|
| 693 |
+
"learning_rate": 6.576337506261314e-06,
|
| 694 |
+
"loss": 0.627,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.3883720930232557,
|
| 699 |
+
"grad_norm": 0.6982282289779738,
|
| 700 |
+
"learning_rate": 6.498081191666549e-06,
|
| 701 |
+
"loss": 0.6428,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.4023255813953488,
|
| 706 |
+
"grad_norm": 0.7592607065836681,
|
| 707 |
+
"learning_rate": 6.419419593900109e-06,
|
| 708 |
+
"loss": 0.635,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.416279069767442,
|
| 713 |
+
"grad_norm": 0.6839742774370206,
|
| 714 |
+
"learning_rate": 6.340373993665607e-06,
|
| 715 |
+
"loss": 0.5776,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.4302325581395348,
|
| 720 |
+
"grad_norm": 0.6731904047016143,
|
| 721 |
+
"learning_rate": 6.2609657755527135e-06,
|
| 722 |
+
"loss": 0.5901,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.4441860465116279,
|
| 727 |
+
"grad_norm": 0.69395359196692,
|
| 728 |
+
"learning_rate": 6.1812164222518626e-06,
|
| 729 |
+
"loss": 0.6094,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.458139534883721,
|
| 734 |
+
"grad_norm": 0.7047360888181433,
|
| 735 |
+
"learning_rate": 6.101147508742456e-06,
|
| 736 |
+
"loss": 0.5819,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.4720930232558138,
|
| 741 |
+
"grad_norm": 0.6583866543259697,
|
| 742 |
+
"learning_rate": 6.020780696456059e-06,
|
| 743 |
+
"loss": 0.5687,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.486046511627907,
|
| 748 |
+
"grad_norm": 0.6606060716042438,
|
| 749 |
+
"learning_rate": 5.940137727416247e-06,
|
| 750 |
+
"loss": 0.6403,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.5,
|
| 755 |
+
"grad_norm": 0.698057421699006,
|
| 756 |
+
"learning_rate": 5.859240418356614e-06,
|
| 757 |
+
"loss": 0.6268,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.5139534883720929,
|
| 762 |
+
"grad_norm": 0.6893661264785836,
|
| 763 |
+
"learning_rate": 5.778110654818602e-06,
|
| 764 |
+
"loss": 0.5974,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.5279069767441862,
|
| 769 |
+
"grad_norm": 0.6925131573803104,
|
| 770 |
+
"learning_rate": 5.696770385230679e-06,
|
| 771 |
+
"loss": 0.5927,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.541860465116279,
|
| 776 |
+
"grad_norm": 0.6245337132253824,
|
| 777 |
+
"learning_rate": 5.615241614970546e-06,
|
| 778 |
+
"loss": 0.5954,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.5558139534883721,
|
| 783 |
+
"grad_norm": 0.6978126500137894,
|
| 784 |
+
"learning_rate": 5.533546400411899e-06,
|
| 785 |
+
"loss": 0.6293,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.5697674418604652,
|
| 790 |
+
"grad_norm": 0.6685702737440643,
|
| 791 |
+
"learning_rate": 5.4517068429574215e-06,
|
| 792 |
+
"loss": 0.5795,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.583720930232558,
|
| 797 |
+
"grad_norm": 0.6886912117050134,
|
| 798 |
+
"learning_rate": 5.3697450830595775e-06,
|
| 799 |
+
"loss": 0.5908,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.5976744186046512,
|
| 804 |
+
"grad_norm": 0.7026779990898988,
|
| 805 |
+
"learning_rate": 5.287683294230855e-06,
|
| 806 |
+
"loss": 0.6626,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.6116279069767443,
|
| 811 |
+
"grad_norm": 0.6541857661723955,
|
| 812 |
+
"learning_rate": 5.20554367704505e-06,
|
| 813 |
+
"loss": 0.5973,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.6255813953488372,
|
| 818 |
+
"grad_norm": 0.6864213567300637,
|
| 819 |
+
"learning_rate": 5.123348453131242e-06,
|
| 820 |
+
"loss": 0.6153,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.6395348837209303,
|
| 825 |
+
"grad_norm": 0.6996659842567946,
|
| 826 |
+
"learning_rate": 5.041119859162068e-06,
|
| 827 |
+
"loss": 0.6875,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.6534883720930234,
|
| 832 |
+
"grad_norm": 0.6559276582075406,
|
| 833 |
+
"learning_rate": 4.958880140837934e-06,
|
| 834 |
+
"loss": 0.628,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.6674418604651162,
|
| 839 |
+
"grad_norm": 0.6759107040741619,
|
| 840 |
+
"learning_rate": 4.876651546868759e-06,
|
| 841 |
+
"loss": 0.6331,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.6813953488372093,
|
| 846 |
+
"grad_norm": 0.6245367486570701,
|
| 847 |
+
"learning_rate": 4.794456322954953e-06,
|
| 848 |
+
"loss": 0.5749,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.6953488372093024,
|
| 853 |
+
"grad_norm": 0.6755673087829257,
|
| 854 |
+
"learning_rate": 4.7123167057691446e-06,
|
| 855 |
+
"loss": 0.6658,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.7093023255813953,
|
| 860 |
+
"grad_norm": 0.6616484608797307,
|
| 861 |
+
"learning_rate": 4.630254916940424e-06,
|
| 862 |
+
"loss": 0.6109,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.7232558139534884,
|
| 867 |
+
"grad_norm": 0.6819020264081828,
|
| 868 |
+
"learning_rate": 4.548293157042581e-06,
|
| 869 |
+
"loss": 0.6293,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.7372093023255815,
|
| 874 |
+
"grad_norm": 0.7243396897916824,
|
| 875 |
+
"learning_rate": 4.466453599588103e-06,
|
| 876 |
+
"loss": 0.6573,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.7511627906976743,
|
| 881 |
+
"grad_norm": 0.6660691959584282,
|
| 882 |
+
"learning_rate": 4.384758385029457e-06,
|
| 883 |
+
"loss": 0.5964,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.7651162790697674,
|
| 888 |
+
"grad_norm": 0.6893445869396398,
|
| 889 |
+
"learning_rate": 4.3032296147693225e-06,
|
| 890 |
+
"loss": 0.6047,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.7790697674418605,
|
| 895 |
+
"grad_norm": 0.6459987584057402,
|
| 896 |
+
"learning_rate": 4.2218893451814005e-06,
|
| 897 |
+
"loss": 0.6016,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.7930232558139534,
|
| 902 |
+
"grad_norm": 0.702552431682566,
|
| 903 |
+
"learning_rate": 4.140759581643386e-06,
|
| 904 |
+
"loss": 0.6244,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.8069767441860465,
|
| 909 |
+
"grad_norm": 0.6598883080299031,
|
| 910 |
+
"learning_rate": 4.059862272583755e-06,
|
| 911 |
+
"loss": 0.6297,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.8209302325581396,
|
| 916 |
+
"grad_norm": 0.6711511252086539,
|
| 917 |
+
"learning_rate": 3.979219303543942e-06,
|
| 918 |
+
"loss": 0.5812,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.8348837209302324,
|
| 923 |
+
"grad_norm": 0.7110881746480423,
|
| 924 |
+
"learning_rate": 3.898852491257547e-06,
|
| 925 |
+
"loss": 0.6068,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.8488372093023255,
|
| 930 |
+
"grad_norm": 0.925220231394644,
|
| 931 |
+
"learning_rate": 3.818783577748138e-06,
|
| 932 |
+
"loss": 0.6044,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.8627906976744186,
|
| 937 |
+
"grad_norm": 0.6512865495463271,
|
| 938 |
+
"learning_rate": 3.7390342244472886e-06,
|
| 939 |
+
"loss": 0.602,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.8767441860465115,
|
| 944 |
+
"grad_norm": 0.6427098152963066,
|
| 945 |
+
"learning_rate": 3.659626006334395e-06,
|
| 946 |
+
"loss": 0.6387,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.8906976744186048,
|
| 951 |
+
"grad_norm": 0.6789356766533332,
|
| 952 |
+
"learning_rate": 3.5805804060998926e-06,
|
| 953 |
+
"loss": 0.5867,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.9046511627906977,
|
| 958 |
+
"grad_norm": 0.6565930246745768,
|
| 959 |
+
"learning_rate": 3.501918808333453e-06,
|
| 960 |
+
"loss": 0.5762,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.9186046511627906,
|
| 965 |
+
"grad_norm": 0.6741456111762378,
|
| 966 |
+
"learning_rate": 3.4236624937386874e-06,
|
| 967 |
+
"loss": 0.6248,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.9325581395348839,
|
| 972 |
+
"grad_norm": 0.611292878790001,
|
| 973 |
+
"learning_rate": 3.3458326333759927e-06,
|
| 974 |
+
"loss": 0.5426,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.9465116279069767,
|
| 979 |
+
"grad_norm": 0.6799947648906923,
|
| 980 |
+
"learning_rate": 3.268450282935026e-06,
|
| 981 |
+
"loss": 0.6036,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.9604651162790696,
|
| 986 |
+
"grad_norm": 0.6551573645826078,
|
| 987 |
+
"learning_rate": 3.1915363770384223e-06,
|
| 988 |
+
"loss": 0.5862,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.974418604651163,
|
| 993 |
+
"grad_norm": 0.6852118798429588,
|
| 994 |
+
"learning_rate": 3.1151117235782346e-06,
|
| 995 |
+
"loss": 0.6466,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.9883720930232558,
|
| 1000 |
+
"grad_norm": 0.6528770811219057,
|
| 1001 |
+
"learning_rate": 3.0391969980866874e-06,
|
| 1002 |
+
"loss": 0.6208,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.0093023255813955,
|
| 1007 |
+
"grad_norm": 0.6456183774964656,
|
| 1008 |
+
"learning_rate": 2.963812738142713e-06,
|
| 1009 |
+
"loss": 0.5486,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.0232558139534884,
|
| 1014 |
+
"grad_norm": 0.6957232768775263,
|
| 1015 |
+
"learning_rate": 2.8889793378158284e-06,
|
| 1016 |
+
"loss": 0.5367,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.0372093023255813,
|
| 1021 |
+
"grad_norm": 0.7308500470080568,
|
| 1022 |
+
"learning_rate": 2.814717042148827e-06,
|
| 1023 |
+
"loss": 0.4895,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.0511627906976746,
|
| 1028 |
+
"grad_norm": 0.6512065562009933,
|
| 1029 |
+
"learning_rate": 2.7410459416807856e-06,
|
| 1030 |
+
"loss": 0.5331,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.0651162790697675,
|
| 1035 |
+
"grad_norm": 0.6229089875521143,
|
| 1036 |
+
"learning_rate": 2.6679859670118785e-06,
|
| 1037 |
+
"loss": 0.5017,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.0790697674418603,
|
| 1042 |
+
"grad_norm": 0.6124709475563379,
|
| 1043 |
+
"learning_rate": 2.5955568834114523e-06,
|
| 1044 |
+
"loss": 0.5306,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.0930232558139537,
|
| 1049 |
+
"grad_norm": 0.6436759263078206,
|
| 1050 |
+
"learning_rate": 2.523778285470835e-06,
|
| 1051 |
+
"loss": 0.5021,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.1069767441860465,
|
| 1056 |
+
"grad_norm": 0.6300072092665709,
|
| 1057 |
+
"learning_rate": 2.452669591802307e-06,
|
| 1058 |
+
"loss": 0.4888,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.1209302325581394,
|
| 1063 |
+
"grad_norm": 0.6452816620049429,
|
| 1064 |
+
"learning_rate": 2.3822500397857016e-06,
|
| 1065 |
+
"loss": 0.5352,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.1348837209302327,
|
| 1070 |
+
"grad_norm": 0.6546455768546999,
|
| 1071 |
+
"learning_rate": 2.3125386803640188e-06,
|
| 1072 |
+
"loss": 0.5395,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.1488372093023256,
|
| 1077 |
+
"grad_norm": 0.6032156311243817,
|
| 1078 |
+
"learning_rate": 2.243554372889479e-06,
|
| 1079 |
+
"loss": 0.5047,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.1627906976744184,
|
| 1084 |
+
"grad_norm": 0.6564942246168397,
|
| 1085 |
+
"learning_rate": 2.1753157800214107e-06,
|
| 1086 |
+
"loss": 0.5332,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.1767441860465118,
|
| 1091 |
+
"grad_norm": 0.6545935023148415,
|
| 1092 |
+
"learning_rate": 2.1078413626773547e-06,
|
| 1093 |
+
"loss": 0.5379,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.1906976744186046,
|
| 1098 |
+
"grad_norm": 0.6490565522460099,
|
| 1099 |
+
"learning_rate": 2.0411493750387423e-06,
|
| 1100 |
+
"loss": 0.5008,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.2046511627906975,
|
| 1105 |
+
"grad_norm": 0.6570991543105086,
|
| 1106 |
+
"learning_rate": 1.9752578596124955e-06,
|
| 1107 |
+
"loss": 0.5207,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.218604651162791,
|
| 1112 |
+
"grad_norm": 0.6785960047205722,
|
| 1113 |
+
"learning_rate": 1.9101846423499113e-06,
|
| 1114 |
+
"loss": 0.5637,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.2325581395348837,
|
| 1119 |
+
"grad_norm": 0.6343793677524651,
|
| 1120 |
+
"learning_rate": 1.8459473278241125e-06,
|
| 1121 |
+
"loss": 0.5105,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.2465116279069766,
|
| 1126 |
+
"grad_norm": 0.6170250813402354,
|
| 1127 |
+
"learning_rate": 1.7825632944674016e-06,
|
| 1128 |
+
"loss": 0.4913,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.26046511627907,
|
| 1133 |
+
"grad_norm": 0.6712608858616463,
|
| 1134 |
+
"learning_rate": 1.7200496898697832e-06,
|
| 1135 |
+
"loss": 0.5385,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.2744186046511627,
|
| 1140 |
+
"grad_norm": 0.6139220986834096,
|
| 1141 |
+
"learning_rate": 1.6584234261399535e-06,
|
| 1142 |
+
"loss": 0.549,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.2883720930232556,
|
| 1147 |
+
"grad_norm": 0.6395378832650639,
|
| 1148 |
+
"learning_rate": 1.5977011753299726e-06,
|
| 1149 |
+
"loss": 0.5189,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.302325581395349,
|
| 1154 |
+
"grad_norm": 0.64789048912411,
|
| 1155 |
+
"learning_rate": 1.5378993649249053e-06,
|
| 1156 |
+
"loss": 0.5158,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.316279069767442,
|
| 1161 |
+
"grad_norm": 0.6213190778163351,
|
| 1162 |
+
"learning_rate": 1.4790341733986085e-06,
|
| 1163 |
+
"loss": 0.5125,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.330232558139535,
|
| 1168 |
+
"grad_norm": 0.6185320002205397,
|
| 1169 |
+
"learning_rate": 1.4211215258368866e-06,
|
| 1170 |
+
"loss": 0.5021,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.344186046511628,
|
| 1175 |
+
"grad_norm": 0.6588110955127657,
|
| 1176 |
+
"learning_rate": 1.3641770896292083e-06,
|
| 1177 |
+
"loss": 0.5016,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.358139534883721,
|
| 1182 |
+
"grad_norm": 0.6086644430365765,
|
| 1183 |
+
"learning_rate": 1.3082162702301276e-06,
|
| 1184 |
+
"loss": 0.5071,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.3720930232558137,
|
| 1189 |
+
"grad_norm": 0.6438720402726894,
|
| 1190 |
+
"learning_rate": 1.2532542069915722e-06,
|
| 1191 |
+
"loss": 0.534,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.386046511627907,
|
| 1196 |
+
"grad_norm": 0.6322988485539291,
|
| 1197 |
+
"learning_rate": 1.1993057690671174e-06,
|
| 1198 |
+
"loss": 0.5222,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.4,
|
| 1203 |
+
"grad_norm": 0.6183513325619755,
|
| 1204 |
+
"learning_rate": 1.1463855513893695e-06,
|
| 1205 |
+
"loss": 0.4919,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.4139534883720932,
|
| 1210 |
+
"grad_norm": 0.6169682332492049,
|
| 1211 |
+
"learning_rate": 1.0945078707215224e-06,
|
| 1212 |
+
"loss": 0.513,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.427906976744186,
|
| 1217 |
+
"grad_norm": 0.6382733959540642,
|
| 1218 |
+
"learning_rate": 1.0436867617841768e-06,
|
| 1219 |
+
"loss": 0.5092,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.441860465116279,
|
| 1224 |
+
"grad_norm": 0.630862281709607,
|
| 1225 |
+
"learning_rate": 9.939359734584552e-07,
|
| 1226 |
+
"loss": 0.5234,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.4558139534883723,
|
| 1231 |
+
"grad_norm": 0.6351853238093808,
|
| 1232 |
+
"learning_rate": 9.452689650664515e-07,
|
| 1233 |
+
"loss": 0.4995,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.469767441860465,
|
| 1238 |
+
"grad_norm": 0.6256987475228005,
|
| 1239 |
+
"learning_rate": 8.976989027300265e-07,
|
| 1240 |
+
"loss": 0.4887,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.483720930232558,
|
| 1245 |
+
"grad_norm": 0.6842807989527501,
|
| 1246 |
+
"learning_rate": 8.512386558088919e-07,
|
| 1247 |
+
"loss": 0.5086,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.4976744186046513,
|
| 1252 |
+
"grad_norm": 0.6316844814615012,
|
| 1253 |
+
"learning_rate": 8.059007934190194e-07,
|
| 1254 |
+
"loss": 0.5027,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.511627906976744,
|
| 1259 |
+
"grad_norm": 0.6073945598367959,
|
| 1260 |
+
"learning_rate": 7.61697581032243e-07,
|
| 1261 |
+
"loss": 0.5533,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.525581395348837,
|
| 1266 |
+
"grad_norm": 0.6207165045297832,
|
| 1267 |
+
"learning_rate": 7.186409771580355e-07,
|
| 1268 |
+
"loss": 0.5255,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.5395348837209304,
|
| 1273 |
+
"grad_norm": 0.6181230727465237,
|
| 1274 |
+
"learning_rate": 6.76742630108298e-07,
|
| 1275 |
+
"loss": 0.5165,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.5534883720930233,
|
| 1280 |
+
"grad_norm": 0.6322294031612018,
|
| 1281 |
+
"learning_rate": 6.360138748461015e-07,
|
| 1282 |
+
"loss": 0.5044,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.567441860465116,
|
| 1287 |
+
"grad_norm": 0.6587999095797966,
|
| 1288 |
+
"learning_rate": 5.964657299191712e-07,
|
| 1289 |
+
"loss": 0.503,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.5813953488372094,
|
| 1294 |
+
"grad_norm": 0.6118678840468961,
|
| 1295 |
+
"learning_rate": 5.581088944789953e-07,
|
| 1296 |
+
"loss": 0.4973,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.5953488372093023,
|
| 1301 |
+
"grad_norm": 0.6499096905728639,
|
| 1302 |
+
"learning_rate": 5.20953745386329e-07,
|
| 1303 |
+
"loss": 0.5195,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.609302325581395,
|
| 1308 |
+
"grad_norm": 0.6288818601403712,
|
| 1309 |
+
"learning_rate": 4.850103344038853e-07,
|
| 1310 |
+
"loss": 0.5092,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.6232558139534885,
|
| 1315 |
+
"grad_norm": 0.586123346403167,
|
| 1316 |
+
"learning_rate": 4.502883854769935e-07,
|
| 1317 |
+
"loss": 0.5069,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.6372093023255814,
|
| 1322 |
+
"grad_norm": 0.6185988812110925,
|
| 1323 |
+
"learning_rate": 4.167972921029262e-07,
|
| 1324 |
+
"loss": 0.5054,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.6511627906976747,
|
| 1329 |
+
"grad_norm": 0.6353396661088835,
|
| 1330 |
+
"learning_rate": 3.8454611478963235e-07,
|
| 1331 |
+
"loss": 0.5259,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.6651162790697676,
|
| 1336 |
+
"grad_norm": 0.6033634241367,
|
| 1337 |
+
"learning_rate": 3.535435786045538e-07,
|
| 1338 |
+
"loss": 0.4995,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.6790697674418604,
|
| 1343 |
+
"grad_norm": 0.5959552484364983,
|
| 1344 |
+
"learning_rate": 3.237980708141919e-07,
|
| 1345 |
+
"loss": 0.5138,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.6930232558139533,
|
| 1350 |
+
"grad_norm": 0.6122358357225203,
|
| 1351 |
+
"learning_rate": 2.9531763861505967e-07,
|
| 1352 |
+
"loss": 0.5264,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.7069767441860466,
|
| 1357 |
+
"grad_norm": 0.6161969475760205,
|
| 1358 |
+
"learning_rate": 2.681099869566328e-07,
|
| 1359 |
+
"loss": 0.5218,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.7209302325581395,
|
| 1364 |
+
"grad_norm": 0.6094009937540151,
|
| 1365 |
+
"learning_rate": 2.4218247645689306e-07,
|
| 1366 |
+
"loss": 0.5103,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.734883720930233,
|
| 1371 |
+
"grad_norm": 0.6457992686372139,
|
| 1372 |
+
"learning_rate": 2.175421214110235e-07,
|
| 1373 |
+
"loss": 0.53,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.7488372093023257,
|
| 1378 |
+
"grad_norm": 0.6069680710832619,
|
| 1379 |
+
"learning_rate": 1.941955878938029e-07,
|
| 1380 |
+
"loss": 0.5097,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.7627906976744185,
|
| 1385 |
+
"grad_norm": 0.632347846663917,
|
| 1386 |
+
"learning_rate": 1.721491919561913e-07,
|
| 1387 |
+
"loss": 0.5038,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.7767441860465114,
|
| 1392 |
+
"grad_norm": 0.6065797290611403,
|
| 1393 |
+
"learning_rate": 1.514088979166256e-07,
|
| 1394 |
+
"loss": 0.4777,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.7906976744186047,
|
| 1399 |
+
"grad_norm": 0.6199549949589643,
|
| 1400 |
+
"learning_rate": 1.3198031674745814e-07,
|
| 1401 |
+
"loss": 0.5037,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.8046511627906976,
|
| 1406 |
+
"grad_norm": 0.5786521555644941,
|
| 1407 |
+
"learning_rate": 1.138687045569975e-07,
|
| 1408 |
+
"loss": 0.5089,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.818604651162791,
|
| 1413 |
+
"grad_norm": 0.6048168103516136,
|
| 1414 |
+
"learning_rate": 9.70789611675449e-08,
|
| 1415 |
+
"loss": 0.5094,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.832558139534884,
|
| 1420 |
+
"grad_norm": 0.6653344807537283,
|
| 1421 |
+
"learning_rate": 8.161562878982399e-08,
|
| 1422 |
+
"loss": 0.4776,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.8465116279069766,
|
| 1427 |
+
"grad_norm": 0.5925645150569813,
|
| 1428 |
+
"learning_rate": 6.74828907941516e-08,
|
| 1429 |
+
"loss": 0.5077,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.8604651162790695,
|
| 1434 |
+
"grad_norm": 0.6361769519335965,
|
| 1435 |
+
"learning_rate": 5.468457057869358e-08,
|
| 1436 |
+
"loss": 0.546,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.874418604651163,
|
| 1441 |
+
"grad_norm": 0.6399221259223546,
|
| 1442 |
+
"learning_rate": 4.322413053509944e-08,
|
| 1443 |
+
"loss": 0.5227,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.8883720930232557,
|
| 1448 |
+
"grad_norm": 0.6105316466062543,
|
| 1449 |
+
"learning_rate": 3.31046711118066e-08,
|
| 1450 |
+
"loss": 0.5323,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 2.902325581395349,
|
| 1455 |
+
"grad_norm": 0.6095910384032711,
|
| 1456 |
+
"learning_rate": 2.4328929975260262e-08,
|
| 1457 |
+
"loss": 0.5091,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.916279069767442,
|
| 1462 |
+
"grad_norm": 0.5949723187727618,
|
| 1463 |
+
"learning_rate": 1.6899281269279756e-08,
|
| 1464 |
+
"loss": 0.5209,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 2.9302325581395348,
|
| 1469 |
+
"grad_norm": 0.5968244471616623,
|
| 1470 |
+
"learning_rate": 1.0817734972768946e-08,
|
| 1471 |
+
"loss": 0.4926,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.9441860465116276,
|
| 1476 |
+
"grad_norm": 0.6428827484769689,
|
| 1477 |
+
"learning_rate": 6.085936355947897e-09,
|
| 1478 |
+
"loss": 0.5449,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 2.958139534883721,
|
| 1483 |
+
"grad_norm": 0.6341403080316971,
|
| 1484 |
+
"learning_rate": 2.7051655352494654e-09,
|
| 1485 |
+
"loss": 0.5251,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.972093023255814,
|
| 1490 |
+
"grad_norm": 0.6104870060581485,
|
| 1491 |
+
"learning_rate": 6.763371270035457e-10,
|
| 1492 |
+
"loss": 0.5439,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 2.986046511627907,
|
| 1497 |
+
"grad_norm": 0.5921345738524789,
|
| 1498 |
+
"learning_rate": 0.0,
|
| 1499 |
+
"loss": 0.516,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.986046511627907,
|
| 1504 |
+
"step": 213,
|
| 1505 |
+
"total_flos": 1.1419933308302131e+17,
|
| 1506 |
+
"train_loss": 0.6310646277917943,
|
| 1507 |
+
"train_runtime": 4071.0206,
|
| 1508 |
+
"train_samples_per_second": 5.059,
|
| 1509 |
+
"train_steps_per_second": 0.052
|
| 1510 |
+
}
|
| 1511 |
+
],
|
| 1512 |
+
"logging_steps": 1,
|
| 1513 |
+
"max_steps": 213,
|
| 1514 |
+
"num_input_tokens_seen": 0,
|
| 1515 |
+
"num_train_epochs": 3,
|
| 1516 |
+
"save_steps": 10,
|
| 1517 |
+
"stateful_callbacks": {
|
| 1518 |
+
"TrainerControl": {
|
| 1519 |
+
"args": {
|
| 1520 |
+
"should_epoch_stop": false,
|
| 1521 |
+
"should_evaluate": false,
|
| 1522 |
+
"should_log": false,
|
| 1523 |
+
"should_save": true,
|
| 1524 |
+
"should_training_stop": true
|
| 1525 |
+
},
|
| 1526 |
+
"attributes": {}
|
| 1527 |
+
}
|
| 1528 |
+
},
|
| 1529 |
+
"total_flos": 1.1419933308302131e+17,
|
| 1530 |
+
"train_batch_size": 2,
|
| 1531 |
+
"trial_name": null,
|
| 1532 |
+
"trial_params": null
|
| 1533 |
+
}
|
training_loss.png
ADDED
|