End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1610 -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: d1_code_gpt_3k
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# d1_code_gpt_3k
|
| 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: d1_code_gpt_3k
|
|
|
|
| 16 |
|
| 17 |
# d1_code_gpt_3k
|
| 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 mlfoundations-dev/d1_code_gpt_3k dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.850632911392405,
|
| 3 |
+
"total_flos": 8.569304900394025e+17,
|
| 4 |
+
"train_loss": 0.5346918581053615,
|
| 5 |
+
"train_runtime": 27285.7344,
|
| 6 |
+
"train_samples_per_second": 0.811,
|
| 7 |
+
"train_steps_per_second": 0.008
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.850632911392405,
|
| 3 |
+
"total_flos": 8.569304900394025e+17,
|
| 4 |
+
"train_loss": 0.5346918581053615,
|
| 5 |
+
"train_runtime": 27285.7344,
|
| 6 |
+
"train_samples_per_second": 0.811,
|
| 7 |
+
"train_steps_per_second": 0.008
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1610 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.850632911392405,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 224,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.030379746835443037,
|
| 13 |
+
"grad_norm": 6.37221868985156,
|
| 14 |
+
"learning_rate": 8.695652173913044e-07,
|
| 15 |
+
"loss": 1.0264,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.060759493670886074,
|
| 20 |
+
"grad_norm": 6.5034802963479175,
|
| 21 |
+
"learning_rate": 1.7391304347826088e-06,
|
| 22 |
+
"loss": 1.0424,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09113924050632911,
|
| 27 |
+
"grad_norm": 6.380172477883636,
|
| 28 |
+
"learning_rate": 2.6086956521739132e-06,
|
| 29 |
+
"loss": 1.0317,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12151898734177215,
|
| 34 |
+
"grad_norm": 5.817576642017238,
|
| 35 |
+
"learning_rate": 3.4782608695652175e-06,
|
| 36 |
+
"loss": 1.0183,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1518987341772152,
|
| 41 |
+
"grad_norm": 4.401190148605733,
|
| 42 |
+
"learning_rate": 4.347826086956522e-06,
|
| 43 |
+
"loss": 0.9624,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18227848101265823,
|
| 48 |
+
"grad_norm": 2.7097647034322696,
|
| 49 |
+
"learning_rate": 5.2173913043478265e-06,
|
| 50 |
+
"loss": 0.959,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21265822784810126,
|
| 55 |
+
"grad_norm": 2.450867089297927,
|
| 56 |
+
"learning_rate": 6.086956521739132e-06,
|
| 57 |
+
"loss": 0.9543,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2430379746835443,
|
| 62 |
+
"grad_norm": 4.04136120197139,
|
| 63 |
+
"learning_rate": 6.956521739130435e-06,
|
| 64 |
+
"loss": 0.9325,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.27341772151898736,
|
| 69 |
+
"grad_norm": 4.409279955403386,
|
| 70 |
+
"learning_rate": 7.82608695652174e-06,
|
| 71 |
+
"loss": 0.9357,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3037974683544304,
|
| 76 |
+
"grad_norm": 3.706439164950476,
|
| 77 |
+
"learning_rate": 8.695652173913044e-06,
|
| 78 |
+
"loss": 0.9084,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3341772151898734,
|
| 83 |
+
"grad_norm": 3.5400094637960513,
|
| 84 |
+
"learning_rate": 9.565217391304349e-06,
|
| 85 |
+
"loss": 0.8941,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.36455696202531646,
|
| 90 |
+
"grad_norm": 2.9424193684317865,
|
| 91 |
+
"learning_rate": 1.0434782608695653e-05,
|
| 92 |
+
"loss": 0.8536,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3949367088607595,
|
| 97 |
+
"grad_norm": 2.1302467308428508,
|
| 98 |
+
"learning_rate": 1.1304347826086957e-05,
|
| 99 |
+
"loss": 0.8482,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4253164556962025,
|
| 104 |
+
"grad_norm": 1.481291511968154,
|
| 105 |
+
"learning_rate": 1.2173913043478263e-05,
|
| 106 |
+
"loss": 0.8522,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.45569620253164556,
|
| 111 |
+
"grad_norm": 1.5948371689967598,
|
| 112 |
+
"learning_rate": 1.3043478260869566e-05,
|
| 113 |
+
"loss": 0.8078,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.4860759493670886,
|
| 118 |
+
"grad_norm": 1.3238320150665204,
|
| 119 |
+
"learning_rate": 1.391304347826087e-05,
|
| 120 |
+
"loss": 0.7868,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5164556962025316,
|
| 125 |
+
"grad_norm": 1.135690386872903,
|
| 126 |
+
"learning_rate": 1.4782608695652174e-05,
|
| 127 |
+
"loss": 0.7968,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5468354430379747,
|
| 132 |
+
"grad_norm": 1.1319959126772645,
|
| 133 |
+
"learning_rate": 1.565217391304348e-05,
|
| 134 |
+
"loss": 0.785,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5772151898734177,
|
| 139 |
+
"grad_norm": 0.9567822025036028,
|
| 140 |
+
"learning_rate": 1.6521739130434785e-05,
|
| 141 |
+
"loss": 0.7757,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6075949367088608,
|
| 146 |
+
"grad_norm": 1.0091484315875876,
|
| 147 |
+
"learning_rate": 1.739130434782609e-05,
|
| 148 |
+
"loss": 0.7549,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6379746835443038,
|
| 153 |
+
"grad_norm": 1.048072111801254,
|
| 154 |
+
"learning_rate": 1.8260869565217393e-05,
|
| 155 |
+
"loss": 0.7415,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6683544303797468,
|
| 160 |
+
"grad_norm": 0.9493693220268987,
|
| 161 |
+
"learning_rate": 1.9130434782608697e-05,
|
| 162 |
+
"loss": 0.7361,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6987341772151898,
|
| 167 |
+
"grad_norm": 0.9262684629013186,
|
| 168 |
+
"learning_rate": 2e-05,
|
| 169 |
+
"loss": 0.7422,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7291139240506329,
|
| 174 |
+
"grad_norm": 0.8741096470542656,
|
| 175 |
+
"learning_rate": 1.999877856940653e-05,
|
| 176 |
+
"loss": 0.7486,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.759493670886076,
|
| 181 |
+
"grad_norm": 0.7873818311147306,
|
| 182 |
+
"learning_rate": 1.999511457600466e-05,
|
| 183 |
+
"loss": 0.7235,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.789873417721519,
|
| 188 |
+
"grad_norm": 0.6420798721747906,
|
| 189 |
+
"learning_rate": 1.9989008914857115e-05,
|
| 190 |
+
"loss": 0.7188,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8202531645569621,
|
| 195 |
+
"grad_norm": 0.7203837580063897,
|
| 196 |
+
"learning_rate": 1.998046307749216e-05,
|
| 197 |
+
"loss": 0.7193,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.850632911392405,
|
| 202 |
+
"grad_norm": 0.8500834268635933,
|
| 203 |
+
"learning_rate": 1.9969479151539238e-05,
|
| 204 |
+
"loss": 0.7127,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8810126582278481,
|
| 209 |
+
"grad_norm": 0.6058389288197817,
|
| 210 |
+
"learning_rate": 1.9956059820218982e-05,
|
| 211 |
+
"loss": 0.7259,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9113924050632911,
|
| 216 |
+
"grad_norm": 0.7078168051247413,
|
| 217 |
+
"learning_rate": 1.9940208361687762e-05,
|
| 218 |
+
"loss": 0.6921,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9417721518987342,
|
| 223 |
+
"grad_norm": 0.7015805895369476,
|
| 224 |
+
"learning_rate": 1.9921928648236855e-05,
|
| 225 |
+
"loss": 0.7197,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9721518987341772,
|
| 230 |
+
"grad_norm": 0.7387848558163191,
|
| 231 |
+
"learning_rate": 1.990122514534651e-05,
|
| 232 |
+
"loss": 0.6959,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.010126582278481,
|
| 237 |
+
"grad_norm": 0.5763144897662861,
|
| 238 |
+
"learning_rate": 1.9878102910595097e-05,
|
| 239 |
+
"loss": 0.6832,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.040506329113924,
|
| 244 |
+
"grad_norm": 0.6437270212433482,
|
| 245 |
+
"learning_rate": 1.985256759242359e-05,
|
| 246 |
+
"loss": 0.6611,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.070886075949367,
|
| 251 |
+
"grad_norm": 0.5953857969572114,
|
| 252 |
+
"learning_rate": 1.982462542875576e-05,
|
| 253 |
+
"loss": 0.6687,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1012658227848102,
|
| 258 |
+
"grad_norm": 0.6001397001527595,
|
| 259 |
+
"learning_rate": 1.979428324547432e-05,
|
| 260 |
+
"loss": 0.6662,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.1316455696202532,
|
| 265 |
+
"grad_norm": 0.6353749730621574,
|
| 266 |
+
"learning_rate": 1.9761548454753455e-05,
|
| 267 |
+
"loss": 0.6533,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1620253164556962,
|
| 272 |
+
"grad_norm": 0.5856471519075721,
|
| 273 |
+
"learning_rate": 1.972642905324813e-05,
|
| 274 |
+
"loss": 0.646,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.1924050632911392,
|
| 279 |
+
"grad_norm": 0.5360846001231565,
|
| 280 |
+
"learning_rate": 1.9688933620140638e-05,
|
| 281 |
+
"loss": 0.6501,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2227848101265824,
|
| 286 |
+
"grad_norm": 0.5361166895681458,
|
| 287 |
+
"learning_rate": 1.96490713150448e-05,
|
| 288 |
+
"loss": 0.6444,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2531645569620253,
|
| 293 |
+
"grad_norm": 0.43296430499453425,
|
| 294 |
+
"learning_rate": 1.9606851875768404e-05,
|
| 295 |
+
"loss": 0.6456,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.2835443037974683,
|
| 300 |
+
"grad_norm": 0.526080735366251,
|
| 301 |
+
"learning_rate": 1.956228561593441e-05,
|
| 302 |
+
"loss": 0.6494,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.3139240506329113,
|
| 307 |
+
"grad_norm": 0.4906847188578764,
|
| 308 |
+
"learning_rate": 1.9515383422461457e-05,
|
| 309 |
+
"loss": 0.6527,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3443037974683545,
|
| 314 |
+
"grad_norm": 0.5680631359585419,
|
| 315 |
+
"learning_rate": 1.9466156752904344e-05,
|
| 316 |
+
"loss": 0.6447,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.3746835443037975,
|
| 321 |
+
"grad_norm": 0.5130133069861942,
|
| 322 |
+
"learning_rate": 1.9414617632655114e-05,
|
| 323 |
+
"loss": 0.6314,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4050632911392404,
|
| 328 |
+
"grad_norm": 0.47327015900025143,
|
| 329 |
+
"learning_rate": 1.9360778652005416e-05,
|
| 330 |
+
"loss": 0.6304,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4354430379746836,
|
| 335 |
+
"grad_norm": 0.6242281508405784,
|
| 336 |
+
"learning_rate": 1.9304652963070868e-05,
|
| 337 |
+
"loss": 0.6398,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4658227848101266,
|
| 342 |
+
"grad_norm": 0.5892904102401884,
|
| 343 |
+
"learning_rate": 1.9246254276578175e-05,
|
| 344 |
+
"loss": 0.6248,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.4962025316455696,
|
| 349 |
+
"grad_norm": 0.6531942163155297,
|
| 350 |
+
"learning_rate": 1.9185596858515797e-05,
|
| 351 |
+
"loss": 0.6404,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5265822784810128,
|
| 356 |
+
"grad_norm": 0.6483567267386434,
|
| 357 |
+
"learning_rate": 1.9122695526648968e-05,
|
| 358 |
+
"loss": 0.6365,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5569620253164556,
|
| 363 |
+
"grad_norm": 0.6983956278308369,
|
| 364 |
+
"learning_rate": 1.905756564689991e-05,
|
| 365 |
+
"loss": 0.643,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.5873417721518988,
|
| 370 |
+
"grad_norm": 0.7862995399359736,
|
| 371 |
+
"learning_rate": 1.8990223129594146e-05,
|
| 372 |
+
"loss": 0.6419,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.6177215189873417,
|
| 377 |
+
"grad_norm": 0.5430925229786947,
|
| 378 |
+
"learning_rate": 1.8920684425573865e-05,
|
| 379 |
+
"loss": 0.6324,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6481012658227847,
|
| 384 |
+
"grad_norm": 0.885437054994622,
|
| 385 |
+
"learning_rate": 1.884896652217917e-05,
|
| 386 |
+
"loss": 0.6356,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.678481012658228,
|
| 391 |
+
"grad_norm": 0.5143264601586393,
|
| 392 |
+
"learning_rate": 1.877508693909831e-05,
|
| 393 |
+
"loss": 0.6333,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.7088607594936709,
|
| 398 |
+
"grad_norm": 0.8076187986221476,
|
| 399 |
+
"learning_rate": 1.8699063724087905e-05,
|
| 400 |
+
"loss": 0.6271,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7392405063291139,
|
| 405 |
+
"grad_norm": 0.6430936572112902,
|
| 406 |
+
"learning_rate": 1.862091544856407e-05,
|
| 407 |
+
"loss": 0.6277,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.769620253164557,
|
| 412 |
+
"grad_norm": 0.4956450969395412,
|
| 413 |
+
"learning_rate": 1.854066120306571e-05,
|
| 414 |
+
"loss": 0.6035,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.8,
|
| 419 |
+
"grad_norm": 0.4950456942853279,
|
| 420 |
+
"learning_rate": 1.8458320592590976e-05,
|
| 421 |
+
"loss": 0.6386,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.830379746835443,
|
| 426 |
+
"grad_norm": 0.6371116080859314,
|
| 427 |
+
"learning_rate": 1.837391373180801e-05,
|
| 428 |
+
"loss": 0.6229,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.8607594936708862,
|
| 433 |
+
"grad_norm": 0.43714150441878535,
|
| 434 |
+
"learning_rate": 1.8287461240141217e-05,
|
| 435 |
+
"loss": 0.6176,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.891139240506329,
|
| 440 |
+
"grad_norm": 0.6453784990074324,
|
| 441 |
+
"learning_rate": 1.8198984236734246e-05,
|
| 442 |
+
"loss": 0.6077,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9215189873417722,
|
| 447 |
+
"grad_norm": 0.5570607723349542,
|
| 448 |
+
"learning_rate": 1.8108504335290852e-05,
|
| 449 |
+
"loss": 0.6352,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9518987341772152,
|
| 454 |
+
"grad_norm": 0.48245224757972355,
|
| 455 |
+
"learning_rate": 1.8016043638794975e-05,
|
| 456 |
+
"loss": 0.637,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.9822784810126581,
|
| 461 |
+
"grad_norm": 0.5721748046329147,
|
| 462 |
+
"learning_rate": 1.7921624734111292e-05,
|
| 463 |
+
"loss": 0.6076,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.020253164556962,
|
| 468 |
+
"grad_norm": 0.5141868421323431,
|
| 469 |
+
"learning_rate": 1.7825270686467567e-05,
|
| 470 |
+
"loss": 0.5894,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.050632911392405,
|
| 475 |
+
"grad_norm": 0.5381266684759717,
|
| 476 |
+
"learning_rate": 1.7727005033820117e-05,
|
| 477 |
+
"loss": 0.5556,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.081012658227848,
|
| 482 |
+
"grad_norm": 0.6016121815074985,
|
| 483 |
+
"learning_rate": 1.762685178110382e-05,
|
| 484 |
+
"loss": 0.5527,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.1113924050632913,
|
| 489 |
+
"grad_norm": 0.6729531958813787,
|
| 490 |
+
"learning_rate": 1.752483539436807e-05,
|
| 491 |
+
"loss": 0.563,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.141772151898734,
|
| 496 |
+
"grad_norm": 0.5102620941806387,
|
| 497 |
+
"learning_rate": 1.7420980794800013e-05,
|
| 498 |
+
"loss": 0.5718,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.1721518987341772,
|
| 503 |
+
"grad_norm": 0.6042146458914103,
|
| 504 |
+
"learning_rate": 1.731531335263669e-05,
|
| 505 |
+
"loss": 0.5608,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.2025316455696204,
|
| 510 |
+
"grad_norm": 0.5894071656765725,
|
| 511 |
+
"learning_rate": 1.720785888096743e-05,
|
| 512 |
+
"loss": 0.5699,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.232911392405063,
|
| 517 |
+
"grad_norm": 0.5775536328227947,
|
| 518 |
+
"learning_rate": 1.7098643629428035e-05,
|
| 519 |
+
"loss": 0.5625,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.2632911392405064,
|
| 524 |
+
"grad_norm": 0.606548495008496,
|
| 525 |
+
"learning_rate": 1.698769427778842e-05,
|
| 526 |
+
"loss": 0.5404,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.293670886075949,
|
| 531 |
+
"grad_norm": 0.6622908431371424,
|
| 532 |
+
"learning_rate": 1.687503792943506e-05,
|
| 533 |
+
"loss": 0.5747,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.3240506329113924,
|
| 538 |
+
"grad_norm": 0.4533766615425194,
|
| 539 |
+
"learning_rate": 1.6760702104750046e-05,
|
| 540 |
+
"loss": 0.5481,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3544303797468356,
|
| 545 |
+
"grad_norm": 0.4811067517469304,
|
| 546 |
+
"learning_rate": 1.664471473438822e-05,
|
| 547 |
+
"loss": 0.5524,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.3848101265822783,
|
| 552 |
+
"grad_norm": 0.5210479774326526,
|
| 553 |
+
"learning_rate": 1.6527104152454096e-05,
|
| 554 |
+
"loss": 0.5372,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.4151898734177215,
|
| 559 |
+
"grad_norm": 0.546680221442232,
|
| 560 |
+
"learning_rate": 1.6407899089580263e-05,
|
| 561 |
+
"loss": 0.5534,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4455696202531647,
|
| 566 |
+
"grad_norm": 0.4607636332203975,
|
| 567 |
+
"learning_rate": 1.628712866590885e-05,
|
| 568 |
+
"loss": 0.556,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.4759493670886075,
|
| 573 |
+
"grad_norm": 0.5340360502609477,
|
| 574 |
+
"learning_rate": 1.6164822383977912e-05,
|
| 575 |
+
"loss": 0.5578,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.5063291139240507,
|
| 580 |
+
"grad_norm": 0.4829659674501385,
|
| 581 |
+
"learning_rate": 1.604101012151436e-05,
|
| 582 |
+
"loss": 0.5586,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.536708860759494,
|
| 587 |
+
"grad_norm": 0.5154687639314293,
|
| 588 |
+
"learning_rate": 1.5915722124135227e-05,
|
| 589 |
+
"loss": 0.5456,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.5670886075949366,
|
| 594 |
+
"grad_norm": 0.4828149719193968,
|
| 595 |
+
"learning_rate": 1.5788988997959115e-05,
|
| 596 |
+
"loss": 0.5485,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.59746835443038,
|
| 601 |
+
"grad_norm": 0.49279954192791947,
|
| 602 |
+
"learning_rate": 1.5660841702129533e-05,
|
| 603 |
+
"loss": 0.57,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.6278481012658226,
|
| 608 |
+
"grad_norm": 0.4742602851060561,
|
| 609 |
+
"learning_rate": 1.5531311541251995e-05,
|
| 610 |
+
"loss": 0.5446,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.6582278481012658,
|
| 615 |
+
"grad_norm": 0.46089627282095114,
|
| 616 |
+
"learning_rate": 1.540043015774676e-05,
|
| 617 |
+
"loss": 0.5432,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.688607594936709,
|
| 622 |
+
"grad_norm": 0.5193521045258312,
|
| 623 |
+
"learning_rate": 1.5268229524119007e-05,
|
| 624 |
+
"loss": 0.5526,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.7189873417721517,
|
| 629 |
+
"grad_norm": 0.599090537230937,
|
| 630 |
+
"learning_rate": 1.513474193514842e-05,
|
| 631 |
+
"loss": 0.5387,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.749367088607595,
|
| 636 |
+
"grad_norm": 0.5392240932844506,
|
| 637 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 638 |
+
"loss": 0.5644,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.779746835443038,
|
| 643 |
+
"grad_norm": 0.6484053666330277,
|
| 644 |
+
"learning_rate": 1.4864036634258112e-05,
|
| 645 |
+
"loss": 0.5529,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.810126582278481,
|
| 650 |
+
"grad_norm": 0.49558973661776645,
|
| 651 |
+
"learning_rate": 1.4726885051885654e-05,
|
| 652 |
+
"loss": 0.5574,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.840506329113924,
|
| 657 |
+
"grad_norm": 0.5477713161891372,
|
| 658 |
+
"learning_rate": 1.4588578757110359e-05,
|
| 659 |
+
"loss": 0.5342,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.8708860759493673,
|
| 664 |
+
"grad_norm": 0.5294733295915529,
|
| 665 |
+
"learning_rate": 1.4449151536240167e-05,
|
| 666 |
+
"loss": 0.56,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.90126582278481,
|
| 671 |
+
"grad_norm": 0.4969749891984834,
|
| 672 |
+
"learning_rate": 1.4308637449409705e-05,
|
| 673 |
+
"loss": 0.5594,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.9316455696202532,
|
| 678 |
+
"grad_norm": 0.4381308375079077,
|
| 679 |
+
"learning_rate": 1.4167070822259868e-05,
|
| 680 |
+
"loss": 0.5439,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.962025316455696,
|
| 685 |
+
"grad_norm": 0.4173526948473983,
|
| 686 |
+
"learning_rate": 1.402448623755254e-05,
|
| 687 |
+
"loss": 0.5461,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.992405063291139,
|
| 692 |
+
"grad_norm": 0.4328409092651217,
|
| 693 |
+
"learning_rate": 1.3880918526722497e-05,
|
| 694 |
+
"loss": 0.5489,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.030379746835443,
|
| 699 |
+
"grad_norm": 0.5284467996380835,
|
| 700 |
+
"learning_rate": 1.3736402761368597e-05,
|
| 701 |
+
"loss": 0.5022,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.060759493670886,
|
| 706 |
+
"grad_norm": 0.4698391974283234,
|
| 707 |
+
"learning_rate": 1.3590974244686248e-05,
|
| 708 |
+
"loss": 0.505,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.091139240506329,
|
| 713 |
+
"grad_norm": 0.611188398162724,
|
| 714 |
+
"learning_rate": 1.344466850284333e-05,
|
| 715 |
+
"loss": 0.506,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.1215189873417724,
|
| 720 |
+
"grad_norm": 0.5181870143762257,
|
| 721 |
+
"learning_rate": 1.3297521276301666e-05,
|
| 722 |
+
"loss": 0.5025,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.151898734177215,
|
| 727 |
+
"grad_norm": 0.4746726810844399,
|
| 728 |
+
"learning_rate": 1.3149568511086104e-05,
|
| 729 |
+
"loss": 0.4962,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.1822784810126583,
|
| 734 |
+
"grad_norm": 0.5222598143525973,
|
| 735 |
+
"learning_rate": 1.300084635000341e-05,
|
| 736 |
+
"loss": 0.4861,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.212658227848101,
|
| 741 |
+
"grad_norm": 0.43490644698600855,
|
| 742 |
+
"learning_rate": 1.2851391123813075e-05,
|
| 743 |
+
"loss": 0.4993,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.2430379746835443,
|
| 748 |
+
"grad_norm": 0.4833019704949156,
|
| 749 |
+
"learning_rate": 1.2701239342352223e-05,
|
| 750 |
+
"loss": 0.4633,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.2734177215189875,
|
| 755 |
+
"grad_norm": 0.5030493664394047,
|
| 756 |
+
"learning_rate": 1.2550427685616767e-05,
|
| 757 |
+
"loss": 0.4863,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.3037974683544302,
|
| 762 |
+
"grad_norm": 0.4335113333888567,
|
| 763 |
+
"learning_rate": 1.239899299480098e-05,
|
| 764 |
+
"loss": 0.4799,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.3341772151898734,
|
| 769 |
+
"grad_norm": 0.4771808496390477,
|
| 770 |
+
"learning_rate": 1.2246972263297718e-05,
|
| 771 |
+
"loss": 0.4739,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.3645569620253166,
|
| 776 |
+
"grad_norm": 0.43288468354190207,
|
| 777 |
+
"learning_rate": 1.2094402627661447e-05,
|
| 778 |
+
"loss": 0.4955,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.3949367088607594,
|
| 783 |
+
"grad_norm": 0.3851443193327133,
|
| 784 |
+
"learning_rate": 1.1941321358536278e-05,
|
| 785 |
+
"loss": 0.492,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.4253164556962026,
|
| 790 |
+
"grad_norm": 0.45280439112604726,
|
| 791 |
+
"learning_rate": 1.1787765851551296e-05,
|
| 792 |
+
"loss": 0.4927,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.4556962025316453,
|
| 797 |
+
"grad_norm": 0.3837500484870304,
|
| 798 |
+
"learning_rate": 1.1633773618185302e-05,
|
| 799 |
+
"loss": 0.4912,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.4860759493670885,
|
| 804 |
+
"grad_norm": 0.46258069326812357,
|
| 805 |
+
"learning_rate": 1.14793822766033e-05,
|
| 806 |
+
"loss": 0.506,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.5164556962025317,
|
| 811 |
+
"grad_norm": 0.4067964537702053,
|
| 812 |
+
"learning_rate": 1.132462954246688e-05,
|
| 813 |
+
"loss": 0.4854,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.546835443037975,
|
| 818 |
+
"grad_norm": 0.42990922993772457,
|
| 819 |
+
"learning_rate": 1.1169553219720828e-05,
|
| 820 |
+
"loss": 0.4929,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.5772151898734177,
|
| 825 |
+
"grad_norm": 0.38635348880117143,
|
| 826 |
+
"learning_rate": 1.1014191191358118e-05,
|
| 827 |
+
"loss": 0.4825,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.607594936708861,
|
| 832 |
+
"grad_norm": 0.3870514422615681,
|
| 833 |
+
"learning_rate": 1.085858141016566e-05,
|
| 834 |
+
"loss": 0.4776,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.6379746835443036,
|
| 839 |
+
"grad_norm": 0.3626145330339587,
|
| 840 |
+
"learning_rate": 1.070276188945293e-05,
|
| 841 |
+
"loss": 0.4771,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.668354430379747,
|
| 846 |
+
"grad_norm": 0.43606645169013275,
|
| 847 |
+
"learning_rate": 1.0546770693765859e-05,
|
| 848 |
+
"loss": 0.4893,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.69873417721519,
|
| 853 |
+
"grad_norm": 0.3799921068914223,
|
| 854 |
+
"learning_rate": 1.0390645929588197e-05,
|
| 855 |
+
"loss": 0.482,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.729113924050633,
|
| 860 |
+
"grad_norm": 0.4170885304781762,
|
| 861 |
+
"learning_rate": 1.0234425736032607e-05,
|
| 862 |
+
"loss": 0.4903,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.759493670886076,
|
| 867 |
+
"grad_norm": 0.6975619010340954,
|
| 868 |
+
"learning_rate": 1.007814827552384e-05,
|
| 869 |
+
"loss": 0.4634,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.7898734177215188,
|
| 874 |
+
"grad_norm": 0.40226723349625243,
|
| 875 |
+
"learning_rate": 9.92185172447616e-06,
|
| 876 |
+
"loss": 0.4851,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.820253164556962,
|
| 881 |
+
"grad_norm": 0.3667501915935181,
|
| 882 |
+
"learning_rate": 9.765574263967397e-06,
|
| 883 |
+
"loss": 0.481,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.850632911392405,
|
| 888 |
+
"grad_norm": 0.3697874278619689,
|
| 889 |
+
"learning_rate": 9.609354070411807e-06,
|
| 890 |
+
"loss": 0.4876,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.8810126582278484,
|
| 895 |
+
"grad_norm": 0.357801393234647,
|
| 896 |
+
"learning_rate": 9.453229306234143e-06,
|
| 897 |
+
"loss": 0.4911,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.911392405063291,
|
| 902 |
+
"grad_norm": 0.39163201253823016,
|
| 903 |
+
"learning_rate": 9.297238110547075e-06,
|
| 904 |
+
"loss": 0.4842,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.9417721518987343,
|
| 909 |
+
"grad_norm": 0.3996876113061803,
|
| 910 |
+
"learning_rate": 9.14141858983434e-06,
|
| 911 |
+
"loss": 0.4838,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.972151898734177,
|
| 916 |
+
"grad_norm": 0.35412118410591714,
|
| 917 |
+
"learning_rate": 8.985808808641883e-06,
|
| 918 |
+
"loss": 0.4864,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.010126582278481,
|
| 923 |
+
"grad_norm": 2.067828413495023,
|
| 924 |
+
"learning_rate": 8.830446780279175e-06,
|
| 925 |
+
"loss": 0.4708,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.040506329113924,
|
| 930 |
+
"grad_norm": 0.9152902583926013,
|
| 931 |
+
"learning_rate": 8.675370457533122e-06,
|
| 932 |
+
"loss": 0.4187,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.0708860759493675,
|
| 937 |
+
"grad_norm": 17.35244540703921,
|
| 938 |
+
"learning_rate": 8.520617723396702e-06,
|
| 939 |
+
"loss": 0.4409,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.10126582278481,
|
| 944 |
+
"grad_norm": 1.6577060257645506,
|
| 945 |
+
"learning_rate": 8.366226381814698e-06,
|
| 946 |
+
"loss": 0.4372,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.131645569620253,
|
| 951 |
+
"grad_norm": 0.5280317347872258,
|
| 952 |
+
"learning_rate": 8.212234148448708e-06,
|
| 953 |
+
"loss": 0.4433,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.162025316455696,
|
| 958 |
+
"grad_norm": 0.6688184987684095,
|
| 959 |
+
"learning_rate": 8.058678641463724e-06,
|
| 960 |
+
"loss": 0.4361,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.192405063291139,
|
| 965 |
+
"grad_norm": 1.0058311734202963,
|
| 966 |
+
"learning_rate": 7.905597372338558e-06,
|
| 967 |
+
"loss": 0.4521,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.222784810126583,
|
| 972 |
+
"grad_norm": 0.4461031074464891,
|
| 973 |
+
"learning_rate": 7.753027736702283e-06,
|
| 974 |
+
"loss": 0.4341,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.253164556962025,
|
| 979 |
+
"grad_norm": 0.5391239885653292,
|
| 980 |
+
"learning_rate": 7.601007005199022e-06,
|
| 981 |
+
"loss": 0.4425,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.283544303797468,
|
| 986 |
+
"grad_norm": 0.5301098206506064,
|
| 987 |
+
"learning_rate": 7.449572314383237e-06,
|
| 988 |
+
"loss": 0.4245,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.313924050632911,
|
| 993 |
+
"grad_norm": 0.46623949539660703,
|
| 994 |
+
"learning_rate": 7.298760657647779e-06,
|
| 995 |
+
"loss": 0.4214,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.3443037974683545,
|
| 1000 |
+
"grad_norm": 0.442414367812498,
|
| 1001 |
+
"learning_rate": 7.148608876186931e-06,
|
| 1002 |
+
"loss": 0.4384,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.374683544303798,
|
| 1007 |
+
"grad_norm": 0.46015866197369026,
|
| 1008 |
+
"learning_rate": 6.999153649996595e-06,
|
| 1009 |
+
"loss": 0.449,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.405063291139241,
|
| 1014 |
+
"grad_norm": 0.4708636486435107,
|
| 1015 |
+
"learning_rate": 6.8504314889138956e-06,
|
| 1016 |
+
"loss": 0.4205,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.435443037974683,
|
| 1021 |
+
"grad_norm": 0.3957689916265755,
|
| 1022 |
+
"learning_rate": 6.702478723698336e-06,
|
| 1023 |
+
"loss": 0.4371,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.465822784810126,
|
| 1028 |
+
"grad_norm": 0.42482708366220556,
|
| 1029 |
+
"learning_rate": 6.555331497156671e-06,
|
| 1030 |
+
"loss": 0.432,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.49620253164557,
|
| 1035 |
+
"grad_norm": 0.48059851550114396,
|
| 1036 |
+
"learning_rate": 6.4090257553137566e-06,
|
| 1037 |
+
"loss": 0.4462,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.526582278481013,
|
| 1042 |
+
"grad_norm": 0.3571362661917008,
|
| 1043 |
+
"learning_rate": 6.263597238631405e-06,
|
| 1044 |
+
"loss": 0.4353,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.556962025316456,
|
| 1049 |
+
"grad_norm": 0.4167219945081819,
|
| 1050 |
+
"learning_rate": 6.119081473277502e-06,
|
| 1051 |
+
"loss": 0.4327,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.587341772151898,
|
| 1056 |
+
"grad_norm": 0.41818988585067385,
|
| 1057 |
+
"learning_rate": 5.975513762447465e-06,
|
| 1058 |
+
"loss": 0.4427,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.6177215189873415,
|
| 1063 |
+
"grad_norm": 0.3925674797200369,
|
| 1064 |
+
"learning_rate": 5.832929177740134e-06,
|
| 1065 |
+
"loss": 0.4364,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.648101265822785,
|
| 1070 |
+
"grad_norm": 0.40374638803129476,
|
| 1071 |
+
"learning_rate": 5.6913625505902966e-06,
|
| 1072 |
+
"loss": 0.4108,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.678481012658228,
|
| 1077 |
+
"grad_norm": 0.388387880997993,
|
| 1078 |
+
"learning_rate": 5.550848463759835e-06,
|
| 1079 |
+
"loss": 0.427,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.708860759493671,
|
| 1084 |
+
"grad_norm": 0.3642305851572156,
|
| 1085 |
+
"learning_rate": 5.411421242889643e-06,
|
| 1086 |
+
"loss": 0.4269,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.739240506329114,
|
| 1091 |
+
"grad_norm": 0.4082645818703638,
|
| 1092 |
+
"learning_rate": 5.273114948114346e-06,
|
| 1093 |
+
"loss": 0.4323,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.769620253164557,
|
| 1098 |
+
"grad_norm": 0.33785963966215543,
|
| 1099 |
+
"learning_rate": 5.135963365741892e-06,
|
| 1100 |
+
"loss": 0.4384,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.8,
|
| 1105 |
+
"grad_norm": 0.37611991947395673,
|
| 1106 |
+
"learning_rate": 5.000000000000003e-06,
|
| 1107 |
+
"loss": 0.4337,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.830379746835443,
|
| 1112 |
+
"grad_norm": 0.35623502983585953,
|
| 1113 |
+
"learning_rate": 4.865258064851579e-06,
|
| 1114 |
+
"loss": 0.4338,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.860759493670886,
|
| 1119 |
+
"grad_norm": 0.3185465826203866,
|
| 1120 |
+
"learning_rate": 4.731770475880995e-06,
|
| 1121 |
+
"loss": 0.4268,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.891139240506329,
|
| 1126 |
+
"grad_norm": 0.3568918726471837,
|
| 1127 |
+
"learning_rate": 4.599569842253244e-06,
|
| 1128 |
+
"loss": 0.4289,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.921518987341772,
|
| 1133 |
+
"grad_norm": 0.3819530518334745,
|
| 1134 |
+
"learning_rate": 4.468688458748006e-06,
|
| 1135 |
+
"loss": 0.4339,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.951898734177215,
|
| 1140 |
+
"grad_norm": 0.3351594331992901,
|
| 1141 |
+
"learning_rate": 4.339158297870469e-06,
|
| 1142 |
+
"loss": 0.4331,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.982278481012658,
|
| 1147 |
+
"grad_norm": 0.3021429578703631,
|
| 1148 |
+
"learning_rate": 4.211011002040885e-06,
|
| 1149 |
+
"loss": 0.4234,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.020253164556962,
|
| 1154 |
+
"grad_norm": 0.4006915974335434,
|
| 1155 |
+
"learning_rate": 4.084277875864776e-06,
|
| 1156 |
+
"loss": 0.4001,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.050632911392405,
|
| 1161 |
+
"grad_norm": 0.36778820176808397,
|
| 1162 |
+
"learning_rate": 3.958989878485644e-06,
|
| 1163 |
+
"loss": 0.4168,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.0810126582278485,
|
| 1168 |
+
"grad_norm": 0.32130448433145914,
|
| 1169 |
+
"learning_rate": 3.83517761602209e-06,
|
| 1170 |
+
"loss": 0.4065,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.111392405063291,
|
| 1175 |
+
"grad_norm": 0.3809966372789622,
|
| 1176 |
+
"learning_rate": 3.712871334091154e-06,
|
| 1177 |
+
"loss": 0.4138,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.141772151898734,
|
| 1182 |
+
"grad_norm": 0.3772267964660737,
|
| 1183 |
+
"learning_rate": 3.592100910419738e-06,
|
| 1184 |
+
"loss": 0.4033,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.172151898734177,
|
| 1189 |
+
"grad_norm": 0.3689981910530363,
|
| 1190 |
+
"learning_rate": 3.4728958475459052e-06,
|
| 1191 |
+
"loss": 0.4068,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.2025316455696204,
|
| 1196 |
+
"grad_norm": 0.35198321548754313,
|
| 1197 |
+
"learning_rate": 3.355285265611784e-06,
|
| 1198 |
+
"loss": 0.3949,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.232911392405064,
|
| 1203 |
+
"grad_norm": 0.35769710232766566,
|
| 1204 |
+
"learning_rate": 3.2392978952499553e-06,
|
| 1205 |
+
"loss": 0.39,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.263291139240506,
|
| 1210 |
+
"grad_norm": 0.33611090570761076,
|
| 1211 |
+
"learning_rate": 3.1249620705649417e-06,
|
| 1212 |
+
"loss": 0.4065,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.293670886075949,
|
| 1217 |
+
"grad_norm": 0.4049939946491654,
|
| 1218 |
+
"learning_rate": 3.0123057222115835e-06,
|
| 1219 |
+
"loss": 0.3778,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.324050632911392,
|
| 1224 |
+
"grad_norm": 0.3210913930686345,
|
| 1225 |
+
"learning_rate": 2.9013563705719673e-06,
|
| 1226 |
+
"loss": 0.3921,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.3544303797468356,
|
| 1231 |
+
"grad_norm": 0.3659685382050678,
|
| 1232 |
+
"learning_rate": 2.7921411190325753e-06,
|
| 1233 |
+
"loss": 0.3979,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.384810126582279,
|
| 1238 |
+
"grad_norm": 0.36117783642633866,
|
| 1239 |
+
"learning_rate": 2.6846866473633126e-06,
|
| 1240 |
+
"loss": 0.4048,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.415189873417722,
|
| 1245 |
+
"grad_norm": 0.30599384631214477,
|
| 1246 |
+
"learning_rate": 2.579019205199992e-06,
|
| 1247 |
+
"loss": 0.3918,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.445569620253164,
|
| 1252 |
+
"grad_norm": 0.3158134459858677,
|
| 1253 |
+
"learning_rate": 2.4751646056319334e-06,
|
| 1254 |
+
"loss": 0.3977,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.4759493670886075,
|
| 1259 |
+
"grad_norm": 0.3415178405429152,
|
| 1260 |
+
"learning_rate": 2.373148218896182e-06,
|
| 1261 |
+
"loss": 0.3941,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.506329113924051,
|
| 1266 |
+
"grad_norm": 0.34056377810339455,
|
| 1267 |
+
"learning_rate": 2.2729949661798876e-06,
|
| 1268 |
+
"loss": 0.3989,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.536708860759494,
|
| 1273 |
+
"grad_norm": 0.32197225533734486,
|
| 1274 |
+
"learning_rate": 2.174729313532433e-06,
|
| 1275 |
+
"loss": 0.3915,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.567088607594937,
|
| 1280 |
+
"grad_norm": 0.32400566343568027,
|
| 1281 |
+
"learning_rate": 2.078375265888707e-06,
|
| 1282 |
+
"loss": 0.3834,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.597468354430379,
|
| 1287 |
+
"grad_norm": 0.2988437918726148,
|
| 1288 |
+
"learning_rate": 1.9839563612050273e-06,
|
| 1289 |
+
"loss": 0.4048,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.627848101265823,
|
| 1294 |
+
"grad_norm": 0.3142881945889849,
|
| 1295 |
+
"learning_rate": 1.8914956647091497e-06,
|
| 1296 |
+
"loss": 0.4044,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.658227848101266,
|
| 1301 |
+
"grad_norm": 0.2912800871138812,
|
| 1302 |
+
"learning_rate": 1.8010157632657544e-06,
|
| 1303 |
+
"loss": 0.3915,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.688607594936709,
|
| 1308 |
+
"grad_norm": 0.331383445209847,
|
| 1309 |
+
"learning_rate": 1.7125387598587862e-06,
|
| 1310 |
+
"loss": 0.3922,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.718987341772152,
|
| 1315 |
+
"grad_norm": 0.295340510609225,
|
| 1316 |
+
"learning_rate": 1.6260862681919965e-06,
|
| 1317 |
+
"loss": 0.3803,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.749367088607595,
|
| 1322 |
+
"grad_norm": 0.2861546159019479,
|
| 1323 |
+
"learning_rate": 1.5416794074090258e-06,
|
| 1324 |
+
"loss": 0.3846,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.779746835443038,
|
| 1329 |
+
"grad_norm": 0.3026296765652463,
|
| 1330 |
+
"learning_rate": 1.459338796934293e-06,
|
| 1331 |
+
"loss": 0.4177,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.810126582278481,
|
| 1336 |
+
"grad_norm": 0.2905750515452127,
|
| 1337 |
+
"learning_rate": 1.3790845514359363e-06,
|
| 1338 |
+
"loss": 0.4132,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.840506329113924,
|
| 1343 |
+
"grad_norm": 0.28908547248165223,
|
| 1344 |
+
"learning_rate": 1.300936275912098e-06,
|
| 1345 |
+
"loss": 0.3877,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.870886075949367,
|
| 1350 |
+
"grad_norm": 0.3120697859896662,
|
| 1351 |
+
"learning_rate": 1.224913060901688e-06,
|
| 1352 |
+
"loss": 0.3857,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.9012658227848105,
|
| 1357 |
+
"grad_norm": 0.2681050832298745,
|
| 1358 |
+
"learning_rate": 1.1510334778208332e-06,
|
| 1359 |
+
"loss": 0.3955,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.931645569620253,
|
| 1364 |
+
"grad_norm": 0.26796636762585163,
|
| 1365 |
+
"learning_rate": 1.0793155744261352e-06,
|
| 1366 |
+
"loss": 0.4222,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.962025316455696,
|
| 1371 |
+
"grad_norm": 0.28122294877877857,
|
| 1372 |
+
"learning_rate": 1.0097768704058542e-06,
|
| 1373 |
+
"loss": 0.3983,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.992405063291139,
|
| 1378 |
+
"grad_norm": 0.28084753163815557,
|
| 1379 |
+
"learning_rate": 9.424343531000968e-07,
|
| 1380 |
+
"loss": 0.3781,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.030379746835443,
|
| 1385 |
+
"grad_norm": 0.3035601235316243,
|
| 1386 |
+
"learning_rate": 8.773044733510338e-07,
|
| 1387 |
+
"loss": 0.3843,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.060759493670886,
|
| 1392 |
+
"grad_norm": 0.286177647003549,
|
| 1393 |
+
"learning_rate": 8.144031414842012e-07,
|
| 1394 |
+
"loss": 0.3553,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.091139240506329,
|
| 1399 |
+
"grad_norm": 0.27499257842697156,
|
| 1400 |
+
"learning_rate": 7.537457234218271e-07,
|
| 1401 |
+
"loss": 0.3878,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.121518987341772,
|
| 1406 |
+
"grad_norm": 0.287831290077759,
|
| 1407 |
+
"learning_rate": 6.953470369291349e-07,
|
| 1408 |
+
"loss": 0.3636,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.151898734177215,
|
| 1413 |
+
"grad_norm": 0.3270864734675839,
|
| 1414 |
+
"learning_rate": 6.392213479945852e-07,
|
| 1415 |
+
"loss": 0.3976,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.182278481012658,
|
| 1420 |
+
"grad_norm": 0.2850933541865749,
|
| 1421 |
+
"learning_rate": 5.853823673448877e-07,
|
| 1422 |
+
"loss": 0.4009,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.2126582278481015,
|
| 1427 |
+
"grad_norm": 0.29495009440796316,
|
| 1428 |
+
"learning_rate": 5.33843247095659e-07,
|
| 1429 |
+
"loss": 0.3956,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.243037974683545,
|
| 1434 |
+
"grad_norm": 0.26822212764266967,
|
| 1435 |
+
"learning_rate": 4.846165775385459e-07,
|
| 1436 |
+
"loss": 0.3896,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.273417721518987,
|
| 1441 |
+
"grad_norm": 0.26953773095218353,
|
| 1442 |
+
"learning_rate": 4.3771438406559173e-07,
|
| 1443 |
+
"loss": 0.3782,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.30379746835443,
|
| 1448 |
+
"grad_norm": 0.26810105671294315,
|
| 1449 |
+
"learning_rate": 3.931481242315993e-07,
|
| 1450 |
+
"loss": 0.3892,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.334177215189873,
|
| 1455 |
+
"grad_norm": 0.2950226684361474,
|
| 1456 |
+
"learning_rate": 3.5092868495520294e-07,
|
| 1457 |
+
"loss": 0.3807,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.364556962025317,
|
| 1462 |
+
"grad_norm": 0.2681133556168342,
|
| 1463 |
+
"learning_rate": 3.110663798593616e-07,
|
| 1464 |
+
"loss": 0.3988,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.39493670886076,
|
| 1469 |
+
"grad_norm": 0.2670123595381677,
|
| 1470 |
+
"learning_rate": 2.735709467518699e-07,
|
| 1471 |
+
"loss": 0.3737,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.425316455696202,
|
| 1476 |
+
"grad_norm": 0.2631719335348355,
|
| 1477 |
+
"learning_rate": 2.384515452465475e-07,
|
| 1478 |
+
"loss": 0.3674,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.455696202531645,
|
| 1483 |
+
"grad_norm": 0.2654377798436416,
|
| 1484 |
+
"learning_rate": 2.0571675452567997e-07,
|
| 1485 |
+
"loss": 0.3816,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.4860759493670885,
|
| 1490 |
+
"grad_norm": 0.2741124110494854,
|
| 1491 |
+
"learning_rate": 1.7537457124423896e-07,
|
| 1492 |
+
"loss": 0.3987,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.516455696202532,
|
| 1497 |
+
"grad_norm": 0.2558742486931433,
|
| 1498 |
+
"learning_rate": 1.474324075764111e-07,
|
| 1499 |
+
"loss": 0.3914,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.546835443037975,
|
| 1504 |
+
"grad_norm": 0.2626289255712206,
|
| 1505 |
+
"learning_rate": 1.2189708940490653e-07,
|
| 1506 |
+
"loss": 0.3719,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.577215189873417,
|
| 1511 |
+
"grad_norm": 0.2576213975257158,
|
| 1512 |
+
"learning_rate": 9.877485465349057e-08,
|
| 1513 |
+
"loss": 0.382,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.6075949367088604,
|
| 1518 |
+
"grad_norm": 0.25520510038953714,
|
| 1519 |
+
"learning_rate": 7.807135176314707e-08,
|
| 1520 |
+
"loss": 0.3755,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.637974683544304,
|
| 1525 |
+
"grad_norm": 0.26587236188019,
|
| 1526 |
+
"learning_rate": 5.979163831223988e-08,
|
| 1527 |
+
"loss": 0.4008,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.668354430379747,
|
| 1532 |
+
"grad_norm": 0.2643841861061346,
|
| 1533 |
+
"learning_rate": 4.394017978101905e-08,
|
| 1534 |
+
"loss": 0.3581,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.69873417721519,
|
| 1539 |
+
"grad_norm": 0.26560598833713933,
|
| 1540 |
+
"learning_rate": 3.0520848460765525e-08,
|
| 1541 |
+
"loss": 0.3832,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.729113924050633,
|
| 1546 |
+
"grad_norm": 0.2619146192545906,
|
| 1547 |
+
"learning_rate": 1.9536922507841227e-08,
|
| 1548 |
+
"loss": 0.3856,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.759493670886076,
|
| 1553 |
+
"grad_norm": 0.2779962742950057,
|
| 1554 |
+
"learning_rate": 1.099108514288627e-08,
|
| 1555 |
+
"loss": 0.3657,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.789873417721519,
|
| 1560 |
+
"grad_norm": 0.25682917386508564,
|
| 1561 |
+
"learning_rate": 4.885423995341088e-09,
|
| 1562 |
+
"loss": 0.3704,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.820253164556962,
|
| 1567 |
+
"grad_norm": 0.2684109452017263,
|
| 1568 |
+
"learning_rate": 1.2214305934699078e-09,
|
| 1569 |
+
"loss": 0.372,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.850632911392405,
|
| 1574 |
+
"grad_norm": 0.25032280703356524,
|
| 1575 |
+
"learning_rate": 0.0,
|
| 1576 |
+
"loss": 0.3908,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 6.850632911392405,
|
| 1581 |
+
"step": 224,
|
| 1582 |
+
"total_flos": 8.569304900394025e+17,
|
| 1583 |
+
"train_loss": 0.5346918581053615,
|
| 1584 |
+
"train_runtime": 27285.7344,
|
| 1585 |
+
"train_samples_per_second": 0.811,
|
| 1586 |
+
"train_steps_per_second": 0.008
|
| 1587 |
+
}
|
| 1588 |
+
],
|
| 1589 |
+
"logging_steps": 1,
|
| 1590 |
+
"max_steps": 224,
|
| 1591 |
+
"num_input_tokens_seen": 0,
|
| 1592 |
+
"num_train_epochs": 7,
|
| 1593 |
+
"save_steps": 500,
|
| 1594 |
+
"stateful_callbacks": {
|
| 1595 |
+
"TrainerControl": {
|
| 1596 |
+
"args": {
|
| 1597 |
+
"should_epoch_stop": false,
|
| 1598 |
+
"should_evaluate": false,
|
| 1599 |
+
"should_log": false,
|
| 1600 |
+
"should_save": true,
|
| 1601 |
+
"should_training_stop": true
|
| 1602 |
+
},
|
| 1603 |
+
"attributes": {}
|
| 1604 |
+
}
|
| 1605 |
+
},
|
| 1606 |
+
"total_flos": 8.569304900394025e+17,
|
| 1607 |
+
"train_batch_size": 1,
|
| 1608 |
+
"trial_name": null,
|
| 1609 |
+
"trial_params": null
|
| 1610 |
+
}
|
training_loss.png
ADDED
|