End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1659 -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: no_pipeline_math_3k
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# no_pipeline_math_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: no_pipeline_math_3k
|
|
|
|
| 16 |
|
| 17 |
# no_pipeline_math_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/no_pipeline_math_3k dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.0,
|
| 3 |
+
"total_flos": 2.2511058919253606e+17,
|
| 4 |
+
"train_loss": 0.25922639719593576,
|
| 5 |
+
"train_runtime": 5246.8102,
|
| 6 |
+
"train_samples_per_second": 4.216,
|
| 7 |
+
"train_steps_per_second": 0.044
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.0,
|
| 3 |
+
"total_flos": 2.2511058919253606e+17,
|
| 4 |
+
"train_loss": 0.25922639719593576,
|
| 5 |
+
"train_runtime": 5246.8102,
|
| 6 |
+
"train_samples_per_second": 4.216,
|
| 7 |
+
"train_steps_per_second": 0.044
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1659 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 7.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 231,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.030303030303030304,
|
| 13 |
+
"grad_norm": 6.955285911020234,
|
| 14 |
+
"learning_rate": 8.333333333333333e-07,
|
| 15 |
+
"loss": 0.8699,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06060606060606061,
|
| 20 |
+
"grad_norm": 7.441039178595325,
|
| 21 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 22 |
+
"loss": 0.9004,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09090909090909091,
|
| 27 |
+
"grad_norm": 7.109328275318153,
|
| 28 |
+
"learning_rate": 2.5e-06,
|
| 29 |
+
"loss": 0.8626,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12121212121212122,
|
| 34 |
+
"grad_norm": 6.588724752631108,
|
| 35 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 36 |
+
"loss": 0.8431,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.15151515151515152,
|
| 41 |
+
"grad_norm": 5.373513004603834,
|
| 42 |
+
"learning_rate": 4.166666666666667e-06,
|
| 43 |
+
"loss": 0.8301,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18181818181818182,
|
| 48 |
+
"grad_norm": 3.4043853998859066,
|
| 49 |
+
"learning_rate": 5e-06,
|
| 50 |
+
"loss": 0.7509,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21212121212121213,
|
| 55 |
+
"grad_norm": 2.8120521919312527,
|
| 56 |
+
"learning_rate": 5.833333333333334e-06,
|
| 57 |
+
"loss": 0.7263,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24242424242424243,
|
| 62 |
+
"grad_norm": 3.711403109617059,
|
| 63 |
+
"learning_rate": 6.666666666666667e-06,
|
| 64 |
+
"loss": 0.7106,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2727272727272727,
|
| 69 |
+
"grad_norm": 3.8596624615612045,
|
| 70 |
+
"learning_rate": 7.500000000000001e-06,
|
| 71 |
+
"loss": 0.7277,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.30303030303030304,
|
| 76 |
+
"grad_norm": 3.538922287541338,
|
| 77 |
+
"learning_rate": 8.333333333333334e-06,
|
| 78 |
+
"loss": 0.6762,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3333333333333333,
|
| 83 |
+
"grad_norm": 3.3803770329911265,
|
| 84 |
+
"learning_rate": 9.166666666666666e-06,
|
| 85 |
+
"loss": 0.6236,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.36363636363636365,
|
| 90 |
+
"grad_norm": 3.0818855435472,
|
| 91 |
+
"learning_rate": 1e-05,
|
| 92 |
+
"loss": 0.6127,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3939393939393939,
|
| 97 |
+
"grad_norm": 2.3753023124660007,
|
| 98 |
+
"learning_rate": 1.0833333333333334e-05,
|
| 99 |
+
"loss": 0.5891,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.42424242424242425,
|
| 104 |
+
"grad_norm": 1.8745417982082844,
|
| 105 |
+
"learning_rate": 1.1666666666666668e-05,
|
| 106 |
+
"loss": 0.5844,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.45454545454545453,
|
| 111 |
+
"grad_norm": 2.3742239969042824,
|
| 112 |
+
"learning_rate": 1.25e-05,
|
| 113 |
+
"loss": 0.6243,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.48484848484848486,
|
| 118 |
+
"grad_norm": 2.7538975200597626,
|
| 119 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 120 |
+
"loss": 0.5731,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5151515151515151,
|
| 125 |
+
"grad_norm": 1.7182833939411697,
|
| 126 |
+
"learning_rate": 1.416666666666667e-05,
|
| 127 |
+
"loss": 0.5813,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5454545454545454,
|
| 132 |
+
"grad_norm": 1.087070578452013,
|
| 133 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 134 |
+
"loss": 0.5459,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5757575757575758,
|
| 139 |
+
"grad_norm": 1.5835128684193447,
|
| 140 |
+
"learning_rate": 1.5833333333333333e-05,
|
| 141 |
+
"loss": 0.5158,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6060606060606061,
|
| 146 |
+
"grad_norm": 1.5234934050673163,
|
| 147 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 148 |
+
"loss": 0.5485,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6363636363636364,
|
| 153 |
+
"grad_norm": 0.9976574153885728,
|
| 154 |
+
"learning_rate": 1.7500000000000002e-05,
|
| 155 |
+
"loss": 0.5289,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6666666666666666,
|
| 160 |
+
"grad_norm": 1.0939314536295377,
|
| 161 |
+
"learning_rate": 1.8333333333333333e-05,
|
| 162 |
+
"loss": 0.5455,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.696969696969697,
|
| 167 |
+
"grad_norm": 1.150968759098677,
|
| 168 |
+
"learning_rate": 1.916666666666667e-05,
|
| 169 |
+
"loss": 0.5565,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7272727272727273,
|
| 174 |
+
"grad_norm": 1.0141197421117834,
|
| 175 |
+
"learning_rate": 2e-05,
|
| 176 |
+
"loss": 0.5312,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7575757575757576,
|
| 181 |
+
"grad_norm": 0.8036309707430053,
|
| 182 |
+
"learning_rate": 1.999884834944106e-05,
|
| 183 |
+
"loss": 0.5184,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7878787878787878,
|
| 188 |
+
"grad_norm": 0.8929941942528171,
|
| 189 |
+
"learning_rate": 1.9995393663024054e-05,
|
| 190 |
+
"loss": 0.5327,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8181818181818182,
|
| 195 |
+
"grad_norm": 0.7649914065820512,
|
| 196 |
+
"learning_rate": 1.9989636736467278e-05,
|
| 197 |
+
"loss": 0.4815,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8484848484848485,
|
| 202 |
+
"grad_norm": 0.7482399920697894,
|
| 203 |
+
"learning_rate": 1.9981578895764272e-05,
|
| 204 |
+
"loss": 0.5123,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8787878787878788,
|
| 209 |
+
"grad_norm": 0.8152884514519559,
|
| 210 |
+
"learning_rate": 1.9971221996878395e-05,
|
| 211 |
+
"loss": 0.5178,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9090909090909091,
|
| 216 |
+
"grad_norm": 0.725342647598433,
|
| 217 |
+
"learning_rate": 1.9958568425315316e-05,
|
| 218 |
+
"loss": 0.4812,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9393939393939394,
|
| 223 |
+
"grad_norm": 0.7441563173742102,
|
| 224 |
+
"learning_rate": 1.9943621095573588e-05,
|
| 225 |
+
"loss": 0.5156,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9696969696969697,
|
| 230 |
+
"grad_norm": 0.7752154845823396,
|
| 231 |
+
"learning_rate": 1.9926383450473344e-05,
|
| 232 |
+
"loss": 0.5164,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0,
|
| 237 |
+
"grad_norm": 0.7198494006916166,
|
| 238 |
+
"learning_rate": 1.9906859460363307e-05,
|
| 239 |
+
"loss": 0.4778,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0303030303030303,
|
| 244 |
+
"grad_norm": 0.8352454069355456,
|
| 245 |
+
"learning_rate": 1.9885053622206305e-05,
|
| 246 |
+
"loss": 0.4217,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0606060606060606,
|
| 251 |
+
"grad_norm": 0.9660187626710007,
|
| 252 |
+
"learning_rate": 1.986097095854347e-05,
|
| 253 |
+
"loss": 0.4161,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0909090909090908,
|
| 258 |
+
"grad_norm": 0.7071157149331548,
|
| 259 |
+
"learning_rate": 1.9834617016337424e-05,
|
| 260 |
+
"loss": 0.4162,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.121212121212121,
|
| 265 |
+
"grad_norm": 1.1005892206492136,
|
| 266 |
+
"learning_rate": 1.9805997865694616e-05,
|
| 267 |
+
"loss": 0.4013,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1515151515151516,
|
| 272 |
+
"grad_norm": 0.9334754936616282,
|
| 273 |
+
"learning_rate": 1.9775120098467212e-05,
|
| 274 |
+
"loss": 0.4249,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.1818181818181819,
|
| 279 |
+
"grad_norm": 0.7816271304035008,
|
| 280 |
+
"learning_rate": 1.9741990826734793e-05,
|
| 281 |
+
"loss": 0.4075,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2121212121212122,
|
| 286 |
+
"grad_norm": 0.6784833145239036,
|
| 287 |
+
"learning_rate": 1.970661768116622e-05,
|
| 288 |
+
"loss": 0.437,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2424242424242424,
|
| 293 |
+
"grad_norm": 0.6513950894158883,
|
| 294 |
+
"learning_rate": 1.9669008809262064e-05,
|
| 295 |
+
"loss": 0.4291,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.2727272727272727,
|
| 300 |
+
"grad_norm": 0.774415542259264,
|
| 301 |
+
"learning_rate": 1.9629172873477995e-05,
|
| 302 |
+
"loss": 0.3911,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.303030303030303,
|
| 307 |
+
"grad_norm": 0.679421098074206,
|
| 308 |
+
"learning_rate": 1.9587119049229558e-05,
|
| 309 |
+
"loss": 0.4229,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3333333333333333,
|
| 314 |
+
"grad_norm": 0.6759155017266889,
|
| 315 |
+
"learning_rate": 1.954285702277879e-05,
|
| 316 |
+
"loss": 0.4081,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.3636363636363638,
|
| 321 |
+
"grad_norm": 0.6932645626121937,
|
| 322 |
+
"learning_rate": 1.9496396989003195e-05,
|
| 323 |
+
"loss": 0.4261,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.393939393939394,
|
| 328 |
+
"grad_norm": 0.6433462051921326,
|
| 329 |
+
"learning_rate": 1.944774964904754e-05,
|
| 330 |
+
"loss": 0.3834,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4242424242424243,
|
| 335 |
+
"grad_norm": 0.6920823864439841,
|
| 336 |
+
"learning_rate": 1.9396926207859085e-05,
|
| 337 |
+
"loss": 0.3941,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4545454545454546,
|
| 342 |
+
"grad_norm": 0.5869584687603564,
|
| 343 |
+
"learning_rate": 1.9343938371606714e-05,
|
| 344 |
+
"loss": 0.4,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.4848484848484849,
|
| 349 |
+
"grad_norm": 0.7652602132876345,
|
| 350 |
+
"learning_rate": 1.9288798344984673e-05,
|
| 351 |
+
"loss": 0.4026,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5151515151515151,
|
| 356 |
+
"grad_norm": 0.6806717081689967,
|
| 357 |
+
"learning_rate": 1.9231518828401458e-05,
|
| 358 |
+
"loss": 0.3959,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5454545454545454,
|
| 363 |
+
"grad_norm": 0.7303868791316439,
|
| 364 |
+
"learning_rate": 1.917211301505453e-05,
|
| 365 |
+
"loss": 0.389,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.5757575757575757,
|
| 370 |
+
"grad_norm": 0.7242459037753521,
|
| 371 |
+
"learning_rate": 1.911059458789152e-05,
|
| 372 |
+
"loss": 0.3946,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.606060606060606,
|
| 377 |
+
"grad_norm": 0.6271112843619893,
|
| 378 |
+
"learning_rate": 1.9046977716458627e-05,
|
| 379 |
+
"loss": 0.3632,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6363636363636362,
|
| 384 |
+
"grad_norm": 0.7737301258721643,
|
| 385 |
+
"learning_rate": 1.8981277053636963e-05,
|
| 386 |
+
"loss": 0.402,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.6666666666666665,
|
| 391 |
+
"grad_norm": 0.6845424366640315,
|
| 392 |
+
"learning_rate": 1.891350773226754e-05,
|
| 393 |
+
"loss": 0.4013,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.696969696969697,
|
| 398 |
+
"grad_norm": 0.5567017944032856,
|
| 399 |
+
"learning_rate": 1.8843685361665724e-05,
|
| 400 |
+
"loss": 0.3934,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7272727272727273,
|
| 405 |
+
"grad_norm": 0.7270095802562017,
|
| 406 |
+
"learning_rate": 1.8771826024025944e-05,
|
| 407 |
+
"loss": 0.3851,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.7575757575757576,
|
| 412 |
+
"grad_norm": 0.5612102341406529,
|
| 413 |
+
"learning_rate": 1.8697946270717468e-05,
|
| 414 |
+
"loss": 0.417,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.7878787878787878,
|
| 419 |
+
"grad_norm": 0.5587450981440957,
|
| 420 |
+
"learning_rate": 1.8622063118472135e-05,
|
| 421 |
+
"loss": 0.404,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8181818181818183,
|
| 426 |
+
"grad_norm": 0.6089274247399353,
|
| 427 |
+
"learning_rate": 1.8544194045464888e-05,
|
| 428 |
+
"loss": 0.3957,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.8484848484848486,
|
| 433 |
+
"grad_norm": 0.5498996754551172,
|
| 434 |
+
"learning_rate": 1.8464356987288012e-05,
|
| 435 |
+
"loss": 0.3944,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.878787878787879,
|
| 440 |
+
"grad_norm": 0.6378506474319554,
|
| 441 |
+
"learning_rate": 1.8382570332820045e-05,
|
| 442 |
+
"loss": 0.4062,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9090909090909092,
|
| 447 |
+
"grad_norm": 0.6820699013876874,
|
| 448 |
+
"learning_rate": 1.8298852919990254e-05,
|
| 449 |
+
"loss": 0.4109,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9393939393939394,
|
| 454 |
+
"grad_norm": 0.6613751546427997,
|
| 455 |
+
"learning_rate": 1.821322403143969e-05,
|
| 456 |
+
"loss": 0.3924,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.9696969696969697,
|
| 461 |
+
"grad_norm": 0.5868159473906754,
|
| 462 |
+
"learning_rate": 1.812570339007983e-05,
|
| 463 |
+
"loss": 0.3834,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.0,
|
| 468 |
+
"grad_norm": 0.5985737313520647,
|
| 469 |
+
"learning_rate": 1.8036311154549783e-05,
|
| 470 |
+
"loss": 0.4171,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.0303030303030303,
|
| 475 |
+
"grad_norm": 0.8195740545881052,
|
| 476 |
+
"learning_rate": 1.7945067914573147e-05,
|
| 477 |
+
"loss": 0.2688,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.0606060606060606,
|
| 482 |
+
"grad_norm": 0.7062499554065702,
|
| 483 |
+
"learning_rate": 1.7851994686215592e-05,
|
| 484 |
+
"loss": 0.3102,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.090909090909091,
|
| 489 |
+
"grad_norm": 0.7980134577908272,
|
| 490 |
+
"learning_rate": 1.77571129070442e-05,
|
| 491 |
+
"loss": 0.2742,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.121212121212121,
|
| 496 |
+
"grad_norm": 0.5686351282877364,
|
| 497 |
+
"learning_rate": 1.766044443118978e-05,
|
| 498 |
+
"loss": 0.2623,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.1515151515151514,
|
| 503 |
+
"grad_norm": 0.7877937175685785,
|
| 504 |
+
"learning_rate": 1.7562011524313187e-05,
|
| 505 |
+
"loss": 0.3029,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.1818181818181817,
|
| 510 |
+
"grad_norm": 0.5633881728706212,
|
| 511 |
+
"learning_rate": 1.7461836858476858e-05,
|
| 512 |
+
"loss": 0.3009,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.212121212121212,
|
| 517 |
+
"grad_norm": 0.6031263751909894,
|
| 518 |
+
"learning_rate": 1.7359943506922775e-05,
|
| 519 |
+
"loss": 0.3018,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.242424242424242,
|
| 524 |
+
"grad_norm": 0.6176095540963793,
|
| 525 |
+
"learning_rate": 1.725635493875799e-05,
|
| 526 |
+
"loss": 0.2914,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.2727272727272725,
|
| 531 |
+
"grad_norm": 0.5147872846790723,
|
| 532 |
+
"learning_rate": 1.7151095013548996e-05,
|
| 533 |
+
"loss": 0.2887,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.303030303030303,
|
| 538 |
+
"grad_norm": 0.6052964176378741,
|
| 539 |
+
"learning_rate": 1.7044187975826126e-05,
|
| 540 |
+
"loss": 0.3025,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3333333333333335,
|
| 545 |
+
"grad_norm": 0.538987033559335,
|
| 546 |
+
"learning_rate": 1.693565844949933e-05,
|
| 547 |
+
"loss": 0.2625,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.3636363636363638,
|
| 552 |
+
"grad_norm": 0.5778684348276075,
|
| 553 |
+
"learning_rate": 1.6825531432186545e-05,
|
| 554 |
+
"loss": 0.264,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.393939393939394,
|
| 559 |
+
"grad_norm": 0.5481681130066359,
|
| 560 |
+
"learning_rate": 1.671383228945597e-05,
|
| 561 |
+
"loss": 0.2703,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4242424242424243,
|
| 566 |
+
"grad_norm": 0.6432693921781116,
|
| 567 |
+
"learning_rate": 1.6600586748983642e-05,
|
| 568 |
+
"loss": 0.2837,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.4545454545454546,
|
| 573 |
+
"grad_norm": 0.5864166237413477,
|
| 574 |
+
"learning_rate": 1.648582089462756e-05,
|
| 575 |
+
"loss": 0.2755,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.484848484848485,
|
| 580 |
+
"grad_norm": 0.5966611067522237,
|
| 581 |
+
"learning_rate": 1.6369561160419783e-05,
|
| 582 |
+
"loss": 0.2598,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.515151515151515,
|
| 587 |
+
"grad_norm": 0.5696369232335121,
|
| 588 |
+
"learning_rate": 1.625183432447789e-05,
|
| 589 |
+
"loss": 0.2803,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.5454545454545454,
|
| 594 |
+
"grad_norm": 0.5231159610229735,
|
| 595 |
+
"learning_rate": 1.6132667502837164e-05,
|
| 596 |
+
"loss": 0.2696,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.5757575757575757,
|
| 601 |
+
"grad_norm": 0.5182632826584535,
|
| 602 |
+
"learning_rate": 1.6012088143204953e-05,
|
| 603 |
+
"loss": 0.2836,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.606060606060606,
|
| 608 |
+
"grad_norm": 0.5890186789657982,
|
| 609 |
+
"learning_rate": 1.589012401863864e-05,
|
| 610 |
+
"loss": 0.2784,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.6363636363636362,
|
| 615 |
+
"grad_norm": 0.567710505482636,
|
| 616 |
+
"learning_rate": 1.5766803221148676e-05,
|
| 617 |
+
"loss": 0.2618,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.6666666666666665,
|
| 622 |
+
"grad_norm": 0.5572197103930109,
|
| 623 |
+
"learning_rate": 1.5642154155228124e-05,
|
| 624 |
+
"loss": 0.2729,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.6969696969696972,
|
| 629 |
+
"grad_norm": 0.6350111818251795,
|
| 630 |
+
"learning_rate": 1.5516205531310272e-05,
|
| 631 |
+
"loss": 0.2986,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.7272727272727275,
|
| 636 |
+
"grad_norm": 0.5899434133869678,
|
| 637 |
+
"learning_rate": 1.538898635915576e-05,
|
| 638 |
+
"loss": 0.2655,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.757575757575758,
|
| 643 |
+
"grad_norm": 0.5278431909713808,
|
| 644 |
+
"learning_rate": 1.526052594117071e-05,
|
| 645 |
+
"loss": 0.2929,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.787878787878788,
|
| 650 |
+
"grad_norm": 0.6136960878846716,
|
| 651 |
+
"learning_rate": 1.513085386565758e-05,
|
| 652 |
+
"loss": 0.2745,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8181818181818183,
|
| 657 |
+
"grad_norm": 0.5793084383935346,
|
| 658 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 659 |
+
"loss": 0.2866,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.8484848484848486,
|
| 664 |
+
"grad_norm": 0.5498377794459244,
|
| 665 |
+
"learning_rate": 1.4867994483783485e-05,
|
| 666 |
+
"loss": 0.2837,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.878787878787879,
|
| 671 |
+
"grad_norm": 0.5988359010382698,
|
| 672 |
+
"learning_rate": 1.4734867721853341e-05,
|
| 673 |
+
"loss": 0.2702,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.909090909090909,
|
| 678 |
+
"grad_norm": 0.5268319798838377,
|
| 679 |
+
"learning_rate": 1.4600650377311523e-05,
|
| 680 |
+
"loss": 0.2472,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.9393939393939394,
|
| 685 |
+
"grad_norm": 0.5022165627213556,
|
| 686 |
+
"learning_rate": 1.4465373364454001e-05,
|
| 687 |
+
"loss": 0.2704,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.9696969696969697,
|
| 692 |
+
"grad_norm": 0.5761546585376526,
|
| 693 |
+
"learning_rate": 1.4329067841650274e-05,
|
| 694 |
+
"loss": 0.2659,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.0,
|
| 699 |
+
"grad_norm": 0.46782846192930244,
|
| 700 |
+
"learning_rate": 1.4191765204166643e-05,
|
| 701 |
+
"loss": 0.2479,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.0303030303030303,
|
| 706 |
+
"grad_norm": 0.7129961579741467,
|
| 707 |
+
"learning_rate": 1.4053497076934948e-05,
|
| 708 |
+
"loss": 0.231,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.0606060606060606,
|
| 713 |
+
"grad_norm": 0.49057301164831774,
|
| 714 |
+
"learning_rate": 1.3914295307268396e-05,
|
| 715 |
+
"loss": 0.1775,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.090909090909091,
|
| 720 |
+
"grad_norm": 0.7759935461555386,
|
| 721 |
+
"learning_rate": 1.3774191957526144e-05,
|
| 722 |
+
"loss": 0.225,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.121212121212121,
|
| 727 |
+
"grad_norm": 0.706420358436487,
|
| 728 |
+
"learning_rate": 1.3633219297728415e-05,
|
| 729 |
+
"loss": 0.2065,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.1515151515151514,
|
| 734 |
+
"grad_norm": 0.5504014857027489,
|
| 735 |
+
"learning_rate": 1.3491409798123687e-05,
|
| 736 |
+
"loss": 0.1936,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.1818181818181817,
|
| 741 |
+
"grad_norm": 0.5706969442123296,
|
| 742 |
+
"learning_rate": 1.3348796121709862e-05,
|
| 743 |
+
"loss": 0.1823,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.212121212121212,
|
| 748 |
+
"grad_norm": 0.6173947476088453,
|
| 749 |
+
"learning_rate": 1.3205411116710973e-05,
|
| 750 |
+
"loss": 0.1865,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.242424242424242,
|
| 755 |
+
"grad_norm": 0.5484188539031339,
|
| 756 |
+
"learning_rate": 1.3061287809011243e-05,
|
| 757 |
+
"loss": 0.196,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.2727272727272725,
|
| 762 |
+
"grad_norm": 0.5728717946594112,
|
| 763 |
+
"learning_rate": 1.291645939454825e-05,
|
| 764 |
+
"loss": 0.2,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.303030303030303,
|
| 769 |
+
"grad_norm": 0.5273279709766339,
|
| 770 |
+
"learning_rate": 1.277095923166689e-05,
|
| 771 |
+
"loss": 0.1913,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.3333333333333335,
|
| 776 |
+
"grad_norm": 0.5594488832775165,
|
| 777 |
+
"learning_rate": 1.2624820833435939e-05,
|
| 778 |
+
"loss": 0.1795,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.3636363636363638,
|
| 783 |
+
"grad_norm": 0.5080147025861013,
|
| 784 |
+
"learning_rate": 1.2478077859929e-05,
|
| 785 |
+
"loss": 0.2025,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.393939393939394,
|
| 790 |
+
"grad_norm": 0.541143047020217,
|
| 791 |
+
"learning_rate": 1.2330764110471567e-05,
|
| 792 |
+
"loss": 0.1864,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.4242424242424243,
|
| 797 |
+
"grad_norm": 0.562993920743498,
|
| 798 |
+
"learning_rate": 1.2182913515856016e-05,
|
| 799 |
+
"loss": 0.2132,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.4545454545454546,
|
| 804 |
+
"grad_norm": 0.5978483575805258,
|
| 805 |
+
"learning_rate": 1.2034560130526341e-05,
|
| 806 |
+
"loss": 0.1869,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.484848484848485,
|
| 811 |
+
"grad_norm": 0.5226554327441967,
|
| 812 |
+
"learning_rate": 1.1885738124734359e-05,
|
| 813 |
+
"loss": 0.2048,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.515151515151515,
|
| 818 |
+
"grad_norm": 0.5387548668211881,
|
| 819 |
+
"learning_rate": 1.1736481776669307e-05,
|
| 820 |
+
"loss": 0.1764,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.5454545454545454,
|
| 825 |
+
"grad_norm": 0.5256689543158756,
|
| 826 |
+
"learning_rate": 1.1586825464562515e-05,
|
| 827 |
+
"loss": 0.1923,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.5757575757575757,
|
| 832 |
+
"grad_norm": 0.5432943472230537,
|
| 833 |
+
"learning_rate": 1.1436803658769082e-05,
|
| 834 |
+
"loss": 0.2146,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.606060606060606,
|
| 839 |
+
"grad_norm": 0.5120326182151331,
|
| 840 |
+
"learning_rate": 1.1286450913828313e-05,
|
| 841 |
+
"loss": 0.1865,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.6363636363636362,
|
| 846 |
+
"grad_norm": 0.4795068148738163,
|
| 847 |
+
"learning_rate": 1.113580186050475e-05,
|
| 848 |
+
"loss": 0.1683,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.6666666666666665,
|
| 853 |
+
"grad_norm": 0.45214877045521085,
|
| 854 |
+
"learning_rate": 1.0984891197811686e-05,
|
| 855 |
+
"loss": 0.1964,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.6969696969696972,
|
| 860 |
+
"grad_norm": 0.5089679969091959,
|
| 861 |
+
"learning_rate": 1.0833753685018935e-05,
|
| 862 |
+
"loss": 0.1801,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.7272727272727275,
|
| 867 |
+
"grad_norm": 0.4770511601662899,
|
| 868 |
+
"learning_rate": 1.0682424133646712e-05,
|
| 869 |
+
"loss": 0.194,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.757575757575758,
|
| 874 |
+
"grad_norm": 0.49018352632265705,
|
| 875 |
+
"learning_rate": 1.0530937399447496e-05,
|
| 876 |
+
"loss": 0.2023,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.787878787878788,
|
| 881 |
+
"grad_norm": 0.49891417126869536,
|
| 882 |
+
"learning_rate": 1.0379328374377715e-05,
|
| 883 |
+
"loss": 0.1875,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.8181818181818183,
|
| 888 |
+
"grad_norm": 0.4757455650970373,
|
| 889 |
+
"learning_rate": 1.0227631978561057e-05,
|
| 890 |
+
"loss": 0.1871,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.8484848484848486,
|
| 895 |
+
"grad_norm": 0.44535590636291156,
|
| 896 |
+
"learning_rate": 1.0075883152245334e-05,
|
| 897 |
+
"loss": 0.1965,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.878787878787879,
|
| 902 |
+
"grad_norm": 0.45879909350398235,
|
| 903 |
+
"learning_rate": 9.92411684775467e-06,
|
| 904 |
+
"loss": 0.175,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.909090909090909,
|
| 909 |
+
"grad_norm": 0.47744572911506666,
|
| 910 |
+
"learning_rate": 9.772368021438943e-06,
|
| 911 |
+
"loss": 0.1948,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.9393939393939394,
|
| 916 |
+
"grad_norm": 0.48094392615107573,
|
| 917 |
+
"learning_rate": 9.620671625622287e-06,
|
| 918 |
+
"loss": 0.1834,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.9696969696969697,
|
| 923 |
+
"grad_norm": 0.4828473685456728,
|
| 924 |
+
"learning_rate": 9.469062600552509e-06,
|
| 925 |
+
"loss": 0.1738,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.0,
|
| 930 |
+
"grad_norm": 0.4718905282545995,
|
| 931 |
+
"learning_rate": 9.317575866353293e-06,
|
| 932 |
+
"loss": 0.1941,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.03030303030303,
|
| 937 |
+
"grad_norm": 0.5591603369489075,
|
| 938 |
+
"learning_rate": 9.166246314981066e-06,
|
| 939 |
+
"loss": 0.1367,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.0606060606060606,
|
| 944 |
+
"grad_norm": 0.4735500518994881,
|
| 945 |
+
"learning_rate": 9.015108802188314e-06,
|
| 946 |
+
"loss": 0.1309,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.090909090909091,
|
| 951 |
+
"grad_norm": 0.3945162964815393,
|
| 952 |
+
"learning_rate": 8.86419813949525e-06,
|
| 953 |
+
"loss": 0.149,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.121212121212121,
|
| 958 |
+
"grad_norm": 0.5335504030219297,
|
| 959 |
+
"learning_rate": 8.71354908617169e-06,
|
| 960 |
+
"loss": 0.1533,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.151515151515151,
|
| 965 |
+
"grad_norm": 0.554966267786691,
|
| 966 |
+
"learning_rate": 8.56319634123092e-06,
|
| 967 |
+
"loss": 0.1437,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.181818181818182,
|
| 972 |
+
"grad_norm": 0.48898997922032944,
|
| 973 |
+
"learning_rate": 8.413174535437486e-06,
|
| 974 |
+
"loss": 0.1489,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.212121212121212,
|
| 979 |
+
"grad_norm": 0.3851954143671289,
|
| 980 |
+
"learning_rate": 8.263518223330698e-06,
|
| 981 |
+
"loss": 0.1261,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.242424242424242,
|
| 986 |
+
"grad_norm": 0.45864789640280124,
|
| 987 |
+
"learning_rate": 8.114261875265643e-06,
|
| 988 |
+
"loss": 0.1255,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.2727272727272725,
|
| 993 |
+
"grad_norm": 0.4527524193004519,
|
| 994 |
+
"learning_rate": 7.965439869473664e-06,
|
| 995 |
+
"loss": 0.1543,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.303030303030303,
|
| 1000 |
+
"grad_norm": 0.4073999225821938,
|
| 1001 |
+
"learning_rate": 7.817086484143987e-06,
|
| 1002 |
+
"loss": 0.139,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.333333333333333,
|
| 1007 |
+
"grad_norm": 0.39261389494801846,
|
| 1008 |
+
"learning_rate": 7.669235889528436e-06,
|
| 1009 |
+
"loss": 0.1257,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.363636363636363,
|
| 1014 |
+
"grad_norm": 0.4575588826846762,
|
| 1015 |
+
"learning_rate": 7.521922140071003e-06,
|
| 1016 |
+
"loss": 0.1343,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.393939393939394,
|
| 1021 |
+
"grad_norm": 0.43876832508333513,
|
| 1022 |
+
"learning_rate": 7.375179166564062e-06,
|
| 1023 |
+
"loss": 0.1172,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.424242424242424,
|
| 1028 |
+
"grad_norm": 0.41691064173333414,
|
| 1029 |
+
"learning_rate": 7.2290407683331154e-06,
|
| 1030 |
+
"loss": 0.1321,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.454545454545454,
|
| 1035 |
+
"grad_norm": 0.40764985584259983,
|
| 1036 |
+
"learning_rate": 7.0835406054517505e-06,
|
| 1037 |
+
"loss": 0.1361,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.484848484848484,
|
| 1042 |
+
"grad_norm": 0.37004839656197375,
|
| 1043 |
+
"learning_rate": 6.93871219098876e-06,
|
| 1044 |
+
"loss": 0.0898,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.515151515151516,
|
| 1049 |
+
"grad_norm": 0.3983250580069294,
|
| 1050 |
+
"learning_rate": 6.79458888328903e-06,
|
| 1051 |
+
"loss": 0.1315,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.545454545454545,
|
| 1056 |
+
"grad_norm": 0.3967068739468456,
|
| 1057 |
+
"learning_rate": 6.651203878290139e-06,
|
| 1058 |
+
"loss": 0.1323,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.575757575757576,
|
| 1063 |
+
"grad_norm": 0.42514016375192504,
|
| 1064 |
+
"learning_rate": 6.508590201876317e-06,
|
| 1065 |
+
"loss": 0.1454,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.606060606060606,
|
| 1070 |
+
"grad_norm": 0.4113082857808772,
|
| 1071 |
+
"learning_rate": 6.366780702271589e-06,
|
| 1072 |
+
"loss": 0.1082,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.636363636363637,
|
| 1077 |
+
"grad_norm": 0.3901457031012471,
|
| 1078 |
+
"learning_rate": 6.225808042473857e-06,
|
| 1079 |
+
"loss": 0.1284,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.666666666666667,
|
| 1084 |
+
"grad_norm": 0.36488454061487613,
|
| 1085 |
+
"learning_rate": 6.085704692731609e-06,
|
| 1086 |
+
"loss": 0.139,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.696969696969697,
|
| 1091 |
+
"grad_norm": 0.41745743900602106,
|
| 1092 |
+
"learning_rate": 5.946502923065054e-06,
|
| 1093 |
+
"loss": 0.1544,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.7272727272727275,
|
| 1098 |
+
"grad_norm": 0.3756471474402877,
|
| 1099 |
+
"learning_rate": 5.8082347958333625e-06,
|
| 1100 |
+
"loss": 0.1275,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.757575757575758,
|
| 1105 |
+
"grad_norm": 0.36520019031697765,
|
| 1106 |
+
"learning_rate": 5.670932158349732e-06,
|
| 1107 |
+
"loss": 0.1117,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.787878787878788,
|
| 1112 |
+
"grad_norm": 0.38832688927383313,
|
| 1113 |
+
"learning_rate": 5.534626635546e-06,
|
| 1114 |
+
"loss": 0.1312,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.818181818181818,
|
| 1119 |
+
"grad_norm": 0.40539798327302723,
|
| 1120 |
+
"learning_rate": 5.399349622688479e-06,
|
| 1121 |
+
"loss": 0.1267,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.848484848484849,
|
| 1126 |
+
"grad_norm": 0.37779606825387413,
|
| 1127 |
+
"learning_rate": 5.2651322781466606e-06,
|
| 1128 |
+
"loss": 0.146,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.878787878787879,
|
| 1133 |
+
"grad_norm": 0.35385327091645463,
|
| 1134 |
+
"learning_rate": 5.132005516216512e-06,
|
| 1135 |
+
"loss": 0.1607,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.909090909090909,
|
| 1140 |
+
"grad_norm": 0.35446080294435106,
|
| 1141 |
+
"learning_rate": 5.000000000000003e-06,
|
| 1142 |
+
"loss": 0.1361,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.9393939393939394,
|
| 1147 |
+
"grad_norm": 0.37164703413823025,
|
| 1148 |
+
"learning_rate": 4.869146134342426e-06,
|
| 1149 |
+
"loss": 0.1266,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.96969696969697,
|
| 1154 |
+
"grad_norm": 0.3730645230829621,
|
| 1155 |
+
"learning_rate": 4.739474058829288e-06,
|
| 1156 |
+
"loss": 0.1151,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.0,
|
| 1161 |
+
"grad_norm": 0.3677955180366353,
|
| 1162 |
+
"learning_rate": 4.611013640844245e-06,
|
| 1163 |
+
"loss": 0.1575,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.03030303030303,
|
| 1168 |
+
"grad_norm": 0.38659427470723917,
|
| 1169 |
+
"learning_rate": 4.483794468689728e-06,
|
| 1170 |
+
"loss": 0.0969,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.0606060606060606,
|
| 1175 |
+
"grad_norm": 0.3745826660690661,
|
| 1176 |
+
"learning_rate": 4.357845844771881e-06,
|
| 1177 |
+
"loss": 0.1062,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.090909090909091,
|
| 1182 |
+
"grad_norm": 0.30623162660117503,
|
| 1183 |
+
"learning_rate": 4.2331967788513295e-06,
|
| 1184 |
+
"loss": 0.0965,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.121212121212121,
|
| 1189 |
+
"grad_norm": 0.26668628394316596,
|
| 1190 |
+
"learning_rate": 4.109875981361363e-06,
|
| 1191 |
+
"loss": 0.089,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.151515151515151,
|
| 1196 |
+
"grad_norm": 0.343182796290794,
|
| 1197 |
+
"learning_rate": 3.987911856795047e-06,
|
| 1198 |
+
"loss": 0.1036,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.181818181818182,
|
| 1203 |
+
"grad_norm": 0.4512425489938098,
|
| 1204 |
+
"learning_rate": 3.867332497162836e-06,
|
| 1205 |
+
"loss": 0.1172,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.212121212121212,
|
| 1210 |
+
"grad_norm": 0.39415631817447083,
|
| 1211 |
+
"learning_rate": 3.748165675522113e-06,
|
| 1212 |
+
"loss": 0.1003,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.242424242424242,
|
| 1217 |
+
"grad_norm": 0.3328008003014252,
|
| 1218 |
+
"learning_rate": 3.630438839580217e-06,
|
| 1219 |
+
"loss": 0.0705,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.2727272727272725,
|
| 1224 |
+
"grad_norm": 0.3372180226738015,
|
| 1225 |
+
"learning_rate": 3.5141791053724405e-06,
|
| 1226 |
+
"loss": 0.0913,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.303030303030303,
|
| 1231 |
+
"grad_norm": 0.31914187491627916,
|
| 1232 |
+
"learning_rate": 3.399413251016359e-06,
|
| 1233 |
+
"loss": 0.1206,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.333333333333333,
|
| 1238 |
+
"grad_norm": 0.30767849508209233,
|
| 1239 |
+
"learning_rate": 3.2861677105440335e-06,
|
| 1240 |
+
"loss": 0.0903,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.363636363636363,
|
| 1245 |
+
"grad_norm": 0.3635221533322771,
|
| 1246 |
+
"learning_rate": 3.174468567813461e-06,
|
| 1247 |
+
"loss": 0.1286,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.393939393939394,
|
| 1252 |
+
"grad_norm": 0.3287109518254317,
|
| 1253 |
+
"learning_rate": 3.0643415505006733e-06,
|
| 1254 |
+
"loss": 0.1061,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.424242424242424,
|
| 1259 |
+
"grad_norm": 0.29669250344847614,
|
| 1260 |
+
"learning_rate": 2.9558120241738786e-06,
|
| 1261 |
+
"loss": 0.0877,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.454545454545454,
|
| 1266 |
+
"grad_norm": 0.292503106556702,
|
| 1267 |
+
"learning_rate": 2.8489049864510053e-06,
|
| 1268 |
+
"loss": 0.0815,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.484848484848484,
|
| 1273 |
+
"grad_norm": 0.264433565061036,
|
| 1274 |
+
"learning_rate": 2.7436450612420098e-06,
|
| 1275 |
+
"loss": 0.1071,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.515151515151516,
|
| 1280 |
+
"grad_norm": 0.3000656689348664,
|
| 1281 |
+
"learning_rate": 2.640056493077231e-06,
|
| 1282 |
+
"loss": 0.1129,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.545454545454545,
|
| 1287 |
+
"grad_norm": 0.2986195036926533,
|
| 1288 |
+
"learning_rate": 2.5381631415231455e-06,
|
| 1289 |
+
"loss": 0.1145,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.575757575757576,
|
| 1294 |
+
"grad_norm": 0.28286786606628167,
|
| 1295 |
+
"learning_rate": 2.4379884756868167e-06,
|
| 1296 |
+
"loss": 0.1016,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.606060606060606,
|
| 1301 |
+
"grad_norm": 0.2855501208614676,
|
| 1302 |
+
"learning_rate": 2.339555568810221e-06,
|
| 1303 |
+
"loss": 0.0848,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.636363636363637,
|
| 1308 |
+
"grad_norm": 0.30332646259992035,
|
| 1309 |
+
"learning_rate": 2.2428870929558012e-06,
|
| 1310 |
+
"loss": 0.0768,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.666666666666667,
|
| 1315 |
+
"grad_norm": 0.274931337533345,
|
| 1316 |
+
"learning_rate": 2.1480053137844115e-06,
|
| 1317 |
+
"loss": 0.0985,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.696969696969697,
|
| 1322 |
+
"grad_norm": 0.2779666708867201,
|
| 1323 |
+
"learning_rate": 2.054932085426856e-06,
|
| 1324 |
+
"loss": 0.0984,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.7272727272727275,
|
| 1329 |
+
"grad_norm": 0.27942525906514387,
|
| 1330 |
+
"learning_rate": 1.963688845450218e-06,
|
| 1331 |
+
"loss": 0.1196,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.757575757575758,
|
| 1336 |
+
"grad_norm": 0.2973668361540923,
|
| 1337 |
+
"learning_rate": 1.8742966099201699e-06,
|
| 1338 |
+
"loss": 0.1209,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.787878787878788,
|
| 1343 |
+
"grad_norm": 0.3163737367641305,
|
| 1344 |
+
"learning_rate": 1.7867759685603115e-06,
|
| 1345 |
+
"loss": 0.1126,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.818181818181818,
|
| 1350 |
+
"grad_norm": 0.2976396232837156,
|
| 1351 |
+
"learning_rate": 1.7011470800097496e-06,
|
| 1352 |
+
"loss": 0.0841,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.848484848484849,
|
| 1357 |
+
"grad_norm": 0.2830331635797014,
|
| 1358 |
+
"learning_rate": 1.6174296671799571e-06,
|
| 1359 |
+
"loss": 0.1037,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.878787878787879,
|
| 1364 |
+
"grad_norm": 0.28824315106684817,
|
| 1365 |
+
"learning_rate": 1.5356430127119915e-06,
|
| 1366 |
+
"loss": 0.0823,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.909090909090909,
|
| 1371 |
+
"grad_norm": 0.28497117356177926,
|
| 1372 |
+
"learning_rate": 1.4558059545351144e-06,
|
| 1373 |
+
"loss": 0.0963,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.9393939393939394,
|
| 1378 |
+
"grad_norm": 0.28469042187281185,
|
| 1379 |
+
"learning_rate": 1.3779368815278648e-06,
|
| 1380 |
+
"loss": 0.0854,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.96969696969697,
|
| 1385 |
+
"grad_norm": 0.27807019385827797,
|
| 1386 |
+
"learning_rate": 1.302053729282533e-06,
|
| 1387 |
+
"loss": 0.1002,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.0,
|
| 1392 |
+
"grad_norm": 0.26785098801575175,
|
| 1393 |
+
"learning_rate": 1.2281739759740575e-06,
|
| 1394 |
+
"loss": 0.1155,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.03030303030303,
|
| 1399 |
+
"grad_norm": 0.29162598201638734,
|
| 1400 |
+
"learning_rate": 1.156314638334277e-06,
|
| 1401 |
+
"loss": 0.0922,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.0606060606060606,
|
| 1406 |
+
"grad_norm": 0.2580771610742117,
|
| 1407 |
+
"learning_rate": 1.086492267732462e-06,
|
| 1408 |
+
"loss": 0.1147,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.090909090909091,
|
| 1413 |
+
"grad_norm": 0.2508767222325874,
|
| 1414 |
+
"learning_rate": 1.01872294636304e-06,
|
| 1415 |
+
"loss": 0.0955,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.121212121212121,
|
| 1420 |
+
"grad_norm": 0.2471964833448436,
|
| 1421 |
+
"learning_rate": 9.530222835413739e-07,
|
| 1422 |
+
"loss": 0.0931,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.151515151515151,
|
| 1427 |
+
"grad_norm": 0.23164025105510516,
|
| 1428 |
+
"learning_rate": 8.894054121084839e-07,
|
| 1429 |
+
"loss": 0.1104,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.181818181818182,
|
| 1434 |
+
"grad_norm": 0.21017896698386224,
|
| 1435 |
+
"learning_rate": 8.278869849454718e-07,
|
| 1436 |
+
"loss": 0.1045,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.212121212121212,
|
| 1441 |
+
"grad_norm": 0.20682193953857264,
|
| 1442 |
+
"learning_rate": 7.684811715985429e-07,
|
| 1443 |
+
"loss": 0.0925,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.242424242424242,
|
| 1448 |
+
"grad_norm": 0.2287415488151266,
|
| 1449 |
+
"learning_rate": 7.1120165501533e-07,
|
| 1450 |
+
"loss": 0.074,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.2727272727272725,
|
| 1455 |
+
"grad_norm": 0.21161857344349783,
|
| 1456 |
+
"learning_rate": 6.560616283932897e-07,
|
| 1457 |
+
"loss": 0.0755,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.303030303030303,
|
| 1462 |
+
"grad_norm": 0.2324100363220123,
|
| 1463 |
+
"learning_rate": 6.030737921409169e-07,
|
| 1464 |
+
"loss": 0.0771,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.333333333333333,
|
| 1469 |
+
"grad_norm": 0.21580743753093612,
|
| 1470 |
+
"learning_rate": 5.522503509524591e-07,
|
| 1471 |
+
"loss": 0.0765,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.363636363636363,
|
| 1476 |
+
"grad_norm": 0.22594061914396948,
|
| 1477 |
+
"learning_rate": 5.036030109968082e-07,
|
| 1478 |
+
"loss": 0.0698,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.393939393939394,
|
| 1483 |
+
"grad_norm": 0.223713564778887,
|
| 1484 |
+
"learning_rate": 4.5714297722121105e-07,
|
| 1485 |
+
"loss": 0.0911,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.424242424242424,
|
| 1490 |
+
"grad_norm": 0.22650430623786036,
|
| 1491 |
+
"learning_rate": 4.128809507704445e-07,
|
| 1492 |
+
"loss": 0.0915,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.454545454545454,
|
| 1497 |
+
"grad_norm": 0.23887129807178004,
|
| 1498 |
+
"learning_rate": 3.708271265220087e-07,
|
| 1499 |
+
"loss": 0.0951,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.484848484848484,
|
| 1504 |
+
"grad_norm": 0.22709835978864928,
|
| 1505 |
+
"learning_rate": 3.309911907379393e-07,
|
| 1506 |
+
"loss": 0.0714,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.515151515151516,
|
| 1511 |
+
"grad_norm": 0.24320007330693988,
|
| 1512 |
+
"learning_rate": 2.9338231883378365e-07,
|
| 1513 |
+
"loss": 0.0919,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.545454545454545,
|
| 1518 |
+
"grad_norm": 0.22420899631754834,
|
| 1519 |
+
"learning_rate": 2.5800917326521013e-07,
|
| 1520 |
+
"loss": 0.0955,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.575757575757576,
|
| 1525 |
+
"grad_norm": 0.2317172157931802,
|
| 1526 |
+
"learning_rate": 2.248799015327907e-07,
|
| 1527 |
+
"loss": 0.0898,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.606060606060606,
|
| 1532 |
+
"grad_norm": 0.2298063528871791,
|
| 1533 |
+
"learning_rate": 1.9400213430538773e-07,
|
| 1534 |
+
"loss": 0.0875,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.636363636363637,
|
| 1539 |
+
"grad_norm": 0.21361359985619743,
|
| 1540 |
+
"learning_rate": 1.6538298366257975e-07,
|
| 1541 |
+
"loss": 0.0776,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.666666666666667,
|
| 1546 |
+
"grad_norm": 0.22665356447562912,
|
| 1547 |
+
"learning_rate": 1.3902904145653094e-07,
|
| 1548 |
+
"loss": 0.0698,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.696969696969697,
|
| 1553 |
+
"grad_norm": 0.21696007072832502,
|
| 1554 |
+
"learning_rate": 1.1494637779369766e-07,
|
| 1555 |
+
"loss": 0.0977,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.7272727272727275,
|
| 1560 |
+
"grad_norm": 0.21793216464699694,
|
| 1561 |
+
"learning_rate": 9.314053963669245e-08,
|
| 1562 |
+
"loss": 0.0729,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.757575757575758,
|
| 1567 |
+
"grad_norm": 0.23660170246935486,
|
| 1568 |
+
"learning_rate": 7.361654952665608e-08,
|
| 1569 |
+
"loss": 0.1075,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.787878787878788,
|
| 1574 |
+
"grad_norm": 0.23961930826265718,
|
| 1575 |
+
"learning_rate": 5.637890442641403e-08,
|
| 1576 |
+
"loss": 0.0876,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 6.818181818181818,
|
| 1581 |
+
"grad_norm": 0.21926479664337387,
|
| 1582 |
+
"learning_rate": 4.143157468468717e-08,
|
| 1583 |
+
"loss": 0.0636,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 6.848484848484849,
|
| 1588 |
+
"grad_norm": 0.21679687829037336,
|
| 1589 |
+
"learning_rate": 2.8778003121607834e-08,
|
| 1590 |
+
"loss": 0.0849,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 6.878787878787879,
|
| 1595 |
+
"grad_norm": 0.23693422307046993,
|
| 1596 |
+
"learning_rate": 1.8421104235727406e-08,
|
| 1597 |
+
"loss": 0.0816,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 6.909090909090909,
|
| 1602 |
+
"grad_norm": 0.2318559705397923,
|
| 1603 |
+
"learning_rate": 1.0363263532724433e-08,
|
| 1604 |
+
"loss": 0.1109,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 6.9393939393939394,
|
| 1609 |
+
"grad_norm": 0.22828784321544407,
|
| 1610 |
+
"learning_rate": 4.606336975948589e-09,
|
| 1611 |
+
"loss": 0.0824,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 6.96969696969697,
|
| 1616 |
+
"grad_norm": 0.20041581186676724,
|
| 1617 |
+
"learning_rate": 1.1516505589381777e-09,
|
| 1618 |
+
"loss": 0.0673,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.0,
|
| 1623 |
+
"grad_norm": 0.24898763457678227,
|
| 1624 |
+
"learning_rate": 0.0,
|
| 1625 |
+
"loss": 0.0905,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.0,
|
| 1630 |
+
"step": 231,
|
| 1631 |
+
"total_flos": 2.2511058919253606e+17,
|
| 1632 |
+
"train_loss": 0.25922639719593576,
|
| 1633 |
+
"train_runtime": 5246.8102,
|
| 1634 |
+
"train_samples_per_second": 4.216,
|
| 1635 |
+
"train_steps_per_second": 0.044
|
| 1636 |
+
}
|
| 1637 |
+
],
|
| 1638 |
+
"logging_steps": 1,
|
| 1639 |
+
"max_steps": 231,
|
| 1640 |
+
"num_input_tokens_seen": 0,
|
| 1641 |
+
"num_train_epochs": 7,
|
| 1642 |
+
"save_steps": 500,
|
| 1643 |
+
"stateful_callbacks": {
|
| 1644 |
+
"TrainerControl": {
|
| 1645 |
+
"args": {
|
| 1646 |
+
"should_epoch_stop": false,
|
| 1647 |
+
"should_evaluate": false,
|
| 1648 |
+
"should_log": false,
|
| 1649 |
+
"should_save": true,
|
| 1650 |
+
"should_training_stop": true
|
| 1651 |
+
},
|
| 1652 |
+
"attributes": {}
|
| 1653 |
+
}
|
| 1654 |
+
},
|
| 1655 |
+
"total_flos": 2.2511058919253606e+17,
|
| 1656 |
+
"train_batch_size": 1,
|
| 1657 |
+
"trial_name": null,
|
| 1658 |
+
"trial_params": null
|
| 1659 |
+
}
|
training_loss.png
ADDED
|