End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1765 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-Coder-14B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: SWE-BENCH-reproduce-verify-validation-plus-gym_qwen_code_14B_lite_gym_validation
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# SWE-BENCH-reproduce-verify-validation-plus-gym_qwen_code_14B_lite_gym_validation
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-14B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-Coder-14B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: SWE-BENCH-reproduce-verify-validation-plus-gym_qwen_code_14B_lite_gym_validation
|
|
|
|
| 16 |
|
| 17 |
# SWE-BENCH-reproduce-verify-validation-plus-gym_qwen_code_14B_lite_gym_validation
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-14B-Instruct) on the SWE-BENCH-reproduce-verify-validation-plus-gym dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9889447236180904,
|
| 3 |
+
"total_flos": 6.810123424602522e+16,
|
| 4 |
+
"train_loss": 0.40526863738773317,
|
| 5 |
+
"train_runtime": 35217.1045,
|
| 6 |
+
"train_samples_per_second": 0.17,
|
| 7 |
+
"train_steps_per_second": 0.007
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9889447236180904,
|
| 3 |
+
"total_flos": 6.810123424602522e+16,
|
| 4 |
+
"train_loss": 0.40526863738773317,
|
| 5 |
+
"train_runtime": 35217.1045,
|
| 6 |
+
"train_samples_per_second": 0.17,
|
| 7 |
+
"train_steps_per_second": 0.007
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1765 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.9889447236180904,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 246,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.012060301507537688,
|
| 14 |
+
"grad_norm": 1.7608720064163208,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.7793,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.024120603015075376,
|
| 21 |
+
"grad_norm": 1.5576972961425781,
|
| 22 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 23 |
+
"loss": 0.7167,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.036180904522613064,
|
| 28 |
+
"grad_norm": 1.4898537397384644,
|
| 29 |
+
"learning_rate": 8.000000000000001e-07,
|
| 30 |
+
"loss": 0.7434,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04824120603015075,
|
| 35 |
+
"grad_norm": 1.3706327676773071,
|
| 36 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 37 |
+
"loss": 0.6949,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.06030150753768844,
|
| 42 |
+
"grad_norm": 1.504503607749939,
|
| 43 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 44 |
+
"loss": 0.7081,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.07236180904522613,
|
| 49 |
+
"grad_norm": 1.5883435010910034,
|
| 50 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 51 |
+
"loss": 0.7598,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.08442211055276382,
|
| 56 |
+
"grad_norm": 1.4544540643692017,
|
| 57 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 58 |
+
"loss": 0.6938,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.0964824120603015,
|
| 63 |
+
"grad_norm": 1.104862928390503,
|
| 64 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 65 |
+
"loss": 0.6629,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.10854271356783919,
|
| 70 |
+
"grad_norm": 1.0699723958969116,
|
| 71 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 72 |
+
"loss": 0.6612,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.12060301507537688,
|
| 77 |
+
"grad_norm": 1.105131983757019,
|
| 78 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 79 |
+
"loss": 0.6916,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.13266331658291458,
|
| 84 |
+
"grad_norm": 0.9772186279296875,
|
| 85 |
+
"learning_rate": 4.000000000000001e-06,
|
| 86 |
+
"loss": 0.6725,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.14472361809045226,
|
| 91 |
+
"grad_norm": 0.9385798573493958,
|
| 92 |
+
"learning_rate": 4.4e-06,
|
| 93 |
+
"loss": 0.6136,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.15678391959798996,
|
| 98 |
+
"grad_norm": 0.8693619966506958,
|
| 99 |
+
"learning_rate": 4.800000000000001e-06,
|
| 100 |
+
"loss": 0.5935,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.16884422110552763,
|
| 105 |
+
"grad_norm": 0.9681348204612732,
|
| 106 |
+
"learning_rate": 5.2e-06,
|
| 107 |
+
"loss": 0.704,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.18090452261306533,
|
| 112 |
+
"grad_norm": 0.9246505498886108,
|
| 113 |
+
"learning_rate": 5.600000000000001e-06,
|
| 114 |
+
"loss": 0.6146,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.192964824120603,
|
| 119 |
+
"grad_norm": 1.019329309463501,
|
| 120 |
+
"learning_rate": 6e-06,
|
| 121 |
+
"loss": 0.6434,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.2050251256281407,
|
| 126 |
+
"grad_norm": 0.8335247039794922,
|
| 127 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 128 |
+
"loss": 0.5955,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.21708542713567838,
|
| 133 |
+
"grad_norm": 0.8235917687416077,
|
| 134 |
+
"learning_rate": 6.800000000000001e-06,
|
| 135 |
+
"loss": 0.6119,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.22914572864321608,
|
| 140 |
+
"grad_norm": 0.7331350445747375,
|
| 141 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 142 |
+
"loss": 0.5668,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.24120603015075376,
|
| 147 |
+
"grad_norm": 0.7560482025146484,
|
| 148 |
+
"learning_rate": 7.600000000000001e-06,
|
| 149 |
+
"loss": 0.6136,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.25326633165829143,
|
| 154 |
+
"grad_norm": 0.7541798949241638,
|
| 155 |
+
"learning_rate": 8.000000000000001e-06,
|
| 156 |
+
"loss": 0.5655,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.26532663316582916,
|
| 161 |
+
"grad_norm": 0.7229031920433044,
|
| 162 |
+
"learning_rate": 8.400000000000001e-06,
|
| 163 |
+
"loss": 0.5869,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.27738693467336684,
|
| 168 |
+
"grad_norm": 0.7003089189529419,
|
| 169 |
+
"learning_rate": 8.8e-06,
|
| 170 |
+
"loss": 0.52,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.2894472361809045,
|
| 175 |
+
"grad_norm": 0.7208999395370483,
|
| 176 |
+
"learning_rate": 9.200000000000002e-06,
|
| 177 |
+
"loss": 0.5961,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3015075376884422,
|
| 182 |
+
"grad_norm": 0.7012421488761902,
|
| 183 |
+
"learning_rate": 9.600000000000001e-06,
|
| 184 |
+
"loss": 0.5569,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.3135678391959799,
|
| 189 |
+
"grad_norm": 0.7341773509979248,
|
| 190 |
+
"learning_rate": 1e-05,
|
| 191 |
+
"loss": 0.6162,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.3256281407035176,
|
| 196 |
+
"grad_norm": 0.7771743535995483,
|
| 197 |
+
"learning_rate": 9.999494817970498e-06,
|
| 198 |
+
"loss": 0.6361,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.33768844221105526,
|
| 203 |
+
"grad_norm": 0.6564878225326538,
|
| 204 |
+
"learning_rate": 9.997979373965542e-06,
|
| 205 |
+
"loss": 0.5519,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.349748743718593,
|
| 210 |
+
"grad_norm": 0.6947225332260132,
|
| 211 |
+
"learning_rate": 9.995453974215164e-06,
|
| 212 |
+
"loss": 0.5622,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.36180904522613067,
|
| 217 |
+
"grad_norm": 0.615100085735321,
|
| 218 |
+
"learning_rate": 9.991919129033994e-06,
|
| 219 |
+
"loss": 0.5429,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.37386934673366834,
|
| 224 |
+
"grad_norm": 0.6487897634506226,
|
| 225 |
+
"learning_rate": 9.987375552718133e-06,
|
| 226 |
+
"loss": 0.5649,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.385929648241206,
|
| 231 |
+
"grad_norm": 0.7198299765586853,
|
| 232 |
+
"learning_rate": 9.981824163400827e-06,
|
| 233 |
+
"loss": 0.5269,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.39798994974874374,
|
| 238 |
+
"grad_norm": 0.6803036332130432,
|
| 239 |
+
"learning_rate": 9.975266082866923e-06,
|
| 240 |
+
"loss": 0.5233,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.4100502512562814,
|
| 245 |
+
"grad_norm": 0.6590858101844788,
|
| 246 |
+
"learning_rate": 9.967702636326195e-06,
|
| 247 |
+
"loss": 0.4995,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.4221105527638191,
|
| 252 |
+
"grad_norm": 0.6516736149787903,
|
| 253 |
+
"learning_rate": 9.959135352145552e-06,
|
| 254 |
+
"loss": 0.5158,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.43417085427135677,
|
| 259 |
+
"grad_norm": 0.6069977283477783,
|
| 260 |
+
"learning_rate": 9.9495659615402e-06,
|
| 261 |
+
"loss": 0.5075,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.4462311557788945,
|
| 266 |
+
"grad_norm": 0.8428781628608704,
|
| 267 |
+
"learning_rate": 9.938996398223802e-06,
|
| 268 |
+
"loss": 0.5744,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.45829145728643217,
|
| 273 |
+
"grad_norm": 0.7003642320632935,
|
| 274 |
+
"learning_rate": 9.927428798017738e-06,
|
| 275 |
+
"loss": 0.5649,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.47035175879396984,
|
| 280 |
+
"grad_norm": 0.5626908540725708,
|
| 281 |
+
"learning_rate": 9.91486549841951e-06,
|
| 282 |
+
"loss": 0.5042,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.4824120603015075,
|
| 287 |
+
"grad_norm": 0.6438745856285095,
|
| 288 |
+
"learning_rate": 9.901309038130392e-06,
|
| 289 |
+
"loss": 0.4814,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.49447236180904525,
|
| 294 |
+
"grad_norm": 0.5857080817222595,
|
| 295 |
+
"learning_rate": 9.886762156542428e-06,
|
| 296 |
+
"loss": 0.5319,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.5065326633165829,
|
| 301 |
+
"grad_norm": 0.579682469367981,
|
| 302 |
+
"learning_rate": 9.871227793184893e-06,
|
| 303 |
+
"loss": 0.4963,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5185929648241207,
|
| 308 |
+
"grad_norm": 0.6599876284599304,
|
| 309 |
+
"learning_rate": 9.854709087130261e-06,
|
| 310 |
+
"loss": 0.5569,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.5306532663316583,
|
| 315 |
+
"grad_norm": 0.5928906202316284,
|
| 316 |
+
"learning_rate": 9.837209376359918e-06,
|
| 317 |
+
"loss": 0.5426,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.542713567839196,
|
| 322 |
+
"grad_norm": 0.5092033743858337,
|
| 323 |
+
"learning_rate": 9.81873219708962e-06,
|
| 324 |
+
"loss": 0.4224,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.5547738693467337,
|
| 329 |
+
"grad_norm": 0.6165465712547302,
|
| 330 |
+
"learning_rate": 9.79928128305494e-06,
|
| 331 |
+
"loss": 0.5157,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.5668341708542713,
|
| 336 |
+
"grad_norm": 0.7021877765655518,
|
| 337 |
+
"learning_rate": 9.778860564756769e-06,
|
| 338 |
+
"loss": 0.6073,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.578894472361809,
|
| 343 |
+
"grad_norm": 0.6414082646369934,
|
| 344 |
+
"learning_rate": 9.757474168667072e-06,
|
| 345 |
+
"loss": 0.5141,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.5909547738693467,
|
| 350 |
+
"grad_norm": 0.5921343564987183,
|
| 351 |
+
"learning_rate": 9.73512641639504e-06,
|
| 352 |
+
"loss": 0.4712,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.6030150753768844,
|
| 357 |
+
"grad_norm": 0.6099699139595032,
|
| 358 |
+
"learning_rate": 9.711821823813812e-06,
|
| 359 |
+
"loss": 0.4844,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.6150753768844222,
|
| 364 |
+
"grad_norm": 0.5854896903038025,
|
| 365 |
+
"learning_rate": 9.68756510014794e-06,
|
| 366 |
+
"loss": 0.5009,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.6271356783919598,
|
| 371 |
+
"grad_norm": 0.650279700756073,
|
| 372 |
+
"learning_rate": 9.66236114702178e-06,
|
| 373 |
+
"loss": 0.5003,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.6391959798994975,
|
| 378 |
+
"grad_norm": 0.6033490896224976,
|
| 379 |
+
"learning_rate": 9.636215057469009e-06,
|
| 380 |
+
"loss": 0.5092,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.6512562814070352,
|
| 385 |
+
"grad_norm": 0.6597270965576172,
|
| 386 |
+
"learning_rate": 9.609132114903458e-06,
|
| 387 |
+
"loss": 0.5529,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6633165829145728,
|
| 392 |
+
"grad_norm": 0.6374630331993103,
|
| 393 |
+
"learning_rate": 9.581117792051487e-06,
|
| 394 |
+
"loss": 0.5637,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.6753768844221105,
|
| 399 |
+
"grad_norm": 0.849519670009613,
|
| 400 |
+
"learning_rate": 9.552177749846083e-06,
|
| 401 |
+
"loss": 0.5119,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.6874371859296482,
|
| 406 |
+
"grad_norm": 0.6256008744239807,
|
| 407 |
+
"learning_rate": 9.522317836282949e-06,
|
| 408 |
+
"loss": 0.5025,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.699497487437186,
|
| 413 |
+
"grad_norm": 0.6780921816825867,
|
| 414 |
+
"learning_rate": 9.491544085238778e-06,
|
| 415 |
+
"loss": 0.5005,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7115577889447237,
|
| 420 |
+
"grad_norm": 0.6347315907478333,
|
| 421 |
+
"learning_rate": 9.459862715251973e-06,
|
| 422 |
+
"loss": 0.5244,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.7236180904522613,
|
| 427 |
+
"grad_norm": 0.6093297600746155,
|
| 428 |
+
"learning_rate": 9.427280128266049e-06,
|
| 429 |
+
"loss": 0.5427,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.735678391959799,
|
| 434 |
+
"grad_norm": 0.7500917315483093,
|
| 435 |
+
"learning_rate": 9.393802908335978e-06,
|
| 436 |
+
"loss": 0.5277,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.7477386934673367,
|
| 441 |
+
"grad_norm": 0.5266271829605103,
|
| 442 |
+
"learning_rate": 9.359437820297716e-06,
|
| 443 |
+
"loss": 0.4353,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7597989949748744,
|
| 448 |
+
"grad_norm": 0.5885184407234192,
|
| 449 |
+
"learning_rate": 9.324191808401235e-06,
|
| 450 |
+
"loss": 0.5307,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.771859296482412,
|
| 455 |
+
"grad_norm": 0.6106560826301575,
|
| 456 |
+
"learning_rate": 9.288071994907262e-06,
|
| 457 |
+
"loss": 0.5196,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.7839195979899497,
|
| 462 |
+
"grad_norm": 0.6581817269325256,
|
| 463 |
+
"learning_rate": 9.251085678648072e-06,
|
| 464 |
+
"loss": 0.5412,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.7959798994974875,
|
| 469 |
+
"grad_norm": 0.6170300841331482,
|
| 470 |
+
"learning_rate": 9.213240333552589e-06,
|
| 471 |
+
"loss": 0.506,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.8080402010050252,
|
| 476 |
+
"grad_norm": 0.5753492712974548,
|
| 477 |
+
"learning_rate": 9.174543607136111e-06,
|
| 478 |
+
"loss": 0.4566,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.8201005025125628,
|
| 483 |
+
"grad_norm": 0.6315353512763977,
|
| 484 |
+
"learning_rate": 9.135003318954954e-06,
|
| 485 |
+
"loss": 0.4965,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.8321608040201005,
|
| 490 |
+
"grad_norm": 0.592758059501648,
|
| 491 |
+
"learning_rate": 9.094627459026326e-06,
|
| 492 |
+
"loss": 0.4126,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.8442211055276382,
|
| 497 |
+
"grad_norm": 0.5657063722610474,
|
| 498 |
+
"learning_rate": 9.053424186213776e-06,
|
| 499 |
+
"loss": 0.4746,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8562814070351759,
|
| 504 |
+
"grad_norm": 0.6852341890335083,
|
| 505 |
+
"learning_rate": 9.011401826578492e-06,
|
| 506 |
+
"loss": 0.4936,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.8683417085427135,
|
| 511 |
+
"grad_norm": 0.5805753469467163,
|
| 512 |
+
"learning_rate": 8.968568871696847e-06,
|
| 513 |
+
"loss": 0.4718,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.8804020100502512,
|
| 518 |
+
"grad_norm": 0.6270849108695984,
|
| 519 |
+
"learning_rate": 8.924933976944474e-06,
|
| 520 |
+
"loss": 0.5179,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.892462311557789,
|
| 525 |
+
"grad_norm": 0.6481483578681946,
|
| 526 |
+
"learning_rate": 8.880505959747245e-06,
|
| 527 |
+
"loss": 0.5252,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.9045226130653267,
|
| 532 |
+
"grad_norm": 0.6944591999053955,
|
| 533 |
+
"learning_rate": 8.835293797799517e-06,
|
| 534 |
+
"loss": 0.5661,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.9165829145728643,
|
| 539 |
+
"grad_norm": 0.5752354860305786,
|
| 540 |
+
"learning_rate": 8.789306627249985e-06,
|
| 541 |
+
"loss": 0.4611,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.928643216080402,
|
| 546 |
+
"grad_norm": 0.6542851328849792,
|
| 547 |
+
"learning_rate": 8.742553740855507e-06,
|
| 548 |
+
"loss": 0.5003,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.9407035175879397,
|
| 553 |
+
"grad_norm": 0.6845901608467102,
|
| 554 |
+
"learning_rate": 8.695044586103297e-06,
|
| 555 |
+
"loss": 0.504,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.9527638190954774,
|
| 560 |
+
"grad_norm": 0.5776346921920776,
|
| 561 |
+
"learning_rate": 8.646788763301842e-06,
|
| 562 |
+
"loss": 0.54,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.964824120603015,
|
| 567 |
+
"grad_norm": 0.5651821494102478,
|
| 568 |
+
"learning_rate": 8.59779602364094e-06,
|
| 569 |
+
"loss": 0.478,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.9768844221105528,
|
| 574 |
+
"grad_norm": 0.6050324440002441,
|
| 575 |
+
"learning_rate": 8.548076267221258e-06,
|
| 576 |
+
"loss": 0.4788,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.9889447236180905,
|
| 581 |
+
"grad_norm": 0.5651063919067383,
|
| 582 |
+
"learning_rate": 8.497639541053769e-06,
|
| 583 |
+
"loss": 0.4821,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.0120603015075378,
|
| 588 |
+
"grad_norm": 0.7770034074783325,
|
| 589 |
+
"learning_rate": 8.446496037029555e-06,
|
| 590 |
+
"loss": 0.9163,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.0241206030150753,
|
| 595 |
+
"grad_norm": 0.6015292406082153,
|
| 596 |
+
"learning_rate": 8.394656089860274e-06,
|
| 597 |
+
"loss": 0.4151,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.0361809045226131,
|
| 602 |
+
"grad_norm": 0.5272526741027832,
|
| 603 |
+
"learning_rate": 8.342130174989819e-06,
|
| 604 |
+
"loss": 0.4027,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.0482412060301507,
|
| 609 |
+
"grad_norm": 0.5854943990707397,
|
| 610 |
+
"learning_rate": 8.288928906477497e-06,
|
| 611 |
+
"loss": 0.3566,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.0603015075376885,
|
| 616 |
+
"grad_norm": 0.5184862613677979,
|
| 617 |
+
"learning_rate": 8.235063034853228e-06,
|
| 618 |
+
"loss": 0.412,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.072361809045226,
|
| 623 |
+
"grad_norm": 0.5804527997970581,
|
| 624 |
+
"learning_rate": 8.180543444945154e-06,
|
| 625 |
+
"loss": 0.3899,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.0844221105527638,
|
| 630 |
+
"grad_norm": 0.6448402404785156,
|
| 631 |
+
"learning_rate": 8.125381153680103e-06,
|
| 632 |
+
"loss": 0.3873,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.0964824120603014,
|
| 637 |
+
"grad_norm": 0.5404708981513977,
|
| 638 |
+
"learning_rate": 8.069587307857377e-06,
|
| 639 |
+
"loss": 0.3368,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.1085427135678392,
|
| 644 |
+
"grad_norm": 0.5709781646728516,
|
| 645 |
+
"learning_rate": 8.013173181896283e-06,
|
| 646 |
+
"loss": 0.4321,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.120603015075377,
|
| 651 |
+
"grad_norm": 0.7116711139678955,
|
| 652 |
+
"learning_rate": 7.95615017555788e-06,
|
| 653 |
+
"loss": 0.3744,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.1326633165829145,
|
| 658 |
+
"grad_norm": 0.5133858323097229,
|
| 659 |
+
"learning_rate": 7.898529811641393e-06,
|
| 660 |
+
"loss": 0.3951,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.1447236180904523,
|
| 665 |
+
"grad_norm": 0.6479538083076477,
|
| 666 |
+
"learning_rate": 7.84032373365578e-06,
|
| 667 |
+
"loss": 0.3973,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.1567839195979899,
|
| 672 |
+
"grad_norm": 0.6897246837615967,
|
| 673 |
+
"learning_rate": 7.781543703466881e-06,
|
| 674 |
+
"loss": 0.381,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.1688442211055277,
|
| 679 |
+
"grad_norm": 0.6539267897605896,
|
| 680 |
+
"learning_rate": 7.722201598920673e-06,
|
| 681 |
+
"loss": 0.3773,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.1809045226130652,
|
| 686 |
+
"grad_norm": 0.6268959641456604,
|
| 687 |
+
"learning_rate": 7.662309411443084e-06,
|
| 688 |
+
"loss": 0.4552,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.192964824120603,
|
| 693 |
+
"grad_norm": 0.7149134278297424,
|
| 694 |
+
"learning_rate": 7.601879243616838e-06,
|
| 695 |
+
"loss": 0.374,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.2050251256281408,
|
| 700 |
+
"grad_norm": 0.646574079990387,
|
| 701 |
+
"learning_rate": 7.540923306735868e-06,
|
| 702 |
+
"loss": 0.396,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.2170854271356784,
|
| 707 |
+
"grad_norm": 0.5743764042854309,
|
| 708 |
+
"learning_rate": 7.479453918337733e-06,
|
| 709 |
+
"loss": 0.405,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.2291457286432161,
|
| 714 |
+
"grad_norm": 0.6699774861335754,
|
| 715 |
+
"learning_rate": 7.417483499714589e-06,
|
| 716 |
+
"loss": 0.3875,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.2412060301507537,
|
| 721 |
+
"grad_norm": 0.5869190692901611,
|
| 722 |
+
"learning_rate": 7.355024573403174e-06,
|
| 723 |
+
"loss": 0.4169,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.2532663316582915,
|
| 728 |
+
"grad_norm": 0.6940112709999084,
|
| 729 |
+
"learning_rate": 7.292089760654352e-06,
|
| 730 |
+
"loss": 0.3653,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.265326633165829,
|
| 735 |
+
"grad_norm": 0.6271514296531677,
|
| 736 |
+
"learning_rate": 7.2286917788826926e-06,
|
| 737 |
+
"loss": 0.3753,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.2773869346733668,
|
| 742 |
+
"grad_norm": 0.5472912192344666,
|
| 743 |
+
"learning_rate": 7.1648434390966356e-06,
|
| 744 |
+
"loss": 0.4224,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.2894472361809046,
|
| 749 |
+
"grad_norm": 0.5858989357948303,
|
| 750 |
+
"learning_rate": 7.100557643309732e-06,
|
| 751 |
+
"loss": 0.3392,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.3015075376884422,
|
| 756 |
+
"grad_norm": 0.5777701139450073,
|
| 757 |
+
"learning_rate": 7.035847381933494e-06,
|
| 758 |
+
"loss": 0.3895,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.31356783919598,
|
| 763 |
+
"grad_norm": 0.6289804577827454,
|
| 764 |
+
"learning_rate": 6.970725731152389e-06,
|
| 765 |
+
"loss": 0.3768,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.3256281407035175,
|
| 770 |
+
"grad_norm": 0.6202074885368347,
|
| 771 |
+
"learning_rate": 6.905205850281502e-06,
|
| 772 |
+
"loss": 0.377,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.3376884422110553,
|
| 777 |
+
"grad_norm": 0.6787943840026855,
|
| 778 |
+
"learning_rate": 6.8393009791073895e-06,
|
| 779 |
+
"loss": 0.3778,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.3497487437185929,
|
| 784 |
+
"grad_norm": 0.6368122100830078,
|
| 785 |
+
"learning_rate": 6.773024435212678e-06,
|
| 786 |
+
"loss": 0.3756,
|
| 787 |
+
"step": 111
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.3618090452261307,
|
| 791 |
+
"grad_norm": 0.6694258451461792,
|
| 792 |
+
"learning_rate": 6.706389611284953e-06,
|
| 793 |
+
"loss": 0.3293,
|
| 794 |
+
"step": 112
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.3738693467336685,
|
| 798 |
+
"grad_norm": 0.6035791635513306,
|
| 799 |
+
"learning_rate": 6.639409972410446e-06,
|
| 800 |
+
"loss": 0.3802,
|
| 801 |
+
"step": 113
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.385929648241206,
|
| 805 |
+
"grad_norm": 0.6572511196136475,
|
| 806 |
+
"learning_rate": 6.57209905335312e-06,
|
| 807 |
+
"loss": 0.4086,
|
| 808 |
+
"step": 114
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.3979899497487438,
|
| 812 |
+
"grad_norm": 0.6025294661521912,
|
| 813 |
+
"learning_rate": 6.504470455819651e-06,
|
| 814 |
+
"loss": 0.3663,
|
| 815 |
+
"step": 115
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.4100502512562814,
|
| 819 |
+
"grad_norm": 0.5874430537223816,
|
| 820 |
+
"learning_rate": 6.436537845710904e-06,
|
| 821 |
+
"loss": 0.3733,
|
| 822 |
+
"step": 116
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.4221105527638191,
|
| 826 |
+
"grad_norm": 0.584042489528656,
|
| 827 |
+
"learning_rate": 6.368314950360416e-06,
|
| 828 |
+
"loss": 0.3619,
|
| 829 |
+
"step": 117
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.4341708542713567,
|
| 833 |
+
"grad_norm": 0.6096173524856567,
|
| 834 |
+
"learning_rate": 6.299815555760478e-06,
|
| 835 |
+
"loss": 0.3958,
|
| 836 |
+
"step": 118
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.4462311557788945,
|
| 840 |
+
"grad_norm": 0.6311055421829224,
|
| 841 |
+
"learning_rate": 6.231053503776363e-06,
|
| 842 |
+
"loss": 0.3428,
|
| 843 |
+
"step": 119
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.4582914572864323,
|
| 847 |
+
"grad_norm": 0.6712104678153992,
|
| 848 |
+
"learning_rate": 6.1620426893492645e-06,
|
| 849 |
+
"loss": 0.3709,
|
| 850 |
+
"step": 120
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.4703517587939698,
|
| 854 |
+
"grad_norm": 0.58527010679245,
|
| 855 |
+
"learning_rate": 6.092797057688496e-06,
|
| 856 |
+
"loss": 0.3549,
|
| 857 |
+
"step": 121
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.4824120603015074,
|
| 861 |
+
"grad_norm": 0.6143395304679871,
|
| 862 |
+
"learning_rate": 6.0233306014535505e-06,
|
| 863 |
+
"loss": 0.3535,
|
| 864 |
+
"step": 122
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.4944723618090452,
|
| 868 |
+
"grad_norm": 0.6603715419769287,
|
| 869 |
+
"learning_rate": 5.953657357926569e-06,
|
| 870 |
+
"loss": 0.3775,
|
| 871 |
+
"step": 123
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.506532663316583,
|
| 875 |
+
"grad_norm": 0.659695565700531,
|
| 876 |
+
"learning_rate": 5.883791406175775e-06,
|
| 877 |
+
"loss": 0.3546,
|
| 878 |
+
"step": 124
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.5185929648241205,
|
| 882 |
+
"grad_norm": 0.6348215937614441,
|
| 883 |
+
"learning_rate": 5.813746864210489e-06,
|
| 884 |
+
"loss": 0.3883,
|
| 885 |
+
"step": 125
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.5306532663316583,
|
| 889 |
+
"grad_norm": 0.7142219543457031,
|
| 890 |
+
"learning_rate": 5.743537886128258e-06,
|
| 891 |
+
"loss": 0.4005,
|
| 892 |
+
"step": 126
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.542713567839196,
|
| 896 |
+
"grad_norm": 0.6406548023223877,
|
| 897 |
+
"learning_rate": 5.673178659254698e-06,
|
| 898 |
+
"loss": 0.3475,
|
| 899 |
+
"step": 127
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.5547738693467337,
|
| 903 |
+
"grad_norm": 0.5952600836753845,
|
| 904 |
+
"learning_rate": 5.6026834012766155e-06,
|
| 905 |
+
"loss": 0.399,
|
| 906 |
+
"step": 128
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.5668341708542712,
|
| 910 |
+
"grad_norm": 0.6538510918617249,
|
| 911 |
+
"learning_rate": 5.532066357369012e-06,
|
| 912 |
+
"loss": 0.3404,
|
| 913 |
+
"step": 129
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.578894472361809,
|
| 917 |
+
"grad_norm": 0.6100224256515503,
|
| 918 |
+
"learning_rate": 5.46134179731651e-06,
|
| 919 |
+
"loss": 0.402,
|
| 920 |
+
"step": 130
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.5909547738693468,
|
| 924 |
+
"grad_norm": 0.6091918349266052,
|
| 925 |
+
"learning_rate": 5.390524012629824e-06,
|
| 926 |
+
"loss": 0.3735,
|
| 927 |
+
"step": 131
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.6030150753768844,
|
| 931 |
+
"grad_norm": 0.6372969746589661,
|
| 932 |
+
"learning_rate": 5.319627313657829e-06,
|
| 933 |
+
"loss": 0.3455,
|
| 934 |
+
"step": 132
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.6150753768844222,
|
| 938 |
+
"grad_norm": 0.5964672565460205,
|
| 939 |
+
"learning_rate": 5.248666026695835e-06,
|
| 940 |
+
"loss": 0.3252,
|
| 941 |
+
"step": 133
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.62713567839196,
|
| 945 |
+
"grad_norm": 0.6028534770011902,
|
| 946 |
+
"learning_rate": 5.177654491090627e-06,
|
| 947 |
+
"loss": 0.407,
|
| 948 |
+
"step": 134
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.6391959798994975,
|
| 952 |
+
"grad_norm": 0.6720948219299316,
|
| 953 |
+
"learning_rate": 5.1066070563428736e-06,
|
| 954 |
+
"loss": 0.3471,
|
| 955 |
+
"step": 135
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.651256281407035,
|
| 959 |
+
"grad_norm": 0.5648738145828247,
|
| 960 |
+
"learning_rate": 5.035538079207488e-06,
|
| 961 |
+
"loss": 0.4174,
|
| 962 |
+
"step": 136
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.6633165829145728,
|
| 966 |
+
"grad_norm": 0.6812758445739746,
|
| 967 |
+
"learning_rate": 4.964461920792512e-06,
|
| 968 |
+
"loss": 0.386,
|
| 969 |
+
"step": 137
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.6753768844221106,
|
| 973 |
+
"grad_norm": 0.6770824790000916,
|
| 974 |
+
"learning_rate": 4.893392943657127e-06,
|
| 975 |
+
"loss": 0.4146,
|
| 976 |
+
"step": 138
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.6874371859296482,
|
| 980 |
+
"grad_norm": 0.7083914875984192,
|
| 981 |
+
"learning_rate": 4.822345508909376e-06,
|
| 982 |
+
"loss": 0.3379,
|
| 983 |
+
"step": 139
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.699497487437186,
|
| 987 |
+
"grad_norm": 0.612530529499054,
|
| 988 |
+
"learning_rate": 4.751333973304166e-06,
|
| 989 |
+
"loss": 0.367,
|
| 990 |
+
"step": 140
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.7115577889447238,
|
| 994 |
+
"grad_norm": 0.6422005891799927,
|
| 995 |
+
"learning_rate": 4.680372686342173e-06,
|
| 996 |
+
"loss": 0.3457,
|
| 997 |
+
"step": 141
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.7236180904522613,
|
| 1001 |
+
"grad_norm": 0.6038451790809631,
|
| 1002 |
+
"learning_rate": 4.609475987370177e-06,
|
| 1003 |
+
"loss": 0.3802,
|
| 1004 |
+
"step": 142
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.735678391959799,
|
| 1008 |
+
"grad_norm": 0.6907718777656555,
|
| 1009 |
+
"learning_rate": 4.53865820268349e-06,
|
| 1010 |
+
"loss": 0.3698,
|
| 1011 |
+
"step": 143
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.7477386934673367,
|
| 1015 |
+
"grad_norm": 0.6266685724258423,
|
| 1016 |
+
"learning_rate": 4.467933642630989e-06,
|
| 1017 |
+
"loss": 0.3741,
|
| 1018 |
+
"step": 144
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.7597989949748745,
|
| 1022 |
+
"grad_norm": 0.6295265555381775,
|
| 1023 |
+
"learning_rate": 4.397316598723385e-06,
|
| 1024 |
+
"loss": 0.3506,
|
| 1025 |
+
"step": 145
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.771859296482412,
|
| 1029 |
+
"grad_norm": 0.5913130044937134,
|
| 1030 |
+
"learning_rate": 4.326821340745304e-06,
|
| 1031 |
+
"loss": 0.3847,
|
| 1032 |
+
"step": 146
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.7839195979899496,
|
| 1036 |
+
"grad_norm": 0.663428008556366,
|
| 1037 |
+
"learning_rate": 4.256462113871741e-06,
|
| 1038 |
+
"loss": 0.39,
|
| 1039 |
+
"step": 147
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.7959798994974876,
|
| 1043 |
+
"grad_norm": 0.7241398692131042,
|
| 1044 |
+
"learning_rate": 4.186253135789511e-06,
|
| 1045 |
+
"loss": 0.375,
|
| 1046 |
+
"step": 148
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.8080402010050252,
|
| 1050 |
+
"grad_norm": 0.5983396172523499,
|
| 1051 |
+
"learning_rate": 4.116208593824227e-06,
|
| 1052 |
+
"loss": 0.3851,
|
| 1053 |
+
"step": 149
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.8201005025125627,
|
| 1057 |
+
"grad_norm": 0.6172501444816589,
|
| 1058 |
+
"learning_rate": 4.046342642073433e-06,
|
| 1059 |
+
"loss": 0.3792,
|
| 1060 |
+
"step": 150
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.8321608040201005,
|
| 1064 |
+
"grad_norm": 0.6548818349838257,
|
| 1065 |
+
"learning_rate": 3.976669398546451e-06,
|
| 1066 |
+
"loss": 0.3225,
|
| 1067 |
+
"step": 151
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.8442211055276383,
|
| 1071 |
+
"grad_norm": 0.586807131767273,
|
| 1072 |
+
"learning_rate": 3.907202942311506e-06,
|
| 1073 |
+
"loss": 0.3902,
|
| 1074 |
+
"step": 152
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.8562814070351759,
|
| 1078 |
+
"grad_norm": 0.7093399167060852,
|
| 1079 |
+
"learning_rate": 3.837957310650738e-06,
|
| 1080 |
+
"loss": 0.389,
|
| 1081 |
+
"step": 153
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.8683417085427134,
|
| 1085 |
+
"grad_norm": 0.7611936926841736,
|
| 1086 |
+
"learning_rate": 3.7689464962236367e-06,
|
| 1087 |
+
"loss": 0.3559,
|
| 1088 |
+
"step": 154
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.8804020100502512,
|
| 1092 |
+
"grad_norm": 0.6062532663345337,
|
| 1093 |
+
"learning_rate": 3.700184444239524e-06,
|
| 1094 |
+
"loss": 0.3671,
|
| 1095 |
+
"step": 155
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.892462311557789,
|
| 1099 |
+
"grad_norm": 0.6498876214027405,
|
| 1100 |
+
"learning_rate": 3.6316850496395863e-06,
|
| 1101 |
+
"loss": 0.3956,
|
| 1102 |
+
"step": 156
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.9045226130653266,
|
| 1106 |
+
"grad_norm": 0.6417592763900757,
|
| 1107 |
+
"learning_rate": 3.563462154289098e-06,
|
| 1108 |
+
"loss": 0.3755,
|
| 1109 |
+
"step": 157
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.9165829145728643,
|
| 1113 |
+
"grad_norm": 0.6421058773994446,
|
| 1114 |
+
"learning_rate": 3.49552954418035e-06,
|
| 1115 |
+
"loss": 0.3806,
|
| 1116 |
+
"step": 158
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.9286432160804021,
|
| 1120 |
+
"grad_norm": 0.6108628511428833,
|
| 1121 |
+
"learning_rate": 3.4279009466468825e-06,
|
| 1122 |
+
"loss": 0.3885,
|
| 1123 |
+
"step": 159
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.9407035175879397,
|
| 1127 |
+
"grad_norm": 0.6215425729751587,
|
| 1128 |
+
"learning_rate": 3.3605900275895565e-06,
|
| 1129 |
+
"loss": 0.3973,
|
| 1130 |
+
"step": 160
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.9527638190954772,
|
| 1134 |
+
"grad_norm": 0.6752203106880188,
|
| 1135 |
+
"learning_rate": 3.2936103887150484e-06,
|
| 1136 |
+
"loss": 0.3687,
|
| 1137 |
+
"step": 161
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.964824120603015,
|
| 1141 |
+
"grad_norm": 0.6283330917358398,
|
| 1142 |
+
"learning_rate": 3.226975564787322e-06,
|
| 1143 |
+
"loss": 0.4121,
|
| 1144 |
+
"step": 162
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.9768844221105528,
|
| 1148 |
+
"grad_norm": 0.6184133887290955,
|
| 1149 |
+
"learning_rate": 3.1606990208926125e-06,
|
| 1150 |
+
"loss": 0.3836,
|
| 1151 |
+
"step": 163
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.9889447236180904,
|
| 1155 |
+
"grad_norm": 0.6305925846099854,
|
| 1156 |
+
"learning_rate": 3.0947941497184985e-06,
|
| 1157 |
+
"loss": 0.3617,
|
| 1158 |
+
"step": 164
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.0120603015075376,
|
| 1162 |
+
"grad_norm": 0.686490535736084,
|
| 1163 |
+
"learning_rate": 3.0292742688476125e-06,
|
| 1164 |
+
"loss": 0.6955,
|
| 1165 |
+
"step": 165
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.0241206030150756,
|
| 1169 |
+
"grad_norm": 0.5885306596755981,
|
| 1170 |
+
"learning_rate": 2.964152618066508e-06,
|
| 1171 |
+
"loss": 0.2988,
|
| 1172 |
+
"step": 166
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.036180904522613,
|
| 1176 |
+
"grad_norm": 0.6542847156524658,
|
| 1177 |
+
"learning_rate": 2.899442356690271e-06,
|
| 1178 |
+
"loss": 0.2977,
|
| 1179 |
+
"step": 167
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.0482412060301507,
|
| 1183 |
+
"grad_norm": 0.6749756336212158,
|
| 1184 |
+
"learning_rate": 2.835156560903365e-06,
|
| 1185 |
+
"loss": 0.2701,
|
| 1186 |
+
"step": 168
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.0603015075376883,
|
| 1190 |
+
"grad_norm": 0.6328144073486328,
|
| 1191 |
+
"learning_rate": 2.771308221117309e-06,
|
| 1192 |
+
"loss": 0.2877,
|
| 1193 |
+
"step": 169
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.0723618090452263,
|
| 1197 |
+
"grad_norm": 0.5921873450279236,
|
| 1198 |
+
"learning_rate": 2.7079102393456503e-06,
|
| 1199 |
+
"loss": 0.2564,
|
| 1200 |
+
"step": 170
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.084422110552764,
|
| 1204 |
+
"grad_norm": 0.629751980304718,
|
| 1205 |
+
"learning_rate": 2.6449754265968263e-06,
|
| 1206 |
+
"loss": 0.255,
|
| 1207 |
+
"step": 171
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.0964824120603014,
|
| 1211 |
+
"grad_norm": 0.5695708990097046,
|
| 1212 |
+
"learning_rate": 2.5825165002854124e-06,
|
| 1213 |
+
"loss": 0.2402,
|
| 1214 |
+
"step": 172
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.1085427135678394,
|
| 1218 |
+
"grad_norm": 0.6583861112594604,
|
| 1219 |
+
"learning_rate": 2.5205460816622684e-06,
|
| 1220 |
+
"loss": 0.2902,
|
| 1221 |
+
"step": 173
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.120603015075377,
|
| 1225 |
+
"grad_norm": 0.6468122601509094,
|
| 1226 |
+
"learning_rate": 2.4590766932641353e-06,
|
| 1227 |
+
"loss": 0.283,
|
| 1228 |
+
"step": 174
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.1326633165829145,
|
| 1232 |
+
"grad_norm": 0.7240746021270752,
|
| 1233 |
+
"learning_rate": 2.3981207563831633e-06,
|
| 1234 |
+
"loss": 0.2608,
|
| 1235 |
+
"step": 175
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.144723618090452,
|
| 1239 |
+
"grad_norm": 0.67857825756073,
|
| 1240 |
+
"learning_rate": 2.3376905885569185e-06,
|
| 1241 |
+
"loss": 0.2821,
|
| 1242 |
+
"step": 176
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.15678391959799,
|
| 1246 |
+
"grad_norm": 0.7460854649543762,
|
| 1247 |
+
"learning_rate": 2.2777984010793264e-06,
|
| 1248 |
+
"loss": 0.2557,
|
| 1249 |
+
"step": 177
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.1688442211055277,
|
| 1253 |
+
"grad_norm": 0.7234517335891724,
|
| 1254 |
+
"learning_rate": 2.2184562965331203e-06,
|
| 1255 |
+
"loss": 0.2807,
|
| 1256 |
+
"step": 178
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.180904522613065,
|
| 1260 |
+
"grad_norm": 0.7631310820579529,
|
| 1261 |
+
"learning_rate": 2.159676266344222e-06,
|
| 1262 |
+
"loss": 0.2861,
|
| 1263 |
+
"step": 179
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.192964824120603,
|
| 1267 |
+
"grad_norm": 0.7467869520187378,
|
| 1268 |
+
"learning_rate": 2.1014701883586087e-06,
|
| 1269 |
+
"loss": 0.2683,
|
| 1270 |
+
"step": 180
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.205025125628141,
|
| 1274 |
+
"grad_norm": 1.1319634914398193,
|
| 1275 |
+
"learning_rate": 2.043849824442124e-06,
|
| 1276 |
+
"loss": 0.2535,
|
| 1277 |
+
"step": 181
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.2170854271356784,
|
| 1281 |
+
"grad_norm": 0.7149626612663269,
|
| 1282 |
+
"learning_rate": 1.9868268181037186e-06,
|
| 1283 |
+
"loss": 0.2965,
|
| 1284 |
+
"step": 182
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.229145728643216,
|
| 1288 |
+
"grad_norm": 0.699482262134552,
|
| 1289 |
+
"learning_rate": 1.9304126921426235e-06,
|
| 1290 |
+
"loss": 0.2691,
|
| 1291 |
+
"step": 183
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.241206030150754,
|
| 1295 |
+
"grad_norm": 0.7378098964691162,
|
| 1296 |
+
"learning_rate": 1.8746188463198983e-06,
|
| 1297 |
+
"loss": 0.2526,
|
| 1298 |
+
"step": 184
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.2532663316582915,
|
| 1302 |
+
"grad_norm": 0.6941034197807312,
|
| 1303 |
+
"learning_rate": 1.8194565550548477e-06,
|
| 1304 |
+
"loss": 0.2913,
|
| 1305 |
+
"step": 185
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.265326633165829,
|
| 1309 |
+
"grad_norm": 0.6785539984703064,
|
| 1310 |
+
"learning_rate": 1.764936965146773e-06,
|
| 1311 |
+
"loss": 0.2481,
|
| 1312 |
+
"step": 186
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.277386934673367,
|
| 1316 |
+
"grad_norm": 0.6490347385406494,
|
| 1317 |
+
"learning_rate": 1.7110710935225055e-06,
|
| 1318 |
+
"loss": 0.2529,
|
| 1319 |
+
"step": 187
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.2894472361809046,
|
| 1323 |
+
"grad_norm": 0.7124976515769958,
|
| 1324 |
+
"learning_rate": 1.6578698250101828e-06,
|
| 1325 |
+
"loss": 0.2787,
|
| 1326 |
+
"step": 188
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.301507537688442,
|
| 1330 |
+
"grad_norm": 0.7103596925735474,
|
| 1331 |
+
"learning_rate": 1.6053439101397257e-06,
|
| 1332 |
+
"loss": 0.2685,
|
| 1333 |
+
"step": 189
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.3135678391959797,
|
| 1337 |
+
"grad_norm": 0.7081505656242371,
|
| 1338 |
+
"learning_rate": 1.5535039629704467e-06,
|
| 1339 |
+
"loss": 0.2717,
|
| 1340 |
+
"step": 190
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.3256281407035178,
|
| 1344 |
+
"grad_norm": 0.7069767117500305,
|
| 1345 |
+
"learning_rate": 1.502360458946232e-06,
|
| 1346 |
+
"loss": 0.2569,
|
| 1347 |
+
"step": 191
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.3376884422110553,
|
| 1351 |
+
"grad_norm": 0.683577835559845,
|
| 1352 |
+
"learning_rate": 1.451923732778745e-06,
|
| 1353 |
+
"loss": 0.2682,
|
| 1354 |
+
"step": 192
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.349748743718593,
|
| 1358 |
+
"grad_norm": 0.6355082392692566,
|
| 1359 |
+
"learning_rate": 1.4022039763590595e-06,
|
| 1360 |
+
"loss": 0.263,
|
| 1361 |
+
"step": 193
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.3618090452261304,
|
| 1365 |
+
"grad_norm": 0.6168832182884216,
|
| 1366 |
+
"learning_rate": 1.3532112366981598e-06,
|
| 1367 |
+
"loss": 0.2759,
|
| 1368 |
+
"step": 194
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.3738693467336685,
|
| 1372 |
+
"grad_norm": 0.6499370336532593,
|
| 1373 |
+
"learning_rate": 1.3049554138967052e-06,
|
| 1374 |
+
"loss": 0.2345,
|
| 1375 |
+
"step": 195
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 2.385929648241206,
|
| 1379 |
+
"grad_norm": 0.6336264610290527,
|
| 1380 |
+
"learning_rate": 1.257446259144494e-06,
|
| 1381 |
+
"loss": 0.2355,
|
| 1382 |
+
"step": 196
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.3979899497487436,
|
| 1386 |
+
"grad_norm": 0.6290974617004395,
|
| 1387 |
+
"learning_rate": 1.210693372750017e-06,
|
| 1388 |
+
"loss": 0.2882,
|
| 1389 |
+
"step": 197
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 2.4100502512562816,
|
| 1393 |
+
"grad_norm": 0.7314269542694092,
|
| 1394 |
+
"learning_rate": 1.1647062022004845e-06,
|
| 1395 |
+
"loss": 0.2363,
|
| 1396 |
+
"step": 198
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.422110552763819,
|
| 1400 |
+
"grad_norm": 0.7161564826965332,
|
| 1401 |
+
"learning_rate": 1.1194940402527566e-06,
|
| 1402 |
+
"loss": 0.3291,
|
| 1403 |
+
"step": 199
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 2.4341708542713567,
|
| 1407 |
+
"grad_norm": 0.7857349514961243,
|
| 1408 |
+
"learning_rate": 1.075066023055527e-06,
|
| 1409 |
+
"loss": 0.2589,
|
| 1410 |
+
"step": 200
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.4462311557788947,
|
| 1414 |
+
"grad_norm": 0.6059231162071228,
|
| 1415 |
+
"learning_rate": 1.0314311283031531e-06,
|
| 1416 |
+
"loss": 0.2608,
|
| 1417 |
+
"step": 201
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.4582914572864323,
|
| 1421 |
+
"grad_norm": 0.70367032289505,
|
| 1422 |
+
"learning_rate": 9.885981734215094e-07,
|
| 1423 |
+
"loss": 0.2862,
|
| 1424 |
+
"step": 202
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.47035175879397,
|
| 1428 |
+
"grad_norm": 0.6592365503311157,
|
| 1429 |
+
"learning_rate": 9.465758137862264e-07,
|
| 1430 |
+
"loss": 0.2493,
|
| 1431 |
+
"step": 203
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 2.4824120603015074,
|
| 1435 |
+
"grad_norm": 0.6650649905204773,
|
| 1436 |
+
"learning_rate": 9.053725409736752e-07,
|
| 1437 |
+
"loss": 0.2684,
|
| 1438 |
+
"step": 204
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.4944723618090454,
|
| 1442 |
+
"grad_norm": 0.7488141059875488,
|
| 1443 |
+
"learning_rate": 8.649966810450472e-07,
|
| 1444 |
+
"loss": 0.3274,
|
| 1445 |
+
"step": 205
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.506532663316583,
|
| 1449 |
+
"grad_norm": 0.7090576887130737,
|
| 1450 |
+
"learning_rate": 8.254563928638892e-07,
|
| 1451 |
+
"loss": 0.2405,
|
| 1452 |
+
"step": 206
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.5185929648241205,
|
| 1456 |
+
"grad_norm": 0.6562180519104004,
|
| 1457 |
+
"learning_rate": 7.86759666447412e-07,
|
| 1458 |
+
"loss": 0.2868,
|
| 1459 |
+
"step": 207
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.530653266331658,
|
| 1463 |
+
"grad_norm": 0.7169398665428162,
|
| 1464 |
+
"learning_rate": 7.489143213519301e-07,
|
| 1465 |
+
"loss": 0.2584,
|
| 1466 |
+
"step": 208
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.542713567839196,
|
| 1470 |
+
"grad_norm": 0.6712690591812134,
|
| 1471 |
+
"learning_rate": 7.119280050927407e-07,
|
| 1472 |
+
"loss": 0.2817,
|
| 1473 |
+
"step": 209
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.5547738693467337,
|
| 1477 |
+
"grad_norm": 0.6303477883338928,
|
| 1478 |
+
"learning_rate": 6.758081915987669e-07,
|
| 1479 |
+
"loss": 0.2781,
|
| 1480 |
+
"step": 210
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.5668341708542712,
|
| 1484 |
+
"grad_norm": 0.6625389456748962,
|
| 1485 |
+
"learning_rate": 6.405621797022848e-07,
|
| 1486 |
+
"loss": 0.2747,
|
| 1487 |
+
"step": 211
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.5788944723618092,
|
| 1491 |
+
"grad_norm": 0.6601296067237854,
|
| 1492 |
+
"learning_rate": 6.061970916640236e-07,
|
| 1493 |
+
"loss": 0.2506,
|
| 1494 |
+
"step": 212
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.590954773869347,
|
| 1498 |
+
"grad_norm": 0.6208679676055908,
|
| 1499 |
+
"learning_rate": 5.727198717339511e-07,
|
| 1500 |
+
"loss": 0.2466,
|
| 1501 |
+
"step": 213
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 2.6030150753768844,
|
| 1505 |
+
"grad_norm": 0.6671913266181946,
|
| 1506 |
+
"learning_rate": 5.401372847480285e-07,
|
| 1507 |
+
"loss": 0.2891,
|
| 1508 |
+
"step": 214
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.6150753768844224,
|
| 1512 |
+
"grad_norm": 0.7155910134315491,
|
| 1513 |
+
"learning_rate": 5.084559147612244e-07,
|
| 1514 |
+
"loss": 0.2778,
|
| 1515 |
+
"step": 215
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 2.62713567839196,
|
| 1519 |
+
"grad_norm": 0.6413914561271667,
|
| 1520 |
+
"learning_rate": 4.776821637170525e-07,
|
| 1521 |
+
"loss": 0.271,
|
| 1522 |
+
"step": 216
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.6391959798994975,
|
| 1526 |
+
"grad_norm": 0.6633293032646179,
|
| 1527 |
+
"learning_rate": 4.4782225015391754e-07,
|
| 1528 |
+
"loss": 0.2594,
|
| 1529 |
+
"step": 217
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 2.651256281407035,
|
| 1533 |
+
"grad_norm": 0.6527714729309082,
|
| 1534 |
+
"learning_rate": 4.1888220794851386e-07,
|
| 1535 |
+
"loss": 0.2789,
|
| 1536 |
+
"step": 218
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.6633165829145726,
|
| 1540 |
+
"grad_norm": 0.7054597735404968,
|
| 1541 |
+
"learning_rate": 3.908678850965425e-07,
|
| 1542 |
+
"loss": 0.2507,
|
| 1543 |
+
"step": 219
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.6753768844221106,
|
| 1547 |
+
"grad_norm": 0.6884572505950928,
|
| 1548 |
+
"learning_rate": 3.6378494253099307e-07,
|
| 1549 |
+
"loss": 0.2765,
|
| 1550 |
+
"step": 220
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.687437185929648,
|
| 1554 |
+
"grad_norm": 0.6827020645141602,
|
| 1555 |
+
"learning_rate": 3.3763885297822153e-07,
|
| 1556 |
+
"loss": 0.2909,
|
| 1557 |
+
"step": 221
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.6994974874371858,
|
| 1561 |
+
"grad_norm": 0.7231439352035522,
|
| 1562 |
+
"learning_rate": 3.1243489985206097e-07,
|
| 1563 |
+
"loss": 0.2519,
|
| 1564 |
+
"step": 222
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.7115577889447238,
|
| 1568 |
+
"grad_norm": 0.6283535361289978,
|
| 1569 |
+
"learning_rate": 2.8817817618618846e-07,
|
| 1570 |
+
"loss": 0.2676,
|
| 1571 |
+
"step": 223
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 2.7236180904522613,
|
| 1575 |
+
"grad_norm": 0.6308393478393555,
|
| 1576 |
+
"learning_rate": 2.648735836049615e-07,
|
| 1577 |
+
"loss": 0.2562,
|
| 1578 |
+
"step": 224
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 2.735678391959799,
|
| 1582 |
+
"grad_norm": 0.6942286491394043,
|
| 1583 |
+
"learning_rate": 2.4252583133292927e-07,
|
| 1584 |
+
"loss": 0.2797,
|
| 1585 |
+
"step": 225
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 2.747738693467337,
|
| 1589 |
+
"grad_norm": 0.6335775256156921,
|
| 1590 |
+
"learning_rate": 2.2113943524323167e-07,
|
| 1591 |
+
"loss": 0.2605,
|
| 1592 |
+
"step": 226
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.7597989949748745,
|
| 1596 |
+
"grad_norm": 0.698582649230957,
|
| 1597 |
+
"learning_rate": 2.007187169450603e-07,
|
| 1598 |
+
"loss": 0.2635,
|
| 1599 |
+
"step": 227
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 2.771859296482412,
|
| 1603 |
+
"grad_norm": 0.6790350675582886,
|
| 1604 |
+
"learning_rate": 1.8126780291038037e-07,
|
| 1605 |
+
"loss": 0.2416,
|
| 1606 |
+
"step": 228
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.7839195979899496,
|
| 1610 |
+
"grad_norm": 0.644870400428772,
|
| 1611 |
+
"learning_rate": 1.6279062364008446e-07,
|
| 1612 |
+
"loss": 0.2616,
|
| 1613 |
+
"step": 229
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 2.7959798994974876,
|
| 1617 |
+
"grad_norm": 0.5776996612548828,
|
| 1618 |
+
"learning_rate": 1.4529091286973994e-07,
|
| 1619 |
+
"loss": 0.2287,
|
| 1620 |
+
"step": 230
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 2.808040201005025,
|
| 1624 |
+
"grad_norm": 0.5627067685127258,
|
| 1625 |
+
"learning_rate": 1.2877220681510927e-07,
|
| 1626 |
+
"loss": 0.2622,
|
| 1627 |
+
"step": 231
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 2.8201005025125627,
|
| 1631 |
+
"grad_norm": 0.6837164759635925,
|
| 1632 |
+
"learning_rate": 1.1323784345757205e-07,
|
| 1633 |
+
"loss": 0.2804,
|
| 1634 |
+
"step": 232
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.8321608040201003,
|
| 1638 |
+
"grad_norm": 0.7210996747016907,
|
| 1639 |
+
"learning_rate": 9.869096186961025e-08,
|
| 1640 |
+
"loss": 0.2608,
|
| 1641 |
+
"step": 233
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 2.8442211055276383,
|
| 1645 |
+
"grad_norm": 0.738089919090271,
|
| 1646 |
+
"learning_rate": 8.513450158049109e-08,
|
| 1647 |
+
"loss": 0.2967,
|
| 1648 |
+
"step": 234
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.856281407035176,
|
| 1652 |
+
"grad_norm": 0.7792130708694458,
|
| 1653 |
+
"learning_rate": 7.257120198226219e-08,
|
| 1654 |
+
"loss": 0.324,
|
| 1655 |
+
"step": 235
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 2.8683417085427134,
|
| 1659 |
+
"grad_norm": 0.752173662185669,
|
| 1660 |
+
"learning_rate": 6.100360177619946e-08,
|
| 1661 |
+
"loss": 0.2671,
|
| 1662 |
+
"step": 236
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 2.8804020100502514,
|
| 1666 |
+
"grad_norm": 0.7220814228057861,
|
| 1667 |
+
"learning_rate": 5.0434038459801213e-08,
|
| 1668 |
+
"loss": 0.2611,
|
| 1669 |
+
"step": 237
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 2.892462311557789,
|
| 1673 |
+
"grad_norm": 0.632590115070343,
|
| 1674 |
+
"learning_rate": 4.086464785444777e-08,
|
| 1675 |
+
"loss": 0.2658,
|
| 1676 |
+
"step": 238
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.9045226130653266,
|
| 1680 |
+
"grad_norm": 0.938666820526123,
|
| 1681 |
+
"learning_rate": 3.229736367380498e-08,
|
| 1682 |
+
"loss": 0.2621,
|
| 1683 |
+
"step": 239
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.9165829145728646,
|
| 1687 |
+
"grad_norm": 0.6873531341552734,
|
| 1688 |
+
"learning_rate": 2.4733917133077378e-08,
|
| 1689 |
+
"loss": 0.2597,
|
| 1690 |
+
"step": 240
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.928643216080402,
|
| 1694 |
+
"grad_norm": 0.655741274356842,
|
| 1695 |
+
"learning_rate": 1.8175836599173545e-08,
|
| 1696 |
+
"loss": 0.254,
|
| 1697 |
+
"step": 241
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.9407035175879397,
|
| 1701 |
+
"grad_norm": 0.6553964614868164,
|
| 1702 |
+
"learning_rate": 1.2624447281867625e-08,
|
| 1703 |
+
"loss": 0.2697,
|
| 1704 |
+
"step": 242
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 2.9527638190954772,
|
| 1708 |
+
"grad_norm": 0.7132120132446289,
|
| 1709 |
+
"learning_rate": 8.080870966008513e-09,
|
| 1710 |
+
"loss": 0.2788,
|
| 1711 |
+
"step": 243
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.964824120603015,
|
| 1715 |
+
"grad_norm": 0.7077562212944031,
|
| 1716 |
+
"learning_rate": 4.546025784837316e-09,
|
| 1717 |
+
"loss": 0.2511,
|
| 1718 |
+
"step": 244
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.976884422110553,
|
| 1722 |
+
"grad_norm": 0.6771462559700012,
|
| 1723 |
+
"learning_rate": 2.0206260344590724e-09,
|
| 1724 |
+
"loss": 0.2715,
|
| 1725 |
+
"step": 245
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.9889447236180904,
|
| 1729 |
+
"grad_norm": 0.5981936454772949,
|
| 1730 |
+
"learning_rate": 5.051820295032262e-10,
|
| 1731 |
+
"loss": 0.2695,
|
| 1732 |
+
"step": 246
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.9889447236180904,
|
| 1736 |
+
"step": 246,
|
| 1737 |
+
"total_flos": 6.810123424602522e+16,
|
| 1738 |
+
"train_loss": 0.40526863738773317,
|
| 1739 |
+
"train_runtime": 35217.1045,
|
| 1740 |
+
"train_samples_per_second": 0.17,
|
| 1741 |
+
"train_steps_per_second": 0.007
|
| 1742 |
+
}
|
| 1743 |
+
],
|
| 1744 |
+
"logging_steps": 1,
|
| 1745 |
+
"max_steps": 246,
|
| 1746 |
+
"num_input_tokens_seen": 0,
|
| 1747 |
+
"num_train_epochs": 3,
|
| 1748 |
+
"save_steps": 200,
|
| 1749 |
+
"stateful_callbacks": {
|
| 1750 |
+
"TrainerControl": {
|
| 1751 |
+
"args": {
|
| 1752 |
+
"should_epoch_stop": false,
|
| 1753 |
+
"should_evaluate": false,
|
| 1754 |
+
"should_log": false,
|
| 1755 |
+
"should_save": true,
|
| 1756 |
+
"should_training_stop": true
|
| 1757 |
+
},
|
| 1758 |
+
"attributes": {}
|
| 1759 |
+
}
|
| 1760 |
+
},
|
| 1761 |
+
"total_flos": 6.810123424602522e+16,
|
| 1762 |
+
"train_batch_size": 1,
|
| 1763 |
+
"trial_name": null,
|
| 1764 |
+
"trial_params": null
|
| 1765 |
+
}
|
training_loss.png
ADDED
|