End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2059 -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-validation_combine_diverse_correct_qwen_code_14B
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# SWE-BENCH-validation_combine_diverse_correct_qwen_code_14B
|
| 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-validation_combine_diverse_correct_qwen_code_14B
|
|
|
|
| 16 |
|
| 17 |
# SWE-BENCH-validation_combine_diverse_correct_qwen_code_14B
|
| 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-validation_combine_diverse_correct dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.993960310612597,
|
| 3 |
+
"total_flos": 1.345368819194921e+17,
|
| 4 |
+
"train_loss": 0.3956852917973366,
|
| 5 |
+
"train_runtime": 35330.3168,
|
| 6 |
+
"train_samples_per_second": 0.393,
|
| 7 |
+
"train_steps_per_second": 0.008
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.993960310612597,
|
| 3 |
+
"total_flos": 1.345368819194921e+17,
|
| 4 |
+
"train_loss": 0.3956852917973366,
|
| 5 |
+
"train_runtime": 35330.3168,
|
| 6 |
+
"train_samples_per_second": 0.393,
|
| 7 |
+
"train_steps_per_second": 0.008
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2059 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.993960310612597,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 288,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.010353753235547885,
|
| 14 |
+
"grad_norm": 1.51571786403656,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.7479,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.02070750647109577,
|
| 21 |
+
"grad_norm": 1.60509192943573,
|
| 22 |
+
"learning_rate": 3.4482758620689656e-07,
|
| 23 |
+
"loss": 0.732,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.031061259706643658,
|
| 28 |
+
"grad_norm": 1.6861212253570557,
|
| 29 |
+
"learning_rate": 6.896551724137931e-07,
|
| 30 |
+
"loss": 0.7576,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04141501294219154,
|
| 35 |
+
"grad_norm": 1.713498830795288,
|
| 36 |
+
"learning_rate": 1.0344827586206898e-06,
|
| 37 |
+
"loss": 0.7226,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.05176876617773943,
|
| 42 |
+
"grad_norm": 1.6940999031066895,
|
| 43 |
+
"learning_rate": 1.3793103448275862e-06,
|
| 44 |
+
"loss": 0.7759,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.062122519413287315,
|
| 49 |
+
"grad_norm": 1.5657801628112793,
|
| 50 |
+
"learning_rate": 1.724137931034483e-06,
|
| 51 |
+
"loss": 0.7468,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0724762726488352,
|
| 56 |
+
"grad_norm": 1.552438497543335,
|
| 57 |
+
"learning_rate": 2.0689655172413796e-06,
|
| 58 |
+
"loss": 0.7563,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.08283002588438308,
|
| 63 |
+
"grad_norm": 1.3916494846343994,
|
| 64 |
+
"learning_rate": 2.4137931034482762e-06,
|
| 65 |
+
"loss": 0.7695,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.09318377911993098,
|
| 70 |
+
"grad_norm": 1.1299245357513428,
|
| 71 |
+
"learning_rate": 2.7586206896551725e-06,
|
| 72 |
+
"loss": 0.7124,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.10353753235547886,
|
| 77 |
+
"grad_norm": 1.0293270349502563,
|
| 78 |
+
"learning_rate": 3.103448275862069e-06,
|
| 79 |
+
"loss": 0.7115,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.11389128559102675,
|
| 84 |
+
"grad_norm": 0.7872817516326904,
|
| 85 |
+
"learning_rate": 3.448275862068966e-06,
|
| 86 |
+
"loss": 0.6389,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.12424503882657463,
|
| 91 |
+
"grad_norm": 0.8512783050537109,
|
| 92 |
+
"learning_rate": 3.793103448275862e-06,
|
| 93 |
+
"loss": 0.6615,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.13459879206212252,
|
| 98 |
+
"grad_norm": 0.8439010381698608,
|
| 99 |
+
"learning_rate": 4.137931034482759e-06,
|
| 100 |
+
"loss": 0.6583,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.1449525452976704,
|
| 105 |
+
"grad_norm": 0.8007126450538635,
|
| 106 |
+
"learning_rate": 4.482758620689656e-06,
|
| 107 |
+
"loss": 0.6128,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.15530629853321828,
|
| 112 |
+
"grad_norm": 0.7570585012435913,
|
| 113 |
+
"learning_rate": 4.8275862068965525e-06,
|
| 114 |
+
"loss": 0.6433,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.16566005176876616,
|
| 119 |
+
"grad_norm": 0.8431252241134644,
|
| 120 |
+
"learning_rate": 5.172413793103449e-06,
|
| 121 |
+
"loss": 0.639,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.17601380500431407,
|
| 126 |
+
"grad_norm": 0.82211834192276,
|
| 127 |
+
"learning_rate": 5.517241379310345e-06,
|
| 128 |
+
"loss": 0.6646,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.18636755823986195,
|
| 133 |
+
"grad_norm": 0.766636848449707,
|
| 134 |
+
"learning_rate": 5.862068965517242e-06,
|
| 135 |
+
"loss": 0.617,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.19672131147540983,
|
| 140 |
+
"grad_norm": 0.7024624943733215,
|
| 141 |
+
"learning_rate": 6.206896551724138e-06,
|
| 142 |
+
"loss": 0.599,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.2070750647109577,
|
| 147 |
+
"grad_norm": 0.6534252166748047,
|
| 148 |
+
"learning_rate": 6.551724137931035e-06,
|
| 149 |
+
"loss": 0.6,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.21742881794650562,
|
| 154 |
+
"grad_norm": 0.4923899471759796,
|
| 155 |
+
"learning_rate": 6.896551724137932e-06,
|
| 156 |
+
"loss": 0.505,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.2277825711820535,
|
| 161 |
+
"grad_norm": 0.8025439977645874,
|
| 162 |
+
"learning_rate": 7.241379310344828e-06,
|
| 163 |
+
"loss": 0.5636,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.23813632441760138,
|
| 168 |
+
"grad_norm": 0.5880887508392334,
|
| 169 |
+
"learning_rate": 7.586206896551724e-06,
|
| 170 |
+
"loss": 0.5808,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.24849007765314926,
|
| 175 |
+
"grad_norm": 0.6298739314079285,
|
| 176 |
+
"learning_rate": 7.93103448275862e-06,
|
| 177 |
+
"loss": 0.617,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.25884383088869717,
|
| 182 |
+
"grad_norm": 0.6893284320831299,
|
| 183 |
+
"learning_rate": 8.275862068965518e-06,
|
| 184 |
+
"loss": 0.6029,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.26919758412424505,
|
| 189 |
+
"grad_norm": 0.5549190640449524,
|
| 190 |
+
"learning_rate": 8.620689655172414e-06,
|
| 191 |
+
"loss": 0.5683,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.27955133735979293,
|
| 196 |
+
"grad_norm": 0.5266641974449158,
|
| 197 |
+
"learning_rate": 8.965517241379312e-06,
|
| 198 |
+
"loss": 0.5317,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.2899050905953408,
|
| 203 |
+
"grad_norm": 0.5857071876525879,
|
| 204 |
+
"learning_rate": 9.310344827586207e-06,
|
| 205 |
+
"loss": 0.5543,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.3002588438308887,
|
| 210 |
+
"grad_norm": 0.5448468327522278,
|
| 211 |
+
"learning_rate": 9.655172413793105e-06,
|
| 212 |
+
"loss": 0.5266,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.31061259706643657,
|
| 217 |
+
"grad_norm": 0.546934187412262,
|
| 218 |
+
"learning_rate": 1e-05,
|
| 219 |
+
"loss": 0.5046,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.32096635030198445,
|
| 224 |
+
"grad_norm": 0.603276252746582,
|
| 225 |
+
"learning_rate": 9.999632180371776e-06,
|
| 226 |
+
"loss": 0.557,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.33132010353753233,
|
| 231 |
+
"grad_norm": 0.5554037690162659,
|
| 232 |
+
"learning_rate": 9.998528775603612e-06,
|
| 233 |
+
"loss": 0.4942,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.34167385677308026,
|
| 238 |
+
"grad_norm": 0.6425076723098755,
|
| 239 |
+
"learning_rate": 9.996689948037081e-06,
|
| 240 |
+
"loss": 0.5955,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.35202761000862814,
|
| 245 |
+
"grad_norm": 0.5685299634933472,
|
| 246 |
+
"learning_rate": 9.994115968214933e-06,
|
| 247 |
+
"loss": 0.5172,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.362381363244176,
|
| 252 |
+
"grad_norm": 0.5078840255737305,
|
| 253 |
+
"learning_rate": 9.990807214841288e-06,
|
| 254 |
+
"loss": 0.4873,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.3727351164797239,
|
| 259 |
+
"grad_norm": 0.504933774471283,
|
| 260 |
+
"learning_rate": 9.98676417472592e-06,
|
| 261 |
+
"loss": 0.4851,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.3830888697152718,
|
| 266 |
+
"grad_norm": 0.5034538507461548,
|
| 267 |
+
"learning_rate": 9.981987442712634e-06,
|
| 268 |
+
"loss": 0.4837,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.39344262295081966,
|
| 273 |
+
"grad_norm": 0.5521865487098694,
|
| 274 |
+
"learning_rate": 9.976477721591746e-06,
|
| 275 |
+
"loss": 0.4959,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.40379637618636754,
|
| 280 |
+
"grad_norm": 0.6357220411300659,
|
| 281 |
+
"learning_rate": 9.97023582199669e-06,
|
| 282 |
+
"loss": 0.5141,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.4141501294219154,
|
| 287 |
+
"grad_norm": 0.6331016421318054,
|
| 288 |
+
"learning_rate": 9.963262662284735e-06,
|
| 289 |
+
"loss": 0.5679,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.4245038826574633,
|
| 294 |
+
"grad_norm": 0.5008139610290527,
|
| 295 |
+
"learning_rate": 9.955559268401893e-06,
|
| 296 |
+
"loss": 0.4824,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.43485763589301124,
|
| 301 |
+
"grad_norm": 0.5969840288162231,
|
| 302 |
+
"learning_rate": 9.947126773731949e-06,
|
| 303 |
+
"loss": 0.5263,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.4452113891285591,
|
| 308 |
+
"grad_norm": 0.6160205602645874,
|
| 309 |
+
"learning_rate": 9.937966418929725e-06,
|
| 310 |
+
"loss": 0.55,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.455565142364107,
|
| 315 |
+
"grad_norm": 0.5549020171165466,
|
| 316 |
+
"learning_rate": 9.928079551738542e-06,
|
| 317 |
+
"loss": 0.4922,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.4659188955996549,
|
| 322 |
+
"grad_norm": 0.48422175645828247,
|
| 323 |
+
"learning_rate": 9.917467626791925e-06,
|
| 324 |
+
"loss": 0.4621,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.47627264883520276,
|
| 329 |
+
"grad_norm": 0.5353500247001648,
|
| 330 |
+
"learning_rate": 9.90613220539959e-06,
|
| 331 |
+
"loss": 0.5049,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.48662640207075064,
|
| 336 |
+
"grad_norm": 0.4994064271450043,
|
| 337 |
+
"learning_rate": 9.89407495531773e-06,
|
| 338 |
+
"loss": 0.4702,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.4969801553062985,
|
| 343 |
+
"grad_norm": 0.5894600749015808,
|
| 344 |
+
"learning_rate": 9.881297650503641e-06,
|
| 345 |
+
"loss": 0.4999,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.5073339085418465,
|
| 350 |
+
"grad_norm": 0.5493610501289368,
|
| 351 |
+
"learning_rate": 9.867802170854724e-06,
|
| 352 |
+
"loss": 0.5259,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.5176876617773943,
|
| 357 |
+
"grad_norm": 0.533173680305481,
|
| 358 |
+
"learning_rate": 9.853590501931905e-06,
|
| 359 |
+
"loss": 0.4437,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.5280414150129422,
|
| 364 |
+
"grad_norm": 0.5561060905456543,
|
| 365 |
+
"learning_rate": 9.838664734667496e-06,
|
| 366 |
+
"loss": 0.5207,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.5383951682484901,
|
| 371 |
+
"grad_norm": 0.46265217661857605,
|
| 372 |
+
"learning_rate": 9.82302706505756e-06,
|
| 373 |
+
"loss": 0.5003,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.548748921484038,
|
| 378 |
+
"grad_norm": 0.46626946330070496,
|
| 379 |
+
"learning_rate": 9.806679793838829e-06,
|
| 380 |
+
"loss": 0.4519,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.5591026747195859,
|
| 385 |
+
"grad_norm": 0.5511159300804138,
|
| 386 |
+
"learning_rate": 9.78962532615019e-06,
|
| 387 |
+
"loss": 0.5469,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.5694564279551337,
|
| 392 |
+
"grad_norm": 0.5796953439712524,
|
| 393 |
+
"learning_rate": 9.771866171178832e-06,
|
| 394 |
+
"loss": 0.5073,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5798101811906816,
|
| 399 |
+
"grad_norm": 0.5126736164093018,
|
| 400 |
+
"learning_rate": 9.753404941791063e-06,
|
| 401 |
+
"loss": 0.4626,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5901639344262295,
|
| 406 |
+
"grad_norm": 0.5270448327064514,
|
| 407 |
+
"learning_rate": 9.734244354147897e-06,
|
| 408 |
+
"loss": 0.5279,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.6005176876617774,
|
| 413 |
+
"grad_norm": 0.4986811578273773,
|
| 414 |
+
"learning_rate": 9.714387227305422e-06,
|
| 415 |
+
"loss": 0.5055,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.6108714408973253,
|
| 420 |
+
"grad_norm": 0.5633049011230469,
|
| 421 |
+
"learning_rate": 9.693836482800044e-06,
|
| 422 |
+
"loss": 0.4659,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.6212251941328731,
|
| 427 |
+
"grad_norm": 0.5301141738891602,
|
| 428 |
+
"learning_rate": 9.672595144218646e-06,
|
| 429 |
+
"loss": 0.4821,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.631578947368421,
|
| 434 |
+
"grad_norm": 0.5037074089050293,
|
| 435 |
+
"learning_rate": 9.65066633675373e-06,
|
| 436 |
+
"loss": 0.4258,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.6419327006039689,
|
| 441 |
+
"grad_norm": 0.595179557800293,
|
| 442 |
+
"learning_rate": 9.628053286743619e-06,
|
| 443 |
+
"loss": 0.4869,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.6522864538395168,
|
| 448 |
+
"grad_norm": 0.4926295280456543,
|
| 449 |
+
"learning_rate": 9.604759321197775e-06,
|
| 450 |
+
"loss": 0.4904,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.6626402070750647,
|
| 455 |
+
"grad_norm": 0.7002621293067932,
|
| 456 |
+
"learning_rate": 9.580787867307293e-06,
|
| 457 |
+
"loss": 0.5157,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.6729939603106126,
|
| 462 |
+
"grad_norm": 0.5082715749740601,
|
| 463 |
+
"learning_rate": 9.55614245194068e-06,
|
| 464 |
+
"loss": 0.4694,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.6833477135461605,
|
| 469 |
+
"grad_norm": 0.6984162926673889,
|
| 470 |
+
"learning_rate": 9.53082670112494e-06,
|
| 471 |
+
"loss": 0.4878,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.6937014667817084,
|
| 476 |
+
"grad_norm": 0.49663907289505005,
|
| 477 |
+
"learning_rate": 9.504844339512096e-06,
|
| 478 |
+
"loss": 0.4975,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.7040552200172563,
|
| 483 |
+
"grad_norm": 0.5490703582763672,
|
| 484 |
+
"learning_rate": 9.478199189831184e-06,
|
| 485 |
+
"loss": 0.4884,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.7144089732528042,
|
| 490 |
+
"grad_norm": 0.5001339316368103,
|
| 491 |
+
"learning_rate": 9.450895172325822e-06,
|
| 492 |
+
"loss": 0.4872,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.724762726488352,
|
| 497 |
+
"grad_norm": 0.5249024033546448,
|
| 498 |
+
"learning_rate": 9.422936304177439e-06,
|
| 499 |
+
"loss": 0.4751,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.7351164797238999,
|
| 504 |
+
"grad_norm": 0.4757773280143738,
|
| 505 |
+
"learning_rate": 9.394326698914229e-06,
|
| 506 |
+
"loss": 0.4412,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.7454702329594478,
|
| 511 |
+
"grad_norm": 0.489673912525177,
|
| 512 |
+
"learning_rate": 9.365070565805941e-06,
|
| 513 |
+
"loss": 0.4475,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.7558239861949957,
|
| 518 |
+
"grad_norm": 0.5110509395599365,
|
| 519 |
+
"learning_rate": 9.335172209244577e-06,
|
| 520 |
+
"loss": 0.4977,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.7661777394305436,
|
| 525 |
+
"grad_norm": 0.5113692283630371,
|
| 526 |
+
"learning_rate": 9.304636028111093e-06,
|
| 527 |
+
"loss": 0.4861,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.7765314926660914,
|
| 532 |
+
"grad_norm": 0.5202982425689697,
|
| 533 |
+
"learning_rate": 9.273466515128209e-06,
|
| 534 |
+
"loss": 0.4793,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.7868852459016393,
|
| 539 |
+
"grad_norm": 0.5220551490783691,
|
| 540 |
+
"learning_rate": 9.241668256199392e-06,
|
| 541 |
+
"loss": 0.4584,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.7972389991371872,
|
| 546 |
+
"grad_norm": 0.5219978094100952,
|
| 547 |
+
"learning_rate": 9.209245929734156e-06,
|
| 548 |
+
"loss": 0.4829,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.8075927523727351,
|
| 553 |
+
"grad_norm": 0.49131661653518677,
|
| 554 |
+
"learning_rate": 9.176204305959727e-06,
|
| 555 |
+
"loss": 0.4953,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.817946505608283,
|
| 560 |
+
"grad_norm": 0.5877750515937805,
|
| 561 |
+
"learning_rate": 9.142548246219212e-06,
|
| 562 |
+
"loss": 0.4603,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.8283002588438308,
|
| 567 |
+
"grad_norm": 0.5228042602539062,
|
| 568 |
+
"learning_rate": 9.108282702256366e-06,
|
| 569 |
+
"loss": 0.4412,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.8386540120793787,
|
| 574 |
+
"grad_norm": 0.505235493183136,
|
| 575 |
+
"learning_rate": 9.073412715487045e-06,
|
| 576 |
+
"loss": 0.4558,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.8490077653149266,
|
| 581 |
+
"grad_norm": 0.5228678584098816,
|
| 582 |
+
"learning_rate": 9.037943416257475e-06,
|
| 583 |
+
"loss": 0.452,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.8593615185504746,
|
| 588 |
+
"grad_norm": 0.4974634647369385,
|
| 589 |
+
"learning_rate": 9.001880023089442e-06,
|
| 590 |
+
"loss": 0.4715,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.8697152717860225,
|
| 595 |
+
"grad_norm": 0.49015647172927856,
|
| 596 |
+
"learning_rate": 8.96522784191249e-06,
|
| 597 |
+
"loss": 0.4535,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.8800690250215704,
|
| 602 |
+
"grad_norm": 0.5977384448051453,
|
| 603 |
+
"learning_rate": 8.927992265283282e-06,
|
| 604 |
+
"loss": 0.4342,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.8904227782571182,
|
| 609 |
+
"grad_norm": 0.4676472842693329,
|
| 610 |
+
"learning_rate": 8.890178771592198e-06,
|
| 611 |
+
"loss": 0.4547,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.9007765314926661,
|
| 616 |
+
"grad_norm": 0.4988389015197754,
|
| 617 |
+
"learning_rate": 8.851792924257316e-06,
|
| 618 |
+
"loss": 0.4548,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.911130284728214,
|
| 623 |
+
"grad_norm": 0.6295838952064514,
|
| 624 |
+
"learning_rate": 8.812840370905872e-06,
|
| 625 |
+
"loss": 0.4541,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.9214840379637619,
|
| 630 |
+
"grad_norm": 0.4959602355957031,
|
| 631 |
+
"learning_rate": 8.773326842543348e-06,
|
| 632 |
+
"loss": 0.497,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.9318377911993098,
|
| 637 |
+
"grad_norm": 0.5431574583053589,
|
| 638 |
+
"learning_rate": 8.733258152710262e-06,
|
| 639 |
+
"loss": 0.4835,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.9421915444348576,
|
| 644 |
+
"grad_norm": 0.5309696793556213,
|
| 645 |
+
"learning_rate": 8.692640196626859e-06,
|
| 646 |
+
"loss": 0.4916,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.9525452976704055,
|
| 651 |
+
"grad_norm": 0.4573999047279358,
|
| 652 |
+
"learning_rate": 8.651478950325739e-06,
|
| 653 |
+
"loss": 0.4552,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.9628990509059534,
|
| 658 |
+
"grad_norm": 0.5374126434326172,
|
| 659 |
+
"learning_rate": 8.609780469772623e-06,
|
| 660 |
+
"loss": 0.4602,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.9732528041415013,
|
| 665 |
+
"grad_norm": 0.5425100326538086,
|
| 666 |
+
"learning_rate": 8.567550889975362e-06,
|
| 667 |
+
"loss": 0.4887,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.9836065573770492,
|
| 672 |
+
"grad_norm": 0.5454104542732239,
|
| 673 |
+
"learning_rate": 8.524796424081291e-06,
|
| 674 |
+
"loss": 0.4663,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.993960310612597,
|
| 679 |
+
"grad_norm": 0.5172043442726135,
|
| 680 |
+
"learning_rate": 8.481523362463111e-06,
|
| 681 |
+
"loss": 0.476,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.010353753235548,
|
| 686 |
+
"grad_norm": 1.4870988130569458,
|
| 687 |
+
"learning_rate": 8.437738071793394e-06,
|
| 688 |
+
"loss": 0.7575,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.0207075064710958,
|
| 693 |
+
"grad_norm": 0.4584636390209198,
|
| 694 |
+
"learning_rate": 8.393446994107876e-06,
|
| 695 |
+
"loss": 0.3624,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.0310612597066438,
|
| 700 |
+
"grad_norm": 0.46639755368232727,
|
| 701 |
+
"learning_rate": 8.348656645857648e-06,
|
| 702 |
+
"loss": 0.3567,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.0414150129421915,
|
| 707 |
+
"grad_norm": 0.49347373843193054,
|
| 708 |
+
"learning_rate": 8.303373616950408e-06,
|
| 709 |
+
"loss": 0.404,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.0517687661777395,
|
| 714 |
+
"grad_norm": 0.45795491337776184,
|
| 715 |
+
"learning_rate": 8.257604569780898e-06,
|
| 716 |
+
"loss": 0.358,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.0621225194132873,
|
| 721 |
+
"grad_norm": 0.4563411474227905,
|
| 722 |
+
"learning_rate": 8.21135623825068e-06,
|
| 723 |
+
"loss": 0.363,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.0724762726488353,
|
| 728 |
+
"grad_norm": 0.49531012773513794,
|
| 729 |
+
"learning_rate": 8.164635426777404e-06,
|
| 730 |
+
"loss": 0.3619,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.082830025884383,
|
| 735 |
+
"grad_norm": 0.5267120599746704,
|
| 736 |
+
"learning_rate": 8.117449009293668e-06,
|
| 737 |
+
"loss": 0.3751,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.093183779119931,
|
| 742 |
+
"grad_norm": 0.5300487875938416,
|
| 743 |
+
"learning_rate": 8.069803928235689e-06,
|
| 744 |
+
"loss": 0.4263,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.1035375323554788,
|
| 749 |
+
"grad_norm": 0.5951116681098938,
|
| 750 |
+
"learning_rate": 8.021707193521865e-06,
|
| 751 |
+
"loss": 0.3849,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.1138912855910268,
|
| 756 |
+
"grad_norm": 0.5268257260322571,
|
| 757 |
+
"learning_rate": 7.973165881521435e-06,
|
| 758 |
+
"loss": 0.3909,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.1242450388265746,
|
| 763 |
+
"grad_norm": 0.5092282891273499,
|
| 764 |
+
"learning_rate": 7.924187134013323e-06,
|
| 765 |
+
"loss": 0.372,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.1345987920621226,
|
| 770 |
+
"grad_norm": 0.4802848696708679,
|
| 771 |
+
"learning_rate": 7.874778157135416e-06,
|
| 772 |
+
"loss": 0.3604,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.1449525452976703,
|
| 777 |
+
"grad_norm": 0.5321550965309143,
|
| 778 |
+
"learning_rate": 7.824946220324313e-06,
|
| 779 |
+
"loss": 0.344,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.1553062985332183,
|
| 784 |
+
"grad_norm": 0.5177443623542786,
|
| 785 |
+
"learning_rate": 7.774698655245802e-06,
|
| 786 |
+
"loss": 0.3529,
|
| 787 |
+
"step": 111
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.165660051768766,
|
| 791 |
+
"grad_norm": 0.568428635597229,
|
| 792 |
+
"learning_rate": 7.724042854716169e-06,
|
| 793 |
+
"loss": 0.3796,
|
| 794 |
+
"step": 112
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.176013805004314,
|
| 798 |
+
"grad_norm": 0.4849674701690674,
|
| 799 |
+
"learning_rate": 7.6729862716145e-06,
|
| 800 |
+
"loss": 0.3657,
|
| 801 |
+
"step": 113
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.186367558239862,
|
| 805 |
+
"grad_norm": 0.5158590078353882,
|
| 806 |
+
"learning_rate": 7.621536417786159e-06,
|
| 807 |
+
"loss": 0.3641,
|
| 808 |
+
"step": 114
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.1967213114754098,
|
| 812 |
+
"grad_norm": 0.5496783256530762,
|
| 813 |
+
"learning_rate": 7.56970086293759e-06,
|
| 814 |
+
"loss": 0.384,
|
| 815 |
+
"step": 115
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.2070750647109576,
|
| 819 |
+
"grad_norm": 0.5463497638702393,
|
| 820 |
+
"learning_rate": 7.5174872335226e-06,
|
| 821 |
+
"loss": 0.3504,
|
| 822 |
+
"step": 116
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.2174288179465056,
|
| 826 |
+
"grad_norm": 0.4986291527748108,
|
| 827 |
+
"learning_rate": 7.464903211620291e-06,
|
| 828 |
+
"loss": 0.3564,
|
| 829 |
+
"step": 117
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.2277825711820536,
|
| 833 |
+
"grad_norm": 0.4728027284145355,
|
| 834 |
+
"learning_rate": 7.4119565338048195e-06,
|
| 835 |
+
"loss": 0.336,
|
| 836 |
+
"step": 118
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.2381363244176014,
|
| 840 |
+
"grad_norm": 0.5206460356712341,
|
| 841 |
+
"learning_rate": 7.358654990007123e-06,
|
| 842 |
+
"loss": 0.3829,
|
| 843 |
+
"step": 119
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.2484900776531493,
|
| 847 |
+
"grad_norm": 0.581432044506073,
|
| 848 |
+
"learning_rate": 7.305006422368811e-06,
|
| 849 |
+
"loss": 0.3567,
|
| 850 |
+
"step": 120
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.2588438308886971,
|
| 854 |
+
"grad_norm": 0.5963355302810669,
|
| 855 |
+
"learning_rate": 7.251018724088367e-06,
|
| 856 |
+
"loss": 0.3624,
|
| 857 |
+
"step": 121
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.269197584124245,
|
| 861 |
+
"grad_norm": 0.575742244720459,
|
| 862 |
+
"learning_rate": 7.196699838259834e-06,
|
| 863 |
+
"loss": 0.358,
|
| 864 |
+
"step": 122
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.2795513373597929,
|
| 868 |
+
"grad_norm": 0.4903321862220764,
|
| 869 |
+
"learning_rate": 7.142057756704168e-06,
|
| 870 |
+
"loss": 0.3766,
|
| 871 |
+
"step": 123
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.2899050905953409,
|
| 875 |
+
"grad_norm": 0.5558143258094788,
|
| 876 |
+
"learning_rate": 7.087100518793421e-06,
|
| 877 |
+
"loss": 0.3702,
|
| 878 |
+
"step": 124
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.3002588438308886,
|
| 882 |
+
"grad_norm": 0.4871218800544739,
|
| 883 |
+
"learning_rate": 7.031836210267915e-06,
|
| 884 |
+
"loss": 0.3891,
|
| 885 |
+
"step": 125
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.3106125970664366,
|
| 889 |
+
"grad_norm": 0.5520638823509216,
|
| 890 |
+
"learning_rate": 6.976272962046619e-06,
|
| 891 |
+
"loss": 0.3843,
|
| 892 |
+
"step": 126
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.3209663503019844,
|
| 896 |
+
"grad_norm": 0.5239480137825012,
|
| 897 |
+
"learning_rate": 6.920418949030856e-06,
|
| 898 |
+
"loss": 0.3673,
|
| 899 |
+
"step": 127
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.3313201035375324,
|
| 903 |
+
"grad_norm": 0.5718082785606384,
|
| 904 |
+
"learning_rate": 6.864282388901544e-06,
|
| 905 |
+
"loss": 0.3498,
|
| 906 |
+
"step": 128
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.3416738567730802,
|
| 910 |
+
"grad_norm": 0.519551694393158,
|
| 911 |
+
"learning_rate": 6.807871540910155e-06,
|
| 912 |
+
"loss": 0.3577,
|
| 913 |
+
"step": 129
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.3520276100086281,
|
| 917 |
+
"grad_norm": 0.5227222442626953,
|
| 918 |
+
"learning_rate": 6.751194704663544e-06,
|
| 919 |
+
"loss": 0.382,
|
| 920 |
+
"step": 130
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.3623813632441761,
|
| 924 |
+
"grad_norm": 0.5666651129722595,
|
| 925 |
+
"learning_rate": 6.694260218902845e-06,
|
| 926 |
+
"loss": 0.3512,
|
| 927 |
+
"step": 131
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.372735116479724,
|
| 931 |
+
"grad_norm": 0.5269270539283752,
|
| 932 |
+
"learning_rate": 6.637076460276612e-06,
|
| 933 |
+
"loss": 0.3925,
|
| 934 |
+
"step": 132
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.3830888697152717,
|
| 938 |
+
"grad_norm": 0.5576265454292297,
|
| 939 |
+
"learning_rate": 6.579651842108381e-06,
|
| 940 |
+
"loss": 0.3616,
|
| 941 |
+
"step": 133
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.3934426229508197,
|
| 945 |
+
"grad_norm": 0.5078628659248352,
|
| 946 |
+
"learning_rate": 6.521994813158834e-06,
|
| 947 |
+
"loss": 0.3601,
|
| 948 |
+
"step": 134
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.4037963761863677,
|
| 952 |
+
"grad_norm": 0.5659478902816772,
|
| 953 |
+
"learning_rate": 6.464113856382752e-06,
|
| 954 |
+
"loss": 0.3847,
|
| 955 |
+
"step": 135
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.4141501294219154,
|
| 959 |
+
"grad_norm": 0.6185561418533325,
|
| 960 |
+
"learning_rate": 6.406017487680938e-06,
|
| 961 |
+
"loss": 0.3572,
|
| 962 |
+
"step": 136
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.4245038826574632,
|
| 966 |
+
"grad_norm": 0.5332062244415283,
|
| 967 |
+
"learning_rate": 6.3477142546472836e-06,
|
| 968 |
+
"loss": 0.3917,
|
| 969 |
+
"step": 137
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.4348576358930112,
|
| 973 |
+
"grad_norm": 0.5554406046867371,
|
| 974 |
+
"learning_rate": 6.28921273531119e-06,
|
| 975 |
+
"loss": 0.3873,
|
| 976 |
+
"step": 138
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.4452113891285592,
|
| 980 |
+
"grad_norm": 0.49953603744506836,
|
| 981 |
+
"learning_rate": 6.230521536875494e-06,
|
| 982 |
+
"loss": 0.3733,
|
| 983 |
+
"step": 139
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.455565142364107,
|
| 987 |
+
"grad_norm": 0.5028650164604187,
|
| 988 |
+
"learning_rate": 6.171649294450113e-06,
|
| 989 |
+
"loss": 0.3769,
|
| 990 |
+
"step": 140
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.465918895599655,
|
| 994 |
+
"grad_norm": 0.4856189489364624,
|
| 995 |
+
"learning_rate": 6.112604669781572e-06,
|
| 996 |
+
"loss": 0.3231,
|
| 997 |
+
"step": 141
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.4762726488352027,
|
| 1001 |
+
"grad_norm": 0.4390804171562195,
|
| 1002 |
+
"learning_rate": 6.053396349978632e-06,
|
| 1003 |
+
"loss": 0.3191,
|
| 1004 |
+
"step": 142
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.4866264020707507,
|
| 1008 |
+
"grad_norm": 0.5393751859664917,
|
| 1009 |
+
"learning_rate": 5.994033046234163e-06,
|
| 1010 |
+
"loss": 0.3905,
|
| 1011 |
+
"step": 143
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.4969801553062985,
|
| 1015 |
+
"grad_norm": 0.5311201810836792,
|
| 1016 |
+
"learning_rate": 5.934523492543489e-06,
|
| 1017 |
+
"loss": 0.3678,
|
| 1018 |
+
"step": 144
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.5073339085418465,
|
| 1022 |
+
"grad_norm": 0.5267587304115295,
|
| 1023 |
+
"learning_rate": 5.874876444419377e-06,
|
| 1024 |
+
"loss": 0.3716,
|
| 1025 |
+
"step": 145
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.5176876617773942,
|
| 1029 |
+
"grad_norm": 0.5269693732261658,
|
| 1030 |
+
"learning_rate": 5.815100677603854e-06,
|
| 1031 |
+
"loss": 0.371,
|
| 1032 |
+
"step": 146
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.5280414150129422,
|
| 1036 |
+
"grad_norm": 0.607457160949707,
|
| 1037 |
+
"learning_rate": 5.75520498677705e-06,
|
| 1038 |
+
"loss": 0.392,
|
| 1039 |
+
"step": 147
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.5383951682484902,
|
| 1043 |
+
"grad_norm": 0.5565496683120728,
|
| 1044 |
+
"learning_rate": 5.695198184263259e-06,
|
| 1045 |
+
"loss": 0.3468,
|
| 1046 |
+
"step": 148
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.548748921484038,
|
| 1050 |
+
"grad_norm": 0.5033702850341797,
|
| 1051 |
+
"learning_rate": 5.635089098734394e-06,
|
| 1052 |
+
"loss": 0.3831,
|
| 1053 |
+
"step": 149
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.5591026747195857,
|
| 1057 |
+
"grad_norm": 0.5453012585639954,
|
| 1058 |
+
"learning_rate": 5.574886573911056e-06,
|
| 1059 |
+
"loss": 0.3618,
|
| 1060 |
+
"step": 150
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.5694564279551337,
|
| 1064 |
+
"grad_norm": 0.6656055450439453,
|
| 1065 |
+
"learning_rate": 5.514599467261363e-06,
|
| 1066 |
+
"loss": 0.3432,
|
| 1067 |
+
"step": 151
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.5798101811906817,
|
| 1071 |
+
"grad_norm": 0.4951886534690857,
|
| 1072 |
+
"learning_rate": 5.454236648697776e-06,
|
| 1073 |
+
"loss": 0.357,
|
| 1074 |
+
"step": 152
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.5901639344262295,
|
| 1078 |
+
"grad_norm": 0.5558794736862183,
|
| 1079 |
+
"learning_rate": 5.3938069992720894e-06,
|
| 1080 |
+
"loss": 0.3948,
|
| 1081 |
+
"step": 153
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.6005176876617773,
|
| 1085 |
+
"grad_norm": 0.6537041664123535,
|
| 1086 |
+
"learning_rate": 5.333319409868777e-06,
|
| 1087 |
+
"loss": 0.3607,
|
| 1088 |
+
"step": 154
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.6108714408973253,
|
| 1092 |
+
"grad_norm": 0.5621142983436584,
|
| 1093 |
+
"learning_rate": 5.272782779896898e-06,
|
| 1094 |
+
"loss": 0.4089,
|
| 1095 |
+
"step": 155
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.6212251941328732,
|
| 1099 |
+
"grad_norm": 0.5526349544525146,
|
| 1100 |
+
"learning_rate": 5.212206015980742e-06,
|
| 1101 |
+
"loss": 0.3829,
|
| 1102 |
+
"step": 156
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.631578947368421,
|
| 1106 |
+
"grad_norm": 0.5542426109313965,
|
| 1107 |
+
"learning_rate": 5.151598030649425e-06,
|
| 1108 |
+
"loss": 0.3695,
|
| 1109 |
+
"step": 157
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.6419327006039688,
|
| 1113 |
+
"grad_norm": 0.509752094745636,
|
| 1114 |
+
"learning_rate": 5.090967741025599e-06,
|
| 1115 |
+
"loss": 0.3551,
|
| 1116 |
+
"step": 158
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.6522864538395168,
|
| 1120 |
+
"grad_norm": 0.48508328199386597,
|
| 1121 |
+
"learning_rate": 5.030324067513499e-06,
|
| 1122 |
+
"loss": 0.355,
|
| 1123 |
+
"step": 159
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.6626402070750648,
|
| 1127 |
+
"grad_norm": 0.554058313369751,
|
| 1128 |
+
"learning_rate": 4.969675932486503e-06,
|
| 1129 |
+
"loss": 0.4016,
|
| 1130 |
+
"step": 160
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.6729939603106128,
|
| 1134 |
+
"grad_norm": 0.519527018070221,
|
| 1135 |
+
"learning_rate": 4.909032258974403e-06,
|
| 1136 |
+
"loss": 0.3739,
|
| 1137 |
+
"step": 161
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.6833477135461605,
|
| 1141 |
+
"grad_norm": 0.5133000016212463,
|
| 1142 |
+
"learning_rate": 4.848401969350577e-06,
|
| 1143 |
+
"loss": 0.3261,
|
| 1144 |
+
"step": 162
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.6937014667817083,
|
| 1148 |
+
"grad_norm": 0.6041058301925659,
|
| 1149 |
+
"learning_rate": 4.78779398401926e-06,
|
| 1150 |
+
"loss": 0.3616,
|
| 1151 |
+
"step": 163
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.7040552200172563,
|
| 1155 |
+
"grad_norm": 0.5437895655632019,
|
| 1156 |
+
"learning_rate": 4.7272172201031055e-06,
|
| 1157 |
+
"loss": 0.3612,
|
| 1158 |
+
"step": 164
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.7144089732528043,
|
| 1162 |
+
"grad_norm": 0.4934717118740082,
|
| 1163 |
+
"learning_rate": 4.666680590131225e-06,
|
| 1164 |
+
"loss": 0.3371,
|
| 1165 |
+
"step": 165
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.724762726488352,
|
| 1169 |
+
"grad_norm": 0.5191500782966614,
|
| 1170 |
+
"learning_rate": 4.606193000727913e-06,
|
| 1171 |
+
"loss": 0.3718,
|
| 1172 |
+
"step": 166
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.7351164797238998,
|
| 1176 |
+
"grad_norm": 0.5798118710517883,
|
| 1177 |
+
"learning_rate": 4.545763351302224e-06,
|
| 1178 |
+
"loss": 0.3807,
|
| 1179 |
+
"step": 167
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.7454702329594478,
|
| 1183 |
+
"grad_norm": 0.5294197201728821,
|
| 1184 |
+
"learning_rate": 4.485400532738638e-06,
|
| 1185 |
+
"loss": 0.4224,
|
| 1186 |
+
"step": 168
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.7558239861949958,
|
| 1190 |
+
"grad_norm": 0.6180734634399414,
|
| 1191 |
+
"learning_rate": 4.425113426088945e-06,
|
| 1192 |
+
"loss": 0.3958,
|
| 1193 |
+
"step": 169
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.7661777394305436,
|
| 1197 |
+
"grad_norm": 0.5860722661018372,
|
| 1198 |
+
"learning_rate": 4.364910901265607e-06,
|
| 1199 |
+
"loss": 0.3577,
|
| 1200 |
+
"step": 170
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.7765314926660913,
|
| 1204 |
+
"grad_norm": 0.9322901964187622,
|
| 1205 |
+
"learning_rate": 4.3048018157367435e-06,
|
| 1206 |
+
"loss": 0.3844,
|
| 1207 |
+
"step": 171
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.7868852459016393,
|
| 1211 |
+
"grad_norm": 0.5235133171081543,
|
| 1212 |
+
"learning_rate": 4.244795013222951e-06,
|
| 1213 |
+
"loss": 0.3444,
|
| 1214 |
+
"step": 172
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.7972389991371873,
|
| 1218 |
+
"grad_norm": 0.4588094651699066,
|
| 1219 |
+
"learning_rate": 4.184899322396147e-06,
|
| 1220 |
+
"loss": 0.3345,
|
| 1221 |
+
"step": 173
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.807592752372735,
|
| 1225 |
+
"grad_norm": 0.5539618730545044,
|
| 1226 |
+
"learning_rate": 4.125123555580624e-06,
|
| 1227 |
+
"loss": 0.3668,
|
| 1228 |
+
"step": 174
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.8179465056082829,
|
| 1232 |
+
"grad_norm": 0.48367369174957275,
|
| 1233 |
+
"learning_rate": 4.0654765074565125e-06,
|
| 1234 |
+
"loss": 0.3581,
|
| 1235 |
+
"step": 175
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.8283002588438308,
|
| 1239 |
+
"grad_norm": 0.48612749576568604,
|
| 1240 |
+
"learning_rate": 4.00596695376584e-06,
|
| 1241 |
+
"loss": 0.3596,
|
| 1242 |
+
"step": 176
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.8386540120793788,
|
| 1246 |
+
"grad_norm": 0.5774437189102173,
|
| 1247 |
+
"learning_rate": 3.94660365002137e-06,
|
| 1248 |
+
"loss": 0.3612,
|
| 1249 |
+
"step": 177
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.8490077653149266,
|
| 1253 |
+
"grad_norm": 0.4907125234603882,
|
| 1254 |
+
"learning_rate": 3.887395330218429e-06,
|
| 1255 |
+
"loss": 0.3428,
|
| 1256 |
+
"step": 178
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.8593615185504746,
|
| 1260 |
+
"grad_norm": 0.573658287525177,
|
| 1261 |
+
"learning_rate": 3.8283507055498886e-06,
|
| 1262 |
+
"loss": 0.3782,
|
| 1263 |
+
"step": 179
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.8697152717860224,
|
| 1267 |
+
"grad_norm": 0.5326614379882812,
|
| 1268 |
+
"learning_rate": 3.7694784631245066e-06,
|
| 1269 |
+
"loss": 0.3894,
|
| 1270 |
+
"step": 180
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.8800690250215704,
|
| 1274 |
+
"grad_norm": 0.5278035998344421,
|
| 1275 |
+
"learning_rate": 3.7107872646888115e-06,
|
| 1276 |
+
"loss": 0.4078,
|
| 1277 |
+
"step": 181
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.8904227782571184,
|
| 1281 |
+
"grad_norm": 0.5404361486434937,
|
| 1282 |
+
"learning_rate": 3.6522857453527172e-06,
|
| 1283 |
+
"loss": 0.3752,
|
| 1284 |
+
"step": 182
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.9007765314926661,
|
| 1288 |
+
"grad_norm": 0.5271287560462952,
|
| 1289 |
+
"learning_rate": 3.5939825123190637e-06,
|
| 1290 |
+
"loss": 0.3941,
|
| 1291 |
+
"step": 183
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.911130284728214,
|
| 1295 |
+
"grad_norm": 0.5372259020805359,
|
| 1296 |
+
"learning_rate": 3.5358861436172487e-06,
|
| 1297 |
+
"loss": 0.3955,
|
| 1298 |
+
"step": 184
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.9214840379637619,
|
| 1302 |
+
"grad_norm": 0.5270196199417114,
|
| 1303 |
+
"learning_rate": 3.478005186841167e-06,
|
| 1304 |
+
"loss": 0.3833,
|
| 1305 |
+
"step": 185
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.9318377911993099,
|
| 1309 |
+
"grad_norm": 0.5605195164680481,
|
| 1310 |
+
"learning_rate": 3.4203481578916197e-06,
|
| 1311 |
+
"loss": 0.376,
|
| 1312 |
+
"step": 186
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.9421915444348576,
|
| 1316 |
+
"grad_norm": 0.5661027431488037,
|
| 1317 |
+
"learning_rate": 3.3629235397233894e-06,
|
| 1318 |
+
"loss": 0.3342,
|
| 1319 |
+
"step": 187
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.9525452976704054,
|
| 1323 |
+
"grad_norm": 0.5460498929023743,
|
| 1324 |
+
"learning_rate": 3.305739781097157e-06,
|
| 1325 |
+
"loss": 0.3724,
|
| 1326 |
+
"step": 188
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.9628990509059534,
|
| 1330 |
+
"grad_norm": 0.5860216021537781,
|
| 1331 |
+
"learning_rate": 3.248805295336458e-06,
|
| 1332 |
+
"loss": 0.3924,
|
| 1333 |
+
"step": 189
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.9732528041415014,
|
| 1337 |
+
"grad_norm": 0.5382038950920105,
|
| 1338 |
+
"learning_rate": 3.192128459089846e-06,
|
| 1339 |
+
"loss": 0.3281,
|
| 1340 |
+
"step": 190
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.9836065573770492,
|
| 1344 |
+
"grad_norm": 0.5436641573905945,
|
| 1345 |
+
"learning_rate": 3.1357176110984578e-06,
|
| 1346 |
+
"loss": 0.4028,
|
| 1347 |
+
"step": 191
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.993960310612597,
|
| 1351 |
+
"grad_norm": 0.49211564660072327,
|
| 1352 |
+
"learning_rate": 3.079581050969146e-06,
|
| 1353 |
+
"loss": 0.3261,
|
| 1354 |
+
"step": 192
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.010353753235548,
|
| 1358 |
+
"grad_norm": 1.7585784196853638,
|
| 1359 |
+
"learning_rate": 3.0237270379533823e-06,
|
| 1360 |
+
"loss": 0.6913,
|
| 1361 |
+
"step": 193
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.020707506471096,
|
| 1365 |
+
"grad_norm": 0.5327117443084717,
|
| 1366 |
+
"learning_rate": 2.968163789732087e-06,
|
| 1367 |
+
"loss": 0.2789,
|
| 1368 |
+
"step": 194
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.0310612597066435,
|
| 1372 |
+
"grad_norm": 0.47956615686416626,
|
| 1373 |
+
"learning_rate": 2.912899481206582e-06,
|
| 1374 |
+
"loss": 0.2981,
|
| 1375 |
+
"step": 195
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 2.0414150129421915,
|
| 1379 |
+
"grad_norm": 0.5536096096038818,
|
| 1380 |
+
"learning_rate": 2.8579422432958316e-06,
|
| 1381 |
+
"loss": 0.298,
|
| 1382 |
+
"step": 196
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.0517687661777395,
|
| 1386 |
+
"grad_norm": 0.5187256932258606,
|
| 1387 |
+
"learning_rate": 2.803300161740166e-06,
|
| 1388 |
+
"loss": 0.2826,
|
| 1389 |
+
"step": 197
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 2.0621225194132875,
|
| 1393 |
+
"grad_norm": 0.5327378511428833,
|
| 1394 |
+
"learning_rate": 2.748981275911633e-06,
|
| 1395 |
+
"loss": 0.2798,
|
| 1396 |
+
"step": 198
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.072476272648835,
|
| 1400 |
+
"grad_norm": 0.49033474922180176,
|
| 1401 |
+
"learning_rate": 2.69499357763119e-06,
|
| 1402 |
+
"loss": 0.2546,
|
| 1403 |
+
"step": 199
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 2.082830025884383,
|
| 1407 |
+
"grad_norm": 0.49919557571411133,
|
| 1408 |
+
"learning_rate": 2.641345009992878e-06,
|
| 1409 |
+
"loss": 0.2816,
|
| 1410 |
+
"step": 200
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.093183779119931,
|
| 1414 |
+
"grad_norm": 0.4898358881473541,
|
| 1415 |
+
"learning_rate": 2.5880434661951826e-06,
|
| 1416 |
+
"loss": 0.2739,
|
| 1417 |
+
"step": 201
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.103537532355479,
|
| 1421 |
+
"grad_norm": 0.5661782622337341,
|
| 1422 |
+
"learning_rate": 2.5350967883797095e-06,
|
| 1423 |
+
"loss": 0.2767,
|
| 1424 |
+
"step": 202
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.1138912855910266,
|
| 1428 |
+
"grad_norm": 0.6124740242958069,
|
| 1429 |
+
"learning_rate": 2.4825127664774008e-06,
|
| 1430 |
+
"loss": 0.2797,
|
| 1431 |
+
"step": 203
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 2.1242450388265746,
|
| 1435 |
+
"grad_norm": 0.605502724647522,
|
| 1436 |
+
"learning_rate": 2.4302991370624106e-06,
|
| 1437 |
+
"loss": 0.2815,
|
| 1438 |
+
"step": 204
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.1345987920621226,
|
| 1442 |
+
"grad_norm": 0.6373029351234436,
|
| 1443 |
+
"learning_rate": 2.3784635822138424e-06,
|
| 1444 |
+
"loss": 0.3051,
|
| 1445 |
+
"step": 205
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.1449525452976705,
|
| 1449 |
+
"grad_norm": 0.6217823028564453,
|
| 1450 |
+
"learning_rate": 2.3270137283855022e-06,
|
| 1451 |
+
"loss": 0.2598,
|
| 1452 |
+
"step": 206
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.155306298533218,
|
| 1456 |
+
"grad_norm": 0.5221353769302368,
|
| 1457 |
+
"learning_rate": 2.2759571452838325e-06,
|
| 1458 |
+
"loss": 0.271,
|
| 1459 |
+
"step": 207
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.165660051768766,
|
| 1463 |
+
"grad_norm": 0.5726423859596252,
|
| 1464 |
+
"learning_rate": 2.2253013447541993e-06,
|
| 1465 |
+
"loss": 0.2848,
|
| 1466 |
+
"step": 208
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.176013805004314,
|
| 1470 |
+
"grad_norm": 0.5945622324943542,
|
| 1471 |
+
"learning_rate": 2.17505377967569e-06,
|
| 1472 |
+
"loss": 0.2781,
|
| 1473 |
+
"step": 209
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.186367558239862,
|
| 1477 |
+
"grad_norm": 0.5583497881889343,
|
| 1478 |
+
"learning_rate": 2.125221842864585e-06,
|
| 1479 |
+
"loss": 0.2716,
|
| 1480 |
+
"step": 210
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.19672131147541,
|
| 1484 |
+
"grad_norm": 0.5198562741279602,
|
| 1485 |
+
"learning_rate": 2.075812865986677e-06,
|
| 1486 |
+
"loss": 0.2491,
|
| 1487 |
+
"step": 211
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.2070750647109576,
|
| 1491 |
+
"grad_norm": 0.5728682279586792,
|
| 1492 |
+
"learning_rate": 2.0268341184785674e-06,
|
| 1493 |
+
"loss": 0.2755,
|
| 1494 |
+
"step": 212
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.2174288179465056,
|
| 1498 |
+
"grad_norm": 0.5961595773696899,
|
| 1499 |
+
"learning_rate": 1.978292806478134e-06,
|
| 1500 |
+
"loss": 0.3165,
|
| 1501 |
+
"step": 213
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 2.2277825711820536,
|
| 1505 |
+
"grad_norm": 0.526508629322052,
|
| 1506 |
+
"learning_rate": 1.930196071764312e-06,
|
| 1507 |
+
"loss": 0.2417,
|
| 1508 |
+
"step": 214
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.2381363244176016,
|
| 1512 |
+
"grad_norm": 0.6173946261405945,
|
| 1513 |
+
"learning_rate": 1.8825509907063328e-06,
|
| 1514 |
+
"loss": 0.2748,
|
| 1515 |
+
"step": 215
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 2.248490077653149,
|
| 1519 |
+
"grad_norm": 0.5461782217025757,
|
| 1520 |
+
"learning_rate": 1.8353645732225977e-06,
|
| 1521 |
+
"loss": 0.2696,
|
| 1522 |
+
"step": 216
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.258843830888697,
|
| 1526 |
+
"grad_norm": 0.5328632593154907,
|
| 1527 |
+
"learning_rate": 1.7886437617493206e-06,
|
| 1528 |
+
"loss": 0.2821,
|
| 1529 |
+
"step": 217
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 2.269197584124245,
|
| 1533 |
+
"grad_norm": 0.5614588260650635,
|
| 1534 |
+
"learning_rate": 1.7423954302191047e-06,
|
| 1535 |
+
"loss": 0.2609,
|
| 1536 |
+
"step": 218
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.279551337359793,
|
| 1540 |
+
"grad_norm": 0.5334609150886536,
|
| 1541 |
+
"learning_rate": 1.6966263830495939e-06,
|
| 1542 |
+
"loss": 0.2792,
|
| 1543 |
+
"step": 219
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.2899050905953406,
|
| 1547 |
+
"grad_norm": 0.5810555815696716,
|
| 1548 |
+
"learning_rate": 1.6513433541423529e-06,
|
| 1549 |
+
"loss": 0.2662,
|
| 1550 |
+
"step": 220
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.3002588438308886,
|
| 1554 |
+
"grad_norm": 0.5375015735626221,
|
| 1555 |
+
"learning_rate": 1.6065530058921253e-06,
|
| 1556 |
+
"loss": 0.2748,
|
| 1557 |
+
"step": 221
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.3106125970664366,
|
| 1561 |
+
"grad_norm": 0.5522162914276123,
|
| 1562 |
+
"learning_rate": 1.562261928206608e-06,
|
| 1563 |
+
"loss": 0.2729,
|
| 1564 |
+
"step": 222
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.3209663503019846,
|
| 1568 |
+
"grad_norm": 0.49842172861099243,
|
| 1569 |
+
"learning_rate": 1.5184766375368914e-06,
|
| 1570 |
+
"loss": 0.2551,
|
| 1571 |
+
"step": 223
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 2.331320103537532,
|
| 1575 |
+
"grad_norm": 0.565407931804657,
|
| 1576 |
+
"learning_rate": 1.4752035759187106e-06,
|
| 1577 |
+
"loss": 0.2984,
|
| 1578 |
+
"step": 224
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 2.34167385677308,
|
| 1582 |
+
"grad_norm": 0.5956223011016846,
|
| 1583 |
+
"learning_rate": 1.4324491100246386e-06,
|
| 1584 |
+
"loss": 0.2909,
|
| 1585 |
+
"step": 225
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 2.352027610008628,
|
| 1589 |
+
"grad_norm": 0.6256874799728394,
|
| 1590 |
+
"learning_rate": 1.390219530227378e-06,
|
| 1591 |
+
"loss": 0.2805,
|
| 1592 |
+
"step": 226
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.362381363244176,
|
| 1596 |
+
"grad_norm": 0.6008132696151733,
|
| 1597 |
+
"learning_rate": 1.348521049674264e-06,
|
| 1598 |
+
"loss": 0.2878,
|
| 1599 |
+
"step": 227
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 2.372735116479724,
|
| 1603 |
+
"grad_norm": 0.5860024094581604,
|
| 1604 |
+
"learning_rate": 1.3073598033731427e-06,
|
| 1605 |
+
"loss": 0.2621,
|
| 1606 |
+
"step": 228
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.3830888697152717,
|
| 1610 |
+
"grad_norm": 0.5847429037094116,
|
| 1611 |
+
"learning_rate": 1.2667418472897386e-06,
|
| 1612 |
+
"loss": 0.2816,
|
| 1613 |
+
"step": 229
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 2.3934426229508197,
|
| 1617 |
+
"grad_norm": 0.5728585124015808,
|
| 1618 |
+
"learning_rate": 1.2266731574566536e-06,
|
| 1619 |
+
"loss": 0.2571,
|
| 1620 |
+
"step": 230
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 2.4037963761863677,
|
| 1624 |
+
"grad_norm": 0.5815078020095825,
|
| 1625 |
+
"learning_rate": 1.1871596290941278e-06,
|
| 1626 |
+
"loss": 0.2601,
|
| 1627 |
+
"step": 231
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 2.414150129421915,
|
| 1631 |
+
"grad_norm": 0.6150763034820557,
|
| 1632 |
+
"learning_rate": 1.1482070757426855e-06,
|
| 1633 |
+
"loss": 0.2579,
|
| 1634 |
+
"step": 232
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.424503882657463,
|
| 1638 |
+
"grad_norm": 0.5849730372428894,
|
| 1639 |
+
"learning_rate": 1.1098212284078037e-06,
|
| 1640 |
+
"loss": 0.2751,
|
| 1641 |
+
"step": 233
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 2.434857635893011,
|
| 1645 |
+
"grad_norm": 0.5246581435203552,
|
| 1646 |
+
"learning_rate": 1.07200773471672e-06,
|
| 1647 |
+
"loss": 0.2879,
|
| 1648 |
+
"step": 234
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.445211389128559,
|
| 1652 |
+
"grad_norm": 0.6034648418426514,
|
| 1653 |
+
"learning_rate": 1.0347721580875125e-06,
|
| 1654 |
+
"loss": 0.2753,
|
| 1655 |
+
"step": 235
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 2.455565142364107,
|
| 1659 |
+
"grad_norm": 0.7165918350219727,
|
| 1660 |
+
"learning_rate": 9.981199769105605e-07,
|
| 1661 |
+
"loss": 0.2737,
|
| 1662 |
+
"step": 236
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 2.4659188955996547,
|
| 1666 |
+
"grad_norm": 0.5597122311592102,
|
| 1667 |
+
"learning_rate": 9.62056583742527e-07,
|
| 1668 |
+
"loss": 0.2794,
|
| 1669 |
+
"step": 237
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 2.4762726488352027,
|
| 1673 |
+
"grad_norm": 0.5976839661598206,
|
| 1674 |
+
"learning_rate": 9.26587284512957e-07,
|
| 1675 |
+
"loss": 0.2887,
|
| 1676 |
+
"step": 238
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.4866264020707507,
|
| 1680 |
+
"grad_norm": 0.5533334016799927,
|
| 1681 |
+
"learning_rate": 8.917172977436356e-07,
|
| 1682 |
+
"loss": 0.2695,
|
| 1683 |
+
"step": 239
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.4969801553062987,
|
| 1687 |
+
"grad_norm": 0.5761386752128601,
|
| 1688 |
+
"learning_rate": 8.574517537807897e-07,
|
| 1689 |
+
"loss": 0.2767,
|
| 1690 |
+
"step": 240
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.5073339085418462,
|
| 1694 |
+
"grad_norm": 0.6337144374847412,
|
| 1695 |
+
"learning_rate": 8.237956940402758e-07,
|
| 1696 |
+
"loss": 0.2919,
|
| 1697 |
+
"step": 241
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.5176876617773942,
|
| 1701 |
+
"grad_norm": 0.5748437643051147,
|
| 1702 |
+
"learning_rate": 7.907540702658456e-07,
|
| 1703 |
+
"loss": 0.2378,
|
| 1704 |
+
"step": 242
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 2.528041415012942,
|
| 1708 |
+
"grad_norm": 0.556391716003418,
|
| 1709 |
+
"learning_rate": 7.583317438006094e-07,
|
| 1710 |
+
"loss": 0.2443,
|
| 1711 |
+
"step": 243
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.53839516824849,
|
| 1715 |
+
"grad_norm": 0.5221635103225708,
|
| 1716 |
+
"learning_rate": 7.265334848717931e-07,
|
| 1717 |
+
"loss": 0.2702,
|
| 1718 |
+
"step": 244
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.548748921484038,
|
| 1722 |
+
"grad_norm": 0.6004324555397034,
|
| 1723 |
+
"learning_rate": 6.953639718889077e-07,
|
| 1724 |
+
"loss": 0.2805,
|
| 1725 |
+
"step": 245
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.5591026747195857,
|
| 1729 |
+
"grad_norm": 0.5894324779510498,
|
| 1730 |
+
"learning_rate": 6.648277907554235e-07,
|
| 1731 |
+
"loss": 0.2729,
|
| 1732 |
+
"step": 246
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.5694564279551337,
|
| 1736 |
+
"grad_norm": 0.5316387414932251,
|
| 1737 |
+
"learning_rate": 6.349294341940593e-07,
|
| 1738 |
+
"loss": 0.2904,
|
| 1739 |
+
"step": 247
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.5798101811906817,
|
| 1743 |
+
"grad_norm": 0.533439576625824,
|
| 1744 |
+
"learning_rate": 6.056733010857713e-07,
|
| 1745 |
+
"loss": 0.2742,
|
| 1746 |
+
"step": 248
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 2.5901639344262293,
|
| 1750 |
+
"grad_norm": 0.5897790193557739,
|
| 1751 |
+
"learning_rate": 5.770636958225617e-07,
|
| 1752 |
+
"loss": 0.2753,
|
| 1753 |
+
"step": 249
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.6005176876617773,
|
| 1757 |
+
"grad_norm": 0.5561836957931519,
|
| 1758 |
+
"learning_rate": 5.491048276741784e-07,
|
| 1759 |
+
"loss": 0.2905,
|
| 1760 |
+
"step": 250
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.6108714408973253,
|
| 1764 |
+
"grad_norm": 0.6526253819465637,
|
| 1765 |
+
"learning_rate": 5.218008101688172e-07,
|
| 1766 |
+
"loss": 0.305,
|
| 1767 |
+
"step": 251
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.6212251941328732,
|
| 1771 |
+
"grad_norm": 0.5558742880821228,
|
| 1772 |
+
"learning_rate": 4.951556604879049e-07,
|
| 1773 |
+
"loss": 0.2827,
|
| 1774 |
+
"step": 252
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.6315789473684212,
|
| 1778 |
+
"grad_norm": 0.6124576330184937,
|
| 1779 |
+
"learning_rate": 4.6917329887506133e-07,
|
| 1780 |
+
"loss": 0.2875,
|
| 1781 |
+
"step": 253
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.641932700603969,
|
| 1785 |
+
"grad_norm": 0.5739328861236572,
|
| 1786 |
+
"learning_rate": 4.43857548059321e-07,
|
| 1787 |
+
"loss": 0.263,
|
| 1788 |
+
"step": 254
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.6522864538395168,
|
| 1792 |
+
"grad_norm": 0.558914303779602,
|
| 1793 |
+
"learning_rate": 4.192121326927073e-07,
|
| 1794 |
+
"loss": 0.283,
|
| 1795 |
+
"step": 255
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.6626402070750648,
|
| 1799 |
+
"grad_norm": 0.5733683109283447,
|
| 1800 |
+
"learning_rate": 3.952406788022267e-07,
|
| 1801 |
+
"loss": 0.2675,
|
| 1802 |
+
"step": 256
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.6729939603106128,
|
| 1806 |
+
"grad_norm": 0.5303119421005249,
|
| 1807 |
+
"learning_rate": 3.71946713256382e-07,
|
| 1808 |
+
"loss": 0.2798,
|
| 1809 |
+
"step": 257
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.6833477135461603,
|
| 1813 |
+
"grad_norm": 0.5494239330291748,
|
| 1814 |
+
"learning_rate": 3.4933366324627183e-07,
|
| 1815 |
+
"loss": 0.2981,
|
| 1816 |
+
"step": 258
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.6937014667817083,
|
| 1820 |
+
"grad_norm": 0.5607611536979675,
|
| 1821 |
+
"learning_rate": 3.274048557813553e-07,
|
| 1822 |
+
"loss": 0.2517,
|
| 1823 |
+
"step": 259
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.7040552200172563,
|
| 1827 |
+
"grad_norm": 0.5573918223381042,
|
| 1828 |
+
"learning_rate": 3.061635171999566e-07,
|
| 1829 |
+
"loss": 0.2628,
|
| 1830 |
+
"step": 260
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.7144089732528043,
|
| 1834 |
+
"grad_norm": 0.5762802958488464,
|
| 1835 |
+
"learning_rate": 2.85612772694579e-07,
|
| 1836 |
+
"loss": 0.2717,
|
| 1837 |
+
"step": 261
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.7247627264883523,
|
| 1841 |
+
"grad_norm": 0.5966189503669739,
|
| 1842 |
+
"learning_rate": 2.6575564585210487e-07,
|
| 1843 |
+
"loss": 0.2682,
|
| 1844 |
+
"step": 262
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.7351164797239,
|
| 1848 |
+
"grad_norm": 0.5517677068710327,
|
| 1849 |
+
"learning_rate": 2.4659505820893827e-07,
|
| 1850 |
+
"loss": 0.2774,
|
| 1851 |
+
"step": 263
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.745470232959448,
|
| 1855 |
+
"grad_norm": 0.5355739593505859,
|
| 1856 |
+
"learning_rate": 2.2813382882116986e-07,
|
| 1857 |
+
"loss": 0.2738,
|
| 1858 |
+
"step": 264
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.755823986194996,
|
| 1862 |
+
"grad_norm": 0.5642960667610168,
|
| 1863 |
+
"learning_rate": 2.1037467384981024e-07,
|
| 1864 |
+
"loss": 0.2721,
|
| 1865 |
+
"step": 265
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.7661777394305433,
|
| 1869 |
+
"grad_norm": 0.5581997632980347,
|
| 1870 |
+
"learning_rate": 1.933202061611722e-07,
|
| 1871 |
+
"loss": 0.2744,
|
| 1872 |
+
"step": 266
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.7765314926660913,
|
| 1876 |
+
"grad_norm": 0.609289288520813,
|
| 1877 |
+
"learning_rate": 1.769729349424415e-07,
|
| 1878 |
+
"loss": 0.2625,
|
| 1879 |
+
"step": 267
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.7868852459016393,
|
| 1883 |
+
"grad_norm": 0.5034677982330322,
|
| 1884 |
+
"learning_rate": 1.6133526533250566e-07,
|
| 1885 |
+
"loss": 0.2459,
|
| 1886 |
+
"step": 268
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.7972389991371873,
|
| 1890 |
+
"grad_norm": 0.5479578375816345,
|
| 1891 |
+
"learning_rate": 1.4640949806809523e-07,
|
| 1892 |
+
"loss": 0.2688,
|
| 1893 |
+
"step": 269
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.8075927523727353,
|
| 1897 |
+
"grad_norm": 0.640988290309906,
|
| 1898 |
+
"learning_rate": 1.3219782914527633e-07,
|
| 1899 |
+
"loss": 0.2562,
|
| 1900 |
+
"step": 270
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.817946505608283,
|
| 1904 |
+
"grad_norm": 0.5685985088348389,
|
| 1905 |
+
"learning_rate": 1.1870234949636072e-07,
|
| 1906 |
+
"loss": 0.2861,
|
| 1907 |
+
"step": 271
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.828300258843831,
|
| 1911 |
+
"grad_norm": 0.5495738983154297,
|
| 1912 |
+
"learning_rate": 1.0592504468227127e-07,
|
| 1913 |
+
"loss": 0.2618,
|
| 1914 |
+
"step": 272
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.838654012079379,
|
| 1918 |
+
"grad_norm": 0.5660983920097351,
|
| 1919 |
+
"learning_rate": 9.386779460041018e-08,
|
| 1920 |
+
"loss": 0.2793,
|
| 1921 |
+
"step": 273
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.8490077653149264,
|
| 1925 |
+
"grad_norm": 0.5590121746063232,
|
| 1926 |
+
"learning_rate": 8.253237320807461e-08,
|
| 1927 |
+
"loss": 0.2808,
|
| 1928 |
+
"step": 274
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.8593615185504744,
|
| 1932 |
+
"grad_norm": 0.5588066577911377,
|
| 1933 |
+
"learning_rate": 7.192044826145772e-08,
|
| 1934 |
+
"loss": 0.2673,
|
| 1935 |
+
"step": 275
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.8697152717860224,
|
| 1939 |
+
"grad_norm": 0.5707908868789673,
|
| 1940 |
+
"learning_rate": 6.203358107027491e-08,
|
| 1941 |
+
"loss": 0.2675,
|
| 1942 |
+
"step": 276
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.8800690250215704,
|
| 1946 |
+
"grad_norm": 0.5975697636604309,
|
| 1947 |
+
"learning_rate": 5.2873226268052026e-08,
|
| 1948 |
+
"loss": 0.2798,
|
| 1949 |
+
"step": 277
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.8904227782571184,
|
| 1953 |
+
"grad_norm": 0.5799188017845154,
|
| 1954 |
+
"learning_rate": 4.444073159810769e-08,
|
| 1955 |
+
"loss": 0.2631,
|
| 1956 |
+
"step": 278
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.9007765314926663,
|
| 1960 |
+
"grad_norm": 0.5713800191879272,
|
| 1961 |
+
"learning_rate": 3.673733771526466e-08,
|
| 1962 |
+
"loss": 0.2769,
|
| 1963 |
+
"step": 279
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.911130284728214,
|
| 1967 |
+
"grad_norm": 0.5764074325561523,
|
| 1968 |
+
"learning_rate": 2.976417800331144e-08,
|
| 1969 |
+
"loss": 0.2695,
|
| 1970 |
+
"step": 280
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.921484037963762,
|
| 1974 |
+
"grad_norm": 0.6320033669471741,
|
| 1975 |
+
"learning_rate": 2.352227840825394e-08,
|
| 1976 |
+
"loss": 0.2908,
|
| 1977 |
+
"step": 281
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.93183779119931,
|
| 1981 |
+
"grad_norm": 0.5747038125991821,
|
| 1982 |
+
"learning_rate": 1.8012557287367394e-08,
|
| 1983 |
+
"loss": 0.2741,
|
| 1984 |
+
"step": 282
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.9421915444348574,
|
| 1988 |
+
"grad_norm": 0.5464159250259399,
|
| 1989 |
+
"learning_rate": 1.3235825274081626e-08,
|
| 1990 |
+
"loss": 0.2374,
|
| 1991 |
+
"step": 283
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.9525452976704054,
|
| 1995 |
+
"grad_norm": 0.6008951663970947,
|
| 1996 |
+
"learning_rate": 9.192785158713691e-09,
|
| 1997 |
+
"loss": 0.2781,
|
| 1998 |
+
"step": 284
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.9628990509059534,
|
| 2002 |
+
"grad_norm": 0.6063032150268555,
|
| 2003 |
+
"learning_rate": 5.884031785068356e-09,
|
| 2004 |
+
"loss": 0.2618,
|
| 2005 |
+
"step": 285
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.9732528041415014,
|
| 2009 |
+
"grad_norm": 0.5779683589935303,
|
| 2010 |
+
"learning_rate": 3.3100519629203353e-09,
|
| 2011 |
+
"loss": 0.2675,
|
| 2012 |
+
"step": 286
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.9836065573770494,
|
| 2016 |
+
"grad_norm": 0.5606078505516052,
|
| 2017 |
+
"learning_rate": 1.471224396389359e-09,
|
| 2018 |
+
"loss": 0.2651,
|
| 2019 |
+
"step": 287
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.993960310612597,
|
| 2023 |
+
"grad_norm": 0.521804928779602,
|
| 2024 |
+
"learning_rate": 3.678196282252966e-10,
|
| 2025 |
+
"loss": 0.2605,
|
| 2026 |
+
"step": 288
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.993960310612597,
|
| 2030 |
+
"step": 288,
|
| 2031 |
+
"total_flos": 1.345368819194921e+17,
|
| 2032 |
+
"train_loss": 0.3956852917973366,
|
| 2033 |
+
"train_runtime": 35330.3168,
|
| 2034 |
+
"train_samples_per_second": 0.393,
|
| 2035 |
+
"train_steps_per_second": 0.008
|
| 2036 |
+
}
|
| 2037 |
+
],
|
| 2038 |
+
"logging_steps": 1,
|
| 2039 |
+
"max_steps": 288,
|
| 2040 |
+
"num_input_tokens_seen": 0,
|
| 2041 |
+
"num_train_epochs": 3,
|
| 2042 |
+
"save_steps": 200,
|
| 2043 |
+
"stateful_callbacks": {
|
| 2044 |
+
"TrainerControl": {
|
| 2045 |
+
"args": {
|
| 2046 |
+
"should_epoch_stop": false,
|
| 2047 |
+
"should_evaluate": false,
|
| 2048 |
+
"should_log": false,
|
| 2049 |
+
"should_save": true,
|
| 2050 |
+
"should_training_stop": true
|
| 2051 |
+
},
|
| 2052 |
+
"attributes": {}
|
| 2053 |
+
}
|
| 2054 |
+
},
|
| 2055 |
+
"total_flos": 1.345368819194921e+17,
|
| 2056 |
+
"train_batch_size": 1,
|
| 2057 |
+
"trial_name": null,
|
| 2058 |
+
"trial_params": null
|
| 2059 |
+
}
|
training_loss.png
ADDED
|