End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +3025 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-Coder-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source_qwen_code_7B
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source_qwen_code_7B
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-Coder-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source_qwen_code_7B
|
|
|
|
| 16 |
|
| 17 |
# SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source_qwen_code_7B
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct) on the SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9836065573770494,
|
| 3 |
+
"total_flos": 7.440485693207347e+16,
|
| 4 |
+
"train_loss": 0.40016700194475235,
|
| 5 |
+
"train_runtime": 24963.7361,
|
| 6 |
+
"train_samples_per_second": 0.411,
|
| 7 |
+
"train_steps_per_second": 0.017
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9836065573770494,
|
| 3 |
+
"total_flos": 7.440485693207347e+16,
|
| 4 |
+
"train_loss": 0.40016700194475235,
|
| 5 |
+
"train_runtime": 24963.7361,
|
| 6 |
+
"train_samples_per_second": 0.411,
|
| 7 |
+
"train_steps_per_second": 0.017
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3025 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.9836065573770494,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 426,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.00702576112412178,
|
| 14 |
+
"grad_norm": 2.574352979660034,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.8248,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.01405152224824356,
|
| 21 |
+
"grad_norm": 2.1349353790283203,
|
| 22 |
+
"learning_rate": 2.3255813953488374e-07,
|
| 23 |
+
"loss": 0.7537,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.02107728337236534,
|
| 28 |
+
"grad_norm": 2.3432695865631104,
|
| 29 |
+
"learning_rate": 4.651162790697675e-07,
|
| 30 |
+
"loss": 0.7967,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.02810304449648712,
|
| 35 |
+
"grad_norm": 2.618528127670288,
|
| 36 |
+
"learning_rate": 6.976744186046513e-07,
|
| 37 |
+
"loss": 0.8444,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0351288056206089,
|
| 42 |
+
"grad_norm": 1.9025769233703613,
|
| 43 |
+
"learning_rate": 9.30232558139535e-07,
|
| 44 |
+
"loss": 0.7037,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.04215456674473068,
|
| 49 |
+
"grad_norm": 2.3843185901641846,
|
| 50 |
+
"learning_rate": 1.1627906976744188e-06,
|
| 51 |
+
"loss": 0.8214,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.04918032786885246,
|
| 56 |
+
"grad_norm": 2.20890736579895,
|
| 57 |
+
"learning_rate": 1.3953488372093025e-06,
|
| 58 |
+
"loss": 0.8083,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.05620608899297424,
|
| 63 |
+
"grad_norm": 1.8575512170791626,
|
| 64 |
+
"learning_rate": 1.6279069767441862e-06,
|
| 65 |
+
"loss": 0.7147,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.06323185011709602,
|
| 70 |
+
"grad_norm": 2.092245101928711,
|
| 71 |
+
"learning_rate": 1.86046511627907e-06,
|
| 72 |
+
"loss": 0.7483,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.0702576112412178,
|
| 77 |
+
"grad_norm": 1.4913474321365356,
|
| 78 |
+
"learning_rate": 2.0930232558139536e-06,
|
| 79 |
+
"loss": 0.7233,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.07728337236533958,
|
| 84 |
+
"grad_norm": 1.4805643558502197,
|
| 85 |
+
"learning_rate": 2.3255813953488376e-06,
|
| 86 |
+
"loss": 0.7878,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.08430913348946135,
|
| 91 |
+
"grad_norm": 1.3100932836532593,
|
| 92 |
+
"learning_rate": 2.558139534883721e-06,
|
| 93 |
+
"loss": 0.76,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.09133489461358314,
|
| 98 |
+
"grad_norm": 1.3169656991958618,
|
| 99 |
+
"learning_rate": 2.790697674418605e-06,
|
| 100 |
+
"loss": 0.731,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.09836065573770492,
|
| 105 |
+
"grad_norm": 1.1664108037948608,
|
| 106 |
+
"learning_rate": 3.0232558139534885e-06,
|
| 107 |
+
"loss": 0.7416,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.1053864168618267,
|
| 112 |
+
"grad_norm": 1.3624343872070312,
|
| 113 |
+
"learning_rate": 3.2558139534883724e-06,
|
| 114 |
+
"loss": 0.6906,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.11241217798594848,
|
| 119 |
+
"grad_norm": 1.2343862056732178,
|
| 120 |
+
"learning_rate": 3.4883720930232564e-06,
|
| 121 |
+
"loss": 0.6636,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.11943793911007025,
|
| 126 |
+
"grad_norm": 1.2548797130584717,
|
| 127 |
+
"learning_rate": 3.72093023255814e-06,
|
| 128 |
+
"loss": 0.6339,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.12646370023419204,
|
| 133 |
+
"grad_norm": 1.347133994102478,
|
| 134 |
+
"learning_rate": 3.953488372093024e-06,
|
| 135 |
+
"loss": 0.7124,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.13348946135831383,
|
| 140 |
+
"grad_norm": 1.074514627456665,
|
| 141 |
+
"learning_rate": 4.186046511627907e-06,
|
| 142 |
+
"loss": 0.6496,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1405152224824356,
|
| 147 |
+
"grad_norm": 1.3441463708877563,
|
| 148 |
+
"learning_rate": 4.418604651162791e-06,
|
| 149 |
+
"loss": 0.7656,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.14754098360655737,
|
| 154 |
+
"grad_norm": 1.066001057624817,
|
| 155 |
+
"learning_rate": 4.651162790697675e-06,
|
| 156 |
+
"loss": 0.6252,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.15456674473067916,
|
| 161 |
+
"grad_norm": 1.124649167060852,
|
| 162 |
+
"learning_rate": 4.883720930232559e-06,
|
| 163 |
+
"loss": 0.646,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.16159250585480095,
|
| 168 |
+
"grad_norm": 1.2024065256118774,
|
| 169 |
+
"learning_rate": 5.116279069767442e-06,
|
| 170 |
+
"loss": 0.7369,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1686182669789227,
|
| 175 |
+
"grad_norm": 0.9406749606132507,
|
| 176 |
+
"learning_rate": 5.348837209302326e-06,
|
| 177 |
+
"loss": 0.6095,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.1756440281030445,
|
| 182 |
+
"grad_norm": 0.9862017035484314,
|
| 183 |
+
"learning_rate": 5.58139534883721e-06,
|
| 184 |
+
"loss": 0.6875,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.18266978922716628,
|
| 189 |
+
"grad_norm": 0.9771542549133301,
|
| 190 |
+
"learning_rate": 5.8139534883720935e-06,
|
| 191 |
+
"loss": 0.614,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.18969555035128804,
|
| 196 |
+
"grad_norm": 0.9741706252098083,
|
| 197 |
+
"learning_rate": 6.046511627906977e-06,
|
| 198 |
+
"loss": 0.6459,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.19672131147540983,
|
| 203 |
+
"grad_norm": 0.9432346224784851,
|
| 204 |
+
"learning_rate": 6.279069767441861e-06,
|
| 205 |
+
"loss": 0.5888,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.20374707259953162,
|
| 210 |
+
"grad_norm": 1.0036011934280396,
|
| 211 |
+
"learning_rate": 6.511627906976745e-06,
|
| 212 |
+
"loss": 0.6198,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.2107728337236534,
|
| 217 |
+
"grad_norm": 1.0411369800567627,
|
| 218 |
+
"learning_rate": 6.744186046511628e-06,
|
| 219 |
+
"loss": 0.6236,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.21779859484777517,
|
| 224 |
+
"grad_norm": 0.9507374167442322,
|
| 225 |
+
"learning_rate": 6.976744186046513e-06,
|
| 226 |
+
"loss": 0.6386,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.22482435597189696,
|
| 231 |
+
"grad_norm": 0.9232994914054871,
|
| 232 |
+
"learning_rate": 7.209302325581395e-06,
|
| 233 |
+
"loss": 0.5525,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.23185011709601874,
|
| 238 |
+
"grad_norm": 0.9739473462104797,
|
| 239 |
+
"learning_rate": 7.44186046511628e-06,
|
| 240 |
+
"loss": 0.616,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.2388758782201405,
|
| 245 |
+
"grad_norm": 1.0418951511383057,
|
| 246 |
+
"learning_rate": 7.674418604651164e-06,
|
| 247 |
+
"loss": 0.6507,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.2459016393442623,
|
| 252 |
+
"grad_norm": 0.9890851974487305,
|
| 253 |
+
"learning_rate": 7.906976744186048e-06,
|
| 254 |
+
"loss": 0.6007,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.2529274004683841,
|
| 259 |
+
"grad_norm": 1.0349009037017822,
|
| 260 |
+
"learning_rate": 8.139534883720931e-06,
|
| 261 |
+
"loss": 0.5901,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.25995316159250587,
|
| 266 |
+
"grad_norm": 0.9471357464790344,
|
| 267 |
+
"learning_rate": 8.372093023255815e-06,
|
| 268 |
+
"loss": 0.6052,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.26697892271662765,
|
| 273 |
+
"grad_norm": 0.9224464297294617,
|
| 274 |
+
"learning_rate": 8.604651162790698e-06,
|
| 275 |
+
"loss": 0.5328,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.27400468384074944,
|
| 280 |
+
"grad_norm": 1.0249232053756714,
|
| 281 |
+
"learning_rate": 8.837209302325582e-06,
|
| 282 |
+
"loss": 0.5395,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.2810304449648712,
|
| 287 |
+
"grad_norm": 0.8731057643890381,
|
| 288 |
+
"learning_rate": 9.069767441860465e-06,
|
| 289 |
+
"loss": 0.5548,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.28805620608899296,
|
| 294 |
+
"grad_norm": 0.9216372966766357,
|
| 295 |
+
"learning_rate": 9.30232558139535e-06,
|
| 296 |
+
"loss": 0.5724,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.29508196721311475,
|
| 301 |
+
"grad_norm": 1.006893277168274,
|
| 302 |
+
"learning_rate": 9.534883720930234e-06,
|
| 303 |
+
"loss": 0.5813,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.30210772833723654,
|
| 308 |
+
"grad_norm": 0.9753144383430481,
|
| 309 |
+
"learning_rate": 9.767441860465117e-06,
|
| 310 |
+
"loss": 0.5982,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3091334894613583,
|
| 315 |
+
"grad_norm": 0.8903833627700806,
|
| 316 |
+
"learning_rate": 1e-05,
|
| 317 |
+
"loss": 0.5235,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.3161592505854801,
|
| 322 |
+
"grad_norm": 0.8771428465843201,
|
| 323 |
+
"learning_rate": 9.99983179466314e-06,
|
| 324 |
+
"loss": 0.5212,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3231850117096019,
|
| 329 |
+
"grad_norm": 0.8775717616081238,
|
| 330 |
+
"learning_rate": 9.999327189969768e-06,
|
| 331 |
+
"loss": 0.5614,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.33021077283372363,
|
| 336 |
+
"grad_norm": 1.019813060760498,
|
| 337 |
+
"learning_rate": 9.998486219870769e-06,
|
| 338 |
+
"loss": 0.6111,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3372365339578454,
|
| 343 |
+
"grad_norm": 0.8810835480690002,
|
| 344 |
+
"learning_rate": 9.997308940948405e-06,
|
| 345 |
+
"loss": 0.5876,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.3442622950819672,
|
| 350 |
+
"grad_norm": 0.9496193528175354,
|
| 351 |
+
"learning_rate": 9.995795432412513e-06,
|
| 352 |
+
"loss": 0.5635,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.351288056206089,
|
| 357 |
+
"grad_norm": 0.8718122839927673,
|
| 358 |
+
"learning_rate": 9.993945796095183e-06,
|
| 359 |
+
"loss": 0.516,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.3583138173302108,
|
| 364 |
+
"grad_norm": 0.9371446371078491,
|
| 365 |
+
"learning_rate": 9.991760156443892e-06,
|
| 366 |
+
"loss": 0.6481,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.36533957845433257,
|
| 371 |
+
"grad_norm": 0.9381894469261169,
|
| 372 |
+
"learning_rate": 9.989238660513141e-06,
|
| 373 |
+
"loss": 0.5661,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.37236533957845436,
|
| 378 |
+
"grad_norm": 0.8230867385864258,
|
| 379 |
+
"learning_rate": 9.98638147795456e-06,
|
| 380 |
+
"loss": 0.5359,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.3793911007025761,
|
| 385 |
+
"grad_norm": 0.8313586115837097,
|
| 386 |
+
"learning_rate": 9.983188801005492e-06,
|
| 387 |
+
"loss": 0.5787,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3864168618266979,
|
| 392 |
+
"grad_norm": 0.9128983616828918,
|
| 393 |
+
"learning_rate": 9.979660844476056e-06,
|
| 394 |
+
"loss": 0.5941,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.39344262295081966,
|
| 399 |
+
"grad_norm": 0.9168326258659363,
|
| 400 |
+
"learning_rate": 9.975797845734699e-06,
|
| 401 |
+
"loss": 0.5344,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.40046838407494145,
|
| 406 |
+
"grad_norm": 0.8436552286148071,
|
| 407 |
+
"learning_rate": 9.971600064692222e-06,
|
| 408 |
+
"loss": 0.5526,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.40749414519906324,
|
| 413 |
+
"grad_norm": 0.910169780254364,
|
| 414 |
+
"learning_rate": 9.967067783784297e-06,
|
| 415 |
+
"loss": 0.5758,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.41451990632318503,
|
| 420 |
+
"grad_norm": 0.9464594721794128,
|
| 421 |
+
"learning_rate": 9.962201307952455e-06,
|
| 422 |
+
"loss": 0.5689,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.4215456674473068,
|
| 427 |
+
"grad_norm": 0.8791322112083435,
|
| 428 |
+
"learning_rate": 9.957000964623585e-06,
|
| 429 |
+
"loss": 0.5275,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.42857142857142855,
|
| 434 |
+
"grad_norm": 0.7610299587249756,
|
| 435 |
+
"learning_rate": 9.951467103687879e-06,
|
| 436 |
+
"loss": 0.5131,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.43559718969555034,
|
| 441 |
+
"grad_norm": 0.8980028629302979,
|
| 442 |
+
"learning_rate": 9.945600097475322e-06,
|
| 443 |
+
"loss": 0.5849,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.4426229508196721,
|
| 448 |
+
"grad_norm": 0.8086323738098145,
|
| 449 |
+
"learning_rate": 9.939400340730611e-06,
|
| 450 |
+
"loss": 0.5466,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.4496487119437939,
|
| 455 |
+
"grad_norm": 0.7332218289375305,
|
| 456 |
+
"learning_rate": 9.932868250586619e-06,
|
| 457 |
+
"loss": 0.4886,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.4566744730679157,
|
| 462 |
+
"grad_norm": 0.804356575012207,
|
| 463 |
+
"learning_rate": 9.926004266536314e-06,
|
| 464 |
+
"loss": 0.5459,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.4637002341920375,
|
| 469 |
+
"grad_norm": 0.8725134134292603,
|
| 470 |
+
"learning_rate": 9.918808850403192e-06,
|
| 471 |
+
"loss": 0.5319,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.4707259953161593,
|
| 476 |
+
"grad_norm": 1.0051013231277466,
|
| 477 |
+
"learning_rate": 9.911282486310214e-06,
|
| 478 |
+
"loss": 0.6205,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.477751756440281,
|
| 483 |
+
"grad_norm": 0.8478599786758423,
|
| 484 |
+
"learning_rate": 9.903425680647225e-06,
|
| 485 |
+
"loss": 0.4813,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.4847775175644028,
|
| 490 |
+
"grad_norm": 1.0624728202819824,
|
| 491 |
+
"learning_rate": 9.895238962036878e-06,
|
| 492 |
+
"loss": 0.5635,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.4918032786885246,
|
| 497 |
+
"grad_norm": 0.9066529273986816,
|
| 498 |
+
"learning_rate": 9.88672288129908e-06,
|
| 499 |
+
"loss": 0.5122,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.49882903981264637,
|
| 504 |
+
"grad_norm": 0.9527250528335571,
|
| 505 |
+
"learning_rate": 9.877878011413924e-06,
|
| 506 |
+
"loss": 0.5768,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5058548009367682,
|
| 511 |
+
"grad_norm": 0.8591968417167664,
|
| 512 |
+
"learning_rate": 9.868704947483134e-06,
|
| 513 |
+
"loss": 0.5522,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.5128805620608899,
|
| 518 |
+
"grad_norm": 0.8399043083190918,
|
| 519 |
+
"learning_rate": 9.859204306690038e-06,
|
| 520 |
+
"loss": 0.5652,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.5199063231850117,
|
| 525 |
+
"grad_norm": 0.8005277514457703,
|
| 526 |
+
"learning_rate": 9.849376728258024e-06,
|
| 527 |
+
"loss": 0.5292,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.5269320843091335,
|
| 532 |
+
"grad_norm": 0.8883223533630371,
|
| 533 |
+
"learning_rate": 9.839222873407553e-06,
|
| 534 |
+
"loss": 0.5303,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.5339578454332553,
|
| 539 |
+
"grad_norm": 0.8855736255645752,
|
| 540 |
+
"learning_rate": 9.828743425311654e-06,
|
| 541 |
+
"loss": 0.5033,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.5409836065573771,
|
| 546 |
+
"grad_norm": 0.7898836135864258,
|
| 547 |
+
"learning_rate": 9.817939089049964e-06,
|
| 548 |
+
"loss": 0.4858,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.5480093676814989,
|
| 553 |
+
"grad_norm": 0.8595909476280212,
|
| 554 |
+
"learning_rate": 9.806810591561295e-06,
|
| 555 |
+
"loss": 0.5326,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.5550351288056206,
|
| 560 |
+
"grad_norm": 1.1396889686584473,
|
| 561 |
+
"learning_rate": 9.795358681594712e-06,
|
| 562 |
+
"loss": 0.5686,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.5620608899297423,
|
| 567 |
+
"grad_norm": 1.002563238143921,
|
| 568 |
+
"learning_rate": 9.783584129659162e-06,
|
| 569 |
+
"loss": 0.5446,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.5690866510538641,
|
| 574 |
+
"grad_norm": 0.8612224459648132,
|
| 575 |
+
"learning_rate": 9.771487727971642e-06,
|
| 576 |
+
"loss": 0.5278,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.5761124121779859,
|
| 581 |
+
"grad_norm": 0.7524361610412598,
|
| 582 |
+
"learning_rate": 9.759070290403873e-06,
|
| 583 |
+
"loss": 0.5276,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.5831381733021077,
|
| 588 |
+
"grad_norm": 0.9360653162002563,
|
| 589 |
+
"learning_rate": 9.746332652427566e-06,
|
| 590 |
+
"loss": 0.5577,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.5901639344262295,
|
| 595 |
+
"grad_norm": 0.9080122113227844,
|
| 596 |
+
"learning_rate": 9.733275671058195e-06,
|
| 597 |
+
"loss": 0.5362,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.5971896955503513,
|
| 602 |
+
"grad_norm": 0.8351691961288452,
|
| 603 |
+
"learning_rate": 9.71990022479734e-06,
|
| 604 |
+
"loss": 0.5188,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.6042154566744731,
|
| 609 |
+
"grad_norm": 0.821236252784729,
|
| 610 |
+
"learning_rate": 9.70620721357358e-06,
|
| 611 |
+
"loss": 0.5184,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.6112412177985949,
|
| 616 |
+
"grad_norm": 0.9071369171142578,
|
| 617 |
+
"learning_rate": 9.69219755868194e-06,
|
| 618 |
+
"loss": 0.5525,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.6182669789227166,
|
| 623 |
+
"grad_norm": 0.9066632986068726,
|
| 624 |
+
"learning_rate": 9.677872202721906e-06,
|
| 625 |
+
"loss": 0.6103,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.6252927400468384,
|
| 630 |
+
"grad_norm": 0.8524773120880127,
|
| 631 |
+
"learning_rate": 9.663232109534011e-06,
|
| 632 |
+
"loss": 0.4937,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.6323185011709602,
|
| 637 |
+
"grad_norm": 0.8702455759048462,
|
| 638 |
+
"learning_rate": 9.648278264134977e-06,
|
| 639 |
+
"loss": 0.4613,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.639344262295082,
|
| 644 |
+
"grad_norm": 0.8476565480232239,
|
| 645 |
+
"learning_rate": 9.633011672651443e-06,
|
| 646 |
+
"loss": 0.5421,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.6463700234192038,
|
| 651 |
+
"grad_norm": 0.7731307744979858,
|
| 652 |
+
"learning_rate": 9.617433362252277e-06,
|
| 653 |
+
"loss": 0.5133,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.6533957845433255,
|
| 658 |
+
"grad_norm": 0.8306621313095093,
|
| 659 |
+
"learning_rate": 9.601544381079457e-06,
|
| 660 |
+
"loss": 0.5665,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.6604215456674473,
|
| 665 |
+
"grad_norm": 0.8650710582733154,
|
| 666 |
+
"learning_rate": 9.585345798177557e-06,
|
| 667 |
+
"loss": 0.4941,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.667447306791569,
|
| 672 |
+
"grad_norm": 0.8899430632591248,
|
| 673 |
+
"learning_rate": 9.56883870342181e-06,
|
| 674 |
+
"loss": 0.4759,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.6744730679156908,
|
| 679 |
+
"grad_norm": 0.8613656759262085,
|
| 680 |
+
"learning_rate": 9.552024207444794e-06,
|
| 681 |
+
"loss": 0.5644,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.6814988290398126,
|
| 686 |
+
"grad_norm": 0.9278618097305298,
|
| 687 |
+
"learning_rate": 9.534903441561693e-06,
|
| 688 |
+
"loss": 0.5615,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.6885245901639344,
|
| 693 |
+
"grad_norm": 0.9220278263092041,
|
| 694 |
+
"learning_rate": 9.517477557694182e-06,
|
| 695 |
+
"loss": 0.5735,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.6955503512880562,
|
| 700 |
+
"grad_norm": 0.8179577589035034,
|
| 701 |
+
"learning_rate": 9.499747728292928e-06,
|
| 702 |
+
"loss": 0.4838,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.702576112412178,
|
| 707 |
+
"grad_norm": 0.903998851776123,
|
| 708 |
+
"learning_rate": 9.481715146258699e-06,
|
| 709 |
+
"loss": 0.583,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.7096018735362998,
|
| 714 |
+
"grad_norm": 0.9776198863983154,
|
| 715 |
+
"learning_rate": 9.463381024862116e-06,
|
| 716 |
+
"loss": 0.5629,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.7166276346604216,
|
| 721 |
+
"grad_norm": 0.830189049243927,
|
| 722 |
+
"learning_rate": 9.444746597662e-06,
|
| 723 |
+
"loss": 0.5186,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.7236533957845434,
|
| 728 |
+
"grad_norm": 0.894983172416687,
|
| 729 |
+
"learning_rate": 9.425813118422393e-06,
|
| 730 |
+
"loss": 0.5929,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.7306791569086651,
|
| 735 |
+
"grad_norm": 0.9019003510475159,
|
| 736 |
+
"learning_rate": 9.406581861028199e-06,
|
| 737 |
+
"loss": 0.4721,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.7377049180327869,
|
| 742 |
+
"grad_norm": 0.8638082146644592,
|
| 743 |
+
"learning_rate": 9.387054119399466e-06,
|
| 744 |
+
"loss": 0.5086,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.7447306791569087,
|
| 749 |
+
"grad_norm": 0.961375892162323,
|
| 750 |
+
"learning_rate": 9.36723120740434e-06,
|
| 751 |
+
"loss": 0.5382,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.7517564402810304,
|
| 756 |
+
"grad_norm": 0.8464214205741882,
|
| 757 |
+
"learning_rate": 9.347114458770656e-06,
|
| 758 |
+
"loss": 0.514,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.7587822014051522,
|
| 763 |
+
"grad_norm": 0.8622469305992126,
|
| 764 |
+
"learning_rate": 9.326705226996207e-06,
|
| 765 |
+
"loss": 0.5179,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.765807962529274,
|
| 770 |
+
"grad_norm": 0.980879545211792,
|
| 771 |
+
"learning_rate": 9.306004885257675e-06,
|
| 772 |
+
"loss": 0.5395,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.7728337236533958,
|
| 777 |
+
"grad_norm": 0.901023268699646,
|
| 778 |
+
"learning_rate": 9.28501482631824e-06,
|
| 779 |
+
"loss": 0.5406,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.7798594847775175,
|
| 784 |
+
"grad_norm": 0.8403215408325195,
|
| 785 |
+
"learning_rate": 9.26373646243388e-06,
|
| 786 |
+
"loss": 0.5313,
|
| 787 |
+
"step": 111
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.7868852459016393,
|
| 791 |
+
"grad_norm": 0.868764340877533,
|
| 792 |
+
"learning_rate": 9.242171225258336e-06,
|
| 793 |
+
"loss": 0.498,
|
| 794 |
+
"step": 112
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.7939110070257611,
|
| 798 |
+
"grad_norm": 0.7906497716903687,
|
| 799 |
+
"learning_rate": 9.220320565746806e-06,
|
| 800 |
+
"loss": 0.5263,
|
| 801 |
+
"step": 113
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.8009367681498829,
|
| 805 |
+
"grad_norm": 0.8329947590827942,
|
| 806 |
+
"learning_rate": 9.198185954058305e-06,
|
| 807 |
+
"loss": 0.4941,
|
| 808 |
+
"step": 114
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.8079625292740047,
|
| 812 |
+
"grad_norm": 0.8946017622947693,
|
| 813 |
+
"learning_rate": 9.175768879456759e-06,
|
| 814 |
+
"loss": 0.4953,
|
| 815 |
+
"step": 115
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.8149882903981265,
|
| 819 |
+
"grad_norm": 0.7878472208976746,
|
| 820 |
+
"learning_rate": 9.153070850210803e-06,
|
| 821 |
+
"loss": 0.4857,
|
| 822 |
+
"step": 116
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.8220140515222483,
|
| 826 |
+
"grad_norm": 0.8317592740058899,
|
| 827 |
+
"learning_rate": 9.130093393492302e-06,
|
| 828 |
+
"loss": 0.508,
|
| 829 |
+
"step": 117
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.8290398126463701,
|
| 833 |
+
"grad_norm": 0.9189662337303162,
|
| 834 |
+
"learning_rate": 9.106838055273589e-06,
|
| 835 |
+
"loss": 0.5354,
|
| 836 |
+
"step": 118
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.8360655737704918,
|
| 840 |
+
"grad_norm": 0.8089627623558044,
|
| 841 |
+
"learning_rate": 9.083306400223465e-06,
|
| 842 |
+
"loss": 0.5021,
|
| 843 |
+
"step": 119
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.8430913348946136,
|
| 847 |
+
"grad_norm": 0.7785940170288086,
|
| 848 |
+
"learning_rate": 9.059500011601919e-06,
|
| 849 |
+
"loss": 0.4905,
|
| 850 |
+
"step": 120
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.8501170960187353,
|
| 854 |
+
"grad_norm": 0.816950261592865,
|
| 855 |
+
"learning_rate": 9.035420491153596e-06,
|
| 856 |
+
"loss": 0.5299,
|
| 857 |
+
"step": 121
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.8571428571428571,
|
| 861 |
+
"grad_norm": 0.794824481010437,
|
| 862 |
+
"learning_rate": 9.011069459000035e-06,
|
| 863 |
+
"loss": 0.4801,
|
| 864 |
+
"step": 122
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.8641686182669789,
|
| 868 |
+
"grad_norm": 0.8010877370834351,
|
| 869 |
+
"learning_rate": 8.986448553530665e-06,
|
| 870 |
+
"loss": 0.492,
|
| 871 |
+
"step": 123
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.8711943793911007,
|
| 875 |
+
"grad_norm": 0.8579709529876709,
|
| 876 |
+
"learning_rate": 8.961559431292562e-06,
|
| 877 |
+
"loss": 0.4774,
|
| 878 |
+
"step": 124
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.8782201405152225,
|
| 882 |
+
"grad_norm": 0.7804529666900635,
|
| 883 |
+
"learning_rate": 8.936403766879003e-06,
|
| 884 |
+
"loss": 0.4615,
|
| 885 |
+
"step": 125
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.8852459016393442,
|
| 889 |
+
"grad_norm": 0.8179356455802917,
|
| 890 |
+
"learning_rate": 8.910983252816794e-06,
|
| 891 |
+
"loss": 0.4974,
|
| 892 |
+
"step": 126
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.892271662763466,
|
| 896 |
+
"grad_norm": 0.9173282384872437,
|
| 897 |
+
"learning_rate": 8.885299599452381e-06,
|
| 898 |
+
"loss": 0.496,
|
| 899 |
+
"step": 127
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.8992974238875878,
|
| 903 |
+
"grad_norm": 0.785065233707428,
|
| 904 |
+
"learning_rate": 8.859354534836797e-06,
|
| 905 |
+
"loss": 0.4637,
|
| 906 |
+
"step": 128
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.9063231850117096,
|
| 910 |
+
"grad_norm": 0.8229131102561951,
|
| 911 |
+
"learning_rate": 8.833149804609372e-06,
|
| 912 |
+
"loss": 0.5046,
|
| 913 |
+
"step": 129
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.9133489461358314,
|
| 917 |
+
"grad_norm": 0.7581368684768677,
|
| 918 |
+
"learning_rate": 8.806687171880298e-06,
|
| 919 |
+
"loss": 0.49,
|
| 920 |
+
"step": 130
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.9203747072599532,
|
| 924 |
+
"grad_norm": 0.8826923966407776,
|
| 925 |
+
"learning_rate": 8.779968417111991e-06,
|
| 926 |
+
"loss": 0.5804,
|
| 927 |
+
"step": 131
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.927400468384075,
|
| 931 |
+
"grad_norm": 0.8899806141853333,
|
| 932 |
+
"learning_rate": 8.752995337999316e-06,
|
| 933 |
+
"loss": 0.5149,
|
| 934 |
+
"step": 132
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.9344262295081968,
|
| 938 |
+
"grad_norm": 0.8376835584640503,
|
| 939 |
+
"learning_rate": 8.725769749348612e-06,
|
| 940 |
+
"loss": 0.5149,
|
| 941 |
+
"step": 133
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.9414519906323185,
|
| 945 |
+
"grad_norm": 0.890254557132721,
|
| 946 |
+
"learning_rate": 8.698293482955605e-06,
|
| 947 |
+
"loss": 0.5291,
|
| 948 |
+
"step": 134
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.9484777517564403,
|
| 952 |
+
"grad_norm": 0.925977349281311,
|
| 953 |
+
"learning_rate": 8.670568387482153e-06,
|
| 954 |
+
"loss": 0.5072,
|
| 955 |
+
"step": 135
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.955503512880562,
|
| 959 |
+
"grad_norm": 0.8560639023780823,
|
| 960 |
+
"learning_rate": 8.642596328331864e-06,
|
| 961 |
+
"loss": 0.4863,
|
| 962 |
+
"step": 136
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.9625292740046838,
|
| 966 |
+
"grad_norm": 0.8338841199874878,
|
| 967 |
+
"learning_rate": 8.614379187524593e-06,
|
| 968 |
+
"loss": 0.4843,
|
| 969 |
+
"step": 137
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.9695550351288056,
|
| 973 |
+
"grad_norm": 0.7932563424110413,
|
| 974 |
+
"learning_rate": 8.585918863569806e-06,
|
| 975 |
+
"loss": 0.5084,
|
| 976 |
+
"step": 138
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.9765807962529274,
|
| 980 |
+
"grad_norm": 0.8398274183273315,
|
| 981 |
+
"learning_rate": 8.55721727133886e-06,
|
| 982 |
+
"loss": 0.5292,
|
| 983 |
+
"step": 139
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.9836065573770492,
|
| 987 |
+
"grad_norm": 0.8969703912734985,
|
| 988 |
+
"learning_rate": 8.528276341936146e-06,
|
| 989 |
+
"loss": 0.499,
|
| 990 |
+
"step": 140
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.990632318501171,
|
| 994 |
+
"grad_norm": 0.9116531014442444,
|
| 995 |
+
"learning_rate": 8.499098022569177e-06,
|
| 996 |
+
"loss": 0.5519,
|
| 997 |
+
"step": 141
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.9976580796252927,
|
| 1001 |
+
"grad_norm": 0.79449063539505,
|
| 1002 |
+
"learning_rate": 8.469684276417568e-06,
|
| 1003 |
+
"loss": 0.493,
|
| 1004 |
+
"step": 142
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.0,
|
| 1008 |
+
"grad_norm": 0.79449063539505,
|
| 1009 |
+
"learning_rate": 8.440037082500953e-06,
|
| 1010 |
+
"loss": 0.5449,
|
| 1011 |
+
"step": 143
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.0070257611241218,
|
| 1015 |
+
"grad_norm": 1.5884685516357422,
|
| 1016 |
+
"learning_rate": 8.410158435545825e-06,
|
| 1017 |
+
"loss": 0.3758,
|
| 1018 |
+
"step": 144
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.0140515222482436,
|
| 1022 |
+
"grad_norm": 0.7242723703384399,
|
| 1023 |
+
"learning_rate": 8.380050345851338e-06,
|
| 1024 |
+
"loss": 0.3874,
|
| 1025 |
+
"step": 145
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.0210772833723654,
|
| 1029 |
+
"grad_norm": 0.8061760663986206,
|
| 1030 |
+
"learning_rate": 8.349714839154035e-06,
|
| 1031 |
+
"loss": 0.3925,
|
| 1032 |
+
"step": 146
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.0281030444964872,
|
| 1036 |
+
"grad_norm": 0.7909725308418274,
|
| 1037 |
+
"learning_rate": 8.319153956491567e-06,
|
| 1038 |
+
"loss": 0.3708,
|
| 1039 |
+
"step": 147
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.035128805620609,
|
| 1043 |
+
"grad_norm": 0.7591220140457153,
|
| 1044 |
+
"learning_rate": 8.288369754065362e-06,
|
| 1045 |
+
"loss": 0.4101,
|
| 1046 |
+
"step": 148
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.0421545667447307,
|
| 1050 |
+
"grad_norm": 0.8976166248321533,
|
| 1051 |
+
"learning_rate": 8.257364303102275e-06,
|
| 1052 |
+
"loss": 0.3982,
|
| 1053 |
+
"step": 149
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.0491803278688525,
|
| 1057 |
+
"grad_norm": 0.8035207390785217,
|
| 1058 |
+
"learning_rate": 8.226139689715233e-06,
|
| 1059 |
+
"loss": 0.4387,
|
| 1060 |
+
"step": 150
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.0562060889929743,
|
| 1064 |
+
"grad_norm": 0.8196238875389099,
|
| 1065 |
+
"learning_rate": 8.19469801476288e-06,
|
| 1066 |
+
"loss": 0.3351,
|
| 1067 |
+
"step": 151
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.063231850117096,
|
| 1071 |
+
"grad_norm": 0.817166268825531,
|
| 1072 |
+
"learning_rate": 8.16304139370823e-06,
|
| 1073 |
+
"loss": 0.4093,
|
| 1074 |
+
"step": 152
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.0702576112412179,
|
| 1078 |
+
"grad_norm": 0.7433443069458008,
|
| 1079 |
+
"learning_rate": 8.131171956476328e-06,
|
| 1080 |
+
"loss": 0.3781,
|
| 1081 |
+
"step": 153
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.0772833723653397,
|
| 1085 |
+
"grad_norm": 0.9531522989273071,
|
| 1086 |
+
"learning_rate": 8.09909184731094e-06,
|
| 1087 |
+
"loss": 0.3705,
|
| 1088 |
+
"step": 154
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.0843091334894615,
|
| 1092 |
+
"grad_norm": 0.8055714964866638,
|
| 1093 |
+
"learning_rate": 8.066803224630295e-06,
|
| 1094 |
+
"loss": 0.4005,
|
| 1095 |
+
"step": 155
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.0913348946135832,
|
| 1099 |
+
"grad_norm": 0.8769353032112122,
|
| 1100 |
+
"learning_rate": 8.034308260881854e-06,
|
| 1101 |
+
"loss": 0.3681,
|
| 1102 |
+
"step": 156
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.098360655737705,
|
| 1106 |
+
"grad_norm": 0.8800336122512817,
|
| 1107 |
+
"learning_rate": 8.00160914239615e-06,
|
| 1108 |
+
"loss": 0.4282,
|
| 1109 |
+
"step": 157
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.1053864168618266,
|
| 1113 |
+
"grad_norm": 0.9371667504310608,
|
| 1114 |
+
"learning_rate": 7.968708069239672e-06,
|
| 1115 |
+
"loss": 0.4118,
|
| 1116 |
+
"step": 158
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.1124121779859484,
|
| 1120 |
+
"grad_norm": 0.8565946221351624,
|
| 1121 |
+
"learning_rate": 7.935607255066867e-06,
|
| 1122 |
+
"loss": 0.3738,
|
| 1123 |
+
"step": 159
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.1194379391100702,
|
| 1127 |
+
"grad_norm": 0.8784101009368896,
|
| 1128 |
+
"learning_rate": 7.902308926971166e-06,
|
| 1129 |
+
"loss": 0.3967,
|
| 1130 |
+
"step": 160
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.126463700234192,
|
| 1134 |
+
"grad_norm": 1.0153743028640747,
|
| 1135 |
+
"learning_rate": 7.868815325335168e-06,
|
| 1136 |
+
"loss": 0.3859,
|
| 1137 |
+
"step": 161
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.1334894613583137,
|
| 1141 |
+
"grad_norm": 0.8309473395347595,
|
| 1142 |
+
"learning_rate": 7.835128703679896e-06,
|
| 1143 |
+
"loss": 0.3819,
|
| 1144 |
+
"step": 162
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.1405152224824355,
|
| 1148 |
+
"grad_norm": 0.9688546657562256,
|
| 1149 |
+
"learning_rate": 7.801251328513164e-06,
|
| 1150 |
+
"loss": 0.369,
|
| 1151 |
+
"step": 163
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.1475409836065573,
|
| 1155 |
+
"grad_norm": 0.9814655780792236,
|
| 1156 |
+
"learning_rate": 7.767185479177092e-06,
|
| 1157 |
+
"loss": 0.3937,
|
| 1158 |
+
"step": 164
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.154566744730679,
|
| 1162 |
+
"grad_norm": 0.8167277574539185,
|
| 1163 |
+
"learning_rate": 7.732933447694748e-06,
|
| 1164 |
+
"loss": 0.3615,
|
| 1165 |
+
"step": 165
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.161592505854801,
|
| 1169 |
+
"grad_norm": 0.8000278472900391,
|
| 1170 |
+
"learning_rate": 7.698497538615928e-06,
|
| 1171 |
+
"loss": 0.3609,
|
| 1172 |
+
"step": 166
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.1686182669789227,
|
| 1176 |
+
"grad_norm": 0.947628915309906,
|
| 1177 |
+
"learning_rate": 7.663880068862106e-06,
|
| 1178 |
+
"loss": 0.3906,
|
| 1179 |
+
"step": 167
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.1756440281030445,
|
| 1183 |
+
"grad_norm": 1.0247867107391357,
|
| 1184 |
+
"learning_rate": 7.629083367570547e-06,
|
| 1185 |
+
"loss": 0.3896,
|
| 1186 |
+
"step": 168
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.1826697892271663,
|
| 1190 |
+
"grad_norm": 0.8357716798782349,
|
| 1191 |
+
"learning_rate": 7.594109775937595e-06,
|
| 1192 |
+
"loss": 0.3945,
|
| 1193 |
+
"step": 169
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.189695550351288,
|
| 1197 |
+
"grad_norm": 0.8735799789428711,
|
| 1198 |
+
"learning_rate": 7.558961647061156e-06,
|
| 1199 |
+
"loss": 0.3352,
|
| 1200 |
+
"step": 170
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.1967213114754098,
|
| 1204 |
+
"grad_norm": 0.8040913343429565,
|
| 1205 |
+
"learning_rate": 7.5236413457823745e-06,
|
| 1206 |
+
"loss": 0.3503,
|
| 1207 |
+
"step": 171
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.2037470725995316,
|
| 1211 |
+
"grad_norm": 0.7816208004951477,
|
| 1212 |
+
"learning_rate": 7.488151248526518e-06,
|
| 1213 |
+
"loss": 0.3538,
|
| 1214 |
+
"step": 172
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.2107728337236534,
|
| 1218 |
+
"grad_norm": 0.9761932492256165,
|
| 1219 |
+
"learning_rate": 7.452493743143092e-06,
|
| 1220 |
+
"loss": 0.4321,
|
| 1221 |
+
"step": 173
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.2177985948477752,
|
| 1225 |
+
"grad_norm": 0.8688777685165405,
|
| 1226 |
+
"learning_rate": 7.416671228745181e-06,
|
| 1227 |
+
"loss": 0.3621,
|
| 1228 |
+
"step": 174
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.224824355971897,
|
| 1232 |
+
"grad_norm": 0.8514328598976135,
|
| 1233 |
+
"learning_rate": 7.380686115548024e-06,
|
| 1234 |
+
"loss": 0.3761,
|
| 1235 |
+
"step": 175
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.2318501170960188,
|
| 1239 |
+
"grad_norm": 0.9079273343086243,
|
| 1240 |
+
"learning_rate": 7.344540824706855e-06,
|
| 1241 |
+
"loss": 0.3883,
|
| 1242 |
+
"step": 176
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.2388758782201406,
|
| 1246 |
+
"grad_norm": 0.8788741230964661,
|
| 1247 |
+
"learning_rate": 7.3082377881540025e-06,
|
| 1248 |
+
"loss": 0.3823,
|
| 1249 |
+
"step": 177
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.2459016393442623,
|
| 1253 |
+
"grad_norm": 0.9224140048027039,
|
| 1254 |
+
"learning_rate": 7.271779448435265e-06,
|
| 1255 |
+
"loss": 0.3878,
|
| 1256 |
+
"step": 178
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.2529274004683841,
|
| 1260 |
+
"grad_norm": 0.9597893953323364,
|
| 1261 |
+
"learning_rate": 7.235168258545569e-06,
|
| 1262 |
+
"loss": 0.396,
|
| 1263 |
+
"step": 179
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.259953161592506,
|
| 1267 |
+
"grad_norm": 0.9831250905990601,
|
| 1268 |
+
"learning_rate": 7.198406681763925e-06,
|
| 1269 |
+
"loss": 0.4016,
|
| 1270 |
+
"step": 180
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.2669789227166277,
|
| 1274 |
+
"grad_norm": 0.7527312636375427,
|
| 1275 |
+
"learning_rate": 7.161497191487693e-06,
|
| 1276 |
+
"loss": 0.3753,
|
| 1277 |
+
"step": 181
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.2740046838407495,
|
| 1281 |
+
"grad_norm": 0.8753437995910645,
|
| 1282 |
+
"learning_rate": 7.124442271066174e-06,
|
| 1283 |
+
"loss": 0.3785,
|
| 1284 |
+
"step": 182
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.281030444964871,
|
| 1288 |
+
"grad_norm": 0.8575742244720459,
|
| 1289 |
+
"learning_rate": 7.087244413633516e-06,
|
| 1290 |
+
"loss": 0.3865,
|
| 1291 |
+
"step": 183
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.288056206088993,
|
| 1295 |
+
"grad_norm": 0.9315704107284546,
|
| 1296 |
+
"learning_rate": 7.049906121940974e-06,
|
| 1297 |
+
"loss": 0.3672,
|
| 1298 |
+
"step": 184
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.2950819672131146,
|
| 1302 |
+
"grad_norm": 0.8974071145057678,
|
| 1303 |
+
"learning_rate": 7.012429908188523e-06,
|
| 1304 |
+
"loss": 0.3426,
|
| 1305 |
+
"step": 185
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.3021077283372366,
|
| 1309 |
+
"grad_norm": 0.8448013663291931,
|
| 1310 |
+
"learning_rate": 6.9748182938558225e-06,
|
| 1311 |
+
"loss": 0.4294,
|
| 1312 |
+
"step": 186
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.3091334894613582,
|
| 1316 |
+
"grad_norm": 0.9331867098808289,
|
| 1317 |
+
"learning_rate": 6.937073809532581e-06,
|
| 1318 |
+
"loss": 0.3471,
|
| 1319 |
+
"step": 187
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.3161592505854802,
|
| 1323 |
+
"grad_norm": 0.8858427405357361,
|
| 1324 |
+
"learning_rate": 6.899198994748274e-06,
|
| 1325 |
+
"loss": 0.3278,
|
| 1326 |
+
"step": 188
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.3231850117096018,
|
| 1330 |
+
"grad_norm": 0.7780877947807312,
|
| 1331 |
+
"learning_rate": 6.861196397801297e-06,
|
| 1332 |
+
"loss": 0.3667,
|
| 1333 |
+
"step": 189
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.3302107728337236,
|
| 1337 |
+
"grad_norm": 0.8207154870033264,
|
| 1338 |
+
"learning_rate": 6.823068575587496e-06,
|
| 1339 |
+
"loss": 0.3489,
|
| 1340 |
+
"step": 190
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.3372365339578454,
|
| 1344 |
+
"grad_norm": 0.8746094703674316,
|
| 1345 |
+
"learning_rate": 6.784818093428144e-06,
|
| 1346 |
+
"loss": 0.3779,
|
| 1347 |
+
"step": 191
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.3442622950819672,
|
| 1351 |
+
"grad_norm": 0.7831175327301025,
|
| 1352 |
+
"learning_rate": 6.746447524897335e-06,
|
| 1353 |
+
"loss": 0.3771,
|
| 1354 |
+
"step": 192
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.351288056206089,
|
| 1358 |
+
"grad_norm": 1.0498249530792236,
|
| 1359 |
+
"learning_rate": 6.70795945164883e-06,
|
| 1360 |
+
"loss": 0.3814,
|
| 1361 |
+
"step": 193
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.3583138173302107,
|
| 1365 |
+
"grad_norm": 1.0242481231689453,
|
| 1366 |
+
"learning_rate": 6.6693564632423626e-06,
|
| 1367 |
+
"loss": 0.3562,
|
| 1368 |
+
"step": 194
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.3653395784543325,
|
| 1372 |
+
"grad_norm": 0.8702220916748047,
|
| 1373 |
+
"learning_rate": 6.630641156969397e-06,
|
| 1374 |
+
"loss": 0.3702,
|
| 1375 |
+
"step": 195
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.3723653395784543,
|
| 1379 |
+
"grad_norm": 0.9588595032691956,
|
| 1380 |
+
"learning_rate": 6.591816137678388e-06,
|
| 1381 |
+
"loss": 0.392,
|
| 1382 |
+
"step": 196
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.379391100702576,
|
| 1386 |
+
"grad_norm": 0.9945947527885437,
|
| 1387 |
+
"learning_rate": 6.552884017599517e-06,
|
| 1388 |
+
"loss": 0.4089,
|
| 1389 |
+
"step": 197
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.3864168618266979,
|
| 1393 |
+
"grad_norm": 0.8764625191688538,
|
| 1394 |
+
"learning_rate": 6.513847416168929e-06,
|
| 1395 |
+
"loss": 0.3266,
|
| 1396 |
+
"step": 198
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.3934426229508197,
|
| 1400 |
+
"grad_norm": 0.8237807750701904,
|
| 1401 |
+
"learning_rate": 6.474708959852504e-06,
|
| 1402 |
+
"loss": 0.347,
|
| 1403 |
+
"step": 199
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.4004683840749415,
|
| 1407 |
+
"grad_norm": 0.8994930982589722,
|
| 1408 |
+
"learning_rate": 6.435471281969133e-06,
|
| 1409 |
+
"loss": 0.3696,
|
| 1410 |
+
"step": 200
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.4074941451990632,
|
| 1414 |
+
"grad_norm": 0.8499532341957092,
|
| 1415 |
+
"learning_rate": 6.396137022513545e-06,
|
| 1416 |
+
"loss": 0.3881,
|
| 1417 |
+
"step": 201
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.414519906323185,
|
| 1421 |
+
"grad_norm": 0.883819043636322,
|
| 1422 |
+
"learning_rate": 6.3567088279786885e-06,
|
| 1423 |
+
"loss": 0.3878,
|
| 1424 |
+
"step": 202
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.4215456674473068,
|
| 1428 |
+
"grad_norm": 1.0152641534805298,
|
| 1429 |
+
"learning_rate": 6.317189351177657e-06,
|
| 1430 |
+
"loss": 0.3853,
|
| 1431 |
+
"step": 203
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.4285714285714286,
|
| 1435 |
+
"grad_norm": 0.8048605918884277,
|
| 1436 |
+
"learning_rate": 6.277581251065217e-06,
|
| 1437 |
+
"loss": 0.4101,
|
| 1438 |
+
"step": 204
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.4355971896955504,
|
| 1442 |
+
"grad_norm": 0.9819334149360657,
|
| 1443 |
+
"learning_rate": 6.237887192558894e-06,
|
| 1444 |
+
"loss": 0.3665,
|
| 1445 |
+
"step": 205
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.4426229508196722,
|
| 1449 |
+
"grad_norm": 0.9762699007987976,
|
| 1450 |
+
"learning_rate": 6.198109846359682e-06,
|
| 1451 |
+
"loss": 0.4379,
|
| 1452 |
+
"step": 206
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.449648711943794,
|
| 1456 |
+
"grad_norm": 0.9064199924468994,
|
| 1457 |
+
"learning_rate": 6.15825188877235e-06,
|
| 1458 |
+
"loss": 0.4014,
|
| 1459 |
+
"step": 207
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.4566744730679158,
|
| 1463 |
+
"grad_norm": 0.9358543157577515,
|
| 1464 |
+
"learning_rate": 6.118316001525368e-06,
|
| 1465 |
+
"loss": 0.3528,
|
| 1466 |
+
"step": 208
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.4637002341920375,
|
| 1470 |
+
"grad_norm": 0.824491024017334,
|
| 1471 |
+
"learning_rate": 6.078304871590485e-06,
|
| 1472 |
+
"loss": 0.375,
|
| 1473 |
+
"step": 209
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.4707259953161593,
|
| 1477 |
+
"grad_norm": 0.8200578093528748,
|
| 1478 |
+
"learning_rate": 6.038221191001935e-06,
|
| 1479 |
+
"loss": 0.4005,
|
| 1480 |
+
"step": 210
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.4777517564402811,
|
| 1484 |
+
"grad_norm": 0.9158371686935425,
|
| 1485 |
+
"learning_rate": 5.998067656675318e-06,
|
| 1486 |
+
"loss": 0.3782,
|
| 1487 |
+
"step": 211
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.4847775175644027,
|
| 1491 |
+
"grad_norm": 0.8998908400535583,
|
| 1492 |
+
"learning_rate": 5.95784697022614e-06,
|
| 1493 |
+
"loss": 0.3677,
|
| 1494 |
+
"step": 212
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.4918032786885247,
|
| 1498 |
+
"grad_norm": 0.8066834211349487,
|
| 1499 |
+
"learning_rate": 5.917561837788046e-06,
|
| 1500 |
+
"loss": 0.3388,
|
| 1501 |
+
"step": 213
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.4988290398126463,
|
| 1505 |
+
"grad_norm": 0.9157138466835022,
|
| 1506 |
+
"learning_rate": 5.877214969830746e-06,
|
| 1507 |
+
"loss": 0.382,
|
| 1508 |
+
"step": 214
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.5058548009367683,
|
| 1512 |
+
"grad_norm": 0.8674065470695496,
|
| 1513 |
+
"learning_rate": 5.836809080977644e-06,
|
| 1514 |
+
"loss": 0.3749,
|
| 1515 |
+
"step": 215
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.5128805620608898,
|
| 1519 |
+
"grad_norm": 0.9985965490341187,
|
| 1520 |
+
"learning_rate": 5.7963468898232026e-06,
|
| 1521 |
+
"loss": 0.4127,
|
| 1522 |
+
"step": 216
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.5199063231850118,
|
| 1526 |
+
"grad_norm": 0.875891387462616,
|
| 1527 |
+
"learning_rate": 5.755831118750016e-06,
|
| 1528 |
+
"loss": 0.372,
|
| 1529 |
+
"step": 217
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.5269320843091334,
|
| 1533 |
+
"grad_norm": 0.8683876991271973,
|
| 1534 |
+
"learning_rate": 5.715264493745652e-06,
|
| 1535 |
+
"loss": 0.3691,
|
| 1536 |
+
"step": 218
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.5339578454332554,
|
| 1540 |
+
"grad_norm": 0.8517021536827087,
|
| 1541 |
+
"learning_rate": 5.6746497442192425e-06,
|
| 1542 |
+
"loss": 0.3539,
|
| 1543 |
+
"step": 219
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.540983606557377,
|
| 1547 |
+
"grad_norm": 0.9395655393600464,
|
| 1548 |
+
"learning_rate": 5.633989602817837e-06,
|
| 1549 |
+
"loss": 0.3917,
|
| 1550 |
+
"step": 220
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.548009367681499,
|
| 1554 |
+
"grad_norm": 1.0133756399154663,
|
| 1555 |
+
"learning_rate": 5.593286805242549e-06,
|
| 1556 |
+
"loss": 0.3833,
|
| 1557 |
+
"step": 221
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.5550351288056206,
|
| 1561 |
+
"grad_norm": 0.9718075394630432,
|
| 1562 |
+
"learning_rate": 5.552544090064487e-06,
|
| 1563 |
+
"loss": 0.3561,
|
| 1564 |
+
"step": 222
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.5620608899297423,
|
| 1568 |
+
"grad_norm": 0.7854673862457275,
|
| 1569 |
+
"learning_rate": 5.5117641985405055e-06,
|
| 1570 |
+
"loss": 0.3654,
|
| 1571 |
+
"step": 223
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.5690866510538641,
|
| 1575 |
+
"grad_norm": 0.8979426026344299,
|
| 1576 |
+
"learning_rate": 5.47094987442876e-06,
|
| 1577 |
+
"loss": 0.385,
|
| 1578 |
+
"step": 224
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.576112412177986,
|
| 1582 |
+
"grad_norm": 0.9698003530502319,
|
| 1583 |
+
"learning_rate": 5.430103863804107e-06,
|
| 1584 |
+
"loss": 0.3924,
|
| 1585 |
+
"step": 225
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.5831381733021077,
|
| 1589 |
+
"grad_norm": 0.7612743973731995,
|
| 1590 |
+
"learning_rate": 5.389228914873334e-06,
|
| 1591 |
+
"loss": 0.3567,
|
| 1592 |
+
"step": 226
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.5901639344262295,
|
| 1596 |
+
"grad_norm": 0.8035783171653748,
|
| 1597 |
+
"learning_rate": 5.348327777790262e-06,
|
| 1598 |
+
"loss": 0.3703,
|
| 1599 |
+
"step": 227
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.5971896955503513,
|
| 1603 |
+
"grad_norm": 0.9705891609191895,
|
| 1604 |
+
"learning_rate": 5.307403204470711e-06,
|
| 1605 |
+
"loss": 0.3945,
|
| 1606 |
+
"step": 228
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.604215456674473,
|
| 1610 |
+
"grad_norm": 0.8569871187210083,
|
| 1611 |
+
"learning_rate": 5.266457948407336e-06,
|
| 1612 |
+
"loss": 0.3756,
|
| 1613 |
+
"step": 229
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.6112412177985949,
|
| 1617 |
+
"grad_norm": 0.8003432154655457,
|
| 1618 |
+
"learning_rate": 5.2254947644843735e-06,
|
| 1619 |
+
"loss": 0.3585,
|
| 1620 |
+
"step": 230
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.6182669789227166,
|
| 1624 |
+
"grad_norm": 0.8991566896438599,
|
| 1625 |
+
"learning_rate": 5.18451640879228e-06,
|
| 1626 |
+
"loss": 0.3177,
|
| 1627 |
+
"step": 231
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.6252927400468384,
|
| 1631 |
+
"grad_norm": 0.7545600533485413,
|
| 1632 |
+
"learning_rate": 5.14352563844231e-06,
|
| 1633 |
+
"loss": 0.338,
|
| 1634 |
+
"step": 232
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.6323185011709602,
|
| 1638 |
+
"grad_norm": 0.8401371836662292,
|
| 1639 |
+
"learning_rate": 5.1025252113809945e-06,
|
| 1640 |
+
"loss": 0.3503,
|
| 1641 |
+
"step": 233
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.639344262295082,
|
| 1645 |
+
"grad_norm": 0.8601613640785217,
|
| 1646 |
+
"learning_rate": 5.061517886204592e-06,
|
| 1647 |
+
"loss": 0.3736,
|
| 1648 |
+
"step": 234
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.6463700234192038,
|
| 1652 |
+
"grad_norm": 0.9229597449302673,
|
| 1653 |
+
"learning_rate": 5.02050642197348e-06,
|
| 1654 |
+
"loss": 0.3645,
|
| 1655 |
+
"step": 235
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.6533957845433256,
|
| 1659 |
+
"grad_norm": 0.848315417766571,
|
| 1660 |
+
"learning_rate": 4.979493578026523e-06,
|
| 1661 |
+
"loss": 0.3935,
|
| 1662 |
+
"step": 236
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.6604215456674472,
|
| 1666 |
+
"grad_norm": 0.8743554353713989,
|
| 1667 |
+
"learning_rate": 4.9384821137954106e-06,
|
| 1668 |
+
"loss": 0.3648,
|
| 1669 |
+
"step": 237
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.6674473067915692,
|
| 1673 |
+
"grad_norm": 0.8939303755760193,
|
| 1674 |
+
"learning_rate": 4.897474788619007e-06,
|
| 1675 |
+
"loss": 0.335,
|
| 1676 |
+
"step": 238
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.6744730679156907,
|
| 1680 |
+
"grad_norm": 0.7993890643119812,
|
| 1681 |
+
"learning_rate": 4.856474361557692e-06,
|
| 1682 |
+
"loss": 0.4,
|
| 1683 |
+
"step": 239
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.6814988290398127,
|
| 1687 |
+
"grad_norm": 0.9421713948249817,
|
| 1688 |
+
"learning_rate": 4.815483591207721e-06,
|
| 1689 |
+
"loss": 0.422,
|
| 1690 |
+
"step": 240
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.6885245901639343,
|
| 1694 |
+
"grad_norm": 0.8530561923980713,
|
| 1695 |
+
"learning_rate": 4.774505235515628e-06,
|
| 1696 |
+
"loss": 0.3576,
|
| 1697 |
+
"step": 241
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.6955503512880563,
|
| 1701 |
+
"grad_norm": 0.9312227368354797,
|
| 1702 |
+
"learning_rate": 4.733542051592665e-06,
|
| 1703 |
+
"loss": 0.3848,
|
| 1704 |
+
"step": 242
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.7025761124121779,
|
| 1708 |
+
"grad_norm": 1.0025192499160767,
|
| 1709 |
+
"learning_rate": 4.69259679552929e-06,
|
| 1710 |
+
"loss": 0.4112,
|
| 1711 |
+
"step": 243
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.7096018735362999,
|
| 1715 |
+
"grad_norm": 0.9068251252174377,
|
| 1716 |
+
"learning_rate": 4.651672222209738e-06,
|
| 1717 |
+
"loss": 0.3632,
|
| 1718 |
+
"step": 244
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.7166276346604215,
|
| 1722 |
+
"grad_norm": 0.843632698059082,
|
| 1723 |
+
"learning_rate": 4.6107710851266695e-06,
|
| 1724 |
+
"loss": 0.3583,
|
| 1725 |
+
"step": 245
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.7236533957845435,
|
| 1729 |
+
"grad_norm": 0.8394821882247925,
|
| 1730 |
+
"learning_rate": 4.5698961361958955e-06,
|
| 1731 |
+
"loss": 0.3598,
|
| 1732 |
+
"step": 246
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.730679156908665,
|
| 1736 |
+
"grad_norm": 1.0585048198699951,
|
| 1737 |
+
"learning_rate": 4.529050125571241e-06,
|
| 1738 |
+
"loss": 0.3331,
|
| 1739 |
+
"step": 247
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.737704918032787,
|
| 1743 |
+
"grad_norm": 0.8787031769752502,
|
| 1744 |
+
"learning_rate": 4.488235801459495e-06,
|
| 1745 |
+
"loss": 0.3934,
|
| 1746 |
+
"step": 248
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.7447306791569086,
|
| 1750 |
+
"grad_norm": 0.9514378309249878,
|
| 1751 |
+
"learning_rate": 4.447455909935513e-06,
|
| 1752 |
+
"loss": 0.3574,
|
| 1753 |
+
"step": 249
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.7517564402810304,
|
| 1757 |
+
"grad_norm": 0.8068012595176697,
|
| 1758 |
+
"learning_rate": 4.4067131947574515e-06,
|
| 1759 |
+
"loss": 0.3703,
|
| 1760 |
+
"step": 250
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.7587822014051522,
|
| 1764 |
+
"grad_norm": 0.9842183589935303,
|
| 1765 |
+
"learning_rate": 4.3660103971821635e-06,
|
| 1766 |
+
"loss": 0.3628,
|
| 1767 |
+
"step": 251
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.765807962529274,
|
| 1771 |
+
"grad_norm": 0.9102012515068054,
|
| 1772 |
+
"learning_rate": 4.3253502557807575e-06,
|
| 1773 |
+
"loss": 0.395,
|
| 1774 |
+
"step": 252
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.7728337236533958,
|
| 1778 |
+
"grad_norm": 0.8737331628799438,
|
| 1779 |
+
"learning_rate": 4.28473550625435e-06,
|
| 1780 |
+
"loss": 0.3949,
|
| 1781 |
+
"step": 253
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.7798594847775175,
|
| 1785 |
+
"grad_norm": 0.8551650643348694,
|
| 1786 |
+
"learning_rate": 4.244168881249986e-06,
|
| 1787 |
+
"loss": 0.3237,
|
| 1788 |
+
"step": 254
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.7868852459016393,
|
| 1792 |
+
"grad_norm": 0.8604973554611206,
|
| 1793 |
+
"learning_rate": 4.203653110176798e-06,
|
| 1794 |
+
"loss": 0.36,
|
| 1795 |
+
"step": 255
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.7939110070257611,
|
| 1799 |
+
"grad_norm": 0.8735095858573914,
|
| 1800 |
+
"learning_rate": 4.163190919022357e-06,
|
| 1801 |
+
"loss": 0.3947,
|
| 1802 |
+
"step": 256
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.800936768149883,
|
| 1806 |
+
"grad_norm": 0.8284133672714233,
|
| 1807 |
+
"learning_rate": 4.122785030169256e-06,
|
| 1808 |
+
"loss": 0.3544,
|
| 1809 |
+
"step": 257
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.8079625292740047,
|
| 1813 |
+
"grad_norm": 0.948117196559906,
|
| 1814 |
+
"learning_rate": 4.082438162211955e-06,
|
| 1815 |
+
"loss": 0.355,
|
| 1816 |
+
"step": 258
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 1.8149882903981265,
|
| 1820 |
+
"grad_norm": 0.8387302756309509,
|
| 1821 |
+
"learning_rate": 4.042153029773861e-06,
|
| 1822 |
+
"loss": 0.378,
|
| 1823 |
+
"step": 259
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.8220140515222483,
|
| 1827 |
+
"grad_norm": 1.0318608283996582,
|
| 1828 |
+
"learning_rate": 4.001932343324683e-06,
|
| 1829 |
+
"loss": 0.4138,
|
| 1830 |
+
"step": 260
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.82903981264637,
|
| 1834 |
+
"grad_norm": 0.9296135306358337,
|
| 1835 |
+
"learning_rate": 3.961778808998066e-06,
|
| 1836 |
+
"loss": 0.4052,
|
| 1837 |
+
"step": 261
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 1.8360655737704918,
|
| 1841 |
+
"grad_norm": 0.8200930953025818,
|
| 1842 |
+
"learning_rate": 3.921695128409517e-06,
|
| 1843 |
+
"loss": 0.3252,
|
| 1844 |
+
"step": 262
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.8430913348946136,
|
| 1848 |
+
"grad_norm": 0.8758650422096252,
|
| 1849 |
+
"learning_rate": 3.8816839984746334e-06,
|
| 1850 |
+
"loss": 0.364,
|
| 1851 |
+
"step": 263
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.8501170960187352,
|
| 1855 |
+
"grad_norm": 0.8040407299995422,
|
| 1856 |
+
"learning_rate": 3.841748111227652e-06,
|
| 1857 |
+
"loss": 0.4083,
|
| 1858 |
+
"step": 264
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.8571428571428572,
|
| 1862 |
+
"grad_norm": 0.9389622807502747,
|
| 1863 |
+
"learning_rate": 3.8018901536403198e-06,
|
| 1864 |
+
"loss": 0.4112,
|
| 1865 |
+
"step": 265
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 1.8641686182669788,
|
| 1869 |
+
"grad_norm": 0.9404301047325134,
|
| 1870 |
+
"learning_rate": 3.762112807441108e-06,
|
| 1871 |
+
"loss": 0.3785,
|
| 1872 |
+
"step": 266
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.8711943793911008,
|
| 1876 |
+
"grad_norm": 0.984291672706604,
|
| 1877 |
+
"learning_rate": 3.7224187489347847e-06,
|
| 1878 |
+
"loss": 0.4063,
|
| 1879 |
+
"step": 267
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 1.8782201405152223,
|
| 1883 |
+
"grad_norm": 0.8425026535987854,
|
| 1884 |
+
"learning_rate": 3.682810648822343e-06,
|
| 1885 |
+
"loss": 0.3901,
|
| 1886 |
+
"step": 268
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.8852459016393444,
|
| 1890 |
+
"grad_norm": 0.9068894982337952,
|
| 1891 |
+
"learning_rate": 3.6432911720213127e-06,
|
| 1892 |
+
"loss": 0.3549,
|
| 1893 |
+
"step": 269
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.892271662763466,
|
| 1897 |
+
"grad_norm": 0.9722284078598022,
|
| 1898 |
+
"learning_rate": 3.6038629774864563e-06,
|
| 1899 |
+
"loss": 0.4069,
|
| 1900 |
+
"step": 270
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.899297423887588,
|
| 1904 |
+
"grad_norm": 0.9396534562110901,
|
| 1905 |
+
"learning_rate": 3.56452871803087e-06,
|
| 1906 |
+
"loss": 0.3223,
|
| 1907 |
+
"step": 271
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.9063231850117095,
|
| 1911 |
+
"grad_norm": 0.949874758720398,
|
| 1912 |
+
"learning_rate": 3.525291040147498e-06,
|
| 1913 |
+
"loss": 0.3538,
|
| 1914 |
+
"step": 272
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.9133489461358315,
|
| 1918 |
+
"grad_norm": 0.9677289128303528,
|
| 1919 |
+
"learning_rate": 3.486152583831072e-06,
|
| 1920 |
+
"loss": 0.3828,
|
| 1921 |
+
"step": 273
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.920374707259953,
|
| 1925 |
+
"grad_norm": 0.8510810136795044,
|
| 1926 |
+
"learning_rate": 3.447115982400485e-06,
|
| 1927 |
+
"loss": 0.4034,
|
| 1928 |
+
"step": 274
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.927400468384075,
|
| 1932 |
+
"grad_norm": 0.8667073845863342,
|
| 1933 |
+
"learning_rate": 3.4081838623216124e-06,
|
| 1934 |
+
"loss": 0.32,
|
| 1935 |
+
"step": 275
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.9344262295081966,
|
| 1939 |
+
"grad_norm": 0.8696290850639343,
|
| 1940 |
+
"learning_rate": 3.3693588430306035e-06,
|
| 1941 |
+
"loss": 0.365,
|
| 1942 |
+
"step": 276
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.9414519906323187,
|
| 1946 |
+
"grad_norm": 0.8641108870506287,
|
| 1947 |
+
"learning_rate": 3.330643536757638e-06,
|
| 1948 |
+
"loss": 0.3681,
|
| 1949 |
+
"step": 277
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 1.9484777517564402,
|
| 1953 |
+
"grad_norm": 0.9129387736320496,
|
| 1954 |
+
"learning_rate": 3.2920405483511702e-06,
|
| 1955 |
+
"loss": 0.3651,
|
| 1956 |
+
"step": 278
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.955503512880562,
|
| 1960 |
+
"grad_norm": 0.8642932772636414,
|
| 1961 |
+
"learning_rate": 3.253552475102668e-06,
|
| 1962 |
+
"loss": 0.3972,
|
| 1963 |
+
"step": 279
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 1.9625292740046838,
|
| 1967 |
+
"grad_norm": 0.8414629697799683,
|
| 1968 |
+
"learning_rate": 3.215181906571858e-06,
|
| 1969 |
+
"loss": 0.3707,
|
| 1970 |
+
"step": 280
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.9695550351288056,
|
| 1974 |
+
"grad_norm": 0.9968050718307495,
|
| 1975 |
+
"learning_rate": 3.1769314244125056e-06,
|
| 1976 |
+
"loss": 0.4179,
|
| 1977 |
+
"step": 281
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.9765807962529274,
|
| 1981 |
+
"grad_norm": 0.8431434035301208,
|
| 1982 |
+
"learning_rate": 3.1388036021987047e-06,
|
| 1983 |
+
"loss": 0.3617,
|
| 1984 |
+
"step": 282
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.9836065573770492,
|
| 1988 |
+
"grad_norm": 0.957405686378479,
|
| 1989 |
+
"learning_rate": 3.100801005251727e-06,
|
| 1990 |
+
"loss": 0.374,
|
| 1991 |
+
"step": 283
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.990632318501171,
|
| 1995 |
+
"grad_norm": 0.8993059992790222,
|
| 1996 |
+
"learning_rate": 3.0629261904674206e-06,
|
| 1997 |
+
"loss": 0.337,
|
| 1998 |
+
"step": 284
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.9976580796252927,
|
| 2002 |
+
"grad_norm": 0.915682315826416,
|
| 2003 |
+
"learning_rate": 3.025181706144178e-06,
|
| 2004 |
+
"loss": 0.4074,
|
| 2005 |
+
"step": 285
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.0,
|
| 2009 |
+
"grad_norm": 0.915682315826416,
|
| 2010 |
+
"learning_rate": 2.987570091811479e-06,
|
| 2011 |
+
"loss": 0.3528,
|
| 2012 |
+
"step": 286
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.0070257611241216,
|
| 2016 |
+
"grad_norm": 1.7454204559326172,
|
| 2017 |
+
"learning_rate": 2.9500938780590276e-06,
|
| 2018 |
+
"loss": 0.294,
|
| 2019 |
+
"step": 287
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.0140515222482436,
|
| 2023 |
+
"grad_norm": 0.90482097864151,
|
| 2024 |
+
"learning_rate": 2.9127555863664857e-06,
|
| 2025 |
+
"loss": 0.2891,
|
| 2026 |
+
"step": 288
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.021077283372365,
|
| 2030 |
+
"grad_norm": 0.8985456228256226,
|
| 2031 |
+
"learning_rate": 2.8755577289338267e-06,
|
| 2032 |
+
"loss": 0.255,
|
| 2033 |
+
"step": 289
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.028103044496487,
|
| 2037 |
+
"grad_norm": 1.0324161052703857,
|
| 2038 |
+
"learning_rate": 2.838502808512309e-06,
|
| 2039 |
+
"loss": 0.275,
|
| 2040 |
+
"step": 290
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.0351288056206087,
|
| 2044 |
+
"grad_norm": 0.8929424285888672,
|
| 2045 |
+
"learning_rate": 2.801593318236078e-06,
|
| 2046 |
+
"loss": 0.2969,
|
| 2047 |
+
"step": 291
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 2.0421545667447307,
|
| 2051 |
+
"grad_norm": 0.9224628806114197,
|
| 2052 |
+
"learning_rate": 2.764831741454432e-06,
|
| 2053 |
+
"loss": 0.2285,
|
| 2054 |
+
"step": 292
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.0491803278688523,
|
| 2058 |
+
"grad_norm": 0.8780328631401062,
|
| 2059 |
+
"learning_rate": 2.7282205515647348e-06,
|
| 2060 |
+
"loss": 0.2345,
|
| 2061 |
+
"step": 293
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.0562060889929743,
|
| 2065 |
+
"grad_norm": 0.838097095489502,
|
| 2066 |
+
"learning_rate": 2.6917622118459975e-06,
|
| 2067 |
+
"loss": 0.2246,
|
| 2068 |
+
"step": 294
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.063231850117096,
|
| 2072 |
+
"grad_norm": 0.7811418175697327,
|
| 2073 |
+
"learning_rate": 2.655459175293146e-06,
|
| 2074 |
+
"loss": 0.2824,
|
| 2075 |
+
"step": 295
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.070257611241218,
|
| 2079 |
+
"grad_norm": 0.9212809205055237,
|
| 2080 |
+
"learning_rate": 2.6193138844519785e-06,
|
| 2081 |
+
"loss": 0.25,
|
| 2082 |
+
"step": 296
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 2.0772833723653394,
|
| 2086 |
+
"grad_norm": 0.8795907497406006,
|
| 2087 |
+
"learning_rate": 2.58332877125482e-06,
|
| 2088 |
+
"loss": 0.2862,
|
| 2089 |
+
"step": 297
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 2.0843091334894615,
|
| 2093 |
+
"grad_norm": 1.2901580333709717,
|
| 2094 |
+
"learning_rate": 2.5475062568569077e-06,
|
| 2095 |
+
"loss": 0.2577,
|
| 2096 |
+
"step": 298
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.091334894613583,
|
| 2100 |
+
"grad_norm": 0.9055448174476624,
|
| 2101 |
+
"learning_rate": 2.511848751473485e-06,
|
| 2102 |
+
"loss": 0.2337,
|
| 2103 |
+
"step": 299
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.098360655737705,
|
| 2107 |
+
"grad_norm": 1.0659441947937012,
|
| 2108 |
+
"learning_rate": 2.476358654217627e-06,
|
| 2109 |
+
"loss": 0.2564,
|
| 2110 |
+
"step": 300
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 2.1053864168618266,
|
| 2114 |
+
"grad_norm": 1.0252413749694824,
|
| 2115 |
+
"learning_rate": 2.4410383529388448e-06,
|
| 2116 |
+
"loss": 0.2536,
|
| 2117 |
+
"step": 301
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 2.1124121779859486,
|
| 2121 |
+
"grad_norm": 0.8777045011520386,
|
| 2122 |
+
"learning_rate": 2.405890224062406e-06,
|
| 2123 |
+
"loss": 0.2225,
|
| 2124 |
+
"step": 302
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 2.11943793911007,
|
| 2128 |
+
"grad_norm": 0.8778117299079895,
|
| 2129 |
+
"learning_rate": 2.370916632429455e-06,
|
| 2130 |
+
"loss": 0.2544,
|
| 2131 |
+
"step": 303
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 2.126463700234192,
|
| 2135 |
+
"grad_norm": 1.0625895261764526,
|
| 2136 |
+
"learning_rate": 2.336119931137897e-06,
|
| 2137 |
+
"loss": 0.258,
|
| 2138 |
+
"step": 304
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.1334894613583137,
|
| 2142 |
+
"grad_norm": 0.9435818195343018,
|
| 2143 |
+
"learning_rate": 2.3015024613840742e-06,
|
| 2144 |
+
"loss": 0.2564,
|
| 2145 |
+
"step": 305
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.1405152224824358,
|
| 2149 |
+
"grad_norm": 0.8762299418449402,
|
| 2150 |
+
"learning_rate": 2.2670665523052534e-06,
|
| 2151 |
+
"loss": 0.2109,
|
| 2152 |
+
"step": 306
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 2.1475409836065573,
|
| 2156 |
+
"grad_norm": 0.8183939456939697,
|
| 2157 |
+
"learning_rate": 2.2328145208229096e-06,
|
| 2158 |
+
"loss": 0.2277,
|
| 2159 |
+
"step": 307
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.1545667447306793,
|
| 2163 |
+
"grad_norm": 0.9453922510147095,
|
| 2164 |
+
"learning_rate": 2.1987486714868384e-06,
|
| 2165 |
+
"loss": 0.2555,
|
| 2166 |
+
"step": 308
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 2.161592505854801,
|
| 2170 |
+
"grad_norm": 0.905980110168457,
|
| 2171 |
+
"learning_rate": 2.164871296320106e-06,
|
| 2172 |
+
"loss": 0.2236,
|
| 2173 |
+
"step": 309
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 2.168618266978923,
|
| 2177 |
+
"grad_norm": 0.7942854166030884,
|
| 2178 |
+
"learning_rate": 2.1311846746648325e-06,
|
| 2179 |
+
"loss": 0.2435,
|
| 2180 |
+
"step": 310
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 2.1756440281030445,
|
| 2184 |
+
"grad_norm": 0.8481754064559937,
|
| 2185 |
+
"learning_rate": 2.097691073028836e-06,
|
| 2186 |
+
"loss": 0.2503,
|
| 2187 |
+
"step": 311
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.1826697892271665,
|
| 2191 |
+
"grad_norm": 0.8431323170661926,
|
| 2192 |
+
"learning_rate": 2.064392744933135e-06,
|
| 2193 |
+
"loss": 0.2242,
|
| 2194 |
+
"step": 312
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 2.189695550351288,
|
| 2198 |
+
"grad_norm": 0.7957973480224609,
|
| 2199 |
+
"learning_rate": 2.0312919307603286e-06,
|
| 2200 |
+
"loss": 0.2317,
|
| 2201 |
+
"step": 313
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.19672131147541,
|
| 2205 |
+
"grad_norm": 0.8730148673057556,
|
| 2206 |
+
"learning_rate": 1.998390857603853e-06,
|
| 2207 |
+
"loss": 0.2259,
|
| 2208 |
+
"step": 314
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 2.2037470725995316,
|
| 2212 |
+
"grad_norm": 0.9832536578178406,
|
| 2213 |
+
"learning_rate": 1.965691739118146e-06,
|
| 2214 |
+
"loss": 0.2129,
|
| 2215 |
+
"step": 315
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 2.210772833723653,
|
| 2219 |
+
"grad_norm": 0.8978882431983948,
|
| 2220 |
+
"learning_rate": 1.9331967753697077e-06,
|
| 2221 |
+
"loss": 0.2167,
|
| 2222 |
+
"step": 316
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.217798594847775,
|
| 2226 |
+
"grad_norm": 1.0279768705368042,
|
| 2227 |
+
"learning_rate": 1.9009081526890622e-06,
|
| 2228 |
+
"loss": 0.2162,
|
| 2229 |
+
"step": 317
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.2248243559718968,
|
| 2233 |
+
"grad_norm": 0.9898053407669067,
|
| 2234 |
+
"learning_rate": 1.8688280435236732e-06,
|
| 2235 |
+
"loss": 0.2652,
|
| 2236 |
+
"step": 318
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.2318501170960188,
|
| 2240 |
+
"grad_norm": 0.9467569589614868,
|
| 2241 |
+
"learning_rate": 1.8369586062917693e-06,
|
| 2242 |
+
"loss": 0.2541,
|
| 2243 |
+
"step": 319
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.2388758782201403,
|
| 2247 |
+
"grad_norm": 0.9604997038841248,
|
| 2248 |
+
"learning_rate": 1.8053019852371195e-06,
|
| 2249 |
+
"loss": 0.2388,
|
| 2250 |
+
"step": 320
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.2459016393442623,
|
| 2254 |
+
"grad_norm": 0.8671456575393677,
|
| 2255 |
+
"learning_rate": 1.7738603102847696e-06,
|
| 2256 |
+
"loss": 0.2688,
|
| 2257 |
+
"step": 321
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.252927400468384,
|
| 2261 |
+
"grad_norm": 1.0521889925003052,
|
| 2262 |
+
"learning_rate": 1.7426356968977265e-06,
|
| 2263 |
+
"loss": 0.2702,
|
| 2264 |
+
"step": 322
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.259953161592506,
|
| 2268 |
+
"grad_norm": 0.9155137538909912,
|
| 2269 |
+
"learning_rate": 1.711630245934638e-06,
|
| 2270 |
+
"loss": 0.2632,
|
| 2271 |
+
"step": 323
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.2669789227166275,
|
| 2275 |
+
"grad_norm": 0.9298955798149109,
|
| 2276 |
+
"learning_rate": 1.6808460435084316e-06,
|
| 2277 |
+
"loss": 0.2485,
|
| 2278 |
+
"step": 324
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.2740046838407495,
|
| 2282 |
+
"grad_norm": 0.8703210353851318,
|
| 2283 |
+
"learning_rate": 1.6502851608459668e-06,
|
| 2284 |
+
"loss": 0.2268,
|
| 2285 |
+
"step": 325
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.281030444964871,
|
| 2289 |
+
"grad_norm": 0.8722267746925354,
|
| 2290 |
+
"learning_rate": 1.6199496541486647e-06,
|
| 2291 |
+
"loss": 0.258,
|
| 2292 |
+
"step": 326
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.288056206088993,
|
| 2296 |
+
"grad_norm": 0.8431782126426697,
|
| 2297 |
+
"learning_rate": 1.589841564454176e-06,
|
| 2298 |
+
"loss": 0.2587,
|
| 2299 |
+
"step": 327
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.2950819672131146,
|
| 2303 |
+
"grad_norm": 0.9284111857414246,
|
| 2304 |
+
"learning_rate": 1.5599629174990482e-06,
|
| 2305 |
+
"loss": 0.2383,
|
| 2306 |
+
"step": 328
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.3021077283372366,
|
| 2310 |
+
"grad_norm": 0.8235135078430176,
|
| 2311 |
+
"learning_rate": 1.5303157235824323e-06,
|
| 2312 |
+
"loss": 0.2379,
|
| 2313 |
+
"step": 329
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.309133489461358,
|
| 2317 |
+
"grad_norm": 0.8774765729904175,
|
| 2318 |
+
"learning_rate": 1.5009019774308249e-06,
|
| 2319 |
+
"loss": 0.2727,
|
| 2320 |
+
"step": 330
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.3161592505854802,
|
| 2324 |
+
"grad_norm": 0.9658228158950806,
|
| 2325 |
+
"learning_rate": 1.471723658063856e-06,
|
| 2326 |
+
"loss": 0.2604,
|
| 2327 |
+
"step": 331
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.323185011709602,
|
| 2331 |
+
"grad_norm": 0.9167743921279907,
|
| 2332 |
+
"learning_rate": 1.4427827286611412e-06,
|
| 2333 |
+
"loss": 0.2106,
|
| 2334 |
+
"step": 332
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 2.330210772833724,
|
| 2338 |
+
"grad_norm": 0.9037270545959473,
|
| 2339 |
+
"learning_rate": 1.4140811364301931e-06,
|
| 2340 |
+
"loss": 0.2499,
|
| 2341 |
+
"step": 333
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 2.3372365339578454,
|
| 2345 |
+
"grad_norm": 0.9935373067855835,
|
| 2346 |
+
"learning_rate": 1.385620812475409e-06,
|
| 2347 |
+
"loss": 0.2772,
|
| 2348 |
+
"step": 334
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 2.3442622950819674,
|
| 2352 |
+
"grad_norm": 0.7985637187957764,
|
| 2353 |
+
"learning_rate": 1.3574036716681366e-06,
|
| 2354 |
+
"loss": 0.2575,
|
| 2355 |
+
"step": 335
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.351288056206089,
|
| 2359 |
+
"grad_norm": 0.8993780612945557,
|
| 2360 |
+
"learning_rate": 1.3294316125178474e-06,
|
| 2361 |
+
"loss": 0.2432,
|
| 2362 |
+
"step": 336
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.358313817330211,
|
| 2366 |
+
"grad_norm": 0.9015387892723083,
|
| 2367 |
+
"learning_rate": 1.301706517044395e-06,
|
| 2368 |
+
"loss": 0.2541,
|
| 2369 |
+
"step": 337
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.3653395784543325,
|
| 2373 |
+
"grad_norm": 0.9596133828163147,
|
| 2374 |
+
"learning_rate": 1.2742302506513894e-06,
|
| 2375 |
+
"loss": 0.2702,
|
| 2376 |
+
"step": 338
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 2.3723653395784545,
|
| 2380 |
+
"grad_norm": 0.9486516714096069,
|
| 2381 |
+
"learning_rate": 1.247004662000686e-06,
|
| 2382 |
+
"loss": 0.2547,
|
| 2383 |
+
"step": 339
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 2.379391100702576,
|
| 2387 |
+
"grad_norm": 1.1242049932479858,
|
| 2388 |
+
"learning_rate": 1.2200315828880094e-06,
|
| 2389 |
+
"loss": 0.2384,
|
| 2390 |
+
"step": 340
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 2.3864168618266977,
|
| 2394 |
+
"grad_norm": 0.9570393562316895,
|
| 2395 |
+
"learning_rate": 1.1933128281197042e-06,
|
| 2396 |
+
"loss": 0.2759,
|
| 2397 |
+
"step": 341
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.3934426229508197,
|
| 2401 |
+
"grad_norm": 0.9540587663650513,
|
| 2402 |
+
"learning_rate": 1.166850195390628e-06,
|
| 2403 |
+
"loss": 0.2709,
|
| 2404 |
+
"step": 342
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 2.4004683840749417,
|
| 2408 |
+
"grad_norm": 1.0115498304367065,
|
| 2409 |
+
"learning_rate": 1.1406454651632042e-06,
|
| 2410 |
+
"loss": 0.2795,
|
| 2411 |
+
"step": 343
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.4074941451990632,
|
| 2415 |
+
"grad_norm": 1.0347506999969482,
|
| 2416 |
+
"learning_rate": 1.1147004005476192e-06,
|
| 2417 |
+
"loss": 0.208,
|
| 2418 |
+
"step": 344
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 2.414519906323185,
|
| 2422 |
+
"grad_norm": 0.988497793674469,
|
| 2423 |
+
"learning_rate": 1.089016747183208e-06,
|
| 2424 |
+
"loss": 0.2472,
|
| 2425 |
+
"step": 345
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 2.421545667447307,
|
| 2429 |
+
"grad_norm": 0.8970013856887817,
|
| 2430 |
+
"learning_rate": 1.063596233120997e-06,
|
| 2431 |
+
"loss": 0.2649,
|
| 2432 |
+
"step": 346
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 2.4285714285714284,
|
| 2436 |
+
"grad_norm": 0.9206312298774719,
|
| 2437 |
+
"learning_rate": 1.03844056870744e-06,
|
| 2438 |
+
"loss": 0.2352,
|
| 2439 |
+
"step": 347
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 2.4355971896955504,
|
| 2443 |
+
"grad_norm": 0.9012823700904846,
|
| 2444 |
+
"learning_rate": 1.013551446469337e-06,
|
| 2445 |
+
"loss": 0.2207,
|
| 2446 |
+
"step": 348
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 2.442622950819672,
|
| 2450 |
+
"grad_norm": 1.0099748373031616,
|
| 2451 |
+
"learning_rate": 9.889305409999656e-07,
|
| 2452 |
+
"loss": 0.298,
|
| 2453 |
+
"step": 349
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 2.449648711943794,
|
| 2457 |
+
"grad_norm": 1.0725970268249512,
|
| 2458 |
+
"learning_rate": 9.64579508846405e-07,
|
| 2459 |
+
"loss": 0.2466,
|
| 2460 |
+
"step": 350
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 2.4566744730679155,
|
| 2464 |
+
"grad_norm": 0.9620457887649536,
|
| 2465 |
+
"learning_rate": 9.40499988398082e-07,
|
| 2466 |
+
"loss": 0.2746,
|
| 2467 |
+
"step": 351
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 2.4637002341920375,
|
| 2471 |
+
"grad_norm": 1.0769095420837402,
|
| 2472 |
+
"learning_rate": 9.166935997765364e-07,
|
| 2473 |
+
"loss": 0.2506,
|
| 2474 |
+
"step": 352
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 2.470725995316159,
|
| 2478 |
+
"grad_norm": 0.9816381931304932,
|
| 2479 |
+
"learning_rate": 8.93161944726414e-07,
|
| 2480 |
+
"loss": 0.2605,
|
| 2481 |
+
"step": 353
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 2.477751756440281,
|
| 2485 |
+
"grad_norm": 0.8115105032920837,
|
| 2486 |
+
"learning_rate": 8.699066065077005e-07,
|
| 2487 |
+
"loss": 0.2515,
|
| 2488 |
+
"step": 354
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 2.4847775175644027,
|
| 2492 |
+
"grad_norm": 0.9756121039390564,
|
| 2493 |
+
"learning_rate": 8.469291497891979e-07,
|
| 2494 |
+
"loss": 0.2223,
|
| 2495 |
+
"step": 355
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 2.4918032786885247,
|
| 2499 |
+
"grad_norm": 0.8359687328338623,
|
| 2500 |
+
"learning_rate": 8.242311205432418e-07,
|
| 2501 |
+
"loss": 0.2633,
|
| 2502 |
+
"step": 356
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 2.4988290398126463,
|
| 2506 |
+
"grad_norm": 1.0315937995910645,
|
| 2507 |
+
"learning_rate": 8.018140459416962e-07,
|
| 2508 |
+
"loss": 0.2505,
|
| 2509 |
+
"step": 357
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 2.5058548009367683,
|
| 2513 |
+
"grad_norm": 0.9899025559425354,
|
| 2514 |
+
"learning_rate": 7.796794342531949e-07,
|
| 2515 |
+
"loss": 0.2504,
|
| 2516 |
+
"step": 358
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.51288056206089,
|
| 2520 |
+
"grad_norm": 0.8839271068572998,
|
| 2521 |
+
"learning_rate": 7.57828774741664e-07,
|
| 2522 |
+
"loss": 0.265,
|
| 2523 |
+
"step": 359
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 2.519906323185012,
|
| 2527 |
+
"grad_norm": 1.0711992979049683,
|
| 2528 |
+
"learning_rate": 7.362635375661225e-07,
|
| 2529 |
+
"loss": 0.285,
|
| 2530 |
+
"step": 360
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 2.5269320843091334,
|
| 2534 |
+
"grad_norm": 1.0341910123825073,
|
| 2535 |
+
"learning_rate": 7.149851736817609e-07,
|
| 2536 |
+
"loss": 0.2219,
|
| 2537 |
+
"step": 361
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 2.5339578454332554,
|
| 2541 |
+
"grad_norm": 0.987358033657074,
|
| 2542 |
+
"learning_rate": 6.939951147423269e-07,
|
| 2543 |
+
"loss": 0.2338,
|
| 2544 |
+
"step": 362
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 2.540983606557377,
|
| 2548 |
+
"grad_norm": 0.9864411950111389,
|
| 2549 |
+
"learning_rate": 6.732947730037936e-07,
|
| 2550 |
+
"loss": 0.2466,
|
| 2551 |
+
"step": 363
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 2.548009367681499,
|
| 2555 |
+
"grad_norm": 1.023818016052246,
|
| 2556 |
+
"learning_rate": 6.52885541229345e-07,
|
| 2557 |
+
"loss": 0.2357,
|
| 2558 |
+
"step": 364
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 2.5550351288056206,
|
| 2562 |
+
"grad_norm": 0.8425894975662231,
|
| 2563 |
+
"learning_rate": 6.327687925956616e-07,
|
| 2564 |
+
"loss": 0.2216,
|
| 2565 |
+
"step": 365
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 2.562060889929742,
|
| 2569 |
+
"grad_norm": 0.9226377606391907,
|
| 2570 |
+
"learning_rate": 6.12945880600535e-07,
|
| 2571 |
+
"loss": 0.2818,
|
| 2572 |
+
"step": 366
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 2.569086651053864,
|
| 2576 |
+
"grad_norm": 0.922722339630127,
|
| 2577 |
+
"learning_rate": 5.93418138971803e-07,
|
| 2578 |
+
"loss": 0.2605,
|
| 2579 |
+
"step": 367
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 2.576112412177986,
|
| 2583 |
+
"grad_norm": 1.0783159732818604,
|
| 2584 |
+
"learning_rate": 5.741868815776081e-07,
|
| 2585 |
+
"loss": 0.257,
|
| 2586 |
+
"step": 368
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 2.5831381733021077,
|
| 2590 |
+
"grad_norm": 0.9857349991798401,
|
| 2591 |
+
"learning_rate": 5.552534023380024e-07,
|
| 2592 |
+
"loss": 0.2133,
|
| 2593 |
+
"step": 369
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 2.5901639344262293,
|
| 2597 |
+
"grad_norm": 0.8876913785934448,
|
| 2598 |
+
"learning_rate": 5.366189751378858e-07,
|
| 2599 |
+
"loss": 0.2728,
|
| 2600 |
+
"step": 370
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 2.5971896955503513,
|
| 2604 |
+
"grad_norm": 1.0224825143814087,
|
| 2605 |
+
"learning_rate": 5.18284853741301e-07,
|
| 2606 |
+
"loss": 0.2375,
|
| 2607 |
+
"step": 371
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 2.6042154566744733,
|
| 2611 |
+
"grad_norm": 0.9141249656677246,
|
| 2612 |
+
"learning_rate": 5.002522717070751e-07,
|
| 2613 |
+
"loss": 0.2436,
|
| 2614 |
+
"step": 372
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 2.611241217798595,
|
| 2618 |
+
"grad_norm": 0.9188523292541504,
|
| 2619 |
+
"learning_rate": 4.8252244230582e-07,
|
| 2620 |
+
"loss": 0.2435,
|
| 2621 |
+
"step": 373
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 2.6182669789227164,
|
| 2625 |
+
"grad_norm": 0.9691129922866821,
|
| 2626 |
+
"learning_rate": 4.6509655843830827e-07,
|
| 2627 |
+
"loss": 0.2444,
|
| 2628 |
+
"step": 374
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 2.6252927400468384,
|
| 2632 |
+
"grad_norm": 0.9464864730834961,
|
| 2633 |
+
"learning_rate": 4.4797579255520585e-07,
|
| 2634 |
+
"loss": 0.2615,
|
| 2635 |
+
"step": 375
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 2.6323185011709604,
|
| 2639 |
+
"grad_norm": 0.7935175895690918,
|
| 2640 |
+
"learning_rate": 4.311612965781903e-07,
|
| 2641 |
+
"loss": 0.2553,
|
| 2642 |
+
"step": 376
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 2.639344262295082,
|
| 2646 |
+
"grad_norm": 0.9593071937561035,
|
| 2647 |
+
"learning_rate": 4.1465420182244476e-07,
|
| 2648 |
+
"loss": 0.2364,
|
| 2649 |
+
"step": 377
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 2.6463700234192036,
|
| 2653 |
+
"grad_norm": 0.9605007171630859,
|
| 2654 |
+
"learning_rate": 3.984556189205441e-07,
|
| 2655 |
+
"loss": 0.2647,
|
| 2656 |
+
"step": 378
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 2.6533957845433256,
|
| 2660 |
+
"grad_norm": 0.9680880904197693,
|
| 2661 |
+
"learning_rate": 3.8256663774772383e-07,
|
| 2662 |
+
"loss": 0.2778,
|
| 2663 |
+
"step": 379
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 2.660421545667447,
|
| 2667 |
+
"grad_norm": 0.9246297478675842,
|
| 2668 |
+
"learning_rate": 3.669883273485575e-07,
|
| 2669 |
+
"loss": 0.2183,
|
| 2670 |
+
"step": 380
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 2.667447306791569,
|
| 2674 |
+
"grad_norm": 0.940924882888794,
|
| 2675 |
+
"learning_rate": 3.5172173586502543e-07,
|
| 2676 |
+
"loss": 0.2297,
|
| 2677 |
+
"step": 381
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 2.6744730679156907,
|
| 2681 |
+
"grad_norm": 0.9774589538574219,
|
| 2682 |
+
"learning_rate": 3.3676789046599045e-07,
|
| 2683 |
+
"loss": 0.28,
|
| 2684 |
+
"step": 382
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 2.6814988290398127,
|
| 2688 |
+
"grad_norm": 0.9266927242279053,
|
| 2689 |
+
"learning_rate": 3.2212779727809504e-07,
|
| 2690 |
+
"loss": 0.2621,
|
| 2691 |
+
"step": 383
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 2.6885245901639343,
|
| 2695 |
+
"grad_norm": 1.0498640537261963,
|
| 2696 |
+
"learning_rate": 3.0780244131806193e-07,
|
| 2697 |
+
"loss": 0.2708,
|
| 2698 |
+
"step": 384
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 2.6955503512880563,
|
| 2702 |
+
"grad_norm": 0.9628897905349731,
|
| 2703 |
+
"learning_rate": 2.937927864264206e-07,
|
| 2704 |
+
"loss": 0.2438,
|
| 2705 |
+
"step": 385
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 2.702576112412178,
|
| 2709 |
+
"grad_norm": 1.0383293628692627,
|
| 2710 |
+
"learning_rate": 2.800997752026596e-07,
|
| 2711 |
+
"loss": 0.2336,
|
| 2712 |
+
"step": 386
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 2.7096018735363,
|
| 2716 |
+
"grad_norm": 0.9352802634239197,
|
| 2717 |
+
"learning_rate": 2.667243289418059e-07,
|
| 2718 |
+
"loss": 0.2046,
|
| 2719 |
+
"step": 387
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 2.7166276346604215,
|
| 2723 |
+
"grad_norm": 0.9982126355171204,
|
| 2724 |
+
"learning_rate": 2.5366734757243496e-07,
|
| 2725 |
+
"loss": 0.2286,
|
| 2726 |
+
"step": 388
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 2.7236533957845435,
|
| 2730 |
+
"grad_norm": 0.8270314931869507,
|
| 2731 |
+
"learning_rate": 2.4092970959612885e-07,
|
| 2732 |
+
"loss": 0.2675,
|
| 2733 |
+
"step": 389
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 2.730679156908665,
|
| 2737 |
+
"grad_norm": 1.0034183263778687,
|
| 2738 |
+
"learning_rate": 2.2851227202836002e-07,
|
| 2739 |
+
"loss": 0.2472,
|
| 2740 |
+
"step": 390
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 2.737704918032787,
|
| 2744 |
+
"grad_norm": 0.860130250453949,
|
| 2745 |
+
"learning_rate": 2.1641587034083756e-07,
|
| 2746 |
+
"loss": 0.2215,
|
| 2747 |
+
"step": 391
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 2.7447306791569086,
|
| 2751 |
+
"grad_norm": 0.8525372743606567,
|
| 2752 |
+
"learning_rate": 2.0464131840528978e-07,
|
| 2753 |
+
"loss": 0.2216,
|
| 2754 |
+
"step": 392
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 2.7517564402810306,
|
| 2758 |
+
"grad_norm": 0.8850567936897278,
|
| 2759 |
+
"learning_rate": 1.9318940843870594e-07,
|
| 2760 |
+
"loss": 0.2656,
|
| 2761 |
+
"step": 393
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 2.758782201405152,
|
| 2765 |
+
"grad_norm": 0.9462560415267944,
|
| 2766 |
+
"learning_rate": 1.8206091095003543e-07,
|
| 2767 |
+
"loss": 0.2344,
|
| 2768 |
+
"step": 394
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 2.7658079625292737,
|
| 2772 |
+
"grad_norm": 0.8583863973617554,
|
| 2773 |
+
"learning_rate": 1.7125657468834656e-07,
|
| 2774 |
+
"loss": 0.2528,
|
| 2775 |
+
"step": 395
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 2.7728337236533958,
|
| 2779 |
+
"grad_norm": 0.9565845131874084,
|
| 2780 |
+
"learning_rate": 1.6077712659244792e-07,
|
| 2781 |
+
"loss": 0.2219,
|
| 2782 |
+
"step": 396
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 2.7798594847775178,
|
| 2786 |
+
"grad_norm": 1.0565779209136963,
|
| 2787 |
+
"learning_rate": 1.5062327174197645e-07,
|
| 2788 |
+
"loss": 0.2828,
|
| 2789 |
+
"step": 397
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 2.7868852459016393,
|
| 2793 |
+
"grad_norm": 0.9745490550994873,
|
| 2794 |
+
"learning_rate": 1.4079569330996412e-07,
|
| 2795 |
+
"loss": 0.2263,
|
| 2796 |
+
"step": 398
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 2.793911007025761,
|
| 2800 |
+
"grad_norm": 0.8957200646400452,
|
| 2801 |
+
"learning_rate": 1.3129505251686603e-07,
|
| 2802 |
+
"loss": 0.2435,
|
| 2803 |
+
"step": 399
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 2.800936768149883,
|
| 2807 |
+
"grad_norm": 0.9497512578964233,
|
| 2808 |
+
"learning_rate": 1.2212198858607694e-07,
|
| 2809 |
+
"loss": 0.2531,
|
| 2810 |
+
"step": 400
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 2.807962529274005,
|
| 2814 |
+
"grad_norm": 0.9695696234703064,
|
| 2815 |
+
"learning_rate": 1.1327711870091963e-07,
|
| 2816 |
+
"loss": 0.2626,
|
| 2817 |
+
"step": 401
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 2.8149882903981265,
|
| 2821 |
+
"grad_norm": 0.9500135183334351,
|
| 2822 |
+
"learning_rate": 1.0476103796312254e-07,
|
| 2823 |
+
"loss": 0.2556,
|
| 2824 |
+
"step": 402
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 2.822014051522248,
|
| 2828 |
+
"grad_norm": 0.8701116442680359,
|
| 2829 |
+
"learning_rate": 9.657431935277629e-08,
|
| 2830 |
+
"loss": 0.245,
|
| 2831 |
+
"step": 403
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 2.82903981264637,
|
| 2835 |
+
"grad_norm": 1.0138100385665894,
|
| 2836 |
+
"learning_rate": 8.871751368978554e-08,
|
| 2837 |
+
"loss": 0.2142,
|
| 2838 |
+
"step": 404
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 2.836065573770492,
|
| 2842 |
+
"grad_norm": 1.0031346082687378,
|
| 2843 |
+
"learning_rate": 8.119114959680929e-08,
|
| 2844 |
+
"loss": 0.2787,
|
| 2845 |
+
"step": 405
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 2.8430913348946136,
|
| 2849 |
+
"grad_norm": 1.0052534341812134,
|
| 2850 |
+
"learning_rate": 7.399573346368871e-08,
|
| 2851 |
+
"loss": 0.2605,
|
| 2852 |
+
"step": 406
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 2.850117096018735,
|
| 2856 |
+
"grad_norm": 0.9534397721290588,
|
| 2857 |
+
"learning_rate": 6.713174941338163e-08,
|
| 2858 |
+
"loss": 0.1808,
|
| 2859 |
+
"step": 407
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 2.857142857142857,
|
| 2863 |
+
"grad_norm": 0.8320339918136597,
|
| 2864 |
+
"learning_rate": 6.05996592693886e-08,
|
| 2865 |
+
"loss": 0.2762,
|
| 2866 |
+
"step": 408
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 2.8641686182669788,
|
| 2870 |
+
"grad_norm": 0.8851465582847595,
|
| 2871 |
+
"learning_rate": 5.439990252467886e-08,
|
| 2872 |
+
"loss": 0.2707,
|
| 2873 |
+
"step": 409
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 2.871194379391101,
|
| 2877 |
+
"grad_norm": 1.0061311721801758,
|
| 2878 |
+
"learning_rate": 4.853289631212066e-08,
|
| 2879 |
+
"loss": 0.2291,
|
| 2880 |
+
"step": 410
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 2.8782201405152223,
|
| 2884 |
+
"grad_norm": 0.8934303522109985,
|
| 2885 |
+
"learning_rate": 4.299903537641703e-08,
|
| 2886 |
+
"loss": 0.2626,
|
| 2887 |
+
"step": 411
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 2.8852459016393444,
|
| 2891 |
+
"grad_norm": 1.021590232849121,
|
| 2892 |
+
"learning_rate": 3.779869204754427e-08,
|
| 2893 |
+
"loss": 0.2321,
|
| 2894 |
+
"step": 412
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 2.892271662763466,
|
| 2898 |
+
"grad_norm": 0.8953171372413635,
|
| 2899 |
+
"learning_rate": 3.2932216215704195e-08,
|
| 2900 |
+
"loss": 0.2318,
|
| 2901 |
+
"step": 413
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 2.899297423887588,
|
| 2905 |
+
"grad_norm": 0.9159654974937439,
|
| 2906 |
+
"learning_rate": 2.8399935307778516e-08,
|
| 2907 |
+
"loss": 0.2313,
|
| 2908 |
+
"step": 414
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 2.9063231850117095,
|
| 2912 |
+
"grad_norm": 0.9971858859062195,
|
| 2913 |
+
"learning_rate": 2.420215426530259e-08,
|
| 2914 |
+
"loss": 0.2179,
|
| 2915 |
+
"step": 415
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 2.9133489461358315,
|
| 2919 |
+
"grad_norm": 0.9292712807655334,
|
| 2920 |
+
"learning_rate": 2.0339155523945164e-08,
|
| 2921 |
+
"loss": 0.2408,
|
| 2922 |
+
"step": 416
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 2.920374707259953,
|
| 2926 |
+
"grad_norm": 0.9429396390914917,
|
| 2927 |
+
"learning_rate": 1.681119899450856e-08,
|
| 2928 |
+
"loss": 0.2313,
|
| 2929 |
+
"step": 417
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 2.927400468384075,
|
| 2933 |
+
"grad_norm": 0.9161145687103271,
|
| 2934 |
+
"learning_rate": 1.3618522045439897e-08,
|
| 2935 |
+
"loss": 0.2418,
|
| 2936 |
+
"step": 418
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 2.9344262295081966,
|
| 2940 |
+
"grad_norm": 0.8718361258506775,
|
| 2941 |
+
"learning_rate": 1.0761339486859424e-08,
|
| 2942 |
+
"loss": 0.1984,
|
| 2943 |
+
"step": 419
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 2.9414519906323187,
|
| 2947 |
+
"grad_norm": 1.0120435953140259,
|
| 2948 |
+
"learning_rate": 8.239843556108739e-09,
|
| 2949 |
+
"loss": 0.2767,
|
| 2950 |
+
"step": 420
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 2.9484777517564402,
|
| 2954 |
+
"grad_norm": 0.9099938273429871,
|
| 2955 |
+
"learning_rate": 6.054203904817812e-09,
|
| 2956 |
+
"loss": 0.2798,
|
| 2957 |
+
"step": 421
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 2.9555035128805622,
|
| 2961 |
+
"grad_norm": 0.9376184344291687,
|
| 2962 |
+
"learning_rate": 4.204567587486885e-09,
|
| 2963 |
+
"loss": 0.2187,
|
| 2964 |
+
"step": 422
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 2.962529274004684,
|
| 2968 |
+
"grad_norm": 0.9267457127571106,
|
| 2969 |
+
"learning_rate": 2.6910590515966117e-09,
|
| 2970 |
+
"loss": 0.2665,
|
| 2971 |
+
"step": 423
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 2.9695550351288054,
|
| 2975 |
+
"grad_norm": 1.0073524713516235,
|
| 2976 |
+
"learning_rate": 1.5137801292325338e-09,
|
| 2977 |
+
"loss": 0.2256,
|
| 2978 |
+
"step": 424
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 2.9765807962529274,
|
| 2982 |
+
"grad_norm": 0.8901485204696655,
|
| 2983 |
+
"learning_rate": 6.728100302327844e-10,
|
| 2984 |
+
"loss": 0.2625,
|
| 2985 |
+
"step": 425
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 2.9836065573770494,
|
| 2989 |
+
"grad_norm": 0.9644309878349304,
|
| 2990 |
+
"learning_rate": 1.6820533686179308e-10,
|
| 2991 |
+
"loss": 0.2242,
|
| 2992 |
+
"step": 426
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 2.9836065573770494,
|
| 2996 |
+
"step": 426,
|
| 2997 |
+
"total_flos": 7.440485693207347e+16,
|
| 2998 |
+
"train_loss": 0.40016700194475235,
|
| 2999 |
+
"train_runtime": 24963.7361,
|
| 3000 |
+
"train_samples_per_second": 0.411,
|
| 3001 |
+
"train_steps_per_second": 0.017
|
| 3002 |
+
}
|
| 3003 |
+
],
|
| 3004 |
+
"logging_steps": 1,
|
| 3005 |
+
"max_steps": 426,
|
| 3006 |
+
"num_input_tokens_seen": 0,
|
| 3007 |
+
"num_train_epochs": 3,
|
| 3008 |
+
"save_steps": 200,
|
| 3009 |
+
"stateful_callbacks": {
|
| 3010 |
+
"TrainerControl": {
|
| 3011 |
+
"args": {
|
| 3012 |
+
"should_epoch_stop": false,
|
| 3013 |
+
"should_evaluate": false,
|
| 3014 |
+
"should_log": false,
|
| 3015 |
+
"should_save": true,
|
| 3016 |
+
"should_training_stop": true
|
| 3017 |
+
},
|
| 3018 |
+
"attributes": {}
|
| 3019 |
+
}
|
| 3020 |
+
},
|
| 3021 |
+
"total_flos": 7.440485693207347e+16,
|
| 3022 |
+
"train_batch_size": 1,
|
| 3023 |
+
"trial_name": null,
|
| 3024 |
+
"trial_params": null
|
| 3025 |
+
}
|
training_loss.png
ADDED
|