End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1408 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-Coder-14B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: SWE-BENCH-loc_500_gen_plus_critic_combine_qwen_code_14B
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# SWE-BENCH-loc_500_gen_plus_critic_combine_qwen_code_14B
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-14B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-Coder-14B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: SWE-BENCH-loc_500_gen_plus_critic_combine_qwen_code_14B
|
|
|
|
| 16 |
|
| 17 |
# SWE-BENCH-loc_500_gen_plus_critic_combine_qwen_code_14B
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-14B-Instruct) on the SWE-BENCH-loc_500_gen_plus_critic_combine dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.986093552465234,
|
| 3 |
+
"total_flos": 2.193163794073518e+17,
|
| 4 |
+
"train_loss": 0.5202219234063075,
|
| 5 |
+
"train_runtime": 39110.2979,
|
| 6 |
+
"train_samples_per_second": 0.243,
|
| 7 |
+
"train_steps_per_second": 0.005
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.986093552465234,
|
| 3 |
+
"total_flos": 2.193163794073518e+17,
|
| 4 |
+
"train_loss": 0.5202219234063075,
|
| 5 |
+
"train_runtime": 39110.2979,
|
| 6 |
+
"train_samples_per_second": 0.243,
|
| 7 |
+
"train_steps_per_second": 0.005
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1408 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.986093552465234,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 195,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.015170670037926675,
|
| 14 |
+
"grad_norm": 1.3418480157852173,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.7433,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.03034134007585335,
|
| 21 |
+
"grad_norm": 1.4127129316329956,
|
| 22 |
+
"learning_rate": 5.000000000000001e-07,
|
| 23 |
+
"loss": 0.761,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.045512010113780026,
|
| 28 |
+
"grad_norm": 1.5653902292251587,
|
| 29 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 30 |
+
"loss": 0.8263,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0606826801517067,
|
| 35 |
+
"grad_norm": 1.3309255838394165,
|
| 36 |
+
"learning_rate": 1.5e-06,
|
| 37 |
+
"loss": 0.7565,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07585335018963338,
|
| 42 |
+
"grad_norm": 1.2815498113632202,
|
| 43 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 44 |
+
"loss": 0.7118,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.09102402022756005,
|
| 49 |
+
"grad_norm": 1.3161976337432861,
|
| 50 |
+
"learning_rate": 2.5e-06,
|
| 51 |
+
"loss": 0.7609,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.10619469026548672,
|
| 56 |
+
"grad_norm": 1.283697485923767,
|
| 57 |
+
"learning_rate": 3e-06,
|
| 58 |
+
"loss": 0.8174,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1213653603034134,
|
| 63 |
+
"grad_norm": 1.020988941192627,
|
| 64 |
+
"learning_rate": 3.5e-06,
|
| 65 |
+
"loss": 0.7995,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.13653603034134007,
|
| 70 |
+
"grad_norm": 0.8449975252151489,
|
| 71 |
+
"learning_rate": 4.000000000000001e-06,
|
| 72 |
+
"loss": 0.7482,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.15170670037926676,
|
| 77 |
+
"grad_norm": 0.9625175595283508,
|
| 78 |
+
"learning_rate": 4.5e-06,
|
| 79 |
+
"loss": 0.7357,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16687737041719342,
|
| 84 |
+
"grad_norm": 1.0360528230667114,
|
| 85 |
+
"learning_rate": 5e-06,
|
| 86 |
+
"loss": 0.7524,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.1820480404551201,
|
| 91 |
+
"grad_norm": 0.8166671395301819,
|
| 92 |
+
"learning_rate": 5.500000000000001e-06,
|
| 93 |
+
"loss": 0.6702,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.1972187104930468,
|
| 98 |
+
"grad_norm": 0.669105052947998,
|
| 99 |
+
"learning_rate": 6e-06,
|
| 100 |
+
"loss": 0.6509,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.21238938053097345,
|
| 105 |
+
"grad_norm": 0.9122142791748047,
|
| 106 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 107 |
+
"loss": 0.7482,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.22756005056890014,
|
| 112 |
+
"grad_norm": 0.8461022973060608,
|
| 113 |
+
"learning_rate": 7e-06,
|
| 114 |
+
"loss": 0.6602,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.2427307206068268,
|
| 119 |
+
"grad_norm": 1.27907133102417,
|
| 120 |
+
"learning_rate": 7.500000000000001e-06,
|
| 121 |
+
"loss": 0.7036,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.2579013906447535,
|
| 126 |
+
"grad_norm": 0.7009077072143555,
|
| 127 |
+
"learning_rate": 8.000000000000001e-06,
|
| 128 |
+
"loss": 0.6934,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.27307206068268014,
|
| 133 |
+
"grad_norm": 0.6173739433288574,
|
| 134 |
+
"learning_rate": 8.5e-06,
|
| 135 |
+
"loss": 0.6572,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.28824273072060685,
|
| 140 |
+
"grad_norm": 0.6167724132537842,
|
| 141 |
+
"learning_rate": 9e-06,
|
| 142 |
+
"loss": 0.646,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.3034134007585335,
|
| 147 |
+
"grad_norm": 0.7092210054397583,
|
| 148 |
+
"learning_rate": 9.5e-06,
|
| 149 |
+
"loss": 0.7002,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.3185840707964602,
|
| 154 |
+
"grad_norm": 0.6299468874931335,
|
| 155 |
+
"learning_rate": 1e-05,
|
| 156 |
+
"loss": 0.6522,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.33375474083438683,
|
| 161 |
+
"grad_norm": 0.5548372268676758,
|
| 162 |
+
"learning_rate": 9.999194339645292e-06,
|
| 163 |
+
"loss": 0.5982,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.34892541087231355,
|
| 168 |
+
"grad_norm": 0.5452980399131775,
|
| 169 |
+
"learning_rate": 9.996777618216608e-06,
|
| 170 |
+
"loss": 0.6258,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.3640960809102402,
|
| 175 |
+
"grad_norm": 0.549773097038269,
|
| 176 |
+
"learning_rate": 9.992750614536606e-06,
|
| 177 |
+
"loss": 0.6204,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.37926675094816686,
|
| 182 |
+
"grad_norm": 0.5395023822784424,
|
| 183 |
+
"learning_rate": 9.987114626364172e-06,
|
| 184 |
+
"loss": 0.6947,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.3944374209860936,
|
| 189 |
+
"grad_norm": 0.49335014820098877,
|
| 190 |
+
"learning_rate": 9.979871469976197e-06,
|
| 191 |
+
"loss": 0.6215,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.40960809102402024,
|
| 196 |
+
"grad_norm": 0.5002505779266357,
|
| 197 |
+
"learning_rate": 9.971023479582258e-06,
|
| 198 |
+
"loss": 0.5899,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.4247787610619469,
|
| 203 |
+
"grad_norm": 0.4934206008911133,
|
| 204 |
+
"learning_rate": 9.960573506572391e-06,
|
| 205 |
+
"loss": 0.6689,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.43994943109987356,
|
| 210 |
+
"grad_norm": 0.4559570848941803,
|
| 211 |
+
"learning_rate": 9.948524918598175e-06,
|
| 212 |
+
"loss": 0.6031,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.45512010113780027,
|
| 217 |
+
"grad_norm": 0.4254303276538849,
|
| 218 |
+
"learning_rate": 9.934881598487478e-06,
|
| 219 |
+
"loss": 0.5164,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.47029077117572693,
|
| 224 |
+
"grad_norm": 0.511420726776123,
|
| 225 |
+
"learning_rate": 9.91964794299315e-06,
|
| 226 |
+
"loss": 0.6158,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4854614412136536,
|
| 231 |
+
"grad_norm": 0.4833051264286041,
|
| 232 |
+
"learning_rate": 9.902828861376101e-06,
|
| 233 |
+
"loss": 0.6044,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.5006321112515802,
|
| 238 |
+
"grad_norm": 0.631668746471405,
|
| 239 |
+
"learning_rate": 9.884429773823238e-06,
|
| 240 |
+
"loss": 0.5945,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.515802781289507,
|
| 245 |
+
"grad_norm": 0.5820228457450867,
|
| 246 |
+
"learning_rate": 9.864456609700726e-06,
|
| 247 |
+
"loss": 0.595,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5309734513274337,
|
| 252 |
+
"grad_norm": 0.4874587655067444,
|
| 253 |
+
"learning_rate": 9.842915805643156e-06,
|
| 254 |
+
"loss": 0.5832,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.5461441213653603,
|
| 259 |
+
"grad_norm": 0.4421619176864624,
|
| 260 |
+
"learning_rate": 9.819814303479268e-06,
|
| 261 |
+
"loss": 0.5983,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.561314791403287,
|
| 266 |
+
"grad_norm": 0.47112926840782166,
|
| 267 |
+
"learning_rate": 9.79515954799483e-06,
|
| 268 |
+
"loss": 0.6261,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.5764854614412137,
|
| 273 |
+
"grad_norm": 0.4074913561344147,
|
| 274 |
+
"learning_rate": 9.768959484533461e-06,
|
| 275 |
+
"loss": 0.5501,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.5916561314791403,
|
| 280 |
+
"grad_norm": 0.46934953331947327,
|
| 281 |
+
"learning_rate": 9.741222556436132e-06,
|
| 282 |
+
"loss": 0.6143,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.606826801517067,
|
| 287 |
+
"grad_norm": 0.470877468585968,
|
| 288 |
+
"learning_rate": 9.711957702320176e-06,
|
| 289 |
+
"loss": 0.5895,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.6219974715549936,
|
| 294 |
+
"grad_norm": 0.5042617321014404,
|
| 295 |
+
"learning_rate": 9.681174353198687e-06,
|
| 296 |
+
"loss": 0.6084,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.6371681415929203,
|
| 301 |
+
"grad_norm": 0.4102831780910492,
|
| 302 |
+
"learning_rate": 9.648882429441258e-06,
|
| 303 |
+
"loss": 0.5517,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.6523388116308471,
|
| 308 |
+
"grad_norm": 0.41832295060157776,
|
| 309 |
+
"learning_rate": 9.615092337576987e-06,
|
| 310 |
+
"loss": 0.5322,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.6675094816687737,
|
| 315 |
+
"grad_norm": 0.4671192467212677,
|
| 316 |
+
"learning_rate": 9.579814966940833e-06,
|
| 317 |
+
"loss": 0.614,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6826801517067004,
|
| 322 |
+
"grad_norm": 0.45350876450538635,
|
| 323 |
+
"learning_rate": 9.543061686164374e-06,
|
| 324 |
+
"loss": 0.6005,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.6978508217446271,
|
| 329 |
+
"grad_norm": 0.46615591645240784,
|
| 330 |
+
"learning_rate": 9.504844339512096e-06,
|
| 331 |
+
"loss": 0.6009,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.7130214917825537,
|
| 336 |
+
"grad_norm": 0.5051223635673523,
|
| 337 |
+
"learning_rate": 9.465175243064428e-06,
|
| 338 |
+
"loss": 0.589,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.7281921618204804,
|
| 343 |
+
"grad_norm": 0.4288654923439026,
|
| 344 |
+
"learning_rate": 9.424067180748692e-06,
|
| 345 |
+
"loss": 0.562,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.7433628318584071,
|
| 350 |
+
"grad_norm": 0.4421122372150421,
|
| 351 |
+
"learning_rate": 9.381533400219319e-06,
|
| 352 |
+
"loss": 0.6247,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.7585335018963337,
|
| 357 |
+
"grad_norm": 0.44172200560569763,
|
| 358 |
+
"learning_rate": 9.337587608588588e-06,
|
| 359 |
+
"loss": 0.6327,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.7737041719342604,
|
| 364 |
+
"grad_norm": 0.43877238035202026,
|
| 365 |
+
"learning_rate": 9.292243968009332e-06,
|
| 366 |
+
"loss": 0.5905,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.7888748419721872,
|
| 371 |
+
"grad_norm": 0.48638173937797546,
|
| 372 |
+
"learning_rate": 9.24551709111097e-06,
|
| 373 |
+
"loss": 0.5787,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.8040455120101138,
|
| 378 |
+
"grad_norm": 0.46385085582733154,
|
| 379 |
+
"learning_rate": 9.197422036290386e-06,
|
| 380 |
+
"loss": 0.6414,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.8192161820480405,
|
| 385 |
+
"grad_norm": 0.47493401169776917,
|
| 386 |
+
"learning_rate": 9.147974302859158e-06,
|
| 387 |
+
"loss": 0.6136,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.8343868520859671,
|
| 392 |
+
"grad_norm": 0.4470069110393524,
|
| 393 |
+
"learning_rate": 9.09718982604866e-06,
|
| 394 |
+
"loss": 0.5687,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.8495575221238938,
|
| 399 |
+
"grad_norm": 0.41280269622802734,
|
| 400 |
+
"learning_rate": 9.045084971874738e-06,
|
| 401 |
+
"loss": 0.5977,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.8647281921618205,
|
| 406 |
+
"grad_norm": 0.4971402883529663,
|
| 407 |
+
"learning_rate": 8.991676531863507e-06,
|
| 408 |
+
"loss": 0.5718,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.8798988621997471,
|
| 413 |
+
"grad_norm": 0.46412622928619385,
|
| 414 |
+
"learning_rate": 8.936981717640061e-06,
|
| 415 |
+
"loss": 0.5713,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.8950695322376738,
|
| 420 |
+
"grad_norm": 0.4311119318008423,
|
| 421 |
+
"learning_rate": 8.881018155381766e-06,
|
| 422 |
+
"loss": 0.6296,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.9102402022756005,
|
| 427 |
+
"grad_norm": 0.45351457595825195,
|
| 428 |
+
"learning_rate": 8.823803880137993e-06,
|
| 429 |
+
"loss": 0.567,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.9254108723135271,
|
| 434 |
+
"grad_norm": 0.4308796226978302,
|
| 435 |
+
"learning_rate": 8.765357330018056e-06,
|
| 436 |
+
"loss": 0.5658,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.9405815423514539,
|
| 441 |
+
"grad_norm": 0.4713445007801056,
|
| 442 |
+
"learning_rate": 8.705697340249275e-06,
|
| 443 |
+
"loss": 0.5758,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.9557522123893806,
|
| 448 |
+
"grad_norm": 0.49278339743614197,
|
| 449 |
+
"learning_rate": 8.644843137107058e-06,
|
| 450 |
+
"loss": 0.5843,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.9709228824273072,
|
| 455 |
+
"grad_norm": 0.44526898860931396,
|
| 456 |
+
"learning_rate": 8.582814331718961e-06,
|
| 457 |
+
"loss": 0.5723,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.9860935524652339,
|
| 462 |
+
"grad_norm": 0.46161067485809326,
|
| 463 |
+
"learning_rate": 8.519630913744726e-06,
|
| 464 |
+
"loss": 0.6162,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.0151706700379266,
|
| 469 |
+
"grad_norm": 0.6008106470108032,
|
| 470 |
+
"learning_rate": 8.455313244934324e-06,
|
| 471 |
+
"loss": 1.1143,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.0303413400758534,
|
| 476 |
+
"grad_norm": 0.4400518834590912,
|
| 477 |
+
"learning_rate": 8.389882052566106e-06,
|
| 478 |
+
"loss": 0.4743,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.04551201011378,
|
| 483 |
+
"grad_norm": 0.403044193983078,
|
| 484 |
+
"learning_rate": 8.32335842276713e-06,
|
| 485 |
+
"loss": 0.483,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.0606826801517066,
|
| 490 |
+
"grad_norm": 1.6882063150405884,
|
| 491 |
+
"learning_rate": 8.255763793717868e-06,
|
| 492 |
+
"loss": 0.5228,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.0758533501896335,
|
| 497 |
+
"grad_norm": 0.5953236818313599,
|
| 498 |
+
"learning_rate": 8.18711994874345e-06,
|
| 499 |
+
"loss": 0.4833,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.09102402022756,
|
| 504 |
+
"grad_norm": 1.2095857858657837,
|
| 505 |
+
"learning_rate": 8.117449009293668e-06,
|
| 506 |
+
"loss": 0.5248,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.1061946902654867,
|
| 511 |
+
"grad_norm": 0.43588605523109436,
|
| 512 |
+
"learning_rate": 8.046773427814043e-06,
|
| 513 |
+
"loss": 0.4946,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.1213653603034135,
|
| 518 |
+
"grad_norm": 0.45902812480926514,
|
| 519 |
+
"learning_rate": 7.975115980510187e-06,
|
| 520 |
+
"loss": 0.5118,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.13653603034134,
|
| 525 |
+
"grad_norm": 0.43903180956840515,
|
| 526 |
+
"learning_rate": 7.902499760007867e-06,
|
| 527 |
+
"loss": 0.5132,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.1517067003792667,
|
| 532 |
+
"grad_norm": 0.4341452717781067,
|
| 533 |
+
"learning_rate": 7.828948167911073e-06,
|
| 534 |
+
"loss": 0.51,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.1668773704171933,
|
| 539 |
+
"grad_norm": 0.4682852029800415,
|
| 540 |
+
"learning_rate": 7.754484907260513e-06,
|
| 541 |
+
"loss": 0.4863,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.1820480404551201,
|
| 546 |
+
"grad_norm": 0.4794589579105377,
|
| 547 |
+
"learning_rate": 7.679133974894984e-06,
|
| 548 |
+
"loss": 0.5083,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.1972187104930467,
|
| 553 |
+
"grad_norm": 0.4567174017429352,
|
| 554 |
+
"learning_rate": 7.602919653718044e-06,
|
| 555 |
+
"loss": 0.5121,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.2123893805309733,
|
| 560 |
+
"grad_norm": 0.45591986179351807,
|
| 561 |
+
"learning_rate": 7.5258665048725065e-06,
|
| 562 |
+
"loss": 0.4804,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.2275600505689002,
|
| 567 |
+
"grad_norm": 0.48482567071914673,
|
| 568 |
+
"learning_rate": 7.447999359825263e-06,
|
| 569 |
+
"loss": 0.4756,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.2427307206068268,
|
| 574 |
+
"grad_norm": 0.44032907485961914,
|
| 575 |
+
"learning_rate": 7.369343312364994e-06,
|
| 576 |
+
"loss": 0.4748,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.2579013906447534,
|
| 581 |
+
"grad_norm": 0.46281248331069946,
|
| 582 |
+
"learning_rate": 7.289923710515338e-06,
|
| 583 |
+
"loss": 0.4969,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.2730720606826802,
|
| 588 |
+
"grad_norm": 0.4520440995693207,
|
| 589 |
+
"learning_rate": 7.2097661483661355e-06,
|
| 590 |
+
"loss": 0.5173,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.2882427307206068,
|
| 595 |
+
"grad_norm": 0.7100924849510193,
|
| 596 |
+
"learning_rate": 7.128896457825364e-06,
|
| 597 |
+
"loss": 0.5154,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.3034134007585334,
|
| 602 |
+
"grad_norm": 0.4339788854122162,
|
| 603 |
+
"learning_rate": 7.047340700294454e-06,
|
| 604 |
+
"loss": 0.4539,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.3185840707964602,
|
| 609 |
+
"grad_norm": 0.5082290172576904,
|
| 610 |
+
"learning_rate": 6.965125158269619e-06,
|
| 611 |
+
"loss": 0.5475,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.3337547408343868,
|
| 616 |
+
"grad_norm": 0.5252617001533508,
|
| 617 |
+
"learning_rate": 6.88227632687196e-06,
|
| 618 |
+
"loss": 0.4785,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.3489254108723134,
|
| 623 |
+
"grad_norm": 0.46251779794692993,
|
| 624 |
+
"learning_rate": 6.798820905309036e-06,
|
| 625 |
+
"loss": 0.4832,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.3640960809102403,
|
| 630 |
+
"grad_norm": 0.44232895970344543,
|
| 631 |
+
"learning_rate": 6.714785788270658e-06,
|
| 632 |
+
"loss": 0.4991,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.3792667509481669,
|
| 637 |
+
"grad_norm": 0.5520949959754944,
|
| 638 |
+
"learning_rate": 6.63019805726171e-06,
|
| 639 |
+
"loss": 0.5147,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.3944374209860935,
|
| 644 |
+
"grad_norm": 0.4751013517379761,
|
| 645 |
+
"learning_rate": 6.545084971874738e-06,
|
| 646 |
+
"loss": 0.468,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.4096080910240203,
|
| 651 |
+
"grad_norm": 0.6363617777824402,
|
| 652 |
+
"learning_rate": 6.459473961005168e-06,
|
| 653 |
+
"loss": 0.4465,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.424778761061947,
|
| 658 |
+
"grad_norm": 0.504291832447052,
|
| 659 |
+
"learning_rate": 6.373392614011952e-06,
|
| 660 |
+
"loss": 0.4599,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.4399494310998735,
|
| 665 |
+
"grad_norm": 0.5495200157165527,
|
| 666 |
+
"learning_rate": 6.286868671826513e-06,
|
| 667 |
+
"loss": 0.4809,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.4551201011378003,
|
| 672 |
+
"grad_norm": 0.45413124561309814,
|
| 673 |
+
"learning_rate": 6.19993001801283e-06,
|
| 674 |
+
"loss": 0.4756,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.470290771175727,
|
| 679 |
+
"grad_norm": 0.48508644104003906,
|
| 680 |
+
"learning_rate": 6.112604669781572e-06,
|
| 681 |
+
"loss": 0.4996,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.4854614412136535,
|
| 686 |
+
"grad_norm": 0.670592725276947,
|
| 687 |
+
"learning_rate": 6.024920768961153e-06,
|
| 688 |
+
"loss": 0.5203,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.5006321112515804,
|
| 693 |
+
"grad_norm": 0.43420279026031494,
|
| 694 |
+
"learning_rate": 5.936906572928625e-06,
|
| 695 |
+
"loss": 0.5112,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.515802781289507,
|
| 700 |
+
"grad_norm": 0.4348045885562897,
|
| 701 |
+
"learning_rate": 5.848590445503345e-06,
|
| 702 |
+
"loss": 0.5077,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.5309734513274336,
|
| 707 |
+
"grad_norm": 0.5098991990089417,
|
| 708 |
+
"learning_rate": 5.760000847806337e-06,
|
| 709 |
+
"loss": 0.5244,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.5461441213653604,
|
| 714 |
+
"grad_norm": 0.4518718123435974,
|
| 715 |
+
"learning_rate": 5.671166329088278e-06,
|
| 716 |
+
"loss": 0.4814,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.561314791403287,
|
| 721 |
+
"grad_norm": 0.4843664765357971,
|
| 722 |
+
"learning_rate": 5.582115517529114e-06,
|
| 723 |
+
"loss": 0.4969,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.5764854614412136,
|
| 728 |
+
"grad_norm": 0.5018794536590576,
|
| 729 |
+
"learning_rate": 5.4928771110122185e-06,
|
| 730 |
+
"loss": 0.4738,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.5916561314791404,
|
| 735 |
+
"grad_norm": 0.4358893930912018,
|
| 736 |
+
"learning_rate": 5.403479867876087e-06,
|
| 737 |
+
"loss": 0.4958,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.606826801517067,
|
| 742 |
+
"grad_norm": 0.45651909708976746,
|
| 743 |
+
"learning_rate": 5.3139525976465675e-06,
|
| 744 |
+
"loss": 0.446,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.6219974715549936,
|
| 749 |
+
"grad_norm": 0.4391672909259796,
|
| 750 |
+
"learning_rate": 5.224324151752575e-06,
|
| 751 |
+
"loss": 0.4789,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.6371681415929205,
|
| 756 |
+
"grad_norm": 0.40504834055900574,
|
| 757 |
+
"learning_rate": 5.134623414228315e-06,
|
| 758 |
+
"loss": 0.4801,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.652338811630847,
|
| 763 |
+
"grad_norm": 0.425678551197052,
|
| 764 |
+
"learning_rate": 5.04487929240499e-06,
|
| 765 |
+
"loss": 0.5132,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.6675094816687737,
|
| 770 |
+
"grad_norm": 0.4834710657596588,
|
| 771 |
+
"learning_rate": 4.955120707595011e-06,
|
| 772 |
+
"loss": 0.4529,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.6826801517067005,
|
| 777 |
+
"grad_norm": 0.3974740207195282,
|
| 778 |
+
"learning_rate": 4.865376585771687e-06,
|
| 779 |
+
"loss": 0.4207,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.697850821744627,
|
| 784 |
+
"grad_norm": 0.39827069640159607,
|
| 785 |
+
"learning_rate": 4.775675848247427e-06,
|
| 786 |
+
"loss": 0.4408,
|
| 787 |
+
"step": 111
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.7130214917825537,
|
| 791 |
+
"grad_norm": 0.45671942830085754,
|
| 792 |
+
"learning_rate": 4.686047402353433e-06,
|
| 793 |
+
"loss": 0.4526,
|
| 794 |
+
"step": 112
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.7281921618204805,
|
| 798 |
+
"grad_norm": 0.46384182572364807,
|
| 799 |
+
"learning_rate": 4.596520132123915e-06,
|
| 800 |
+
"loss": 0.4952,
|
| 801 |
+
"step": 113
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.7433628318584071,
|
| 805 |
+
"grad_norm": 0.5297421813011169,
|
| 806 |
+
"learning_rate": 4.507122888987782e-06,
|
| 807 |
+
"loss": 0.514,
|
| 808 |
+
"step": 114
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.7585335018963337,
|
| 812 |
+
"grad_norm": 0.4392913579940796,
|
| 813 |
+
"learning_rate": 4.417884482470887e-06,
|
| 814 |
+
"loss": 0.4494,
|
| 815 |
+
"step": 115
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.7737041719342606,
|
| 819 |
+
"grad_norm": 0.44901785254478455,
|
| 820 |
+
"learning_rate": 4.3288336709117246e-06,
|
| 821 |
+
"loss": 0.4812,
|
| 822 |
+
"step": 116
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.7888748419721872,
|
| 826 |
+
"grad_norm": 0.43553972244262695,
|
| 827 |
+
"learning_rate": 4.239999152193664e-06,
|
| 828 |
+
"loss": 0.5324,
|
| 829 |
+
"step": 117
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.8040455120101138,
|
| 833 |
+
"grad_norm": 0.47039350867271423,
|
| 834 |
+
"learning_rate": 4.1514095544966556e-06,
|
| 835 |
+
"loss": 0.4907,
|
| 836 |
+
"step": 118
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.8192161820480406,
|
| 840 |
+
"grad_norm": 0.45996445417404175,
|
| 841 |
+
"learning_rate": 4.063093427071376e-06,
|
| 842 |
+
"loss": 0.53,
|
| 843 |
+
"step": 119
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.834386852085967,
|
| 847 |
+
"grad_norm": 0.4866219460964203,
|
| 848 |
+
"learning_rate": 3.975079231038848e-06,
|
| 849 |
+
"loss": 0.4665,
|
| 850 |
+
"step": 120
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.8495575221238938,
|
| 854 |
+
"grad_norm": 0.4342859387397766,
|
| 855 |
+
"learning_rate": 3.887395330218429e-06,
|
| 856 |
+
"loss": 0.5168,
|
| 857 |
+
"step": 121
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.8647281921618206,
|
| 861 |
+
"grad_norm": 0.45591673254966736,
|
| 862 |
+
"learning_rate": 3.8000699819871704e-06,
|
| 863 |
+
"loss": 0.5011,
|
| 864 |
+
"step": 122
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.879898862199747,
|
| 868 |
+
"grad_norm": 0.40730446577072144,
|
| 869 |
+
"learning_rate": 3.7131313281734895e-06,
|
| 870 |
+
"loss": 0.4997,
|
| 871 |
+
"step": 123
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.8950695322376738,
|
| 875 |
+
"grad_norm": 0.43347927927970886,
|
| 876 |
+
"learning_rate": 3.62660738598805e-06,
|
| 877 |
+
"loss": 0.4689,
|
| 878 |
+
"step": 124
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.9102402022756007,
|
| 882 |
+
"grad_norm": 0.4671528935432434,
|
| 883 |
+
"learning_rate": 3.540526038994834e-06,
|
| 884 |
+
"loss": 0.5054,
|
| 885 |
+
"step": 125
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.925410872313527,
|
| 889 |
+
"grad_norm": 0.47119274735450745,
|
| 890 |
+
"learning_rate": 3.4549150281252635e-06,
|
| 891 |
+
"loss": 0.4463,
|
| 892 |
+
"step": 126
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.9405815423514539,
|
| 896 |
+
"grad_norm": 0.46681082248687744,
|
| 897 |
+
"learning_rate": 3.3698019427382912e-06,
|
| 898 |
+
"loss": 0.4955,
|
| 899 |
+
"step": 127
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.9557522123893807,
|
| 903 |
+
"grad_norm": 0.46624964475631714,
|
| 904 |
+
"learning_rate": 3.2852142117293435e-06,
|
| 905 |
+
"loss": 0.4706,
|
| 906 |
+
"step": 128
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.970922882427307,
|
| 910 |
+
"grad_norm": 0.48810526728630066,
|
| 911 |
+
"learning_rate": 3.2011790946909673e-06,
|
| 912 |
+
"loss": 0.5306,
|
| 913 |
+
"step": 129
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.986093552465234,
|
| 917 |
+
"grad_norm": 0.49850064516067505,
|
| 918 |
+
"learning_rate": 3.11772367312804e-06,
|
| 919 |
+
"loss": 0.4863,
|
| 920 |
+
"step": 130
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.015170670037927,
|
| 924 |
+
"grad_norm": 0.5622653961181641,
|
| 925 |
+
"learning_rate": 3.0348748417303826e-06,
|
| 926 |
+
"loss": 0.9358,
|
| 927 |
+
"step": 131
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.030341340075853,
|
| 931 |
+
"grad_norm": 0.4461542069911957,
|
| 932 |
+
"learning_rate": 2.9526592997055488e-06,
|
| 933 |
+
"loss": 0.4256,
|
| 934 |
+
"step": 132
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.04551201011378,
|
| 938 |
+
"grad_norm": 0.4193766713142395,
|
| 939 |
+
"learning_rate": 2.871103542174637e-06,
|
| 940 |
+
"loss": 0.3838,
|
| 941 |
+
"step": 133
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.060682680151707,
|
| 945 |
+
"grad_norm": 0.47868236899375916,
|
| 946 |
+
"learning_rate": 2.790233851633868e-06,
|
| 947 |
+
"loss": 0.4796,
|
| 948 |
+
"step": 134
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.0758533501896332,
|
| 952 |
+
"grad_norm": 0.4564484655857086,
|
| 953 |
+
"learning_rate": 2.7100762894846633e-06,
|
| 954 |
+
"loss": 0.4175,
|
| 955 |
+
"step": 135
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.09102402022756,
|
| 959 |
+
"grad_norm": 0.42239049077033997,
|
| 960 |
+
"learning_rate": 2.6306566876350072e-06,
|
| 961 |
+
"loss": 0.4059,
|
| 962 |
+
"step": 136
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.106194690265487,
|
| 966 |
+
"grad_norm": 0.4516836404800415,
|
| 967 |
+
"learning_rate": 2.55200064017474e-06,
|
| 968 |
+
"loss": 0.4155,
|
| 969 |
+
"step": 137
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.1213653603034133,
|
| 973 |
+
"grad_norm": 0.41927093267440796,
|
| 974 |
+
"learning_rate": 2.4741334951274948e-06,
|
| 975 |
+
"loss": 0.4085,
|
| 976 |
+
"step": 138
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.13653603034134,
|
| 980 |
+
"grad_norm": 0.42958956956863403,
|
| 981 |
+
"learning_rate": 2.3970803462819586e-06,
|
| 982 |
+
"loss": 0.413,
|
| 983 |
+
"step": 139
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.151706700379267,
|
| 987 |
+
"grad_norm": 0.43164947628974915,
|
| 988 |
+
"learning_rate": 2.320866025105016e-06,
|
| 989 |
+
"loss": 0.3982,
|
| 990 |
+
"step": 140
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.1668773704171933,
|
| 994 |
+
"grad_norm": 0.4566631317138672,
|
| 995 |
+
"learning_rate": 2.245515092739488e-06,
|
| 996 |
+
"loss": 0.4189,
|
| 997 |
+
"step": 141
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.18204804045512,
|
| 1001 |
+
"grad_norm": 0.4045684039592743,
|
| 1002 |
+
"learning_rate": 2.171051832088928e-06,
|
| 1003 |
+
"loss": 0.4274,
|
| 1004 |
+
"step": 142
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.197218710493047,
|
| 1008 |
+
"grad_norm": 0.46526044607162476,
|
| 1009 |
+
"learning_rate": 2.097500239992132e-06,
|
| 1010 |
+
"loss": 0.4309,
|
| 1011 |
+
"step": 143
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.2123893805309733,
|
| 1015 |
+
"grad_norm": 0.4466479420661926,
|
| 1016 |
+
"learning_rate": 2.0248840194898155e-06,
|
| 1017 |
+
"loss": 0.428,
|
| 1018 |
+
"step": 144
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.2275600505689,
|
| 1022 |
+
"grad_norm": 0.4933473765850067,
|
| 1023 |
+
"learning_rate": 1.95322657218596e-06,
|
| 1024 |
+
"loss": 0.4206,
|
| 1025 |
+
"step": 145
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.242730720606827,
|
| 1029 |
+
"grad_norm": 0.45855018496513367,
|
| 1030 |
+
"learning_rate": 1.8825509907063328e-06,
|
| 1031 |
+
"loss": 0.4085,
|
| 1032 |
+
"step": 146
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.2579013906447534,
|
| 1036 |
+
"grad_norm": 0.4341532588005066,
|
| 1037 |
+
"learning_rate": 1.8128800512565514e-06,
|
| 1038 |
+
"loss": 0.3596,
|
| 1039 |
+
"step": 147
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.27307206068268,
|
| 1043 |
+
"grad_norm": 0.4618159234523773,
|
| 1044 |
+
"learning_rate": 1.7442362062821323e-06,
|
| 1045 |
+
"loss": 0.4303,
|
| 1046 |
+
"step": 148
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.288242730720607,
|
| 1050 |
+
"grad_norm": 0.523358166217804,
|
| 1051 |
+
"learning_rate": 1.6766415772328732e-06,
|
| 1052 |
+
"loss": 0.4721,
|
| 1053 |
+
"step": 149
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.3034134007585334,
|
| 1057 |
+
"grad_norm": 0.47894734144210815,
|
| 1058 |
+
"learning_rate": 1.610117947433897e-06,
|
| 1059 |
+
"loss": 0.3937,
|
| 1060 |
+
"step": 150
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.3185840707964602,
|
| 1064 |
+
"grad_norm": 0.4730558395385742,
|
| 1065 |
+
"learning_rate": 1.544686755065677e-06,
|
| 1066 |
+
"loss": 0.4465,
|
| 1067 |
+
"step": 151
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.3337547408343866,
|
| 1071 |
+
"grad_norm": 0.5242443680763245,
|
| 1072 |
+
"learning_rate": 1.4803690862552755e-06,
|
| 1073 |
+
"loss": 0.4034,
|
| 1074 |
+
"step": 152
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.3489254108723134,
|
| 1078 |
+
"grad_norm": 0.44352981448173523,
|
| 1079 |
+
"learning_rate": 1.4171856682810386e-06,
|
| 1080 |
+
"loss": 0.3969,
|
| 1081 |
+
"step": 153
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.3640960809102403,
|
| 1085 |
+
"grad_norm": 0.45506805181503296,
|
| 1086 |
+
"learning_rate": 1.3551568628929434e-06,
|
| 1087 |
+
"loss": 0.421,
|
| 1088 |
+
"step": 154
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.379266750948167,
|
| 1092 |
+
"grad_norm": 0.45182856917381287,
|
| 1093 |
+
"learning_rate": 1.2943026597507268e-06,
|
| 1094 |
+
"loss": 0.4364,
|
| 1095 |
+
"step": 155
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.3944374209860935,
|
| 1099 |
+
"grad_norm": 0.5200023055076599,
|
| 1100 |
+
"learning_rate": 1.234642669981946e-06,
|
| 1101 |
+
"loss": 0.4293,
|
| 1102 |
+
"step": 156
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.4096080910240203,
|
| 1106 |
+
"grad_norm": 0.5059816837310791,
|
| 1107 |
+
"learning_rate": 1.1761961198620081e-06,
|
| 1108 |
+
"loss": 0.4273,
|
| 1109 |
+
"step": 157
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.4247787610619467,
|
| 1113 |
+
"grad_norm": 0.5020045638084412,
|
| 1114 |
+
"learning_rate": 1.118981844618236e-06,
|
| 1115 |
+
"loss": 0.4646,
|
| 1116 |
+
"step": 158
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.4399494310998735,
|
| 1120 |
+
"grad_norm": 0.4282710552215576,
|
| 1121 |
+
"learning_rate": 1.06301828235994e-06,
|
| 1122 |
+
"loss": 0.3575,
|
| 1123 |
+
"step": 159
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.4551201011378003,
|
| 1127 |
+
"grad_norm": 0.4831399619579315,
|
| 1128 |
+
"learning_rate": 1.0083234681364934e-06,
|
| 1129 |
+
"loss": 0.4138,
|
| 1130 |
+
"step": 160
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.470290771175727,
|
| 1134 |
+
"grad_norm": 0.4688436985015869,
|
| 1135 |
+
"learning_rate": 9.549150281252633e-07,
|
| 1136 |
+
"loss": 0.3955,
|
| 1137 |
+
"step": 161
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.4854614412136535,
|
| 1141 |
+
"grad_norm": 0.46004679799079895,
|
| 1142 |
+
"learning_rate": 9.028101739513406e-07,
|
| 1143 |
+
"loss": 0.4322,
|
| 1144 |
+
"step": 162
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.5006321112515804,
|
| 1148 |
+
"grad_norm": 0.5005855560302734,
|
| 1149 |
+
"learning_rate": 8.520256971408453e-07,
|
| 1150 |
+
"loss": 0.403,
|
| 1151 |
+
"step": 163
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.5158027812895067,
|
| 1155 |
+
"grad_norm": 0.45457518100738525,
|
| 1156 |
+
"learning_rate": 8.025779637096138e-07,
|
| 1157 |
+
"loss": 0.4441,
|
| 1158 |
+
"step": 164
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.5309734513274336,
|
| 1162 |
+
"grad_norm": 0.48755598068237305,
|
| 1163 |
+
"learning_rate": 7.544829088890326e-07,
|
| 1164 |
+
"loss": 0.4202,
|
| 1165 |
+
"step": 165
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.5461441213653604,
|
| 1169 |
+
"grad_norm": 0.4439229965209961,
|
| 1170 |
+
"learning_rate": 7.077560319906696e-07,
|
| 1171 |
+
"loss": 0.385,
|
| 1172 |
+
"step": 166
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.561314791403287,
|
| 1176 |
+
"grad_norm": 0.47930651903152466,
|
| 1177 |
+
"learning_rate": 6.624123914114122e-07,
|
| 1178 |
+
"loss": 0.4306,
|
| 1179 |
+
"step": 167
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.5764854614412136,
|
| 1183 |
+
"grad_norm": 0.4874081611633301,
|
| 1184 |
+
"learning_rate": 6.184665997806832e-07,
|
| 1185 |
+
"loss": 0.4068,
|
| 1186 |
+
"step": 168
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.5916561314791404,
|
| 1190 |
+
"grad_norm": 0.4433707296848297,
|
| 1191 |
+
"learning_rate": 5.759328192513075e-07,
|
| 1192 |
+
"loss": 0.4043,
|
| 1193 |
+
"step": 169
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.606826801517067,
|
| 1197 |
+
"grad_norm": 0.425663024187088,
|
| 1198 |
+
"learning_rate": 5.348247569355736e-07,
|
| 1199 |
+
"loss": 0.3979,
|
| 1200 |
+
"step": 170
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.6219974715549936,
|
| 1204 |
+
"grad_norm": 0.4746127128601074,
|
| 1205 |
+
"learning_rate": 4.951556604879049e-07,
|
| 1206 |
+
"loss": 0.4319,
|
| 1207 |
+
"step": 171
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.6371681415929205,
|
| 1211 |
+
"grad_norm": 0.8202550411224365,
|
| 1212 |
+
"learning_rate": 4.569383138356276e-07,
|
| 1213 |
+
"loss": 0.3562,
|
| 1214 |
+
"step": 172
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.6523388116308473,
|
| 1218 |
+
"grad_norm": 0.48485466837882996,
|
| 1219 |
+
"learning_rate": 4.201850330591678e-07,
|
| 1220 |
+
"loss": 0.446,
|
| 1221 |
+
"step": 173
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.6675094816687737,
|
| 1225 |
+
"grad_norm": 0.5126750469207764,
|
| 1226 |
+
"learning_rate": 3.8490766242301356e-07,
|
| 1227 |
+
"loss": 0.4152,
|
| 1228 |
+
"step": 174
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.6826801517067005,
|
| 1232 |
+
"grad_norm": 0.4601971209049225,
|
| 1233 |
+
"learning_rate": 3.511175705587433e-07,
|
| 1234 |
+
"loss": 0.4524,
|
| 1235 |
+
"step": 175
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.697850821744627,
|
| 1239 |
+
"grad_norm": 0.4501294493675232,
|
| 1240 |
+
"learning_rate": 3.18825646801314e-07,
|
| 1241 |
+
"loss": 0.4194,
|
| 1242 |
+
"step": 176
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.7130214917825537,
|
| 1246 |
+
"grad_norm": 0.483879029750824,
|
| 1247 |
+
"learning_rate": 2.8804229767982637e-07,
|
| 1248 |
+
"loss": 0.422,
|
| 1249 |
+
"step": 177
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.7281921618204805,
|
| 1253 |
+
"grad_norm": 0.5405057668685913,
|
| 1254 |
+
"learning_rate": 2.587774435638679e-07,
|
| 1255 |
+
"loss": 0.3963,
|
| 1256 |
+
"step": 178
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.7433628318584073,
|
| 1260 |
+
"grad_norm": 0.4544113576412201,
|
| 1261 |
+
"learning_rate": 2.3104051546654016e-07,
|
| 1262 |
+
"loss": 0.4196,
|
| 1263 |
+
"step": 179
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.7585335018963337,
|
| 1267 |
+
"grad_norm": 0.4182121753692627,
|
| 1268 |
+
"learning_rate": 2.0484045200517222e-07,
|
| 1269 |
+
"loss": 0.3744,
|
| 1270 |
+
"step": 180
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.7737041719342606,
|
| 1274 |
+
"grad_norm": 0.45946115255355835,
|
| 1275 |
+
"learning_rate": 1.801856965207338e-07,
|
| 1276 |
+
"loss": 0.4358,
|
| 1277 |
+
"step": 181
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.788874841972187,
|
| 1281 |
+
"grad_norm": 0.5174111127853394,
|
| 1282 |
+
"learning_rate": 1.5708419435684463e-07,
|
| 1283 |
+
"loss": 0.3991,
|
| 1284 |
+
"step": 182
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.8040455120101138,
|
| 1288 |
+
"grad_norm": 0.511076807975769,
|
| 1289 |
+
"learning_rate": 1.3554339029927532e-07,
|
| 1290 |
+
"loss": 0.4408,
|
| 1291 |
+
"step": 183
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.8192161820480406,
|
| 1295 |
+
"grad_norm": 0.47969406843185425,
|
| 1296 |
+
"learning_rate": 1.1557022617676217e-07,
|
| 1297 |
+
"loss": 0.4113,
|
| 1298 |
+
"step": 184
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.834386852085967,
|
| 1302 |
+
"grad_norm": 0.4797833263874054,
|
| 1303 |
+
"learning_rate": 9.717113862389993e-08,
|
| 1304 |
+
"loss": 0.4321,
|
| 1305 |
+
"step": 185
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.849557522123894,
|
| 1309 |
+
"grad_norm": 0.4134206473827362,
|
| 1310 |
+
"learning_rate": 8.035205700685167e-08,
|
| 1311 |
+
"loss": 0.383,
|
| 1312 |
+
"step": 186
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.8647281921618206,
|
| 1316 |
+
"grad_norm": 0.4146213233470917,
|
| 1317 |
+
"learning_rate": 6.511840151252169e-08,
|
| 1318 |
+
"loss": 0.4215,
|
| 1319 |
+
"step": 187
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.879898862199747,
|
| 1323 |
+
"grad_norm": 0.45892632007598877,
|
| 1324 |
+
"learning_rate": 5.1475081401825553e-08,
|
| 1325 |
+
"loss": 0.4055,
|
| 1326 |
+
"step": 188
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.895069532237674,
|
| 1330 |
+
"grad_norm": 0.44647830724716187,
|
| 1331 |
+
"learning_rate": 3.9426493427611177e-08,
|
| 1332 |
+
"loss": 0.416,
|
| 1333 |
+
"step": 189
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.9102402022756007,
|
| 1337 |
+
"grad_norm": 0.4823377728462219,
|
| 1338 |
+
"learning_rate": 2.8976520417742794e-08,
|
| 1339 |
+
"loss": 0.4056,
|
| 1340 |
+
"step": 190
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.925410872313527,
|
| 1344 |
+
"grad_norm": 0.47122204303741455,
|
| 1345 |
+
"learning_rate": 2.012853002380466e-08,
|
| 1346 |
+
"loss": 0.4114,
|
| 1347 |
+
"step": 191
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.940581542351454,
|
| 1351 |
+
"grad_norm": 0.5014013648033142,
|
| 1352 |
+
"learning_rate": 1.2885373635829756e-08,
|
| 1353 |
+
"loss": 0.3915,
|
| 1354 |
+
"step": 192
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.9557522123893807,
|
| 1358 |
+
"grad_norm": 0.5270026326179504,
|
| 1359 |
+
"learning_rate": 7.249385463395375e-09,
|
| 1360 |
+
"loss": 0.4121,
|
| 1361 |
+
"step": 193
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.970922882427307,
|
| 1365 |
+
"grad_norm": 0.46766987442970276,
|
| 1366 |
+
"learning_rate": 3.2223817833931803e-09,
|
| 1367 |
+
"loss": 0.3977,
|
| 1368 |
+
"step": 194
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.986093552465234,
|
| 1372 |
+
"grad_norm": 0.4500380754470825,
|
| 1373 |
+
"learning_rate": 8.056603547090813e-10,
|
| 1374 |
+
"loss": 0.407,
|
| 1375 |
+
"step": 195
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 2.986093552465234,
|
| 1379 |
+
"step": 195,
|
| 1380 |
+
"total_flos": 2.193163794073518e+17,
|
| 1381 |
+
"train_loss": 0.5202219234063075,
|
| 1382 |
+
"train_runtime": 39110.2979,
|
| 1383 |
+
"train_samples_per_second": 0.243,
|
| 1384 |
+
"train_steps_per_second": 0.005
|
| 1385 |
+
}
|
| 1386 |
+
],
|
| 1387 |
+
"logging_steps": 1,
|
| 1388 |
+
"max_steps": 195,
|
| 1389 |
+
"num_input_tokens_seen": 0,
|
| 1390 |
+
"num_train_epochs": 3,
|
| 1391 |
+
"save_steps": 200,
|
| 1392 |
+
"stateful_callbacks": {
|
| 1393 |
+
"TrainerControl": {
|
| 1394 |
+
"args": {
|
| 1395 |
+
"should_epoch_stop": false,
|
| 1396 |
+
"should_evaluate": false,
|
| 1397 |
+
"should_log": false,
|
| 1398 |
+
"should_save": true,
|
| 1399 |
+
"should_training_stop": true
|
| 1400 |
+
},
|
| 1401 |
+
"attributes": {}
|
| 1402 |
+
}
|
| 1403 |
+
},
|
| 1404 |
+
"total_flos": 2.193163794073518e+17,
|
| 1405 |
+
"train_batch_size": 1,
|
| 1406 |
+
"trial_name": null,
|
| 1407 |
+
"trial_params": null
|
| 1408 |
+
}
|
training_loss.png
ADDED
|