End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1303 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-Coder-32B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: SWE-BENCH-5k-first-2000-claude-search-replace-generation_qwen_code_32B_5k_first_2000_generation
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# SWE-BENCH-5k-first-2000-claude-search-replace-generation_qwen_code_32B_5k_first_2000_generation
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-Coder-32B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: SWE-BENCH-5k-first-2000-claude-search-replace-generation_qwen_code_32B_5k_first_2000_generation
|
|
|
|
| 16 |
|
| 17 |
# SWE-BENCH-5k-first-2000-claude-search-replace-generation_qwen_code_32B_5k_first_2000_generation
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct) on the SWE-BENCH-5k-first-2000-claude-search-replace-generation dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.96398891966759,
|
| 3 |
+
"total_flos": 6.743893969836442e+16,
|
| 4 |
+
"train_loss": 0.3866574793226189,
|
| 5 |
+
"train_runtime": 24143.75,
|
| 6 |
+
"train_samples_per_second": 0.179,
|
| 7 |
+
"train_steps_per_second": 0.007
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.96398891966759,
|
| 3 |
+
"total_flos": 6.743893969836442e+16,
|
| 4 |
+
"train_loss": 0.3866574793226189,
|
| 5 |
+
"train_runtime": 24143.75,
|
| 6 |
+
"train_samples_per_second": 0.179,
|
| 7 |
+
"train_steps_per_second": 0.007
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1303 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.96398891966759,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 180,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.01662049861495845,
|
| 14 |
+
"grad_norm": 1.9652302265167236,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.7152,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0332409972299169,
|
| 21 |
+
"grad_norm": 2.135629177093506,
|
| 22 |
+
"learning_rate": 5.555555555555555e-07,
|
| 23 |
+
"loss": 0.7024,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.04986149584487535,
|
| 28 |
+
"grad_norm": 2.365844964981079,
|
| 29 |
+
"learning_rate": 1.111111111111111e-06,
|
| 30 |
+
"loss": 0.7755,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0664819944598338,
|
| 35 |
+
"grad_norm": 1.939900517463684,
|
| 36 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 37 |
+
"loss": 0.7134,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08310249307479224,
|
| 42 |
+
"grad_norm": 1.8507870435714722,
|
| 43 |
+
"learning_rate": 2.222222222222222e-06,
|
| 44 |
+
"loss": 0.6644,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.0997229916897507,
|
| 49 |
+
"grad_norm": 1.8390847444534302,
|
| 50 |
+
"learning_rate": 2.7777777777777783e-06,
|
| 51 |
+
"loss": 0.7306,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.11634349030470914,
|
| 56 |
+
"grad_norm": 1.2149966955184937,
|
| 57 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 58 |
+
"loss": 0.5377,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1329639889196676,
|
| 63 |
+
"grad_norm": 1.203329086303711,
|
| 64 |
+
"learning_rate": 3.88888888888889e-06,
|
| 65 |
+
"loss": 0.6448,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.14958448753462603,
|
| 70 |
+
"grad_norm": 1.1259090900421143,
|
| 71 |
+
"learning_rate": 4.444444444444444e-06,
|
| 72 |
+
"loss": 0.6041,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.16620498614958448,
|
| 77 |
+
"grad_norm": 0.9785488247871399,
|
| 78 |
+
"learning_rate": 5e-06,
|
| 79 |
+
"loss": 0.6802,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.18282548476454294,
|
| 84 |
+
"grad_norm": 0.7702904343605042,
|
| 85 |
+
"learning_rate": 5.555555555555557e-06,
|
| 86 |
+
"loss": 0.5737,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.1994459833795014,
|
| 91 |
+
"grad_norm": 0.7972448468208313,
|
| 92 |
+
"learning_rate": 6.111111111111112e-06,
|
| 93 |
+
"loss": 0.6071,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.21606648199445982,
|
| 98 |
+
"grad_norm": 0.8643639087677002,
|
| 99 |
+
"learning_rate": 6.666666666666667e-06,
|
| 100 |
+
"loss": 0.5645,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.23268698060941828,
|
| 105 |
+
"grad_norm": 0.822340190410614,
|
| 106 |
+
"learning_rate": 7.222222222222223e-06,
|
| 107 |
+
"loss": 0.5512,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.24930747922437674,
|
| 112 |
+
"grad_norm": 1.0604660511016846,
|
| 113 |
+
"learning_rate": 7.77777777777778e-06,
|
| 114 |
+
"loss": 0.5875,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.2659279778393352,
|
| 119 |
+
"grad_norm": 0.8126739263534546,
|
| 120 |
+
"learning_rate": 8.333333333333334e-06,
|
| 121 |
+
"loss": 0.5601,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.28254847645429365,
|
| 126 |
+
"grad_norm": 0.7240079641342163,
|
| 127 |
+
"learning_rate": 8.888888888888888e-06,
|
| 128 |
+
"loss": 0.5724,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.29916897506925205,
|
| 133 |
+
"grad_norm": 0.6566236615180969,
|
| 134 |
+
"learning_rate": 9.444444444444445e-06,
|
| 135 |
+
"loss": 0.5535,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.3157894736842105,
|
| 140 |
+
"grad_norm": 0.7229272723197937,
|
| 141 |
+
"learning_rate": 1e-05,
|
| 142 |
+
"loss": 0.5413,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.33240997229916897,
|
| 147 |
+
"grad_norm": 0.6160261034965515,
|
| 148 |
+
"learning_rate": 9.999059852242508e-06,
|
| 149 |
+
"loss": 0.4809,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.3490304709141274,
|
| 154 |
+
"grad_norm": 0.5426657199859619,
|
| 155 |
+
"learning_rate": 9.996239762521152e-06,
|
| 156 |
+
"loss": 0.4453,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.3656509695290859,
|
| 161 |
+
"grad_norm": 0.6986624002456665,
|
| 162 |
+
"learning_rate": 9.991540791356342e-06,
|
| 163 |
+
"loss": 0.5704,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.38227146814404434,
|
| 168 |
+
"grad_norm": 0.6466948986053467,
|
| 169 |
+
"learning_rate": 9.98496470583896e-06,
|
| 170 |
+
"loss": 0.5222,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.3988919667590028,
|
| 175 |
+
"grad_norm": 0.5881003141403198,
|
| 176 |
+
"learning_rate": 9.976513978965829e-06,
|
| 177 |
+
"loss": 0.4903,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.4155124653739612,
|
| 182 |
+
"grad_norm": 0.5835773348808289,
|
| 183 |
+
"learning_rate": 9.966191788709716e-06,
|
| 184 |
+
"loss": 0.4936,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.43213296398891965,
|
| 189 |
+
"grad_norm": 0.5974717736244202,
|
| 190 |
+
"learning_rate": 9.954002016824226e-06,
|
| 191 |
+
"loss": 0.544,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.4487534626038781,
|
| 196 |
+
"grad_norm": 0.6126233339309692,
|
| 197 |
+
"learning_rate": 9.939949247384046e-06,
|
| 198 |
+
"loss": 0.5313,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.46537396121883656,
|
| 203 |
+
"grad_norm": 0.5605891942977905,
|
| 204 |
+
"learning_rate": 9.924038765061042e-06,
|
| 205 |
+
"loss": 0.5121,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.481994459833795,
|
| 210 |
+
"grad_norm": 0.523395299911499,
|
| 211 |
+
"learning_rate": 9.906276553136924e-06,
|
| 212 |
+
"loss": 0.4705,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.4986149584487535,
|
| 217 |
+
"grad_norm": 0.5597982406616211,
|
| 218 |
+
"learning_rate": 9.886669291253178e-06,
|
| 219 |
+
"loss": 0.4951,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5152354570637119,
|
| 224 |
+
"grad_norm": 0.5273374915122986,
|
| 225 |
+
"learning_rate": 9.86522435289912e-06,
|
| 226 |
+
"loss": 0.4763,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.5318559556786704,
|
| 231 |
+
"grad_norm": 0.5255304574966431,
|
| 232 |
+
"learning_rate": 9.841949802639031e-06,
|
| 233 |
+
"loss": 0.5133,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.5484764542936288,
|
| 238 |
+
"grad_norm": 0.8223831057548523,
|
| 239 |
+
"learning_rate": 9.816854393079402e-06,
|
| 240 |
+
"loss": 0.4865,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5650969529085873,
|
| 245 |
+
"grad_norm": 0.4619203805923462,
|
| 246 |
+
"learning_rate": 9.789947561577445e-06,
|
| 247 |
+
"loss": 0.4631,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5817174515235457,
|
| 252 |
+
"grad_norm": 0.4974648654460907,
|
| 253 |
+
"learning_rate": 9.761239426692077e-06,
|
| 254 |
+
"loss": 0.5039,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.5983379501385041,
|
| 259 |
+
"grad_norm": 0.5178198218345642,
|
| 260 |
+
"learning_rate": 9.730740784378755e-06,
|
| 261 |
+
"loss": 0.4618,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.6149584487534626,
|
| 266 |
+
"grad_norm": 0.5592218637466431,
|
| 267 |
+
"learning_rate": 9.698463103929542e-06,
|
| 268 |
+
"loss": 0.4777,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.631578947368421,
|
| 273 |
+
"grad_norm": 0.4956098198890686,
|
| 274 |
+
"learning_rate": 9.664418523660004e-06,
|
| 275 |
+
"loss": 0.4925,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.6481994459833795,
|
| 280 |
+
"grad_norm": 0.48805150389671326,
|
| 281 |
+
"learning_rate": 9.628619846344453e-06,
|
| 282 |
+
"loss": 0.4423,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.6648199445983379,
|
| 287 |
+
"grad_norm": 0.5749639868736267,
|
| 288 |
+
"learning_rate": 9.591080534401371e-06,
|
| 289 |
+
"loss": 0.55,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.6814404432132964,
|
| 294 |
+
"grad_norm": 0.7393980622291565,
|
| 295 |
+
"learning_rate": 9.551814704830734e-06,
|
| 296 |
+
"loss": 0.426,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.6980609418282548,
|
| 301 |
+
"grad_norm": 0.5011327862739563,
|
| 302 |
+
"learning_rate": 9.51083712390519e-06,
|
| 303 |
+
"loss": 0.4628,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.7146814404432132,
|
| 308 |
+
"grad_norm": 0.572926938533783,
|
| 309 |
+
"learning_rate": 9.468163201617063e-06,
|
| 310 |
+
"loss": 0.527,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.7313019390581718,
|
| 315 |
+
"grad_norm": 0.5243227481842041,
|
| 316 |
+
"learning_rate": 9.423808985883289e-06,
|
| 317 |
+
"loss": 0.5115,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.7479224376731302,
|
| 322 |
+
"grad_norm": 0.5271593928337097,
|
| 323 |
+
"learning_rate": 9.377791156510456e-06,
|
| 324 |
+
"loss": 0.4921,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.7645429362880887,
|
| 329 |
+
"grad_norm": 0.5143831968307495,
|
| 330 |
+
"learning_rate": 9.330127018922195e-06,
|
| 331 |
+
"loss": 0.4842,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.7811634349030471,
|
| 336 |
+
"grad_norm": 0.5135733485221863,
|
| 337 |
+
"learning_rate": 9.280834497651334e-06,
|
| 338 |
+
"loss": 0.4939,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.7977839335180056,
|
| 343 |
+
"grad_norm": 0.5173041820526123,
|
| 344 |
+
"learning_rate": 9.229932129599206e-06,
|
| 345 |
+
"loss": 0.4819,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.814404432132964,
|
| 350 |
+
"grad_norm": 0.570851743221283,
|
| 351 |
+
"learning_rate": 9.177439057064684e-06,
|
| 352 |
+
"loss": 0.5439,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.8310249307479224,
|
| 357 |
+
"grad_norm": 0.552671492099762,
|
| 358 |
+
"learning_rate": 9.123375020545534e-06,
|
| 359 |
+
"loss": 0.4669,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.8476454293628809,
|
| 364 |
+
"grad_norm": 0.5668032765388489,
|
| 365 |
+
"learning_rate": 9.067760351314838e-06,
|
| 366 |
+
"loss": 0.5138,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.8642659279778393,
|
| 371 |
+
"grad_norm": 0.48532989621162415,
|
| 372 |
+
"learning_rate": 9.01061596377522e-06,
|
| 373 |
+
"loss": 0.4827,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.8808864265927978,
|
| 378 |
+
"grad_norm": 0.4953126311302185,
|
| 379 |
+
"learning_rate": 8.951963347593797e-06,
|
| 380 |
+
"loss": 0.4273,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.8975069252077562,
|
| 385 |
+
"grad_norm": 0.5042351484298706,
|
| 386 |
+
"learning_rate": 8.891824559620801e-06,
|
| 387 |
+
"loss": 0.5311,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.9141274238227147,
|
| 392 |
+
"grad_norm": 0.532244086265564,
|
| 393 |
+
"learning_rate": 8.83022221559489e-06,
|
| 394 |
+
"loss": 0.5364,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.9307479224376731,
|
| 399 |
+
"grad_norm": 0.5507211089134216,
|
| 400 |
+
"learning_rate": 8.767179481638303e-06,
|
| 401 |
+
"loss": 0.5264,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.9473684210526315,
|
| 406 |
+
"grad_norm": 0.5117627382278442,
|
| 407 |
+
"learning_rate": 8.702720065545024e-06,
|
| 408 |
+
"loss": 0.4994,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.96398891966759,
|
| 413 |
+
"grad_norm": 0.6424684524536133,
|
| 414 |
+
"learning_rate": 8.636868207865244e-06,
|
| 415 |
+
"loss": 0.5321,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.9806094182825484,
|
| 420 |
+
"grad_norm": 0.5632804036140442,
|
| 421 |
+
"learning_rate": 8.569648672789496e-06,
|
| 422 |
+
"loss": 0.5354,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.997229916897507,
|
| 427 |
+
"grad_norm": 0.5519580841064453,
|
| 428 |
+
"learning_rate": 8.501086738835843e-06,
|
| 429 |
+
"loss": 0.5502,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.0,
|
| 434 |
+
"grad_norm": 0.5519580841064453,
|
| 435 |
+
"learning_rate": 8.43120818934367e-06,
|
| 436 |
+
"loss": 0.4298,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.0166204986149585,
|
| 441 |
+
"grad_norm": 1.4024403095245361,
|
| 442 |
+
"learning_rate": 8.360039302777614e-06,
|
| 443 |
+
"loss": 0.3848,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.0332409972299168,
|
| 448 |
+
"grad_norm": 0.4745033085346222,
|
| 449 |
+
"learning_rate": 8.28760684284532e-06,
|
| 450 |
+
"loss": 0.4,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.0498614958448753,
|
| 455 |
+
"grad_norm": 0.5079669952392578,
|
| 456 |
+
"learning_rate": 8.213938048432697e-06,
|
| 457 |
+
"loss": 0.3824,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.0664819944598338,
|
| 462 |
+
"grad_norm": 0.49697190523147583,
|
| 463 |
+
"learning_rate": 8.139060623360494e-06,
|
| 464 |
+
"loss": 0.4243,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.0831024930747923,
|
| 469 |
+
"grad_norm": 0.4616394639015198,
|
| 470 |
+
"learning_rate": 8.063002725966014e-06,
|
| 471 |
+
"loss": 0.3888,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.0997229916897506,
|
| 476 |
+
"grad_norm": 0.4260391294956207,
|
| 477 |
+
"learning_rate": 7.985792958513932e-06,
|
| 478 |
+
"loss": 0.3406,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.1163434903047091,
|
| 483 |
+
"grad_norm": 0.47153493762016296,
|
| 484 |
+
"learning_rate": 7.907460356440133e-06,
|
| 485 |
+
"loss": 0.3636,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.1329639889196677,
|
| 490 |
+
"grad_norm": 0.5076174139976501,
|
| 491 |
+
"learning_rate": 7.828034377432694e-06,
|
| 492 |
+
"loss": 0.4166,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.149584487534626,
|
| 497 |
+
"grad_norm": 0.5310080647468567,
|
| 498 |
+
"learning_rate": 7.747544890354031e-06,
|
| 499 |
+
"loss": 0.4311,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.1662049861495845,
|
| 504 |
+
"grad_norm": 0.5010002851486206,
|
| 505 |
+
"learning_rate": 7.666022164008458e-06,
|
| 506 |
+
"loss": 0.3193,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.182825484764543,
|
| 511 |
+
"grad_norm": 0.49259936809539795,
|
| 512 |
+
"learning_rate": 7.5834968557593155e-06,
|
| 513 |
+
"loss": 0.3456,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.1994459833795015,
|
| 518 |
+
"grad_norm": 0.5213885307312012,
|
| 519 |
+
"learning_rate": 7.500000000000001e-06,
|
| 520 |
+
"loss": 0.3615,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.2160664819944598,
|
| 525 |
+
"grad_norm": 0.512752115726471,
|
| 526 |
+
"learning_rate": 7.415562996483193e-06,
|
| 527 |
+
"loss": 0.3569,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.2326869806094183,
|
| 532 |
+
"grad_norm": 0.5139035582542419,
|
| 533 |
+
"learning_rate": 7.330217598512696e-06,
|
| 534 |
+
"loss": 0.3859,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.2493074792243768,
|
| 539 |
+
"grad_norm": 0.5561084151268005,
|
| 540 |
+
"learning_rate": 7.243995901002312e-06,
|
| 541 |
+
"loss": 0.363,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.2659279778393353,
|
| 546 |
+
"grad_norm": 0.49844229221343994,
|
| 547 |
+
"learning_rate": 7.156930328406268e-06,
|
| 548 |
+
"loss": 0.3648,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.2825484764542936,
|
| 553 |
+
"grad_norm": 0.5111745595932007,
|
| 554 |
+
"learning_rate": 7.069053622525697e-06,
|
| 555 |
+
"loss": 0.3453,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.299168975069252,
|
| 560 |
+
"grad_norm": 0.5968831777572632,
|
| 561 |
+
"learning_rate": 6.980398830195785e-06,
|
| 562 |
+
"loss": 0.3601,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.3157894736842106,
|
| 567 |
+
"grad_norm": 0.3998188376426697,
|
| 568 |
+
"learning_rate": 6.890999290858213e-06,
|
| 569 |
+
"loss": 0.2965,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.332409972299169,
|
| 574 |
+
"grad_norm": 0.5044348239898682,
|
| 575 |
+
"learning_rate": 6.800888624023552e-06,
|
| 576 |
+
"loss": 0.3579,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.3490304709141274,
|
| 581 |
+
"grad_norm": 0.499636709690094,
|
| 582 |
+
"learning_rate": 6.710100716628345e-06,
|
| 583 |
+
"loss": 0.3751,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.365650969529086,
|
| 588 |
+
"grad_norm": 0.5045871734619141,
|
| 589 |
+
"learning_rate": 6.618669710291607e-06,
|
| 590 |
+
"loss": 0.3782,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.3822714681440442,
|
| 595 |
+
"grad_norm": 0.5296726822853088,
|
| 596 |
+
"learning_rate": 6.526629988475567e-06,
|
| 597 |
+
"loss": 0.413,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.3988919667590027,
|
| 602 |
+
"grad_norm": 0.5541542768478394,
|
| 603 |
+
"learning_rate": 6.434016163555452e-06,
|
| 604 |
+
"loss": 0.4176,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.4155124653739612,
|
| 609 |
+
"grad_norm": 0.52264803647995,
|
| 610 |
+
"learning_rate": 6.340863063803187e-06,
|
| 611 |
+
"loss": 0.3687,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.4321329639889195,
|
| 616 |
+
"grad_norm": 0.5726013779640198,
|
| 617 |
+
"learning_rate": 6.247205720289907e-06,
|
| 618 |
+
"loss": 0.4127,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.448753462603878,
|
| 623 |
+
"grad_norm": 0.5129911303520203,
|
| 624 |
+
"learning_rate": 6.153079353712201e-06,
|
| 625 |
+
"loss": 0.3608,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.4653739612188366,
|
| 630 |
+
"grad_norm": 0.5869404673576355,
|
| 631 |
+
"learning_rate": 6.058519361147055e-06,
|
| 632 |
+
"loss": 0.369,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.481994459833795,
|
| 637 |
+
"grad_norm": 0.4603992998600006,
|
| 638 |
+
"learning_rate": 5.9635613027404495e-06,
|
| 639 |
+
"loss": 0.2792,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.4986149584487536,
|
| 644 |
+
"grad_norm": 0.433829128742218,
|
| 645 |
+
"learning_rate": 5.8682408883346535e-06,
|
| 646 |
+
"loss": 0.2935,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.5152354570637119,
|
| 651 |
+
"grad_norm": 0.4892548620700836,
|
| 652 |
+
"learning_rate": 5.772593964039203e-06,
|
| 653 |
+
"loss": 0.3591,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.5318559556786704,
|
| 658 |
+
"grad_norm": 0.4414325952529907,
|
| 659 |
+
"learning_rate": 5.6766564987506564e-06,
|
| 660 |
+
"loss": 0.3312,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.548476454293629,
|
| 665 |
+
"grad_norm": 0.5104185938835144,
|
| 666 |
+
"learning_rate": 5.5804645706261515e-06,
|
| 667 |
+
"loss": 0.3524,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.5650969529085872,
|
| 672 |
+
"grad_norm": 0.46491438150405884,
|
| 673 |
+
"learning_rate": 5.484054353515896e-06,
|
| 674 |
+
"loss": 0.3127,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.5817174515235457,
|
| 679 |
+
"grad_norm": 0.5037529468536377,
|
| 680 |
+
"learning_rate": 5.387462103359655e-06,
|
| 681 |
+
"loss": 0.3549,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.5983379501385042,
|
| 686 |
+
"grad_norm": 0.456927090883255,
|
| 687 |
+
"learning_rate": 5.290724144552379e-06,
|
| 688 |
+
"loss": 0.3583,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.6149584487534625,
|
| 693 |
+
"grad_norm": 0.48146891593933105,
|
| 694 |
+
"learning_rate": 5.193876856284085e-06,
|
| 695 |
+
"loss": 0.3485,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.631578947368421,
|
| 700 |
+
"grad_norm": 0.45695117115974426,
|
| 701 |
+
"learning_rate": 5.096956658859122e-06,
|
| 702 |
+
"loss": 0.3325,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.6481994459833795,
|
| 707 |
+
"grad_norm": 0.46289077401161194,
|
| 708 |
+
"learning_rate": 5e-06,
|
| 709 |
+
"loss": 0.3461,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.6648199445983378,
|
| 714 |
+
"grad_norm": 0.5340746641159058,
|
| 715 |
+
"learning_rate": 4.903043341140879e-06,
|
| 716 |
+
"loss": 0.3856,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.6814404432132966,
|
| 721 |
+
"grad_norm": 0.433956503868103,
|
| 722 |
+
"learning_rate": 4.806123143715916e-06,
|
| 723 |
+
"loss": 0.3166,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.6980609418282548,
|
| 728 |
+
"grad_norm": 0.4446304440498352,
|
| 729 |
+
"learning_rate": 4.7092758554476215e-06,
|
| 730 |
+
"loss": 0.3378,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.7146814404432131,
|
| 735 |
+
"grad_norm": 0.5027093291282654,
|
| 736 |
+
"learning_rate": 4.6125378966403465e-06,
|
| 737 |
+
"loss": 0.3915,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.7313019390581719,
|
| 742 |
+
"grad_norm": 0.5546647310256958,
|
| 743 |
+
"learning_rate": 4.515945646484105e-06,
|
| 744 |
+
"loss": 0.3484,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.7479224376731302,
|
| 749 |
+
"grad_norm": 0.49674123525619507,
|
| 750 |
+
"learning_rate": 4.4195354293738484e-06,
|
| 751 |
+
"loss": 0.3501,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.7645429362880887,
|
| 756 |
+
"grad_norm": 0.5134773850440979,
|
| 757 |
+
"learning_rate": 4.323343501249346e-06,
|
| 758 |
+
"loss": 0.3818,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.7811634349030472,
|
| 763 |
+
"grad_norm": 0.5111790299415588,
|
| 764 |
+
"learning_rate": 4.227406035960798e-06,
|
| 765 |
+
"loss": 0.4027,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.7977839335180055,
|
| 770 |
+
"grad_norm": 0.5103554129600525,
|
| 771 |
+
"learning_rate": 4.131759111665349e-06,
|
| 772 |
+
"loss": 0.3295,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.814404432132964,
|
| 777 |
+
"grad_norm": 0.48488280177116394,
|
| 778 |
+
"learning_rate": 4.036438697259551e-06,
|
| 779 |
+
"loss": 0.3339,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.8310249307479225,
|
| 784 |
+
"grad_norm": 0.4840296506881714,
|
| 785 |
+
"learning_rate": 3.941480638852948e-06,
|
| 786 |
+
"loss": 0.3519,
|
| 787 |
+
"step": 111
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.8476454293628808,
|
| 791 |
+
"grad_norm": 0.4919949471950531,
|
| 792 |
+
"learning_rate": 3.8469206462878e-06,
|
| 793 |
+
"loss": 0.328,
|
| 794 |
+
"step": 112
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.8642659279778393,
|
| 798 |
+
"grad_norm": 0.5291365385055542,
|
| 799 |
+
"learning_rate": 3.752794279710094e-06,
|
| 800 |
+
"loss": 0.3753,
|
| 801 |
+
"step": 113
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.8808864265927978,
|
| 805 |
+
"grad_norm": 0.4807715117931366,
|
| 806 |
+
"learning_rate": 3.6591369361968127e-06,
|
| 807 |
+
"loss": 0.393,
|
| 808 |
+
"step": 114
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.897506925207756,
|
| 812 |
+
"grad_norm": 0.4700012803077698,
|
| 813 |
+
"learning_rate": 3.5659838364445505e-06,
|
| 814 |
+
"loss": 0.3182,
|
| 815 |
+
"step": 115
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.9141274238227148,
|
| 819 |
+
"grad_norm": 1.0692706108093262,
|
| 820 |
+
"learning_rate": 3.473370011524435e-06,
|
| 821 |
+
"loss": 0.3463,
|
| 822 |
+
"step": 116
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.9307479224376731,
|
| 826 |
+
"grad_norm": 0.49183958768844604,
|
| 827 |
+
"learning_rate": 3.3813302897083955e-06,
|
| 828 |
+
"loss": 0.3694,
|
| 829 |
+
"step": 117
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.9473684210526314,
|
| 833 |
+
"grad_norm": 0.5577133893966675,
|
| 834 |
+
"learning_rate": 3.289899283371657e-06,
|
| 835 |
+
"loss": 0.3693,
|
| 836 |
+
"step": 118
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.9639889196675901,
|
| 840 |
+
"grad_norm": 0.47118237614631653,
|
| 841 |
+
"learning_rate": 3.1991113759764493e-06,
|
| 842 |
+
"loss": 0.3325,
|
| 843 |
+
"step": 119
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.9806094182825484,
|
| 847 |
+
"grad_norm": 0.44954901933670044,
|
| 848 |
+
"learning_rate": 3.1090007091417884e-06,
|
| 849 |
+
"loss": 0.3497,
|
| 850 |
+
"step": 120
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.997229916897507,
|
| 854 |
+
"grad_norm": 0.5316449403762817,
|
| 855 |
+
"learning_rate": 3.019601169804216e-06,
|
| 856 |
+
"loss": 0.4239,
|
| 857 |
+
"step": 121
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.0,
|
| 861 |
+
"grad_norm": 0.5316449403762817,
|
| 862 |
+
"learning_rate": 2.9309463774743047e-06,
|
| 863 |
+
"loss": 0.302,
|
| 864 |
+
"step": 122
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.0166204986149583,
|
| 868 |
+
"grad_norm": 1.3086326122283936,
|
| 869 |
+
"learning_rate": 2.843069671593734e-06,
|
| 870 |
+
"loss": 0.2255,
|
| 871 |
+
"step": 123
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 2.033240997229917,
|
| 875 |
+
"grad_norm": 0.4746488928794861,
|
| 876 |
+
"learning_rate": 2.7560040989976894e-06,
|
| 877 |
+
"loss": 0.2275,
|
| 878 |
+
"step": 124
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.0498614958448753,
|
| 882 |
+
"grad_norm": 0.4944143295288086,
|
| 883 |
+
"learning_rate": 2.6697824014873076e-06,
|
| 884 |
+
"loss": 0.2648,
|
| 885 |
+
"step": 125
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 2.0664819944598336,
|
| 889 |
+
"grad_norm": 0.5195774435997009,
|
| 890 |
+
"learning_rate": 2.5844370035168077e-06,
|
| 891 |
+
"loss": 0.2707,
|
| 892 |
+
"step": 126
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.0831024930747923,
|
| 896 |
+
"grad_norm": 0.885553240776062,
|
| 897 |
+
"learning_rate": 2.5000000000000015e-06,
|
| 898 |
+
"loss": 0.2764,
|
| 899 |
+
"step": 127
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.0997229916897506,
|
| 903 |
+
"grad_norm": 0.5028234124183655,
|
| 904 |
+
"learning_rate": 2.4165031442406857e-06,
|
| 905 |
+
"loss": 0.2503,
|
| 906 |
+
"step": 128
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.1163434903047094,
|
| 910 |
+
"grad_norm": 0.4780957102775574,
|
| 911 |
+
"learning_rate": 2.333977835991545e-06,
|
| 912 |
+
"loss": 0.2406,
|
| 913 |
+
"step": 129
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.1329639889196677,
|
| 917 |
+
"grad_norm": 0.46052825450897217,
|
| 918 |
+
"learning_rate": 2.2524551096459703e-06,
|
| 919 |
+
"loss": 0.2155,
|
| 920 |
+
"step": 130
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.149584487534626,
|
| 924 |
+
"grad_norm": 0.6180452704429626,
|
| 925 |
+
"learning_rate": 2.171965622567308e-06,
|
| 926 |
+
"loss": 0.2787,
|
| 927 |
+
"step": 131
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.1662049861495847,
|
| 931 |
+
"grad_norm": 0.6939100027084351,
|
| 932 |
+
"learning_rate": 2.0925396435598665e-06,
|
| 933 |
+
"loss": 0.246,
|
| 934 |
+
"step": 132
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.182825484764543,
|
| 938 |
+
"grad_norm": 0.6042692065238953,
|
| 939 |
+
"learning_rate": 2.0142070414860704e-06,
|
| 940 |
+
"loss": 0.2609,
|
| 941 |
+
"step": 133
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.1994459833795013,
|
| 945 |
+
"grad_norm": 0.7851183414459229,
|
| 946 |
+
"learning_rate": 1.936997274033986e-06,
|
| 947 |
+
"loss": 0.2876,
|
| 948 |
+
"step": 134
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.21606648199446,
|
| 952 |
+
"grad_norm": 0.5801565051078796,
|
| 953 |
+
"learning_rate": 1.8609393766395083e-06,
|
| 954 |
+
"loss": 0.288,
|
| 955 |
+
"step": 135
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.2326869806094183,
|
| 959 |
+
"grad_norm": 0.5398533940315247,
|
| 960 |
+
"learning_rate": 1.7860619515673034e-06,
|
| 961 |
+
"loss": 0.2958,
|
| 962 |
+
"step": 136
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.2493074792243766,
|
| 966 |
+
"grad_norm": 0.48142921924591064,
|
| 967 |
+
"learning_rate": 1.7123931571546826e-06,
|
| 968 |
+
"loss": 0.2506,
|
| 969 |
+
"step": 137
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.2659279778393353,
|
| 973 |
+
"grad_norm": 0.48484477400779724,
|
| 974 |
+
"learning_rate": 1.639960697222388e-06,
|
| 975 |
+
"loss": 0.2166,
|
| 976 |
+
"step": 138
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.2825484764542936,
|
| 980 |
+
"grad_norm": 0.4676513075828552,
|
| 981 |
+
"learning_rate": 1.5687918106563326e-06,
|
| 982 |
+
"loss": 0.2558,
|
| 983 |
+
"step": 139
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.299168975069252,
|
| 987 |
+
"grad_norm": 0.5008206963539124,
|
| 988 |
+
"learning_rate": 1.4989132611641576e-06,
|
| 989 |
+
"loss": 0.2315,
|
| 990 |
+
"step": 140
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.3157894736842106,
|
| 994 |
+
"grad_norm": 0.5055615901947021,
|
| 995 |
+
"learning_rate": 1.4303513272105057e-06,
|
| 996 |
+
"loss": 0.278,
|
| 997 |
+
"step": 141
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.332409972299169,
|
| 1001 |
+
"grad_norm": 0.5048314332962036,
|
| 1002 |
+
"learning_rate": 1.3631317921347564e-06,
|
| 1003 |
+
"loss": 0.2469,
|
| 1004 |
+
"step": 142
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.349030470914127,
|
| 1008 |
+
"grad_norm": 0.4561052620410919,
|
| 1009 |
+
"learning_rate": 1.297279934454978e-06,
|
| 1010 |
+
"loss": 0.2363,
|
| 1011 |
+
"step": 143
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.365650969529086,
|
| 1015 |
+
"grad_norm": 0.4409971237182617,
|
| 1016 |
+
"learning_rate": 1.2328205183616964e-06,
|
| 1017 |
+
"loss": 0.2582,
|
| 1018 |
+
"step": 144
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.3822714681440442,
|
| 1022 |
+
"grad_norm": 0.5186073780059814,
|
| 1023 |
+
"learning_rate": 1.1697777844051105e-06,
|
| 1024 |
+
"loss": 0.2354,
|
| 1025 |
+
"step": 145
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.398891966759003,
|
| 1029 |
+
"grad_norm": 0.4931983947753906,
|
| 1030 |
+
"learning_rate": 1.1081754403792e-06,
|
| 1031 |
+
"loss": 0.2628,
|
| 1032 |
+
"step": 146
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.4155124653739612,
|
| 1036 |
+
"grad_norm": 0.4725812077522278,
|
| 1037 |
+
"learning_rate": 1.0480366524062041e-06,
|
| 1038 |
+
"loss": 0.2465,
|
| 1039 |
+
"step": 147
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.4321329639889195,
|
| 1043 |
+
"grad_norm": 0.459830641746521,
|
| 1044 |
+
"learning_rate": 9.893840362247809e-07,
|
| 1045 |
+
"loss": 0.2494,
|
| 1046 |
+
"step": 148
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.4487534626038783,
|
| 1050 |
+
"grad_norm": 0.45882484316825867,
|
| 1051 |
+
"learning_rate": 9.322396486851626e-07,
|
| 1052 |
+
"loss": 0.2572,
|
| 1053 |
+
"step": 149
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.4653739612188366,
|
| 1057 |
+
"grad_norm": 0.4628044664859772,
|
| 1058 |
+
"learning_rate": 8.766249794544662e-07,
|
| 1059 |
+
"loss": 0.2473,
|
| 1060 |
+
"step": 150
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.481994459833795,
|
| 1064 |
+
"grad_norm": 0.43482884764671326,
|
| 1065 |
+
"learning_rate": 8.225609429353187e-07,
|
| 1066 |
+
"loss": 0.2334,
|
| 1067 |
+
"step": 151
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.4986149584487536,
|
| 1071 |
+
"grad_norm": 0.5092786550521851,
|
| 1072 |
+
"learning_rate": 7.700678704007947e-07,
|
| 1073 |
+
"loss": 0.2464,
|
| 1074 |
+
"step": 152
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.515235457063712,
|
| 1078 |
+
"grad_norm": 0.5002970695495605,
|
| 1079 |
+
"learning_rate": 7.191655023486682e-07,
|
| 1080 |
+
"loss": 0.2386,
|
| 1081 |
+
"step": 153
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.5318559556786706,
|
| 1085 |
+
"grad_norm": 0.44085896015167236,
|
| 1086 |
+
"learning_rate": 6.698729810778065e-07,
|
| 1087 |
+
"loss": 0.2231,
|
| 1088 |
+
"step": 154
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.548476454293629,
|
| 1092 |
+
"grad_norm": 0.4750898480415344,
|
| 1093 |
+
"learning_rate": 6.222088434895462e-07,
|
| 1094 |
+
"loss": 0.2746,
|
| 1095 |
+
"step": 155
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.565096952908587,
|
| 1099 |
+
"grad_norm": 0.5058760643005371,
|
| 1100 |
+
"learning_rate": 5.76191014116711e-07,
|
| 1101 |
+
"loss": 0.2753,
|
| 1102 |
+
"step": 156
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.581717451523546,
|
| 1106 |
+
"grad_norm": 0.4807314872741699,
|
| 1107 |
+
"learning_rate": 5.318367983829393e-07,
|
| 1108 |
+
"loss": 0.2295,
|
| 1109 |
+
"step": 157
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.598337950138504,
|
| 1113 |
+
"grad_norm": 0.4975450336933136,
|
| 1114 |
+
"learning_rate": 4.891628760948114e-07,
|
| 1115 |
+
"loss": 0.2623,
|
| 1116 |
+
"step": 158
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.6149584487534625,
|
| 1120 |
+
"grad_norm": 0.44517505168914795,
|
| 1121 |
+
"learning_rate": 4.481852951692672e-07,
|
| 1122 |
+
"loss": 0.2505,
|
| 1123 |
+
"step": 159
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.6315789473684212,
|
| 1127 |
+
"grad_norm": 0.526871919631958,
|
| 1128 |
+
"learning_rate": 4.089194655986306e-07,
|
| 1129 |
+
"loss": 0.2944,
|
| 1130 |
+
"step": 160
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.6481994459833795,
|
| 1134 |
+
"grad_norm": 0.5860976576805115,
|
| 1135 |
+
"learning_rate": 3.7138015365554834e-07,
|
| 1136 |
+
"loss": 0.2929,
|
| 1137 |
+
"step": 161
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.664819944598338,
|
| 1141 |
+
"grad_norm": 0.5570012927055359,
|
| 1142 |
+
"learning_rate": 3.355814763399973e-07,
|
| 1143 |
+
"loss": 0.2669,
|
| 1144 |
+
"step": 162
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.6814404432132966,
|
| 1148 |
+
"grad_norm": 0.46305856108665466,
|
| 1149 |
+
"learning_rate": 3.015368960704584e-07,
|
| 1150 |
+
"loss": 0.2464,
|
| 1151 |
+
"step": 163
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.698060941828255,
|
| 1155 |
+
"grad_norm": 0.49931517243385315,
|
| 1156 |
+
"learning_rate": 2.6925921562124867e-07,
|
| 1157 |
+
"loss": 0.233,
|
| 1158 |
+
"step": 164
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.714681440443213,
|
| 1162 |
+
"grad_norm": 0.4253719449043274,
|
| 1163 |
+
"learning_rate": 2.3876057330792344e-07,
|
| 1164 |
+
"loss": 0.2115,
|
| 1165 |
+
"step": 165
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.731301939058172,
|
| 1169 |
+
"grad_norm": 0.46956562995910645,
|
| 1170 |
+
"learning_rate": 2.1005243842255552e-07,
|
| 1171 |
+
"loss": 0.2419,
|
| 1172 |
+
"step": 166
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.74792243767313,
|
| 1176 |
+
"grad_norm": 0.47405821084976196,
|
| 1177 |
+
"learning_rate": 1.8314560692059836e-07,
|
| 1178 |
+
"loss": 0.2442,
|
| 1179 |
+
"step": 167
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.7645429362880884,
|
| 1183 |
+
"grad_norm": 0.5373594164848328,
|
| 1184 |
+
"learning_rate": 1.5805019736097105e-07,
|
| 1185 |
+
"loss": 0.304,
|
| 1186 |
+
"step": 168
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.781163434903047,
|
| 1190 |
+
"grad_norm": 0.49911409616470337,
|
| 1191 |
+
"learning_rate": 1.3477564710088097e-07,
|
| 1192 |
+
"loss": 0.2604,
|
| 1193 |
+
"step": 169
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.7977839335180055,
|
| 1197 |
+
"grad_norm": 0.524211585521698,
|
| 1198 |
+
"learning_rate": 1.1333070874682217e-07,
|
| 1199 |
+
"loss": 0.2319,
|
| 1200 |
+
"step": 170
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.8144044321329638,
|
| 1204 |
+
"grad_norm": 0.49799832701683044,
|
| 1205 |
+
"learning_rate": 9.372344686307655e-08,
|
| 1206 |
+
"loss": 0.2648,
|
| 1207 |
+
"step": 171
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.8310249307479225,
|
| 1211 |
+
"grad_norm": 0.4979800581932068,
|
| 1212 |
+
"learning_rate": 7.59612349389599e-08,
|
| 1213 |
+
"loss": 0.2671,
|
| 1214 |
+
"step": 172
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.847645429362881,
|
| 1218 |
+
"grad_norm": 0.5030661225318909,
|
| 1219 |
+
"learning_rate": 6.005075261595495e-08,
|
| 1220 |
+
"loss": 0.2219,
|
| 1221 |
+
"step": 173
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.864265927977839,
|
| 1225 |
+
"grad_norm": 0.4839530885219574,
|
| 1226 |
+
"learning_rate": 4.599798317577342e-08,
|
| 1227 |
+
"loss": 0.2981,
|
| 1228 |
+
"step": 174
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.880886426592798,
|
| 1232 |
+
"grad_norm": 0.49113729596138,
|
| 1233 |
+
"learning_rate": 3.3808211290284886e-08,
|
| 1234 |
+
"loss": 0.2574,
|
| 1235 |
+
"step": 175
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.897506925207756,
|
| 1239 |
+
"grad_norm": 0.5154249668121338,
|
| 1240 |
+
"learning_rate": 2.3486021034170857e-08,
|
| 1241 |
+
"loss": 0.2584,
|
| 1242 |
+
"step": 176
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.914127423822715,
|
| 1246 |
+
"grad_norm": 0.46952885389328003,
|
| 1247 |
+
"learning_rate": 1.5035294161039882e-08,
|
| 1248 |
+
"loss": 0.2785,
|
| 1249 |
+
"step": 177
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.930747922437673,
|
| 1253 |
+
"grad_norm": 0.49860695004463196,
|
| 1254 |
+
"learning_rate": 8.459208643659122e-09,
|
| 1255 |
+
"loss": 0.2572,
|
| 1256 |
+
"step": 178
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.9473684210526314,
|
| 1260 |
+
"grad_norm": 0.5341483354568481,
|
| 1261 |
+
"learning_rate": 3.760237478849793e-09,
|
| 1262 |
+
"loss": 0.2964,
|
| 1263 |
+
"step": 179
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.96398891966759,
|
| 1267 |
+
"grad_norm": 0.5575993061065674,
|
| 1268 |
+
"learning_rate": 9.401477574932927e-10,
|
| 1269 |
+
"loss": 0.2896,
|
| 1270 |
+
"step": 180
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.96398891966759,
|
| 1274 |
+
"step": 180,
|
| 1275 |
+
"total_flos": 6.743893969836442e+16,
|
| 1276 |
+
"train_loss": 0.3866574793226189,
|
| 1277 |
+
"train_runtime": 24143.75,
|
| 1278 |
+
"train_samples_per_second": 0.179,
|
| 1279 |
+
"train_steps_per_second": 0.007
|
| 1280 |
+
}
|
| 1281 |
+
],
|
| 1282 |
+
"logging_steps": 1,
|
| 1283 |
+
"max_steps": 180,
|
| 1284 |
+
"num_input_tokens_seen": 0,
|
| 1285 |
+
"num_train_epochs": 3,
|
| 1286 |
+
"save_steps": 100,
|
| 1287 |
+
"stateful_callbacks": {
|
| 1288 |
+
"TrainerControl": {
|
| 1289 |
+
"args": {
|
| 1290 |
+
"should_epoch_stop": false,
|
| 1291 |
+
"should_evaluate": false,
|
| 1292 |
+
"should_log": false,
|
| 1293 |
+
"should_save": true,
|
| 1294 |
+
"should_training_stop": true
|
| 1295 |
+
},
|
| 1296 |
+
"attributes": {}
|
| 1297 |
+
}
|
| 1298 |
+
},
|
| 1299 |
+
"total_flos": 6.743893969836442e+16,
|
| 1300 |
+
"train_batch_size": 1,
|
| 1301 |
+
"trial_name": null,
|
| 1302 |
+
"trial_params": null
|
| 1303 |
+
}
|
training_loss.png
ADDED
|