End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2919 -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-combine-all_combine-all-32b
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# SWE-BENCH-combine-all_combine-all-32b
|
| 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-combine-all_combine-all-32b
|
|
|
|
| 16 |
|
| 17 |
# SWE-BENCH-combine-all_combine-all-32b
|
| 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-combine-all dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9848024316109423,
|
| 3 |
+
"total_flos": 3.260394272163103e+17,
|
| 4 |
+
"train_loss": 0.4436677635586175,
|
| 5 |
+
"train_runtime": 142049.243,
|
| 6 |
+
"train_samples_per_second": 0.139,
|
| 7 |
+
"train_steps_per_second": 0.003
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9848024316109423,
|
| 3 |
+
"total_flos": 3.260394272163103e+17,
|
| 4 |
+
"train_loss": 0.4436677635586175,
|
| 5 |
+
"train_runtime": 142049.243,
|
| 6 |
+
"train_samples_per_second": 0.139,
|
| 7 |
+
"train_steps_per_second": 0.003
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2919 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9848024316109423,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 411,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00729483282674772,
|
| 13 |
+
"grad_norm": 0.14541301131248474,
|
| 14 |
+
"learning_rate": 2.3809523809523811e-07,
|
| 15 |
+
"loss": 0.7602,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01458966565349544,
|
| 20 |
+
"grad_norm": 0.1567784547805786,
|
| 21 |
+
"learning_rate": 4.7619047619047623e-07,
|
| 22 |
+
"loss": 0.8215,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02188449848024316,
|
| 27 |
+
"grad_norm": 0.1584789901971817,
|
| 28 |
+
"learning_rate": 7.142857142857143e-07,
|
| 29 |
+
"loss": 0.8269,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.02917933130699088,
|
| 34 |
+
"grad_norm": 0.157843217253685,
|
| 35 |
+
"learning_rate": 9.523809523809525e-07,
|
| 36 |
+
"loss": 0.7909,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.0364741641337386,
|
| 41 |
+
"grad_norm": 0.1607961654663086,
|
| 42 |
+
"learning_rate": 1.1904761904761906e-06,
|
| 43 |
+
"loss": 0.8198,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04376899696048632,
|
| 48 |
+
"grad_norm": 0.15348272025585175,
|
| 49 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 50 |
+
"loss": 0.7687,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05106382978723404,
|
| 55 |
+
"grad_norm": 0.1496104598045349,
|
| 56 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 57 |
+
"loss": 0.7942,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.05835866261398176,
|
| 62 |
+
"grad_norm": 0.2734036445617676,
|
| 63 |
+
"learning_rate": 1.904761904761905e-06,
|
| 64 |
+
"loss": 0.847,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06565349544072949,
|
| 69 |
+
"grad_norm": 0.13683773577213287,
|
| 70 |
+
"learning_rate": 2.1428571428571427e-06,
|
| 71 |
+
"loss": 0.754,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0729483282674772,
|
| 76 |
+
"grad_norm": 0.11306589841842651,
|
| 77 |
+
"learning_rate": 2.380952380952381e-06,
|
| 78 |
+
"loss": 0.6991,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08024316109422493,
|
| 83 |
+
"grad_norm": 0.12233421206474304,
|
| 84 |
+
"learning_rate": 2.6190476190476192e-06,
|
| 85 |
+
"loss": 0.7829,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08753799392097264,
|
| 90 |
+
"grad_norm": 0.10262873023748398,
|
| 91 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 92 |
+
"loss": 0.7048,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09483282674772037,
|
| 97 |
+
"grad_norm": 0.10435234010219574,
|
| 98 |
+
"learning_rate": 3.0952380952380957e-06,
|
| 99 |
+
"loss": 0.768,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.10212765957446808,
|
| 104 |
+
"grad_norm": 0.0735386535525322,
|
| 105 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 106 |
+
"loss": 0.6147,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1094224924012158,
|
| 111 |
+
"grad_norm": 0.07339954376220703,
|
| 112 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 113 |
+
"loss": 0.7452,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.11671732522796352,
|
| 118 |
+
"grad_norm": 0.06846445798873901,
|
| 119 |
+
"learning_rate": 3.80952380952381e-06,
|
| 120 |
+
"loss": 0.7383,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.12401215805471125,
|
| 125 |
+
"grad_norm": 0.07185480743646622,
|
| 126 |
+
"learning_rate": 4.047619047619048e-06,
|
| 127 |
+
"loss": 0.7081,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.13130699088145897,
|
| 132 |
+
"grad_norm": 0.06281375139951706,
|
| 133 |
+
"learning_rate": 4.2857142857142855e-06,
|
| 134 |
+
"loss": 0.6171,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1386018237082067,
|
| 139 |
+
"grad_norm": 0.08302997052669525,
|
| 140 |
+
"learning_rate": 4.523809523809524e-06,
|
| 141 |
+
"loss": 0.6136,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.1458966565349544,
|
| 146 |
+
"grad_norm": 0.07521834969520569,
|
| 147 |
+
"learning_rate": 4.761904761904762e-06,
|
| 148 |
+
"loss": 0.6401,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.15319148936170213,
|
| 153 |
+
"grad_norm": 0.07346966117620468,
|
| 154 |
+
"learning_rate": 5e-06,
|
| 155 |
+
"loss": 0.6575,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.16048632218844985,
|
| 160 |
+
"grad_norm": 0.06220546364784241,
|
| 161 |
+
"learning_rate": 5.2380952380952384e-06,
|
| 162 |
+
"loss": 0.585,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.16778115501519758,
|
| 167 |
+
"grad_norm": 0.06210927292704582,
|
| 168 |
+
"learning_rate": 5.476190476190477e-06,
|
| 169 |
+
"loss": 0.6116,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.17507598784194528,
|
| 174 |
+
"grad_norm": 0.06617508083581924,
|
| 175 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 176 |
+
"loss": 0.6591,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.182370820668693,
|
| 181 |
+
"grad_norm": 0.06115543842315674,
|
| 182 |
+
"learning_rate": 5.9523809523809525e-06,
|
| 183 |
+
"loss": 0.6164,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.18966565349544073,
|
| 188 |
+
"grad_norm": 0.05512455105781555,
|
| 189 |
+
"learning_rate": 6.1904761904761914e-06,
|
| 190 |
+
"loss": 0.6131,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.19696048632218846,
|
| 195 |
+
"grad_norm": 0.05426128953695297,
|
| 196 |
+
"learning_rate": 6.4285714285714295e-06,
|
| 197 |
+
"loss": 0.6233,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.20425531914893616,
|
| 202 |
+
"grad_norm": 0.04495101794600487,
|
| 203 |
+
"learning_rate": 6.666666666666667e-06,
|
| 204 |
+
"loss": 0.6017,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2115501519756839,
|
| 209 |
+
"grad_norm": 0.052700527012348175,
|
| 210 |
+
"learning_rate": 6.9047619047619055e-06,
|
| 211 |
+
"loss": 0.6209,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2188449848024316,
|
| 216 |
+
"grad_norm": 0.05274520814418793,
|
| 217 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 218 |
+
"loss": 0.562,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.22613981762917934,
|
| 223 |
+
"grad_norm": 0.0418085902929306,
|
| 224 |
+
"learning_rate": 7.380952380952382e-06,
|
| 225 |
+
"loss": 0.5356,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.23343465045592704,
|
| 230 |
+
"grad_norm": 0.04744059965014458,
|
| 231 |
+
"learning_rate": 7.61904761904762e-06,
|
| 232 |
+
"loss": 0.5759,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.24072948328267477,
|
| 237 |
+
"grad_norm": 0.051624756306409836,
|
| 238 |
+
"learning_rate": 7.857142857142858e-06,
|
| 239 |
+
"loss": 0.5989,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2480243161094225,
|
| 244 |
+
"grad_norm": 0.04632480815052986,
|
| 245 |
+
"learning_rate": 8.095238095238097e-06,
|
| 246 |
+
"loss": 0.6036,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.2553191489361702,
|
| 251 |
+
"grad_norm": 0.040394943207502365,
|
| 252 |
+
"learning_rate": 8.333333333333334e-06,
|
| 253 |
+
"loss": 0.5439,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.26261398176291795,
|
| 258 |
+
"grad_norm": 0.047632846981287,
|
| 259 |
+
"learning_rate": 8.571428571428571e-06,
|
| 260 |
+
"loss": 0.6187,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.26990881458966565,
|
| 265 |
+
"grad_norm": 0.04498811811208725,
|
| 266 |
+
"learning_rate": 8.80952380952381e-06,
|
| 267 |
+
"loss": 0.5686,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2772036474164134,
|
| 272 |
+
"grad_norm": 0.04858787730336189,
|
| 273 |
+
"learning_rate": 9.047619047619049e-06,
|
| 274 |
+
"loss": 0.6224,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.2844984802431611,
|
| 279 |
+
"grad_norm": 0.04534129053354263,
|
| 280 |
+
"learning_rate": 9.285714285714288e-06,
|
| 281 |
+
"loss": 0.576,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2917933130699088,
|
| 286 |
+
"grad_norm": 0.04878037050366402,
|
| 287 |
+
"learning_rate": 9.523809523809525e-06,
|
| 288 |
+
"loss": 0.5956,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.29908814589665655,
|
| 293 |
+
"grad_norm": 0.044632136821746826,
|
| 294 |
+
"learning_rate": 9.761904761904762e-06,
|
| 295 |
+
"loss": 0.5748,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.30638297872340425,
|
| 300 |
+
"grad_norm": 0.041874803602695465,
|
| 301 |
+
"learning_rate": 1e-05,
|
| 302 |
+
"loss": 0.5752,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.31367781155015195,
|
| 307 |
+
"grad_norm": 0.041942398995161057,
|
| 308 |
+
"learning_rate": 9.999818789066164e-06,
|
| 309 |
+
"loss": 0.5759,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3209726443768997,
|
| 314 |
+
"grad_norm": 0.045705121010541916,
|
| 315 |
+
"learning_rate": 9.999275169399614e-06,
|
| 316 |
+
"loss": 0.5962,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.3282674772036474,
|
| 321 |
+
"grad_norm": 0.043411824852228165,
|
| 322 |
+
"learning_rate": 9.998369180404283e-06,
|
| 323 |
+
"loss": 0.54,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.33556231003039516,
|
| 328 |
+
"grad_norm": 0.03998137265443802,
|
| 329 |
+
"learning_rate": 9.997100887750215e-06,
|
| 330 |
+
"loss": 0.5874,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.34285714285714286,
|
| 335 |
+
"grad_norm": 0.047370899468660355,
|
| 336 |
+
"learning_rate": 9.995470383368808e-06,
|
| 337 |
+
"loss": 0.6051,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.35015197568389056,
|
| 342 |
+
"grad_norm": 0.04455406963825226,
|
| 343 |
+
"learning_rate": 9.993477785446151e-06,
|
| 344 |
+
"loss": 0.5604,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3574468085106383,
|
| 349 |
+
"grad_norm": 0.043418001383543015,
|
| 350 |
+
"learning_rate": 9.991123238414455e-06,
|
| 351 |
+
"loss": 0.5555,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.364741641337386,
|
| 356 |
+
"grad_norm": 0.03939136862754822,
|
| 357 |
+
"learning_rate": 9.988406912941591e-06,
|
| 358 |
+
"loss": 0.5493,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.3720364741641337,
|
| 363 |
+
"grad_norm": 0.04485655948519707,
|
| 364 |
+
"learning_rate": 9.985329005918702e-06,
|
| 365 |
+
"loss": 0.5804,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.37933130699088147,
|
| 370 |
+
"grad_norm": 0.0435781255364418,
|
| 371 |
+
"learning_rate": 9.981889740445958e-06,
|
| 372 |
+
"loss": 0.5617,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.38662613981762917,
|
| 377 |
+
"grad_norm": 0.03838958591222763,
|
| 378 |
+
"learning_rate": 9.978089365816357e-06,
|
| 379 |
+
"loss": 0.5481,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.3939209726443769,
|
| 384 |
+
"grad_norm": 0.03926938772201538,
|
| 385 |
+
"learning_rate": 9.973928157497675e-06,
|
| 386 |
+
"loss": 0.5195,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.4012158054711246,
|
| 391 |
+
"grad_norm": 0.049530286341905594,
|
| 392 |
+
"learning_rate": 9.969406417112489e-06,
|
| 393 |
+
"loss": 0.5854,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.4085106382978723,
|
| 398 |
+
"grad_norm": 0.08943431824445724,
|
| 399 |
+
"learning_rate": 9.964524472416319e-06,
|
| 400 |
+
"loss": 0.5706,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4158054711246201,
|
| 405 |
+
"grad_norm": 0.04114034026861191,
|
| 406 |
+
"learning_rate": 9.959282677273869e-06,
|
| 407 |
+
"loss": 0.4923,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.4231003039513678,
|
| 412 |
+
"grad_norm": 0.03834295645356178,
|
| 413 |
+
"learning_rate": 9.953681411633376e-06,
|
| 414 |
+
"loss": 0.5151,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.43039513677811553,
|
| 419 |
+
"grad_norm": 0.03940470516681671,
|
| 420 |
+
"learning_rate": 9.947721081499068e-06,
|
| 421 |
+
"loss": 0.5274,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4376899696048632,
|
| 426 |
+
"grad_norm": 0.05276661738753319,
|
| 427 |
+
"learning_rate": 9.941402118901743e-06,
|
| 428 |
+
"loss": 0.5312,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.4449848024316109,
|
| 433 |
+
"grad_norm": 0.04116562008857727,
|
| 434 |
+
"learning_rate": 9.934724981867447e-06,
|
| 435 |
+
"loss": 0.5073,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.4522796352583587,
|
| 440 |
+
"grad_norm": 0.039049182087183,
|
| 441 |
+
"learning_rate": 9.927690154384273e-06,
|
| 442 |
+
"loss": 0.5367,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.4595744680851064,
|
| 447 |
+
"grad_norm": 0.042383261024951935,
|
| 448 |
+
"learning_rate": 9.920298146367287e-06,
|
| 449 |
+
"loss": 0.5232,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.4668693009118541,
|
| 454 |
+
"grad_norm": 0.04153553023934364,
|
| 455 |
+
"learning_rate": 9.912549493621555e-06,
|
| 456 |
+
"loss": 0.5438,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.47416413373860183,
|
| 461 |
+
"grad_norm": 0.04116344451904297,
|
| 462 |
+
"learning_rate": 9.904444757803322e-06,
|
| 463 |
+
"loss": 0.4803,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.48145896656534953,
|
| 468 |
+
"grad_norm": 0.06467548757791519,
|
| 469 |
+
"learning_rate": 9.895984526379282e-06,
|
| 470 |
+
"loss": 0.5554,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.4887537993920973,
|
| 475 |
+
"grad_norm": 0.04420805722475052,
|
| 476 |
+
"learning_rate": 9.887169412584012e-06,
|
| 477 |
+
"loss": 0.5659,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.496048632218845,
|
| 482 |
+
"grad_norm": 0.04072507843375206,
|
| 483 |
+
"learning_rate": 9.878000055375512e-06,
|
| 484 |
+
"loss": 0.486,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.5033434650455927,
|
| 489 |
+
"grad_norm": 0.04508865624666214,
|
| 490 |
+
"learning_rate": 9.868477119388897e-06,
|
| 491 |
+
"loss": 0.5284,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.5106382978723404,
|
| 496 |
+
"grad_norm": 0.04231835529208183,
|
| 497 |
+
"learning_rate": 9.858601294888212e-06,
|
| 498 |
+
"loss": 0.5185,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.5179331306990882,
|
| 503 |
+
"grad_norm": 0.03981684520840645,
|
| 504 |
+
"learning_rate": 9.848373297716414e-06,
|
| 505 |
+
"loss": 0.5246,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.5252279635258359,
|
| 510 |
+
"grad_norm": 0.045293230563402176,
|
| 511 |
+
"learning_rate": 9.837793869243468e-06,
|
| 512 |
+
"loss": 0.5403,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.5325227963525836,
|
| 517 |
+
"grad_norm": 0.0415407195687294,
|
| 518 |
+
"learning_rate": 9.826863776312621e-06,
|
| 519 |
+
"loss": 0.568,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.5398176291793313,
|
| 524 |
+
"grad_norm": 0.04549698531627655,
|
| 525 |
+
"learning_rate": 9.815583811184809e-06,
|
| 526 |
+
"loss": 0.5547,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.547112462006079,
|
| 531 |
+
"grad_norm": 0.03895876184105873,
|
| 532 |
+
"learning_rate": 9.803954791481239e-06,
|
| 533 |
+
"loss": 0.5374,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5544072948328268,
|
| 538 |
+
"grad_norm": 0.046192847192287445,
|
| 539 |
+
"learning_rate": 9.79197756012412e-06,
|
| 540 |
+
"loss": 0.5561,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.5617021276595745,
|
| 545 |
+
"grad_norm": 0.03921407088637352,
|
| 546 |
+
"learning_rate": 9.779652985275562e-06,
|
| 547 |
+
"loss": 0.5488,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5689969604863222,
|
| 552 |
+
"grad_norm": 0.037232838571071625,
|
| 553 |
+
"learning_rate": 9.766981960274653e-06,
|
| 554 |
+
"loss": 0.4963,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5762917933130699,
|
| 559 |
+
"grad_norm": 0.05492810904979706,
|
| 560 |
+
"learning_rate": 9.753965403572703e-06,
|
| 561 |
+
"loss": 0.5621,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.5835866261398176,
|
| 566 |
+
"grad_norm": 0.04202823340892792,
|
| 567 |
+
"learning_rate": 9.740604258666668e-06,
|
| 568 |
+
"loss": 0.5479,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5908814589665653,
|
| 573 |
+
"grad_norm": 0.04189832881093025,
|
| 574 |
+
"learning_rate": 9.726899494030768e-06,
|
| 575 |
+
"loss": 0.5802,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5981762917933131,
|
| 580 |
+
"grad_norm": 0.039709825068712234,
|
| 581 |
+
"learning_rate": 9.712852103046281e-06,
|
| 582 |
+
"loss": 0.5166,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.6054711246200608,
|
| 587 |
+
"grad_norm": 0.04080045223236084,
|
| 588 |
+
"learning_rate": 9.698463103929542e-06,
|
| 589 |
+
"loss": 0.5289,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.6127659574468085,
|
| 594 |
+
"grad_norm": 0.038535572588443756,
|
| 595 |
+
"learning_rate": 9.68373353965814e-06,
|
| 596 |
+
"loss": 0.5352,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.6200607902735562,
|
| 601 |
+
"grad_norm": 0.04705570638179779,
|
| 602 |
+
"learning_rate": 9.66866447789531e-06,
|
| 603 |
+
"loss": 0.5235,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.6273556231003039,
|
| 608 |
+
"grad_norm": 0.042207516729831696,
|
| 609 |
+
"learning_rate": 9.65325701091256e-06,
|
| 610 |
+
"loss": 0.5147,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.6346504559270517,
|
| 615 |
+
"grad_norm": 0.04210168495774269,
|
| 616 |
+
"learning_rate": 9.637512255510475e-06,
|
| 617 |
+
"loss": 0.5241,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.6419452887537994,
|
| 622 |
+
"grad_norm": 0.03785989060997963,
|
| 623 |
+
"learning_rate": 9.62143135293779e-06,
|
| 624 |
+
"loss": 0.5429,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.6492401215805471,
|
| 629 |
+
"grad_norm": 0.04283512756228447,
|
| 630 |
+
"learning_rate": 9.605015468808651e-06,
|
| 631 |
+
"loss": 0.5242,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.6565349544072948,
|
| 636 |
+
"grad_norm": 0.043273307383060455,
|
| 637 |
+
"learning_rate": 9.588265793018141e-06,
|
| 638 |
+
"loss": 0.5455,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.6638297872340425,
|
| 643 |
+
"grad_norm": 0.04218590632081032,
|
| 644 |
+
"learning_rate": 9.571183539656011e-06,
|
| 645 |
+
"loss": 0.5778,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.6711246200607903,
|
| 650 |
+
"grad_norm": 0.03844400867819786,
|
| 651 |
+
"learning_rate": 9.553769946918698e-06,
|
| 652 |
+
"loss": 0.5233,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.678419452887538,
|
| 657 |
+
"grad_norm": 0.04001948982477188,
|
| 658 |
+
"learning_rate": 9.536026277019562e-06,
|
| 659 |
+
"loss": 0.5156,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6857142857142857,
|
| 664 |
+
"grad_norm": 0.04228726401925087,
|
| 665 |
+
"learning_rate": 9.517953816097396e-06,
|
| 666 |
+
"loss": 0.5138,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6930091185410334,
|
| 671 |
+
"grad_norm": 0.03879157081246376,
|
| 672 |
+
"learning_rate": 9.499553874123213e-06,
|
| 673 |
+
"loss": 0.4926,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.7003039513677811,
|
| 678 |
+
"grad_norm": 0.04016513749957085,
|
| 679 |
+
"learning_rate": 9.480827784805278e-06,
|
| 680 |
+
"loss": 0.497,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.7075987841945289,
|
| 685 |
+
"grad_norm": 0.03983764350414276,
|
| 686 |
+
"learning_rate": 9.461776905492446e-06,
|
| 687 |
+
"loss": 0.4852,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.7148936170212766,
|
| 692 |
+
"grad_norm": 0.056514669209718704,
|
| 693 |
+
"learning_rate": 9.442402617075765e-06,
|
| 694 |
+
"loss": 0.5288,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.7221884498480243,
|
| 699 |
+
"grad_norm": 0.046206481754779816,
|
| 700 |
+
"learning_rate": 9.422706323888398e-06,
|
| 701 |
+
"loss": 0.5418,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.729483282674772,
|
| 706 |
+
"grad_norm": 0.0474584735929966,
|
| 707 |
+
"learning_rate": 9.402689453603815e-06,
|
| 708 |
+
"loss": 0.5531,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.7367781155015197,
|
| 713 |
+
"grad_norm": 0.037484850734472275,
|
| 714 |
+
"learning_rate": 9.382353457132318e-06,
|
| 715 |
+
"loss": 0.4869,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.7440729483282674,
|
| 720 |
+
"grad_norm": 0.03749077394604683,
|
| 721 |
+
"learning_rate": 9.361699808515877e-06,
|
| 722 |
+
"loss": 0.5275,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.7513677811550152,
|
| 727 |
+
"grad_norm": 0.038470759987831116,
|
| 728 |
+
"learning_rate": 9.340730004821266e-06,
|
| 729 |
+
"loss": 0.5044,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.7586626139817629,
|
| 734 |
+
"grad_norm": 0.038027700036764145,
|
| 735 |
+
"learning_rate": 9.31944556603157e-06,
|
| 736 |
+
"loss": 0.5025,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.7659574468085106,
|
| 741 |
+
"grad_norm": 0.046422988176345825,
|
| 742 |
+
"learning_rate": 9.297848034936007e-06,
|
| 743 |
+
"loss": 0.5341,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.7732522796352583,
|
| 748 |
+
"grad_norm": 0.043657850474119186,
|
| 749 |
+
"learning_rate": 9.275938977018082e-06,
|
| 750 |
+
"loss": 0.5085,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.780547112462006,
|
| 755 |
+
"grad_norm": 0.04235101863741875,
|
| 756 |
+
"learning_rate": 9.253719980342134e-06,
|
| 757 |
+
"loss": 0.5397,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.7878419452887538,
|
| 762 |
+
"grad_norm": 0.04412844404578209,
|
| 763 |
+
"learning_rate": 9.231192655438222e-06,
|
| 764 |
+
"loss": 0.5522,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.7951367781155015,
|
| 769 |
+
"grad_norm": 0.04110129550099373,
|
| 770 |
+
"learning_rate": 9.208358635185372e-06,
|
| 771 |
+
"loss": 0.5785,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.8024316109422492,
|
| 776 |
+
"grad_norm": 0.03757128119468689,
|
| 777 |
+
"learning_rate": 9.185219574693242e-06,
|
| 778 |
+
"loss": 0.4777,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.8097264437689969,
|
| 783 |
+
"grad_norm": 0.03927014395594597,
|
| 784 |
+
"learning_rate": 9.161777151182137e-06,
|
| 785 |
+
"loss": 0.526,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.8170212765957446,
|
| 790 |
+
"grad_norm": 0.03983665257692337,
|
| 791 |
+
"learning_rate": 9.138033063861436e-06,
|
| 792 |
+
"loss": 0.5138,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.8243161094224924,
|
| 797 |
+
"grad_norm": 0.038819894194602966,
|
| 798 |
+
"learning_rate": 9.113989033806434e-06,
|
| 799 |
+
"loss": 0.494,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.8316109422492401,
|
| 804 |
+
"grad_norm": 0.05275421962141991,
|
| 805 |
+
"learning_rate": 9.089646803833589e-06,
|
| 806 |
+
"loss": 0.539,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.8389057750759878,
|
| 811 |
+
"grad_norm": 0.04078809916973114,
|
| 812 |
+
"learning_rate": 9.06500813837419e-06,
|
| 813 |
+
"loss": 0.4778,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.8462006079027355,
|
| 818 |
+
"grad_norm": 0.040415696799755096,
|
| 819 |
+
"learning_rate": 9.040074823346466e-06,
|
| 820 |
+
"loss": 0.5443,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.8534954407294832,
|
| 825 |
+
"grad_norm": 0.03834336996078491,
|
| 826 |
+
"learning_rate": 9.014848666026138e-06,
|
| 827 |
+
"loss": 0.4945,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.8607902735562311,
|
| 832 |
+
"grad_norm": 0.04293690249323845,
|
| 833 |
+
"learning_rate": 8.989331494915417e-06,
|
| 834 |
+
"loss": 0.5404,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.8680851063829788,
|
| 839 |
+
"grad_norm": 0.04116823151707649,
|
| 840 |
+
"learning_rate": 8.963525159610465e-06,
|
| 841 |
+
"loss": 0.5274,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.8753799392097265,
|
| 846 |
+
"grad_norm": 0.04143698886036873,
|
| 847 |
+
"learning_rate": 8.937431530667329e-06,
|
| 848 |
+
"loss": 0.4916,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.8826747720364742,
|
| 853 |
+
"grad_norm": 0.039400726556777954,
|
| 854 |
+
"learning_rate": 8.911052499466358e-06,
|
| 855 |
+
"loss": 0.5573,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.8899696048632219,
|
| 860 |
+
"grad_norm": 0.035739775747060776,
|
| 861 |
+
"learning_rate": 8.884389978075098e-06,
|
| 862 |
+
"loss": 0.4961,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.8972644376899696,
|
| 867 |
+
"grad_norm": 0.04682913422584534,
|
| 868 |
+
"learning_rate": 8.857445899109716e-06,
|
| 869 |
+
"loss": 0.4712,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.9045592705167174,
|
| 874 |
+
"grad_norm": 0.04531010612845421,
|
| 875 |
+
"learning_rate": 8.83022221559489e-06,
|
| 876 |
+
"loss": 0.5242,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.9118541033434651,
|
| 881 |
+
"grad_norm": 0.04528380185365677,
|
| 882 |
+
"learning_rate": 8.80272090082227e-06,
|
| 883 |
+
"loss": 0.5506,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.9191489361702128,
|
| 888 |
+
"grad_norm": 0.03824332728981972,
|
| 889 |
+
"learning_rate": 8.774943948207427e-06,
|
| 890 |
+
"loss": 0.4581,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.9264437689969605,
|
| 895 |
+
"grad_norm": 0.03896916285157204,
|
| 896 |
+
"learning_rate": 8.746893371145367e-06,
|
| 897 |
+
"loss": 0.5504,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.9337386018237082,
|
| 902 |
+
"grad_norm": 0.03746696934103966,
|
| 903 |
+
"learning_rate": 8.718571202864598e-06,
|
| 904 |
+
"loss": 0.4589,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.941033434650456,
|
| 909 |
+
"grad_norm": 0.04142184555530548,
|
| 910 |
+
"learning_rate": 8.689979496279747e-06,
|
| 911 |
+
"loss": 0.5299,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.9483282674772037,
|
| 916 |
+
"grad_norm": 0.03700762987136841,
|
| 917 |
+
"learning_rate": 8.661120323842751e-06,
|
| 918 |
+
"loss": 0.5159,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.9556231003039514,
|
| 923 |
+
"grad_norm": 0.036684855818748474,
|
| 924 |
+
"learning_rate": 8.631995777392645e-06,
|
| 925 |
+
"loss": 0.4854,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.9629179331306991,
|
| 930 |
+
"grad_norm": 0.06939133256673813,
|
| 931 |
+
"learning_rate": 8.602607968003935e-06,
|
| 932 |
+
"loss": 0.5101,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.9702127659574468,
|
| 937 |
+
"grad_norm": 0.039062708616256714,
|
| 938 |
+
"learning_rate": 8.572959025833573e-06,
|
| 939 |
+
"loss": 0.5005,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.9775075987841946,
|
| 944 |
+
"grad_norm": 0.04555986076593399,
|
| 945 |
+
"learning_rate": 8.543051099966558e-06,
|
| 946 |
+
"loss": 0.5465,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.9848024316109423,
|
| 951 |
+
"grad_norm": 0.04333364591002464,
|
| 952 |
+
"learning_rate": 8.512886358260162e-06,
|
| 953 |
+
"loss": 0.5237,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.99209726443769,
|
| 958 |
+
"grad_norm": 0.04095487669110298,
|
| 959 |
+
"learning_rate": 8.482466987186785e-06,
|
| 960 |
+
"loss": 0.5335,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.9993920972644377,
|
| 965 |
+
"grad_norm": 0.0442386157810688,
|
| 966 |
+
"learning_rate": 8.451795191675488e-06,
|
| 967 |
+
"loss": 0.5107,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.0,
|
| 972 |
+
"grad_norm": 0.0442386157810688,
|
| 973 |
+
"learning_rate": 8.420873194952153e-06,
|
| 974 |
+
"loss": 0.465,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.0072948328267477,
|
| 979 |
+
"grad_norm": 0.1445915699005127,
|
| 980 |
+
"learning_rate": 8.38970323837834e-06,
|
| 981 |
+
"loss": 0.4704,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.0145896656534954,
|
| 986 |
+
"grad_norm": 0.042022328823804855,
|
| 987 |
+
"learning_rate": 8.358287581288824e-06,
|
| 988 |
+
"loss": 0.4282,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.021884498480243,
|
| 993 |
+
"grad_norm": 0.04201134666800499,
|
| 994 |
+
"learning_rate": 8.326628500827826e-06,
|
| 995 |
+
"loss": 0.4539,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.0291793313069908,
|
| 1000 |
+
"grad_norm": 0.04877388849854469,
|
| 1001 |
+
"learning_rate": 8.294728291783967e-06,
|
| 1002 |
+
"loss": 0.4641,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.0364741641337385,
|
| 1007 |
+
"grad_norm": 0.046164825558662415,
|
| 1008 |
+
"learning_rate": 8.262589266423908e-06,
|
| 1009 |
+
"loss": 0.419,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.0437689969604864,
|
| 1014 |
+
"grad_norm": 0.041141681373119354,
|
| 1015 |
+
"learning_rate": 8.230213754324773e-06,
|
| 1016 |
+
"loss": 0.4224,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.0510638297872341,
|
| 1021 |
+
"grad_norm": 0.03967837244272232,
|
| 1022 |
+
"learning_rate": 8.19760410220527e-06,
|
| 1023 |
+
"loss": 0.4268,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.0583586626139818,
|
| 1028 |
+
"grad_norm": 0.05634555220603943,
|
| 1029 |
+
"learning_rate": 8.16476267375561e-06,
|
| 1030 |
+
"loss": 0.4025,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.0656534954407295,
|
| 1035 |
+
"grad_norm": 0.041606318205595016,
|
| 1036 |
+
"learning_rate": 8.131691849466154e-06,
|
| 1037 |
+
"loss": 0.4335,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.0729483282674772,
|
| 1042 |
+
"grad_norm": 0.03656647726893425,
|
| 1043 |
+
"learning_rate": 8.098394026454886e-06,
|
| 1044 |
+
"loss": 0.456,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.080243161094225,
|
| 1049 |
+
"grad_norm": 0.041005730628967285,
|
| 1050 |
+
"learning_rate": 8.064871618293647e-06,
|
| 1051 |
+
"loss": 0.3925,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.0875379939209726,
|
| 1056 |
+
"grad_norm": 0.04722120240330696,
|
| 1057 |
+
"learning_rate": 8.031127054833192e-06,
|
| 1058 |
+
"loss": 0.3981,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.0948328267477203,
|
| 1063 |
+
"grad_norm": 0.043071143329143524,
|
| 1064 |
+
"learning_rate": 7.997162782027061e-06,
|
| 1065 |
+
"loss": 0.4296,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.102127659574468,
|
| 1070 |
+
"grad_norm": 0.04518291354179382,
|
| 1071 |
+
"learning_rate": 7.962981261754295e-06,
|
| 1072 |
+
"loss": 0.4376,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.1094224924012157,
|
| 1077 |
+
"grad_norm": 0.04998685419559479,
|
| 1078 |
+
"learning_rate": 7.928584971640974e-06,
|
| 1079 |
+
"loss": 0.452,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.1167173252279636,
|
| 1084 |
+
"grad_norm": 0.04469837248325348,
|
| 1085 |
+
"learning_rate": 7.893976404880643e-06,
|
| 1086 |
+
"loss": 0.4316,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.1240121580547113,
|
| 1091 |
+
"grad_norm": 0.040255557745695114,
|
| 1092 |
+
"learning_rate": 7.859158070053578e-06,
|
| 1093 |
+
"loss": 0.4378,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.131306990881459,
|
| 1098 |
+
"grad_norm": 0.04467206820845604,
|
| 1099 |
+
"learning_rate": 7.824132490944968e-06,
|
| 1100 |
+
"loss": 0.4215,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.1386018237082067,
|
| 1105 |
+
"grad_norm": 0.03889721632003784,
|
| 1106 |
+
"learning_rate": 7.788902206361974e-06,
|
| 1107 |
+
"loss": 0.4257,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.1458966565349544,
|
| 1112 |
+
"grad_norm": 0.04140063747763634,
|
| 1113 |
+
"learning_rate": 7.753469769949701e-06,
|
| 1114 |
+
"loss": 0.4434,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.1531914893617021,
|
| 1119 |
+
"grad_norm": 0.039931997656822205,
|
| 1120 |
+
"learning_rate": 7.717837750006106e-06,
|
| 1121 |
+
"loss": 0.41,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.1604863221884498,
|
| 1126 |
+
"grad_norm": 0.03909624367952347,
|
| 1127 |
+
"learning_rate": 7.682008729295834e-06,
|
| 1128 |
+
"loss": 0.3904,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.1677811550151975,
|
| 1133 |
+
"grad_norm": 0.0401025116443634,
|
| 1134 |
+
"learning_rate": 7.645985304863004e-06,
|
| 1135 |
+
"loss": 0.4618,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.1750759878419452,
|
| 1140 |
+
"grad_norm": 0.04733911529183388,
|
| 1141 |
+
"learning_rate": 7.609770087842969e-06,
|
| 1142 |
+
"loss": 0.4247,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.182370820668693,
|
| 1147 |
+
"grad_norm": 0.037687744945287704,
|
| 1148 |
+
"learning_rate": 7.573365703273045e-06,
|
| 1149 |
+
"loss": 0.4071,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.1896656534954406,
|
| 1154 |
+
"grad_norm": 0.039216116070747375,
|
| 1155 |
+
"learning_rate": 7.536774789902246e-06,
|
| 1156 |
+
"loss": 0.4259,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.1969604863221885,
|
| 1161 |
+
"grad_norm": 0.0408397912979126,
|
| 1162 |
+
"learning_rate": 7.500000000000001e-06,
|
| 1163 |
+
"loss": 0.394,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.2042553191489362,
|
| 1168 |
+
"grad_norm": 0.04507288709282875,
|
| 1169 |
+
"learning_rate": 7.463043999163919e-06,
|
| 1170 |
+
"loss": 0.4605,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.211550151975684,
|
| 1175 |
+
"grad_norm": 0.03989469259977341,
|
| 1176 |
+
"learning_rate": 7.4259094661265685e-06,
|
| 1177 |
+
"loss": 0.4285,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.2188449848024316,
|
| 1182 |
+
"grad_norm": 0.0407419353723526,
|
| 1183 |
+
"learning_rate": 7.388599092561315e-06,
|
| 1184 |
+
"loss": 0.4204,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.2261398176291793,
|
| 1189 |
+
"grad_norm": 0.040525760501623154,
|
| 1190 |
+
"learning_rate": 7.351115582887212e-06,
|
| 1191 |
+
"loss": 0.4253,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.233434650455927,
|
| 1196 |
+
"grad_norm": 0.04370498284697533,
|
| 1197 |
+
"learning_rate": 7.313461654072974e-06,
|
| 1198 |
+
"loss": 0.4071,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.2407294832826747,
|
| 1203 |
+
"grad_norm": 0.0392344668507576,
|
| 1204 |
+
"learning_rate": 7.2756400354400445e-06,
|
| 1205 |
+
"loss": 0.4093,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.2480243161094224,
|
| 1210 |
+
"grad_norm": 0.03849213197827339,
|
| 1211 |
+
"learning_rate": 7.237653468464756e-06,
|
| 1212 |
+
"loss": 0.4157,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.2553191489361701,
|
| 1217 |
+
"grad_norm": 0.04228688403964043,
|
| 1218 |
+
"learning_rate": 7.199504706579617e-06,
|
| 1219 |
+
"loss": 0.482,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.262613981762918,
|
| 1224 |
+
"grad_norm": 0.037325162440538406,
|
| 1225 |
+
"learning_rate": 7.161196514973735e-06,
|
| 1226 |
+
"loss": 0.4224,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.2699088145896655,
|
| 1231 |
+
"grad_norm": 0.047044239938259125,
|
| 1232 |
+
"learning_rate": 7.122731670392381e-06,
|
| 1233 |
+
"loss": 0.4249,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.2772036474164135,
|
| 1238 |
+
"grad_norm": 0.04322784021496773,
|
| 1239 |
+
"learning_rate": 7.0841129609357165e-06,
|
| 1240 |
+
"loss": 0.4051,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.2844984802431612,
|
| 1245 |
+
"grad_norm": 0.041998326778411865,
|
| 1246 |
+
"learning_rate": 7.045343185856701e-06,
|
| 1247 |
+
"loss": 0.4106,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.2917933130699089,
|
| 1252 |
+
"grad_norm": 0.040727648884058,
|
| 1253 |
+
"learning_rate": 7.006425155358195e-06,
|
| 1254 |
+
"loss": 0.4427,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.2990881458966566,
|
| 1259 |
+
"grad_norm": 0.04059009999036789,
|
| 1260 |
+
"learning_rate": 6.967361690389258e-06,
|
| 1261 |
+
"loss": 0.437,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.3063829787234043,
|
| 1266 |
+
"grad_norm": 0.042023915797472,
|
| 1267 |
+
"learning_rate": 6.92815562244068e-06,
|
| 1268 |
+
"loss": 0.4315,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.313677811550152,
|
| 1273 |
+
"grad_norm": 0.04910752549767494,
|
| 1274 |
+
"learning_rate": 6.888809793339729e-06,
|
| 1275 |
+
"loss": 0.4436,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.3209726443768997,
|
| 1280 |
+
"grad_norm": 0.04180140420794487,
|
| 1281 |
+
"learning_rate": 6.849327055044182e-06,
|
| 1282 |
+
"loss": 0.3948,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.3282674772036474,
|
| 1287 |
+
"grad_norm": 0.03989269211888313,
|
| 1288 |
+
"learning_rate": 6.80971026943559e-06,
|
| 1289 |
+
"loss": 0.3929,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.335562310030395,
|
| 1294 |
+
"grad_norm": 0.04497074335813522,
|
| 1295 |
+
"learning_rate": 6.769962308111839e-06,
|
| 1296 |
+
"loss": 0.4429,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.342857142857143,
|
| 1301 |
+
"grad_norm": 0.04516409710049629,
|
| 1302 |
+
"learning_rate": 6.7300860521790034e-06,
|
| 1303 |
+
"loss": 0.4363,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.3501519756838904,
|
| 1308 |
+
"grad_norm": 0.041362229734659195,
|
| 1309 |
+
"learning_rate": 6.690084392042514e-06,
|
| 1310 |
+
"loss": 0.4058,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.3574468085106384,
|
| 1315 |
+
"grad_norm": 0.04281953349709511,
|
| 1316 |
+
"learning_rate": 6.649960227197648e-06,
|
| 1317 |
+
"loss": 0.423,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.364741641337386,
|
| 1322 |
+
"grad_norm": 0.046076931059360504,
|
| 1323 |
+
"learning_rate": 6.609716466019356e-06,
|
| 1324 |
+
"loss": 0.4427,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.3720364741641338,
|
| 1329 |
+
"grad_norm": 0.03960058465600014,
|
| 1330 |
+
"learning_rate": 6.569356025551454e-06,
|
| 1331 |
+
"loss": 0.4193,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.3793313069908815,
|
| 1336 |
+
"grad_norm": 0.044169649481773376,
|
| 1337 |
+
"learning_rate": 6.5288818312951886e-06,
|
| 1338 |
+
"loss": 0.4034,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.3866261398176292,
|
| 1343 |
+
"grad_norm": 0.04062066227197647,
|
| 1344 |
+
"learning_rate": 6.4882968169971734e-06,
|
| 1345 |
+
"loss": 0.4018,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.3939209726443769,
|
| 1350 |
+
"grad_norm": 0.04406093806028366,
|
| 1351 |
+
"learning_rate": 6.447603924436744e-06,
|
| 1352 |
+
"loss": 0.4498,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.4012158054711246,
|
| 1357 |
+
"grad_norm": 0.04197722300887108,
|
| 1358 |
+
"learning_rate": 6.406806103212725e-06,
|
| 1359 |
+
"loss": 0.4356,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.4085106382978723,
|
| 1364 |
+
"grad_norm": 0.04061530530452728,
|
| 1365 |
+
"learning_rate": 6.365906310529631e-06,
|
| 1366 |
+
"loss": 0.4441,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.41580547112462,
|
| 1371 |
+
"grad_norm": 0.046513479202985764,
|
| 1372 |
+
"learning_rate": 6.32490751098331e-06,
|
| 1373 |
+
"loss": 0.4166,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.4231003039513679,
|
| 1378 |
+
"grad_norm": 0.03948912024497986,
|
| 1379 |
+
"learning_rate": 6.2838126763460635e-06,
|
| 1380 |
+
"loss": 0.4478,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.4303951367781156,
|
| 1385 |
+
"grad_norm": 0.04548676684498787,
|
| 1386 |
+
"learning_rate": 6.2426247853512355e-06,
|
| 1387 |
+
"loss": 0.4653,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.4376899696048633,
|
| 1392 |
+
"grad_norm": 0.041050177067518234,
|
| 1393 |
+
"learning_rate": 6.2013468234773034e-06,
|
| 1394 |
+
"loss": 0.3953,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.444984802431611,
|
| 1399 |
+
"grad_norm": 0.03936685994267464,
|
| 1400 |
+
"learning_rate": 6.1599817827314744e-06,
|
| 1401 |
+
"loss": 0.4256,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.4522796352583587,
|
| 1406 |
+
"grad_norm": 0.041237395256757736,
|
| 1407 |
+
"learning_rate": 6.118532661432812e-06,
|
| 1408 |
+
"loss": 0.3892,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.4595744680851064,
|
| 1413 |
+
"grad_norm": 0.043174393475055695,
|
| 1414 |
+
"learning_rate": 6.077002463994908e-06,
|
| 1415 |
+
"loss": 0.4174,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.466869300911854,
|
| 1420 |
+
"grad_norm": 0.04198073223233223,
|
| 1421 |
+
"learning_rate": 6.035394200708104e-06,
|
| 1422 |
+
"loss": 0.4278,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.4741641337386018,
|
| 1427 |
+
"grad_norm": 0.045515723526477814,
|
| 1428 |
+
"learning_rate": 5.993710887521302e-06,
|
| 1429 |
+
"loss": 0.4346,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.4814589665653495,
|
| 1434 |
+
"grad_norm": 0.04443354532122612,
|
| 1435 |
+
"learning_rate": 5.951955545823342e-06,
|
| 1436 |
+
"loss": 0.4116,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.4887537993920974,
|
| 1441 |
+
"grad_norm": 0.04223044961690903,
|
| 1442 |
+
"learning_rate": 5.910131202224011e-06,
|
| 1443 |
+
"loss": 0.3844,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.4960486322188449,
|
| 1448 |
+
"grad_norm": 0.04305846244096756,
|
| 1449 |
+
"learning_rate": 5.8682408883346535e-06,
|
| 1450 |
+
"loss": 0.4414,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.5033434650455928,
|
| 1455 |
+
"grad_norm": 0.04148327186703682,
|
| 1456 |
+
"learning_rate": 5.826287640548425e-06,
|
| 1457 |
+
"loss": 0.4327,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.5106382978723403,
|
| 1462 |
+
"grad_norm": 0.0433870293200016,
|
| 1463 |
+
"learning_rate": 5.784274499820214e-06,
|
| 1464 |
+
"loss": 0.3787,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.5179331306990882,
|
| 1469 |
+
"grad_norm": 0.041102319955825806,
|
| 1470 |
+
"learning_rate": 5.742204511446203e-06,
|
| 1471 |
+
"loss": 0.4189,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.525227963525836,
|
| 1476 |
+
"grad_norm": 0.04175707325339317,
|
| 1477 |
+
"learning_rate": 5.7000807248431466e-06,
|
| 1478 |
+
"loss": 0.427,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.5325227963525836,
|
| 1483 |
+
"grad_norm": 0.04286907613277435,
|
| 1484 |
+
"learning_rate": 5.657906193327325e-06,
|
| 1485 |
+
"loss": 0.4,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.5398176291793313,
|
| 1490 |
+
"grad_norm": 0.04246861860156059,
|
| 1491 |
+
"learning_rate": 5.615683973893235e-06,
|
| 1492 |
+
"loss": 0.4097,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.547112462006079,
|
| 1497 |
+
"grad_norm": 0.03898885101079941,
|
| 1498 |
+
"learning_rate": 5.573417126992004e-06,
|
| 1499 |
+
"loss": 0.4237,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.554407294832827,
|
| 1504 |
+
"grad_norm": 0.04554813727736473,
|
| 1505 |
+
"learning_rate": 5.5311087163095475e-06,
|
| 1506 |
+
"loss": 0.436,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.5617021276595744,
|
| 1511 |
+
"grad_norm": 0.04189833253622055,
|
| 1512 |
+
"learning_rate": 5.4887618085445094e-06,
|
| 1513 |
+
"loss": 0.4121,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.5689969604863223,
|
| 1518 |
+
"grad_norm": 0.05306672677397728,
|
| 1519 |
+
"learning_rate": 5.446379473185972e-06,
|
| 1520 |
+
"loss": 0.4015,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.5762917933130698,
|
| 1525 |
+
"grad_norm": 0.04060041531920433,
|
| 1526 |
+
"learning_rate": 5.403964782290962e-06,
|
| 1527 |
+
"loss": 0.3967,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.5835866261398177,
|
| 1532 |
+
"grad_norm": 0.045451849699020386,
|
| 1533 |
+
"learning_rate": 5.361520810261779e-06,
|
| 1534 |
+
"loss": 0.4161,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.5908814589665652,
|
| 1539 |
+
"grad_norm": 0.043955542147159576,
|
| 1540 |
+
"learning_rate": 5.319050633623141e-06,
|
| 1541 |
+
"loss": 0.4205,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.598176291793313,
|
| 1546 |
+
"grad_norm": 0.040733452886343,
|
| 1547 |
+
"learning_rate": 5.276557330799203e-06,
|
| 1548 |
+
"loss": 0.4165,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.6054711246200608,
|
| 1553 |
+
"grad_norm": 0.04190356284379959,
|
| 1554 |
+
"learning_rate": 5.234043981890395e-06,
|
| 1555 |
+
"loss": 0.4515,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.6127659574468085,
|
| 1560 |
+
"grad_norm": 0.037713076919317245,
|
| 1561 |
+
"learning_rate": 5.191513668450178e-06,
|
| 1562 |
+
"loss": 0.4131,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.6200607902735562,
|
| 1567 |
+
"grad_norm": 0.038250233978033066,
|
| 1568 |
+
"learning_rate": 5.1489694732616805e-06,
|
| 1569 |
+
"loss": 0.4028,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.627355623100304,
|
| 1574 |
+
"grad_norm": 0.039751507341861725,
|
| 1575 |
+
"learning_rate": 5.106414480114238e-06,
|
| 1576 |
+
"loss": 0.4121,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.6346504559270518,
|
| 1581 |
+
"grad_norm": 0.044864848256111145,
|
| 1582 |
+
"learning_rate": 5.06385177357987e-06,
|
| 1583 |
+
"loss": 0.4708,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.6419452887537993,
|
| 1588 |
+
"grad_norm": 0.04169140383601189,
|
| 1589 |
+
"learning_rate": 5.021284438789694e-06,
|
| 1590 |
+
"loss": 0.425,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.6492401215805472,
|
| 1595 |
+
"grad_norm": 0.04238287732005119,
|
| 1596 |
+
"learning_rate": 4.9787155612103076e-06,
|
| 1597 |
+
"loss": 0.409,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.6565349544072947,
|
| 1602 |
+
"grad_norm": 0.03984750807285309,
|
| 1603 |
+
"learning_rate": 4.936148226420133e-06,
|
| 1604 |
+
"loss": 0.4451,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.6638297872340426,
|
| 1609 |
+
"grad_norm": 0.03823258727788925,
|
| 1610 |
+
"learning_rate": 4.893585519885764e-06,
|
| 1611 |
+
"loss": 0.4318,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.6711246200607903,
|
| 1616 |
+
"grad_norm": 0.043166667222976685,
|
| 1617 |
+
"learning_rate": 4.851030526738321e-06,
|
| 1618 |
+
"loss": 0.4348,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.678419452887538,
|
| 1623 |
+
"grad_norm": 0.04118693992495537,
|
| 1624 |
+
"learning_rate": 4.808486331549824e-06,
|
| 1625 |
+
"loss": 0.435,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.6857142857142857,
|
| 1630 |
+
"grad_norm": 0.040095556527376175,
|
| 1631 |
+
"learning_rate": 4.765956018109607e-06,
|
| 1632 |
+
"loss": 0.4506,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.6930091185410334,
|
| 1637 |
+
"grad_norm": 0.04523642733693123,
|
| 1638 |
+
"learning_rate": 4.7234426692007985e-06,
|
| 1639 |
+
"loss": 0.4394,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.7003039513677811,
|
| 1644 |
+
"grad_norm": 0.041244085878133774,
|
| 1645 |
+
"learning_rate": 4.680949366376858e-06,
|
| 1646 |
+
"loss": 0.4698,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.7075987841945288,
|
| 1651 |
+
"grad_norm": 0.04374610632658005,
|
| 1652 |
+
"learning_rate": 4.638479189738224e-06,
|
| 1653 |
+
"loss": 0.4129,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.7148936170212767,
|
| 1658 |
+
"grad_norm": 0.040487710386514664,
|
| 1659 |
+
"learning_rate": 4.596035217709039e-06,
|
| 1660 |
+
"loss": 0.4362,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.7221884498480242,
|
| 1665 |
+
"grad_norm": 0.044370926916599274,
|
| 1666 |
+
"learning_rate": 4.553620526814029e-06,
|
| 1667 |
+
"loss": 0.4155,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.7294832826747721,
|
| 1672 |
+
"grad_norm": 0.04036295786499977,
|
| 1673 |
+
"learning_rate": 4.511238191455491e-06,
|
| 1674 |
+
"loss": 0.4214,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.7367781155015196,
|
| 1679 |
+
"grad_norm": 0.03773313760757446,
|
| 1680 |
+
"learning_rate": 4.468891283690454e-06,
|
| 1681 |
+
"loss": 0.4298,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.7440729483282675,
|
| 1686 |
+
"grad_norm": 0.045683182775974274,
|
| 1687 |
+
"learning_rate": 4.426582873007999e-06,
|
| 1688 |
+
"loss": 0.4485,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.7513677811550152,
|
| 1693 |
+
"grad_norm": 0.04686903581023216,
|
| 1694 |
+
"learning_rate": 4.384316026106766e-06,
|
| 1695 |
+
"loss": 0.4303,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.758662613981763,
|
| 1700 |
+
"grad_norm": 0.045155324041843414,
|
| 1701 |
+
"learning_rate": 4.342093806672678e-06,
|
| 1702 |
+
"loss": 0.4409,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.7659574468085106,
|
| 1707 |
+
"grad_norm": 0.0418829619884491,
|
| 1708 |
+
"learning_rate": 4.299919275156857e-06,
|
| 1709 |
+
"loss": 0.4149,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.7732522796352583,
|
| 1714 |
+
"grad_norm": 0.041985101997852325,
|
| 1715 |
+
"learning_rate": 4.2577954885537985e-06,
|
| 1716 |
+
"loss": 0.4293,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.780547112462006,
|
| 1721 |
+
"grad_norm": 0.042692556977272034,
|
| 1722 |
+
"learning_rate": 4.215725500179788e-06,
|
| 1723 |
+
"loss": 0.4258,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.7878419452887537,
|
| 1728 |
+
"grad_norm": 0.04013342410326004,
|
| 1729 |
+
"learning_rate": 4.173712359451576e-06,
|
| 1730 |
+
"loss": 0.4015,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.7951367781155017,
|
| 1735 |
+
"grad_norm": 0.038998380303382874,
|
| 1736 |
+
"learning_rate": 4.131759111665349e-06,
|
| 1737 |
+
"loss": 0.4596,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.8024316109422491,
|
| 1742 |
+
"grad_norm": 0.039829254150390625,
|
| 1743 |
+
"learning_rate": 4.0898687977759895e-06,
|
| 1744 |
+
"loss": 0.4128,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.809726443768997,
|
| 1749 |
+
"grad_norm": 0.04312862455844879,
|
| 1750 |
+
"learning_rate": 4.048044454176658e-06,
|
| 1751 |
+
"loss": 0.4243,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.8170212765957445,
|
| 1756 |
+
"grad_norm": 0.04533419758081436,
|
| 1757 |
+
"learning_rate": 4.0062891124787e-06,
|
| 1758 |
+
"loss": 0.4414,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.8243161094224924,
|
| 1763 |
+
"grad_norm": 0.0438460148870945,
|
| 1764 |
+
"learning_rate": 3.964605799291897e-06,
|
| 1765 |
+
"loss": 0.4553,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.8316109422492401,
|
| 1770 |
+
"grad_norm": 0.0429726168513298,
|
| 1771 |
+
"learning_rate": 3.922997536005094e-06,
|
| 1772 |
+
"loss": 0.4311,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.8389057750759878,
|
| 1777 |
+
"grad_norm": 0.039493922144174576,
|
| 1778 |
+
"learning_rate": 3.88146733856719e-06,
|
| 1779 |
+
"loss": 0.4387,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.8462006079027355,
|
| 1784 |
+
"grad_norm": 0.04514075070619583,
|
| 1785 |
+
"learning_rate": 3.840018217268527e-06,
|
| 1786 |
+
"loss": 0.4442,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.8534954407294832,
|
| 1791 |
+
"grad_norm": 0.04080420732498169,
|
| 1792 |
+
"learning_rate": 3.7986531765226965e-06,
|
| 1793 |
+
"loss": 0.3884,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.8607902735562312,
|
| 1798 |
+
"grad_norm": 0.04457089304924011,
|
| 1799 |
+
"learning_rate": 3.757375214648764e-06,
|
| 1800 |
+
"loss": 0.3804,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.8680851063829786,
|
| 1805 |
+
"grad_norm": 0.044411323964595795,
|
| 1806 |
+
"learning_rate": 3.716187323653939e-06,
|
| 1807 |
+
"loss": 0.4164,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.8753799392097266,
|
| 1812 |
+
"grad_norm": 0.04527450352907181,
|
| 1813 |
+
"learning_rate": 3.675092489016693e-06,
|
| 1814 |
+
"loss": 0.4402,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.882674772036474,
|
| 1819 |
+
"grad_norm": 0.039380993694067,
|
| 1820 |
+
"learning_rate": 3.6340936894703717e-06,
|
| 1821 |
+
"loss": 0.4329,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.889969604863222,
|
| 1826 |
+
"grad_norm": 0.04769477993249893,
|
| 1827 |
+
"learning_rate": 3.593193896787277e-06,
|
| 1828 |
+
"loss": 0.4447,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.8972644376899694,
|
| 1833 |
+
"grad_norm": 0.04017976298928261,
|
| 1834 |
+
"learning_rate": 3.5523960755632573e-06,
|
| 1835 |
+
"loss": 0.4066,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.9045592705167174,
|
| 1840 |
+
"grad_norm": 0.04179855436086655,
|
| 1841 |
+
"learning_rate": 3.5117031830028274e-06,
|
| 1842 |
+
"loss": 0.4048,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.911854103343465,
|
| 1847 |
+
"grad_norm": 0.041397638618946075,
|
| 1848 |
+
"learning_rate": 3.4711181687048114e-06,
|
| 1849 |
+
"loss": 0.4296,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.9191489361702128,
|
| 1854 |
+
"grad_norm": 0.05517794191837311,
|
| 1855 |
+
"learning_rate": 3.4306439744485453e-06,
|
| 1856 |
+
"loss": 0.4266,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.9264437689969605,
|
| 1861 |
+
"grad_norm": 0.04322275519371033,
|
| 1862 |
+
"learning_rate": 3.3902835339806463e-06,
|
| 1863 |
+
"loss": 0.4084,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.9337386018237082,
|
| 1868 |
+
"grad_norm": 0.04079868271946907,
|
| 1869 |
+
"learning_rate": 3.3500397728023536e-06,
|
| 1870 |
+
"loss": 0.3987,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.941033434650456,
|
| 1875 |
+
"grad_norm": 0.040178705006837845,
|
| 1876 |
+
"learning_rate": 3.309915607957487e-06,
|
| 1877 |
+
"loss": 0.3508,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.9483282674772036,
|
| 1882 |
+
"grad_norm": 0.042311254888772964,
|
| 1883 |
+
"learning_rate": 3.2699139478209987e-06,
|
| 1884 |
+
"loss": 0.4315,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.9556231003039515,
|
| 1889 |
+
"grad_norm": 0.041265442967414856,
|
| 1890 |
+
"learning_rate": 3.2300376918881628e-06,
|
| 1891 |
+
"loss": 0.4096,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.962917933130699,
|
| 1896 |
+
"grad_norm": 0.035929929465055466,
|
| 1897 |
+
"learning_rate": 3.19028973056441e-06,
|
| 1898 |
+
"loss": 0.3872,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.9702127659574469,
|
| 1903 |
+
"grad_norm": 0.04031127318739891,
|
| 1904 |
+
"learning_rate": 3.150672944955818e-06,
|
| 1905 |
+
"loss": 0.4299,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.9775075987841946,
|
| 1910 |
+
"grad_norm": 0.043629132211208344,
|
| 1911 |
+
"learning_rate": 3.111190206660273e-06,
|
| 1912 |
+
"loss": 0.4371,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.9848024316109423,
|
| 1917 |
+
"grad_norm": 0.03935433551669121,
|
| 1918 |
+
"learning_rate": 3.0718443775593233e-06,
|
| 1919 |
+
"loss": 0.3912,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.99209726443769,
|
| 1924 |
+
"grad_norm": 0.04069478437304497,
|
| 1925 |
+
"learning_rate": 3.0326383096107424e-06,
|
| 1926 |
+
"loss": 0.416,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.9993920972644377,
|
| 1931 |
+
"grad_norm": 0.05225847661495209,
|
| 1932 |
+
"learning_rate": 2.993574844641807e-06,
|
| 1933 |
+
"loss": 0.3656,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.0,
|
| 1938 |
+
"grad_norm": 0.05225847661495209,
|
| 1939 |
+
"learning_rate": 2.9546568141433007e-06,
|
| 1940 |
+
"loss": 0.5271,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.007294832826748,
|
| 1945 |
+
"grad_norm": 0.15690822899341583,
|
| 1946 |
+
"learning_rate": 2.915887039064287e-06,
|
| 1947 |
+
"loss": 0.3677,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.0145896656534954,
|
| 1952 |
+
"grad_norm": 0.040792327374219894,
|
| 1953 |
+
"learning_rate": 2.8772683296076197e-06,
|
| 1954 |
+
"loss": 0.3493,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.0218844984802433,
|
| 1959 |
+
"grad_norm": 0.042940009385347366,
|
| 1960 |
+
"learning_rate": 2.838803485026265e-06,
|
| 1961 |
+
"loss": 0.3622,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 2.029179331306991,
|
| 1966 |
+
"grad_norm": 0.03872222825884819,
|
| 1967 |
+
"learning_rate": 2.800495293420384e-06,
|
| 1968 |
+
"loss": 0.3204,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 2.0364741641337387,
|
| 1973 |
+
"grad_norm": 0.03756758198142052,
|
| 1974 |
+
"learning_rate": 2.762346531535246e-06,
|
| 1975 |
+
"loss": 0.3158,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.043768996960486,
|
| 1980 |
+
"grad_norm": 0.04466132074594498,
|
| 1981 |
+
"learning_rate": 2.724359964559958e-06,
|
| 1982 |
+
"loss": 0.3638,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.051063829787234,
|
| 1987 |
+
"grad_norm": 0.04124055802822113,
|
| 1988 |
+
"learning_rate": 2.686538345927027e-06,
|
| 1989 |
+
"loss": 0.3231,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 2.0583586626139816,
|
| 1994 |
+
"grad_norm": 0.04917893931269646,
|
| 1995 |
+
"learning_rate": 2.6488844171127903e-06,
|
| 1996 |
+
"loss": 0.3683,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 2.0656534954407295,
|
| 2001 |
+
"grad_norm": 0.04118992015719414,
|
| 2002 |
+
"learning_rate": 2.611400907438685e-06,
|
| 2003 |
+
"loss": 0.294,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 2.072948328267477,
|
| 2008 |
+
"grad_norm": 0.04872892051935196,
|
| 2009 |
+
"learning_rate": 2.574090533873431e-06,
|
| 2010 |
+
"loss": 0.3156,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 2.080243161094225,
|
| 2015 |
+
"grad_norm": 0.04159025847911835,
|
| 2016 |
+
"learning_rate": 2.5369560008360826e-06,
|
| 2017 |
+
"loss": 0.3467,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.087537993920973,
|
| 2022 |
+
"grad_norm": 0.04014930874109268,
|
| 2023 |
+
"learning_rate": 2.5000000000000015e-06,
|
| 2024 |
+
"loss": 0.3303,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.0948328267477203,
|
| 2029 |
+
"grad_norm": 0.04120990261435509,
|
| 2030 |
+
"learning_rate": 2.4632252100977567e-06,
|
| 2031 |
+
"loss": 0.3086,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 2.1021276595744682,
|
| 2036 |
+
"grad_norm": 0.04079623147845268,
|
| 2037 |
+
"learning_rate": 2.426634296726955e-06,
|
| 2038 |
+
"loss": 0.3364,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.1094224924012157,
|
| 2043 |
+
"grad_norm": 0.04097681865096092,
|
| 2044 |
+
"learning_rate": 2.3902299121570332e-06,
|
| 2045 |
+
"loss": 0.3254,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 2.1167173252279636,
|
| 2050 |
+
"grad_norm": 0.04335152730345726,
|
| 2051 |
+
"learning_rate": 2.354014695136997e-06,
|
| 2052 |
+
"loss": 0.2875,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.124012158054711,
|
| 2057 |
+
"grad_norm": 0.041609250009059906,
|
| 2058 |
+
"learning_rate": 2.317991270704167e-06,
|
| 2059 |
+
"loss": 0.3585,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.131306990881459,
|
| 2064 |
+
"grad_norm": 0.03962058201432228,
|
| 2065 |
+
"learning_rate": 2.282162249993895e-06,
|
| 2066 |
+
"loss": 0.3094,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.1386018237082065,
|
| 2071 |
+
"grad_norm": 0.041423097252845764,
|
| 2072 |
+
"learning_rate": 2.2465302300503012e-06,
|
| 2073 |
+
"loss": 0.358,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.1458966565349544,
|
| 2078 |
+
"grad_norm": 0.044281214475631714,
|
| 2079 |
+
"learning_rate": 2.211097793638029e-06,
|
| 2080 |
+
"loss": 0.3575,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.153191489361702,
|
| 2085 |
+
"grad_norm": 0.04825511574745178,
|
| 2086 |
+
"learning_rate": 2.175867509055033e-06,
|
| 2087 |
+
"loss": 0.3364,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.16048632218845,
|
| 2092 |
+
"grad_norm": 0.04354681074619293,
|
| 2093 |
+
"learning_rate": 2.1408419299464245e-06,
|
| 2094 |
+
"loss": 0.2979,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.1677811550151977,
|
| 2099 |
+
"grad_norm": 0.042697008699178696,
|
| 2100 |
+
"learning_rate": 2.106023595119358e-06,
|
| 2101 |
+
"loss": 0.3356,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.1750759878419452,
|
| 2106 |
+
"grad_norm": 0.047276780009269714,
|
| 2107 |
+
"learning_rate": 2.071415028359026e-06,
|
| 2108 |
+
"loss": 0.3634,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.182370820668693,
|
| 2113 |
+
"grad_norm": 0.041072778403759,
|
| 2114 |
+
"learning_rate": 2.037018738245707e-06,
|
| 2115 |
+
"loss": 0.3362,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.1896656534954406,
|
| 2120 |
+
"grad_norm": 0.04692791774868965,
|
| 2121 |
+
"learning_rate": 2.0028372179729405e-06,
|
| 2122 |
+
"loss": 0.3511,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.1969604863221885,
|
| 2127 |
+
"grad_norm": 0.04538114741444588,
|
| 2128 |
+
"learning_rate": 1.9688729451668116e-06,
|
| 2129 |
+
"loss": 0.3336,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.204255319148936,
|
| 2134 |
+
"grad_norm": 0.044003862887620926,
|
| 2135 |
+
"learning_rate": 1.935128381706355e-06,
|
| 2136 |
+
"loss": 0.3349,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.211550151975684,
|
| 2141 |
+
"grad_norm": 0.045857448130846024,
|
| 2142 |
+
"learning_rate": 1.901605973545116e-06,
|
| 2143 |
+
"loss": 0.3537,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.2188449848024314,
|
| 2148 |
+
"grad_norm": 0.04272821545600891,
|
| 2149 |
+
"learning_rate": 1.8683081505338468e-06,
|
| 2150 |
+
"loss": 0.3373,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.2261398176291793,
|
| 2155 |
+
"grad_norm": 0.04273563250899315,
|
| 2156 |
+
"learning_rate": 1.8352373262443918e-06,
|
| 2157 |
+
"loss": 0.3436,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.2334346504559273,
|
| 2162 |
+
"grad_norm": 0.058361783623695374,
|
| 2163 |
+
"learning_rate": 1.8023958977947303e-06,
|
| 2164 |
+
"loss": 0.3245,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.2407294832826747,
|
| 2169 |
+
"grad_norm": 0.04002346843481064,
|
| 2170 |
+
"learning_rate": 1.7697862456752273e-06,
|
| 2171 |
+
"loss": 0.3317,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.2480243161094227,
|
| 2176 |
+
"grad_norm": 0.039896223694086075,
|
| 2177 |
+
"learning_rate": 1.7374107335760937e-06,
|
| 2178 |
+
"loss": 0.2976,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.25531914893617,
|
| 2183 |
+
"grad_norm": 0.04140615463256836,
|
| 2184 |
+
"learning_rate": 1.7052717082160348e-06,
|
| 2185 |
+
"loss": 0.3178,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.262613981762918,
|
| 2190 |
+
"grad_norm": 0.04329473525285721,
|
| 2191 |
+
"learning_rate": 1.6733714991721738e-06,
|
| 2192 |
+
"loss": 0.3187,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.2699088145896655,
|
| 2197 |
+
"grad_norm": 0.04440492019057274,
|
| 2198 |
+
"learning_rate": 1.6417124187111778e-06,
|
| 2199 |
+
"loss": 0.3194,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.2772036474164135,
|
| 2204 |
+
"grad_norm": 0.03996637463569641,
|
| 2205 |
+
"learning_rate": 1.610296761621662e-06,
|
| 2206 |
+
"loss": 0.3504,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.284498480243161,
|
| 2211 |
+
"grad_norm": 0.04870344325900078,
|
| 2212 |
+
"learning_rate": 1.5791268050478487e-06,
|
| 2213 |
+
"loss": 0.3599,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.291793313069909,
|
| 2218 |
+
"grad_norm": 0.049775756895542145,
|
| 2219 |
+
"learning_rate": 1.5482048083245116e-06,
|
| 2220 |
+
"loss": 0.3051,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.2990881458966568,
|
| 2225 |
+
"grad_norm": 0.04166199639439583,
|
| 2226 |
+
"learning_rate": 1.517533012813217e-06,
|
| 2227 |
+
"loss": 0.3383,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.3063829787234043,
|
| 2232 |
+
"grad_norm": 0.04483890160918236,
|
| 2233 |
+
"learning_rate": 1.4871136417398407e-06,
|
| 2234 |
+
"loss": 0.3302,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.3136778115501517,
|
| 2239 |
+
"grad_norm": 0.04276059940457344,
|
| 2240 |
+
"learning_rate": 1.4569489000334435e-06,
|
| 2241 |
+
"loss": 0.3615,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.3209726443768997,
|
| 2246 |
+
"grad_norm": 0.04011907801032066,
|
| 2247 |
+
"learning_rate": 1.427040974166427e-06,
|
| 2248 |
+
"loss": 0.3139,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.3282674772036476,
|
| 2253 |
+
"grad_norm": 0.043183084577322006,
|
| 2254 |
+
"learning_rate": 1.3973920319960654e-06,
|
| 2255 |
+
"loss": 0.3327,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.335562310030395,
|
| 2260 |
+
"grad_norm": 0.045110031962394714,
|
| 2261 |
+
"learning_rate": 1.3680042226073554e-06,
|
| 2262 |
+
"loss": 0.3183,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.342857142857143,
|
| 2267 |
+
"grad_norm": 0.04653245955705643,
|
| 2268 |
+
"learning_rate": 1.3388796761572493e-06,
|
| 2269 |
+
"loss": 0.3411,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.3501519756838904,
|
| 2274 |
+
"grad_norm": 0.04192928597331047,
|
| 2275 |
+
"learning_rate": 1.310020503720254e-06,
|
| 2276 |
+
"loss": 0.3722,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.3574468085106384,
|
| 2281 |
+
"grad_norm": 0.04330296441912651,
|
| 2282 |
+
"learning_rate": 1.2814287971354023e-06,
|
| 2283 |
+
"loss": 0.325,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.364741641337386,
|
| 2288 |
+
"grad_norm": 0.04404173046350479,
|
| 2289 |
+
"learning_rate": 1.253106628854635e-06,
|
| 2290 |
+
"loss": 0.3247,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.3720364741641338,
|
| 2295 |
+
"grad_norm": 0.04104992374777794,
|
| 2296 |
+
"learning_rate": 1.2250560517925747e-06,
|
| 2297 |
+
"loss": 0.3079,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.3793313069908812,
|
| 2302 |
+
"grad_norm": 0.04262121394276619,
|
| 2303 |
+
"learning_rate": 1.197279099177731e-06,
|
| 2304 |
+
"loss": 0.3446,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.386626139817629,
|
| 2309 |
+
"grad_norm": 0.04929178208112717,
|
| 2310 |
+
"learning_rate": 1.1697777844051105e-06,
|
| 2311 |
+
"loss": 0.3501,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.393920972644377,
|
| 2316 |
+
"grad_norm": 0.04329733923077583,
|
| 2317 |
+
"learning_rate": 1.1425541008902852e-06,
|
| 2318 |
+
"loss": 0.3213,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.4012158054711246,
|
| 2323 |
+
"grad_norm": 0.04333839192986488,
|
| 2324 |
+
"learning_rate": 1.1156100219249022e-06,
|
| 2325 |
+
"loss": 0.3232,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.4085106382978725,
|
| 2330 |
+
"grad_norm": 0.04259442910552025,
|
| 2331 |
+
"learning_rate": 1.0889475005336447e-06,
|
| 2332 |
+
"loss": 0.3632,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.41580547112462,
|
| 2337 |
+
"grad_norm": 0.04376016557216644,
|
| 2338 |
+
"learning_rate": 1.0625684693326727e-06,
|
| 2339 |
+
"loss": 0.3355,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.423100303951368,
|
| 2344 |
+
"grad_norm": 0.04070465639233589,
|
| 2345 |
+
"learning_rate": 1.0364748403895368e-06,
|
| 2346 |
+
"loss": 0.341,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.4303951367781154,
|
| 2351 |
+
"grad_norm": 0.041908472776412964,
|
| 2352 |
+
"learning_rate": 1.0106685050845838e-06,
|
| 2353 |
+
"loss": 0.3383,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.4376899696048633,
|
| 2358 |
+
"grad_norm": 0.04458033666014671,
|
| 2359 |
+
"learning_rate": 9.851513339738627e-07,
|
| 2360 |
+
"loss": 0.3246,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.4449848024316108,
|
| 2365 |
+
"grad_norm": 0.04246847331523895,
|
| 2366 |
+
"learning_rate": 9.599251766535344e-07,
|
| 2367 |
+
"loss": 0.3418,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.4522796352583587,
|
| 2372 |
+
"grad_norm": 0.04456906393170357,
|
| 2373 |
+
"learning_rate": 9.349918616258113e-07,
|
| 2374 |
+
"loss": 0.3536,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.4595744680851066,
|
| 2379 |
+
"grad_norm": 0.041079938411712646,
|
| 2380 |
+
"learning_rate": 9.10353196166412e-07,
|
| 2381 |
+
"loss": 0.3278,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.466869300911854,
|
| 2386 |
+
"grad_norm": 0.049959778785705566,
|
| 2387 |
+
"learning_rate": 8.860109661935673e-07,
|
| 2388 |
+
"loss": 0.3417,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.474164133738602,
|
| 2393 |
+
"grad_norm": 0.044310178607702255,
|
| 2394 |
+
"learning_rate": 8.619669361385663e-07,
|
| 2395 |
+
"loss": 0.3148,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.4814589665653495,
|
| 2400 |
+
"grad_norm": 0.04187872260808945,
|
| 2401 |
+
"learning_rate": 8.382228488178639e-07,
|
| 2402 |
+
"loss": 0.3392,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.4887537993920974,
|
| 2407 |
+
"grad_norm": 0.04105791822075844,
|
| 2408 |
+
"learning_rate": 8.147804253067581e-07,
|
| 2409 |
+
"loss": 0.3273,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.496048632218845,
|
| 2414 |
+
"grad_norm": 0.039138007909059525,
|
| 2415 |
+
"learning_rate": 7.916413648146282e-07,
|
| 2416 |
+
"loss": 0.3112,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.503343465045593,
|
| 2421 |
+
"grad_norm": 0.04459141194820404,
|
| 2422 |
+
"learning_rate": 7.6880734456178e-07,
|
| 2423 |
+
"loss": 0.3463,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 2.5106382978723403,
|
| 2428 |
+
"grad_norm": 0.043715398758649826,
|
| 2429 |
+
"learning_rate": 7.462800196578662e-07,
|
| 2430 |
+
"loss": 0.3439,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 2.517933130699088,
|
| 2435 |
+
"grad_norm": 0.043532464653253555,
|
| 2436 |
+
"learning_rate": 7.240610229819195e-07,
|
| 2437 |
+
"loss": 0.3303,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 2.525227963525836,
|
| 2442 |
+
"grad_norm": 0.04273553937673569,
|
| 2443 |
+
"learning_rate": 7.021519650639952e-07,
|
| 2444 |
+
"loss": 0.3171,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.5325227963525836,
|
| 2449 |
+
"grad_norm": 0.05441723391413689,
|
| 2450 |
+
"learning_rate": 6.805544339684295e-07,
|
| 2451 |
+
"loss": 0.3239,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.539817629179331,
|
| 2456 |
+
"grad_norm": 0.04585114121437073,
|
| 2457 |
+
"learning_rate": 6.592699951787362e-07,
|
| 2458 |
+
"loss": 0.3378,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.547112462006079,
|
| 2463 |
+
"grad_norm": 0.04242338612675667,
|
| 2464 |
+
"learning_rate": 6.383001914841252e-07,
|
| 2465 |
+
"loss": 0.2992,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 2.554407294832827,
|
| 2470 |
+
"grad_norm": 0.046155836433172226,
|
| 2471 |
+
"learning_rate": 6.17646542867682e-07,
|
| 2472 |
+
"loss": 0.3503,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 2.5617021276595744,
|
| 2477 |
+
"grad_norm": 0.04374154284596443,
|
| 2478 |
+
"learning_rate": 5.973105463961864e-07,
|
| 2479 |
+
"loss": 0.3385,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 2.5689969604863223,
|
| 2484 |
+
"grad_norm": 0.04297053441405296,
|
| 2485 |
+
"learning_rate": 5.772936761116027e-07,
|
| 2486 |
+
"loss": 0.3396,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.57629179331307,
|
| 2491 |
+
"grad_norm": 0.04214682802557945,
|
| 2492 |
+
"learning_rate": 5.575973829242365e-07,
|
| 2493 |
+
"loss": 0.3373,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 2.5835866261398177,
|
| 2498 |
+
"grad_norm": 0.04097369685769081,
|
| 2499 |
+
"learning_rate": 5.382230945075556e-07,
|
| 2500 |
+
"loss": 0.3386,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.590881458966565,
|
| 2505 |
+
"grad_norm": 0.042690787464380264,
|
| 2506 |
+
"learning_rate": 5.191722151947227e-07,
|
| 2507 |
+
"loss": 0.3319,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 2.598176291793313,
|
| 2512 |
+
"grad_norm": 0.040518004447221756,
|
| 2513 |
+
"learning_rate": 5.004461258767873e-07,
|
| 2514 |
+
"loss": 0.3174,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 2.6054711246200606,
|
| 2519 |
+
"grad_norm": 0.04100370407104492,
|
| 2520 |
+
"learning_rate": 4.820461839026047e-07,
|
| 2521 |
+
"loss": 0.34,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 2.6127659574468085,
|
| 2526 |
+
"grad_norm": 0.04036758467555046,
|
| 2527 |
+
"learning_rate": 4.639737229804403e-07,
|
| 2528 |
+
"loss": 0.3351,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 2.6200607902735564,
|
| 2533 |
+
"grad_norm": 0.04206588491797447,
|
| 2534 |
+
"learning_rate": 4.4623005308130243e-07,
|
| 2535 |
+
"loss": 0.3244,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 2.627355623100304,
|
| 2540 |
+
"grad_norm": 0.04280061274766922,
|
| 2541 |
+
"learning_rate": 4.2881646034398926e-07,
|
| 2542 |
+
"loss": 0.3065,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 2.634650455927052,
|
| 2547 |
+
"grad_norm": 0.04229553043842316,
|
| 2548 |
+
"learning_rate": 4.1173420698186027e-07,
|
| 2549 |
+
"loss": 0.3306,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 2.6419452887537993,
|
| 2554 |
+
"grad_norm": 0.044544368982315063,
|
| 2555 |
+
"learning_rate": 3.9498453119134917e-07,
|
| 2556 |
+
"loss": 0.3514,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 2.6492401215805472,
|
| 2561 |
+
"grad_norm": 0.045995116233825684,
|
| 2562 |
+
"learning_rate": 3.7856864706221187e-07,
|
| 2563 |
+
"loss": 0.3498,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 2.6565349544072947,
|
| 2568 |
+
"grad_norm": 0.048596058040857315,
|
| 2569 |
+
"learning_rate": 3.6248774448952695e-07,
|
| 2570 |
+
"loss": 0.3358,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 2.6638297872340426,
|
| 2575 |
+
"grad_norm": 0.04591159150004387,
|
| 2576 |
+
"learning_rate": 3.467429890874424e-07,
|
| 2577 |
+
"loss": 0.3129,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 2.67112462006079,
|
| 2582 |
+
"grad_norm": 0.041663773357868195,
|
| 2583 |
+
"learning_rate": 3.313355221046888e-07,
|
| 2584 |
+
"loss": 0.3213,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 2.678419452887538,
|
| 2589 |
+
"grad_norm": 0.04105694591999054,
|
| 2590 |
+
"learning_rate": 3.1626646034186084e-07,
|
| 2591 |
+
"loss": 0.345,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 2.685714285714286,
|
| 2596 |
+
"grad_norm": 0.044980090111494064,
|
| 2597 |
+
"learning_rate": 3.015368960704584e-07,
|
| 2598 |
+
"loss": 0.3265,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 2.6930091185410334,
|
| 2603 |
+
"grad_norm": 0.04313720017671585,
|
| 2604 |
+
"learning_rate": 2.871478969537206e-07,
|
| 2605 |
+
"loss": 0.3434,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 2.700303951367781,
|
| 2610 |
+
"grad_norm": 0.04387833923101425,
|
| 2611 |
+
"learning_rate": 2.7310050596923323e-07,
|
| 2612 |
+
"loss": 0.33,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 2.707598784194529,
|
| 2617 |
+
"grad_norm": 0.04352449253201485,
|
| 2618 |
+
"learning_rate": 2.593957413333331e-07,
|
| 2619 |
+
"loss": 0.3315,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 2.7148936170212767,
|
| 2624 |
+
"grad_norm": 0.04267902672290802,
|
| 2625 |
+
"learning_rate": 2.4603459642729867e-07,
|
| 2626 |
+
"loss": 0.3574,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 2.722188449848024,
|
| 2631 |
+
"grad_norm": 0.046177759766578674,
|
| 2632 |
+
"learning_rate": 2.330180397253473e-07,
|
| 2633 |
+
"loss": 0.3428,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 2.729483282674772,
|
| 2638 |
+
"grad_norm": 0.0405619777739048,
|
| 2639 |
+
"learning_rate": 2.2034701472443854e-07,
|
| 2640 |
+
"loss": 0.2903,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 2.7367781155015196,
|
| 2645 |
+
"grad_norm": 0.04294833540916443,
|
| 2646 |
+
"learning_rate": 2.0802243987588068e-07,
|
| 2647 |
+
"loss": 0.3664,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 2.7440729483282675,
|
| 2652 |
+
"grad_norm": 0.048284079879522324,
|
| 2653 |
+
"learning_rate": 1.9604520851876196e-07,
|
| 2654 |
+
"loss": 0.3294,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 2.7513677811550155,
|
| 2659 |
+
"grad_norm": 0.04744973033666611,
|
| 2660 |
+
"learning_rate": 1.8441618881519186e-07,
|
| 2661 |
+
"loss": 0.321,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 2.758662613981763,
|
| 2666 |
+
"grad_norm": 0.03917940333485603,
|
| 2667 |
+
"learning_rate": 1.7313622368738014e-07,
|
| 2668 |
+
"loss": 0.307,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 2.7659574468085104,
|
| 2673 |
+
"grad_norm": 0.04587104544043541,
|
| 2674 |
+
"learning_rate": 1.6220613075653201e-07,
|
| 2675 |
+
"loss": 0.3464,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 2.7732522796352583,
|
| 2680 |
+
"grad_norm": 0.042470064014196396,
|
| 2681 |
+
"learning_rate": 1.51626702283586e-07,
|
| 2682 |
+
"loss": 0.2907,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 2.7805471124620063,
|
| 2687 |
+
"grad_norm": 0.048345521092414856,
|
| 2688 |
+
"learning_rate": 1.4139870511178767e-07,
|
| 2689 |
+
"loss": 0.3481,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 2.7878419452887537,
|
| 2694 |
+
"grad_norm": 0.038930460810661316,
|
| 2695 |
+
"learning_rate": 1.3152288061110518e-07,
|
| 2696 |
+
"loss": 0.2987,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 2.7951367781155017,
|
| 2701 |
+
"grad_norm": 0.04472014680504799,
|
| 2702 |
+
"learning_rate": 1.2199994462448906e-07,
|
| 2703 |
+
"loss": 0.3612,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 2.802431610942249,
|
| 2708 |
+
"grad_norm": 0.05124653875827789,
|
| 2709 |
+
"learning_rate": 1.1283058741598962e-07,
|
| 2710 |
+
"loss": 0.3051,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 2.809726443768997,
|
| 2715 |
+
"grad_norm": 0.038610782474279404,
|
| 2716 |
+
"learning_rate": 1.0401547362071939e-07,
|
| 2717 |
+
"loss": 0.3362,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 2.8170212765957445,
|
| 2722 |
+
"grad_norm": 0.042344819754362106,
|
| 2723 |
+
"learning_rate": 9.555524219667989e-08,
|
| 2724 |
+
"loss": 0.3206,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 2.8243161094224924,
|
| 2729 |
+
"grad_norm": 0.046877894550561905,
|
| 2730 |
+
"learning_rate": 8.745050637844532e-08,
|
| 2731 |
+
"loss": 0.3332,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 2.83161094224924,
|
| 2736 |
+
"grad_norm": 0.04104023799300194,
|
| 2737 |
+
"learning_rate": 7.970185363271432e-08,
|
| 2738 |
+
"loss": 0.2941,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 2.838905775075988,
|
| 2743 |
+
"grad_norm": 0.04339218884706497,
|
| 2744 |
+
"learning_rate": 7.230984561572729e-08,
|
| 2745 |
+
"loss": 0.3409,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 2.8462006079027358,
|
| 2750 |
+
"grad_norm": 0.047086864709854126,
|
| 2751 |
+
"learning_rate": 6.527501813255344e-08,
|
| 2752 |
+
"loss": 0.3282,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 2.8534954407294832,
|
| 2757 |
+
"grad_norm": 0.04140612855553627,
|
| 2758 |
+
"learning_rate": 5.8597881098257924e-08,
|
| 2759 |
+
"loss": 0.3706,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 2.860790273556231,
|
| 2764 |
+
"grad_norm": 0.045086655765771866,
|
| 2765 |
+
"learning_rate": 5.227891850093314e-08,
|
| 2766 |
+
"loss": 0.3159,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 2.8680851063829786,
|
| 2771 |
+
"grad_norm": 0.04669662564992905,
|
| 2772 |
+
"learning_rate": 4.631858836662562e-08,
|
| 2773 |
+
"loss": 0.3212,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 2.8753799392097266,
|
| 2778 |
+
"grad_norm": 0.048425789922475815,
|
| 2779 |
+
"learning_rate": 4.071732272613149e-08,
|
| 2780 |
+
"loss": 0.3781,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 2.882674772036474,
|
| 2785 |
+
"grad_norm": 0.0414654016494751,
|
| 2786 |
+
"learning_rate": 3.5475527583681005e-08,
|
| 2787 |
+
"loss": 0.3487,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 2.889969604863222,
|
| 2792 |
+
"grad_norm": 0.04240646958351135,
|
| 2793 |
+
"learning_rate": 3.059358288751202e-08,
|
| 2794 |
+
"loss": 0.3063,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 2.8972644376899694,
|
| 2799 |
+
"grad_norm": 0.047270409762859344,
|
| 2800 |
+
"learning_rate": 2.6071842502326526e-08,
|
| 2801 |
+
"loss": 0.3352,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 2.9045592705167174,
|
| 2806 |
+
"grad_norm": 0.043454963713884354,
|
| 2807 |
+
"learning_rate": 2.1910634183644475e-08,
|
| 2808 |
+
"loss": 0.3442,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 2.9118541033434653,
|
| 2813 |
+
"grad_norm": 0.045333363115787506,
|
| 2814 |
+
"learning_rate": 1.811025955404333e-08,
|
| 2815 |
+
"loss": 0.3196,
|
| 2816 |
+
"step": 401
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 2.9191489361702128,
|
| 2820 |
+
"grad_norm": 0.04678984358906746,
|
| 2821 |
+
"learning_rate": 1.4670994081297796e-08,
|
| 2822 |
+
"loss": 0.3319,
|
| 2823 |
+
"step": 402
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 2.9264437689969602,
|
| 2827 |
+
"grad_norm": 0.05119337886571884,
|
| 2828 |
+
"learning_rate": 1.159308705841078e-08,
|
| 2829 |
+
"loss": 0.3614,
|
| 2830 |
+
"step": 403
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 2.933738601823708,
|
| 2834 |
+
"grad_norm": 0.048367924988269806,
|
| 2835 |
+
"learning_rate": 8.87676158554507e-09,
|
| 2836 |
+
"loss": 0.3615,
|
| 2837 |
+
"step": 404
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 2.941033434650456,
|
| 2841 |
+
"grad_norm": 0.04424307495355606,
|
| 2842 |
+
"learning_rate": 6.5222145538501595e-09,
|
| 2843 |
+
"loss": 0.3027,
|
| 2844 |
+
"step": 405
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 2.9483282674772036,
|
| 2848 |
+
"grad_norm": 0.0484929159283638,
|
| 2849 |
+
"learning_rate": 4.5296166311931125e-09,
|
| 2850 |
+
"loss": 0.3259,
|
| 2851 |
+
"step": 406
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 2.9556231003039515,
|
| 2855 |
+
"grad_norm": 0.04274160414934158,
|
| 2856 |
+
"learning_rate": 2.899112249786229e-09,
|
| 2857 |
+
"loss": 0.3219,
|
| 2858 |
+
"step": 407
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 2.962917933130699,
|
| 2862 |
+
"grad_norm": 0.04609229788184166,
|
| 2863 |
+
"learning_rate": 1.6308195957182028e-09,
|
| 2864 |
+
"loss": 0.3074,
|
| 2865 |
+
"step": 408
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 2.970212765957447,
|
| 2869 |
+
"grad_norm": 0.05541510134935379,
|
| 2870 |
+
"learning_rate": 7.24830600386528e-10,
|
| 2871 |
+
"loss": 0.3853,
|
| 2872 |
+
"step": 409
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 2.977507598784195,
|
| 2876 |
+
"grad_norm": 0.0421581007540226,
|
| 2877 |
+
"learning_rate": 1.812109338367174e-10,
|
| 2878 |
+
"loss": 0.3192,
|
| 2879 |
+
"step": 410
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 2.9848024316109423,
|
| 2883 |
+
"grad_norm": 0.04373352229595184,
|
| 2884 |
+
"learning_rate": 0.0,
|
| 2885 |
+
"loss": 0.3766,
|
| 2886 |
+
"step": 411
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 2.9848024316109423,
|
| 2890 |
+
"step": 411,
|
| 2891 |
+
"total_flos": 3.260394272163103e+17,
|
| 2892 |
+
"train_loss": 0.4436677635586175,
|
| 2893 |
+
"train_runtime": 142049.243,
|
| 2894 |
+
"train_samples_per_second": 0.139,
|
| 2895 |
+
"train_steps_per_second": 0.003
|
| 2896 |
+
}
|
| 2897 |
+
],
|
| 2898 |
+
"logging_steps": 1,
|
| 2899 |
+
"max_steps": 411,
|
| 2900 |
+
"num_input_tokens_seen": 0,
|
| 2901 |
+
"num_train_epochs": 3,
|
| 2902 |
+
"save_steps": 100,
|
| 2903 |
+
"stateful_callbacks": {
|
| 2904 |
+
"TrainerControl": {
|
| 2905 |
+
"args": {
|
| 2906 |
+
"should_epoch_stop": false,
|
| 2907 |
+
"should_evaluate": false,
|
| 2908 |
+
"should_log": false,
|
| 2909 |
+
"should_save": true,
|
| 2910 |
+
"should_training_stop": true
|
| 2911 |
+
},
|
| 2912 |
+
"attributes": {}
|
| 2913 |
+
}
|
| 2914 |
+
},
|
| 2915 |
+
"total_flos": 3.260394272163103e+17,
|
| 2916 |
+
"train_batch_size": 1,
|
| 2917 |
+
"trial_name": null,
|
| 2918 |
+
"trial_params": null
|
| 2919 |
+
}
|
training_loss.png
ADDED
|