Yichen Feng commited on
End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1156 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen3.5-35B-A3B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: Qwen3.5-35B-A3B-SFT-artarena_sft-LR1.0e-6-EPOCHS3-LF
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# Qwen3.5-35B-A3B-SFT-artarena_sft-LR1.0e-6-EPOCHS3-LF
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen3.5-35B-A3B](https://huggingface.co/Qwen/Qwen3.5-35B-A3B) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen3.5-35B-A3B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: Qwen3.5-35B-A3B-SFT-artarena_sft-LR1.0e-6-EPOCHS3-LF
|
|
|
|
| 16 |
|
| 17 |
# Qwen3.5-35B-A3B-SFT-artarena_sft-LR1.0e-6-EPOCHS3-LF
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen3.5-35B-A3B](https://huggingface.co/Qwen/Qwen3.5-35B-A3B) on the artarena_sft dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 23335512768512.0,
|
| 4 |
+
"train_loss": 0.8809327138294963,
|
| 5 |
+
"train_runtime": 1440.1859,
|
| 6 |
+
"train_samples_per_second": 3.485,
|
| 7 |
+
"train_steps_per_second": 0.11
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 23335512768512.0,
|
| 4 |
+
"train_loss": 0.8809327138294963,
|
| 5 |
+
"train_runtime": 1440.1859,
|
| 6 |
+
"train_samples_per_second": 3.485,
|
| 7 |
+
"train_steps_per_second": 0.11
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1156 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 159,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.018867924528301886,
|
| 14 |
+
"grad_norm": 15.087599797333977,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 1.3308970928192139,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.03773584905660377,
|
| 21 |
+
"grad_norm": 13.905749218469701,
|
| 22 |
+
"learning_rate": 1e-07,
|
| 23 |
+
"loss": 1.285962700843811,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.05660377358490566,
|
| 28 |
+
"grad_norm": 12.056842971309038,
|
| 29 |
+
"learning_rate": 2e-07,
|
| 30 |
+
"loss": 1.3014495372772217,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.07547169811320754,
|
| 35 |
+
"grad_norm": 13.244790508016711,
|
| 36 |
+
"learning_rate": 3e-07,
|
| 37 |
+
"loss": 1.306698203086853,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09433962264150944,
|
| 42 |
+
"grad_norm": 13.343404572220207,
|
| 43 |
+
"learning_rate": 4e-07,
|
| 44 |
+
"loss": 1.316192388534546,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.11320754716981132,
|
| 49 |
+
"grad_norm": 11.45228080585811,
|
| 50 |
+
"learning_rate": 5e-07,
|
| 51 |
+
"loss": 1.3045374155044556,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.1320754716981132,
|
| 56 |
+
"grad_norm": 14.318691524925551,
|
| 57 |
+
"learning_rate": 6e-07,
|
| 58 |
+
"loss": 1.3311705589294434,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1509433962264151,
|
| 63 |
+
"grad_norm": 12.050953431316902,
|
| 64 |
+
"learning_rate": 7e-07,
|
| 65 |
+
"loss": 1.2908077239990234,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.16981132075471697,
|
| 70 |
+
"grad_norm": 10.724349690276135,
|
| 71 |
+
"learning_rate": 8e-07,
|
| 72 |
+
"loss": 1.3058435916900635,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.18867924528301888,
|
| 77 |
+
"grad_norm": 9.473154789049158,
|
| 78 |
+
"learning_rate": 9e-07,
|
| 79 |
+
"loss": 1.2856130599975586,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.20754716981132076,
|
| 84 |
+
"grad_norm": 6.5491980764342275,
|
| 85 |
+
"learning_rate": 1e-06,
|
| 86 |
+
"loss": 1.2199636697769165,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.22641509433962265,
|
| 91 |
+
"grad_norm": 6.462810113541478,
|
| 92 |
+
"learning_rate": 9.99888864929809e-07,
|
| 93 |
+
"loss": 1.1673463582992554,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.24528301886792453,
|
| 98 |
+
"grad_norm": 6.9296822460672045,
|
| 99 |
+
"learning_rate": 9.995555091232516e-07,
|
| 100 |
+
"loss": 1.1699671745300293,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2641509433962264,
|
| 105 |
+
"grad_norm": 6.0515568106146596,
|
| 106 |
+
"learning_rate": 9.990000807704114e-07,
|
| 107 |
+
"loss": 1.1814613342285156,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.2830188679245283,
|
| 112 |
+
"grad_norm": 4.743020637878028,
|
| 113 |
+
"learning_rate": 9.982228267815643e-07,
|
| 114 |
+
"loss": 1.0652694702148438,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3018867924528302,
|
| 119 |
+
"grad_norm": 4.526630266791274,
|
| 120 |
+
"learning_rate": 9.972240926774166e-07,
|
| 121 |
+
"loss": 1.0635337829589844,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.32075471698113206,
|
| 126 |
+
"grad_norm": 4.609514753545406,
|
| 127 |
+
"learning_rate": 9.96004322435508e-07,
|
| 128 |
+
"loss": 1.0902111530303955,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.33962264150943394,
|
| 133 |
+
"grad_norm": 4.34097054100359,
|
| 134 |
+
"learning_rate": 9.945640582928437e-07,
|
| 135 |
+
"loss": 1.06702721118927,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.3584905660377358,
|
| 140 |
+
"grad_norm": 3.864434007517437,
|
| 141 |
+
"learning_rate": 9.9290394050485e-07,
|
| 142 |
+
"loss": 1.0476477146148682,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.37735849056603776,
|
| 147 |
+
"grad_norm": 3.8969527857656656,
|
| 148 |
+
"learning_rate": 9.91024707060755e-07,
|
| 149 |
+
"loss": 1.0617330074310303,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.39622641509433965,
|
| 154 |
+
"grad_norm": 3.911924948199517,
|
| 155 |
+
"learning_rate": 9.889271933555212e-07,
|
| 156 |
+
"loss": 1.07832932472229,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.41509433962264153,
|
| 161 |
+
"grad_norm": 3.833567236811609,
|
| 162 |
+
"learning_rate": 9.8661233181848e-07,
|
| 163 |
+
"loss": 1.0324124097824097,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.4339622641509434,
|
| 168 |
+
"grad_norm": 3.6955004487569396,
|
| 169 |
+
"learning_rate": 9.840811514988293e-07,
|
| 170 |
+
"loss": 0.9815853834152222,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.4528301886792453,
|
| 175 |
+
"grad_norm": 4.023404125176107,
|
| 176 |
+
"learning_rate": 9.813347776081788e-07,
|
| 177 |
+
"loss": 1.0266845226287842,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.4716981132075472,
|
| 182 |
+
"grad_norm": 3.712461139695743,
|
| 183 |
+
"learning_rate": 9.78374431020349e-07,
|
| 184 |
+
"loss": 1.0085935592651367,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.49056603773584906,
|
| 189 |
+
"grad_norm": 3.7543864084874596,
|
| 190 |
+
"learning_rate": 9.752014277286431e-07,
|
| 191 |
+
"loss": 0.9968965649604797,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5094339622641509,
|
| 196 |
+
"grad_norm": 3.8046734306564467,
|
| 197 |
+
"learning_rate": 9.718171782608353e-07,
|
| 198 |
+
"loss": 0.9803509712219238,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5283018867924528,
|
| 203 |
+
"grad_norm": 3.6105782650336433,
|
| 204 |
+
"learning_rate": 9.682231870521345e-07,
|
| 205 |
+
"loss": 0.9759021997451782,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5471698113207547,
|
| 210 |
+
"grad_norm": 3.3896428780092753,
|
| 211 |
+
"learning_rate": 9.644210517764013e-07,
|
| 212 |
+
"loss": 0.9812103509902954,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5660377358490566,
|
| 217 |
+
"grad_norm": 3.118079780719029,
|
| 218 |
+
"learning_rate": 9.60412462635919e-07,
|
| 219 |
+
"loss": 0.9091012477874756,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5849056603773585,
|
| 224 |
+
"grad_norm": 3.3662986364845,
|
| 225 |
+
"learning_rate": 9.561992016100291e-07,
|
| 226 |
+
"loss": 0.9503388404846191,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.6037735849056604,
|
| 231 |
+
"grad_norm": 2.9779547004368196,
|
| 232 |
+
"learning_rate": 9.517831416629716e-07,
|
| 233 |
+
"loss": 0.9247981309890747,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6226415094339622,
|
| 238 |
+
"grad_norm": 3.468415170701323,
|
| 239 |
+
"learning_rate": 9.471662459112745e-07,
|
| 240 |
+
"loss": 0.9473499655723572,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.6415094339622641,
|
| 245 |
+
"grad_norm": 2.8573918489427688,
|
| 246 |
+
"learning_rate": 9.423505667510723e-07,
|
| 247 |
+
"loss": 0.9340516328811646,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.660377358490566,
|
| 252 |
+
"grad_norm": 2.949529150108781,
|
| 253 |
+
"learning_rate": 9.373382449457303e-07,
|
| 254 |
+
"loss": 0.9248940348625183,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6792452830188679,
|
| 259 |
+
"grad_norm": 2.9658340262784697,
|
| 260 |
+
"learning_rate": 9.321315086741915e-07,
|
| 261 |
+
"loss": 0.9420664310455322,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.6981132075471698,
|
| 266 |
+
"grad_norm": 3.019712899281778,
|
| 267 |
+
"learning_rate": 9.267326725404598e-07,
|
| 268 |
+
"loss": 0.9231287240982056,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.7169811320754716,
|
| 273 |
+
"grad_norm": 2.827563138085356,
|
| 274 |
+
"learning_rate": 9.21144136544666e-07,
|
| 275 |
+
"loss": 0.9293084740638733,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7358490566037735,
|
| 280 |
+
"grad_norm": 3.126960585054511,
|
| 281 |
+
"learning_rate": 9.153683850161705e-07,
|
| 282 |
+
"loss": 0.9372609853744507,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.7547169811320755,
|
| 287 |
+
"grad_norm": 2.7757572634358456,
|
| 288 |
+
"learning_rate": 9.094079855091797e-07,
|
| 289 |
+
"loss": 0.9204014539718628,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.7735849056603774,
|
| 294 |
+
"grad_norm": 2.86268897243828,
|
| 295 |
+
"learning_rate": 9.032655876613635e-07,
|
| 296 |
+
"loss": 0.9143469333648682,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.7924528301886793,
|
| 301 |
+
"grad_norm": 2.899411491265449,
|
| 302 |
+
"learning_rate": 8.96943922015986e-07,
|
| 303 |
+
"loss": 0.901626467704773,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.8113207547169812,
|
| 308 |
+
"grad_norm": 3.0296165470958494,
|
| 309 |
+
"learning_rate": 8.90445798808068e-07,
|
| 310 |
+
"loss": 0.9193109273910522,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.8301886792452831,
|
| 315 |
+
"grad_norm": 2.832066082274235,
|
| 316 |
+
"learning_rate": 8.837741067151249e-07,
|
| 317 |
+
"loss": 0.9078618288040161,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.8490566037735849,
|
| 322 |
+
"grad_norm": 2.9792386000035083,
|
| 323 |
+
"learning_rate": 8.769318115730328e-07,
|
| 324 |
+
"loss": 0.9032235145568848,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.8679245283018868,
|
| 329 |
+
"grad_norm": 2.8570785041355373,
|
| 330 |
+
"learning_rate": 8.699219550575952e-07,
|
| 331 |
+
"loss": 0.8799638152122498,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8867924528301887,
|
| 336 |
+
"grad_norm": 2.8898604537645185,
|
| 337 |
+
"learning_rate": 8.627476533323956e-07,
|
| 338 |
+
"loss": 0.9072629809379578,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.9056603773584906,
|
| 343 |
+
"grad_norm": 2.819489131324746,
|
| 344 |
+
"learning_rate": 8.554120956635374e-07,
|
| 345 |
+
"loss": 0.879642128944397,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9245283018867925,
|
| 350 |
+
"grad_norm": 2.884576949261456,
|
| 351 |
+
"learning_rate": 8.479185430018858e-07,
|
| 352 |
+
"loss": 0.9129672050476074,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.9433962264150944,
|
| 357 |
+
"grad_norm": 2.8206974490824663,
|
| 358 |
+
"learning_rate": 8.402703265334454e-07,
|
| 359 |
+
"loss": 0.9072036147117615,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.9622641509433962,
|
| 364 |
+
"grad_norm": 2.8666837714043414,
|
| 365 |
+
"learning_rate": 8.324708461985124e-07,
|
| 366 |
+
"loss": 0.8936312198638916,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.9811320754716981,
|
| 371 |
+
"grad_norm": 2.75278105425475,
|
| 372 |
+
"learning_rate": 8.245235691802643e-07,
|
| 373 |
+
"loss": 0.886029839515686,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.0,
|
| 378 |
+
"grad_norm": 2.9063116637756807,
|
| 379 |
+
"learning_rate": 8.164320283634585e-07,
|
| 380 |
+
"loss": 0.886949360370636,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.0188679245283019,
|
| 385 |
+
"grad_norm": 2.8027377644406104,
|
| 386 |
+
"learning_rate": 8.081998207639212e-07,
|
| 387 |
+
"loss": 0.8734487891197205,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.0377358490566038,
|
| 392 |
+
"grad_norm": 2.975237594360833,
|
| 393 |
+
"learning_rate": 7.998306059295302e-07,
|
| 394 |
+
"loss": 0.8541756868362427,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0566037735849056,
|
| 399 |
+
"grad_norm": 2.7212092257296785,
|
| 400 |
+
"learning_rate": 7.913281043133977e-07,
|
| 401 |
+
"loss": 0.855162501335144,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0754716981132075,
|
| 406 |
+
"grad_norm": 4.004522306787069,
|
| 407 |
+
"learning_rate": 7.826960956199794e-07,
|
| 408 |
+
"loss": 0.8469276428222656,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0943396226415094,
|
| 413 |
+
"grad_norm": 2.789521379215554,
|
| 414 |
+
"learning_rate": 7.739384171248434e-07,
|
| 415 |
+
"loss": 0.8612252473831177,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1132075471698113,
|
| 420 |
+
"grad_norm": 3.0001618191920008,
|
| 421 |
+
"learning_rate": 7.650589619688468e-07,
|
| 422 |
+
"loss": 0.8504967093467712,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.1320754716981132,
|
| 427 |
+
"grad_norm": 2.803340918384437,
|
| 428 |
+
"learning_rate": 7.560616774274774e-07,
|
| 429 |
+
"loss": 0.8487892150878906,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.150943396226415,
|
| 434 |
+
"grad_norm": 2.7872996717171112,
|
| 435 |
+
"learning_rate": 7.469505631561317e-07,
|
| 436 |
+
"loss": 0.8430064916610718,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.169811320754717,
|
| 441 |
+
"grad_norm": 2.767338948376076,
|
| 442 |
+
"learning_rate": 7.377296694121058e-07,
|
| 443 |
+
"loss": 0.834577202796936,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.1886792452830188,
|
| 448 |
+
"grad_norm": 2.7744551402453883,
|
| 449 |
+
"learning_rate": 7.284030952540936e-07,
|
| 450 |
+
"loss": 0.8389214277267456,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.2075471698113207,
|
| 455 |
+
"grad_norm": 2.94391173341089,
|
| 456 |
+
"learning_rate": 7.189749867199898e-07,
|
| 457 |
+
"loss": 0.8442764282226562,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.2264150943396226,
|
| 462 |
+
"grad_norm": 2.9244734720758285,
|
| 463 |
+
"learning_rate": 7.094495349838092e-07,
|
| 464 |
+
"loss": 0.802047848701477,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2452830188679245,
|
| 469 |
+
"grad_norm": 2.997891576167027,
|
| 470 |
+
"learning_rate": 6.998309744925411e-07,
|
| 471 |
+
"loss": 0.8562427163124084,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2641509433962264,
|
| 476 |
+
"grad_norm": 2.7454101056544618,
|
| 477 |
+
"learning_rate": 6.901235810837667e-07,
|
| 478 |
+
"loss": 0.8214827179908752,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2830188679245282,
|
| 483 |
+
"grad_norm": 2.9952605769764853,
|
| 484 |
+
"learning_rate": 6.803316700848778e-07,
|
| 485 |
+
"loss": 0.7995479702949524,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.3018867924528301,
|
| 490 |
+
"grad_norm": 2.86683247629566,
|
| 491 |
+
"learning_rate": 6.704595943947385e-07,
|
| 492 |
+
"loss": 0.8077808022499084,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.320754716981132,
|
| 497 |
+
"grad_norm": 2.7702979738330322,
|
| 498 |
+
"learning_rate": 6.605117425486481e-07,
|
| 499 |
+
"loss": 0.8417398929595947,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3396226415094339,
|
| 504 |
+
"grad_norm": 2.725158428984504,
|
| 505 |
+
"learning_rate": 6.504925367674594e-07,
|
| 506 |
+
"loss": 0.8494030833244324,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.3584905660377358,
|
| 511 |
+
"grad_norm": 2.8106277256279255,
|
| 512 |
+
"learning_rate": 6.40406430991723e-07,
|
| 513 |
+
"loss": 0.8620424866676331,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.3773584905660377,
|
| 518 |
+
"grad_norm": 2.818628329932316,
|
| 519 |
+
"learning_rate": 6.302579089017327e-07,
|
| 520 |
+
"loss": 0.8398749232292175,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3962264150943398,
|
| 525 |
+
"grad_norm": 2.745904001646307,
|
| 526 |
+
"learning_rate": 6.200514819243475e-07,
|
| 527 |
+
"loss": 0.8420323133468628,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.4150943396226414,
|
| 532 |
+
"grad_norm": 2.7850840819985416,
|
| 533 |
+
"learning_rate": 6.097916872274814e-07,
|
| 534 |
+
"loss": 0.8359158635139465,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.4339622641509435,
|
| 539 |
+
"grad_norm": 2.793048578545994,
|
| 540 |
+
"learning_rate": 5.994830857031499e-07,
|
| 541 |
+
"loss": 0.8336814641952515,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.4528301886792452,
|
| 546 |
+
"grad_norm": 2.8505241824701826,
|
| 547 |
+
"learning_rate": 5.891302599399684e-07,
|
| 548 |
+
"loss": 0.7930982112884521,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4716981132075473,
|
| 553 |
+
"grad_norm": 2.6769256052426615,
|
| 554 |
+
"learning_rate": 5.78737812186009e-07,
|
| 555 |
+
"loss": 0.8192281723022461,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.490566037735849,
|
| 560 |
+
"grad_norm": 2.7762595596745916,
|
| 561 |
+
"learning_rate": 5.683103623029134e-07,
|
| 562 |
+
"loss": 0.8389377593994141,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.509433962264151,
|
| 567 |
+
"grad_norm": 2.8899154085340166,
|
| 568 |
+
"learning_rate": 5.578525457121806e-07,
|
| 569 |
+
"loss": 0.8256187438964844,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.5283018867924527,
|
| 574 |
+
"grad_norm": 2.7720983651750917,
|
| 575 |
+
"learning_rate": 5.473690113345342e-07,
|
| 576 |
+
"loss": 0.8473238945007324,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.5471698113207548,
|
| 581 |
+
"grad_norm": 2.8065774463241495,
|
| 582 |
+
"learning_rate": 5.368644195232895e-07,
|
| 583 |
+
"loss": 0.8165145516395569,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.5660377358490565,
|
| 588 |
+
"grad_norm": 2.9614754969968016,
|
| 589 |
+
"learning_rate": 5.263434399926398e-07,
|
| 590 |
+
"loss": 0.8529609441757202,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.5849056603773586,
|
| 595 |
+
"grad_norm": 2.90447128441676,
|
| 596 |
+
"learning_rate": 5.158107497417794e-07,
|
| 597 |
+
"loss": 0.8249980211257935,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.6037735849056602,
|
| 602 |
+
"grad_norm": 2.7563670691746767,
|
| 603 |
+
"learning_rate": 5.052710309757898e-07,
|
| 604 |
+
"loss": 0.7900608777999878,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.6226415094339623,
|
| 609 |
+
"grad_norm": 2.781624786647774,
|
| 610 |
+
"learning_rate": 4.947289690242102e-07,
|
| 611 |
+
"loss": 0.7917711734771729,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.641509433962264,
|
| 616 |
+
"grad_norm": 2.8227831992064165,
|
| 617 |
+
"learning_rate": 4.841892502582205e-07,
|
| 618 |
+
"loss": 0.8228881359100342,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6603773584905661,
|
| 623 |
+
"grad_norm": 3.0626612203128687,
|
| 624 |
+
"learning_rate": 4.736565600073602e-07,
|
| 625 |
+
"loss": 0.8176588416099548,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.6792452830188678,
|
| 630 |
+
"grad_norm": 2.7691999193756316,
|
| 631 |
+
"learning_rate": 4.6313558047671047e-07,
|
| 632 |
+
"loss": 0.8315557837486267,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.6981132075471699,
|
| 637 |
+
"grad_norm": 2.9603416787137276,
|
| 638 |
+
"learning_rate": 4.5263098866546586e-07,
|
| 639 |
+
"loss": 0.8079712390899658,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.7169811320754715,
|
| 644 |
+
"grad_norm": 2.7648310195075023,
|
| 645 |
+
"learning_rate": 4.421474542878194e-07,
|
| 646 |
+
"loss": 0.7854694128036499,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.7358490566037736,
|
| 651 |
+
"grad_norm": 2.9565749840190736,
|
| 652 |
+
"learning_rate": 4.316896376970866e-07,
|
| 653 |
+
"loss": 0.8382487297058105,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.7547169811320755,
|
| 658 |
+
"grad_norm": 2.904524931485949,
|
| 659 |
+
"learning_rate": 4.2126218781399114e-07,
|
| 660 |
+
"loss": 0.8337287902832031,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.7735849056603774,
|
| 665 |
+
"grad_norm": 2.9419686201700794,
|
| 666 |
+
"learning_rate": 4.1086974006003154e-07,
|
| 667 |
+
"loss": 0.8450314402580261,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.7924528301886793,
|
| 672 |
+
"grad_norm": 2.738066358519684,
|
| 673 |
+
"learning_rate": 4.0051691429685023e-07,
|
| 674 |
+
"loss": 0.7846765518188477,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.8113207547169812,
|
| 679 |
+
"grad_norm": 2.7276079074380895,
|
| 680 |
+
"learning_rate": 3.902083127725186e-07,
|
| 681 |
+
"loss": 0.814504861831665,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.830188679245283,
|
| 686 |
+
"grad_norm": 2.8093937971147835,
|
| 687 |
+
"learning_rate": 3.799485180756525e-07,
|
| 688 |
+
"loss": 0.8011671304702759,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.849056603773585,
|
| 693 |
+
"grad_norm": 2.842796846086812,
|
| 694 |
+
"learning_rate": 3.697420910982672e-07,
|
| 695 |
+
"loss": 0.8165295124053955,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.8679245283018868,
|
| 700 |
+
"grad_norm": 2.8189503982268977,
|
| 701 |
+
"learning_rate": 3.5959356900827687e-07,
|
| 702 |
+
"loss": 0.8199301958084106,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.8867924528301887,
|
| 707 |
+
"grad_norm": 2.910644604198592,
|
| 708 |
+
"learning_rate": 3.4950746323254063e-07,
|
| 709 |
+
"loss": 0.8019869327545166,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.9056603773584906,
|
| 714 |
+
"grad_norm": 2.863904675767849,
|
| 715 |
+
"learning_rate": 3.394882574513519e-07,
|
| 716 |
+
"loss": 0.8060827255249023,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.9245283018867925,
|
| 721 |
+
"grad_norm": 2.8904123754351723,
|
| 722 |
+
"learning_rate": 3.295404056052616e-07,
|
| 723 |
+
"loss": 0.8078351020812988,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.9433962264150944,
|
| 728 |
+
"grad_norm": 2.8850916542883778,
|
| 729 |
+
"learning_rate": 3.1966832991512225e-07,
|
| 730 |
+
"loss": 0.8068495988845825,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.9622641509433962,
|
| 735 |
+
"grad_norm": 2.9528533111592865,
|
| 736 |
+
"learning_rate": 3.0987641891623315e-07,
|
| 737 |
+
"loss": 0.8184278011322021,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.9811320754716981,
|
| 742 |
+
"grad_norm": 2.869159446180868,
|
| 743 |
+
"learning_rate": 3.0016902550745895e-07,
|
| 744 |
+
"loss": 0.8299746513366699,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 2.0,
|
| 749 |
+
"grad_norm": 2.778568933671074,
|
| 750 |
+
"learning_rate": 2.9055046501619083e-07,
|
| 751 |
+
"loss": 0.785747766494751,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 2.018867924528302,
|
| 756 |
+
"grad_norm": 2.9408610818195062,
|
| 757 |
+
"learning_rate": 2.810250132800103e-07,
|
| 758 |
+
"loss": 0.7670397758483887,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 2.0377358490566038,
|
| 763 |
+
"grad_norm": 2.6257935800346694,
|
| 764 |
+
"learning_rate": 2.715969047459066e-07,
|
| 765 |
+
"loss": 0.7878092527389526,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 2.056603773584906,
|
| 770 |
+
"grad_norm": 3.058449053263793,
|
| 771 |
+
"learning_rate": 2.6227033058789403e-07,
|
| 772 |
+
"loss": 0.7904379367828369,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 2.0754716981132075,
|
| 777 |
+
"grad_norm": 2.88973427193669,
|
| 778 |
+
"learning_rate": 2.5304943684386825e-07,
|
| 779 |
+
"loss": 0.8011707067489624,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.0943396226415096,
|
| 784 |
+
"grad_norm": 2.723021754211135,
|
| 785 |
+
"learning_rate": 2.439383225725225e-07,
|
| 786 |
+
"loss": 0.7658779621124268,
|
| 787 |
+
"step": 111
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 2.1132075471698113,
|
| 791 |
+
"grad_norm": 2.787460559434829,
|
| 792 |
+
"learning_rate": 2.3494103803115318e-07,
|
| 793 |
+
"loss": 0.7720337510108948,
|
| 794 |
+
"step": 112
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.1320754716981134,
|
| 798 |
+
"grad_norm": 2.7422069166294802,
|
| 799 |
+
"learning_rate": 2.2606158287515658e-07,
|
| 800 |
+
"loss": 0.7842212915420532,
|
| 801 |
+
"step": 113
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 2.150943396226415,
|
| 805 |
+
"grad_norm": 3.381034950183202,
|
| 806 |
+
"learning_rate": 2.1730390438002056e-07,
|
| 807 |
+
"loss": 0.7690730094909668,
|
| 808 |
+
"step": 114
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.169811320754717,
|
| 812 |
+
"grad_norm": 2.7764924352985663,
|
| 813 |
+
"learning_rate": 2.0867189568660236e-07,
|
| 814 |
+
"loss": 0.7737655639648438,
|
| 815 |
+
"step": 115
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 2.188679245283019,
|
| 819 |
+
"grad_norm": 2.8245587551592264,
|
| 820 |
+
"learning_rate": 2.0016939407046986e-07,
|
| 821 |
+
"loss": 0.7852470278739929,
|
| 822 |
+
"step": 116
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 2.207547169811321,
|
| 826 |
+
"grad_norm": 3.429004827616326,
|
| 827 |
+
"learning_rate": 1.9180017923607883e-07,
|
| 828 |
+
"loss": 0.7893455624580383,
|
| 829 |
+
"step": 117
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 2.2264150943396226,
|
| 833 |
+
"grad_norm": 3.1969648790899408,
|
| 834 |
+
"learning_rate": 1.835679716365417e-07,
|
| 835 |
+
"loss": 0.7634609937667847,
|
| 836 |
+
"step": 118
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.2452830188679247,
|
| 840 |
+
"grad_norm": 2.70318214433158,
|
| 841 |
+
"learning_rate": 1.7547643081973578e-07,
|
| 842 |
+
"loss": 0.7859703898429871,
|
| 843 |
+
"step": 119
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 2.2641509433962264,
|
| 847 |
+
"grad_norm": 2.961996890522788,
|
| 848 |
+
"learning_rate": 1.6752915380148768e-07,
|
| 849 |
+
"loss": 0.7709099650382996,
|
| 850 |
+
"step": 120
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.2830188679245285,
|
| 854 |
+
"grad_norm": 2.8177889556978095,
|
| 855 |
+
"learning_rate": 1.5972967346655448e-07,
|
| 856 |
+
"loss": 0.7789061069488525,
|
| 857 |
+
"step": 121
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.30188679245283,
|
| 861 |
+
"grad_norm": 3.320024417308839,
|
| 862 |
+
"learning_rate": 1.5208145699811415e-07,
|
| 863 |
+
"loss": 0.7862054705619812,
|
| 864 |
+
"step": 122
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.3207547169811322,
|
| 868 |
+
"grad_norm": 2.8631784669698415,
|
| 869 |
+
"learning_rate": 1.4458790433646263e-07,
|
| 870 |
+
"loss": 0.7816888689994812,
|
| 871 |
+
"step": 123
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 2.339622641509434,
|
| 875 |
+
"grad_norm": 2.902161614336072,
|
| 876 |
+
"learning_rate": 1.3725234666760427e-07,
|
| 877 |
+
"loss": 0.7391059398651123,
|
| 878 |
+
"step": 124
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.358490566037736,
|
| 882 |
+
"grad_norm": 2.882470659827849,
|
| 883 |
+
"learning_rate": 1.3007804494240476e-07,
|
| 884 |
+
"loss": 0.7627633810043335,
|
| 885 |
+
"step": 125
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 2.3773584905660377,
|
| 889 |
+
"grad_norm": 2.8433427591245284,
|
| 890 |
+
"learning_rate": 1.2306818842696715e-07,
|
| 891 |
+
"loss": 0.7769066095352173,
|
| 892 |
+
"step": 126
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.3962264150943398,
|
| 896 |
+
"grad_norm": 2.8617729260756573,
|
| 897 |
+
"learning_rate": 1.1622589328487503e-07,
|
| 898 |
+
"loss": 0.7934216856956482,
|
| 899 |
+
"step": 127
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.4150943396226414,
|
| 903 |
+
"grad_norm": 2.8509595069990823,
|
| 904 |
+
"learning_rate": 1.0955420119193198e-07,
|
| 905 |
+
"loss": 0.7673547863960266,
|
| 906 |
+
"step": 128
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.4339622641509435,
|
| 910 |
+
"grad_norm": 2.874293982355328,
|
| 911 |
+
"learning_rate": 1.03056077984014e-07,
|
| 912 |
+
"loss": 0.7849991917610168,
|
| 913 |
+
"step": 129
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.452830188679245,
|
| 917 |
+
"grad_norm": 3.0937215388279,
|
| 918 |
+
"learning_rate": 9.673441233863661e-08,
|
| 919 |
+
"loss": 0.7473263740539551,
|
| 920 |
+
"step": 130
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.4716981132075473,
|
| 924 |
+
"grad_norm": 2.9292035796935054,
|
| 925 |
+
"learning_rate": 9.059201449082043e-08,
|
| 926 |
+
"loss": 0.784021258354187,
|
| 927 |
+
"step": 131
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.490566037735849,
|
| 931 |
+
"grad_norm": 2.810444173384006,
|
| 932 |
+
"learning_rate": 8.463161498382949e-08,
|
| 933 |
+
"loss": 0.7882828712463379,
|
| 934 |
+
"step": 132
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.509433962264151,
|
| 938 |
+
"grad_norm": 2.829313317652292,
|
| 939 |
+
"learning_rate": 7.885586345533396e-08,
|
| 940 |
+
"loss": 0.7572199702262878,
|
| 941 |
+
"step": 133
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.5283018867924527,
|
| 945 |
+
"grad_norm": 2.6656369607187567,
|
| 946 |
+
"learning_rate": 7.326732745954e-08,
|
| 947 |
+
"loss": 0.7826784253120422,
|
| 948 |
+
"step": 134
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.547169811320755,
|
| 952 |
+
"grad_norm": 2.7036355808226173,
|
| 953 |
+
"learning_rate": 6.786849132580841e-08,
|
| 954 |
+
"loss": 0.7726486325263977,
|
| 955 |
+
"step": 135
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.5660377358490565,
|
| 959 |
+
"grad_norm": 2.805033772692598,
|
| 960 |
+
"learning_rate": 6.266175505426957e-08,
|
| 961 |
+
"loss": 0.7736940383911133,
|
| 962 |
+
"step": 136
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.5849056603773586,
|
| 966 |
+
"grad_norm": 2.8181269221147396,
|
| 967 |
+
"learning_rate": 5.7649433248927794e-08,
|
| 968 |
+
"loss": 0.7888213396072388,
|
| 969 |
+
"step": 137
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.6037735849056602,
|
| 973 |
+
"grad_norm": 2.9760303324315256,
|
| 974 |
+
"learning_rate": 5.283375408872537e-08,
|
| 975 |
+
"loss": 0.7611340284347534,
|
| 976 |
+
"step": 138
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.6226415094339623,
|
| 980 |
+
"grad_norm": 2.828152013200315,
|
| 981 |
+
"learning_rate": 4.821685833702849e-08,
|
| 982 |
+
"loss": 0.779454231262207,
|
| 983 |
+
"step": 139
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.641509433962264,
|
| 987 |
+
"grad_norm": 2.8581322420761786,
|
| 988 |
+
"learning_rate": 4.3800798389970863e-08,
|
| 989 |
+
"loss": 0.769560694694519,
|
| 990 |
+
"step": 140
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.660377358490566,
|
| 994 |
+
"grad_norm": 2.8125888801619103,
|
| 995 |
+
"learning_rate": 3.958753736408105e-08,
|
| 996 |
+
"loss": 0.7890896797180176,
|
| 997 |
+
"step": 141
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.6792452830188678,
|
| 1001 |
+
"grad_norm": 2.757727954638762,
|
| 1002 |
+
"learning_rate": 3.557894822359864e-08,
|
| 1003 |
+
"loss": 0.7476776838302612,
|
| 1004 |
+
"step": 142
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.69811320754717,
|
| 1008 |
+
"grad_norm": 2.802525331124496,
|
| 1009 |
+
"learning_rate": 3.1776812947865384e-08,
|
| 1010 |
+
"loss": 0.7551087737083435,
|
| 1011 |
+
"step": 143
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.7169811320754715,
|
| 1015 |
+
"grad_norm": 3.172109709327269,
|
| 1016 |
+
"learning_rate": 2.818282173916453e-08,
|
| 1017 |
+
"loss": 0.7675119638442993,
|
| 1018 |
+
"step": 144
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.7358490566037736,
|
| 1022 |
+
"grad_norm": 2.836017838014085,
|
| 1023 |
+
"learning_rate": 2.4798572271356843e-08,
|
| 1024 |
+
"loss": 0.7670686841011047,
|
| 1025 |
+
"step": 145
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.7547169811320753,
|
| 1029 |
+
"grad_norm": 2.9198667506437905,
|
| 1030 |
+
"learning_rate": 2.162556897965101e-08,
|
| 1031 |
+
"loss": 0.7993500828742981,
|
| 1032 |
+
"step": 146
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.7735849056603774,
|
| 1036 |
+
"grad_norm": 2.795471164301072,
|
| 1037 |
+
"learning_rate": 1.8665222391821166e-08,
|
| 1038 |
+
"loss": 0.7754116654396057,
|
| 1039 |
+
"step": 147
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.7924528301886795,
|
| 1043 |
+
"grad_norm": 2.7725526525432787,
|
| 1044 |
+
"learning_rate": 1.5918848501170644e-08,
|
| 1045 |
+
"loss": 0.7710179090499878,
|
| 1046 |
+
"step": 148
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.811320754716981,
|
| 1050 |
+
"grad_norm": 2.784214561225124,
|
| 1051 |
+
"learning_rate": 1.3387668181519818e-08,
|
| 1052 |
+
"loss": 0.7384580969810486,
|
| 1053 |
+
"step": 149
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.830188679245283,
|
| 1057 |
+
"grad_norm": 2.8847249743481833,
|
| 1058 |
+
"learning_rate": 1.1072806644478738e-08,
|
| 1059 |
+
"loss": 0.7740883827209473,
|
| 1060 |
+
"step": 150
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.849056603773585,
|
| 1064 |
+
"grad_norm": 2.8315645307075945,
|
| 1065 |
+
"learning_rate": 8.975292939244927e-09,
|
| 1066 |
+
"loss": 0.7919697165489197,
|
| 1067 |
+
"step": 151
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.867924528301887,
|
| 1071 |
+
"grad_norm": 2.9085892225722034,
|
| 1072 |
+
"learning_rate": 7.096059495149853e-09,
|
| 1073 |
+
"loss": 0.781722903251648,
|
| 1074 |
+
"step": 152
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.8867924528301887,
|
| 1078 |
+
"grad_norm": 2.7506543384708224,
|
| 1079 |
+
"learning_rate": 5.435941707156388e-09,
|
| 1080 |
+
"loss": 0.7471998929977417,
|
| 1081 |
+
"step": 153
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.9056603773584904,
|
| 1085 |
+
"grad_norm": 2.8426972222396136,
|
| 1086 |
+
"learning_rate": 3.995677564492039e-09,
|
| 1087 |
+
"loss": 0.7751771807670593,
|
| 1088 |
+
"step": 154
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.9245283018867925,
|
| 1092 |
+
"grad_norm": 2.844363880881091,
|
| 1093 |
+
"learning_rate": 2.7759073225832597e-09,
|
| 1094 |
+
"loss": 0.7668254375457764,
|
| 1095 |
+
"step": 155
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.9433962264150946,
|
| 1099 |
+
"grad_norm": 3.278094344932399,
|
| 1100 |
+
"learning_rate": 1.7771732184357901e-09,
|
| 1101 |
+
"loss": 0.7961957454681396,
|
| 1102 |
+
"step": 156
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.9622641509433962,
|
| 1106 |
+
"grad_norm": 2.9897635623753955,
|
| 1107 |
+
"learning_rate": 9.999192295886971e-10,
|
| 1108 |
+
"loss": 0.7848834991455078,
|
| 1109 |
+
"step": 157
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.981132075471698,
|
| 1113 |
+
"grad_norm": 2.748244107712091,
|
| 1114 |
+
"learning_rate": 4.4449087674847117e-10,
|
| 1115 |
+
"loss": 0.777495801448822,
|
| 1116 |
+
"step": 158
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 3.0,
|
| 1120 |
+
"grad_norm": 2.9554977361208974,
|
| 1121 |
+
"learning_rate": 1.1113507019094858e-10,
|
| 1122 |
+
"loss": 0.7618961334228516,
|
| 1123 |
+
"step": 159
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 3.0,
|
| 1127 |
+
"step": 159,
|
| 1128 |
+
"total_flos": 23335512768512.0,
|
| 1129 |
+
"train_loss": 0.8809327138294963,
|
| 1130 |
+
"train_runtime": 1440.1859,
|
| 1131 |
+
"train_samples_per_second": 3.485,
|
| 1132 |
+
"train_steps_per_second": 0.11
|
| 1133 |
+
}
|
| 1134 |
+
],
|
| 1135 |
+
"logging_steps": 1.0,
|
| 1136 |
+
"max_steps": 159,
|
| 1137 |
+
"num_input_tokens_seen": 0,
|
| 1138 |
+
"num_train_epochs": 3,
|
| 1139 |
+
"save_steps": 999999,
|
| 1140 |
+
"stateful_callbacks": {
|
| 1141 |
+
"TrainerControl": {
|
| 1142 |
+
"args": {
|
| 1143 |
+
"should_epoch_stop": false,
|
| 1144 |
+
"should_evaluate": false,
|
| 1145 |
+
"should_log": false,
|
| 1146 |
+
"should_save": true,
|
| 1147 |
+
"should_training_stop": true
|
| 1148 |
+
},
|
| 1149 |
+
"attributes": {}
|
| 1150 |
+
}
|
| 1151 |
+
},
|
| 1152 |
+
"total_flos": 23335512768512.0,
|
| 1153 |
+
"train_batch_size": 4,
|
| 1154 |
+
"trial_name": null,
|
| 1155 |
+
"trial_params": null
|
| 1156 |
+
}
|
training_loss.png
ADDED
|