End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +3129 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: KodCode-V1-SFT-R1_300k_batch_size_512
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# KodCode-V1-SFT-R1_300k_batch_size_512
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: KodCode-V1-SFT-R1_300k_batch_size_512
|
|
|
|
| 16 |
|
| 17 |
# KodCode-V1-SFT-R1_300k_batch_size_512
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/KodCode-V1-SFT-R1 dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.991097922848665,
|
| 3 |
+
"total_flos": 7397230226964480.0,
|
| 4 |
+
"train_loss": 0.6955865539120438,
|
| 5 |
+
"train_runtime": 47632.2951,
|
| 6 |
+
"train_samples_per_second": 4.753,
|
| 7 |
+
"train_steps_per_second": 0.009
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.991097922848665,
|
| 3 |
+
"total_flos": 7397230226964480.0,
|
| 4 |
+
"train_loss": 0.6955865539120438,
|
| 5 |
+
"train_runtime": 47632.2951,
|
| 6 |
+
"train_samples_per_second": 4.753,
|
| 7 |
+
"train_steps_per_second": 0.009
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.991097922848665,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 441,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.006782534972445952,
|
| 13 |
+
"grad_norm": 7.137413186349268,
|
| 14 |
+
"learning_rate": 1.777777777777778e-06,
|
| 15 |
+
"loss": 1.1988,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.013565069944891903,
|
| 20 |
+
"grad_norm": 7.105029708492786,
|
| 21 |
+
"learning_rate": 3.555555555555556e-06,
|
| 22 |
+
"loss": 1.1926,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.020347604917337857,
|
| 27 |
+
"grad_norm": 6.975628069653311,
|
| 28 |
+
"learning_rate": 5.333333333333334e-06,
|
| 29 |
+
"loss": 1.1857,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.027130139889783807,
|
| 34 |
+
"grad_norm": 5.073330744823269,
|
| 35 |
+
"learning_rate": 7.111111111111112e-06,
|
| 36 |
+
"loss": 1.1299,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03391267486222976,
|
| 41 |
+
"grad_norm": 3.0855210826655557,
|
| 42 |
+
"learning_rate": 8.888888888888888e-06,
|
| 43 |
+
"loss": 1.0926,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.040695209834675714,
|
| 48 |
+
"grad_norm": 5.847010298730299,
|
| 49 |
+
"learning_rate": 1.0666666666666667e-05,
|
| 50 |
+
"loss": 1.0767,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04747774480712166,
|
| 55 |
+
"grad_norm": 5.667170104694723,
|
| 56 |
+
"learning_rate": 1.2444444444444446e-05,
|
| 57 |
+
"loss": 1.0742,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.05426027977956761,
|
| 62 |
+
"grad_norm": 5.164747156326828,
|
| 63 |
+
"learning_rate": 1.4222222222222224e-05,
|
| 64 |
+
"loss": 1.0239,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06104281475201356,
|
| 69 |
+
"grad_norm": 4.055520122890776,
|
| 70 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 71 |
+
"loss": 1.0123,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.06782534972445951,
|
| 76 |
+
"grad_norm": 2.250978802714125,
|
| 77 |
+
"learning_rate": 1.7777777777777777e-05,
|
| 78 |
+
"loss": 0.9613,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07460788469690546,
|
| 83 |
+
"grad_norm": 2.553100004726335,
|
| 84 |
+
"learning_rate": 1.9555555555555557e-05,
|
| 85 |
+
"loss": 0.9432,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08139041966935143,
|
| 90 |
+
"grad_norm": 2.324385101768056,
|
| 91 |
+
"learning_rate": 2.1333333333333335e-05,
|
| 92 |
+
"loss": 0.9219,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.08817295464179738,
|
| 97 |
+
"grad_norm": 2.135936923815521,
|
| 98 |
+
"learning_rate": 2.3111111111111112e-05,
|
| 99 |
+
"loss": 0.8906,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.09495548961424333,
|
| 104 |
+
"grad_norm": 1.9227504072028807,
|
| 105 |
+
"learning_rate": 2.4888888888888893e-05,
|
| 106 |
+
"loss": 0.8878,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.10173802458668928,
|
| 111 |
+
"grad_norm": 1.2751512941755272,
|
| 112 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 113 |
+
"loss": 0.8858,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.10852055955913523,
|
| 118 |
+
"grad_norm": 1.3275331460158257,
|
| 119 |
+
"learning_rate": 2.8444444444444447e-05,
|
| 120 |
+
"loss": 0.845,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.11530309453158118,
|
| 125 |
+
"grad_norm": 1.3969086473134111,
|
| 126 |
+
"learning_rate": 3.0222222222222225e-05,
|
| 127 |
+
"loss": 0.8413,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.12208562950402713,
|
| 132 |
+
"grad_norm": 1.2596435113586295,
|
| 133 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 134 |
+
"loss": 0.8427,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1288681644764731,
|
| 139 |
+
"grad_norm": 1.0919111628660803,
|
| 140 |
+
"learning_rate": 3.377777777777778e-05,
|
| 141 |
+
"loss": 0.8427,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.13565069944891903,
|
| 146 |
+
"grad_norm": 2.1726063042436725,
|
| 147 |
+
"learning_rate": 3.555555555555555e-05,
|
| 148 |
+
"loss": 0.8361,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.142433234421365,
|
| 153 |
+
"grad_norm": 1.128837783426394,
|
| 154 |
+
"learning_rate": 3.733333333333334e-05,
|
| 155 |
+
"loss": 0.8125,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.14921576939381093,
|
| 160 |
+
"grad_norm": 2.0673361900007405,
|
| 161 |
+
"learning_rate": 3.9111111111111115e-05,
|
| 162 |
+
"loss": 0.8096,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.1559983043662569,
|
| 167 |
+
"grad_norm": 1.6404001379510031,
|
| 168 |
+
"learning_rate": 4.088888888888889e-05,
|
| 169 |
+
"loss": 0.7992,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.16278083933870285,
|
| 174 |
+
"grad_norm": 1.868383646234072,
|
| 175 |
+
"learning_rate": 4.266666666666667e-05,
|
| 176 |
+
"loss": 0.8177,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.1695633743111488,
|
| 181 |
+
"grad_norm": 1.4754626309706587,
|
| 182 |
+
"learning_rate": 4.444444444444445e-05,
|
| 183 |
+
"loss": 0.8192,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.17634590928359475,
|
| 188 |
+
"grad_norm": 1.9030805002187652,
|
| 189 |
+
"learning_rate": 4.6222222222222224e-05,
|
| 190 |
+
"loss": 0.8082,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.1831284442560407,
|
| 195 |
+
"grad_norm": 1.0440198672277559,
|
| 196 |
+
"learning_rate": 4.8e-05,
|
| 197 |
+
"loss": 0.7944,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.18991097922848665,
|
| 202 |
+
"grad_norm": 2.596599541792705,
|
| 203 |
+
"learning_rate": 4.9777777777777785e-05,
|
| 204 |
+
"loss": 0.8017,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.1966935142009326,
|
| 209 |
+
"grad_norm": 1.9546381445372338,
|
| 210 |
+
"learning_rate": 5.155555555555556e-05,
|
| 211 |
+
"loss": 0.7945,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.20347604917337855,
|
| 216 |
+
"grad_norm": 1.751189632085305,
|
| 217 |
+
"learning_rate": 5.333333333333333e-05,
|
| 218 |
+
"loss": 0.7945,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2102585841458245,
|
| 223 |
+
"grad_norm": 1.7584486355585562,
|
| 224 |
+
"learning_rate": 5.511111111111112e-05,
|
| 225 |
+
"loss": 0.7844,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.21704111911827045,
|
| 230 |
+
"grad_norm": 1.0584793475130616,
|
| 231 |
+
"learning_rate": 5.6888888888888895e-05,
|
| 232 |
+
"loss": 0.7812,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.22382365409071642,
|
| 237 |
+
"grad_norm": 2.196048787381429,
|
| 238 |
+
"learning_rate": 5.8666666666666665e-05,
|
| 239 |
+
"loss": 0.8027,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.23060618906316235,
|
| 244 |
+
"grad_norm": 1.6456458122301723,
|
| 245 |
+
"learning_rate": 6.044444444444445e-05,
|
| 246 |
+
"loss": 0.7876,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.23738872403560832,
|
| 251 |
+
"grad_norm": 1.9008326288925532,
|
| 252 |
+
"learning_rate": 6.222222222222223e-05,
|
| 253 |
+
"loss": 0.7992,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.24417125900805425,
|
| 258 |
+
"grad_norm": 2.7375256976568405,
|
| 259 |
+
"learning_rate": 6.400000000000001e-05,
|
| 260 |
+
"loss": 0.7788,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.2509537939805002,
|
| 265 |
+
"grad_norm": 1.7512537641089396,
|
| 266 |
+
"learning_rate": 6.577777777777777e-05,
|
| 267 |
+
"loss": 0.7907,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2577363289529462,
|
| 272 |
+
"grad_norm": 2.0707248009544568,
|
| 273 |
+
"learning_rate": 6.755555555555557e-05,
|
| 274 |
+
"loss": 0.7894,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.2645188639253921,
|
| 279 |
+
"grad_norm": 2.5069295004911054,
|
| 280 |
+
"learning_rate": 6.933333333333334e-05,
|
| 281 |
+
"loss": 0.7795,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.27130139889783805,
|
| 286 |
+
"grad_norm": 1.769408811743829,
|
| 287 |
+
"learning_rate": 7.11111111111111e-05,
|
| 288 |
+
"loss": 0.7823,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.278083933870284,
|
| 293 |
+
"grad_norm": 3.264548001005339,
|
| 294 |
+
"learning_rate": 7.28888888888889e-05,
|
| 295 |
+
"loss": 0.7768,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.28486646884273,
|
| 300 |
+
"grad_norm": 2.4056236262328885,
|
| 301 |
+
"learning_rate": 7.466666666666667e-05,
|
| 302 |
+
"loss": 0.7841,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.29164900381517594,
|
| 307 |
+
"grad_norm": 2.5559829162282597,
|
| 308 |
+
"learning_rate": 7.644444444444445e-05,
|
| 309 |
+
"loss": 0.7818,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.29843153878762185,
|
| 314 |
+
"grad_norm": 1.9829026032666919,
|
| 315 |
+
"learning_rate": 7.822222222222223e-05,
|
| 316 |
+
"loss": 0.7641,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.3052140737600678,
|
| 321 |
+
"grad_norm": 1.9013014130141392,
|
| 322 |
+
"learning_rate": 8e-05,
|
| 323 |
+
"loss": 0.7673,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.3119966087325138,
|
| 328 |
+
"grad_norm": 1.3239983020470356,
|
| 329 |
+
"learning_rate": 7.999874125693333e-05,
|
| 330 |
+
"loss": 0.7792,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.31877914370495974,
|
| 335 |
+
"grad_norm": 2.093841653755095,
|
| 336 |
+
"learning_rate": 7.999496510695501e-05,
|
| 337 |
+
"loss": 0.7703,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.3255616786774057,
|
| 342 |
+
"grad_norm": 2.750870577254638,
|
| 343 |
+
"learning_rate": 7.998867178772517e-05,
|
| 344 |
+
"loss": 0.7712,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3323442136498516,
|
| 349 |
+
"grad_norm": 1.4125603524830153,
|
| 350 |
+
"learning_rate": 7.997986169532741e-05,
|
| 351 |
+
"loss": 0.7637,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.3391267486222976,
|
| 356 |
+
"grad_norm": 2.524537816103899,
|
| 357 |
+
"learning_rate": 7.996853538424387e-05,
|
| 358 |
+
"loss": 0.7827,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.34590928359474354,
|
| 363 |
+
"grad_norm": 1.7058887152668756,
|
| 364 |
+
"learning_rate": 7.995469356732033e-05,
|
| 365 |
+
"loss": 0.7689,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.3526918185671895,
|
| 370 |
+
"grad_norm": 1.9625755711375161,
|
| 371 |
+
"learning_rate": 7.993833711572133e-05,
|
| 372 |
+
"loss": 0.7764,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.3594743535396354,
|
| 377 |
+
"grad_norm": 1.9608607243300042,
|
| 378 |
+
"learning_rate": 7.991946705887539e-05,
|
| 379 |
+
"loss": 0.7583,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.3662568885120814,
|
| 384 |
+
"grad_norm": 1.5775528005817034,
|
| 385 |
+
"learning_rate": 7.989808458441014e-05,
|
| 386 |
+
"loss": 0.755,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.37303942348452734,
|
| 391 |
+
"grad_norm": 1.9419695989547225,
|
| 392 |
+
"learning_rate": 7.98741910380777e-05,
|
| 393 |
+
"loss": 0.7467,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.3798219584569733,
|
| 398 |
+
"grad_norm": 1.4275820221414393,
|
| 399 |
+
"learning_rate": 7.984778792366983e-05,
|
| 400 |
+
"loss": 0.7476,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.38660449342941927,
|
| 405 |
+
"grad_norm": 1.4199172439377408,
|
| 406 |
+
"learning_rate": 7.981887690292339e-05,
|
| 407 |
+
"loss": 0.7477,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.3933870284018652,
|
| 412 |
+
"grad_norm": 1.8664000400986023,
|
| 413 |
+
"learning_rate": 7.978745979541574e-05,
|
| 414 |
+
"loss": 0.7456,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.40016956337431114,
|
| 419 |
+
"grad_norm": 1.0582378742113208,
|
| 420 |
+
"learning_rate": 7.975353857845017e-05,
|
| 421 |
+
"loss": 0.7453,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4069520983467571,
|
| 426 |
+
"grad_norm": 2.6856740014169875,
|
| 427 |
+
"learning_rate": 7.971711538693153e-05,
|
| 428 |
+
"loss": 0.7652,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.41373463331920307,
|
| 433 |
+
"grad_norm": 2.2330393685649894,
|
| 434 |
+
"learning_rate": 7.967819251323182e-05,
|
| 435 |
+
"loss": 0.7468,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.420517168291649,
|
| 440 |
+
"grad_norm": 1.3266578444651185,
|
| 441 |
+
"learning_rate": 7.963677240704588e-05,
|
| 442 |
+
"loss": 0.7476,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.42729970326409494,
|
| 447 |
+
"grad_norm": 1.337167566676173,
|
| 448 |
+
"learning_rate": 7.959285767523732e-05,
|
| 449 |
+
"loss": 0.7486,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.4340822382365409,
|
| 454 |
+
"grad_norm": 1.7695088306451494,
|
| 455 |
+
"learning_rate": 7.954645108167432e-05,
|
| 456 |
+
"loss": 0.7542,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.44086477320898687,
|
| 461 |
+
"grad_norm": 1.3827391907399533,
|
| 462 |
+
"learning_rate": 7.949755554705577e-05,
|
| 463 |
+
"loss": 0.7517,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.44764730818143283,
|
| 468 |
+
"grad_norm": 1.354435012969295,
|
| 469 |
+
"learning_rate": 7.944617414872747e-05,
|
| 470 |
+
"loss": 0.7589,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.45442984315387874,
|
| 475 |
+
"grad_norm": 0.9204507675004047,
|
| 476 |
+
"learning_rate": 7.939231012048833e-05,
|
| 477 |
+
"loss": 0.7345,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.4612123781263247,
|
| 482 |
+
"grad_norm": 1.4388297723137893,
|
| 483 |
+
"learning_rate": 7.933596685238697e-05,
|
| 484 |
+
"loss": 0.7478,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.46799491309877067,
|
| 489 |
+
"grad_norm": 1.6753627019958215,
|
| 490 |
+
"learning_rate": 7.927714789050826e-05,
|
| 491 |
+
"loss": 0.732,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.47477744807121663,
|
| 496 |
+
"grad_norm": 0.919762144543453,
|
| 497 |
+
"learning_rate": 7.921585693675029e-05,
|
| 498 |
+
"loss": 0.7385,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.48155998304366254,
|
| 503 |
+
"grad_norm": 2.2697133943634427,
|
| 504 |
+
"learning_rate": 7.915209784859116e-05,
|
| 505 |
+
"loss": 0.7334,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.4883425180161085,
|
| 510 |
+
"grad_norm": 1.2442834324078464,
|
| 511 |
+
"learning_rate": 7.908587463884638e-05,
|
| 512 |
+
"loss": 0.7501,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.49512505298855447,
|
| 517 |
+
"grad_norm": 2.0455012466197995,
|
| 518 |
+
"learning_rate": 7.90171914754163e-05,
|
| 519 |
+
"loss": 0.7323,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.5019075879610004,
|
| 524 |
+
"grad_norm": 1.466104092423094,
|
| 525 |
+
"learning_rate": 7.894605268102365e-05,
|
| 526 |
+
"loss": 0.7284,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.5086901229334464,
|
| 531 |
+
"grad_norm": 1.4621681457693403,
|
| 532 |
+
"learning_rate": 7.887246273294167e-05,
|
| 533 |
+
"loss": 0.7345,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5154726579058924,
|
| 538 |
+
"grad_norm": 1.5491014708068618,
|
| 539 |
+
"learning_rate": 7.87964262627122e-05,
|
| 540 |
+
"loss": 0.7346,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.5222551928783383,
|
| 545 |
+
"grad_norm": 1.1091999421133278,
|
| 546 |
+
"learning_rate": 7.871794805585427e-05,
|
| 547 |
+
"loss": 0.7313,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5290377278507842,
|
| 552 |
+
"grad_norm": 1.4167418001519376,
|
| 553 |
+
"learning_rate": 7.863703305156273e-05,
|
| 554 |
+
"loss": 0.7217,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5358202628232301,
|
| 559 |
+
"grad_norm": 1.2449973723473307,
|
| 560 |
+
"learning_rate": 7.855368634239769e-05,
|
| 561 |
+
"loss": 0.7338,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.5426027977956761,
|
| 566 |
+
"grad_norm": 1.2279144181708292,
|
| 567 |
+
"learning_rate": 7.846791317396373e-05,
|
| 568 |
+
"loss": 0.7275,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5493853327681221,
|
| 573 |
+
"grad_norm": 1.1622952382317697,
|
| 574 |
+
"learning_rate": 7.837971894457991e-05,
|
| 575 |
+
"loss": 0.7459,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.556167867740568,
|
| 580 |
+
"grad_norm": 1.16239709067931,
|
| 581 |
+
"learning_rate": 7.828910920493995e-05,
|
| 582 |
+
"loss": 0.7176,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.562950402713014,
|
| 587 |
+
"grad_norm": 1.4393853572209945,
|
| 588 |
+
"learning_rate": 7.819608965776295e-05,
|
| 589 |
+
"loss": 0.7378,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.56973293768546,
|
| 594 |
+
"grad_norm": 0.8723694343578424,
|
| 595 |
+
"learning_rate": 7.810066615743443e-05,
|
| 596 |
+
"loss": 0.7235,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.5765154726579059,
|
| 601 |
+
"grad_norm": 1.4627194527050393,
|
| 602 |
+
"learning_rate": 7.800284470963783e-05,
|
| 603 |
+
"loss": 0.7245,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.5832980076303519,
|
| 608 |
+
"grad_norm": 1.5007942983664493,
|
| 609 |
+
"learning_rate": 7.790263147097661e-05,
|
| 610 |
+
"loss": 0.7358,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.5900805426027977,
|
| 615 |
+
"grad_norm": 1.0492585064630457,
|
| 616 |
+
"learning_rate": 7.780003274858674e-05,
|
| 617 |
+
"loss": 0.7222,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.5968630775752437,
|
| 622 |
+
"grad_norm": 1.4048579687963796,
|
| 623 |
+
"learning_rate": 7.769505499973977e-05,
|
| 624 |
+
"loss": 0.716,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.6036456125476897,
|
| 629 |
+
"grad_norm": 1.0188682627658208,
|
| 630 |
+
"learning_rate": 7.758770483143634e-05,
|
| 631 |
+
"loss": 0.7268,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.6104281475201356,
|
| 636 |
+
"grad_norm": 1.1982575150619237,
|
| 637 |
+
"learning_rate": 7.747798899999048e-05,
|
| 638 |
+
"loss": 0.7407,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.6172106824925816,
|
| 643 |
+
"grad_norm": 0.9182390822096019,
|
| 644 |
+
"learning_rate": 7.736591441060427e-05,
|
| 645 |
+
"loss": 0.7326,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.6239932174650276,
|
| 650 |
+
"grad_norm": 1.4475324042423592,
|
| 651 |
+
"learning_rate": 7.725148811693336e-05,
|
| 652 |
+
"loss": 0.7336,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.6307757524374735,
|
| 657 |
+
"grad_norm": 1.1065799129734808,
|
| 658 |
+
"learning_rate": 7.71347173206429e-05,
|
| 659 |
+
"loss": 0.7173,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6375582874099195,
|
| 664 |
+
"grad_norm": 1.237603778756235,
|
| 665 |
+
"learning_rate": 7.701560937095445e-05,
|
| 666 |
+
"loss": 0.7303,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6443408223823655,
|
| 671 |
+
"grad_norm": 1.1120475384075377,
|
| 672 |
+
"learning_rate": 7.689417176418327e-05,
|
| 673 |
+
"loss": 0.7185,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6511233573548114,
|
| 678 |
+
"grad_norm": 1.3783747990085908,
|
| 679 |
+
"learning_rate": 7.677041214326663e-05,
|
| 680 |
+
"loss": 0.7357,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.6579058923272573,
|
| 685 |
+
"grad_norm": 1.0644720482330112,
|
| 686 |
+
"learning_rate": 7.664433829728279e-05,
|
| 687 |
+
"loss": 0.7155,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.6646884272997032,
|
| 692 |
+
"grad_norm": 1.4446573818002777,
|
| 693 |
+
"learning_rate": 7.651595816096071e-05,
|
| 694 |
+
"loss": 0.7109,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.6714709622721492,
|
| 699 |
+
"grad_norm": 1.0329689312285972,
|
| 700 |
+
"learning_rate": 7.638527981418075e-05,
|
| 701 |
+
"loss": 0.7333,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.6782534972445952,
|
| 706 |
+
"grad_norm": 1.2999144870004373,
|
| 707 |
+
"learning_rate": 7.625231148146601e-05,
|
| 708 |
+
"loss": 0.7097,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.6850360322170411,
|
| 713 |
+
"grad_norm": 1.1368825847990134,
|
| 714 |
+
"learning_rate": 7.611706153146486e-05,
|
| 715 |
+
"loss": 0.7128,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.6918185671894871,
|
| 720 |
+
"grad_norm": 1.1174260276126868,
|
| 721 |
+
"learning_rate": 7.597953847642413e-05,
|
| 722 |
+
"loss": 0.7241,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.698601102161933,
|
| 727 |
+
"grad_norm": 0.8874864224204202,
|
| 728 |
+
"learning_rate": 7.583975097165344e-05,
|
| 729 |
+
"loss": 0.7189,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.705383637134379,
|
| 734 |
+
"grad_norm": 0.6432301304209681,
|
| 735 |
+
"learning_rate": 7.56977078149804e-05,
|
| 736 |
+
"loss": 0.7145,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.712166172106825,
|
| 741 |
+
"grad_norm": 0.9183282978967895,
|
| 742 |
+
"learning_rate": 7.555341794619695e-05,
|
| 743 |
+
"loss": 0.7107,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.7189487070792708,
|
| 748 |
+
"grad_norm": 1.0095498465315442,
|
| 749 |
+
"learning_rate": 7.540689044649666e-05,
|
| 750 |
+
"loss": 0.703,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.7257312420517168,
|
| 755 |
+
"grad_norm": 1.26183318694208,
|
| 756 |
+
"learning_rate": 7.525813453790328e-05,
|
| 757 |
+
"loss": 0.7127,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.7325137770241628,
|
| 762 |
+
"grad_norm": 0.6972659136766679,
|
| 763 |
+
"learning_rate": 7.510715958269023e-05,
|
| 764 |
+
"loss": 0.7052,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.7392963119966087,
|
| 769 |
+
"grad_norm": 0.7185478814495369,
|
| 770 |
+
"learning_rate": 7.49539750827914e-05,
|
| 771 |
+
"loss": 0.7109,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.7460788469690547,
|
| 776 |
+
"grad_norm": 0.9112632508609093,
|
| 777 |
+
"learning_rate": 7.479859067920317e-05,
|
| 778 |
+
"loss": 0.7037,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.7528613819415007,
|
| 783 |
+
"grad_norm": 0.8720981679213001,
|
| 784 |
+
"learning_rate": 7.464101615137756e-05,
|
| 785 |
+
"loss": 0.7075,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.7596439169139466,
|
| 790 |
+
"grad_norm": 1.0271120509299871,
|
| 791 |
+
"learning_rate": 7.448126141660678e-05,
|
| 792 |
+
"loss": 0.7204,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.7664264518863926,
|
| 797 |
+
"grad_norm": 1.016829719094229,
|
| 798 |
+
"learning_rate": 7.431933652939909e-05,
|
| 799 |
+
"loss": 0.7129,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.7732089868588385,
|
| 804 |
+
"grad_norm": 1.0366544063351542,
|
| 805 |
+
"learning_rate": 7.415525168084593e-05,
|
| 806 |
+
"loss": 0.708,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.7799915218312844,
|
| 811 |
+
"grad_norm": 1.144762721052778,
|
| 812 |
+
"learning_rate": 7.398901719798059e-05,
|
| 813 |
+
"loss": 0.7102,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.7867740568037304,
|
| 818 |
+
"grad_norm": 1.0391721353507355,
|
| 819 |
+
"learning_rate": 7.382064354312818e-05,
|
| 820 |
+
"loss": 0.7098,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.7935565917761763,
|
| 825 |
+
"grad_norm": 0.9632167130265382,
|
| 826 |
+
"learning_rate": 7.365014131324725e-05,
|
| 827 |
+
"loss": 0.7029,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.8003391267486223,
|
| 832 |
+
"grad_norm": 1.227602079185617,
|
| 833 |
+
"learning_rate": 7.34775212392628e-05,
|
| 834 |
+
"loss": 0.6972,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.8071216617210683,
|
| 839 |
+
"grad_norm": 0.8826250475798045,
|
| 840 |
+
"learning_rate": 7.330279418539086e-05,
|
| 841 |
+
"loss": 0.7018,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.8139041966935142,
|
| 846 |
+
"grad_norm": 0.5567261188617991,
|
| 847 |
+
"learning_rate": 7.312597114845483e-05,
|
| 848 |
+
"loss": 0.6981,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.8206867316659602,
|
| 853 |
+
"grad_norm": 0.8765091262196973,
|
| 854 |
+
"learning_rate": 7.294706325719331e-05,
|
| 855 |
+
"loss": 0.7195,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.8274692666384061,
|
| 860 |
+
"grad_norm": 0.9742989954487304,
|
| 861 |
+
"learning_rate": 7.276608177155968e-05,
|
| 862 |
+
"loss": 0.7033,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.8342518016108521,
|
| 867 |
+
"grad_norm": 0.9120816045559339,
|
| 868 |
+
"learning_rate": 7.258303808201343e-05,
|
| 869 |
+
"loss": 0.6984,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.841034336583298,
|
| 874 |
+
"grad_norm": 0.8325780076464987,
|
| 875 |
+
"learning_rate": 7.239794370880335e-05,
|
| 876 |
+
"loss": 0.6921,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.8478168715557439,
|
| 881 |
+
"grad_norm": 0.6155944829472941,
|
| 882 |
+
"learning_rate": 7.221081030124235e-05,
|
| 883 |
+
"loss": 0.6988,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.8545994065281899,
|
| 888 |
+
"grad_norm": 0.7511218044788109,
|
| 889 |
+
"learning_rate": 7.202164963697442e-05,
|
| 890 |
+
"loss": 0.6972,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.8613819415006359,
|
| 895 |
+
"grad_norm": 0.7293772573099593,
|
| 896 |
+
"learning_rate": 7.183047362123329e-05,
|
| 897 |
+
"loss": 0.7004,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.8681644764730818,
|
| 902 |
+
"grad_norm": 0.46098198038922594,
|
| 903 |
+
"learning_rate": 7.163729428609318e-05,
|
| 904 |
+
"loss": 0.6956,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.8749470114455278,
|
| 909 |
+
"grad_norm": 0.7377483747880154,
|
| 910 |
+
"learning_rate": 7.144212378971151e-05,
|
| 911 |
+
"loss": 0.7089,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.8817295464179737,
|
| 916 |
+
"grad_norm": 0.8138881211490862,
|
| 917 |
+
"learning_rate": 7.124497441556374e-05,
|
| 918 |
+
"loss": 0.7025,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.8885120813904197,
|
| 923 |
+
"grad_norm": 0.7428361558045745,
|
| 924 |
+
"learning_rate": 7.104585857167028e-05,
|
| 925 |
+
"loss": 0.7093,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.8952946163628657,
|
| 930 |
+
"grad_norm": 0.8252776283861111,
|
| 931 |
+
"learning_rate": 7.084478878981552e-05,
|
| 932 |
+
"loss": 0.7,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.9020771513353115,
|
| 937 |
+
"grad_norm": 0.8110674797944571,
|
| 938 |
+
"learning_rate": 7.064177772475912e-05,
|
| 939 |
+
"loss": 0.695,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.9088596863077575,
|
| 944 |
+
"grad_norm": 0.6529878802072909,
|
| 945 |
+
"learning_rate": 7.043683815343967e-05,
|
| 946 |
+
"loss": 0.6938,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.9156422212802034,
|
| 951 |
+
"grad_norm": 0.7848292838747539,
|
| 952 |
+
"learning_rate": 7.022998297417034e-05,
|
| 953 |
+
"loss": 0.6924,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.9224247562526494,
|
| 958 |
+
"grad_norm": 1.1207835540553364,
|
| 959 |
+
"learning_rate": 7.00212252058273e-05,
|
| 960 |
+
"loss": 0.6988,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.9292072912250954,
|
| 965 |
+
"grad_norm": 0.9343138621109743,
|
| 966 |
+
"learning_rate": 6.98105779870302e-05,
|
| 967 |
+
"loss": 0.6966,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.9359898261975413,
|
| 972 |
+
"grad_norm": 0.6897906744897139,
|
| 973 |
+
"learning_rate": 6.959805457531536e-05,
|
| 974 |
+
"loss": 0.6995,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.9427723611699873,
|
| 979 |
+
"grad_norm": 0.48700498757406885,
|
| 980 |
+
"learning_rate": 6.938366834630133e-05,
|
| 981 |
+
"loss": 0.689,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.9495548961424333,
|
| 986 |
+
"grad_norm": 0.460000455610998,
|
| 987 |
+
"learning_rate": 6.916743279284709e-05,
|
| 988 |
+
"loss": 0.699,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.9563374311148792,
|
| 993 |
+
"grad_norm": 0.6228357519746542,
|
| 994 |
+
"learning_rate": 6.89493615242028e-05,
|
| 995 |
+
"loss": 0.6987,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.9631199660873251,
|
| 1000 |
+
"grad_norm": 0.6004863488992443,
|
| 1001 |
+
"learning_rate": 6.872946826515338e-05,
|
| 1002 |
+
"loss": 0.7078,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.969902501059771,
|
| 1007 |
+
"grad_norm": 0.38322005162459927,
|
| 1008 |
+
"learning_rate": 6.850776685515453e-05,
|
| 1009 |
+
"loss": 0.6951,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.976685036032217,
|
| 1014 |
+
"grad_norm": 0.4778442733404826,
|
| 1015 |
+
"learning_rate": 6.828427124746191e-05,
|
| 1016 |
+
"loss": 0.7059,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.983467571004663,
|
| 1021 |
+
"grad_norm": 0.4078839327460854,
|
| 1022 |
+
"learning_rate": 6.805899550825285e-05,
|
| 1023 |
+
"loss": 0.7034,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9902501059771089,
|
| 1028 |
+
"grad_norm": 0.3930697959309545,
|
| 1029 |
+
"learning_rate": 6.78319538157411e-05,
|
| 1030 |
+
"loss": 0.688,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9970326409495549,
|
| 1035 |
+
"grad_norm": 0.471709925781642,
|
| 1036 |
+
"learning_rate": 6.760316045928449e-05,
|
| 1037 |
+
"loss": 0.6996,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.0038151759220009,
|
| 1042 |
+
"grad_norm": 0.5236980599221154,
|
| 1043 |
+
"learning_rate": 6.737262983848554e-05,
|
| 1044 |
+
"loss": 1.0973,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.0105977108944468,
|
| 1049 |
+
"grad_norm": 0.7531817832475667,
|
| 1050 |
+
"learning_rate": 6.714037646228529e-05,
|
| 1051 |
+
"loss": 0.6724,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.0173802458668928,
|
| 1056 |
+
"grad_norm": 0.8924803206824311,
|
| 1057 |
+
"learning_rate": 6.690641494805011e-05,
|
| 1058 |
+
"loss": 0.6904,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.0241627808393388,
|
| 1063 |
+
"grad_norm": 0.9578752031258555,
|
| 1064 |
+
"learning_rate": 6.667076002065168e-05,
|
| 1065 |
+
"loss": 0.6763,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.0309453158117847,
|
| 1070 |
+
"grad_norm": 0.9584132989076937,
|
| 1071 |
+
"learning_rate": 6.643342651154028e-05,
|
| 1072 |
+
"loss": 0.6928,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.0377278507842307,
|
| 1077 |
+
"grad_norm": 0.6631156370651635,
|
| 1078 |
+
"learning_rate": 6.619442935781141e-05,
|
| 1079 |
+
"loss": 0.6914,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.0445103857566767,
|
| 1084 |
+
"grad_norm": 0.5101036583421924,
|
| 1085 |
+
"learning_rate": 6.595378360126555e-05,
|
| 1086 |
+
"loss": 0.6778,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.0512929207291226,
|
| 1091 |
+
"grad_norm": 0.6836210135721171,
|
| 1092 |
+
"learning_rate": 6.571150438746157e-05,
|
| 1093 |
+
"loss": 0.6776,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.0580754557015684,
|
| 1098 |
+
"grad_norm": 0.5297506092844936,
|
| 1099 |
+
"learning_rate": 6.546760696476354e-05,
|
| 1100 |
+
"loss": 0.6792,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.0648579906740143,
|
| 1105 |
+
"grad_norm": 0.6440705198875945,
|
| 1106 |
+
"learning_rate": 6.52221066833809e-05,
|
| 1107 |
+
"loss": 0.6813,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.0716405256464603,
|
| 1112 |
+
"grad_norm": 0.5649902540029019,
|
| 1113 |
+
"learning_rate": 6.497501899440255e-05,
|
| 1114 |
+
"loss": 0.6758,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.0784230606189062,
|
| 1119 |
+
"grad_norm": 0.5451745835283485,
|
| 1120 |
+
"learning_rate": 6.472635944882421e-05,
|
| 1121 |
+
"loss": 0.6846,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.0852055955913522,
|
| 1126 |
+
"grad_norm": 0.6742930881134588,
|
| 1127 |
+
"learning_rate": 6.447614369656987e-05,
|
| 1128 |
+
"loss": 0.6659,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.0919881305637982,
|
| 1133 |
+
"grad_norm": 0.40739982620953924,
|
| 1134 |
+
"learning_rate": 6.422438748550667e-05,
|
| 1135 |
+
"loss": 0.6688,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.0987706655362441,
|
| 1140 |
+
"grad_norm": 0.520705689346025,
|
| 1141 |
+
"learning_rate": 6.397110666045388e-05,
|
| 1142 |
+
"loss": 0.675,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.10555320050869,
|
| 1147 |
+
"grad_norm": 0.4809784234416612,
|
| 1148 |
+
"learning_rate": 6.371631716218563e-05,
|
| 1149 |
+
"loss": 0.6714,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.112335735481136,
|
| 1154 |
+
"grad_norm": 0.6328561377367079,
|
| 1155 |
+
"learning_rate": 6.346003502642762e-05,
|
| 1156 |
+
"loss": 0.6818,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.119118270453582,
|
| 1161 |
+
"grad_norm": 0.6964171119263963,
|
| 1162 |
+
"learning_rate": 6.320227638284793e-05,
|
| 1163 |
+
"loss": 0.6663,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.125900805426028,
|
| 1168 |
+
"grad_norm": 0.5303565775080002,
|
| 1169 |
+
"learning_rate": 6.294305745404185e-05,
|
| 1170 |
+
"loss": 0.6784,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.132683340398474,
|
| 1175 |
+
"grad_norm": 0.44306420532590796,
|
| 1176 |
+
"learning_rate": 6.268239455451083e-05,
|
| 1177 |
+
"loss": 0.6867,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.13946587537092,
|
| 1182 |
+
"grad_norm": 0.3414530565574301,
|
| 1183 |
+
"learning_rate": 6.242030408963576e-05,
|
| 1184 |
+
"loss": 0.6672,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.1462484103433659,
|
| 1189 |
+
"grad_norm": 0.40031871777699174,
|
| 1190 |
+
"learning_rate": 6.215680255464442e-05,
|
| 1191 |
+
"loss": 0.6651,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.1530309453158119,
|
| 1196 |
+
"grad_norm": 0.277522312209641,
|
| 1197 |
+
"learning_rate": 6.18919065335733e-05,
|
| 1198 |
+
"loss": 0.6669,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.1598134802882578,
|
| 1203 |
+
"grad_norm": 0.35702823672803347,
|
| 1204 |
+
"learning_rate": 6.162563269822391e-05,
|
| 1205 |
+
"loss": 0.6636,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.1665960152607038,
|
| 1210 |
+
"grad_norm": 0.29212513135494134,
|
| 1211 |
+
"learning_rate": 6.135799780711345e-05,
|
| 1212 |
+
"loss": 0.675,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.1733785502331497,
|
| 1217 |
+
"grad_norm": 0.32422232694491343,
|
| 1218 |
+
"learning_rate": 6.10890187044201e-05,
|
| 1219 |
+
"loss": 0.6833,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.1801610852055955,
|
| 1224 |
+
"grad_norm": 0.34358927362334013,
|
| 1225 |
+
"learning_rate": 6.0818712318922894e-05,
|
| 1226 |
+
"loss": 0.6682,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.1869436201780414,
|
| 1231 |
+
"grad_norm": 0.24681935714411127,
|
| 1232 |
+
"learning_rate": 6.054709566293627e-05,
|
| 1233 |
+
"loss": 0.683,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.1937261551504874,
|
| 1238 |
+
"grad_norm": 0.3149112104902362,
|
| 1239 |
+
"learning_rate": 6.0274185831239325e-05,
|
| 1240 |
+
"loss": 0.6638,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.2005086901229334,
|
| 1245 |
+
"grad_norm": 0.34401534901495145,
|
| 1246 |
+
"learning_rate": 6.000000000000001e-05,
|
| 1247 |
+
"loss": 0.6725,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.2072912250953793,
|
| 1252 |
+
"grad_norm": 0.323484399780841,
|
| 1253 |
+
"learning_rate": 5.972455542569402e-05,
|
| 1254 |
+
"loss": 0.6692,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.2140737600678253,
|
| 1259 |
+
"grad_norm": 0.30391385199629817,
|
| 1260 |
+
"learning_rate": 5.944786944401875e-05,
|
| 1261 |
+
"loss": 0.6623,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.2208562950402713,
|
| 1266 |
+
"grad_norm": 0.26879513471231475,
|
| 1267 |
+
"learning_rate": 5.916995946880228e-05,
|
| 1268 |
+
"loss": 0.6842,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.2276388300127172,
|
| 1273 |
+
"grad_norm": 0.29492095831429094,
|
| 1274 |
+
"learning_rate": 5.889084299090732e-05,
|
| 1275 |
+
"loss": 0.6786,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.2344213649851632,
|
| 1280 |
+
"grad_norm": 0.3788666762600595,
|
| 1281 |
+
"learning_rate": 5.861053757713043e-05,
|
| 1282 |
+
"loss": 0.6726,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.2412038999576092,
|
| 1287 |
+
"grad_norm": 0.4398437908833291,
|
| 1288 |
+
"learning_rate": 5.832906086909642e-05,
|
| 1289 |
+
"loss": 0.6742,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.2479864349300551,
|
| 1294 |
+
"grad_norm": 0.5383171396673415,
|
| 1295 |
+
"learning_rate": 5.8046430582148034e-05,
|
| 1296 |
+
"loss": 0.6636,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.254768969902501,
|
| 1301 |
+
"grad_norm": 0.6287710260233805,
|
| 1302 |
+
"learning_rate": 5.776266450423097e-05,
|
| 1303 |
+
"loss": 0.6745,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.261551504874947,
|
| 1308 |
+
"grad_norm": 0.692244364864307,
|
| 1309 |
+
"learning_rate": 5.747778049477438e-05,
|
| 1310 |
+
"loss": 0.692,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.268334039847393,
|
| 1315 |
+
"grad_norm": 0.5173002490296247,
|
| 1316 |
+
"learning_rate": 5.7191796483566874e-05,
|
| 1317 |
+
"loss": 0.6851,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.275116574819839,
|
| 1322 |
+
"grad_norm": 0.3170872618606908,
|
| 1323 |
+
"learning_rate": 5.6904730469627985e-05,
|
| 1324 |
+
"loss": 0.6694,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.281899109792285,
|
| 1329 |
+
"grad_norm": 0.3605888836619855,
|
| 1330 |
+
"learning_rate": 5.661660052007547e-05,
|
| 1331 |
+
"loss": 0.6674,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.288681644764731,
|
| 1336 |
+
"grad_norm": 0.5359207812719183,
|
| 1337 |
+
"learning_rate": 5.632742476898813e-05,
|
| 1338 |
+
"loss": 0.6574,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.2954641797371766,
|
| 1343 |
+
"grad_norm": 0.5493258534047062,
|
| 1344 |
+
"learning_rate": 5.6037221416264554e-05,
|
| 1345 |
+
"loss": 0.6651,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.3022467147096228,
|
| 1350 |
+
"grad_norm": 0.3583558907183342,
|
| 1351 |
+
"learning_rate": 5.574600872647766e-05,
|
| 1352 |
+
"loss": 0.6741,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.3090292496820686,
|
| 1357 |
+
"grad_norm": 0.38616308348313483,
|
| 1358 |
+
"learning_rate": 5.5453805027725145e-05,
|
| 1359 |
+
"loss": 0.6735,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.3158117846545148,
|
| 1364 |
+
"grad_norm": 0.5232887516865083,
|
| 1365 |
+
"learning_rate": 5.516062871047602e-05,
|
| 1366 |
+
"loss": 0.6648,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.3225943196269605,
|
| 1371 |
+
"grad_norm": 0.5001634228373406,
|
| 1372 |
+
"learning_rate": 5.48664982264131e-05,
|
| 1373 |
+
"loss": 0.682,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.3293768545994065,
|
| 1378 |
+
"grad_norm": 0.4545741719631783,
|
| 1379 |
+
"learning_rate": 5.4571432087271775e-05,
|
| 1380 |
+
"loss": 0.6748,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.3361593895718524,
|
| 1385 |
+
"grad_norm": 0.42741557026919574,
|
| 1386 |
+
"learning_rate": 5.427544886367488e-05,
|
| 1387 |
+
"loss": 0.677,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.3429419245442984,
|
| 1392 |
+
"grad_norm": 0.31094524488846675,
|
| 1393 |
+
"learning_rate": 5.397856718396394e-05,
|
| 1394 |
+
"loss": 0.6652,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.3497244595167444,
|
| 1399 |
+
"grad_norm": 0.24780800036037465,
|
| 1400 |
+
"learning_rate": 5.368080573302676e-05,
|
| 1401 |
+
"loss": 0.6664,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.3565069944891903,
|
| 1406 |
+
"grad_norm": 0.3181527932049964,
|
| 1407 |
+
"learning_rate": 5.3382183251121415e-05,
|
| 1408 |
+
"loss": 0.6675,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.3632895294616363,
|
| 1413 |
+
"grad_norm": 0.29745491627360243,
|
| 1414 |
+
"learning_rate": 5.3082718532696874e-05,
|
| 1415 |
+
"loss": 0.6706,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.3700720644340822,
|
| 1420 |
+
"grad_norm": 0.2584376200240668,
|
| 1421 |
+
"learning_rate": 5.2782430425210004e-05,
|
| 1422 |
+
"loss": 0.6631,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.3768545994065282,
|
| 1427 |
+
"grad_norm": 0.28291959973268804,
|
| 1428 |
+
"learning_rate": 5.2481337827939486e-05,
|
| 1429 |
+
"loss": 0.665,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.3836371343789742,
|
| 1434 |
+
"grad_norm": 0.222511659896141,
|
| 1435 |
+
"learning_rate": 5.217945969079629e-05,
|
| 1436 |
+
"loss": 0.6737,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.3904196693514201,
|
| 1441 |
+
"grad_norm": 0.1904985337395551,
|
| 1442 |
+
"learning_rate": 5.1876815013131e-05,
|
| 1443 |
+
"loss": 0.6773,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.397202204323866,
|
| 1448 |
+
"grad_norm": 0.2556162510484352,
|
| 1449 |
+
"learning_rate": 5.157342284253812e-05,
|
| 1450 |
+
"loss": 0.674,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.403984739296312,
|
| 1455 |
+
"grad_norm": 0.2384701949692735,
|
| 1456 |
+
"learning_rate": 5.1269302273657195e-05,
|
| 1457 |
+
"loss": 0.6817,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.410767274268758,
|
| 1462 |
+
"grad_norm": 0.20411764282852152,
|
| 1463 |
+
"learning_rate": 5.0964472446971114e-05,
|
| 1464 |
+
"loss": 0.6536,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.417549809241204,
|
| 1469 |
+
"grad_norm": 0.20834830910241572,
|
| 1470 |
+
"learning_rate": 5.06589525476014e-05,
|
| 1471 |
+
"loss": 0.6694,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.4243323442136497,
|
| 1476 |
+
"grad_norm": 0.22097226372451154,
|
| 1477 |
+
"learning_rate": 5.0352761804100835e-05,
|
| 1478 |
+
"loss": 0.6607,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.431114879186096,
|
| 1483 |
+
"grad_norm": 0.23524605314054492,
|
| 1484 |
+
"learning_rate": 5.004591948724317e-05,
|
| 1485 |
+
"loss": 0.6591,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.4378974141585417,
|
| 1490 |
+
"grad_norm": 0.32464444157336164,
|
| 1491 |
+
"learning_rate": 4.9738444908810365e-05,
|
| 1492 |
+
"loss": 0.6698,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.4446799491309876,
|
| 1497 |
+
"grad_norm": 0.32718322363813607,
|
| 1498 |
+
"learning_rate": 4.94303574203771e-05,
|
| 1499 |
+
"loss": 0.6543,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.4514624841034336,
|
| 1504 |
+
"grad_norm": 0.2459546193619208,
|
| 1505 |
+
"learning_rate": 4.9121676412092874e-05,
|
| 1506 |
+
"loss": 0.6696,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.4582450190758796,
|
| 1511 |
+
"grad_norm": 0.1878400255657447,
|
| 1512 |
+
"learning_rate": 4.881242131146163e-05,
|
| 1513 |
+
"loss": 0.6745,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.4650275540483255,
|
| 1518 |
+
"grad_norm": 0.21340996885255564,
|
| 1519 |
+
"learning_rate": 4.8502611582119065e-05,
|
| 1520 |
+
"loss": 0.6702,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.4718100890207715,
|
| 1525 |
+
"grad_norm": 0.1930500360689927,
|
| 1526 |
+
"learning_rate": 4.819226672260763e-05,
|
| 1527 |
+
"loss": 0.6739,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.4785926239932174,
|
| 1532 |
+
"grad_norm": 0.19466196179996254,
|
| 1533 |
+
"learning_rate": 4.788140626514933e-05,
|
| 1534 |
+
"loss": 0.6678,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.4853751589656634,
|
| 1539 |
+
"grad_norm": 0.1802358321365316,
|
| 1540 |
+
"learning_rate": 4.7570049774416414e-05,
|
| 1541 |
+
"loss": 0.6578,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.4921576939381094,
|
| 1546 |
+
"grad_norm": 0.21190165053247084,
|
| 1547 |
+
"learning_rate": 4.7258216846300106e-05,
|
| 1548 |
+
"loss": 0.6687,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.4989402289105553,
|
| 1553 |
+
"grad_norm": 0.2025332247212839,
|
| 1554 |
+
"learning_rate": 4.694592710667723e-05,
|
| 1555 |
+
"loss": 0.6695,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.5057227638830013,
|
| 1560 |
+
"grad_norm": 0.18244954655230586,
|
| 1561 |
+
"learning_rate": 4.663320021017497e-05,
|
| 1562 |
+
"loss": 0.6655,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.5125052988554473,
|
| 1567 |
+
"grad_norm": 0.2269278837073207,
|
| 1568 |
+
"learning_rate": 4.6320055838934e-05,
|
| 1569 |
+
"loss": 0.6727,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.5192878338278932,
|
| 1574 |
+
"grad_norm": 0.2259027791570099,
|
| 1575 |
+
"learning_rate": 4.6006513701369616e-05,
|
| 1576 |
+
"loss": 0.6606,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.526070368800339,
|
| 1581 |
+
"grad_norm": 0.2477760784045823,
|
| 1582 |
+
"learning_rate": 4.5692593530931416e-05,
|
| 1583 |
+
"loss": 0.6759,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.5328529037727852,
|
| 1588 |
+
"grad_norm": 0.2603331861037767,
|
| 1589 |
+
"learning_rate": 4.5378315084861276e-05,
|
| 1590 |
+
"loss": 0.6624,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.539635438745231,
|
| 1595 |
+
"grad_norm": 0.2506086748418563,
|
| 1596 |
+
"learning_rate": 4.506369814294998e-05,
|
| 1597 |
+
"loss": 0.6599,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.546417973717677,
|
| 1602 |
+
"grad_norm": 0.26765221421169105,
|
| 1603 |
+
"learning_rate": 4.474876250629221e-05,
|
| 1604 |
+
"loss": 0.6764,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.5532005086901228,
|
| 1609 |
+
"grad_norm": 0.17936756384362929,
|
| 1610 |
+
"learning_rate": 4.4433527996040443e-05,
|
| 1611 |
+
"loss": 0.6643,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.559983043662569,
|
| 1616 |
+
"grad_norm": 0.1830215522438223,
|
| 1617 |
+
"learning_rate": 4.411801445215739e-05,
|
| 1618 |
+
"loss": 0.6644,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.5667655786350148,
|
| 1623 |
+
"grad_norm": 0.20385050238488933,
|
| 1624 |
+
"learning_rate": 4.38022417321673e-05,
|
| 1625 |
+
"loss": 0.6559,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.573548113607461,
|
| 1630 |
+
"grad_norm": 0.21314261341134186,
|
| 1631 |
+
"learning_rate": 4.348622970990634e-05,
|
| 1632 |
+
"loss": 0.6733,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.5803306485799067,
|
| 1637 |
+
"grad_norm": 0.23351272270245574,
|
| 1638 |
+
"learning_rate": 4.316999827427154e-05,
|
| 1639 |
+
"loss": 0.6609,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.5871131835523526,
|
| 1644 |
+
"grad_norm": 0.24621369232426915,
|
| 1645 |
+
"learning_rate": 4.2853567327969296e-05,
|
| 1646 |
+
"loss": 0.6681,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.5938957185247986,
|
| 1651 |
+
"grad_norm": 0.30216358602341464,
|
| 1652 |
+
"learning_rate": 4.2536956786262585e-05,
|
| 1653 |
+
"loss": 0.6822,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.6006782534972446,
|
| 1658 |
+
"grad_norm": 0.2721373138715926,
|
| 1659 |
+
"learning_rate": 4.222018657571761e-05,
|
| 1660 |
+
"loss": 0.6636,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.6074607884696905,
|
| 1665 |
+
"grad_norm": 0.23208696219893624,
|
| 1666 |
+
"learning_rate": 4.19032766329497e-05,
|
| 1667 |
+
"loss": 0.6701,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.6142433234421365,
|
| 1672 |
+
"grad_norm": 0.32166382447684766,
|
| 1673 |
+
"learning_rate": 4.1586246903368496e-05,
|
| 1674 |
+
"loss": 0.6786,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.6210258584145825,
|
| 1679 |
+
"grad_norm": 0.24677680563722856,
|
| 1680 |
+
"learning_rate": 4.126911733992272e-05,
|
| 1681 |
+
"loss": 0.6791,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.6278083933870284,
|
| 1686 |
+
"grad_norm": 0.18800235990418823,
|
| 1687 |
+
"learning_rate": 4.0951907901844296e-05,
|
| 1688 |
+
"loss": 0.6591,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.6345909283594744,
|
| 1693 |
+
"grad_norm": 0.18086364097944294,
|
| 1694 |
+
"learning_rate": 4.063463855339232e-05,
|
| 1695 |
+
"loss": 0.6607,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.6413734633319204,
|
| 1700 |
+
"grad_norm": 0.1816725478249145,
|
| 1701 |
+
"learning_rate": 4.031732926259639e-05,
|
| 1702 |
+
"loss": 0.6636,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.6481559983043663,
|
| 1707 |
+
"grad_norm": 0.2179460901904272,
|
| 1708 |
+
"learning_rate": 4e-05,
|
| 1709 |
+
"loss": 0.6582,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.654938533276812,
|
| 1714 |
+
"grad_norm": 0.2408536943475412,
|
| 1715 |
+
"learning_rate": 3.9682670737403624e-05,
|
| 1716 |
+
"loss": 0.6581,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.6617210682492582,
|
| 1721 |
+
"grad_norm": 0.19000091863164711,
|
| 1722 |
+
"learning_rate": 3.9365361446607684e-05,
|
| 1723 |
+
"loss": 0.6597,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.668503603221704,
|
| 1728 |
+
"grad_norm": 0.1604164527075533,
|
| 1729 |
+
"learning_rate": 3.904809209815571e-05,
|
| 1730 |
+
"loss": 0.6609,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.6752861381941502,
|
| 1735 |
+
"grad_norm": 0.20780114182132747,
|
| 1736 |
+
"learning_rate": 3.87308826600773e-05,
|
| 1737 |
+
"loss": 0.6588,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.682068673166596,
|
| 1742 |
+
"grad_norm": 0.21684423863552219,
|
| 1743 |
+
"learning_rate": 3.841375309663151e-05,
|
| 1744 |
+
"loss": 0.6688,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.688851208139042,
|
| 1749 |
+
"grad_norm": 0.21044006155176154,
|
| 1750 |
+
"learning_rate": 3.809672336705031e-05,
|
| 1751 |
+
"loss": 0.6711,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.6956337431114878,
|
| 1756 |
+
"grad_norm": 0.1911074700409385,
|
| 1757 |
+
"learning_rate": 3.7779813424282404e-05,
|
| 1758 |
+
"loss": 0.676,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.702416278083934,
|
| 1763 |
+
"grad_norm": 0.20477853050090425,
|
| 1764 |
+
"learning_rate": 3.746304321373742e-05,
|
| 1765 |
+
"loss": 0.6694,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.7091988130563798,
|
| 1770 |
+
"grad_norm": 0.2337501696824405,
|
| 1771 |
+
"learning_rate": 3.714643267203071e-05,
|
| 1772 |
+
"loss": 0.6545,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.7159813480288257,
|
| 1777 |
+
"grad_norm": 0.174865590490607,
|
| 1778 |
+
"learning_rate": 3.683000172572846e-05,
|
| 1779 |
+
"loss": 0.6766,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.7227638830012717,
|
| 1784 |
+
"grad_norm": 0.21609429620617512,
|
| 1785 |
+
"learning_rate": 3.6513770290093674e-05,
|
| 1786 |
+
"loss": 0.6574,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.7295464179737177,
|
| 1791 |
+
"grad_norm": 0.24191345936732944,
|
| 1792 |
+
"learning_rate": 3.6197758267832705e-05,
|
| 1793 |
+
"loss": 0.6635,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.7363289529461636,
|
| 1798 |
+
"grad_norm": 0.21998311775904844,
|
| 1799 |
+
"learning_rate": 3.5881985547842627e-05,
|
| 1800 |
+
"loss": 0.6676,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.7431114879186096,
|
| 1805 |
+
"grad_norm": 0.23724435100092056,
|
| 1806 |
+
"learning_rate": 3.556647200395956e-05,
|
| 1807 |
+
"loss": 0.6624,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.7498940228910556,
|
| 1812 |
+
"grad_norm": 0.240215596344218,
|
| 1813 |
+
"learning_rate": 3.5251237493707804e-05,
|
| 1814 |
+
"loss": 0.6526,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.7566765578635015,
|
| 1819 |
+
"grad_norm": 0.19872673849172526,
|
| 1820 |
+
"learning_rate": 3.493630185705003e-05,
|
| 1821 |
+
"loss": 0.6752,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.7634590928359475,
|
| 1826 |
+
"grad_norm": 0.23055129025202933,
|
| 1827 |
+
"learning_rate": 3.462168491513873e-05,
|
| 1828 |
+
"loss": 0.6537,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.7702416278083932,
|
| 1833 |
+
"grad_norm": 0.18151089866178272,
|
| 1834 |
+
"learning_rate": 3.4307406469068604e-05,
|
| 1835 |
+
"loss": 0.6699,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.7770241627808394,
|
| 1840 |
+
"grad_norm": 0.21466329451747204,
|
| 1841 |
+
"learning_rate": 3.3993486298630384e-05,
|
| 1842 |
+
"loss": 0.6505,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.7838066977532852,
|
| 1847 |
+
"grad_norm": 0.22222597418898904,
|
| 1848 |
+
"learning_rate": 3.367994416106601e-05,
|
| 1849 |
+
"loss": 0.6646,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.7905892327257313,
|
| 1854 |
+
"grad_norm": 0.17502941394521884,
|
| 1855 |
+
"learning_rate": 3.3366799789825044e-05,
|
| 1856 |
+
"loss": 0.6641,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.797371767698177,
|
| 1861 |
+
"grad_norm": 0.24125561495280676,
|
| 1862 |
+
"learning_rate": 3.305407289332279e-05,
|
| 1863 |
+
"loss": 0.6591,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.8041543026706233,
|
| 1868 |
+
"grad_norm": 0.16946358730913183,
|
| 1869 |
+
"learning_rate": 3.27417831536999e-05,
|
| 1870 |
+
"loss": 0.6607,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.810936837643069,
|
| 1875 |
+
"grad_norm": 0.2021667757386069,
|
| 1876 |
+
"learning_rate": 3.2429950225583606e-05,
|
| 1877 |
+
"loss": 0.6578,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.8177193726155152,
|
| 1882 |
+
"grad_norm": 0.21522045832395853,
|
| 1883 |
+
"learning_rate": 3.2118593734850686e-05,
|
| 1884 |
+
"loss": 0.6666,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.824501907587961,
|
| 1889 |
+
"grad_norm": 0.1955912817171962,
|
| 1890 |
+
"learning_rate": 3.180773327739238e-05,
|
| 1891 |
+
"loss": 0.6666,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.831284442560407,
|
| 1896 |
+
"grad_norm": 0.17543010393599448,
|
| 1897 |
+
"learning_rate": 3.1497388417880935e-05,
|
| 1898 |
+
"loss": 0.6689,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.8380669775328529,
|
| 1903 |
+
"grad_norm": 0.16672656731926996,
|
| 1904 |
+
"learning_rate": 3.118757868853838e-05,
|
| 1905 |
+
"loss": 0.6618,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.8448495125052988,
|
| 1910 |
+
"grad_norm": 0.17242532075303685,
|
| 1911 |
+
"learning_rate": 3.087832358790715e-05,
|
| 1912 |
+
"loss": 0.6516,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.8516320474777448,
|
| 1917 |
+
"grad_norm": 0.15372909956958672,
|
| 1918 |
+
"learning_rate": 3.0569642579622905e-05,
|
| 1919 |
+
"loss": 0.6632,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.8584145824501908,
|
| 1924 |
+
"grad_norm": 0.15847716517543572,
|
| 1925 |
+
"learning_rate": 3.0261555091189648e-05,
|
| 1926 |
+
"loss": 0.6652,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.8651971174226367,
|
| 1931 |
+
"grad_norm": 0.16812640824945566,
|
| 1932 |
+
"learning_rate": 2.9954080512756836e-05,
|
| 1933 |
+
"loss": 0.6544,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.8719796523950827,
|
| 1938 |
+
"grad_norm": 0.14101975804977546,
|
| 1939 |
+
"learning_rate": 2.9647238195899168e-05,
|
| 1940 |
+
"loss": 0.6744,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.8787621873675286,
|
| 1945 |
+
"grad_norm": 0.16388464384133353,
|
| 1946 |
+
"learning_rate": 2.9341047452398607e-05,
|
| 1947 |
+
"loss": 0.655,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.8855447223399746,
|
| 1952 |
+
"grad_norm": 0.17548344056861861,
|
| 1953 |
+
"learning_rate": 2.9035527553028906e-05,
|
| 1954 |
+
"loss": 0.6719,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 1.8923272573124206,
|
| 1959 |
+
"grad_norm": 0.1422728380203633,
|
| 1960 |
+
"learning_rate": 2.873069772634281e-05,
|
| 1961 |
+
"loss": 0.6614,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.8991097922848663,
|
| 1966 |
+
"grad_norm": 0.16674659478989506,
|
| 1967 |
+
"learning_rate": 2.8426577157461897e-05,
|
| 1968 |
+
"loss": 0.6564,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.9058923272573125,
|
| 1973 |
+
"grad_norm": 0.15682390820225428,
|
| 1974 |
+
"learning_rate": 2.8123184986869022e-05,
|
| 1975 |
+
"loss": 0.6683,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.9126748622297582,
|
| 1980 |
+
"grad_norm": 0.1734396291500367,
|
| 1981 |
+
"learning_rate": 2.7820540309203728e-05,
|
| 1982 |
+
"loss": 0.6655,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.9194573972022044,
|
| 1987 |
+
"grad_norm": 0.1467568096698091,
|
| 1988 |
+
"learning_rate": 2.751866217206052e-05,
|
| 1989 |
+
"loss": 0.6605,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.9262399321746502,
|
| 1994 |
+
"grad_norm": 0.17404162787939303,
|
| 1995 |
+
"learning_rate": 2.721756957479001e-05,
|
| 1996 |
+
"loss": 0.6451,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.9330224671470964,
|
| 2001 |
+
"grad_norm": 0.16431154167988282,
|
| 2002 |
+
"learning_rate": 2.691728146730314e-05,
|
| 2003 |
+
"loss": 0.6717,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.939805002119542,
|
| 2008 |
+
"grad_norm": 0.1311637148885735,
|
| 2009 |
+
"learning_rate": 2.6617816748878595e-05,
|
| 2010 |
+
"loss": 0.6779,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 1.9465875370919883,
|
| 2015 |
+
"grad_norm": 0.15022746394032438,
|
| 2016 |
+
"learning_rate": 2.6319194266973256e-05,
|
| 2017 |
+
"loss": 0.6642,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.953370072064434,
|
| 2022 |
+
"grad_norm": 0.13388633011558382,
|
| 2023 |
+
"learning_rate": 2.6021432816036073e-05,
|
| 2024 |
+
"loss": 0.6673,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.96015260703688,
|
| 2029 |
+
"grad_norm": 0.15469686766213264,
|
| 2030 |
+
"learning_rate": 2.5724551136325132e-05,
|
| 2031 |
+
"loss": 0.6641,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.966935142009326,
|
| 2036 |
+
"grad_norm": 0.13266678943395285,
|
| 2037 |
+
"learning_rate": 2.5428567912728225e-05,
|
| 2038 |
+
"loss": 0.6524,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 1.973717676981772,
|
| 2043 |
+
"grad_norm": 0.14468475238604925,
|
| 2044 |
+
"learning_rate": 2.5133501773586905e-05,
|
| 2045 |
+
"loss": 0.6671,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.9805002119542179,
|
| 2050 |
+
"grad_norm": 0.14376432351647073,
|
| 2051 |
+
"learning_rate": 2.483937128952399e-05,
|
| 2052 |
+
"loss": 0.6609,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 1.9872827469266638,
|
| 2057 |
+
"grad_norm": 0.159716122840683,
|
| 2058 |
+
"learning_rate": 2.4546194972274852e-05,
|
| 2059 |
+
"loss": 0.67,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.9940652818991098,
|
| 2064 |
+
"grad_norm": 0.13840709480527438,
|
| 2065 |
+
"learning_rate": 2.425399127352235e-05,
|
| 2066 |
+
"loss": 0.6613,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.0008478168715556,
|
| 2071 |
+
"grad_norm": 0.23463615437325822,
|
| 2072 |
+
"learning_rate": 2.3962778583735463e-05,
|
| 2073 |
+
"loss": 1.0436,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.0076303518440017,
|
| 2078 |
+
"grad_norm": 0.17334268341845893,
|
| 2079 |
+
"learning_rate": 2.3672575231011888e-05,
|
| 2080 |
+
"loss": 0.6518,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.0144128868164475,
|
| 2085 |
+
"grad_norm": 0.16319357693293915,
|
| 2086 |
+
"learning_rate": 2.338339947992455e-05,
|
| 2087 |
+
"loss": 0.6541,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.0211954217888937,
|
| 2092 |
+
"grad_norm": 0.1642296066509337,
|
| 2093 |
+
"learning_rate": 2.3095269530372032e-05,
|
| 2094 |
+
"loss": 0.642,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.0279779567613394,
|
| 2099 |
+
"grad_norm": 0.14714181360085615,
|
| 2100 |
+
"learning_rate": 2.2808203516433136e-05,
|
| 2101 |
+
"loss": 0.6585,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.0347604917337856,
|
| 2106 |
+
"grad_norm": 0.17243791907969416,
|
| 2107 |
+
"learning_rate": 2.2522219505225627e-05,
|
| 2108 |
+
"loss": 0.6367,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.0415430267062313,
|
| 2113 |
+
"grad_norm": 0.15616193111829843,
|
| 2114 |
+
"learning_rate": 2.2237335495769035e-05,
|
| 2115 |
+
"loss": 0.6435,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.0483255616786775,
|
| 2120 |
+
"grad_norm": 0.16853105336890806,
|
| 2121 |
+
"learning_rate": 2.1953569417851983e-05,
|
| 2122 |
+
"loss": 0.6512,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.0551080966511233,
|
| 2127 |
+
"grad_norm": 0.1491939764011879,
|
| 2128 |
+
"learning_rate": 2.1670939130903585e-05,
|
| 2129 |
+
"loss": 0.6421,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.0618906316235694,
|
| 2134 |
+
"grad_norm": 0.15064138598058863,
|
| 2135 |
+
"learning_rate": 2.1389462422869576e-05,
|
| 2136 |
+
"loss": 0.6449,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.068673166596015,
|
| 2141 |
+
"grad_norm": 0.14394176360070499,
|
| 2142 |
+
"learning_rate": 2.11091570090927e-05,
|
| 2143 |
+
"loss": 0.645,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.0754557015684614,
|
| 2148 |
+
"grad_norm": 0.14131247697973692,
|
| 2149 |
+
"learning_rate": 2.0830040531197744e-05,
|
| 2150 |
+
"loss": 0.6486,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.082238236540907,
|
| 2155 |
+
"grad_norm": 0.13739301329908918,
|
| 2156 |
+
"learning_rate": 2.055213055598126e-05,
|
| 2157 |
+
"loss": 0.6294,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.0890207715133533,
|
| 2162 |
+
"grad_norm": 0.14567983415785407,
|
| 2163 |
+
"learning_rate": 2.027544457430599e-05,
|
| 2164 |
+
"loss": 0.6533,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.095803306485799,
|
| 2169 |
+
"grad_norm": 0.12128606904438433,
|
| 2170 |
+
"learning_rate": 2.0000000000000012e-05,
|
| 2171 |
+
"loss": 0.6439,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.1025858414582452,
|
| 2176 |
+
"grad_norm": 0.13694286534266129,
|
| 2177 |
+
"learning_rate": 1.9725814168760688e-05,
|
| 2178 |
+
"loss": 0.6326,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.109368376430691,
|
| 2183 |
+
"grad_norm": 0.11981049040505881,
|
| 2184 |
+
"learning_rate": 1.9452904337063757e-05,
|
| 2185 |
+
"loss": 0.6294,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.1161509114031367,
|
| 2190 |
+
"grad_norm": 0.14362025503542125,
|
| 2191 |
+
"learning_rate": 1.9181287681077116e-05,
|
| 2192 |
+
"loss": 0.6391,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.122933446375583,
|
| 2197 |
+
"grad_norm": 0.13487152713963088,
|
| 2198 |
+
"learning_rate": 1.8910981295579903e-05,
|
| 2199 |
+
"loss": 0.6363,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.1297159813480286,
|
| 2204 |
+
"grad_norm": 0.13488125276633461,
|
| 2205 |
+
"learning_rate": 1.864200219288656e-05,
|
| 2206 |
+
"loss": 0.6499,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.136498516320475,
|
| 2211 |
+
"grad_norm": 0.11819804560452499,
|
| 2212 |
+
"learning_rate": 1.8374367301776112e-05,
|
| 2213 |
+
"loss": 0.6484,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.1432810512929206,
|
| 2218 |
+
"grad_norm": 0.14333365446921437,
|
| 2219 |
+
"learning_rate": 1.8108093466426713e-05,
|
| 2220 |
+
"loss": 0.6517,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.1500635862653668,
|
| 2225 |
+
"grad_norm": 0.11619579623506068,
|
| 2226 |
+
"learning_rate": 1.7843197445355593e-05,
|
| 2227 |
+
"loss": 0.6424,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.1568461212378125,
|
| 2232 |
+
"grad_norm": 0.11214714038576876,
|
| 2233 |
+
"learning_rate": 1.7579695910364235e-05,
|
| 2234 |
+
"loss": 0.6327,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.1636286562102587,
|
| 2239 |
+
"grad_norm": 0.11836436894373582,
|
| 2240 |
+
"learning_rate": 1.7317605445489178e-05,
|
| 2241 |
+
"loss": 0.6332,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.1704111911827044,
|
| 2246 |
+
"grad_norm": 0.12190048890581265,
|
| 2247 |
+
"learning_rate": 1.7056942545958167e-05,
|
| 2248 |
+
"loss": 0.6664,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.1771937261551506,
|
| 2253 |
+
"grad_norm": 0.12475488169306143,
|
| 2254 |
+
"learning_rate": 1.679772361715208e-05,
|
| 2255 |
+
"loss": 0.6411,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.1839762611275964,
|
| 2260 |
+
"grad_norm": 0.13291841960547746,
|
| 2261 |
+
"learning_rate": 1.6539964973572388e-05,
|
| 2262 |
+
"loss": 0.6454,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.1907587961000425,
|
| 2267 |
+
"grad_norm": 0.12175988658418246,
|
| 2268 |
+
"learning_rate": 1.6283682837814388e-05,
|
| 2269 |
+
"loss": 0.646,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.1975413310724883,
|
| 2274 |
+
"grad_norm": 0.11016625987010713,
|
| 2275 |
+
"learning_rate": 1.6028893339546122e-05,
|
| 2276 |
+
"loss": 0.6326,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.2043238660449345,
|
| 2281 |
+
"grad_norm": 0.1108749966998608,
|
| 2282 |
+
"learning_rate": 1.5775612514493343e-05,
|
| 2283 |
+
"loss": 0.6461,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.21110640101738,
|
| 2288 |
+
"grad_norm": 0.1143880178225703,
|
| 2289 |
+
"learning_rate": 1.552385630343014e-05,
|
| 2290 |
+
"loss": 0.6372,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.2178889359898264,
|
| 2295 |
+
"grad_norm": 0.11233236110610757,
|
| 2296 |
+
"learning_rate": 1.527364055117579e-05,
|
| 2297 |
+
"loss": 0.6286,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.224671470962272,
|
| 2302 |
+
"grad_norm": 0.11855577442735764,
|
| 2303 |
+
"learning_rate": 1.502498100559747e-05,
|
| 2304 |
+
"loss": 0.6439,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.231454005934718,
|
| 2309 |
+
"grad_norm": 0.12612061766992566,
|
| 2310 |
+
"learning_rate": 1.4777893316619114e-05,
|
| 2311 |
+
"loss": 0.6369,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.238236540907164,
|
| 2316 |
+
"grad_norm": 0.10643244567354232,
|
| 2317 |
+
"learning_rate": 1.4532393035236477e-05,
|
| 2318 |
+
"loss": 0.6385,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.24501907587961,
|
| 2323 |
+
"grad_norm": 0.1173067781996439,
|
| 2324 |
+
"learning_rate": 1.4288495612538427e-05,
|
| 2325 |
+
"loss": 0.6339,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.251801610852056,
|
| 2330 |
+
"grad_norm": 0.11386078478518308,
|
| 2331 |
+
"learning_rate": 1.404621639873447e-05,
|
| 2332 |
+
"loss": 0.6397,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.2585841458245017,
|
| 2337 |
+
"grad_norm": 0.09826930234307879,
|
| 2338 |
+
"learning_rate": 1.3805570642188602e-05,
|
| 2339 |
+
"loss": 0.645,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.265366680796948,
|
| 2344 |
+
"grad_norm": 0.11305991403115098,
|
| 2345 |
+
"learning_rate": 1.3566573488459729e-05,
|
| 2346 |
+
"loss": 0.6449,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.2721492157693937,
|
| 2351 |
+
"grad_norm": 0.11019252840128578,
|
| 2352 |
+
"learning_rate": 1.3329239979348341e-05,
|
| 2353 |
+
"loss": 0.64,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.27893175074184,
|
| 2358 |
+
"grad_norm": 0.11234819489299995,
|
| 2359 |
+
"learning_rate": 1.30935850519499e-05,
|
| 2360 |
+
"loss": 0.6514,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.2857142857142856,
|
| 2365 |
+
"grad_norm": 0.1112071135535242,
|
| 2366 |
+
"learning_rate": 1.2859623537714719e-05,
|
| 2367 |
+
"loss": 0.6551,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.2924968206867318,
|
| 2372 |
+
"grad_norm": 0.131859623404786,
|
| 2373 |
+
"learning_rate": 1.262737016151447e-05,
|
| 2374 |
+
"loss": 0.6299,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.2992793556591775,
|
| 2379 |
+
"grad_norm": 0.11621464001477737,
|
| 2380 |
+
"learning_rate": 1.2396839540715528e-05,
|
| 2381 |
+
"loss": 0.6316,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.3060618906316237,
|
| 2386 |
+
"grad_norm": 0.10479280516442933,
|
| 2387 |
+
"learning_rate": 1.2168046184258896e-05,
|
| 2388 |
+
"loss": 0.6284,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.3128444256040694,
|
| 2393 |
+
"grad_norm": 0.10786907647373126,
|
| 2394 |
+
"learning_rate": 1.1941004491747145e-05,
|
| 2395 |
+
"loss": 0.6436,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.3196269605765156,
|
| 2400 |
+
"grad_norm": 0.10782220614805556,
|
| 2401 |
+
"learning_rate": 1.1715728752538103e-05,
|
| 2402 |
+
"loss": 0.644,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.3264094955489614,
|
| 2407 |
+
"grad_norm": 0.11055552541032848,
|
| 2408 |
+
"learning_rate": 1.1492233144845497e-05,
|
| 2409 |
+
"loss": 0.654,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.3331920305214076,
|
| 2414 |
+
"grad_norm": 0.11811567288889502,
|
| 2415 |
+
"learning_rate": 1.1270531734846642e-05,
|
| 2416 |
+
"loss": 0.6383,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.3399745654938533,
|
| 2421 |
+
"grad_norm": 0.09614027593231717,
|
| 2422 |
+
"learning_rate": 1.1050638475797193e-05,
|
| 2423 |
+
"loss": 0.6486,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 2.3467571004662995,
|
| 2428 |
+
"grad_norm": 0.11607457015859846,
|
| 2429 |
+
"learning_rate": 1.083256720715292e-05,
|
| 2430 |
+
"loss": 0.6404,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 2.3535396354387452,
|
| 2435 |
+
"grad_norm": 0.10091742280729821,
|
| 2436 |
+
"learning_rate": 1.0616331653698673e-05,
|
| 2437 |
+
"loss": 0.6532,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 2.360322170411191,
|
| 2442 |
+
"grad_norm": 0.09008242164464271,
|
| 2443 |
+
"learning_rate": 1.0401945424684653e-05,
|
| 2444 |
+
"loss": 0.6345,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.367104705383637,
|
| 2449 |
+
"grad_norm": 0.11007729141322087,
|
| 2450 |
+
"learning_rate": 1.0189422012969814e-05,
|
| 2451 |
+
"loss": 0.6482,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.373887240356083,
|
| 2456 |
+
"grad_norm": 0.08713293928424415,
|
| 2457 |
+
"learning_rate": 9.978774794172717e-06,
|
| 2458 |
+
"loss": 0.6243,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.380669775328529,
|
| 2463 |
+
"grad_norm": 0.09736881200508285,
|
| 2464 |
+
"learning_rate": 9.770017025829675e-06,
|
| 2465 |
+
"loss": 0.6464,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 2.387452310300975,
|
| 2470 |
+
"grad_norm": 0.09620789674440036,
|
| 2471 |
+
"learning_rate": 9.563161846560342e-06,
|
| 2472 |
+
"loss": 0.6399,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 2.394234845273421,
|
| 2477 |
+
"grad_norm": 0.1013119106642466,
|
| 2478 |
+
"learning_rate": 9.358222275240884e-06,
|
| 2479 |
+
"loss": 0.6369,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 2.4010173802458668,
|
| 2484 |
+
"grad_norm": 0.0916584684509351,
|
| 2485 |
+
"learning_rate": 9.155211210184495e-06,
|
| 2486 |
+
"loss": 0.6389,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.407799915218313,
|
| 2491 |
+
"grad_norm": 0.10238057837102806,
|
| 2492 |
+
"learning_rate": 8.954141428329723e-06,
|
| 2493 |
+
"loss": 0.6377,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 2.4145824501907587,
|
| 2498 |
+
"grad_norm": 0.10324400627592539,
|
| 2499 |
+
"learning_rate": 8.755025584436266e-06,
|
| 2500 |
+
"loss": 0.6376,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.421364985163205,
|
| 2505 |
+
"grad_norm": 0.10057048511977558,
|
| 2506 |
+
"learning_rate": 8.55787621028851e-06,
|
| 2507 |
+
"loss": 0.6506,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 2.4281475201356506,
|
| 2512 |
+
"grad_norm": 0.0920558216849332,
|
| 2513 |
+
"learning_rate": 8.362705713906836e-06,
|
| 2514 |
+
"loss": 0.6313,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 2.434930055108097,
|
| 2519 |
+
"grad_norm": 0.10610753513988694,
|
| 2520 |
+
"learning_rate": 8.169526378766713e-06,
|
| 2521 |
+
"loss": 0.6429,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 2.4417125900805425,
|
| 2526 |
+
"grad_norm": 0.10421896294296241,
|
| 2527 |
+
"learning_rate": 7.978350363025588e-06,
|
| 2528 |
+
"loss": 0.6407,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 2.4484951250529887,
|
| 2533 |
+
"grad_norm": 0.09993739787650231,
|
| 2534 |
+
"learning_rate": 7.789189698757656e-06,
|
| 2535 |
+
"loss": 0.6348,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 2.4552776600254345,
|
| 2540 |
+
"grad_norm": 0.09901740077583913,
|
| 2541 |
+
"learning_rate": 7.602056291196671e-06,
|
| 2542 |
+
"loss": 0.6459,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 2.4620601949978806,
|
| 2547 |
+
"grad_norm": 0.10236110912446401,
|
| 2548 |
+
"learning_rate": 7.416961917986572e-06,
|
| 2549 |
+
"loss": 0.6509,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 2.4688427299703264,
|
| 2554 |
+
"grad_norm": 0.08705071344819373,
|
| 2555 |
+
"learning_rate": 7.233918228440324e-06,
|
| 2556 |
+
"loss": 0.6348,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 2.4756252649427726,
|
| 2561 |
+
"grad_norm": 0.11881375541999774,
|
| 2562 |
+
"learning_rate": 7.052936742806693e-06,
|
| 2563 |
+
"loss": 0.646,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 2.4824077999152183,
|
| 2568 |
+
"grad_norm": 0.10091840020224621,
|
| 2569 |
+
"learning_rate": 6.874028851545174e-06,
|
| 2570 |
+
"loss": 0.64,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 2.489190334887664,
|
| 2575 |
+
"grad_norm": 0.0828156389233352,
|
| 2576 |
+
"learning_rate": 6.697205814609149e-06,
|
| 2577 |
+
"loss": 0.6324,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 2.4959728698601102,
|
| 2582 |
+
"grad_norm": 0.09511789444296942,
|
| 2583 |
+
"learning_rate": 6.522478760737212e-06,
|
| 2584 |
+
"loss": 0.6398,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 2.5027554048325564,
|
| 2589 |
+
"grad_norm": 0.10690963856134064,
|
| 2590 |
+
"learning_rate": 6.349858686752748e-06,
|
| 2591 |
+
"loss": 0.6441,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 2.509537939805002,
|
| 2596 |
+
"grad_norm": 0.10170196889030197,
|
| 2597 |
+
"learning_rate": 6.17935645687183e-06,
|
| 2598 |
+
"loss": 0.6307,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 2.516320474777448,
|
| 2603 |
+
"grad_norm": 0.0873576460247102,
|
| 2604 |
+
"learning_rate": 6.010982802019429e-06,
|
| 2605 |
+
"loss": 0.6382,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 2.523103009749894,
|
| 2610 |
+
"grad_norm": 0.10193237211341288,
|
| 2611 |
+
"learning_rate": 5.8447483191540784e-06,
|
| 2612 |
+
"loss": 0.6427,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 2.52988554472234,
|
| 2617 |
+
"grad_norm": 0.09337413798566872,
|
| 2618 |
+
"learning_rate": 5.680663470600918e-06,
|
| 2619 |
+
"loss": 0.6479,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 2.536668079694786,
|
| 2624 |
+
"grad_norm": 0.08453984752561841,
|
| 2625 |
+
"learning_rate": 5.518738583393228e-06,
|
| 2626 |
+
"loss": 0.6508,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 2.5434506146672318,
|
| 2631 |
+
"grad_norm": 0.08158348255933849,
|
| 2632 |
+
"learning_rate": 5.358983848622452e-06,
|
| 2633 |
+
"loss": 0.6372,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 2.550233149639678,
|
| 2638 |
+
"grad_norm": 0.09556063525511022,
|
| 2639 |
+
"learning_rate": 5.201409320796842e-06,
|
| 2640 |
+
"loss": 0.6397,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 2.5570156846121237,
|
| 2645 |
+
"grad_norm": 0.09622696560627544,
|
| 2646 |
+
"learning_rate": 5.046024917208603e-06,
|
| 2647 |
+
"loss": 0.6386,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 2.56379821958457,
|
| 2652 |
+
"grad_norm": 0.09148441726550699,
|
| 2653 |
+
"learning_rate": 4.892840417309775e-06,
|
| 2654 |
+
"loss": 0.6489,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 2.5705807545570156,
|
| 2659 |
+
"grad_norm": 0.08606094930168158,
|
| 2660 |
+
"learning_rate": 4.7418654620967216e-06,
|
| 2661 |
+
"loss": 0.6527,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 2.577363289529462,
|
| 2666 |
+
"grad_norm": 0.08364373433134552,
|
| 2667 |
+
"learning_rate": 4.593109553503343e-06,
|
| 2668 |
+
"loss": 0.6428,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 2.5841458245019076,
|
| 2673 |
+
"grad_norm": 0.0843513293417055,
|
| 2674 |
+
"learning_rate": 4.446582053803066e-06,
|
| 2675 |
+
"loss": 0.6408,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 2.5909283594743533,
|
| 2680 |
+
"grad_norm": 0.09217062871361449,
|
| 2681 |
+
"learning_rate": 4.302292185019603e-06,
|
| 2682 |
+
"loss": 0.6496,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 2.5977108944467995,
|
| 2687 |
+
"grad_norm": 0.08394208095826587,
|
| 2688 |
+
"learning_rate": 4.16024902834657e-06,
|
| 2689 |
+
"loss": 0.6407,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 2.6044934294192457,
|
| 2694 |
+
"grad_norm": 0.07675967715086138,
|
| 2695 |
+
"learning_rate": 4.020461523575873e-06,
|
| 2696 |
+
"loss": 0.6497,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 2.6112759643916914,
|
| 2701 |
+
"grad_norm": 0.08253304493993292,
|
| 2702 |
+
"learning_rate": 3.882938468535158e-06,
|
| 2703 |
+
"loss": 0.6473,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 2.618058499364137,
|
| 2708 |
+
"grad_norm": 0.08837276182393047,
|
| 2709 |
+
"learning_rate": 3.747688518534003e-06,
|
| 2710 |
+
"loss": 0.6467,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 2.6248410343365833,
|
| 2715 |
+
"grad_norm": 0.08553692728277926,
|
| 2716 |
+
"learning_rate": 3.6147201858192627e-06,
|
| 2717 |
+
"loss": 0.6271,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 2.6316235693090295,
|
| 2722 |
+
"grad_norm": 0.08672709388396345,
|
| 2723 |
+
"learning_rate": 3.4840418390392895e-06,
|
| 2724 |
+
"loss": 0.6417,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 2.6384061042814753,
|
| 2729 |
+
"grad_norm": 0.08408576121900961,
|
| 2730 |
+
"learning_rate": 3.3556617027172168e-06,
|
| 2731 |
+
"loss": 0.6474,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 2.645188639253921,
|
| 2736 |
+
"grad_norm": 0.08946340052273392,
|
| 2737 |
+
"learning_rate": 3.2295878567333784e-06,
|
| 2738 |
+
"loss": 0.6475,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 2.651971174226367,
|
| 2743 |
+
"grad_norm": 0.10659211899169968,
|
| 2744 |
+
"learning_rate": 3.1058282358167503e-06,
|
| 2745 |
+
"loss": 0.6478,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 2.658753709198813,
|
| 2750 |
+
"grad_norm": 0.08095162713105776,
|
| 2751 |
+
"learning_rate": 2.984390629045555e-06,
|
| 2752 |
+
"loss": 0.6469,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 2.665536244171259,
|
| 2757 |
+
"grad_norm": 0.08002747021341323,
|
| 2758 |
+
"learning_rate": 2.8652826793570975e-06,
|
| 2759 |
+
"loss": 0.625,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 2.672318779143705,
|
| 2764 |
+
"grad_norm": 0.08060733679400718,
|
| 2765 |
+
"learning_rate": 2.7485118830666535e-06,
|
| 2766 |
+
"loss": 0.6419,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 2.679101314116151,
|
| 2771 |
+
"grad_norm": 0.07939209749073067,
|
| 2772 |
+
"learning_rate": 2.634085589395734e-06,
|
| 2773 |
+
"loss": 0.6396,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 2.685883849088597,
|
| 2778 |
+
"grad_norm": 0.08689462211707911,
|
| 2779 |
+
"learning_rate": 2.5220110000095366e-06,
|
| 2780 |
+
"loss": 0.6453,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 2.692666384061043,
|
| 2785 |
+
"grad_norm": 0.08614923525446444,
|
| 2786 |
+
"learning_rate": 2.4122951685636674e-06,
|
| 2787 |
+
"loss": 0.6473,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 2.6994489190334887,
|
| 2792 |
+
"grad_norm": 0.08063088657355912,
|
| 2793 |
+
"learning_rate": 2.3049450002602394e-06,
|
| 2794 |
+
"loss": 0.627,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 2.706231454005935,
|
| 2799 |
+
"grad_norm": 0.08042306304216008,
|
| 2800 |
+
"learning_rate": 2.199967251413262e-06,
|
| 2801 |
+
"loss": 0.6411,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 2.7130139889783806,
|
| 2806 |
+
"grad_norm": 0.07728604571452127,
|
| 2807 |
+
"learning_rate": 2.0973685290234024e-06,
|
| 2808 |
+
"loss": 0.6307,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 2.7197965239508264,
|
| 2813 |
+
"grad_norm": 0.08086443240008037,
|
| 2814 |
+
"learning_rate": 1.997155290362187e-06,
|
| 2815 |
+
"loss": 0.6452,
|
| 2816 |
+
"step": 401
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 2.7265790589232726,
|
| 2820 |
+
"grad_norm": 0.07595495043741796,
|
| 2821 |
+
"learning_rate": 1.8993338425655805e-06,
|
| 2822 |
+
"loss": 0.6472,
|
| 2823 |
+
"step": 402
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 2.7333615938957188,
|
| 2827 |
+
"grad_norm": 0.08185565508209801,
|
| 2828 |
+
"learning_rate": 1.8039103422370452e-06,
|
| 2829 |
+
"loss": 0.6341,
|
| 2830 |
+
"step": 403
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 2.7401441288681645,
|
| 2834 |
+
"grad_norm": 0.0849805109945081,
|
| 2835 |
+
"learning_rate": 1.7108907950600516e-06,
|
| 2836 |
+
"loss": 0.6292,
|
| 2837 |
+
"step": 404
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 2.7469266638406102,
|
| 2841 |
+
"grad_norm": 0.07973338132333087,
|
| 2842 |
+
"learning_rate": 1.6202810554201099e-06,
|
| 2843 |
+
"loss": 0.6437,
|
| 2844 |
+
"step": 405
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 2.7537091988130564,
|
| 2848 |
+
"grad_norm": 0.07790884660042009,
|
| 2849 |
+
"learning_rate": 1.532086826036281e-06,
|
| 2850 |
+
"loss": 0.6456,
|
| 2851 |
+
"step": 406
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 2.760491733785502,
|
| 2855 |
+
"grad_norm": 0.08873860153969253,
|
| 2856 |
+
"learning_rate": 1.4463136576023185e-06,
|
| 2857 |
+
"loss": 0.6502,
|
| 2858 |
+
"step": 407
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 2.7672742687579484,
|
| 2862 |
+
"grad_norm": 0.08041178968421962,
|
| 2863 |
+
"learning_rate": 1.3629669484372722e-06,
|
| 2864 |
+
"loss": 0.6469,
|
| 2865 |
+
"step": 408
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 2.774056803730394,
|
| 2869 |
+
"grad_norm": 0.08336326759495126,
|
| 2870 |
+
"learning_rate": 1.2820519441457502e-06,
|
| 2871 |
+
"loss": 0.6431,
|
| 2872 |
+
"step": 409
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 2.7808393387028403,
|
| 2876 |
+
"grad_norm": 0.07304922054760385,
|
| 2877 |
+
"learning_rate": 1.2035737372877931e-06,
|
| 2878 |
+
"loss": 0.6456,
|
| 2879 |
+
"step": 410
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 2.787621873675286,
|
| 2883 |
+
"grad_norm": 0.07545648509921901,
|
| 2884 |
+
"learning_rate": 1.127537267058334e-06,
|
| 2885 |
+
"loss": 0.6459,
|
| 2886 |
+
"step": 411
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 2.794404408647732,
|
| 2890 |
+
"grad_norm": 0.08168214003465647,
|
| 2891 |
+
"learning_rate": 1.0539473189763583e-06,
|
| 2892 |
+
"loss": 0.6429,
|
| 2893 |
+
"step": 412
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 2.801186943620178,
|
| 2897 |
+
"grad_norm": 0.07428190943785873,
|
| 2898 |
+
"learning_rate": 9.828085245837183e-07,
|
| 2899 |
+
"loss": 0.6341,
|
| 2900 |
+
"step": 413
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 2.807969478592624,
|
| 2904 |
+
"grad_norm": 0.0753237289047626,
|
| 2905 |
+
"learning_rate": 9.141253611536238e-07,
|
| 2906 |
+
"loss": 0.6494,
|
| 2907 |
+
"step": 414
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 2.81475201356507,
|
| 2911 |
+
"grad_norm": 0.07502561278928772,
|
| 2912 |
+
"learning_rate": 8.479021514088548e-07,
|
| 2913 |
+
"loss": 0.6424,
|
| 2914 |
+
"step": 415
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 2.821534548537516,
|
| 2918 |
+
"grad_norm": 0.07379263477638442,
|
| 2919 |
+
"learning_rate": 7.841430632497204e-07,
|
| 2920 |
+
"loss": 0.6426,
|
| 2921 |
+
"step": 416
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 2.828317083509962,
|
| 2925 |
+
"grad_norm": 0.07233921710316674,
|
| 2926 |
+
"learning_rate": 7.228521094917318e-07,
|
| 2927 |
+
"loss": 0.6438,
|
| 2928 |
+
"step": 417
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 2.835099618482408,
|
| 2932 |
+
"grad_norm": 0.07416362872430317,
|
| 2933 |
+
"learning_rate": 6.640331476130435e-07,
|
| 2934 |
+
"loss": 0.6405,
|
| 2935 |
+
"step": 418
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 2.8418821534548537,
|
| 2939 |
+
"grad_norm": 0.07700076601803095,
|
| 2940 |
+
"learning_rate": 6.076898795116792e-07,
|
| 2941 |
+
"loss": 0.6489,
|
| 2942 |
+
"step": 419
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 2.8486646884272995,
|
| 2946 |
+
"grad_norm": 0.07609683257441903,
|
| 2947 |
+
"learning_rate": 5.538258512725403e-07,
|
| 2948 |
+
"loss": 0.6209,
|
| 2949 |
+
"step": 420
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 2.8554472233997457,
|
| 2953 |
+
"grad_norm": 0.07737901602583444,
|
| 2954 |
+
"learning_rate": 5.024444529442285e-07,
|
| 2955 |
+
"loss": 0.6447,
|
| 2956 |
+
"step": 421
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 2.862229758372192,
|
| 2960 |
+
"grad_norm": 0.0759561526572301,
|
| 2961 |
+
"learning_rate": 4.5354891832569245e-07,
|
| 2962 |
+
"loss": 0.6327,
|
| 2963 |
+
"step": 422
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 2.8690122933446376,
|
| 2967 |
+
"grad_norm": 0.07540799112373622,
|
| 2968 |
+
"learning_rate": 4.0714232476269265e-07,
|
| 2969 |
+
"loss": 0.6357,
|
| 2970 |
+
"step": 423
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 2.8757948283170833,
|
| 2974 |
+
"grad_norm": 0.07261245642997437,
|
| 2975 |
+
"learning_rate": 3.632275929541207e-07,
|
| 2976 |
+
"loss": 0.6384,
|
| 2977 |
+
"step": 424
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 2.8825773632895295,
|
| 2981 |
+
"grad_norm": 0.07730392178003218,
|
| 2982 |
+
"learning_rate": 3.218074867681864e-07,
|
| 2983 |
+
"loss": 0.6343,
|
| 2984 |
+
"step": 425
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 2.8893598982619753,
|
| 2988 |
+
"grad_norm": 0.07481289428504849,
|
| 2989 |
+
"learning_rate": 2.8288461306846817e-07,
|
| 2990 |
+
"loss": 0.6418,
|
| 2991 |
+
"step": 426
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 2.8961424332344214,
|
| 2995 |
+
"grad_norm": 0.07365288746369081,
|
| 2996 |
+
"learning_rate": 2.464614215498351e-07,
|
| 2997 |
+
"loss": 0.6399,
|
| 2998 |
+
"step": 427
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 2.902924968206867,
|
| 3002 |
+
"grad_norm": 0.07293600568995603,
|
| 3003 |
+
"learning_rate": 2.1254020458427281e-07,
|
| 3004 |
+
"loss": 0.645,
|
| 3005 |
+
"step": 428
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 2.9097075031793134,
|
| 3009 |
+
"grad_norm": 0.07299059860739354,
|
| 3010 |
+
"learning_rate": 1.8112309707661647e-07,
|
| 3011 |
+
"loss": 0.6366,
|
| 3012 |
+
"step": 429
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 2.916490038151759,
|
| 3016 |
+
"grad_norm": 0.07762603812234999,
|
| 3017 |
+
"learning_rate": 1.522120763301782e-07,
|
| 3018 |
+
"loss": 0.6468,
|
| 3019 |
+
"step": 430
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 2.9232725731242053,
|
| 3023 |
+
"grad_norm": 0.07121635004426441,
|
| 3024 |
+
"learning_rate": 1.2580896192230906e-07,
|
| 3025 |
+
"loss": 0.6457,
|
| 3026 |
+
"step": 431
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 2.930055108096651,
|
| 3030 |
+
"grad_norm": 0.07189028443702318,
|
| 3031 |
+
"learning_rate": 1.019154155898594e-07,
|
| 3032 |
+
"loss": 0.648,
|
| 3033 |
+
"step": 432
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 2.9368376430690972,
|
| 3037 |
+
"grad_norm": 0.07237541136245566,
|
| 3038 |
+
"learning_rate": 8.053294112462696e-08,
|
| 3039 |
+
"loss": 0.6406,
|
| 3040 |
+
"step": 433
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 2.943620178041543,
|
| 3044 |
+
"grad_norm": 0.07059922854737963,
|
| 3045 |
+
"learning_rate": 6.166288427867706e-08,
|
| 3046 |
+
"loss": 0.6349,
|
| 3047 |
+
"step": 434
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 2.9504027130139887,
|
| 3051 |
+
"grad_norm": 0.07048464732580594,
|
| 3052 |
+
"learning_rate": 4.530643267968149e-08,
|
| 3053 |
+
"loss": 0.6532,
|
| 3054 |
+
"step": 435
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 2.957185247986435,
|
| 3058 |
+
"grad_norm": 0.0714872818335937,
|
| 3059 |
+
"learning_rate": 3.1464615756133796e-08,
|
| 3060 |
+
"loss": 0.6323,
|
| 3061 |
+
"step": 436
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 2.963967782958881,
|
| 3065 |
+
"grad_norm": 0.07124837526302868,
|
| 3066 |
+
"learning_rate": 2.01383046725967e-08,
|
| 3067 |
+
"loss": 0.6495,
|
| 3068 |
+
"step": 437
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 2.970750317931327,
|
| 3072 |
+
"grad_norm": 0.0731063695999088,
|
| 3073 |
+
"learning_rate": 1.1328212274839267e-08,
|
| 3074 |
+
"loss": 0.6516,
|
| 3075 |
+
"step": 438
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 2.9775328529037726,
|
| 3079 |
+
"grad_norm": 0.07121633404132001,
|
| 3080 |
+
"learning_rate": 5.034893044997269e-09,
|
| 3081 |
+
"loss": 0.6377,
|
| 3082 |
+
"step": 439
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 2.9843153878762188,
|
| 3086 |
+
"grad_norm": 0.0754923808775944,
|
| 3087 |
+
"learning_rate": 1.2587430666766153e-09,
|
| 3088 |
+
"loss": 0.6443,
|
| 3089 |
+
"step": 440
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 2.991097922848665,
|
| 3093 |
+
"grad_norm": 0.07154903528443775,
|
| 3094 |
+
"learning_rate": 0.0,
|
| 3095 |
+
"loss": 0.6409,
|
| 3096 |
+
"step": 441
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 2.991097922848665,
|
| 3100 |
+
"step": 441,
|
| 3101 |
+
"total_flos": 7397230226964480.0,
|
| 3102 |
+
"train_loss": 0.6955865539120438,
|
| 3103 |
+
"train_runtime": 47632.2951,
|
| 3104 |
+
"train_samples_per_second": 4.753,
|
| 3105 |
+
"train_steps_per_second": 0.009
|
| 3106 |
+
}
|
| 3107 |
+
],
|
| 3108 |
+
"logging_steps": 1,
|
| 3109 |
+
"max_steps": 441,
|
| 3110 |
+
"num_input_tokens_seen": 0,
|
| 3111 |
+
"num_train_epochs": 3,
|
| 3112 |
+
"save_steps": 500,
|
| 3113 |
+
"stateful_callbacks": {
|
| 3114 |
+
"TrainerControl": {
|
| 3115 |
+
"args": {
|
| 3116 |
+
"should_epoch_stop": false,
|
| 3117 |
+
"should_evaluate": false,
|
| 3118 |
+
"should_log": false,
|
| 3119 |
+
"should_save": true,
|
| 3120 |
+
"should_training_stop": true
|
| 3121 |
+
},
|
| 3122 |
+
"attributes": {}
|
| 3123 |
+
}
|
| 3124 |
+
},
|
| 3125 |
+
"total_flos": 7397230226964480.0,
|
| 3126 |
+
"train_batch_size": 1,
|
| 3127 |
+
"trial_name": null,
|
| 3128 |
+
"trial_params": null
|
| 3129 |
+
}
|
training_loss.png
ADDED
|