End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2772 -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: nemo_nano_code_10k
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# nemo_nano_code_10k
|
| 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: nemo_nano_code_10k
|
|
|
|
| 16 |
|
| 17 |
# nemo_nano_code_10k
|
| 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/nemo_nano_code_10k dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.992,
|
| 3 |
+
"total_flos": 8.727962333646684e+17,
|
| 4 |
+
"train_loss": 0.3591920976837476,
|
| 5 |
+
"train_runtime": 40468.1043,
|
| 6 |
+
"train_samples_per_second": 1.236,
|
| 7 |
+
"train_steps_per_second": 0.01
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.992,
|
| 3 |
+
"total_flos": 8.727962333646684e+17,
|
| 4 |
+
"train_loss": 0.3591920976837476,
|
| 5 |
+
"train_runtime": 40468.1043,
|
| 6 |
+
"train_samples_per_second": 1.236,
|
| 7 |
+
"train_steps_per_second": 0.01
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2772 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.992,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 390,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0128,
|
| 13 |
+
"grad_norm": 7.123053613417962,
|
| 14 |
+
"learning_rate": 1.0256410256410257e-06,
|
| 15 |
+
"loss": 1.035,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0256,
|
| 20 |
+
"grad_norm": 7.0772773660633055,
|
| 21 |
+
"learning_rate": 2.0512820512820513e-06,
|
| 22 |
+
"loss": 1.0268,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0384,
|
| 27 |
+
"grad_norm": 7.224069482168115,
|
| 28 |
+
"learning_rate": 3.0769230769230774e-06,
|
| 29 |
+
"loss": 0.9939,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0512,
|
| 34 |
+
"grad_norm": 6.406171402894334,
|
| 35 |
+
"learning_rate": 4.102564102564103e-06,
|
| 36 |
+
"loss": 0.9844,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.064,
|
| 41 |
+
"grad_norm": 5.056161097746435,
|
| 42 |
+
"learning_rate": 5.128205128205128e-06,
|
| 43 |
+
"loss": 0.9174,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0768,
|
| 48 |
+
"grad_norm": 2.9985257852509144,
|
| 49 |
+
"learning_rate": 6.153846153846155e-06,
|
| 50 |
+
"loss": 0.8926,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.0896,
|
| 55 |
+
"grad_norm": 2.412737062597919,
|
| 56 |
+
"learning_rate": 7.17948717948718e-06,
|
| 57 |
+
"loss": 0.8616,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.1024,
|
| 62 |
+
"grad_norm": 3.134868397104596,
|
| 63 |
+
"learning_rate": 8.205128205128205e-06,
|
| 64 |
+
"loss": 0.8082,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1152,
|
| 69 |
+
"grad_norm": 3.3159704097546556,
|
| 70 |
+
"learning_rate": 9.230769230769232e-06,
|
| 71 |
+
"loss": 0.8059,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.128,
|
| 76 |
+
"grad_norm": 3.448587893229317,
|
| 77 |
+
"learning_rate": 1.0256410256410256e-05,
|
| 78 |
+
"loss": 0.7758,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1408,
|
| 83 |
+
"grad_norm": 3.397591217036032,
|
| 84 |
+
"learning_rate": 1.1282051282051283e-05,
|
| 85 |
+
"loss": 0.7898,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.1536,
|
| 90 |
+
"grad_norm": 2.1006434943489696,
|
| 91 |
+
"learning_rate": 1.230769230769231e-05,
|
| 92 |
+
"loss": 0.7259,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1664,
|
| 97 |
+
"grad_norm": 1.8515813858287578,
|
| 98 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 99 |
+
"loss": 0.7181,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.1792,
|
| 104 |
+
"grad_norm": 2.2466859323801636,
|
| 105 |
+
"learning_rate": 1.435897435897436e-05,
|
| 106 |
+
"loss": 0.7167,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.192,
|
| 111 |
+
"grad_norm": 1.7693235334444182,
|
| 112 |
+
"learning_rate": 1.5384615384615387e-05,
|
| 113 |
+
"loss": 0.6574,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.2048,
|
| 118 |
+
"grad_norm": 1.5573639796973424,
|
| 119 |
+
"learning_rate": 1.641025641025641e-05,
|
| 120 |
+
"loss": 0.6196,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2176,
|
| 125 |
+
"grad_norm": 1.087312438452788,
|
| 126 |
+
"learning_rate": 1.7435897435897438e-05,
|
| 127 |
+
"loss": 0.6714,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.2304,
|
| 132 |
+
"grad_norm": 1.43755497127367,
|
| 133 |
+
"learning_rate": 1.8461538461538465e-05,
|
| 134 |
+
"loss": 0.6283,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2432,
|
| 139 |
+
"grad_norm": 1.3816641907233016,
|
| 140 |
+
"learning_rate": 1.9487179487179488e-05,
|
| 141 |
+
"loss": 0.6396,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.256,
|
| 146 |
+
"grad_norm": 1.3335939002054764,
|
| 147 |
+
"learning_rate": 2.0512820512820512e-05,
|
| 148 |
+
"loss": 0.647,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2688,
|
| 153 |
+
"grad_norm": 1.2691529171427989,
|
| 154 |
+
"learning_rate": 2.153846153846154e-05,
|
| 155 |
+
"loss": 0.6408,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.2816,
|
| 160 |
+
"grad_norm": 1.2491274574482119,
|
| 161 |
+
"learning_rate": 2.2564102564102566e-05,
|
| 162 |
+
"loss": 0.6054,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2944,
|
| 167 |
+
"grad_norm": 1.0155946986612912,
|
| 168 |
+
"learning_rate": 2.3589743589743593e-05,
|
| 169 |
+
"loss": 0.615,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.3072,
|
| 174 |
+
"grad_norm": 0.9523725194587307,
|
| 175 |
+
"learning_rate": 2.461538461538462e-05,
|
| 176 |
+
"loss": 0.565,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.32,
|
| 181 |
+
"grad_norm": 0.9357744643738358,
|
| 182 |
+
"learning_rate": 2.5641025641025646e-05,
|
| 183 |
+
"loss": 0.6183,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.3328,
|
| 188 |
+
"grad_norm": 0.8682158439022564,
|
| 189 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 190 |
+
"loss": 0.5538,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3456,
|
| 195 |
+
"grad_norm": 1.0352335204653185,
|
| 196 |
+
"learning_rate": 2.7692307692307694e-05,
|
| 197 |
+
"loss": 0.5848,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.3584,
|
| 202 |
+
"grad_norm": 1.1647136824799569,
|
| 203 |
+
"learning_rate": 2.871794871794872e-05,
|
| 204 |
+
"loss": 0.6,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3712,
|
| 209 |
+
"grad_norm": 0.9488986144363175,
|
| 210 |
+
"learning_rate": 2.9743589743589747e-05,
|
| 211 |
+
"loss": 0.6148,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.384,
|
| 216 |
+
"grad_norm": 0.9286632723576274,
|
| 217 |
+
"learning_rate": 3.0769230769230774e-05,
|
| 218 |
+
"loss": 0.5697,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3968,
|
| 223 |
+
"grad_norm": 1.0085816233375569,
|
| 224 |
+
"learning_rate": 3.1794871794871795e-05,
|
| 225 |
+
"loss": 0.5953,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.4096,
|
| 230 |
+
"grad_norm": 0.8751939356347122,
|
| 231 |
+
"learning_rate": 3.282051282051282e-05,
|
| 232 |
+
"loss": 0.5419,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.4224,
|
| 237 |
+
"grad_norm": 0.7817707345623733,
|
| 238 |
+
"learning_rate": 3.384615384615385e-05,
|
| 239 |
+
"loss": 0.5688,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.4352,
|
| 244 |
+
"grad_norm": 0.977757879353551,
|
| 245 |
+
"learning_rate": 3.4871794871794875e-05,
|
| 246 |
+
"loss": 0.5559,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.448,
|
| 251 |
+
"grad_norm": 0.8358613617680577,
|
| 252 |
+
"learning_rate": 3.58974358974359e-05,
|
| 253 |
+
"loss": 0.5521,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.4608,
|
| 258 |
+
"grad_norm": 1.0786205356480556,
|
| 259 |
+
"learning_rate": 3.692307692307693e-05,
|
| 260 |
+
"loss": 0.55,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4736,
|
| 265 |
+
"grad_norm": 0.8778605192965089,
|
| 266 |
+
"learning_rate": 3.794871794871795e-05,
|
| 267 |
+
"loss": 0.5394,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.4864,
|
| 272 |
+
"grad_norm": 0.9668049439409904,
|
| 273 |
+
"learning_rate": 3.8974358974358976e-05,
|
| 274 |
+
"loss": 0.586,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.4992,
|
| 279 |
+
"grad_norm": 1.3231030163531852,
|
| 280 |
+
"learning_rate": 4e-05,
|
| 281 |
+
"loss": 0.548,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.512,
|
| 286 |
+
"grad_norm": 1.0433463303841382,
|
| 287 |
+
"learning_rate": 3.9999198907597046e-05,
|
| 288 |
+
"loss": 0.5507,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.5248,
|
| 293 |
+
"grad_norm": 1.041525340076879,
|
| 294 |
+
"learning_rate": 3.9996795694563096e-05,
|
| 295 |
+
"loss": 0.5599,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.5376,
|
| 300 |
+
"grad_norm": 0.9144604553010471,
|
| 301 |
+
"learning_rate": 3.999279055341771e-05,
|
| 302 |
+
"loss": 0.5518,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5504,
|
| 307 |
+
"grad_norm": 0.8332232775924605,
|
| 308 |
+
"learning_rate": 3.998718380500971e-05,
|
| 309 |
+
"loss": 0.5781,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5632,
|
| 314 |
+
"grad_norm": 0.7041324823131757,
|
| 315 |
+
"learning_rate": 3.997997589849145e-05,
|
| 316 |
+
"loss": 0.5689,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.576,
|
| 321 |
+
"grad_norm": 0.8879449244016675,
|
| 322 |
+
"learning_rate": 3.9971167411282835e-05,
|
| 323 |
+
"loss": 0.5588,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5888,
|
| 328 |
+
"grad_norm": 0.8426817508013171,
|
| 329 |
+
"learning_rate": 3.99607590490251e-05,
|
| 330 |
+
"loss": 0.5555,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.6016,
|
| 335 |
+
"grad_norm": 0.6676361437634917,
|
| 336 |
+
"learning_rate": 3.9948751645524235e-05,
|
| 337 |
+
"loss": 0.5341,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.6144,
|
| 342 |
+
"grad_norm": 0.8563536971964707,
|
| 343 |
+
"learning_rate": 3.9935146162684206e-05,
|
| 344 |
+
"loss": 0.5613,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.6272,
|
| 349 |
+
"grad_norm": 0.8584610175472877,
|
| 350 |
+
"learning_rate": 3.9919943690429906e-05,
|
| 351 |
+
"loss": 0.5804,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.64,
|
| 356 |
+
"grad_norm": 0.7711829326268597,
|
| 357 |
+
"learning_rate": 3.9903145446619837e-05,
|
| 358 |
+
"loss": 0.5346,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6528,
|
| 363 |
+
"grad_norm": 0.7917394963364137,
|
| 364 |
+
"learning_rate": 3.9884752776948564e-05,
|
| 365 |
+
"loss": 0.5348,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.6656,
|
| 370 |
+
"grad_norm": 0.739050002695369,
|
| 371 |
+
"learning_rate": 3.9864767154838864e-05,
|
| 372 |
+
"loss": 0.5273,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6784,
|
| 377 |
+
"grad_norm": 0.7777915961382682,
|
| 378 |
+
"learning_rate": 3.9843190181323744e-05,
|
| 379 |
+
"loss": 0.5411,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.6912,
|
| 384 |
+
"grad_norm": 0.8334856806710838,
|
| 385 |
+
"learning_rate": 3.982002358491817e-05,
|
| 386 |
+
"loss": 0.549,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.704,
|
| 391 |
+
"grad_norm": 0.7359023269093001,
|
| 392 |
+
"learning_rate": 3.979526922148058e-05,
|
| 393 |
+
"loss": 0.54,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.7168,
|
| 398 |
+
"grad_norm": 0.8093918944005913,
|
| 399 |
+
"learning_rate": 3.9768929074064206e-05,
|
| 400 |
+
"loss": 0.5355,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.7296,
|
| 405 |
+
"grad_norm": 0.6928900027986061,
|
| 406 |
+
"learning_rate": 3.9741005252758255e-05,
|
| 407 |
+
"loss": 0.5287,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.7424,
|
| 412 |
+
"grad_norm": 0.8182310889995922,
|
| 413 |
+
"learning_rate": 3.971149999451886e-05,
|
| 414 |
+
"loss": 0.5252,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.7552,
|
| 419 |
+
"grad_norm": 0.7612498920899996,
|
| 420 |
+
"learning_rate": 3.9680415662989806e-05,
|
| 421 |
+
"loss": 0.5548,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.768,
|
| 426 |
+
"grad_norm": 0.6940465259964024,
|
| 427 |
+
"learning_rate": 3.9647754748313294e-05,
|
| 428 |
+
"loss": 0.5156,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.7808,
|
| 433 |
+
"grad_norm": 0.801234736067753,
|
| 434 |
+
"learning_rate": 3.96135198669304e-05,
|
| 435 |
+
"loss": 0.5478,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7936,
|
| 440 |
+
"grad_norm": 0.784729558104407,
|
| 441 |
+
"learning_rate": 3.957771376137144e-05,
|
| 442 |
+
"loss": 0.5576,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.8064,
|
| 447 |
+
"grad_norm": 0.5777461221723659,
|
| 448 |
+
"learning_rate": 3.954033930003634e-05,
|
| 449 |
+
"loss": 0.5236,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.8192,
|
| 454 |
+
"grad_norm": 0.8675379106211997,
|
| 455 |
+
"learning_rate": 3.9501399476964806e-05,
|
| 456 |
+
"loss": 0.5267,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.832,
|
| 461 |
+
"grad_norm": 0.5604568731066945,
|
| 462 |
+
"learning_rate": 3.946089741159648e-05,
|
| 463 |
+
"loss": 0.5155,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.8448,
|
| 468 |
+
"grad_norm": 0.8432224301089873,
|
| 469 |
+
"learning_rate": 3.9418836348521045e-05,
|
| 470 |
+
"loss": 0.5451,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.8576,
|
| 475 |
+
"grad_norm": 0.6038651170676773,
|
| 476 |
+
"learning_rate": 3.937521965721831e-05,
|
| 477 |
+
"loss": 0.5237,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.8704,
|
| 482 |
+
"grad_norm": 0.6641840953558849,
|
| 483 |
+
"learning_rate": 3.933005083178828e-05,
|
| 484 |
+
"loss": 0.5294,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8832,
|
| 489 |
+
"grad_norm": 0.734143907965988,
|
| 490 |
+
"learning_rate": 3.928333349067125e-05,
|
| 491 |
+
"loss": 0.5486,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.896,
|
| 496 |
+
"grad_norm": 0.5735415008946284,
|
| 497 |
+
"learning_rate": 3.923507137635792e-05,
|
| 498 |
+
"loss": 0.5264,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.9088,
|
| 503 |
+
"grad_norm": 0.6085720729928222,
|
| 504 |
+
"learning_rate": 3.9185268355089606e-05,
|
| 505 |
+
"loss": 0.5454,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.9216,
|
| 510 |
+
"grad_norm": 0.6621583223498013,
|
| 511 |
+
"learning_rate": 3.913392841654851e-05,
|
| 512 |
+
"loss": 0.5429,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.9344,
|
| 517 |
+
"grad_norm": 0.6573766001506345,
|
| 518 |
+
"learning_rate": 3.9081055673538093e-05,
|
| 519 |
+
"loss": 0.5039,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.9472,
|
| 524 |
+
"grad_norm": 0.7271563978013613,
|
| 525 |
+
"learning_rate": 3.902665436165364e-05,
|
| 526 |
+
"loss": 0.5625,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.96,
|
| 531 |
+
"grad_norm": 0.674099804328331,
|
| 532 |
+
"learning_rate": 3.897072883894291e-05,
|
| 533 |
+
"loss": 0.528,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.9728,
|
| 538 |
+
"grad_norm": 0.8193293277891587,
|
| 539 |
+
"learning_rate": 3.8913283585557054e-05,
|
| 540 |
+
"loss": 0.5408,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.9856,
|
| 545 |
+
"grad_norm": 0.5700699436836789,
|
| 546 |
+
"learning_rate": 3.885432320339167e-05,
|
| 547 |
+
"loss": 0.5196,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.9984,
|
| 552 |
+
"grad_norm": 0.8965864045994495,
|
| 553 |
+
"learning_rate": 3.879385241571817e-05,
|
| 554 |
+
"loss": 0.5182,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.0112,
|
| 559 |
+
"grad_norm": 0.7259677942199666,
|
| 560 |
+
"learning_rate": 3.873187606680543e-05,
|
| 561 |
+
"loss": 0.4344,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.024,
|
| 566 |
+
"grad_norm": 0.5617989102718455,
|
| 567 |
+
"learning_rate": 3.866839912153168e-05,
|
| 568 |
+
"loss": 0.4505,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.0368,
|
| 573 |
+
"grad_norm": 0.7408990103547018,
|
| 574 |
+
"learning_rate": 3.860342666498677e-05,
|
| 575 |
+
"loss": 0.4174,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.0496,
|
| 580 |
+
"grad_norm": 0.7922127365264651,
|
| 581 |
+
"learning_rate": 3.853696390206484e-05,
|
| 582 |
+
"loss": 0.4371,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.0624,
|
| 587 |
+
"grad_norm": 0.6688534746939121,
|
| 588 |
+
"learning_rate": 3.846901615704734e-05,
|
| 589 |
+
"loss": 0.4596,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.0752,
|
| 594 |
+
"grad_norm": 0.8041015092496046,
|
| 595 |
+
"learning_rate": 3.839958887317649e-05,
|
| 596 |
+
"loss": 0.4218,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.088,
|
| 601 |
+
"grad_norm": 1.1209555919181473,
|
| 602 |
+
"learning_rate": 3.832868761221926e-05,
|
| 603 |
+
"loss": 0.4398,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.1008,
|
| 608 |
+
"grad_norm": 0.8422613372900545,
|
| 609 |
+
"learning_rate": 3.825631805402182e-05,
|
| 610 |
+
"loss": 0.465,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.1136,
|
| 615 |
+
"grad_norm": 0.8984807579222902,
|
| 616 |
+
"learning_rate": 3.818248599605448e-05,
|
| 617 |
+
"loss": 0.4332,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.1264,
|
| 622 |
+
"grad_norm": 0.793206969966527,
|
| 623 |
+
"learning_rate": 3.810719735294731e-05,
|
| 624 |
+
"loss": 0.4314,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.1392,
|
| 629 |
+
"grad_norm": 0.7144674540191216,
|
| 630 |
+
"learning_rate": 3.8030458156016326e-05,
|
| 631 |
+
"loss": 0.4473,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.152,
|
| 636 |
+
"grad_norm": 0.6942414134654957,
|
| 637 |
+
"learning_rate": 3.795227455278029e-05,
|
| 638 |
+
"loss": 0.442,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.1648,
|
| 643 |
+
"grad_norm": 0.6102690387788114,
|
| 644 |
+
"learning_rate": 3.787265280646825e-05,
|
| 645 |
+
"loss": 0.4468,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.1776,
|
| 650 |
+
"grad_norm": 0.5474438543454925,
|
| 651 |
+
"learning_rate": 3.7791599295517825e-05,
|
| 652 |
+
"loss": 0.4112,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.1904,
|
| 657 |
+
"grad_norm": 0.6710179331192792,
|
| 658 |
+
"learning_rate": 3.7709120513064196e-05,
|
| 659 |
+
"loss": 0.4248,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.2032,
|
| 664 |
+
"grad_norm": 0.5394903877190298,
|
| 665 |
+
"learning_rate": 3.762522306641998e-05,
|
| 666 |
+
"loss": 0.4108,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.216,
|
| 671 |
+
"grad_norm": 0.5478834461703908,
|
| 672 |
+
"learning_rate": 3.7539913676545874e-05,
|
| 673 |
+
"loss": 0.4048,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.2288000000000001,
|
| 678 |
+
"grad_norm": 0.515358739981039,
|
| 679 |
+
"learning_rate": 3.745319917751229e-05,
|
| 680 |
+
"loss": 0.4086,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.2416,
|
| 685 |
+
"grad_norm": 0.5157574756013145,
|
| 686 |
+
"learning_rate": 3.736508651595188e-05,
|
| 687 |
+
"loss": 0.4065,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.2544,
|
| 692 |
+
"grad_norm": 0.5711115163156624,
|
| 693 |
+
"learning_rate": 3.727558275050301e-05,
|
| 694 |
+
"loss": 0.4209,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.2671999999999999,
|
| 699 |
+
"grad_norm": 0.5206777381940013,
|
| 700 |
+
"learning_rate": 3.718469505124434e-05,
|
| 701 |
+
"loss": 0.4219,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.28,
|
| 706 |
+
"grad_norm": 0.6676554045255457,
|
| 707 |
+
"learning_rate": 3.709243069912041e-05,
|
| 708 |
+
"loss": 0.4196,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.2928,
|
| 713 |
+
"grad_norm": 0.6492543000211944,
|
| 714 |
+
"learning_rate": 3.699879708535838e-05,
|
| 715 |
+
"loss": 0.4394,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.3056,
|
| 720 |
+
"grad_norm": 0.7472697211133067,
|
| 721 |
+
"learning_rate": 3.69038017108759e-05,
|
| 722 |
+
"loss": 0.41,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.3184,
|
| 727 |
+
"grad_norm": 0.7750383959530253,
|
| 728 |
+
"learning_rate": 3.680745218568026e-05,
|
| 729 |
+
"loss": 0.4132,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.3312,
|
| 734 |
+
"grad_norm": 0.7216163320449477,
|
| 735 |
+
"learning_rate": 3.6709756228258735e-05,
|
| 736 |
+
"loss": 0.4186,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.3439999999999999,
|
| 741 |
+
"grad_norm": 0.6785583085049647,
|
| 742 |
+
"learning_rate": 3.6610721664960236e-05,
|
| 743 |
+
"loss": 0.4237,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.3568,
|
| 748 |
+
"grad_norm": 0.8125785776285916,
|
| 749 |
+
"learning_rate": 3.65103564293684e-05,
|
| 750 |
+
"loss": 0.411,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.3696,
|
| 755 |
+
"grad_norm": 0.6399981959781915,
|
| 756 |
+
"learning_rate": 3.640866856166601e-05,
|
| 757 |
+
"loss": 0.3894,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.3824,
|
| 762 |
+
"grad_norm": 0.6183279788017838,
|
| 763 |
+
"learning_rate": 3.6305666207990886e-05,
|
| 764 |
+
"loss": 0.4489,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.3952,
|
| 769 |
+
"grad_norm": 0.6022650902452937,
|
| 770 |
+
"learning_rate": 3.6201357619783336e-05,
|
| 771 |
+
"loss": 0.438,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.408,
|
| 776 |
+
"grad_norm": 0.5619529171553805,
|
| 777 |
+
"learning_rate": 3.609575115312511e-05,
|
| 778 |
+
"loss": 0.4053,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.4208,
|
| 783 |
+
"grad_norm": 0.5749506015054412,
|
| 784 |
+
"learning_rate": 3.598885526807003e-05,
|
| 785 |
+
"loss": 0.4022,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.4336,
|
| 790 |
+
"grad_norm": 0.5169290481579895,
|
| 791 |
+
"learning_rate": 3.5880678527966224e-05,
|
| 792 |
+
"loss": 0.415,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.4464000000000001,
|
| 797 |
+
"grad_norm": 0.5799994069099426,
|
| 798 |
+
"learning_rate": 3.577122959877017e-05,
|
| 799 |
+
"loss": 0.3921,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.4592,
|
| 804 |
+
"grad_norm": 0.5735944670713298,
|
| 805 |
+
"learning_rate": 3.566051724835245e-05,
|
| 806 |
+
"loss": 0.4204,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.472,
|
| 811 |
+
"grad_norm": 0.6062368015564766,
|
| 812 |
+
"learning_rate": 3.554855034579532e-05,
|
| 813 |
+
"loss": 0.4134,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.4848,
|
| 818 |
+
"grad_norm": 0.5789142958185306,
|
| 819 |
+
"learning_rate": 3.5435337860682304e-05,
|
| 820 |
+
"loss": 0.4355,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.4976,
|
| 825 |
+
"grad_norm": 0.7856515301224941,
|
| 826 |
+
"learning_rate": 3.532088886237956e-05,
|
| 827 |
+
"loss": 0.4097,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.5104,
|
| 832 |
+
"grad_norm": 0.5341522241422082,
|
| 833 |
+
"learning_rate": 3.520521251930941e-05,
|
| 834 |
+
"loss": 0.4263,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.5232,
|
| 839 |
+
"grad_norm": 0.776377968902034,
|
| 840 |
+
"learning_rate": 3.5088318098215805e-05,
|
| 841 |
+
"loss": 0.4158,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.536,
|
| 846 |
+
"grad_norm": 0.6064221541044117,
|
| 847 |
+
"learning_rate": 3.497021496342203e-05,
|
| 848 |
+
"loss": 0.435,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.5488,
|
| 853 |
+
"grad_norm": 0.6092258566217099,
|
| 854 |
+
"learning_rate": 3.485091257608047e-05,
|
| 855 |
+
"loss": 0.3884,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.5615999999999999,
|
| 860 |
+
"grad_norm": 0.6331147798170179,
|
| 861 |
+
"learning_rate": 3.473042049341474e-05,
|
| 862 |
+
"loss": 0.4089,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.5744,
|
| 867 |
+
"grad_norm": 0.6135277354958222,
|
| 868 |
+
"learning_rate": 3.4608748367954064e-05,
|
| 869 |
+
"loss": 0.4335,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.5872000000000002,
|
| 874 |
+
"grad_norm": 0.6066501434888445,
|
| 875 |
+
"learning_rate": 3.4485905946759965e-05,
|
| 876 |
+
"loss": 0.4181,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.6,
|
| 881 |
+
"grad_norm": 0.6070219835049349,
|
| 882 |
+
"learning_rate": 3.4361903070645484e-05,
|
| 883 |
+
"loss": 0.4181,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.6128,
|
| 888 |
+
"grad_norm": 0.6149227608995402,
|
| 889 |
+
"learning_rate": 3.423674967338681e-05,
|
| 890 |
+
"loss": 0.4076,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.6256,
|
| 895 |
+
"grad_norm": 0.5786879888544758,
|
| 896 |
+
"learning_rate": 3.411045578092754e-05,
|
| 897 |
+
"loss": 0.4032,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.6383999999999999,
|
| 902 |
+
"grad_norm": 0.6784538128892833,
|
| 903 |
+
"learning_rate": 3.398303151057543e-05,
|
| 904 |
+
"loss": 0.4443,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.6512,
|
| 909 |
+
"grad_norm": 0.551938572366439,
|
| 910 |
+
"learning_rate": 3.385448707019199e-05,
|
| 911 |
+
"loss": 0.4182,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.6640000000000001,
|
| 916 |
+
"grad_norm": 0.5862418832912583,
|
| 917 |
+
"learning_rate": 3.372483275737468e-05,
|
| 918 |
+
"loss": 0.4497,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.6768,
|
| 923 |
+
"grad_norm": 0.5767556346413378,
|
| 924 |
+
"learning_rate": 3.359407895863199e-05,
|
| 925 |
+
"loss": 0.417,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.6896,
|
| 930 |
+
"grad_norm": 0.5016799355677755,
|
| 931 |
+
"learning_rate": 3.34622361485514e-05,
|
| 932 |
+
"loss": 0.4331,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.7024,
|
| 937 |
+
"grad_norm": 0.48341334713379586,
|
| 938 |
+
"learning_rate": 3.332931488896029e-05,
|
| 939 |
+
"loss": 0.3904,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.7151999999999998,
|
| 944 |
+
"grad_norm": 0.5116327887424218,
|
| 945 |
+
"learning_rate": 3.319532582807977e-05,
|
| 946 |
+
"loss": 0.3866,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.728,
|
| 951 |
+
"grad_norm": 0.5402684337267992,
|
| 952 |
+
"learning_rate": 3.30602796996717e-05,
|
| 953 |
+
"loss": 0.4218,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.7408000000000001,
|
| 958 |
+
"grad_norm": 0.532197016605654,
|
| 959 |
+
"learning_rate": 3.2924187322178865e-05,
|
| 960 |
+
"loss": 0.4022,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.7536,
|
| 965 |
+
"grad_norm": 0.4837115469846966,
|
| 966 |
+
"learning_rate": 3.278705959785821e-05,
|
| 967 |
+
"loss": 0.4315,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.7664,
|
| 972 |
+
"grad_norm": 0.5119148643249282,
|
| 973 |
+
"learning_rate": 3.2648907511907544e-05,
|
| 974 |
+
"loss": 0.4285,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.7792,
|
| 979 |
+
"grad_norm": 0.5411357051537327,
|
| 980 |
+
"learning_rate": 3.250974213158555e-05,
|
| 981 |
+
"loss": 0.4089,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.792,
|
| 986 |
+
"grad_norm": 0.5213559100963829,
|
| 987 |
+
"learning_rate": 3.23695746053251e-05,
|
| 988 |
+
"loss": 0.4007,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.8048,
|
| 993 |
+
"grad_norm": 0.5155346041131477,
|
| 994 |
+
"learning_rate": 3.222841616184025e-05,
|
| 995 |
+
"loss": 0.4152,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.8176,
|
| 1000 |
+
"grad_norm": 0.6997329954050828,
|
| 1001 |
+
"learning_rate": 3.208627810922665e-05,
|
| 1002 |
+
"loss": 0.4661,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.8304,
|
| 1007 |
+
"grad_norm": 0.5308502643781506,
|
| 1008 |
+
"learning_rate": 3.194317183405573e-05,
|
| 1009 |
+
"loss": 0.4244,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.8432,
|
| 1014 |
+
"grad_norm": 0.6642345888916624,
|
| 1015 |
+
"learning_rate": 3.1799108800462466e-05,
|
| 1016 |
+
"loss": 0.4549,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.8559999999999999,
|
| 1021 |
+
"grad_norm": 0.4732956139724008,
|
| 1022 |
+
"learning_rate": 3.1654100549227024e-05,
|
| 1023 |
+
"loss": 0.4234,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.8688,
|
| 1028 |
+
"grad_norm": 0.5083639542501994,
|
| 1029 |
+
"learning_rate": 3.1508158696850275e-05,
|
| 1030 |
+
"loss": 0.4166,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.8816000000000002,
|
| 1035 |
+
"grad_norm": 0.5289728887488433,
|
| 1036 |
+
"learning_rate": 3.136129493462312e-05,
|
| 1037 |
+
"loss": 0.4464,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.8944,
|
| 1042 |
+
"grad_norm": 0.49961557479356383,
|
| 1043 |
+
"learning_rate": 3.121352102768998e-05,
|
| 1044 |
+
"loss": 0.4208,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.9072,
|
| 1049 |
+
"grad_norm": 0.5271156302946176,
|
| 1050 |
+
"learning_rate": 3.106484881410628e-05,
|
| 1051 |
+
"loss": 0.4494,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.92,
|
| 1056 |
+
"grad_norm": 0.5058965958517215,
|
| 1057 |
+
"learning_rate": 3.091529020389009e-05,
|
| 1058 |
+
"loss": 0.3939,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.9327999999999999,
|
| 1063 |
+
"grad_norm": 0.521319317182331,
|
| 1064 |
+
"learning_rate": 3.076485717806808e-05,
|
| 1065 |
+
"loss": 0.4327,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.9456,
|
| 1070 |
+
"grad_norm": 0.5242493173094706,
|
| 1071 |
+
"learning_rate": 3.061356178771564e-05,
|
| 1072 |
+
"loss": 0.4037,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.9584000000000001,
|
| 1077 |
+
"grad_norm": 0.5416936483204631,
|
| 1078 |
+
"learning_rate": 3.0461416152991555e-05,
|
| 1079 |
+
"loss": 0.4323,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.9712,
|
| 1084 |
+
"grad_norm": 0.5187949446347746,
|
| 1085 |
+
"learning_rate": 3.0308432462167045e-05,
|
| 1086 |
+
"loss": 0.4386,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.984,
|
| 1091 |
+
"grad_norm": 0.49016993717471097,
|
| 1092 |
+
"learning_rate": 3.015462297064936e-05,
|
| 1093 |
+
"loss": 0.4056,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.9968,
|
| 1098 |
+
"grad_norm": 0.5050271626448363,
|
| 1099 |
+
"learning_rate": 3.0000000000000004e-05,
|
| 1100 |
+
"loss": 0.4443,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.0096,
|
| 1105 |
+
"grad_norm": 0.5064690410359016,
|
| 1106 |
+
"learning_rate": 2.98445759369477e-05,
|
| 1107 |
+
"loss": 0.3304,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.0224,
|
| 1112 |
+
"grad_norm": 0.5069317306390775,
|
| 1113 |
+
"learning_rate": 2.9688363232396056e-05,
|
| 1114 |
+
"loss": 0.3227,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.0352,
|
| 1119 |
+
"grad_norm": 0.6126269132868637,
|
| 1120 |
+
"learning_rate": 2.9531374400426158e-05,
|
| 1121 |
+
"loss": 0.3164,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.048,
|
| 1126 |
+
"grad_norm": 0.5939372011513905,
|
| 1127 |
+
"learning_rate": 2.9373622017294075e-05,
|
| 1128 |
+
"loss": 0.3216,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.0608,
|
| 1133 |
+
"grad_norm": 0.5458305751982532,
|
| 1134 |
+
"learning_rate": 2.9215118720423375e-05,
|
| 1135 |
+
"loss": 0.3257,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.0736,
|
| 1140 |
+
"grad_norm": 0.563608692727318,
|
| 1141 |
+
"learning_rate": 2.9055877207392752e-05,
|
| 1142 |
+
"loss": 0.3115,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.0864,
|
| 1147 |
+
"grad_norm": 0.6117949539451228,
|
| 1148 |
+
"learning_rate": 2.8895910234918828e-05,
|
| 1149 |
+
"loss": 0.3327,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.0992,
|
| 1154 |
+
"grad_norm": 0.5801355464656908,
|
| 1155 |
+
"learning_rate": 2.873523061783426e-05,
|
| 1156 |
+
"loss": 0.3573,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.112,
|
| 1161 |
+
"grad_norm": 0.673612125082193,
|
| 1162 |
+
"learning_rate": 2.8573851228061084e-05,
|
| 1163 |
+
"loss": 0.3151,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.1248,
|
| 1168 |
+
"grad_norm": 0.5056513633085493,
|
| 1169 |
+
"learning_rate": 2.8411784993579633e-05,
|
| 1170 |
+
"loss": 0.3041,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.1376,
|
| 1175 |
+
"grad_norm": 0.562126042299196,
|
| 1176 |
+
"learning_rate": 2.8249044897392814e-05,
|
| 1177 |
+
"loss": 0.328,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.1504,
|
| 1182 |
+
"grad_norm": 0.5426169830161941,
|
| 1183 |
+
"learning_rate": 2.80856439764861e-05,
|
| 1184 |
+
"loss": 0.3104,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.1632,
|
| 1189 |
+
"grad_norm": 0.5036234765639455,
|
| 1190 |
+
"learning_rate": 2.792159532078314e-05,
|
| 1191 |
+
"loss": 0.2858,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.176,
|
| 1196 |
+
"grad_norm": 0.5359038654252825,
|
| 1197 |
+
"learning_rate": 2.77569120720971e-05,
|
| 1198 |
+
"loss": 0.3093,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.1888,
|
| 1203 |
+
"grad_norm": 0.5339507279296722,
|
| 1204 |
+
"learning_rate": 2.7591607423077932e-05,
|
| 1205 |
+
"loss": 0.3299,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.2016,
|
| 1210 |
+
"grad_norm": 0.5877580107006486,
|
| 1211 |
+
"learning_rate": 2.7425694616155474e-05,
|
| 1212 |
+
"loss": 0.3115,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.2144,
|
| 1217 |
+
"grad_norm": 0.6243690115560702,
|
| 1218 |
+
"learning_rate": 2.7259186942478656e-05,
|
| 1219 |
+
"loss": 0.3399,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.2272,
|
| 1224 |
+
"grad_norm": 0.5171025610907465,
|
| 1225 |
+
"learning_rate": 2.7092097740850712e-05,
|
| 1226 |
+
"loss": 0.3247,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.24,
|
| 1231 |
+
"grad_norm": 0.5675806120504167,
|
| 1232 |
+
"learning_rate": 2.692444039666066e-05,
|
| 1233 |
+
"loss": 0.3295,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.2528,
|
| 1238 |
+
"grad_norm": 0.5274736445648369,
|
| 1239 |
+
"learning_rate": 2.6756228340810946e-05,
|
| 1240 |
+
"loss": 0.2975,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.2656,
|
| 1245 |
+
"grad_norm": 0.4714012998161742,
|
| 1246 |
+
"learning_rate": 2.6587475048641596e-05,
|
| 1247 |
+
"loss": 0.3012,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.2784,
|
| 1252 |
+
"grad_norm": 0.4711115049913107,
|
| 1253 |
+
"learning_rate": 2.6418194038850634e-05,
|
| 1254 |
+
"loss": 0.3281,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.2912,
|
| 1259 |
+
"grad_norm": 0.5196283761716489,
|
| 1260 |
+
"learning_rate": 2.624839887241115e-05,
|
| 1261 |
+
"loss": 0.3391,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.304,
|
| 1266 |
+
"grad_norm": 0.46814357438824156,
|
| 1267 |
+
"learning_rate": 2.607810315148494e-05,
|
| 1268 |
+
"loss": 0.3204,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.3168,
|
| 1273 |
+
"grad_norm": 0.5299865417151409,
|
| 1274 |
+
"learning_rate": 2.5907320518332827e-05,
|
| 1275 |
+
"loss": 0.3245,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.3296,
|
| 1280 |
+
"grad_norm": 0.49288322858537664,
|
| 1281 |
+
"learning_rate": 2.5736064654221808e-05,
|
| 1282 |
+
"loss": 0.2716,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.3424,
|
| 1287 |
+
"grad_norm": 0.4200100521701435,
|
| 1288 |
+
"learning_rate": 2.5564349278329056e-05,
|
| 1289 |
+
"loss": 0.3143,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.3552,
|
| 1294 |
+
"grad_norm": 0.5189496713485224,
|
| 1295 |
+
"learning_rate": 2.539218814664288e-05,
|
| 1296 |
+
"loss": 0.2789,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.368,
|
| 1301 |
+
"grad_norm": 0.498057989067044,
|
| 1302 |
+
"learning_rate": 2.521959505086075e-05,
|
| 1303 |
+
"loss": 0.3333,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.3808,
|
| 1308 |
+
"grad_norm": 0.4557883186933527,
|
| 1309 |
+
"learning_rate": 2.5046583817284437e-05,
|
| 1310 |
+
"loss": 0.2891,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.3936,
|
| 1315 |
+
"grad_norm": 0.49762496414917196,
|
| 1316 |
+
"learning_rate": 2.487316830571244e-05,
|
| 1317 |
+
"loss": 0.3164,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.4064,
|
| 1322 |
+
"grad_norm": 0.43591658685187756,
|
| 1323 |
+
"learning_rate": 2.4699362408329646e-05,
|
| 1324 |
+
"loss": 0.3253,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.4192,
|
| 1329 |
+
"grad_norm": 0.42371649722843396,
|
| 1330 |
+
"learning_rate": 2.4525180048594452e-05,
|
| 1331 |
+
"loss": 0.3354,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.432,
|
| 1336 |
+
"grad_norm": 0.4568181264814724,
|
| 1337 |
+
"learning_rate": 2.435063518012335e-05,
|
| 1338 |
+
"loss": 0.3024,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.4448,
|
| 1343 |
+
"grad_norm": 0.44045937237350474,
|
| 1344 |
+
"learning_rate": 2.4175741785573177e-05,
|
| 1345 |
+
"loss": 0.3077,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.4576000000000002,
|
| 1350 |
+
"grad_norm": 0.5019590154900115,
|
| 1351 |
+
"learning_rate": 2.4000513875520892e-05,
|
| 1352 |
+
"loss": 0.3451,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.4704,
|
| 1357 |
+
"grad_norm": 0.4676089477205968,
|
| 1358 |
+
"learning_rate": 2.3824965487341247e-05,
|
| 1359 |
+
"loss": 0.3203,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.4832,
|
| 1364 |
+
"grad_norm": 0.48372854045562635,
|
| 1365 |
+
"learning_rate": 2.3649110684082258e-05,
|
| 1366 |
+
"loss": 0.3093,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.496,
|
| 1371 |
+
"grad_norm": 0.4636540082215299,
|
| 1372 |
+
"learning_rate": 2.3472963553338614e-05,
|
| 1373 |
+
"loss": 0.3099,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.5088,
|
| 1378 |
+
"grad_norm": 0.4680925787550491,
|
| 1379 |
+
"learning_rate": 2.3296538206123134e-05,
|
| 1380 |
+
"loss": 0.316,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.5216,
|
| 1385 |
+
"grad_norm": 0.4074596062272102,
|
| 1386 |
+
"learning_rate": 2.311984877573636e-05,
|
| 1387 |
+
"loss": 0.285,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.5343999999999998,
|
| 1392 |
+
"grad_norm": 0.569069284545878,
|
| 1393 |
+
"learning_rate": 2.2942909416634326e-05,
|
| 1394 |
+
"loss": 0.3146,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.5472,
|
| 1399 |
+
"grad_norm": 0.4025723787987535,
|
| 1400 |
+
"learning_rate": 2.2765734303294666e-05,
|
| 1401 |
+
"loss": 0.3267,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.56,
|
| 1406 |
+
"grad_norm": 0.524781949080461,
|
| 1407 |
+
"learning_rate": 2.2588337629081107e-05,
|
| 1408 |
+
"loss": 0.3088,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.5728,
|
| 1413 |
+
"grad_norm": 0.4111354409664677,
|
| 1414 |
+
"learning_rate": 2.2410733605106462e-05,
|
| 1415 |
+
"loss": 0.3376,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.5856,
|
| 1420 |
+
"grad_norm": 0.4729581743067788,
|
| 1421 |
+
"learning_rate": 2.2232936459094158e-05,
|
| 1422 |
+
"loss": 0.2933,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.5984,
|
| 1427 |
+
"grad_norm": 0.4617925423325259,
|
| 1428 |
+
"learning_rate": 2.205496043423849e-05,
|
| 1429 |
+
"loss": 0.3325,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.6112,
|
| 1434 |
+
"grad_norm": 0.39451304381006896,
|
| 1435 |
+
"learning_rate": 2.1876819788063586e-05,
|
| 1436 |
+
"loss": 0.3081,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.624,
|
| 1441 |
+
"grad_norm": 0.42970773208284047,
|
| 1442 |
+
"learning_rate": 2.16985287912813e-05,
|
| 1443 |
+
"loss": 0.3046,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.6368,
|
| 1448 |
+
"grad_norm": 0.4355064840165267,
|
| 1449 |
+
"learning_rate": 2.1520101726647922e-05,
|
| 1450 |
+
"loss": 0.3257,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 2.6496,
|
| 1455 |
+
"grad_norm": 0.42568187175267763,
|
| 1456 |
+
"learning_rate": 2.1341552887820048e-05,
|
| 1457 |
+
"loss": 0.3064,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.6624,
|
| 1462 |
+
"grad_norm": 0.4151687815115344,
|
| 1463 |
+
"learning_rate": 2.1162896578209517e-05,
|
| 1464 |
+
"loss": 0.3181,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 2.6752000000000002,
|
| 1469 |
+
"grad_norm": 0.4391363195391212,
|
| 1470 |
+
"learning_rate": 2.0984147109837564e-05,
|
| 1471 |
+
"loss": 0.3229,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.6879999999999997,
|
| 1476 |
+
"grad_norm": 0.4096181340419125,
|
| 1477 |
+
"learning_rate": 2.0805318802188307e-05,
|
| 1478 |
+
"loss": 0.2911,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 2.7008,
|
| 1483 |
+
"grad_norm": 0.4259956871485725,
|
| 1484 |
+
"learning_rate": 2.0626425981061608e-05,
|
| 1485 |
+
"loss": 0.306,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.7136,
|
| 1490 |
+
"grad_norm": 0.449411215139247,
|
| 1491 |
+
"learning_rate": 2.0447482977425465e-05,
|
| 1492 |
+
"loss": 0.3097,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 2.7264,
|
| 1497 |
+
"grad_norm": 0.42360453698815276,
|
| 1498 |
+
"learning_rate": 2.0268504126267952e-05,
|
| 1499 |
+
"loss": 0.3016,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.7392,
|
| 1504 |
+
"grad_norm": 0.442503174431085,
|
| 1505 |
+
"learning_rate": 2.008950376544887e-05,
|
| 1506 |
+
"loss": 0.3503,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 2.752,
|
| 1511 |
+
"grad_norm": 0.49740172938571353,
|
| 1512 |
+
"learning_rate": 1.9910496234551132e-05,
|
| 1513 |
+
"loss": 0.3472,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 2.7648,
|
| 1518 |
+
"grad_norm": 0.4095347985713004,
|
| 1519 |
+
"learning_rate": 1.9731495873732055e-05,
|
| 1520 |
+
"loss": 0.2887,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 2.7776,
|
| 1525 |
+
"grad_norm": 0.437748245736374,
|
| 1526 |
+
"learning_rate": 1.9552517022574542e-05,
|
| 1527 |
+
"loss": 0.3284,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.7904,
|
| 1532 |
+
"grad_norm": 0.451773557564148,
|
| 1533 |
+
"learning_rate": 1.93735740189384e-05,
|
| 1534 |
+
"loss": 0.2817,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 2.8032,
|
| 1539 |
+
"grad_norm": 0.44626247584864404,
|
| 1540 |
+
"learning_rate": 1.9194681197811703e-05,
|
| 1541 |
+
"loss": 0.3248,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.816,
|
| 1546 |
+
"grad_norm": 0.3795144617967381,
|
| 1547 |
+
"learning_rate": 1.901585289016244e-05,
|
| 1548 |
+
"loss": 0.3196,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 2.8288,
|
| 1553 |
+
"grad_norm": 0.44134742506699864,
|
| 1554 |
+
"learning_rate": 1.8837103421790486e-05,
|
| 1555 |
+
"loss": 0.317,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 2.8416,
|
| 1560 |
+
"grad_norm": 0.40250009217964594,
|
| 1561 |
+
"learning_rate": 1.8658447112179952e-05,
|
| 1562 |
+
"loss": 0.3236,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 2.8544,
|
| 1567 |
+
"grad_norm": 0.39575373068255826,
|
| 1568 |
+
"learning_rate": 1.8479898273352084e-05,
|
| 1569 |
+
"loss": 0.3058,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.8672,
|
| 1574 |
+
"grad_norm": 0.4673278705380792,
|
| 1575 |
+
"learning_rate": 1.83014712087187e-05,
|
| 1576 |
+
"loss": 0.3096,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 2.88,
|
| 1581 |
+
"grad_norm": 0.38581239124535677,
|
| 1582 |
+
"learning_rate": 1.8123180211936417e-05,
|
| 1583 |
+
"loss": 0.3086,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 2.8928000000000003,
|
| 1588 |
+
"grad_norm": 0.4616630201048585,
|
| 1589 |
+
"learning_rate": 1.794503956576152e-05,
|
| 1590 |
+
"loss": 0.3359,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 2.9055999999999997,
|
| 1595 |
+
"grad_norm": 0.4080695642416766,
|
| 1596 |
+
"learning_rate": 1.776706354090585e-05,
|
| 1597 |
+
"loss": 0.2987,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 2.9184,
|
| 1602 |
+
"grad_norm": 0.41582733102834524,
|
| 1603 |
+
"learning_rate": 1.758926639489354e-05,
|
| 1604 |
+
"loss": 0.2632,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 2.9312,
|
| 1609 |
+
"grad_norm": 0.44094210958109037,
|
| 1610 |
+
"learning_rate": 1.7411662370918893e-05,
|
| 1611 |
+
"loss": 0.3186,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.944,
|
| 1616 |
+
"grad_norm": 0.38341759079987514,
|
| 1617 |
+
"learning_rate": 1.7234265696705344e-05,
|
| 1618 |
+
"loss": 0.2867,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.9568,
|
| 1623 |
+
"grad_norm": 0.5166138259487162,
|
| 1624 |
+
"learning_rate": 1.7057090583365678e-05,
|
| 1625 |
+
"loss": 0.3205,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.9696,
|
| 1630 |
+
"grad_norm": 0.40798833673821217,
|
| 1631 |
+
"learning_rate": 1.6880151224263646e-05,
|
| 1632 |
+
"loss": 0.3321,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.9824,
|
| 1637 |
+
"grad_norm": 0.44371243751880507,
|
| 1638 |
+
"learning_rate": 1.6703461793876876e-05,
|
| 1639 |
+
"loss": 0.3062,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.9952,
|
| 1644 |
+
"grad_norm": 0.4760656739021338,
|
| 1645 |
+
"learning_rate": 1.6527036446661396e-05,
|
| 1646 |
+
"loss": 0.3247,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 3.008,
|
| 1651 |
+
"grad_norm": 0.37692665832853156,
|
| 1652 |
+
"learning_rate": 1.635088931591775e-05,
|
| 1653 |
+
"loss": 0.3053,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 3.0208,
|
| 1658 |
+
"grad_norm": 0.44891524452502357,
|
| 1659 |
+
"learning_rate": 1.6175034512658753e-05,
|
| 1660 |
+
"loss": 0.2354,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 3.0336,
|
| 1665 |
+
"grad_norm": 0.40636197979518823,
|
| 1666 |
+
"learning_rate": 1.5999486124479115e-05,
|
| 1667 |
+
"loss": 0.2556,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 3.0464,
|
| 1672 |
+
"grad_norm": 0.5092020726269557,
|
| 1673 |
+
"learning_rate": 1.5824258214426833e-05,
|
| 1674 |
+
"loss": 0.2339,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 3.0592,
|
| 1679 |
+
"grad_norm": 0.38167315277048636,
|
| 1680 |
+
"learning_rate": 1.5649364819876655e-05,
|
| 1681 |
+
"loss": 0.2263,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 3.072,
|
| 1686 |
+
"grad_norm": 0.4400989736622662,
|
| 1687 |
+
"learning_rate": 1.547481995140556e-05,
|
| 1688 |
+
"loss": 0.223,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 3.0848,
|
| 1693 |
+
"grad_norm": 0.4734524336864123,
|
| 1694 |
+
"learning_rate": 1.5300637591670357e-05,
|
| 1695 |
+
"loss": 0.2227,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 3.0976,
|
| 1700 |
+
"grad_norm": 0.37181553094968756,
|
| 1701 |
+
"learning_rate": 1.5126831694287564e-05,
|
| 1702 |
+
"loss": 0.2029,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 3.1104,
|
| 1707 |
+
"grad_norm": 0.44528707231170533,
|
| 1708 |
+
"learning_rate": 1.4953416182715566e-05,
|
| 1709 |
+
"loss": 0.2418,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 3.1232,
|
| 1714 |
+
"grad_norm": 0.41680948338939605,
|
| 1715 |
+
"learning_rate": 1.478040494913926e-05,
|
| 1716 |
+
"loss": 0.2471,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 3.136,
|
| 1721 |
+
"grad_norm": 0.36293617588468985,
|
| 1722 |
+
"learning_rate": 1.460781185335713e-05,
|
| 1723 |
+
"loss": 0.2724,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 3.1488,
|
| 1728 |
+
"grad_norm": 0.40426767585101725,
|
| 1729 |
+
"learning_rate": 1.443565072167095e-05,
|
| 1730 |
+
"loss": 0.2678,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 3.1616,
|
| 1735 |
+
"grad_norm": 0.3819598492232231,
|
| 1736 |
+
"learning_rate": 1.4263935345778202e-05,
|
| 1737 |
+
"loss": 0.2139,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 3.1744,
|
| 1742 |
+
"grad_norm": 0.39989671859500364,
|
| 1743 |
+
"learning_rate": 1.409267948166718e-05,
|
| 1744 |
+
"loss": 0.2423,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 3.1872,
|
| 1749 |
+
"grad_norm": 0.34641336727696975,
|
| 1750 |
+
"learning_rate": 1.3921896848515064e-05,
|
| 1751 |
+
"loss": 0.2397,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 3.2,
|
| 1756 |
+
"grad_norm": 0.3893290034117696,
|
| 1757 |
+
"learning_rate": 1.3751601127588849e-05,
|
| 1758 |
+
"loss": 0.2664,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 3.2128,
|
| 1763 |
+
"grad_norm": 0.40917604434187654,
|
| 1764 |
+
"learning_rate": 1.3581805961149371e-05,
|
| 1765 |
+
"loss": 0.2403,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 3.2256,
|
| 1770 |
+
"grad_norm": 0.34424063354262885,
|
| 1771 |
+
"learning_rate": 1.341252495135841e-05,
|
| 1772 |
+
"loss": 0.217,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 3.2384,
|
| 1777 |
+
"grad_norm": 0.36331987383874076,
|
| 1778 |
+
"learning_rate": 1.324377165918906e-05,
|
| 1779 |
+
"loss": 0.2367,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 3.2512,
|
| 1784 |
+
"grad_norm": 0.3566422219423312,
|
| 1785 |
+
"learning_rate": 1.3075559603339354e-05,
|
| 1786 |
+
"loss": 0.2381,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 3.2640000000000002,
|
| 1791 |
+
"grad_norm": 0.31017581514496495,
|
| 1792 |
+
"learning_rate": 1.2907902259149287e-05,
|
| 1793 |
+
"loss": 0.2157,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 3.2768,
|
| 1798 |
+
"grad_norm": 0.367186904196843,
|
| 1799 |
+
"learning_rate": 1.274081305752135e-05,
|
| 1800 |
+
"loss": 0.257,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 3.2896,
|
| 1805 |
+
"grad_norm": 0.3337440900953406,
|
| 1806 |
+
"learning_rate": 1.2574305383844528e-05,
|
| 1807 |
+
"loss": 0.2055,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 3.3024,
|
| 1812 |
+
"grad_norm": 0.3336237356155655,
|
| 1813 |
+
"learning_rate": 1.2408392576922075e-05,
|
| 1814 |
+
"loss": 0.2511,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 3.3152,
|
| 1819 |
+
"grad_norm": 0.300651420561767,
|
| 1820 |
+
"learning_rate": 1.2243087927902905e-05,
|
| 1821 |
+
"loss": 0.2038,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 3.328,
|
| 1826 |
+
"grad_norm": 0.31752616451541105,
|
| 1827 |
+
"learning_rate": 1.2078404679216864e-05,
|
| 1828 |
+
"loss": 0.2405,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 3.3407999999999998,
|
| 1833 |
+
"grad_norm": 0.3218267761965695,
|
| 1834 |
+
"learning_rate": 1.1914356023513904e-05,
|
| 1835 |
+
"loss": 0.1992,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 3.3536,
|
| 1840 |
+
"grad_norm": 0.2918458747111414,
|
| 1841 |
+
"learning_rate": 1.1750955102607193e-05,
|
| 1842 |
+
"loss": 0.2507,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 3.3664,
|
| 1847 |
+
"grad_norm": 0.3093907692361296,
|
| 1848 |
+
"learning_rate": 1.1588215006420374e-05,
|
| 1849 |
+
"loss": 0.2374,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 3.3792,
|
| 1854 |
+
"grad_norm": 0.3045693552016105,
|
| 1855 |
+
"learning_rate": 1.1426148771938915e-05,
|
| 1856 |
+
"loss": 0.1928,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 3.392,
|
| 1861 |
+
"grad_norm": 0.36473327946790146,
|
| 1862 |
+
"learning_rate": 1.1264769382165748e-05,
|
| 1863 |
+
"loss": 0.2501,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 3.4048,
|
| 1868 |
+
"grad_norm": 0.32912667678835017,
|
| 1869 |
+
"learning_rate": 1.110408976508118e-05,
|
| 1870 |
+
"loss": 0.2301,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 3.4176,
|
| 1875 |
+
"grad_norm": 0.30186839540587557,
|
| 1876 |
+
"learning_rate": 1.094412279260726e-05,
|
| 1877 |
+
"loss": 0.2612,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 3.4304,
|
| 1882 |
+
"grad_norm": 0.3165181319078008,
|
| 1883 |
+
"learning_rate": 1.0784881279576635e-05,
|
| 1884 |
+
"loss": 0.2615,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 3.4432,
|
| 1889 |
+
"grad_norm": 0.3189059514287575,
|
| 1890 |
+
"learning_rate": 1.0626377982705929e-05,
|
| 1891 |
+
"loss": 0.2722,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 3.456,
|
| 1896 |
+
"grad_norm": 0.32188428613390385,
|
| 1897 |
+
"learning_rate": 1.0468625599573842e-05,
|
| 1898 |
+
"loss": 0.24,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 3.4688,
|
| 1903 |
+
"grad_norm": 0.3060768488900926,
|
| 1904 |
+
"learning_rate": 1.0311636767603952e-05,
|
| 1905 |
+
"loss": 0.2105,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 3.4816,
|
| 1910 |
+
"grad_norm": 0.33207198798670173,
|
| 1911 |
+
"learning_rate": 1.0155424063052306e-05,
|
| 1912 |
+
"loss": 0.232,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 3.4944,
|
| 1917 |
+
"grad_norm": 0.3128637275072241,
|
| 1918 |
+
"learning_rate": 1.0000000000000006e-05,
|
| 1919 |
+
"loss": 0.2707,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 3.5072,
|
| 1924 |
+
"grad_norm": 0.299423652903677,
|
| 1925 |
+
"learning_rate": 9.84537702935065e-06,
|
| 1926 |
+
"loss": 0.2483,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 3.52,
|
| 1931 |
+
"grad_norm": 0.3194043723836402,
|
| 1932 |
+
"learning_rate": 9.691567537832964e-06,
|
| 1933 |
+
"loss": 0.2459,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 3.5328,
|
| 1938 |
+
"grad_norm": 0.29134874604257066,
|
| 1939 |
+
"learning_rate": 9.538583847008452e-06,
|
| 1940 |
+
"loss": 0.2383,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 3.5456,
|
| 1945 |
+
"grad_norm": 0.3174469505841507,
|
| 1946 |
+
"learning_rate": 9.386438212284372e-06,
|
| 1947 |
+
"loss": 0.2171,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 3.5584,
|
| 1952 |
+
"grad_norm": 0.33469109498385036,
|
| 1953 |
+
"learning_rate": 9.235142821931928e-06,
|
| 1954 |
+
"loss": 0.2649,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 3.5712,
|
| 1959 |
+
"grad_norm": 0.3388488536987889,
|
| 1960 |
+
"learning_rate": 9.084709796109907e-06,
|
| 1961 |
+
"loss": 0.2554,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 3.584,
|
| 1966 |
+
"grad_norm": 0.3232370072303198,
|
| 1967 |
+
"learning_rate": 8.93515118589373e-06,
|
| 1968 |
+
"loss": 0.2554,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 3.5968,
|
| 1973 |
+
"grad_norm": 0.3126656112165131,
|
| 1974 |
+
"learning_rate": 8.786478972310023e-06,
|
| 1975 |
+
"loss": 0.2175,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 3.6096,
|
| 1980 |
+
"grad_norm": 0.3216261570369998,
|
| 1981 |
+
"learning_rate": 8.638705065376887e-06,
|
| 1982 |
+
"loss": 0.2652,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 3.6224,
|
| 1987 |
+
"grad_norm": 0.37207664870644563,
|
| 1988 |
+
"learning_rate": 8.491841303149728e-06,
|
| 1989 |
+
"loss": 0.2495,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 3.6352,
|
| 1994 |
+
"grad_norm": 0.3613796739265539,
|
| 1995 |
+
"learning_rate": 8.345899450772975e-06,
|
| 1996 |
+
"loss": 0.2416,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 3.648,
|
| 2001 |
+
"grad_norm": 0.3021810760971353,
|
| 2002 |
+
"learning_rate": 8.200891199537549e-06,
|
| 2003 |
+
"loss": 0.2686,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 3.6608,
|
| 2008 |
+
"grad_norm": 0.32371976254247287,
|
| 2009 |
+
"learning_rate": 8.056828165944282e-06,
|
| 2010 |
+
"loss": 0.2195,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 3.6736,
|
| 2015 |
+
"grad_norm": 0.37351534686010873,
|
| 2016 |
+
"learning_rate": 7.913721890773354e-06,
|
| 2017 |
+
"loss": 0.2026,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 3.6864,
|
| 2022 |
+
"grad_norm": 0.3257786273828465,
|
| 2023 |
+
"learning_rate": 7.771583838159756e-06,
|
| 2024 |
+
"loss": 0.2626,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 3.6992000000000003,
|
| 2029 |
+
"grad_norm": 0.3129432224019586,
|
| 2030 |
+
"learning_rate": 7.630425394674903e-06,
|
| 2031 |
+
"loss": 0.2849,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 3.7119999999999997,
|
| 2036 |
+
"grad_norm": 0.2971460101083159,
|
| 2037 |
+
"learning_rate": 7.49025786841445e-06,
|
| 2038 |
+
"loss": 0.2538,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 3.7248,
|
| 2043 |
+
"grad_norm": 0.30682223576241,
|
| 2044 |
+
"learning_rate": 7.3510924880924575e-06,
|
| 2045 |
+
"loss": 0.2281,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 3.7376,
|
| 2050 |
+
"grad_norm": 0.29253199059925383,
|
| 2051 |
+
"learning_rate": 7.212940402141808e-06,
|
| 2052 |
+
"loss": 0.2374,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 3.7504,
|
| 2057 |
+
"grad_norm": 0.3198286958387637,
|
| 2058 |
+
"learning_rate": 7.075812677821145e-06,
|
| 2059 |
+
"loss": 0.237,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 3.7632,
|
| 2064 |
+
"grad_norm": 0.29830559498606257,
|
| 2065 |
+
"learning_rate": 6.939720300328303e-06,
|
| 2066 |
+
"loss": 0.208,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 3.776,
|
| 2071 |
+
"grad_norm": 0.32641408005682787,
|
| 2072 |
+
"learning_rate": 6.8046741719202385e-06,
|
| 2073 |
+
"loss": 0.2191,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 3.7888,
|
| 2078 |
+
"grad_norm": 0.3054231611897762,
|
| 2079 |
+
"learning_rate": 6.67068511103971e-06,
|
| 2080 |
+
"loss": 0.2557,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 3.8016,
|
| 2085 |
+
"grad_norm": 0.28819830207122954,
|
| 2086 |
+
"learning_rate": 6.537763851448593e-06,
|
| 2087 |
+
"loss": 0.2408,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 3.8144,
|
| 2092 |
+
"grad_norm": 0.2808978866688667,
|
| 2093 |
+
"learning_rate": 6.4059210413680175e-06,
|
| 2094 |
+
"loss": 0.2305,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 3.8272,
|
| 2099 |
+
"grad_norm": 0.330429061190732,
|
| 2100 |
+
"learning_rate": 6.275167242625331e-06,
|
| 2101 |
+
"loss": 0.1908,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 3.84,
|
| 2106 |
+
"grad_norm": 0.2951130069921395,
|
| 2107 |
+
"learning_rate": 6.145512929808013e-06,
|
| 2108 |
+
"loss": 0.2497,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 3.8528000000000002,
|
| 2113 |
+
"grad_norm": 0.3041071353261243,
|
| 2114 |
+
"learning_rate": 6.016968489424572e-06,
|
| 2115 |
+
"loss": 0.2475,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 3.8656,
|
| 2120 |
+
"grad_norm": 0.2915523771205442,
|
| 2121 |
+
"learning_rate": 5.889544219072465e-06,
|
| 2122 |
+
"loss": 0.2604,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 3.8784,
|
| 2127 |
+
"grad_norm": 0.3175214205339179,
|
| 2128 |
+
"learning_rate": 5.7632503266131925e-06,
|
| 2129 |
+
"loss": 0.2876,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 3.8912,
|
| 2134 |
+
"grad_norm": 0.30033055222130695,
|
| 2135 |
+
"learning_rate": 5.638096929354522e-06,
|
| 2136 |
+
"loss": 0.2354,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 3.904,
|
| 2141 |
+
"grad_norm": 0.2894674764716848,
|
| 2142 |
+
"learning_rate": 5.514094053240035e-06,
|
| 2143 |
+
"loss": 0.2176,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 3.9168,
|
| 2148 |
+
"grad_norm": 0.3026228226715849,
|
| 2149 |
+
"learning_rate": 5.39125163204594e-06,
|
| 2150 |
+
"loss": 0.2451,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 3.9295999999999998,
|
| 2155 |
+
"grad_norm": 0.28666004209756163,
|
| 2156 |
+
"learning_rate": 5.269579506585259e-06,
|
| 2157 |
+
"loss": 0.2294,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 3.9424,
|
| 2162 |
+
"grad_norm": 0.29243736051982305,
|
| 2163 |
+
"learning_rate": 5.149087423919541e-06,
|
| 2164 |
+
"loss": 0.2331,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 3.9552,
|
| 2169 |
+
"grad_norm": 0.3050396561468016,
|
| 2170 |
+
"learning_rate": 5.029785036577976e-06,
|
| 2171 |
+
"loss": 0.2356,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 3.968,
|
| 2176 |
+
"grad_norm": 0.2927321190412269,
|
| 2177 |
+
"learning_rate": 4.911681901784198e-06,
|
| 2178 |
+
"loss": 0.2654,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 3.9808,
|
| 2183 |
+
"grad_norm": 0.29687933189950727,
|
| 2184 |
+
"learning_rate": 4.794787480690597e-06,
|
| 2185 |
+
"loss": 0.2461,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 3.9936,
|
| 2190 |
+
"grad_norm": 0.29055653537661325,
|
| 2191 |
+
"learning_rate": 4.679111137620442e-06,
|
| 2192 |
+
"loss": 0.2637,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 4.0064,
|
| 2197 |
+
"grad_norm": 0.29827242289036976,
|
| 2198 |
+
"learning_rate": 4.5646621393177e-06,
|
| 2199 |
+
"loss": 0.2403,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 4.0192,
|
| 2204 |
+
"grad_norm": 0.3622615155513866,
|
| 2205 |
+
"learning_rate": 4.451449654204685e-06,
|
| 2206 |
+
"loss": 0.2213,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 4.032,
|
| 2211 |
+
"grad_norm": 0.2651792180825393,
|
| 2212 |
+
"learning_rate": 4.339482751647557e-06,
|
| 2213 |
+
"loss": 0.21,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 4.0448,
|
| 2218 |
+
"grad_norm": 0.24169218154190175,
|
| 2219 |
+
"learning_rate": 4.228770401229824e-06,
|
| 2220 |
+
"loss": 0.2208,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 4.0576,
|
| 2225 |
+
"grad_norm": 0.26414429883004253,
|
| 2226 |
+
"learning_rate": 4.119321472033779e-06,
|
| 2227 |
+
"loss": 0.233,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 4.0704,
|
| 2232 |
+
"grad_norm": 0.3130091718310116,
|
| 2233 |
+
"learning_rate": 4.011144731929981e-06,
|
| 2234 |
+
"loss": 0.2,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 4.0832,
|
| 2239 |
+
"grad_norm": 0.29855366750237733,
|
| 2240 |
+
"learning_rate": 3.904248846874894e-06,
|
| 2241 |
+
"loss": 0.2143,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 4.096,
|
| 2246 |
+
"grad_norm": 0.3136541550744826,
|
| 2247 |
+
"learning_rate": 3.7986423802166705e-06,
|
| 2248 |
+
"loss": 0.2069,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 4.1088,
|
| 2253 |
+
"grad_norm": 0.30133265725520775,
|
| 2254 |
+
"learning_rate": 3.694333792009115e-06,
|
| 2255 |
+
"loss": 0.2021,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 4.1216,
|
| 2260 |
+
"grad_norm": 0.2736643470524421,
|
| 2261 |
+
"learning_rate": 3.5913314383339937e-06,
|
| 2262 |
+
"loss": 0.2328,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 4.1344,
|
| 2267 |
+
"grad_norm": 0.28096293851377707,
|
| 2268 |
+
"learning_rate": 3.4896435706316e-06,
|
| 2269 |
+
"loss": 0.1806,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 4.1472,
|
| 2274 |
+
"grad_norm": 0.26794706282661696,
|
| 2275 |
+
"learning_rate": 3.3892783350397675e-06,
|
| 2276 |
+
"loss": 0.1988,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 4.16,
|
| 2281 |
+
"grad_norm": 0.24801672827728208,
|
| 2282 |
+
"learning_rate": 3.290243771741275e-06,
|
| 2283 |
+
"loss": 0.189,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 4.1728,
|
| 2288 |
+
"grad_norm": 0.20979065647430015,
|
| 2289 |
+
"learning_rate": 3.1925478143197418e-06,
|
| 2290 |
+
"loss": 0.1924,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 4.1856,
|
| 2295 |
+
"grad_norm": 0.22875275171944212,
|
| 2296 |
+
"learning_rate": 3.0961982891241083e-06,
|
| 2297 |
+
"loss": 0.2322,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 4.1984,
|
| 2302 |
+
"grad_norm": 0.3008955800785413,
|
| 2303 |
+
"learning_rate": 3.001202914641628e-06,
|
| 2304 |
+
"loss": 0.2167,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 4.2112,
|
| 2309 |
+
"grad_norm": 0.22864015562176315,
|
| 2310 |
+
"learning_rate": 2.907569300879596e-06,
|
| 2311 |
+
"loss": 0.2093,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 4.224,
|
| 2316 |
+
"grad_norm": 0.23366402630095842,
|
| 2317 |
+
"learning_rate": 2.815304948755664e-06,
|
| 2318 |
+
"loss": 0.1797,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 4.2368,
|
| 2323 |
+
"grad_norm": 0.21896877565916695,
|
| 2324 |
+
"learning_rate": 2.7244172494969978e-06,
|
| 2325 |
+
"loss": 0.193,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 4.2496,
|
| 2330 |
+
"grad_norm": 0.22517194394168735,
|
| 2331 |
+
"learning_rate": 2.6349134840481294e-06,
|
| 2332 |
+
"loss": 0.2052,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 4.2624,
|
| 2337 |
+
"grad_norm": 0.22296050511729035,
|
| 2338 |
+
"learning_rate": 2.546800822487714e-06,
|
| 2339 |
+
"loss": 0.2118,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 4.2752,
|
| 2344 |
+
"grad_norm": 0.22480523270915104,
|
| 2345 |
+
"learning_rate": 2.4600863234541338e-06,
|
| 2346 |
+
"loss": 0.2103,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 4.288,
|
| 2351 |
+
"grad_norm": 0.22770961722797545,
|
| 2352 |
+
"learning_rate": 2.374776933580025e-06,
|
| 2353 |
+
"loss": 0.2106,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 4.3008,
|
| 2358 |
+
"grad_norm": 0.22433400468628836,
|
| 2359 |
+
"learning_rate": 2.2908794869358044e-06,
|
| 2360 |
+
"loss": 0.2187,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 4.3136,
|
| 2365 |
+
"grad_norm": 0.23655045540430727,
|
| 2366 |
+
"learning_rate": 2.2084007044821764e-06,
|
| 2367 |
+
"loss": 0.1602,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 4.3264,
|
| 2372 |
+
"grad_norm": 0.2265796633153063,
|
| 2373 |
+
"learning_rate": 2.127347193531757e-06,
|
| 2374 |
+
"loss": 0.2122,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 4.3392,
|
| 2379 |
+
"grad_norm": 0.2215699108063643,
|
| 2380 |
+
"learning_rate": 2.0477254472197237e-06,
|
| 2381 |
+
"loss": 0.1759,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 4.352,
|
| 2386 |
+
"grad_norm": 0.22588332957618415,
|
| 2387 |
+
"learning_rate": 1.96954184398368e-06,
|
| 2388 |
+
"loss": 0.2067,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 4.3648,
|
| 2393 |
+
"grad_norm": 0.21736889134394882,
|
| 2394 |
+
"learning_rate": 1.8928026470526917e-06,
|
| 2395 |
+
"loss": 0.1962,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 4.3776,
|
| 2400 |
+
"grad_norm": 0.23020162042601253,
|
| 2401 |
+
"learning_rate": 1.817514003945524e-06,
|
| 2402 |
+
"loss": 0.201,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 4.3904,
|
| 2407 |
+
"grad_norm": 0.21569896695509316,
|
| 2408 |
+
"learning_rate": 1.743681945978184e-06,
|
| 2409 |
+
"loss": 0.1873,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 4.4032,
|
| 2414 |
+
"grad_norm": 0.22261015084745364,
|
| 2415 |
+
"learning_rate": 1.6713123877807413e-06,
|
| 2416 |
+
"loss": 0.2103,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 4.416,
|
| 2421 |
+
"grad_norm": 0.20959081547567032,
|
| 2422 |
+
"learning_rate": 1.6004111268235156e-06,
|
| 2423 |
+
"loss": 0.1897,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 4.4288,
|
| 2428 |
+
"grad_norm": 0.2245061712954699,
|
| 2429 |
+
"learning_rate": 1.5309838429526714e-06,
|
| 2430 |
+
"loss": 0.1392,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 4.4416,
|
| 2435 |
+
"grad_norm": 0.22011225244684698,
|
| 2436 |
+
"learning_rate": 1.4630360979351644e-06,
|
| 2437 |
+
"loss": 0.2109,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 4.4544,
|
| 2442 |
+
"grad_norm": 0.21542343216558188,
|
| 2443 |
+
"learning_rate": 1.396573335013236e-06,
|
| 2444 |
+
"loss": 0.2324,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 4.4672,
|
| 2449 |
+
"grad_norm": 0.2238750665819962,
|
| 2450 |
+
"learning_rate": 1.3316008784683265e-06,
|
| 2451 |
+
"loss": 0.1742,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 4.48,
|
| 2456 |
+
"grad_norm": 0.20408276618504737,
|
| 2457 |
+
"learning_rate": 1.2681239331945695e-06,
|
| 2458 |
+
"loss": 0.1978,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 4.4928,
|
| 2463 |
+
"grad_norm": 0.21869312114432388,
|
| 2464 |
+
"learning_rate": 1.2061475842818337e-06,
|
| 2465 |
+
"loss": 0.1657,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 4.5056,
|
| 2470 |
+
"grad_norm": 0.21558851650925154,
|
| 2471 |
+
"learning_rate": 1.1456767966083393e-06,
|
| 2472 |
+
"loss": 0.1587,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 4.5184,
|
| 2477 |
+
"grad_norm": 0.21526876311324686,
|
| 2478 |
+
"learning_rate": 1.086716414442952e-06,
|
| 2479 |
+
"loss": 0.2101,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 4.5312,
|
| 2484 |
+
"grad_norm": 0.23669085539753498,
|
| 2485 |
+
"learning_rate": 1.0292711610570904e-06,
|
| 2486 |
+
"loss": 0.2049,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 4.5440000000000005,
|
| 2491 |
+
"grad_norm": 0.2147383641646149,
|
| 2492 |
+
"learning_rate": 9.733456383463658e-07,
|
| 2493 |
+
"loss": 0.2012,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 4.5568,
|
| 2498 |
+
"grad_norm": 0.21993975256883413,
|
| 2499 |
+
"learning_rate": 9.189443264619102e-07,
|
| 2500 |
+
"loss": 0.159,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 4.5696,
|
| 2505 |
+
"grad_norm": 0.20704280436682335,
|
| 2506 |
+
"learning_rate": 8.660715834514977e-07,
|
| 2507 |
+
"loss": 0.2085,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 4.5824,
|
| 2512 |
+
"grad_norm": 0.21921381853186572,
|
| 2513 |
+
"learning_rate": 8.147316449103959e-07,
|
| 2514 |
+
"loss": 0.2124,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 4.5952,
|
| 2519 |
+
"grad_norm": 0.21300778947324445,
|
| 2520 |
+
"learning_rate": 7.649286236420806e-07,
|
| 2521 |
+
"loss": 0.2301,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 4.608,
|
| 2526 |
+
"grad_norm": 0.203882908460394,
|
| 2527 |
+
"learning_rate": 7.166665093287539e-07,
|
| 2528 |
+
"loss": 0.1693,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 4.6208,
|
| 2533 |
+
"grad_norm": 0.21817271507009403,
|
| 2534 |
+
"learning_rate": 6.69949168211721e-07,
|
| 2535 |
+
"loss": 0.1993,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 4.6336,
|
| 2540 |
+
"grad_norm": 0.211988722708523,
|
| 2541 |
+
"learning_rate": 6.247803427816945e-07,
|
| 2542 |
+
"loss": 0.2301,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 4.6464,
|
| 2547 |
+
"grad_norm": 0.22603033808260806,
|
| 2548 |
+
"learning_rate": 5.811636514789598e-07,
|
| 2549 |
+
"loss": 0.1919,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 4.6592,
|
| 2554 |
+
"grad_norm": 0.20179555800371948,
|
| 2555 |
+
"learning_rate": 5.391025884035239e-07,
|
| 2556 |
+
"loss": 0.1928,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 4.672,
|
| 2561 |
+
"grad_norm": 0.21557542653649922,
|
| 2562 |
+
"learning_rate": 4.986005230351954e-07,
|
| 2563 |
+
"loss": 0.1949,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 4.6848,
|
| 2568 |
+
"grad_norm": 0.20474732215155866,
|
| 2569 |
+
"learning_rate": 4.5966069996365993e-07,
|
| 2570 |
+
"loss": 0.2339,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 4.6975999999999996,
|
| 2575 |
+
"grad_norm": 0.2165988646827934,
|
| 2576 |
+
"learning_rate": 4.22286238628562e-07,
|
| 2577 |
+
"loss": 0.241,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 4.7104,
|
| 2582 |
+
"grad_norm": 0.21668548464775883,
|
| 2583 |
+
"learning_rate": 3.8648013306960664e-07,
|
| 2584 |
+
"loss": 0.1941,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 4.7232,
|
| 2589 |
+
"grad_norm": 0.2597288042736453,
|
| 2590 |
+
"learning_rate": 3.522452516867048e-07,
|
| 2591 |
+
"loss": 0.1644,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 4.736,
|
| 2596 |
+
"grad_norm": 0.21747406103057232,
|
| 2597 |
+
"learning_rate": 3.1958433701019697e-07,
|
| 2598 |
+
"loss": 0.201,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 4.7488,
|
| 2603 |
+
"grad_norm": 0.21474318980184318,
|
| 2604 |
+
"learning_rate": 2.8850000548115155e-07,
|
| 2605 |
+
"loss": 0.2217,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 4.7616,
|
| 2610 |
+
"grad_norm": 0.21648982898785513,
|
| 2611 |
+
"learning_rate": 2.5899474724174313e-07,
|
| 2612 |
+
"loss": 0.2083,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 4.7744,
|
| 2617 |
+
"grad_norm": 0.22446238548571779,
|
| 2618 |
+
"learning_rate": 2.3107092593579905e-07,
|
| 2619 |
+
"loss": 0.2531,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 4.7872,
|
| 2624 |
+
"grad_norm": 0.209769555619485,
|
| 2625 |
+
"learning_rate": 2.0473077851942858e-07,
|
| 2626 |
+
"loss": 0.177,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 4.8,
|
| 2631 |
+
"grad_norm": 0.21029622810200657,
|
| 2632 |
+
"learning_rate": 1.799764150818306e-07,
|
| 2633 |
+
"loss": 0.219,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 4.8128,
|
| 2638 |
+
"grad_norm": 0.2065844651417679,
|
| 2639 |
+
"learning_rate": 1.5680981867625566e-07,
|
| 2640 |
+
"loss": 0.1967,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 4.8256,
|
| 2645 |
+
"grad_norm": 0.1961007926786603,
|
| 2646 |
+
"learning_rate": 1.3523284516113955e-07,
|
| 2647 |
+
"loss": 0.2279,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 4.8384,
|
| 2652 |
+
"grad_norm": 0.21632855311757646,
|
| 2653 |
+
"learning_rate": 1.1524722305144231e-07,
|
| 2654 |
+
"loss": 0.2055,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 4.8512,
|
| 2659 |
+
"grad_norm": 0.21324193299133418,
|
| 2660 |
+
"learning_rate": 9.685455338016347e-08,
|
| 2661 |
+
"loss": 0.2086,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 4.864,
|
| 2666 |
+
"grad_norm": 0.21767593623161713,
|
| 2667 |
+
"learning_rate": 8.005630957010014e-08,
|
| 2668 |
+
"loss": 0.1838,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 4.8768,
|
| 2673 |
+
"grad_norm": 0.20857092252520282,
|
| 2674 |
+
"learning_rate": 6.485383731580142e-08,
|
| 2675 |
+
"loss": 0.1987,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 4.8896,
|
| 2680 |
+
"grad_norm": 0.21623493864374912,
|
| 2681 |
+
"learning_rate": 5.1248354475768034e-08,
|
| 2682 |
+
"loss": 0.1694,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 4.9024,
|
| 2687 |
+
"grad_norm": 0.22065768387382806,
|
| 2688 |
+
"learning_rate": 3.924095097489922e-08,
|
| 2689 |
+
"loss": 0.1833,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 4.9152000000000005,
|
| 2694 |
+
"grad_norm": 0.2246620696428788,
|
| 2695 |
+
"learning_rate": 2.8832588717164766e-08,
|
| 2696 |
+
"loss": 0.1769,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 4.928,
|
| 2701 |
+
"grad_norm": 0.2130616454697376,
|
| 2702 |
+
"learning_rate": 2.0024101508555604e-08,
|
| 2703 |
+
"loss": 0.2304,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 4.9408,
|
| 2708 |
+
"grad_norm": 0.21365009157227124,
|
| 2709 |
+
"learning_rate": 1.281619499029274e-08,
|
| 2710 |
+
"loss": 0.2475,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 4.9536,
|
| 2715 |
+
"grad_norm": 0.1985785825291255,
|
| 2716 |
+
"learning_rate": 7.209446582292501e-09,
|
| 2717 |
+
"loss": 0.1996,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 4.9664,
|
| 2722 |
+
"grad_norm": 0.20073536666154784,
|
| 2723 |
+
"learning_rate": 3.2043054369057523e-09,
|
| 2724 |
+
"loss": 0.1874,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 4.9792,
|
| 2729 |
+
"grad_norm": 0.21381598752008804,
|
| 2730 |
+
"learning_rate": 8.010924029533406e-10,
|
| 2731 |
+
"loss": 0.1781,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 4.992,
|
| 2736 |
+
"grad_norm": 0.21157933427703016,
|
| 2737 |
+
"learning_rate": 0.0,
|
| 2738 |
+
"loss": 0.1731,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 4.992,
|
| 2743 |
+
"step": 390,
|
| 2744 |
+
"total_flos": 8.727962333646684e+17,
|
| 2745 |
+
"train_loss": 0.3591920976837476,
|
| 2746 |
+
"train_runtime": 40468.1043,
|
| 2747 |
+
"train_samples_per_second": 1.236,
|
| 2748 |
+
"train_steps_per_second": 0.01
|
| 2749 |
+
}
|
| 2750 |
+
],
|
| 2751 |
+
"logging_steps": 1,
|
| 2752 |
+
"max_steps": 390,
|
| 2753 |
+
"num_input_tokens_seen": 0,
|
| 2754 |
+
"num_train_epochs": 5,
|
| 2755 |
+
"save_steps": 500,
|
| 2756 |
+
"stateful_callbacks": {
|
| 2757 |
+
"TrainerControl": {
|
| 2758 |
+
"args": {
|
| 2759 |
+
"should_epoch_stop": false,
|
| 2760 |
+
"should_evaluate": false,
|
| 2761 |
+
"should_log": false,
|
| 2762 |
+
"should_save": true,
|
| 2763 |
+
"should_training_stop": true
|
| 2764 |
+
},
|
| 2765 |
+
"attributes": {}
|
| 2766 |
+
}
|
| 2767 |
+
},
|
| 2768 |
+
"total_flos": 8.727962333646684e+17,
|
| 2769 |
+
"train_batch_size": 1,
|
| 2770 |
+
"trial_name": null,
|
| 2771 |
+
"trial_params": null
|
| 2772 |
+
}
|
training_loss.png
ADDED
|