End of training
Browse files- README.md +3 -2
- all_results.json +12 -0
- eval_results.json +7 -0
- runs/May10_13-13-54_d96183b7b29f/events.out.tfevents.1715358763.d96183b7b29f.515.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +3060 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -2,6 +2,7 @@
|
|
| 2 |
library_name: peft
|
| 3 |
tags:
|
| 4 |
- llama-factory
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
base_model: ytcheng/llama-3-8B-pretrain
|
| 7 |
model-index:
|
|
@@ -14,9 +15,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# llama-3-8B-lora-sft
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [ytcheng/llama-3-8B-pretrain](https://huggingface.co/ytcheng/llama-3-8B-pretrain) on
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss: 1.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 2 |
library_name: peft
|
| 3 |
tags:
|
| 4 |
- llama-factory
|
| 5 |
+
- lora
|
| 6 |
- generated_from_trainer
|
| 7 |
base_model: ytcheng/llama-3-8B-pretrain
|
| 8 |
model-index:
|
|
|
|
| 15 |
|
| 16 |
# llama-3-8B-lora-sft
|
| 17 |
|
| 18 |
+
This model is a fine-tuned version of [ytcheng/llama-3-8B-pretrain](https://huggingface.co/ytcheng/llama-3-8B-pretrain) on the sm_question dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 1.9779
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9986300118732303,
|
| 3 |
+
"eval_loss": 1.9779435396194458,
|
| 4 |
+
"eval_runtime": 105.0149,
|
| 5 |
+
"eval_samples_per_second": 11.589,
|
| 6 |
+
"eval_steps_per_second": 11.589,
|
| 7 |
+
"total_flos": 1.3798092749438976e+17,
|
| 8 |
+
"train_loss": 2.123500373163651,
|
| 9 |
+
"train_runtime": 11771.2816,
|
| 10 |
+
"train_samples_per_second": 2.79,
|
| 11 |
+
"train_steps_per_second": 0.349
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9986300118732303,
|
| 3 |
+
"eval_loss": 1.9779435396194458,
|
| 4 |
+
"eval_runtime": 105.0149,
|
| 5 |
+
"eval_samples_per_second": 11.589,
|
| 6 |
+
"eval_steps_per_second": 11.589
|
| 7 |
+
}
|
runs/May10_13-13-54_d96183b7b29f/events.out.tfevents.1715358763.d96183b7b29f.515.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06ae20d0bae59236f3e098af45e3696f55523f028e2b14ed0e902d1ed2f6eae0
|
| 3 |
+
size 359
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9986300118732303,
|
| 3 |
+
"total_flos": 1.3798092749438976e+17,
|
| 4 |
+
"train_loss": 2.123500373163651,
|
| 5 |
+
"train_runtime": 11771.2816,
|
| 6 |
+
"train_samples_per_second": 2.79,
|
| 7 |
+
"train_steps_per_second": 0.349
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3060 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9986300118732303,
|
| 5 |
+
"eval_steps": 200,
|
| 6 |
+
"global_step": 4104,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0073066033427710295,
|
| 13 |
+
"grad_norm": 1.7583225965499878,
|
| 14 |
+
"learning_rate": 2.1897810218978103e-06,
|
| 15 |
+
"loss": 3.655,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.014613206685542059,
|
| 20 |
+
"grad_norm": NaN,
|
| 21 |
+
"learning_rate": 4.379562043795621e-06,
|
| 22 |
+
"loss": 4.0596,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02191981002831309,
|
| 27 |
+
"grad_norm": 2.3730218410491943,
|
| 28 |
+
"learning_rate": 6.812652068126521e-06,
|
| 29 |
+
"loss": 4.2837,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.029226413371084118,
|
| 34 |
+
"grad_norm": 3.1703426837921143,
|
| 35 |
+
"learning_rate": 9.24574209245742e-06,
|
| 36 |
+
"loss": 3.8455,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03653301671385514,
|
| 41 |
+
"grad_norm": 2.987401247024536,
|
| 42 |
+
"learning_rate": 1.1678832116788322e-05,
|
| 43 |
+
"loss": 3.5357,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04383962005662618,
|
| 48 |
+
"grad_norm": 4.219663619995117,
|
| 49 |
+
"learning_rate": 1.411192214111922e-05,
|
| 50 |
+
"loss": 3.6343,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05114622339939721,
|
| 55 |
+
"grad_norm": 4.096388816833496,
|
| 56 |
+
"learning_rate": 1.654501216545012e-05,
|
| 57 |
+
"loss": 3.4301,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.058452826742168236,
|
| 62 |
+
"grad_norm": 10.184355735778809,
|
| 63 |
+
"learning_rate": 1.8734793187347933e-05,
|
| 64 |
+
"loss": 3.2771,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06575943008493926,
|
| 69 |
+
"grad_norm": 8.512189865112305,
|
| 70 |
+
"learning_rate": 2.1167883211678834e-05,
|
| 71 |
+
"loss": 2.9386,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.07306603342771029,
|
| 76 |
+
"grad_norm": 5.223749160766602,
|
| 77 |
+
"learning_rate": 2.360097323600973e-05,
|
| 78 |
+
"loss": 2.6416,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08037263677048133,
|
| 83 |
+
"grad_norm": 5.041982173919678,
|
| 84 |
+
"learning_rate": 2.6034063260340636e-05,
|
| 85 |
+
"loss": 2.78,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08767924011325236,
|
| 90 |
+
"grad_norm": 8.065332412719727,
|
| 91 |
+
"learning_rate": 2.8467153284671533e-05,
|
| 92 |
+
"loss": 2.7657,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09498584345602339,
|
| 97 |
+
"grad_norm": 2.3226640224456787,
|
| 98 |
+
"learning_rate": 3.0900243309002434e-05,
|
| 99 |
+
"loss": 2.7316,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.10229244679879441,
|
| 104 |
+
"grad_norm": 3.4772486686706543,
|
| 105 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 106 |
+
"loss": 2.6205,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.10959905014156544,
|
| 111 |
+
"grad_norm": 2.8939208984375,
|
| 112 |
+
"learning_rate": 3.5766423357664236e-05,
|
| 113 |
+
"loss": 2.648,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.11690565348433647,
|
| 118 |
+
"grad_norm": 2.951481342315674,
|
| 119 |
+
"learning_rate": 3.819951338199514e-05,
|
| 120 |
+
"loss": 2.6972,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1242122568271075,
|
| 125 |
+
"grad_norm": 2.277855157852173,
|
| 126 |
+
"learning_rate": 4.063260340632604e-05,
|
| 127 |
+
"loss": 2.4787,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.13151886016987852,
|
| 132 |
+
"grad_norm": 9.363456726074219,
|
| 133 |
+
"learning_rate": 4.306569343065693e-05,
|
| 134 |
+
"loss": 2.6347,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.13882546351264954,
|
| 139 |
+
"grad_norm": 3.6133177280426025,
|
| 140 |
+
"learning_rate": 4.549878345498784e-05,
|
| 141 |
+
"loss": 2.656,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.14613206685542057,
|
| 146 |
+
"grad_norm": 4.037402153015137,
|
| 147 |
+
"learning_rate": 4.793187347931874e-05,
|
| 148 |
+
"loss": 2.5512,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.14613206685542057,
|
| 153 |
+
"eval_loss": 2.5883829593658447,
|
| 154 |
+
"eval_runtime": 109.0986,
|
| 155 |
+
"eval_samples_per_second": 11.155,
|
| 156 |
+
"eval_steps_per_second": 11.155,
|
| 157 |
+
"step": 200
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.15343867019819163,
|
| 161 |
+
"grad_norm": 3.737877368927002,
|
| 162 |
+
"learning_rate": 5.036496350364964e-05,
|
| 163 |
+
"loss": 2.5724,
|
| 164 |
+
"step": 210
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.16074527354096266,
|
| 168 |
+
"grad_norm": 3.6852433681488037,
|
| 169 |
+
"learning_rate": 5.279805352798054e-05,
|
| 170 |
+
"loss": 2.6093,
|
| 171 |
+
"step": 220
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1680518768837337,
|
| 175 |
+
"grad_norm": 4.3811936378479,
|
| 176 |
+
"learning_rate": 5.5231143552311436e-05,
|
| 177 |
+
"loss": 2.5968,
|
| 178 |
+
"step": 230
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.17535848022650471,
|
| 182 |
+
"grad_norm": 4.019372940063477,
|
| 183 |
+
"learning_rate": 5.766423357664234e-05,
|
| 184 |
+
"loss": 2.7364,
|
| 185 |
+
"step": 240
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.18266508356927574,
|
| 189 |
+
"grad_norm": 4.277368068695068,
|
| 190 |
+
"learning_rate": 6.0097323600973245e-05,
|
| 191 |
+
"loss": 2.7445,
|
| 192 |
+
"step": 250
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.18997168691204677,
|
| 196 |
+
"grad_norm": 11.767447471618652,
|
| 197 |
+
"learning_rate": 6.253041362530415e-05,
|
| 198 |
+
"loss": 2.6712,
|
| 199 |
+
"step": 260
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.1972782902548178,
|
| 203 |
+
"grad_norm": 3.2719593048095703,
|
| 204 |
+
"learning_rate": 6.496350364963504e-05,
|
| 205 |
+
"loss": 2.5184,
|
| 206 |
+
"step": 270
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.20458489359758883,
|
| 210 |
+
"grad_norm": 3.923417091369629,
|
| 211 |
+
"learning_rate": 6.739659367396593e-05,
|
| 212 |
+
"loss": 2.6089,
|
| 213 |
+
"step": 280
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.21189149694035986,
|
| 217 |
+
"grad_norm": 5.958123207092285,
|
| 218 |
+
"learning_rate": 6.982968369829684e-05,
|
| 219 |
+
"loss": 2.5422,
|
| 220 |
+
"step": 290
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.2191981002831309,
|
| 224 |
+
"grad_norm": 4.775500774383545,
|
| 225 |
+
"learning_rate": 7.226277372262774e-05,
|
| 226 |
+
"loss": 2.3889,
|
| 227 |
+
"step": 300
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.22650470362590192,
|
| 231 |
+
"grad_norm": 7.674183368682861,
|
| 232 |
+
"learning_rate": 7.469586374695864e-05,
|
| 233 |
+
"loss": 2.4629,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.23381130696867294,
|
| 238 |
+
"grad_norm": 5.1322760581970215,
|
| 239 |
+
"learning_rate": 7.712895377128954e-05,
|
| 240 |
+
"loss": 2.5144,
|
| 241 |
+
"step": 320
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.24111791031144397,
|
| 245 |
+
"grad_norm": 5.469252586364746,
|
| 246 |
+
"learning_rate": 7.956204379562045e-05,
|
| 247 |
+
"loss": 2.5871,
|
| 248 |
+
"step": 330
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.248424513654215,
|
| 252 |
+
"grad_norm": 4.472785472869873,
|
| 253 |
+
"learning_rate": 8.199513381995134e-05,
|
| 254 |
+
"loss": 2.4764,
|
| 255 |
+
"step": 340
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.25573111699698603,
|
| 259 |
+
"grad_norm": 7.12663459777832,
|
| 260 |
+
"learning_rate": 8.442822384428223e-05,
|
| 261 |
+
"loss": 2.5562,
|
| 262 |
+
"step": 350
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.26303772033975703,
|
| 266 |
+
"grad_norm": 3.2579259872436523,
|
| 267 |
+
"learning_rate": 8.686131386861314e-05,
|
| 268 |
+
"loss": 2.4707,
|
| 269 |
+
"step": 360
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.2703443236825281,
|
| 273 |
+
"grad_norm": 4.707943916320801,
|
| 274 |
+
"learning_rate": 8.929440389294405e-05,
|
| 275 |
+
"loss": 2.5207,
|
| 276 |
+
"step": 370
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.2776509270252991,
|
| 280 |
+
"grad_norm": 8.757339477539062,
|
| 281 |
+
"learning_rate": 9.172749391727494e-05,
|
| 282 |
+
"loss": 2.5424,
|
| 283 |
+
"step": 380
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.28495753036807014,
|
| 287 |
+
"grad_norm": 6.260244369506836,
|
| 288 |
+
"learning_rate": 9.416058394160584e-05,
|
| 289 |
+
"loss": 2.5552,
|
| 290 |
+
"step": 390
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.29226413371084115,
|
| 294 |
+
"grad_norm": 3.4638991355895996,
|
| 295 |
+
"learning_rate": 9.659367396593674e-05,
|
| 296 |
+
"loss": 2.3861,
|
| 297 |
+
"step": 400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.29226413371084115,
|
| 301 |
+
"eval_loss": 2.477276563644409,
|
| 302 |
+
"eval_runtime": 107.9591,
|
| 303 |
+
"eval_samples_per_second": 11.273,
|
| 304 |
+
"eval_steps_per_second": 11.273,
|
| 305 |
+
"step": 400
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.2995707370536122,
|
| 309 |
+
"grad_norm": 4.783712863922119,
|
| 310 |
+
"learning_rate": 9.902676399026765e-05,
|
| 311 |
+
"loss": 2.5961,
|
| 312 |
+
"step": 410
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.30687734039638326,
|
| 316 |
+
"grad_norm": 6.2626729011535645,
|
| 317 |
+
"learning_rate": 9.999934869757279e-05,
|
| 318 |
+
"loss": 2.5863,
|
| 319 |
+
"step": 420
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.31418394373915426,
|
| 323 |
+
"grad_norm": 4.87408971786499,
|
| 324 |
+
"learning_rate": 9.999536857752013e-05,
|
| 325 |
+
"loss": 2.3565,
|
| 326 |
+
"step": 430
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.3214905470819253,
|
| 330 |
+
"grad_norm": 3.7447562217712402,
|
| 331 |
+
"learning_rate": 9.998777045977545e-05,
|
| 332 |
+
"loss": 2.4354,
|
| 333 |
+
"step": 440
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.3287971504246963,
|
| 337 |
+
"grad_norm": 5.334987640380859,
|
| 338 |
+
"learning_rate": 9.997655489418913e-05,
|
| 339 |
+
"loss": 2.412,
|
| 340 |
+
"step": 450
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.3361037537674674,
|
| 344 |
+
"grad_norm": 5.347410678863525,
|
| 345 |
+
"learning_rate": 9.996172269239417e-05,
|
| 346 |
+
"loss": 2.4486,
|
| 347 |
+
"step": 460
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.3434103571102384,
|
| 351 |
+
"grad_norm": 5.353787899017334,
|
| 352 |
+
"learning_rate": 9.99432749277474e-05,
|
| 353 |
+
"loss": 2.3383,
|
| 354 |
+
"step": 470
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.35071696045300943,
|
| 358 |
+
"grad_norm": 5.850295543670654,
|
| 359 |
+
"learning_rate": 9.992121293525189e-05,
|
| 360 |
+
"loss": 2.3838,
|
| 361 |
+
"step": 480
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.35802356379578043,
|
| 365 |
+
"grad_norm": 4.7518792152404785,
|
| 366 |
+
"learning_rate": 9.98955383114603e-05,
|
| 367 |
+
"loss": 2.427,
|
| 368 |
+
"step": 490
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.3653301671385515,
|
| 372 |
+
"grad_norm": 7.027746677398682,
|
| 373 |
+
"learning_rate": 9.986625291435933e-05,
|
| 374 |
+
"loss": 2.608,
|
| 375 |
+
"step": 500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.3726367704813225,
|
| 379 |
+
"grad_norm": 5.194112300872803,
|
| 380 |
+
"learning_rate": 9.983335886323523e-05,
|
| 381 |
+
"loss": 2.5837,
|
| 382 |
+
"step": 510
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.37994337382409354,
|
| 386 |
+
"grad_norm": 5.106868743896484,
|
| 387 |
+
"learning_rate": 9.979685853852057e-05,
|
| 388 |
+
"loss": 2.4636,
|
| 389 |
+
"step": 520
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.38724997716686455,
|
| 393 |
+
"grad_norm": 4.584632396697998,
|
| 394 |
+
"learning_rate": 9.975675458162177e-05,
|
| 395 |
+
"loss": 2.3892,
|
| 396 |
+
"step": 530
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.3945565805096356,
|
| 400 |
+
"grad_norm": 3.168287992477417,
|
| 401 |
+
"learning_rate": 9.971304989472819e-05,
|
| 402 |
+
"loss": 2.4344,
|
| 403 |
+
"step": 540
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.4018631838524066,
|
| 407 |
+
"grad_norm": 5.590039253234863,
|
| 408 |
+
"learning_rate": 9.966574764060186e-05,
|
| 409 |
+
"loss": 2.377,
|
| 410 |
+
"step": 550
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.40916978719517766,
|
| 414 |
+
"grad_norm": 3.9408605098724365,
|
| 415 |
+
"learning_rate": 9.961485124234881e-05,
|
| 416 |
+
"loss": 2.4833,
|
| 417 |
+
"step": 560
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.41647639053794866,
|
| 421 |
+
"grad_norm": 3.6905062198638916,
|
| 422 |
+
"learning_rate": 9.956036438317124e-05,
|
| 423 |
+
"loss": 2.5575,
|
| 424 |
+
"step": 570
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.4237829938807197,
|
| 428 |
+
"grad_norm": 5.282037258148193,
|
| 429 |
+
"learning_rate": 9.9502291006101e-05,
|
| 430 |
+
"loss": 2.438,
|
| 431 |
+
"step": 580
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.4310895972234907,
|
| 435 |
+
"grad_norm": 3.3918325901031494,
|
| 436 |
+
"learning_rate": 9.944063531371423e-05,
|
| 437 |
+
"loss": 2.5227,
|
| 438 |
+
"step": 590
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.4383962005662618,
|
| 442 |
+
"grad_norm": 3.809112787246704,
|
| 443 |
+
"learning_rate": 9.937540176782732e-05,
|
| 444 |
+
"loss": 2.5938,
|
| 445 |
+
"step": 600
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.4383962005662618,
|
| 449 |
+
"eval_loss": 2.399010181427002,
|
| 450 |
+
"eval_runtime": 107.2947,
|
| 451 |
+
"eval_samples_per_second": 11.343,
|
| 452 |
+
"eval_steps_per_second": 11.343,
|
| 453 |
+
"step": 600
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.4457028039090328,
|
| 457 |
+
"grad_norm": 6.199471473693848,
|
| 458 |
+
"learning_rate": 9.930659508917388e-05,
|
| 459 |
+
"loss": 2.4369,
|
| 460 |
+
"step": 610
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.45300940725180383,
|
| 464 |
+
"grad_norm": 4.105823993682861,
|
| 465 |
+
"learning_rate": 9.923422025706323e-05,
|
| 466 |
+
"loss": 2.5947,
|
| 467 |
+
"step": 620
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.46031601059457483,
|
| 471 |
+
"grad_norm": 4.27271842956543,
|
| 472 |
+
"learning_rate": 9.915828250902004e-05,
|
| 473 |
+
"loss": 2.3976,
|
| 474 |
+
"step": 630
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.4676226139373459,
|
| 478 |
+
"grad_norm": 3.6033222675323486,
|
| 479 |
+
"learning_rate": 9.907878734040525e-05,
|
| 480 |
+
"loss": 2.3318,
|
| 481 |
+
"step": 640
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.4749292172801169,
|
| 485 |
+
"grad_norm": 3.8373398780822754,
|
| 486 |
+
"learning_rate": 9.89957405040185e-05,
|
| 487 |
+
"loss": 2.4319,
|
| 488 |
+
"step": 650
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.48223582062288795,
|
| 492 |
+
"grad_norm": 3.295011520385742,
|
| 493 |
+
"learning_rate": 9.890914800968171e-05,
|
| 494 |
+
"loss": 2.2978,
|
| 495 |
+
"step": 660
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.48954242396565895,
|
| 499 |
+
"grad_norm": 3.935311794281006,
|
| 500 |
+
"learning_rate": 9.88190161238042e-05,
|
| 501 |
+
"loss": 2.4859,
|
| 502 |
+
"step": 670
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 0.49684902730843,
|
| 506 |
+
"grad_norm": 4.868900299072266,
|
| 507 |
+
"learning_rate": 9.872535136892926e-05,
|
| 508 |
+
"loss": 2.3433,
|
| 509 |
+
"step": 680
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.5041556306512011,
|
| 513 |
+
"grad_norm": 6.791601181030273,
|
| 514 |
+
"learning_rate": 9.862816052326209e-05,
|
| 515 |
+
"loss": 2.3252,
|
| 516 |
+
"step": 690
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.5114622339939721,
|
| 520 |
+
"grad_norm": 2.8761675357818604,
|
| 521 |
+
"learning_rate": 9.852745062017927e-05,
|
| 522 |
+
"loss": 2.4546,
|
| 523 |
+
"step": 700
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.5187688373367431,
|
| 527 |
+
"grad_norm": 7.200977802276611,
|
| 528 |
+
"learning_rate": 9.84338089316251e-05,
|
| 529 |
+
"loss": 2.462,
|
| 530 |
+
"step": 710
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.5260754406795141,
|
| 534 |
+
"grad_norm": 5.1391825675964355,
|
| 535 |
+
"learning_rate": 9.832643310805385e-05,
|
| 536 |
+
"loss": 2.601,
|
| 537 |
+
"step": 720
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.5333820440222852,
|
| 541 |
+
"grad_norm": 3.5819122791290283,
|
| 542 |
+
"learning_rate": 9.821556006207133e-05,
|
| 543 |
+
"loss": 2.3209,
|
| 544 |
+
"step": 730
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.5406886473650562,
|
| 548 |
+
"grad_norm": 3.3951661586761475,
|
| 549 |
+
"learning_rate": 9.810119781718924e-05,
|
| 550 |
+
"loss": 2.1652,
|
| 551 |
+
"step": 740
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.5479952507078272,
|
| 555 |
+
"grad_norm": 5.344549655914307,
|
| 556 |
+
"learning_rate": 9.798335464942094e-05,
|
| 557 |
+
"loss": 2.3954,
|
| 558 |
+
"step": 750
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.5553018540505982,
|
| 562 |
+
"grad_norm": 5.5213446617126465,
|
| 563 |
+
"learning_rate": 9.786203908668255e-05,
|
| 564 |
+
"loss": 2.1831,
|
| 565 |
+
"step": 760
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.5626084573933693,
|
| 569 |
+
"grad_norm": 4.242595195770264,
|
| 570 |
+
"learning_rate": 9.773725990817575e-05,
|
| 571 |
+
"loss": 2.4479,
|
| 572 |
+
"step": 770
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.5699150607361403,
|
| 576 |
+
"grad_norm": 5.442046165466309,
|
| 577 |
+
"learning_rate": 9.76090261437526e-05,
|
| 578 |
+
"loss": 2.2424,
|
| 579 |
+
"step": 780
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5772216640789113,
|
| 583 |
+
"grad_norm": 5.669281482696533,
|
| 584 |
+
"learning_rate": 9.747734707326195e-05,
|
| 585 |
+
"loss": 2.1442,
|
| 586 |
+
"step": 790
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.5845282674216823,
|
| 590 |
+
"grad_norm": 6.425075531005859,
|
| 591 |
+
"learning_rate": 9.734223222587792e-05,
|
| 592 |
+
"loss": 2.3384,
|
| 593 |
+
"step": 800
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5845282674216823,
|
| 597 |
+
"eval_loss": 2.348578691482544,
|
| 598 |
+
"eval_runtime": 108.5263,
|
| 599 |
+
"eval_samples_per_second": 11.214,
|
| 600 |
+
"eval_steps_per_second": 11.214,
|
| 601 |
+
"step": 800
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.5918348707644534,
|
| 605 |
+
"grad_norm": 5.118699073791504,
|
| 606 |
+
"learning_rate": 9.720369137941034e-05,
|
| 607 |
+
"loss": 2.4253,
|
| 608 |
+
"step": 810
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.5991414741072244,
|
| 612 |
+
"grad_norm": 6.51192569732666,
|
| 613 |
+
"learning_rate": 9.706173455959715e-05,
|
| 614 |
+
"loss": 2.5238,
|
| 615 |
+
"step": 820
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.6064480774499954,
|
| 619 |
+
"grad_norm": 4.507262706756592,
|
| 620 |
+
"learning_rate": 9.69163720393788e-05,
|
| 621 |
+
"loss": 2.4986,
|
| 622 |
+
"step": 830
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 0.6137546807927665,
|
| 626 |
+
"grad_norm": 5.540140151977539,
|
| 627 |
+
"learning_rate": 9.676761433815498e-05,
|
| 628 |
+
"loss": 2.2241,
|
| 629 |
+
"step": 840
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.6210612841355375,
|
| 633 |
+
"grad_norm": 3.475051164627075,
|
| 634 |
+
"learning_rate": 9.661547222102323e-05,
|
| 635 |
+
"loss": 2.3395,
|
| 636 |
+
"step": 850
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 0.6283678874783085,
|
| 640 |
+
"grad_norm": 5.291077136993408,
|
| 641 |
+
"learning_rate": 9.645995669799995e-05,
|
| 642 |
+
"loss": 2.2988,
|
| 643 |
+
"step": 860
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.6356744908210795,
|
| 647 |
+
"grad_norm": 3.973806858062744,
|
| 648 |
+
"learning_rate": 9.630107902322367e-05,
|
| 649 |
+
"loss": 2.3554,
|
| 650 |
+
"step": 870
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.6429810941638506,
|
| 654 |
+
"grad_norm": 4.875363349914551,
|
| 655 |
+
"learning_rate": 9.613885069414061e-05,
|
| 656 |
+
"loss": 2.3115,
|
| 657 |
+
"step": 880
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.6502876975066216,
|
| 661 |
+
"grad_norm": 4.581294059753418,
|
| 662 |
+
"learning_rate": 9.597328345067259e-05,
|
| 663 |
+
"loss": 2.1619,
|
| 664 |
+
"step": 890
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 0.6575943008493926,
|
| 668 |
+
"grad_norm": 5.785318374633789,
|
| 669 |
+
"learning_rate": 9.580438927436756e-05,
|
| 670 |
+
"loss": 2.3814,
|
| 671 |
+
"step": 900
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.6649009041921636,
|
| 675 |
+
"grad_norm": 8.992704391479492,
|
| 676 |
+
"learning_rate": 9.563218038753246e-05,
|
| 677 |
+
"loss": 2.5302,
|
| 678 |
+
"step": 910
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.6722075075349347,
|
| 682 |
+
"grad_norm": 8.763484954833984,
|
| 683 |
+
"learning_rate": 9.545666925234873e-05,
|
| 684 |
+
"loss": 2.5443,
|
| 685 |
+
"step": 920
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6795141108777057,
|
| 689 |
+
"grad_norm": 4.8247456550598145,
|
| 690 |
+
"learning_rate": 9.52778685699705e-05,
|
| 691 |
+
"loss": 2.269,
|
| 692 |
+
"step": 930
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.6868207142204767,
|
| 696 |
+
"grad_norm": 4.4191131591796875,
|
| 697 |
+
"learning_rate": 9.509579127960543e-05,
|
| 698 |
+
"loss": 2.0768,
|
| 699 |
+
"step": 940
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.6941273175632477,
|
| 703 |
+
"grad_norm": 3.7724087238311768,
|
| 704 |
+
"learning_rate": 9.491045055757836e-05,
|
| 705 |
+
"loss": 2.2629,
|
| 706 |
+
"step": 950
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.7014339209060189,
|
| 710 |
+
"grad_norm": 4.011133193969727,
|
| 711 |
+
"learning_rate": 9.472185981637775e-05,
|
| 712 |
+
"loss": 2.3676,
|
| 713 |
+
"step": 960
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.7087405242487899,
|
| 717 |
+
"grad_norm": 3.6976683139801025,
|
| 718 |
+
"learning_rate": 9.45300327036851e-05,
|
| 719 |
+
"loss": 2.3744,
|
| 720 |
+
"step": 970
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.7160471275915609,
|
| 724 |
+
"grad_norm": 4.968256950378418,
|
| 725 |
+
"learning_rate": 9.433498310138728e-05,
|
| 726 |
+
"loss": 2.1978,
|
| 727 |
+
"step": 980
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.7233537309343319,
|
| 731 |
+
"grad_norm": 32.12665939331055,
|
| 732 |
+
"learning_rate": 9.413672512457197e-05,
|
| 733 |
+
"loss": 2.1205,
|
| 734 |
+
"step": 990
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.730660334277103,
|
| 738 |
+
"grad_norm": 3.214895248413086,
|
| 739 |
+
"learning_rate": 9.393527312050618e-05,
|
| 740 |
+
"loss": 2.3232,
|
| 741 |
+
"step": 1000
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.730660334277103,
|
| 745 |
+
"eval_loss": 2.3093771934509277,
|
| 746 |
+
"eval_runtime": 109.2454,
|
| 747 |
+
"eval_samples_per_second": 11.14,
|
| 748 |
+
"eval_steps_per_second": 11.14,
|
| 749 |
+
"step": 1000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.737966937619874,
|
| 753 |
+
"grad_norm": 5.2641072273254395,
|
| 754 |
+
"learning_rate": 9.373064166759803e-05,
|
| 755 |
+
"loss": 2.2691,
|
| 756 |
+
"step": 1010
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 0.745273540962645,
|
| 760 |
+
"grad_norm": 4.43721866607666,
|
| 761 |
+
"learning_rate": 9.352284557434166e-05,
|
| 762 |
+
"loss": 2.3682,
|
| 763 |
+
"step": 1020
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 0.752580144305416,
|
| 767 |
+
"grad_norm": 3.769683599472046,
|
| 768 |
+
"learning_rate": 9.331189987824569e-05,
|
| 769 |
+
"loss": 2.2732,
|
| 770 |
+
"step": 1030
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.7598867476481871,
|
| 774 |
+
"grad_norm": 5.2646002769470215,
|
| 775 |
+
"learning_rate": 9.309781984474497e-05,
|
| 776 |
+
"loss": 2.1909,
|
| 777 |
+
"step": 1040
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.7671933509909581,
|
| 781 |
+
"grad_norm": 4.442707538604736,
|
| 782 |
+
"learning_rate": 9.288062096609588e-05,
|
| 783 |
+
"loss": 2.2229,
|
| 784 |
+
"step": 1050
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 0.7744999543337291,
|
| 788 |
+
"grad_norm": 7.33398962020874,
|
| 789 |
+
"learning_rate": 9.266031896025516e-05,
|
| 790 |
+
"loss": 2.3366,
|
| 791 |
+
"step": 1060
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.7818065576765001,
|
| 795 |
+
"grad_norm": 4.366443634033203,
|
| 796 |
+
"learning_rate": 9.243692976974254e-05,
|
| 797 |
+
"loss": 2.0555,
|
| 798 |
+
"step": 1070
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.7891131610192712,
|
| 802 |
+
"grad_norm": 6.721280097961426,
|
| 803 |
+
"learning_rate": 9.221046956048696e-05,
|
| 804 |
+
"loss": 2.2303,
|
| 805 |
+
"step": 1080
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.7964197643620422,
|
| 809 |
+
"grad_norm": 4.3517746925354,
|
| 810 |
+
"learning_rate": 9.198095472065668e-05,
|
| 811 |
+
"loss": 2.3249,
|
| 812 |
+
"step": 1090
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.8037263677048132,
|
| 816 |
+
"grad_norm": 5.425954818725586,
|
| 817 |
+
"learning_rate": 9.174840185947345e-05,
|
| 818 |
+
"loss": 2.2346,
|
| 819 |
+
"step": 1100
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.8110329710475842,
|
| 823 |
+
"grad_norm": 4.331856727600098,
|
| 824 |
+
"learning_rate": 9.151282780601039e-05,
|
| 825 |
+
"loss": 2.4686,
|
| 826 |
+
"step": 1110
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.8183395743903553,
|
| 830 |
+
"grad_norm": 7.6655707359313965,
|
| 831 |
+
"learning_rate": 9.127424960797424e-05,
|
| 832 |
+
"loss": 2.3503,
|
| 833 |
+
"step": 1120
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.8256461777331263,
|
| 837 |
+
"grad_norm": 3.6036481857299805,
|
| 838 |
+
"learning_rate": 9.103268453047165e-05,
|
| 839 |
+
"loss": 2.3279,
|
| 840 |
+
"step": 1130
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 0.8329527810758973,
|
| 844 |
+
"grad_norm": 7.087502479553223,
|
| 845 |
+
"learning_rate": 9.078815005475974e-05,
|
| 846 |
+
"loss": 2.4316,
|
| 847 |
+
"step": 1140
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 0.8402593844186683,
|
| 851 |
+
"grad_norm": 3.372032880783081,
|
| 852 |
+
"learning_rate": 9.054066387698104e-05,
|
| 853 |
+
"loss": 2.3761,
|
| 854 |
+
"step": 1150
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.8475659877614394,
|
| 858 |
+
"grad_norm": 5.309089660644531,
|
| 859 |
+
"learning_rate": 9.02902439068829e-05,
|
| 860 |
+
"loss": 2.4221,
|
| 861 |
+
"step": 1160
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.8548725911042104,
|
| 865 |
+
"grad_norm": 4.797391414642334,
|
| 866 |
+
"learning_rate": 9.003690826652143e-05,
|
| 867 |
+
"loss": 2.2968,
|
| 868 |
+
"step": 1170
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 0.8621791944469814,
|
| 872 |
+
"grad_norm": 3.347250461578369,
|
| 873 |
+
"learning_rate": 8.978067528895003e-05,
|
| 874 |
+
"loss": 2.2259,
|
| 875 |
+
"step": 1180
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 0.8694857977897524,
|
| 879 |
+
"grad_norm": 2.894286870956421,
|
| 880 |
+
"learning_rate": 8.95215635168927e-05,
|
| 881 |
+
"loss": 2.3369,
|
| 882 |
+
"step": 1190
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.8767924011325235,
|
| 886 |
+
"grad_norm": 5.936110019683838,
|
| 887 |
+
"learning_rate": 8.925959170140218e-05,
|
| 888 |
+
"loss": 2.2603,
|
| 889 |
+
"step": 1200
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.8767924011325235,
|
| 893 |
+
"eval_loss": 2.2486681938171387,
|
| 894 |
+
"eval_runtime": 109.2268,
|
| 895 |
+
"eval_samples_per_second": 11.142,
|
| 896 |
+
"eval_steps_per_second": 11.142,
|
| 897 |
+
"step": 1200
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.8840990044752945,
|
| 901 |
+
"grad_norm": 4.270968437194824,
|
| 902 |
+
"learning_rate": 8.899477880050306e-05,
|
| 903 |
+
"loss": 2.3473,
|
| 904 |
+
"step": 1210
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 0.8914056078180655,
|
| 908 |
+
"grad_norm": 5.154277324676514,
|
| 909 |
+
"learning_rate": 8.872714397781965e-05,
|
| 910 |
+
"loss": 2.3085,
|
| 911 |
+
"step": 1220
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.8987122111608367,
|
| 915 |
+
"grad_norm": 4.050198078155518,
|
| 916 |
+
"learning_rate": 8.84567066011894e-05,
|
| 917 |
+
"loss": 2.2946,
|
| 918 |
+
"step": 1230
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.9060188145036077,
|
| 922 |
+
"grad_norm": 11.016803741455078,
|
| 923 |
+
"learning_rate": 8.818348624126122e-05,
|
| 924 |
+
"loss": 2.1233,
|
| 925 |
+
"step": 1240
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.9133254178463787,
|
| 929 |
+
"grad_norm": 4.9638447761535645,
|
| 930 |
+
"learning_rate": 8.790750267007918e-05,
|
| 931 |
+
"loss": 2.1703,
|
| 932 |
+
"step": 1250
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.9206320211891497,
|
| 936 |
+
"grad_norm": 5.055572032928467,
|
| 937 |
+
"learning_rate": 8.762877585965172e-05,
|
| 938 |
+
"loss": 2.3752,
|
| 939 |
+
"step": 1260
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.9279386245319208,
|
| 943 |
+
"grad_norm": 5.373875617980957,
|
| 944 |
+
"learning_rate": 8.734732598050637e-05,
|
| 945 |
+
"loss": 2.2983,
|
| 946 |
+
"step": 1270
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 0.9352452278746918,
|
| 950 |
+
"grad_norm": 4.6786723136901855,
|
| 951 |
+
"learning_rate": 8.706317340022997e-05,
|
| 952 |
+
"loss": 2.2051,
|
| 953 |
+
"step": 1280
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.9425518312174628,
|
| 957 |
+
"grad_norm": 3.9016830921173096,
|
| 958 |
+
"learning_rate": 8.677633868199487e-05,
|
| 959 |
+
"loss": 2.1745,
|
| 960 |
+
"step": 1290
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.9498584345602338,
|
| 964 |
+
"grad_norm": 3.027470350265503,
|
| 965 |
+
"learning_rate": 8.648684258307076e-05,
|
| 966 |
+
"loss": 2.2724,
|
| 967 |
+
"step": 1300
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 0.9571650379030049,
|
| 971 |
+
"grad_norm": 4.489301681518555,
|
| 972 |
+
"learning_rate": 8.619470605332253e-05,
|
| 973 |
+
"loss": 2.2487,
|
| 974 |
+
"step": 1310
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.9644716412457759,
|
| 978 |
+
"grad_norm": 4.5948686599731445,
|
| 979 |
+
"learning_rate": 8.589995023369429e-05,
|
| 980 |
+
"loss": 2.2639,
|
| 981 |
+
"step": 1320
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.9717782445885469,
|
| 985 |
+
"grad_norm": 5.749096393585205,
|
| 986 |
+
"learning_rate": 8.560259645467928e-05,
|
| 987 |
+
"loss": 2.2842,
|
| 988 |
+
"step": 1330
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 0.9790848479313179,
|
| 992 |
+
"grad_norm": 3.869464635848999,
|
| 993 |
+
"learning_rate": 8.53026662347765e-05,
|
| 994 |
+
"loss": 2.3613,
|
| 995 |
+
"step": 1340
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.986391451274089,
|
| 999 |
+
"grad_norm": 6.761131763458252,
|
| 1000 |
+
"learning_rate": 8.500018127893329e-05,
|
| 1001 |
+
"loss": 2.1473,
|
| 1002 |
+
"step": 1350
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 0.99369805461686,
|
| 1006 |
+
"grad_norm": 3.230255365371704,
|
| 1007 |
+
"learning_rate": 8.469516347697473e-05,
|
| 1008 |
+
"loss": 2.2302,
|
| 1009 |
+
"step": 1360
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 1.001004657959631,
|
| 1013 |
+
"grad_norm": 4.553093433380127,
|
| 1014 |
+
"learning_rate": 8.438763490201946e-05,
|
| 1015 |
+
"loss": 2.1864,
|
| 1016 |
+
"step": 1370
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.0083112613024021,
|
| 1020 |
+
"grad_norm": 4.498405456542969,
|
| 1021 |
+
"learning_rate": 8.407761780888244e-05,
|
| 1022 |
+
"loss": 2.209,
|
| 1023 |
+
"step": 1380
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 1.015617864645173,
|
| 1027 |
+
"grad_norm": 5.025854110717773,
|
| 1028 |
+
"learning_rate": 8.37651346324643e-05,
|
| 1029 |
+
"loss": 2.2477,
|
| 1030 |
+
"step": 1390
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.0229244679879441,
|
| 1034 |
+
"grad_norm": 4.791236400604248,
|
| 1035 |
+
"learning_rate": 8.345020798612791e-05,
|
| 1036 |
+
"loss": 2.1007,
|
| 1037 |
+
"step": 1400
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 1.0229244679879441,
|
| 1041 |
+
"eval_loss": 2.2063653469085693,
|
| 1042 |
+
"eval_runtime": 108.5908,
|
| 1043 |
+
"eval_samples_per_second": 11.207,
|
| 1044 |
+
"eval_steps_per_second": 11.207,
|
| 1045 |
+
"step": 1400
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.0302310713307152,
|
| 1049 |
+
"grad_norm": 4.35936164855957,
|
| 1050 |
+
"learning_rate": 8.313286066006187e-05,
|
| 1051 |
+
"loss": 2.1783,
|
| 1052 |
+
"step": 1410
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.0375376746734861,
|
| 1056 |
+
"grad_norm": 4.893729209899902,
|
| 1057 |
+
"learning_rate": 8.28131156196313e-05,
|
| 1058 |
+
"loss": 2.0583,
|
| 1059 |
+
"step": 1420
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.0448442780162572,
|
| 1063 |
+
"grad_norm": 9.001852989196777,
|
| 1064 |
+
"learning_rate": 8.249099600371591e-05,
|
| 1065 |
+
"loss": 2.0463,
|
| 1066 |
+
"step": 1430
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.0521508813590281,
|
| 1070 |
+
"grad_norm": 4.704784393310547,
|
| 1071 |
+
"learning_rate": 8.216652512303543e-05,
|
| 1072 |
+
"loss": 2.255,
|
| 1073 |
+
"step": 1440
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.0594574847017992,
|
| 1077 |
+
"grad_norm": 4.346590995788574,
|
| 1078 |
+
"learning_rate": 8.183972645846283e-05,
|
| 1079 |
+
"loss": 2.2898,
|
| 1080 |
+
"step": 1450
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.0667640880445703,
|
| 1084 |
+
"grad_norm": 6.388704299926758,
|
| 1085 |
+
"learning_rate": 8.1510623659325e-05,
|
| 1086 |
+
"loss": 2.0351,
|
| 1087 |
+
"step": 1460
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.0740706913873412,
|
| 1091 |
+
"grad_norm": 5.978973388671875,
|
| 1092 |
+
"learning_rate": 8.117924054169133e-05,
|
| 1093 |
+
"loss": 2.091,
|
| 1094 |
+
"step": 1470
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.0813772947301123,
|
| 1098 |
+
"grad_norm": 3.6497621536254883,
|
| 1099 |
+
"learning_rate": 8.084560108665024e-05,
|
| 1100 |
+
"loss": 2.0603,
|
| 1101 |
+
"step": 1480
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.0886838980728835,
|
| 1105 |
+
"grad_norm": 5.628828048706055,
|
| 1106 |
+
"learning_rate": 8.050972943857375e-05,
|
| 1107 |
+
"loss": 2.1513,
|
| 1108 |
+
"step": 1490
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.0959905014156544,
|
| 1112 |
+
"grad_norm": 15.277828216552734,
|
| 1113 |
+
"learning_rate": 8.017164990337026e-05,
|
| 1114 |
+
"loss": 2.3273,
|
| 1115 |
+
"step": 1500
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.1032971047584255,
|
| 1119 |
+
"grad_norm": 7.737220764160156,
|
| 1120 |
+
"learning_rate": 7.983138694672552e-05,
|
| 1121 |
+
"loss": 2.2664,
|
| 1122 |
+
"step": 1510
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.1106037081011966,
|
| 1126 |
+
"grad_norm": 6.418485164642334,
|
| 1127 |
+
"learning_rate": 7.948896519233225e-05,
|
| 1128 |
+
"loss": 2.1581,
|
| 1129 |
+
"step": 1520
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.1179103114439675,
|
| 1133 |
+
"grad_norm": 3.803318738937378,
|
| 1134 |
+
"learning_rate": 7.914440942010807e-05,
|
| 1135 |
+
"loss": 2.2859,
|
| 1136 |
+
"step": 1530
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.1252169147867386,
|
| 1140 |
+
"grad_norm": 3.083667278289795,
|
| 1141 |
+
"learning_rate": 7.879774456440243e-05,
|
| 1142 |
+
"loss": 2.1394,
|
| 1143 |
+
"step": 1540
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.1325235181295095,
|
| 1147 |
+
"grad_norm": 13.434645652770996,
|
| 1148 |
+
"learning_rate": 7.844899571219202e-05,
|
| 1149 |
+
"loss": 2.0644,
|
| 1150 |
+
"step": 1550
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.1398301214722806,
|
| 1154 |
+
"grad_norm": 5.204081058502197,
|
| 1155 |
+
"learning_rate": 7.809818810126545e-05,
|
| 1156 |
+
"loss": 2.055,
|
| 1157 |
+
"step": 1560
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.1471367248150517,
|
| 1161 |
+
"grad_norm": 4.114535808563232,
|
| 1162 |
+
"learning_rate": 7.774534711839677e-05,
|
| 1163 |
+
"loss": 2.0104,
|
| 1164 |
+
"step": 1570
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.1544433281578226,
|
| 1168 |
+
"grad_norm": 5.424691677093506,
|
| 1169 |
+
"learning_rate": 7.73904982975084e-05,
|
| 1170 |
+
"loss": 2.0453,
|
| 1171 |
+
"step": 1580
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.1617499315005937,
|
| 1175 |
+
"grad_norm": 4.465676784515381,
|
| 1176 |
+
"learning_rate": 7.703366731782327e-05,
|
| 1177 |
+
"loss": 2.2777,
|
| 1178 |
+
"step": 1590
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.1690565348433646,
|
| 1182 |
+
"grad_norm": 3.521860361099243,
|
| 1183 |
+
"learning_rate": 7.667488000200649e-05,
|
| 1184 |
+
"loss": 2.0635,
|
| 1185 |
+
"step": 1600
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.1690565348433646,
|
| 1189 |
+
"eval_loss": 2.1665337085723877,
|
| 1190 |
+
"eval_runtime": 109.3334,
|
| 1191 |
+
"eval_samples_per_second": 11.131,
|
| 1192 |
+
"eval_steps_per_second": 11.131,
|
| 1193 |
+
"step": 1600
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.1763631381861357,
|
| 1197 |
+
"grad_norm": 5.549830913543701,
|
| 1198 |
+
"learning_rate": 7.631416231429672e-05,
|
| 1199 |
+
"loss": 2.0994,
|
| 1200 |
+
"step": 1610
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.1836697415289068,
|
| 1204 |
+
"grad_norm": 6.577401638031006,
|
| 1205 |
+
"learning_rate": 7.595154035862715e-05,
|
| 1206 |
+
"loss": 2.0379,
|
| 1207 |
+
"step": 1620
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.1909763448716777,
|
| 1211 |
+
"grad_norm": 6.858834743499756,
|
| 1212 |
+
"learning_rate": 7.558704037673648e-05,
|
| 1213 |
+
"loss": 2.1925,
|
| 1214 |
+
"step": 1630
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.1982829482144488,
|
| 1218 |
+
"grad_norm": 4.788407325744629,
|
| 1219 |
+
"learning_rate": 7.522068874626988e-05,
|
| 1220 |
+
"loss": 2.1162,
|
| 1221 |
+
"step": 1640
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.20558955155722,
|
| 1225 |
+
"grad_norm": 3.5293450355529785,
|
| 1226 |
+
"learning_rate": 7.48525119788702e-05,
|
| 1227 |
+
"loss": 2.1279,
|
| 1228 |
+
"step": 1650
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.2128961548999908,
|
| 1232 |
+
"grad_norm": 4.138194561004639,
|
| 1233 |
+
"learning_rate": 7.448253671825927e-05,
|
| 1234 |
+
"loss": 2.1242,
|
| 1235 |
+
"step": 1660
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.220202758242762,
|
| 1239 |
+
"grad_norm": 3.9674646854400635,
|
| 1240 |
+
"learning_rate": 7.411078973830987e-05,
|
| 1241 |
+
"loss": 2.1451,
|
| 1242 |
+
"step": 1670
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.227509361585533,
|
| 1246 |
+
"grad_norm": 10.673015594482422,
|
| 1247 |
+
"learning_rate": 7.373729794110826e-05,
|
| 1248 |
+
"loss": 2.0227,
|
| 1249 |
+
"step": 1680
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.234815964928304,
|
| 1253 |
+
"grad_norm": 5.9692511558532715,
|
| 1254 |
+
"learning_rate": 7.33620883550072e-05,
|
| 1255 |
+
"loss": 2.191,
|
| 1256 |
+
"step": 1690
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.242122568271075,
|
| 1260 |
+
"grad_norm": 8.75203800201416,
|
| 1261 |
+
"learning_rate": 7.298518813267015e-05,
|
| 1262 |
+
"loss": 1.9689,
|
| 1263 |
+
"step": 1700
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.249429171613846,
|
| 1267 |
+
"grad_norm": 6.571184158325195,
|
| 1268 |
+
"learning_rate": 7.260662454910621e-05,
|
| 1269 |
+
"loss": 2.0869,
|
| 1270 |
+
"step": 1710
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.256735774956617,
|
| 1274 |
+
"grad_norm": 6.252017498016357,
|
| 1275 |
+
"learning_rate": 7.222642499969646e-05,
|
| 1276 |
+
"loss": 2.0596,
|
| 1277 |
+
"step": 1720
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.2640423782993881,
|
| 1281 |
+
"grad_norm": 7.060279369354248,
|
| 1282 |
+
"learning_rate": 7.184461699821126e-05,
|
| 1283 |
+
"loss": 2.211,
|
| 1284 |
+
"step": 1730
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.271348981642159,
|
| 1288 |
+
"grad_norm": 5.095629692077637,
|
| 1289 |
+
"learning_rate": 7.14612281748193e-05,
|
| 1290 |
+
"loss": 2.0639,
|
| 1291 |
+
"step": 1740
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.2786555849849301,
|
| 1295 |
+
"grad_norm": 3.8237876892089844,
|
| 1296 |
+
"learning_rate": 7.107628627408813e-05,
|
| 1297 |
+
"loss": 2.0824,
|
| 1298 |
+
"step": 1750
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.285962188327701,
|
| 1302 |
+
"grad_norm": 4.18397331237793,
|
| 1303 |
+
"learning_rate": 7.068981915297626e-05,
|
| 1304 |
+
"loss": 2.0253,
|
| 1305 |
+
"step": 1760
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.2932687916704722,
|
| 1309 |
+
"grad_norm": 4.53864049911499,
|
| 1310 |
+
"learning_rate": 7.030185477881726e-05,
|
| 1311 |
+
"loss": 2.1168,
|
| 1312 |
+
"step": 1770
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.3005753950132433,
|
| 1316 |
+
"grad_norm": 6.684192657470703,
|
| 1317 |
+
"learning_rate": 6.991242122729597e-05,
|
| 1318 |
+
"loss": 1.9231,
|
| 1319 |
+
"step": 1780
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.3078819983560144,
|
| 1323 |
+
"grad_norm": 6.962904453277588,
|
| 1324 |
+
"learning_rate": 6.952154668041666e-05,
|
| 1325 |
+
"loss": 1.9549,
|
| 1326 |
+
"step": 1790
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.3151886016987853,
|
| 1330 |
+
"grad_norm": 8.408102035522461,
|
| 1331 |
+
"learning_rate": 6.91292594244636e-05,
|
| 1332 |
+
"loss": 2.1212,
|
| 1333 |
+
"step": 1800
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.3151886016987853,
|
| 1337 |
+
"eval_loss": 2.135313034057617,
|
| 1338 |
+
"eval_runtime": 108.3855,
|
| 1339 |
+
"eval_samples_per_second": 11.228,
|
| 1340 |
+
"eval_steps_per_second": 11.228,
|
| 1341 |
+
"step": 1800
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.3224952050415564,
|
| 1345 |
+
"grad_norm": 4.266064167022705,
|
| 1346 |
+
"learning_rate": 6.873558784795412e-05,
|
| 1347 |
+
"loss": 1.9563,
|
| 1348 |
+
"step": 1810
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 1.3298018083843273,
|
| 1352 |
+
"grad_norm": 5.321630001068115,
|
| 1353 |
+
"learning_rate": 6.834056043958419e-05,
|
| 1354 |
+
"loss": 2.3073,
|
| 1355 |
+
"step": 1820
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.3371084117270984,
|
| 1359 |
+
"grad_norm": 5.667231559753418,
|
| 1360 |
+
"learning_rate": 6.794420578616679e-05,
|
| 1361 |
+
"loss": 2.0931,
|
| 1362 |
+
"step": 1830
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.3444150150698695,
|
| 1366 |
+
"grad_norm": 5.461520671844482,
|
| 1367 |
+
"learning_rate": 6.754655257056322e-05,
|
| 1368 |
+
"loss": 2.0288,
|
| 1369 |
+
"step": 1840
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 1.3517216184126404,
|
| 1373 |
+
"grad_norm": 3.88820743560791,
|
| 1374 |
+
"learning_rate": 6.71476295696073e-05,
|
| 1375 |
+
"loss": 2.0259,
|
| 1376 |
+
"step": 1850
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.3590282217554115,
|
| 1380 |
+
"grad_norm": 5.630788803100586,
|
| 1381 |
+
"learning_rate": 6.674746565202309e-05,
|
| 1382 |
+
"loss": 1.9281,
|
| 1383 |
+
"step": 1860
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 1.3663348250981824,
|
| 1387 |
+
"grad_norm": 3.890331745147705,
|
| 1388 |
+
"learning_rate": 6.634608977633555e-05,
|
| 1389 |
+
"loss": 2.1578,
|
| 1390 |
+
"step": 1870
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.3736414284409535,
|
| 1394 |
+
"grad_norm": 3.679614782333374,
|
| 1395 |
+
"learning_rate": 6.594353098877503e-05,
|
| 1396 |
+
"loss": 2.1456,
|
| 1397 |
+
"step": 1880
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 1.3809480317837246,
|
| 1401 |
+
"grad_norm": 5.798458099365234,
|
| 1402 |
+
"learning_rate": 6.553981842117526e-05,
|
| 1403 |
+
"loss": 2.0121,
|
| 1404 |
+
"step": 1890
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 1.3882546351264955,
|
| 1408 |
+
"grad_norm": 4.401270389556885,
|
| 1409 |
+
"learning_rate": 6.513498128886515e-05,
|
| 1410 |
+
"loss": 2.0316,
|
| 1411 |
+
"step": 1900
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 1.3955612384692666,
|
| 1415 |
+
"grad_norm": 6.3201823234558105,
|
| 1416 |
+
"learning_rate": 6.472904888855463e-05,
|
| 1417 |
+
"loss": 2.1175,
|
| 1418 |
+
"step": 1910
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.4028678418120375,
|
| 1422 |
+
"grad_norm": 5.530188083648682,
|
| 1423 |
+
"learning_rate": 6.432205059621449e-05,
|
| 1424 |
+
"loss": 2.0955,
|
| 1425 |
+
"step": 1920
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 1.4101744451548086,
|
| 1429 |
+
"grad_norm": 5.705748558044434,
|
| 1430 |
+
"learning_rate": 6.391401586495059e-05,
|
| 1431 |
+
"loss": 2.0269,
|
| 1432 |
+
"step": 1930
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 1.4174810484975797,
|
| 1436 |
+
"grad_norm": 5.124378204345703,
|
| 1437 |
+
"learning_rate": 6.350497422287236e-05,
|
| 1438 |
+
"loss": 2.2386,
|
| 1439 |
+
"step": 1940
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.4247876518403508,
|
| 1443 |
+
"grad_norm": 7.351918697357178,
|
| 1444 |
+
"learning_rate": 6.309495527095606e-05,
|
| 1445 |
+
"loss": 2.1977,
|
| 1446 |
+
"step": 1950
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 1.4320942551831217,
|
| 1450 |
+
"grad_norm": 5.076743125915527,
|
| 1451 |
+
"learning_rate": 6.268398868090255e-05,
|
| 1452 |
+
"loss": 2.117,
|
| 1453 |
+
"step": 1960
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 1.4394008585258928,
|
| 1457 |
+
"grad_norm": 3.8619604110717773,
|
| 1458 |
+
"learning_rate": 6.227210419299014e-05,
|
| 1459 |
+
"loss": 1.9848,
|
| 1460 |
+
"step": 1970
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 1.4467074618686637,
|
| 1464 |
+
"grad_norm": 3.990159034729004,
|
| 1465 |
+
"learning_rate": 6.185933161392228e-05,
|
| 1466 |
+
"loss": 2.0853,
|
| 1467 |
+
"step": 1980
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.4540140652114348,
|
| 1471 |
+
"grad_norm": 6.388422966003418,
|
| 1472 |
+
"learning_rate": 6.144570081467066e-05,
|
| 1473 |
+
"loss": 2.0883,
|
| 1474 |
+
"step": 1990
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 1.461320668554206,
|
| 1478 |
+
"grad_norm": 5.564253807067871,
|
| 1479 |
+
"learning_rate": 6.103124172831346e-05,
|
| 1480 |
+
"loss": 2.038,
|
| 1481 |
+
"step": 2000
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.461320668554206,
|
| 1485 |
+
"eval_loss": 2.0880391597747803,
|
| 1486 |
+
"eval_runtime": 109.0271,
|
| 1487 |
+
"eval_samples_per_second": 11.162,
|
| 1488 |
+
"eval_steps_per_second": 11.162,
|
| 1489 |
+
"step": 2000
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 1.4686272718969768,
|
| 1493 |
+
"grad_norm": 6.205707550048828,
|
| 1494 |
+
"learning_rate": 6.061598434786926e-05,
|
| 1495 |
+
"loss": 1.9301,
|
| 1496 |
+
"step": 2010
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 1.475933875239748,
|
| 1500 |
+
"grad_norm": 7.6283111572265625,
|
| 1501 |
+
"learning_rate": 6.019995872412649e-05,
|
| 1502 |
+
"loss": 2.0155,
|
| 1503 |
+
"step": 2020
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 1.4832404785825188,
|
| 1507 |
+
"grad_norm": 6.749846458435059,
|
| 1508 |
+
"learning_rate": 5.9783194963468784e-05,
|
| 1509 |
+
"loss": 1.9461,
|
| 1510 |
+
"step": 2030
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 1.49054708192529,
|
| 1514 |
+
"grad_norm": 5.651824951171875,
|
| 1515 |
+
"learning_rate": 5.936572322569629e-05,
|
| 1516 |
+
"loss": 2.0335,
|
| 1517 |
+
"step": 2040
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 1.497853685268061,
|
| 1521 |
+
"grad_norm": 11.750208854675293,
|
| 1522 |
+
"learning_rate": 5.894757372184309e-05,
|
| 1523 |
+
"loss": 2.0556,
|
| 1524 |
+
"step": 2050
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 1.5051602886108322,
|
| 1528 |
+
"grad_norm": 5.419318675994873,
|
| 1529 |
+
"learning_rate": 5.852877671199091e-05,
|
| 1530 |
+
"loss": 1.9466,
|
| 1531 |
+
"step": 2060
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 1.512466891953603,
|
| 1535 |
+
"grad_norm": 4.80405330657959,
|
| 1536 |
+
"learning_rate": 5.810936250307935e-05,
|
| 1537 |
+
"loss": 2.1046,
|
| 1538 |
+
"step": 2070
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 1.519773495296374,
|
| 1542 |
+
"grad_norm": 11.513900756835938,
|
| 1543 |
+
"learning_rate": 5.768936144671261e-05,
|
| 1544 |
+
"loss": 1.8582,
|
| 1545 |
+
"step": 2080
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 1.527080098639145,
|
| 1549 |
+
"grad_norm": 5.375672817230225,
|
| 1550 |
+
"learning_rate": 5.7268803936963124e-05,
|
| 1551 |
+
"loss": 2.0872,
|
| 1552 |
+
"step": 2090
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 1.5343867019819162,
|
| 1556 |
+
"grad_norm": 4.7279863357543945,
|
| 1557 |
+
"learning_rate": 5.6847720408171946e-05,
|
| 1558 |
+
"loss": 2.0174,
|
| 1559 |
+
"step": 2100
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 1.5416933053246873,
|
| 1563 |
+
"grad_norm": 4.411227226257324,
|
| 1564 |
+
"learning_rate": 5.642614133274641e-05,
|
| 1565 |
+
"loss": 2.0424,
|
| 1566 |
+
"step": 2110
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 1.5489999086674582,
|
| 1570 |
+
"grad_norm": 7.084948539733887,
|
| 1571 |
+
"learning_rate": 5.600409721895488e-05,
|
| 1572 |
+
"loss": 2.2553,
|
| 1573 |
+
"step": 2120
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 1.5563065120102293,
|
| 1577 |
+
"grad_norm": 5.153254508972168,
|
| 1578 |
+
"learning_rate": 5.558161860871899e-05,
|
| 1579 |
+
"loss": 1.9618,
|
| 1580 |
+
"step": 2130
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 1.5636131153530002,
|
| 1584 |
+
"grad_norm": 8.986722946166992,
|
| 1585 |
+
"learning_rate": 5.515873607540346e-05,
|
| 1586 |
+
"loss": 2.0533,
|
| 1587 |
+
"step": 2140
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 1.5709197186957713,
|
| 1591 |
+
"grad_norm": 6.175150394439697,
|
| 1592 |
+
"learning_rate": 5.473548022160354e-05,
|
| 1593 |
+
"loss": 2.0438,
|
| 1594 |
+
"step": 2150
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 1.5782263220385424,
|
| 1598 |
+
"grad_norm": 8.26689624786377,
|
| 1599 |
+
"learning_rate": 5.431188167693044e-05,
|
| 1600 |
+
"loss": 2.0011,
|
| 1601 |
+
"step": 2160
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 1.5855329253813135,
|
| 1605 |
+
"grad_norm": 4.056985855102539,
|
| 1606 |
+
"learning_rate": 5.388797109579479e-05,
|
| 1607 |
+
"loss": 1.9823,
|
| 1608 |
+
"step": 2170
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 1.5928395287240844,
|
| 1612 |
+
"grad_norm": 5.051248073577881,
|
| 1613 |
+
"learning_rate": 5.346377915518821e-05,
|
| 1614 |
+
"loss": 1.9149,
|
| 1615 |
+
"step": 2180
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 1.6001461320668553,
|
| 1619 |
+
"grad_norm": 5.093779563903809,
|
| 1620 |
+
"learning_rate": 5.3039336552463414e-05,
|
| 1621 |
+
"loss": 2.0657,
|
| 1622 |
+
"step": 2190
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 1.6074527354096264,
|
| 1626 |
+
"grad_norm": 9.080710411071777,
|
| 1627 |
+
"learning_rate": 5.261467400311266e-05,
|
| 1628 |
+
"loss": 1.9656,
|
| 1629 |
+
"step": 2200
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.6074527354096264,
|
| 1633 |
+
"eval_loss": 2.061772108078003,
|
| 1634 |
+
"eval_runtime": 108.1317,
|
| 1635 |
+
"eval_samples_per_second": 11.255,
|
| 1636 |
+
"eval_steps_per_second": 11.255,
|
| 1637 |
+
"step": 2200
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 1.6147593387523975,
|
| 1641 |
+
"grad_norm": 5.32938814163208,
|
| 1642 |
+
"learning_rate": 5.2189822238545017e-05,
|
| 1643 |
+
"loss": 1.9647,
|
| 1644 |
+
"step": 2210
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 1.6220659420951686,
|
| 1648 |
+
"grad_norm": 7.389804840087891,
|
| 1649 |
+
"learning_rate": 5.176481200386245e-05,
|
| 1650 |
+
"loss": 1.904,
|
| 1651 |
+
"step": 2220
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 1.6293725454379395,
|
| 1655 |
+
"grad_norm": 4.834039688110352,
|
| 1656 |
+
"learning_rate": 5.1339674055634826e-05,
|
| 1657 |
+
"loss": 1.9067,
|
| 1658 |
+
"step": 2230
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 1.6366791487807104,
|
| 1662 |
+
"grad_norm": 5.643000602722168,
|
| 1663 |
+
"learning_rate": 5.0914439159674244e-05,
|
| 1664 |
+
"loss": 2.2128,
|
| 1665 |
+
"step": 2240
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 1.6439857521234815,
|
| 1669 |
+
"grad_norm": 4.886800765991211,
|
| 1670 |
+
"learning_rate": 5.048913808880861e-05,
|
| 1671 |
+
"loss": 1.9497,
|
| 1672 |
+
"step": 2250
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 1.6512923554662526,
|
| 1676 |
+
"grad_norm": 12.440683364868164,
|
| 1677 |
+
"learning_rate": 5.006380162065465e-05,
|
| 1678 |
+
"loss": 1.9488,
|
| 1679 |
+
"step": 2260
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 1.6585989588090237,
|
| 1683 |
+
"grad_norm": 7.266822814941406,
|
| 1684 |
+
"learning_rate": 4.963846053539071e-05,
|
| 1685 |
+
"loss": 2.1738,
|
| 1686 |
+
"step": 2270
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 1.6659055621517946,
|
| 1690 |
+
"grad_norm": 9.061182975769043,
|
| 1691 |
+
"learning_rate": 4.9213145613529194e-05,
|
| 1692 |
+
"loss": 2.0644,
|
| 1693 |
+
"step": 2280
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 1.6732121654945658,
|
| 1697 |
+
"grad_norm": 4.660227298736572,
|
| 1698 |
+
"learning_rate": 4.878788763368921e-05,
|
| 1699 |
+
"loss": 1.8194,
|
| 1700 |
+
"step": 2290
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 1.6805187688373366,
|
| 1704 |
+
"grad_norm": 6.212920665740967,
|
| 1705 |
+
"learning_rate": 4.836271737036916e-05,
|
| 1706 |
+
"loss": 2.1629,
|
| 1707 |
+
"step": 2300
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 1.6878253721801078,
|
| 1711 |
+
"grad_norm": 5.760697841644287,
|
| 1712 |
+
"learning_rate": 4.7937665591719664e-05,
|
| 1713 |
+
"loss": 1.9379,
|
| 1714 |
+
"step": 2310
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 1.6951319755228789,
|
| 1718 |
+
"grad_norm": 6.021162033081055,
|
| 1719 |
+
"learning_rate": 4.7512763057317014e-05,
|
| 1720 |
+
"loss": 1.9756,
|
| 1721 |
+
"step": 2320
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 1.70243857886565,
|
| 1725 |
+
"grad_norm": 4.001003265380859,
|
| 1726 |
+
"learning_rate": 4.70880405159372e-05,
|
| 1727 |
+
"loss": 1.9218,
|
| 1728 |
+
"step": 2330
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 1.7097451822084209,
|
| 1732 |
+
"grad_norm": 6.9795002937316895,
|
| 1733 |
+
"learning_rate": 4.666352870333072e-05,
|
| 1734 |
+
"loss": 2.1045,
|
| 1735 |
+
"step": 2340
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.7170517855511918,
|
| 1739 |
+
"grad_norm": 5.518786907196045,
|
| 1740 |
+
"learning_rate": 4.623925833999832e-05,
|
| 1741 |
+
"loss": 2.0515,
|
| 1742 |
+
"step": 2350
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 1.7243583888939629,
|
| 1746 |
+
"grad_norm": 3.8339755535125732,
|
| 1747 |
+
"learning_rate": 4.5815260128967894e-05,
|
| 1748 |
+
"loss": 1.9462,
|
| 1749 |
+
"step": 2360
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.731664992236734,
|
| 1753 |
+
"grad_norm": 4.569764137268066,
|
| 1754 |
+
"learning_rate": 4.539156475357257e-05,
|
| 1755 |
+
"loss": 2.0918,
|
| 1756 |
+
"step": 2370
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 1.738971595579505,
|
| 1760 |
+
"grad_norm": 4.515227317810059,
|
| 1761 |
+
"learning_rate": 4.496820287523027e-05,
|
| 1762 |
+
"loss": 2.1785,
|
| 1763 |
+
"step": 2380
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 1.746278198922276,
|
| 1767 |
+
"grad_norm": 4.677797794342041,
|
| 1768 |
+
"learning_rate": 4.454520513122484e-05,
|
| 1769 |
+
"loss": 2.1057,
|
| 1770 |
+
"step": 2390
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 1.7535848022650469,
|
| 1774 |
+
"grad_norm": 8.1099214553833,
|
| 1775 |
+
"learning_rate": 4.412260213248898e-05,
|
| 1776 |
+
"loss": 2.0285,
|
| 1777 |
+
"step": 2400
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 1.7535848022650469,
|
| 1781 |
+
"eval_loss": 2.0392701625823975,
|
| 1782 |
+
"eval_runtime": 108.4121,
|
| 1783 |
+
"eval_samples_per_second": 11.226,
|
| 1784 |
+
"eval_steps_per_second": 11.226,
|
| 1785 |
+
"step": 2400
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 1.760891405607818,
|
| 1789 |
+
"grad_norm": 5.997411727905273,
|
| 1790 |
+
"learning_rate": 4.370042446138897e-05,
|
| 1791 |
+
"loss": 2.0577,
|
| 1792 |
+
"step": 2410
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 1.768198008950589,
|
| 1796 |
+
"grad_norm": 4.918084144592285,
|
| 1797 |
+
"learning_rate": 4.3278702669511506e-05,
|
| 1798 |
+
"loss": 1.9363,
|
| 1799 |
+
"step": 2420
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 1.7755046122933602,
|
| 1803 |
+
"grad_norm": 5.5613932609558105,
|
| 1804 |
+
"learning_rate": 4.285746727545291e-05,
|
| 1805 |
+
"loss": 2.041,
|
| 1806 |
+
"step": 2430
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 1.7828112156361313,
|
| 1810 |
+
"grad_norm": 11.325989723205566,
|
| 1811 |
+
"learning_rate": 4.2436748762610465e-05,
|
| 1812 |
+
"loss": 1.9016,
|
| 1813 |
+
"step": 2440
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 1.7901178189789022,
|
| 1817 |
+
"grad_norm": 5.7209296226501465,
|
| 1818 |
+
"learning_rate": 4.201657757697651e-05,
|
| 1819 |
+
"loss": 2.0692,
|
| 1820 |
+
"step": 2450
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 1.797424422321673,
|
| 1824 |
+
"grad_norm": 7.241997718811035,
|
| 1825 |
+
"learning_rate": 4.159698412493515e-05,
|
| 1826 |
+
"loss": 1.9591,
|
| 1827 |
+
"step": 2460
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 1.8047310256644442,
|
| 1831 |
+
"grad_norm": 4.004190921783447,
|
| 1832 |
+
"learning_rate": 4.117799877106181e-05,
|
| 1833 |
+
"loss": 2.0189,
|
| 1834 |
+
"step": 2470
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 1.8120376290072153,
|
| 1838 |
+
"grad_norm": 7.879773139953613,
|
| 1839 |
+
"learning_rate": 4.075965183592592e-05,
|
| 1840 |
+
"loss": 1.9538,
|
| 1841 |
+
"step": 2480
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 1.8193442323499864,
|
| 1845 |
+
"grad_norm": 5.945032596588135,
|
| 1846 |
+
"learning_rate": 4.034197359389666e-05,
|
| 1847 |
+
"loss": 1.853,
|
| 1848 |
+
"step": 2490
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 1.8266508356927573,
|
| 1852 |
+
"grad_norm": 6.213232517242432,
|
| 1853 |
+
"learning_rate": 3.992499427095213e-05,
|
| 1854 |
+
"loss": 2.102,
|
| 1855 |
+
"step": 2500
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 1.8339574390355282,
|
| 1859 |
+
"grad_norm": 4.097675323486328,
|
| 1860 |
+
"learning_rate": 3.950874404249199e-05,
|
| 1861 |
+
"loss": 2.1149,
|
| 1862 |
+
"step": 2510
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 1.8412640423782993,
|
| 1866 |
+
"grad_norm": 3.8982291221618652,
|
| 1867 |
+
"learning_rate": 3.9093253031153755e-05,
|
| 1868 |
+
"loss": 1.961,
|
| 1869 |
+
"step": 2520
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 1.8485706457210704,
|
| 1873 |
+
"grad_norm": 4.43918514251709,
|
| 1874 |
+
"learning_rate": 3.8678551304632965e-05,
|
| 1875 |
+
"loss": 1.8826,
|
| 1876 |
+
"step": 2530
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 1.8558772490638415,
|
| 1880 |
+
"grad_norm": 6.397246360778809,
|
| 1881 |
+
"learning_rate": 3.8264668873507245e-05,
|
| 1882 |
+
"loss": 1.9244,
|
| 1883 |
+
"step": 2540
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 1.8631838524066124,
|
| 1887 |
+
"grad_norm": 4.929690361022949,
|
| 1888 |
+
"learning_rate": 3.7851635689064546e-05,
|
| 1889 |
+
"loss": 1.8184,
|
| 1890 |
+
"step": 2550
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 1.8704904557493836,
|
| 1894 |
+
"grad_norm": 4.7890801429748535,
|
| 1895 |
+
"learning_rate": 3.743948164113567e-05,
|
| 1896 |
+
"loss": 1.9089,
|
| 1897 |
+
"step": 2560
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 1.8777970590921544,
|
| 1901 |
+
"grad_norm": 5.918452262878418,
|
| 1902 |
+
"learning_rate": 3.702823655593128e-05,
|
| 1903 |
+
"loss": 2.0868,
|
| 1904 |
+
"step": 2570
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 1.8851036624349256,
|
| 1908 |
+
"grad_norm": 5.867265701293945,
|
| 1909 |
+
"learning_rate": 3.6617930193883384e-05,
|
| 1910 |
+
"loss": 2.0125,
|
| 1911 |
+
"step": 2580
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 1.8924102657776967,
|
| 1915 |
+
"grad_norm": 4.536275863647461,
|
| 1916 |
+
"learning_rate": 3.62085922474918e-05,
|
| 1917 |
+
"loss": 1.9122,
|
| 1918 |
+
"step": 2590
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 1.8997168691204678,
|
| 1922 |
+
"grad_norm": 5.288751602172852,
|
| 1923 |
+
"learning_rate": 3.580025233917529e-05,
|
| 1924 |
+
"loss": 1.9932,
|
| 1925 |
+
"step": 2600
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 1.8997168691204678,
|
| 1929 |
+
"eval_loss": 2.0270345211029053,
|
| 1930 |
+
"eval_runtime": 111.0925,
|
| 1931 |
+
"eval_samples_per_second": 10.955,
|
| 1932 |
+
"eval_steps_per_second": 10.955,
|
| 1933 |
+
"step": 2600
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 1.9070234724632387,
|
| 1937 |
+
"grad_norm": 5.767890453338623,
|
| 1938 |
+
"learning_rate": 3.5392940019127977e-05,
|
| 1939 |
+
"loss": 1.9772,
|
| 1940 |
+
"step": 2610
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 1.9143300758060096,
|
| 1944 |
+
"grad_norm": 5.185894012451172,
|
| 1945 |
+
"learning_rate": 3.498668476318083e-05,
|
| 1946 |
+
"loss": 1.6921,
|
| 1947 |
+
"step": 2620
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 1.9216366791487807,
|
| 1951 |
+
"grad_norm": 6.62992525100708,
|
| 1952 |
+
"learning_rate": 3.458151597066863e-05,
|
| 1953 |
+
"loss": 1.9544,
|
| 1954 |
+
"step": 2630
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 1.9289432824915518,
|
| 1958 |
+
"grad_norm": 11.95433235168457,
|
| 1959 |
+
"learning_rate": 3.417746296230244e-05,
|
| 1960 |
+
"loss": 2.0189,
|
| 1961 |
+
"step": 2640
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 1.936249885834323,
|
| 1965 |
+
"grad_norm": 4.77971076965332,
|
| 1966 |
+
"learning_rate": 3.3774554978047756e-05,
|
| 1967 |
+
"loss": 1.9203,
|
| 1968 |
+
"step": 2650
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 1.9435564891770938,
|
| 1972 |
+
"grad_norm": 7.271909236907959,
|
| 1973 |
+
"learning_rate": 3.337282117500847e-05,
|
| 1974 |
+
"loss": 2.0945,
|
| 1975 |
+
"step": 2660
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 1.9508630925198647,
|
| 1979 |
+
"grad_norm": 8.73035717010498,
|
| 1980 |
+
"learning_rate": 3.297229062531696e-05,
|
| 1981 |
+
"loss": 2.0409,
|
| 1982 |
+
"step": 2670
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 1.9581696958626358,
|
| 1986 |
+
"grad_norm": 9.147926330566406,
|
| 1987 |
+
"learning_rate": 3.257299231403014e-05,
|
| 1988 |
+
"loss": 1.9463,
|
| 1989 |
+
"step": 2680
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 1.965476299205407,
|
| 1993 |
+
"grad_norm": 29.067060470581055,
|
| 1994 |
+
"learning_rate": 3.217495513703198e-05,
|
| 1995 |
+
"loss": 2.0646,
|
| 1996 |
+
"step": 2690
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 1.972782902548178,
|
| 2000 |
+
"grad_norm": 17.448034286499023,
|
| 2001 |
+
"learning_rate": 3.177820789894234e-05,
|
| 2002 |
+
"loss": 1.8981,
|
| 2003 |
+
"step": 2700
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 1.980089505890949,
|
| 2007 |
+
"grad_norm": 5.893951892852783,
|
| 2008 |
+
"learning_rate": 3.138277931103254e-05,
|
| 2009 |
+
"loss": 1.7993,
|
| 2010 |
+
"step": 2710
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 1.98739610923372,
|
| 2014 |
+
"grad_norm": 8.20900821685791,
|
| 2015 |
+
"learning_rate": 3.09886979891476e-05,
|
| 2016 |
+
"loss": 2.1449,
|
| 2017 |
+
"step": 2720
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 1.994702712576491,
|
| 2021 |
+
"grad_norm": 5.447132587432861,
|
| 2022 |
+
"learning_rate": 3.059599245163538e-05,
|
| 2023 |
+
"loss": 1.8473,
|
| 2024 |
+
"step": 2730
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 2.002009315919262,
|
| 2028 |
+
"grad_norm": 6.67828893661499,
|
| 2029 |
+
"learning_rate": 3.0204691117282856e-05,
|
| 2030 |
+
"loss": 1.9341,
|
| 2031 |
+
"step": 2740
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 2.009315919262033,
|
| 2035 |
+
"grad_norm": 7.406982898712158,
|
| 2036 |
+
"learning_rate": 2.981482230325946e-05,
|
| 2037 |
+
"loss": 1.9778,
|
| 2038 |
+
"step": 2750
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 2.0166225226048042,
|
| 2042 |
+
"grad_norm": 4.272141456604004,
|
| 2043 |
+
"learning_rate": 2.9426414223067978e-05,
|
| 2044 |
+
"loss": 1.8293,
|
| 2045 |
+
"step": 2760
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 2.023929125947575,
|
| 2049 |
+
"grad_norm": 5.6061320304870605,
|
| 2050 |
+
"learning_rate": 2.9039494984502734e-05,
|
| 2051 |
+
"loss": 1.7844,
|
| 2052 |
+
"step": 2770
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 2.031235729290346,
|
| 2056 |
+
"grad_norm": 6.46597957611084,
|
| 2057 |
+
"learning_rate": 2.865409258761557e-05,
|
| 2058 |
+
"loss": 1.8147,
|
| 2059 |
+
"step": 2780
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 2.038542332633117,
|
| 2063 |
+
"grad_norm": 6.094344139099121,
|
| 2064 |
+
"learning_rate": 2.8270234922689597e-05,
|
| 2065 |
+
"loss": 1.7082,
|
| 2066 |
+
"step": 2790
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 2.0458489359758882,
|
| 2070 |
+
"grad_norm": 5.04010534286499,
|
| 2071 |
+
"learning_rate": 2.788794976822077e-05,
|
| 2072 |
+
"loss": 1.6477,
|
| 2073 |
+
"step": 2800
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 2.0458489359758882,
|
| 2077 |
+
"eval_loss": 2.0164308547973633,
|
| 2078 |
+
"eval_runtime": 111.1374,
|
| 2079 |
+
"eval_samples_per_second": 10.95,
|
| 2080 |
+
"eval_steps_per_second": 10.95,
|
| 2081 |
+
"step": 2800
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.0531555393186594,
|
| 2085 |
+
"grad_norm": 13.458788871765137,
|
| 2086 |
+
"learning_rate": 2.7507264788907783e-05,
|
| 2087 |
+
"loss": 1.7778,
|
| 2088 |
+
"step": 2810
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.0604621426614305,
|
| 2092 |
+
"grad_norm": 6.946293830871582,
|
| 2093 |
+
"learning_rate": 2.712820753364998e-05,
|
| 2094 |
+
"loss": 1.6629,
|
| 2095 |
+
"step": 2820
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.067768746004201,
|
| 2099 |
+
"grad_norm": 8.358539581298828,
|
| 2100 |
+
"learning_rate": 2.6750805433553728e-05,
|
| 2101 |
+
"loss": 2.0157,
|
| 2102 |
+
"step": 2830
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.0750753493469722,
|
| 2106 |
+
"grad_norm": 6.011590480804443,
|
| 2107 |
+
"learning_rate": 2.637508579994741e-05,
|
| 2108 |
+
"loss": 1.9037,
|
| 2109 |
+
"step": 2840
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.0823819526897434,
|
| 2113 |
+
"grad_norm": 5.1315226554870605,
|
| 2114 |
+
"learning_rate": 2.6001075822404864e-05,
|
| 2115 |
+
"loss": 1.8682,
|
| 2116 |
+
"step": 2850
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.0896885560325145,
|
| 2120 |
+
"grad_norm": 5.693300247192383,
|
| 2121 |
+
"learning_rate": 2.5628802566777904e-05,
|
| 2122 |
+
"loss": 1.8367,
|
| 2123 |
+
"step": 2860
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.0969951593752856,
|
| 2127 |
+
"grad_norm": 5.308597087860107,
|
| 2128 |
+
"learning_rate": 2.5258292973237536e-05,
|
| 2129 |
+
"loss": 1.9468,
|
| 2130 |
+
"step": 2870
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.1043017627180562,
|
| 2134 |
+
"grad_norm": 6.057744979858398,
|
| 2135 |
+
"learning_rate": 2.4889573854324443e-05,
|
| 2136 |
+
"loss": 1.7683,
|
| 2137 |
+
"step": 2880
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.1116083660608274,
|
| 2141 |
+
"grad_norm": 5.609562397003174,
|
| 2142 |
+
"learning_rate": 2.452267189300864e-05,
|
| 2143 |
+
"loss": 1.8463,
|
| 2144 |
+
"step": 2890
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.1189149694035985,
|
| 2148 |
+
"grad_norm": 4.729313850402832,
|
| 2149 |
+
"learning_rate": 2.415761364075857e-05,
|
| 2150 |
+
"loss": 1.9935,
|
| 2151 |
+
"step": 2900
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.1262215727463696,
|
| 2155 |
+
"grad_norm": 4.285727500915527,
|
| 2156 |
+
"learning_rate": 2.3794425515619535e-05,
|
| 2157 |
+
"loss": 1.8125,
|
| 2158 |
+
"step": 2910
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.1335281760891407,
|
| 2162 |
+
"grad_norm": 5.974366664886475,
|
| 2163 |
+
"learning_rate": 2.343313380030207e-05,
|
| 2164 |
+
"loss": 1.8855,
|
| 2165 |
+
"step": 2920
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.140834779431912,
|
| 2169 |
+
"grad_norm": 5.998205661773682,
|
| 2170 |
+
"learning_rate": 2.30737646402798e-05,
|
| 2171 |
+
"loss": 2.0122,
|
| 2172 |
+
"step": 2930
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.1481413827746825,
|
| 2176 |
+
"grad_norm": 7.216914176940918,
|
| 2177 |
+
"learning_rate": 2.271634404189752e-05,
|
| 2178 |
+
"loss": 1.8831,
|
| 2179 |
+
"step": 2940
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.1554479861174536,
|
| 2183 |
+
"grad_norm": 6.571489334106445,
|
| 2184 |
+
"learning_rate": 2.2360897870489055e-05,
|
| 2185 |
+
"loss": 1.6656,
|
| 2186 |
+
"step": 2950
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.1627545894602247,
|
| 2190 |
+
"grad_norm": 5.754743576049805,
|
| 2191 |
+
"learning_rate": 2.2007451848505627e-05,
|
| 2192 |
+
"loss": 1.7651,
|
| 2193 |
+
"step": 2960
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.170061192802996,
|
| 2197 |
+
"grad_norm": 4.261959552764893,
|
| 2198 |
+
"learning_rate": 2.1656031553654272e-05,
|
| 2199 |
+
"loss": 2.0409,
|
| 2200 |
+
"step": 2970
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.177367796145767,
|
| 2204 |
+
"grad_norm": 4.679163932800293,
|
| 2205 |
+
"learning_rate": 2.1306662417046968e-05,
|
| 2206 |
+
"loss": 1.7555,
|
| 2207 |
+
"step": 2980
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.1846743994885376,
|
| 2211 |
+
"grad_norm": 5.427217483520508,
|
| 2212 |
+
"learning_rate": 2.0959369721360183e-05,
|
| 2213 |
+
"loss": 2.0074,
|
| 2214 |
+
"step": 2990
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.1919810028313087,
|
| 2218 |
+
"grad_norm": 8.53897476196289,
|
| 2219 |
+
"learning_rate": 2.0614178599005356e-05,
|
| 2220 |
+
"loss": 1.5985,
|
| 2221 |
+
"step": 3000
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.1919810028313087,
|
| 2225 |
+
"eval_loss": 2.0024373531341553,
|
| 2226 |
+
"eval_runtime": 110.7503,
|
| 2227 |
+
"eval_samples_per_second": 10.989,
|
| 2228 |
+
"eval_steps_per_second": 10.989,
|
| 2229 |
+
"step": 3000
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.19928760617408,
|
| 2233 |
+
"grad_norm": 8.221793174743652,
|
| 2234 |
+
"learning_rate": 2.0271114030310035e-05,
|
| 2235 |
+
"loss": 1.8105,
|
| 2236 |
+
"step": 3010
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.206594209516851,
|
| 2240 |
+
"grad_norm": 5.500391006469727,
|
| 2241 |
+
"learning_rate": 1.9930200841710193e-05,
|
| 2242 |
+
"loss": 1.844,
|
| 2243 |
+
"step": 3020
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.213900812859622,
|
| 2247 |
+
"grad_norm": 8.051352500915527,
|
| 2248 |
+
"learning_rate": 1.9591463703953672e-05,
|
| 2249 |
+
"loss": 1.6881,
|
| 2250 |
+
"step": 3030
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.221207416202393,
|
| 2254 |
+
"grad_norm": 4.4952898025512695,
|
| 2255 |
+
"learning_rate": 1.9254927130314726e-05,
|
| 2256 |
+
"loss": 1.8373,
|
| 2257 |
+
"step": 3040
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.228514019545164,
|
| 2261 |
+
"grad_norm": 7.687475204467773,
|
| 2262 |
+
"learning_rate": 1.8920615474820152e-05,
|
| 2263 |
+
"loss": 1.9104,
|
| 2264 |
+
"step": 3050
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.235820622887935,
|
| 2268 |
+
"grad_norm": 5.942209243774414,
|
| 2269 |
+
"learning_rate": 1.8588552930486915e-05,
|
| 2270 |
+
"loss": 1.9455,
|
| 2271 |
+
"step": 3060
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.243127226230706,
|
| 2275 |
+
"grad_norm": 4.695751667022705,
|
| 2276 |
+
"learning_rate": 1.8258763527571243e-05,
|
| 2277 |
+
"loss": 1.7007,
|
| 2278 |
+
"step": 3070
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.250433829573477,
|
| 2282 |
+
"grad_norm": 5.661278247833252,
|
| 2283 |
+
"learning_rate": 1.7931271131829758e-05,
|
| 2284 |
+
"loss": 1.7467,
|
| 2285 |
+
"step": 3080
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.257740432916248,
|
| 2289 |
+
"grad_norm": 7.149540901184082,
|
| 2290 |
+
"learning_rate": 1.7606099442792373e-05,
|
| 2291 |
+
"loss": 1.6398,
|
| 2292 |
+
"step": 3090
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.265047036259019,
|
| 2296 |
+
"grad_norm": 7.075028419494629,
|
| 2297 |
+
"learning_rate": 1.728327199204716e-05,
|
| 2298 |
+
"loss": 1.7911,
|
| 2299 |
+
"step": 3100
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.27235363960179,
|
| 2303 |
+
"grad_norm": 5.169941425323486,
|
| 2304 |
+
"learning_rate": 1.696281214153757e-05,
|
| 2305 |
+
"loss": 1.9487,
|
| 2306 |
+
"step": 3110
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.279660242944561,
|
| 2310 |
+
"grad_norm": 8.544054985046387,
|
| 2311 |
+
"learning_rate": 1.664474308187167e-05,
|
| 2312 |
+
"loss": 1.9077,
|
| 2313 |
+
"step": 3120
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.2869668462873323,
|
| 2317 |
+
"grad_norm": 5.160293102264404,
|
| 2318 |
+
"learning_rate": 1.6329087830644053e-05,
|
| 2319 |
+
"loss": 1.7878,
|
| 2320 |
+
"step": 3130
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.2942734496301034,
|
| 2324 |
+
"grad_norm": 7.8444342613220215,
|
| 2325 |
+
"learning_rate": 1.6015869230769992e-05,
|
| 2326 |
+
"loss": 1.6692,
|
| 2327 |
+
"step": 3140
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.301580052972874,
|
| 2331 |
+
"grad_norm": 5.798057556152344,
|
| 2332 |
+
"learning_rate": 1.5705109948832526e-05,
|
| 2333 |
+
"loss": 1.8317,
|
| 2334 |
+
"step": 3150
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 2.308886656315645,
|
| 2338 |
+
"grad_norm": 7.343387126922607,
|
| 2339 |
+
"learning_rate": 1.5396832473442e-05,
|
| 2340 |
+
"loss": 1.9332,
|
| 2341 |
+
"step": 3160
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 2.3161932596584163,
|
| 2345 |
+
"grad_norm": 5.731598377227783,
|
| 2346 |
+
"learning_rate": 1.5091059113608785e-05,
|
| 2347 |
+
"loss": 1.6953,
|
| 2348 |
+
"step": 3170
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 2.3234998630011874,
|
| 2352 |
+
"grad_norm": 7.075622081756592,
|
| 2353 |
+
"learning_rate": 1.4787811997128737e-05,
|
| 2354 |
+
"loss": 1.8049,
|
| 2355 |
+
"step": 3180
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.3308064663439585,
|
| 2359 |
+
"grad_norm": 9.426651000976562,
|
| 2360 |
+
"learning_rate": 1.4487113068981934e-05,
|
| 2361 |
+
"loss": 2.002,
|
| 2362 |
+
"step": 3190
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.338113069686729,
|
| 2366 |
+
"grad_norm": 6.5062360763549805,
|
| 2367 |
+
"learning_rate": 1.418898408974456e-05,
|
| 2368 |
+
"loss": 1.8708,
|
| 2369 |
+
"step": 3200
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.338113069686729,
|
| 2373 |
+
"eval_loss": 1.9918181896209717,
|
| 2374 |
+
"eval_runtime": 110.5621,
|
| 2375 |
+
"eval_samples_per_second": 11.007,
|
| 2376 |
+
"eval_steps_per_second": 11.007,
|
| 2377 |
+
"step": 3200
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 2.3454196730295003,
|
| 2381 |
+
"grad_norm": 5.592414379119873,
|
| 2382 |
+
"learning_rate": 1.3893446634014257e-05,
|
| 2383 |
+
"loss": 1.8714,
|
| 2384 |
+
"step": 3210
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 2.3527262763722714,
|
| 2388 |
+
"grad_norm": 5.582223892211914,
|
| 2389 |
+
"learning_rate": 1.3600522088848689e-05,
|
| 2390 |
+
"loss": 1.7765,
|
| 2391 |
+
"step": 3220
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 2.3600328797150425,
|
| 2395 |
+
"grad_norm": 7.002323150634766,
|
| 2396 |
+
"learning_rate": 1.3310231652217997e-05,
|
| 2397 |
+
"loss": 1.9461,
|
| 2398 |
+
"step": 3230
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 2.3673394830578136,
|
| 2402 |
+
"grad_norm": 12.038192749023438,
|
| 2403 |
+
"learning_rate": 1.3022596331470632e-05,
|
| 2404 |
+
"loss": 1.8854,
|
| 2405 |
+
"step": 3240
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 2.3746460864005847,
|
| 2409 |
+
"grad_norm": 7.399435520172119,
|
| 2410 |
+
"learning_rate": 1.2737636941813196e-05,
|
| 2411 |
+
"loss": 1.9169,
|
| 2412 |
+
"step": 3250
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 2.3819526897433554,
|
| 2416 |
+
"grad_norm": 8.563164710998535,
|
| 2417 |
+
"learning_rate": 1.245537410480414e-05,
|
| 2418 |
+
"loss": 1.9562,
|
| 2419 |
+
"step": 3260
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 2.3892592930861265,
|
| 2423 |
+
"grad_norm": 5.194084167480469,
|
| 2424 |
+
"learning_rate": 1.2175828246861359e-05,
|
| 2425 |
+
"loss": 1.7997,
|
| 2426 |
+
"step": 3270
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 2.3965658964288976,
|
| 2430 |
+
"grad_norm": 6.553465366363525,
|
| 2431 |
+
"learning_rate": 1.1899019597784117e-05,
|
| 2432 |
+
"loss": 1.8751,
|
| 2433 |
+
"step": 3280
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 2.4038724997716687,
|
| 2437 |
+
"grad_norm": 5.508111476898193,
|
| 2438 |
+
"learning_rate": 1.1624968189288965e-05,
|
| 2439 |
+
"loss": 1.7808,
|
| 2440 |
+
"step": 3290
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 2.41117910311444,
|
| 2444 |
+
"grad_norm": 6.692938327789307,
|
| 2445 |
+
"learning_rate": 1.1353693853560216e-05,
|
| 2446 |
+
"loss": 1.6511,
|
| 2447 |
+
"step": 3300
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 2.4184857064572105,
|
| 2451 |
+
"grad_norm": 6.976585388183594,
|
| 2452 |
+
"learning_rate": 1.1085216221814665e-05,
|
| 2453 |
+
"loss": 1.8939,
|
| 2454 |
+
"step": 3310
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 2.4257923097999816,
|
| 2458 |
+
"grad_norm": 7.673839569091797,
|
| 2459 |
+
"learning_rate": 1.0819554722881048e-05,
|
| 2460 |
+
"loss": 2.0006,
|
| 2461 |
+
"step": 3320
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 2.4330989131427527,
|
| 2465 |
+
"grad_norm": 6.984090805053711,
|
| 2466 |
+
"learning_rate": 1.055672858179393e-05,
|
| 2467 |
+
"loss": 1.7107,
|
| 2468 |
+
"step": 3330
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 2.440405516485524,
|
| 2472 |
+
"grad_norm": 6.236724853515625,
|
| 2473 |
+
"learning_rate": 1.0296756818402531e-05,
|
| 2474 |
+
"loss": 1.6792,
|
| 2475 |
+
"step": 3340
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 2.447712119828295,
|
| 2479 |
+
"grad_norm": 7.661273002624512,
|
| 2480 |
+
"learning_rate": 1.0039658245994277e-05,
|
| 2481 |
+
"loss": 1.818,
|
| 2482 |
+
"step": 3350
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 2.455018723171066,
|
| 2486 |
+
"grad_norm": 8.403165817260742,
|
| 2487 |
+
"learning_rate": 9.78545146993342e-06,
|
| 2488 |
+
"loss": 1.9927,
|
| 2489 |
+
"step": 3360
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 2.4623253265138367,
|
| 2493 |
+
"grad_norm": 5.08246374130249,
|
| 2494 |
+
"learning_rate": 9.534154886314517e-06,
|
| 2495 |
+
"loss": 1.7398,
|
| 2496 |
+
"step": 3370
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 2.469631929856608,
|
| 2500 |
+
"grad_norm": 7.482173919677734,
|
| 2501 |
+
"learning_rate": 9.28578668063127e-06,
|
| 2502 |
+
"loss": 1.9105,
|
| 2503 |
+
"step": 3380
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 2.476938533199379,
|
| 2507 |
+
"grad_norm": 4.506008148193359,
|
| 2508 |
+
"learning_rate": 9.040364826460423e-06,
|
| 2509 |
+
"loss": 1.9258,
|
| 2510 |
+
"step": 3390
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 2.48424513654215,
|
| 2514 |
+
"grad_norm": 6.290830612182617,
|
| 2515 |
+
"learning_rate": 8.797907084161155e-06,
|
| 2516 |
+
"loss": 1.8533,
|
| 2517 |
+
"step": 3400
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 2.48424513654215,
|
| 2521 |
+
"eval_loss": 1.9872702360153198,
|
| 2522 |
+
"eval_runtime": 110.4675,
|
| 2523 |
+
"eval_samples_per_second": 11.017,
|
| 2524 |
+
"eval_steps_per_second": 11.017,
|
| 2525 |
+
"step": 3400
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 2.491551739884921,
|
| 2529 |
+
"grad_norm": 5.8204522132873535,
|
| 2530 |
+
"learning_rate": 8.558430999589723e-06,
|
| 2531 |
+
"loss": 1.9186,
|
| 2532 |
+
"step": 3410
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 2.498858343227692,
|
| 2536 |
+
"grad_norm": 7.549562931060791,
|
| 2537 |
+
"learning_rate": 8.321953902829842e-06,
|
| 2538 |
+
"loss": 1.6548,
|
| 2539 |
+
"step": 3420
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 2.506164946570463,
|
| 2543 |
+
"grad_norm": 5.260471820831299,
|
| 2544 |
+
"learning_rate": 8.08849290693846e-06,
|
| 2545 |
+
"loss": 1.877,
|
| 2546 |
+
"step": 3430
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 2.513471549913234,
|
| 2550 |
+
"grad_norm": 7.633718967437744,
|
| 2551 |
+
"learning_rate": 7.85806490670739e-06,
|
| 2552 |
+
"loss": 1.7109,
|
| 2553 |
+
"step": 3440
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 2.520778153256005,
|
| 2557 |
+
"grad_norm": 6.902381420135498,
|
| 2558 |
+
"learning_rate": 7.630686577440722e-06,
|
| 2559 |
+
"loss": 1.8875,
|
| 2560 |
+
"step": 3450
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 2.5280847565987763,
|
| 2564 |
+
"grad_norm": 8.275242805480957,
|
| 2565 |
+
"learning_rate": 7.406374373748004e-06,
|
| 2566 |
+
"loss": 1.8563,
|
| 2567 |
+
"step": 3460
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 2.5353913599415474,
|
| 2571 |
+
"grad_norm": 5.76878023147583,
|
| 2572 |
+
"learning_rate": 7.185144528353583e-06,
|
| 2573 |
+
"loss": 1.7962,
|
| 2574 |
+
"step": 3470
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 2.542697963284318,
|
| 2578 |
+
"grad_norm": 4.979648590087891,
|
| 2579 |
+
"learning_rate": 6.967013050921795e-06,
|
| 2580 |
+
"loss": 1.9158,
|
| 2581 |
+
"step": 3480
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 2.550004566627089,
|
| 2585 |
+
"grad_norm": 9.704955101013184,
|
| 2586 |
+
"learning_rate": 6.751995726898464e-06,
|
| 2587 |
+
"loss": 2.0237,
|
| 2588 |
+
"step": 3490
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 2.5573111699698603,
|
| 2592 |
+
"grad_norm": 9.094025611877441,
|
| 2593 |
+
"learning_rate": 6.540108116368515e-06,
|
| 2594 |
+
"loss": 1.8791,
|
| 2595 |
+
"step": 3500
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 2.5646177733126314,
|
| 2599 |
+
"grad_norm": 6.252310752868652,
|
| 2600 |
+
"learning_rate": 6.33136555293003e-06,
|
| 2601 |
+
"loss": 1.8576,
|
| 2602 |
+
"step": 3510
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 2.571924376655402,
|
| 2606 |
+
"grad_norm": 7.935026168823242,
|
| 2607 |
+
"learning_rate": 6.125783142584479e-06,
|
| 2608 |
+
"loss": 1.9725,
|
| 2609 |
+
"step": 3520
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 2.579230979998173,
|
| 2613 |
+
"grad_norm": 6.289064884185791,
|
| 2614 |
+
"learning_rate": 5.923375762643668e-06,
|
| 2615 |
+
"loss": 1.7741,
|
| 2616 |
+
"step": 3530
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 2.5865375833409443,
|
| 2620 |
+
"grad_norm": 6.07642126083374,
|
| 2621 |
+
"learning_rate": 5.724158060653029e-06,
|
| 2622 |
+
"loss": 1.8716,
|
| 2623 |
+
"step": 3540
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 2.5938441866837154,
|
| 2627 |
+
"grad_norm": 7.241240501403809,
|
| 2628 |
+
"learning_rate": 5.528144453331696e-06,
|
| 2629 |
+
"loss": 1.7577,
|
| 2630 |
+
"step": 3550
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 2.6011507900264865,
|
| 2634 |
+
"grad_norm": 5.838751316070557,
|
| 2635 |
+
"learning_rate": 5.335349125529154e-06,
|
| 2636 |
+
"loss": 1.8637,
|
| 2637 |
+
"step": 3560
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 2.6084573933692576,
|
| 2641 |
+
"grad_norm": 7.343444347381592,
|
| 2642 |
+
"learning_rate": 5.14578602919879e-06,
|
| 2643 |
+
"loss": 1.8605,
|
| 2644 |
+
"step": 3570
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 2.6157639967120287,
|
| 2648 |
+
"grad_norm": 5.384237289428711,
|
| 2649 |
+
"learning_rate": 4.959468882388163e-06,
|
| 2650 |
+
"loss": 1.739,
|
| 2651 |
+
"step": 3580
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 2.6230706000547994,
|
| 2655 |
+
"grad_norm": 8.187921524047852,
|
| 2656 |
+
"learning_rate": 4.776411168246353e-06,
|
| 2657 |
+
"loss": 1.8639,
|
| 2658 |
+
"step": 3590
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 2.6303772033975705,
|
| 2662 |
+
"grad_norm": 5.8128838539123535,
|
| 2663 |
+
"learning_rate": 4.596626134048176e-06,
|
| 2664 |
+
"loss": 1.7357,
|
| 2665 |
+
"step": 3600
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 2.6303772033975705,
|
| 2669 |
+
"eval_loss": 1.9813354015350342,
|
| 2670 |
+
"eval_runtime": 104.4364,
|
| 2671 |
+
"eval_samples_per_second": 11.653,
|
| 2672 |
+
"eval_steps_per_second": 11.653,
|
| 2673 |
+
"step": 3600
|
| 2674 |
+
},
|
| 2675 |
+
{
|
| 2676 |
+
"epoch": 2.6376838067403416,
|
| 2677 |
+
"grad_norm": 5.993869304656982,
|
| 2678 |
+
"learning_rate": 4.420126790235552e-06,
|
| 2679 |
+
"loss": 1.9172,
|
| 2680 |
+
"step": 3610
|
| 2681 |
+
},
|
| 2682 |
+
{
|
| 2683 |
+
"epoch": 2.6449904100831128,
|
| 2684 |
+
"grad_norm": 6.651885509490967,
|
| 2685 |
+
"learning_rate": 4.246925909475957e-06,
|
| 2686 |
+
"loss": 1.7701,
|
| 2687 |
+
"step": 3620
|
| 2688 |
+
},
|
| 2689 |
+
{
|
| 2690 |
+
"epoch": 2.6522970134258834,
|
| 2691 |
+
"grad_norm": 5.599200248718262,
|
| 2692 |
+
"learning_rate": 4.077036025738118e-06,
|
| 2693 |
+
"loss": 1.9576,
|
| 2694 |
+
"step": 3630
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"epoch": 2.6596036167686545,
|
| 2698 |
+
"grad_norm": 5.870293140411377,
|
| 2699 |
+
"learning_rate": 3.910469433385017e-06,
|
| 2700 |
+
"loss": 1.7158,
|
| 2701 |
+
"step": 3640
|
| 2702 |
+
},
|
| 2703 |
+
{
|
| 2704 |
+
"epoch": 2.6669102201114256,
|
| 2705 |
+
"grad_norm": 7.64784574508667,
|
| 2706 |
+
"learning_rate": 3.7472381862840967e-06,
|
| 2707 |
+
"loss": 1.8767,
|
| 2708 |
+
"step": 3650
|
| 2709 |
+
},
|
| 2710 |
+
{
|
| 2711 |
+
"epoch": 2.6742168234541968,
|
| 2712 |
+
"grad_norm": 5.308115005493164,
|
| 2713 |
+
"learning_rate": 3.5873540969350415e-06,
|
| 2714 |
+
"loss": 1.9414,
|
| 2715 |
+
"step": 3660
|
| 2716 |
+
},
|
| 2717 |
+
{
|
| 2718 |
+
"epoch": 2.681523426796968,
|
| 2719 |
+
"grad_norm": 6.556251525878906,
|
| 2720 |
+
"learning_rate": 3.430828735614916e-06,
|
| 2721 |
+
"loss": 1.8932,
|
| 2722 |
+
"step": 3670
|
| 2723 |
+
},
|
| 2724 |
+
{
|
| 2725 |
+
"epoch": 2.688830030139739,
|
| 2726 |
+
"grad_norm": 5.682919025421143,
|
| 2727 |
+
"learning_rate": 3.277673429540862e-06,
|
| 2728 |
+
"loss": 1.8576,
|
| 2729 |
+
"step": 3680
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 2.69613663348251,
|
| 2733 |
+
"grad_norm": 5.832954406738281,
|
| 2734 |
+
"learning_rate": 3.1278992620503877e-06,
|
| 2735 |
+
"loss": 1.6419,
|
| 2736 |
+
"step": 3690
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 2.7034432368252808,
|
| 2740 |
+
"grad_norm": 5.423207759857178,
|
| 2741 |
+
"learning_rate": 2.9815170717993115e-06,
|
| 2742 |
+
"loss": 1.7736,
|
| 2743 |
+
"step": 3700
|
| 2744 |
+
},
|
| 2745 |
+
{
|
| 2746 |
+
"epoch": 2.710749840168052,
|
| 2747 |
+
"grad_norm": 5.549076080322266,
|
| 2748 |
+
"learning_rate": 2.83853745197738e-06,
|
| 2749 |
+
"loss": 1.9699,
|
| 2750 |
+
"step": 3710
|
| 2751 |
+
},
|
| 2752 |
+
{
|
| 2753 |
+
"epoch": 2.718056443510823,
|
| 2754 |
+
"grad_norm": 5.261001110076904,
|
| 2755 |
+
"learning_rate": 2.6989707495417292e-06,
|
| 2756 |
+
"loss": 1.5671,
|
| 2757 |
+
"step": 3720
|
| 2758 |
+
},
|
| 2759 |
+
{
|
| 2760 |
+
"epoch": 2.725363046853594,
|
| 2761 |
+
"grad_norm": 9.049291610717773,
|
| 2762 |
+
"learning_rate": 2.5628270644680265e-06,
|
| 2763 |
+
"loss": 1.9909,
|
| 2764 |
+
"step": 3730
|
| 2765 |
+
},
|
| 2766 |
+
{
|
| 2767 |
+
"epoch": 2.7326696501963648,
|
| 2768 |
+
"grad_norm": 5.768558979034424,
|
| 2769 |
+
"learning_rate": 2.430116249019665e-06,
|
| 2770 |
+
"loss": 1.8207,
|
| 2771 |
+
"step": 3740
|
| 2772 |
+
},
|
| 2773 |
+
{
|
| 2774 |
+
"epoch": 2.739976253539136,
|
| 2775 |
+
"grad_norm": 7.210824012756348,
|
| 2776 |
+
"learning_rate": 2.3008479070346867e-06,
|
| 2777 |
+
"loss": 1.8219,
|
| 2778 |
+
"step": 3750
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"epoch": 2.747282856881907,
|
| 2782 |
+
"grad_norm": 5.107162952423096,
|
| 2783 |
+
"learning_rate": 2.1750313932308806e-06,
|
| 2784 |
+
"loss": 1.7551,
|
| 2785 |
+
"step": 3760
|
| 2786 |
+
},
|
| 2787 |
+
{
|
| 2788 |
+
"epoch": 2.754589460224678,
|
| 2789 |
+
"grad_norm": 6.917966365814209,
|
| 2790 |
+
"learning_rate": 2.0526758125287427e-06,
|
| 2791 |
+
"loss": 1.6674,
|
| 2792 |
+
"step": 3770
|
| 2793 |
+
},
|
| 2794 |
+
{
|
| 2795 |
+
"epoch": 2.761896063567449,
|
| 2796 |
+
"grad_norm": 5.870244026184082,
|
| 2797 |
+
"learning_rate": 1.933790019392634e-06,
|
| 2798 |
+
"loss": 1.8,
|
| 2799 |
+
"step": 3780
|
| 2800 |
+
},
|
| 2801 |
+
{
|
| 2802 |
+
"epoch": 2.7692026669102203,
|
| 2803 |
+
"grad_norm": 5.093700408935547,
|
| 2804 |
+
"learning_rate": 1.8183826171899677e-06,
|
| 2805 |
+
"loss": 1.7592,
|
| 2806 |
+
"step": 3790
|
| 2807 |
+
},
|
| 2808 |
+
{
|
| 2809 |
+
"epoch": 2.776509270252991,
|
| 2810 |
+
"grad_norm": 9.029096603393555,
|
| 2811 |
+
"learning_rate": 1.7064619575686336e-06,
|
| 2812 |
+
"loss": 1.8041,
|
| 2813 |
+
"step": 3800
|
| 2814 |
+
},
|
| 2815 |
+
{
|
| 2816 |
+
"epoch": 2.776509270252991,
|
| 2817 |
+
"eval_loss": 1.9782235622406006,
|
| 2818 |
+
"eval_runtime": 105.0311,
|
| 2819 |
+
"eval_samples_per_second": 11.587,
|
| 2820 |
+
"eval_steps_per_second": 11.587,
|
| 2821 |
+
"step": 3800
|
| 2822 |
+
},
|
| 2823 |
+
{
|
| 2824 |
+
"epoch": 2.783815873595762,
|
| 2825 |
+
"grad_norm": 5.48701810836792,
|
| 2826 |
+
"learning_rate": 1.5980361398526267e-06,
|
| 2827 |
+
"loss": 2.1588,
|
| 2828 |
+
"step": 3810
|
| 2829 |
+
},
|
| 2830 |
+
{
|
| 2831 |
+
"epoch": 2.791122476938533,
|
| 2832 |
+
"grad_norm": 8.91283893585205,
|
| 2833 |
+
"learning_rate": 1.4931130104559154e-06,
|
| 2834 |
+
"loss": 1.8274,
|
| 2835 |
+
"step": 3820
|
| 2836 |
+
},
|
| 2837 |
+
{
|
| 2838 |
+
"epoch": 2.7984290802813043,
|
| 2839 |
+
"grad_norm": 7.846653938293457,
|
| 2840 |
+
"learning_rate": 1.3917001623146186e-06,
|
| 2841 |
+
"loss": 1.9029,
|
| 2842 |
+
"step": 3830
|
| 2843 |
+
},
|
| 2844 |
+
{
|
| 2845 |
+
"epoch": 2.805735683624075,
|
| 2846 |
+
"grad_norm": 5.458700656890869,
|
| 2847 |
+
"learning_rate": 1.2938049343375502e-06,
|
| 2848 |
+
"loss": 1.8208,
|
| 2849 |
+
"step": 3840
|
| 2850 |
+
},
|
| 2851 |
+
{
|
| 2852 |
+
"epoch": 2.813042286966846,
|
| 2853 |
+
"grad_norm": 6.338927745819092,
|
| 2854 |
+
"learning_rate": 1.1994344108750833e-06,
|
| 2855 |
+
"loss": 1.808,
|
| 2856 |
+
"step": 3850
|
| 2857 |
+
},
|
| 2858 |
+
{
|
| 2859 |
+
"epoch": 2.820348890309617,
|
| 2860 |
+
"grad_norm": 6.470076560974121,
|
| 2861 |
+
"learning_rate": 1.108595421206532e-06,
|
| 2862 |
+
"loss": 1.8503,
|
| 2863 |
+
"step": 3860
|
| 2864 |
+
},
|
| 2865 |
+
{
|
| 2866 |
+
"epoch": 2.8276554936523883,
|
| 2867 |
+
"grad_norm": 4.856273174285889,
|
| 2868 |
+
"learning_rate": 1.021294539045914e-06,
|
| 2869 |
+
"loss": 1.9037,
|
| 2870 |
+
"step": 3870
|
| 2871 |
+
},
|
| 2872 |
+
{
|
| 2873 |
+
"epoch": 2.8349620969951594,
|
| 2874 |
+
"grad_norm": 6.1774139404296875,
|
| 2875 |
+
"learning_rate": 9.375380820662194e-07,
|
| 2876 |
+
"loss": 1.8673,
|
| 2877 |
+
"step": 3880
|
| 2878 |
+
},
|
| 2879 |
+
{
|
| 2880 |
+
"epoch": 2.8422687003379306,
|
| 2881 |
+
"grad_norm": 5.6335296630859375,
|
| 2882 |
+
"learning_rate": 8.57332111442255e-07,
|
| 2883 |
+
"loss": 1.6094,
|
| 2884 |
+
"step": 3890
|
| 2885 |
+
},
|
| 2886 |
+
{
|
| 2887 |
+
"epoch": 2.8495753036807017,
|
| 2888 |
+
"grad_norm": 6.399806022644043,
|
| 2889 |
+
"learning_rate": 7.806824314119832e-07,
|
| 2890 |
+
"loss": 1.8876,
|
| 2891 |
+
"step": 3900
|
| 2892 |
+
},
|
| 2893 |
+
{
|
| 2894 |
+
"epoch": 2.8568819070234723,
|
| 2895 |
+
"grad_norm": 8.844847679138184,
|
| 2896 |
+
"learning_rate": 7.075945888565194e-07,
|
| 2897 |
+
"loss": 1.9354,
|
| 2898 |
+
"step": 3910
|
| 2899 |
+
},
|
| 2900 |
+
{
|
| 2901 |
+
"epoch": 2.8641885103662434,
|
| 2902 |
+
"grad_norm": 7.803627014160156,
|
| 2903 |
+
"learning_rate": 6.380738728986924e-07,
|
| 2904 |
+
"loss": 1.7139,
|
| 2905 |
+
"step": 3920
|
| 2906 |
+
},
|
| 2907 |
+
{
|
| 2908 |
+
"epoch": 2.8714951137090146,
|
| 2909 |
+
"grad_norm": 8.815073013305664,
|
| 2910 |
+
"learning_rate": 5.721253145203165e-07,
|
| 2911 |
+
"loss": 1.8449,
|
| 2912 |
+
"step": 3930
|
| 2913 |
+
},
|
| 2914 |
+
{
|
| 2915 |
+
"epoch": 2.8788017170517857,
|
| 2916 |
+
"grad_norm": 5.23136568069458,
|
| 2917 |
+
"learning_rate": 5.097536861981e-07,
|
| 2918 |
+
"loss": 1.774,
|
| 2919 |
+
"step": 3940
|
| 2920 |
+
},
|
| 2921 |
+
{
|
| 2922 |
+
"epoch": 2.8861083203945563,
|
| 2923 |
+
"grad_norm": 5.498495101928711,
|
| 2924 |
+
"learning_rate": 4.5096350155827693e-07,
|
| 2925 |
+
"loss": 1.7747,
|
| 2926 |
+
"step": 3950
|
| 2927 |
+
},
|
| 2928 |
+
{
|
| 2929 |
+
"epoch": 2.8934149237373274,
|
| 2930 |
+
"grad_norm": 6.538133144378662,
|
| 2931 |
+
"learning_rate": 3.957590150499735e-07,
|
| 2932 |
+
"loss": 1.8453,
|
| 2933 |
+
"step": 3960
|
| 2934 |
+
},
|
| 2935 |
+
{
|
| 2936 |
+
"epoch": 2.9007215270800986,
|
| 2937 |
+
"grad_norm": 6.259313106536865,
|
| 2938 |
+
"learning_rate": 3.441442216373436e-07,
|
| 2939 |
+
"loss": 1.9312,
|
| 2940 |
+
"step": 3970
|
| 2941 |
+
},
|
| 2942 |
+
{
|
| 2943 |
+
"epoch": 2.9080281304228697,
|
| 2944 |
+
"grad_norm": 5.1868486404418945,
|
| 2945 |
+
"learning_rate": 2.9612285651042795e-07,
|
| 2946 |
+
"loss": 1.8508,
|
| 2947 |
+
"step": 3980
|
| 2948 |
+
},
|
| 2949 |
+
{
|
| 2950 |
+
"epoch": 2.915334733765641,
|
| 2951 |
+
"grad_norm": 6.202200412750244,
|
| 2952 |
+
"learning_rate": 2.5169839481489764e-07,
|
| 2953 |
+
"loss": 1.8832,
|
| 2954 |
+
"step": 3990
|
| 2955 |
+
},
|
| 2956 |
+
{
|
| 2957 |
+
"epoch": 2.922641337108412,
|
| 2958 |
+
"grad_norm": 5.868015289306641,
|
| 2959 |
+
"learning_rate": 2.1087405140053362e-07,
|
| 2960 |
+
"loss": 1.7998,
|
| 2961 |
+
"step": 4000
|
| 2962 |
+
},
|
| 2963 |
+
{
|
| 2964 |
+
"epoch": 2.922641337108412,
|
| 2965 |
+
"eval_loss": 1.9779986143112183,
|
| 2966 |
+
"eval_runtime": 103.4275,
|
| 2967 |
+
"eval_samples_per_second": 11.767,
|
| 2968 |
+
"eval_steps_per_second": 11.767,
|
| 2969 |
+
"step": 4000
|
| 2970 |
+
},
|
| 2971 |
+
{
|
| 2972 |
+
"epoch": 2.929947940451183,
|
| 2973 |
+
"grad_norm": 9.322286605834961,
|
| 2974 |
+
"learning_rate": 1.736527805885957e-07,
|
| 2975 |
+
"loss": 1.8784,
|
| 2976 |
+
"step": 4010
|
| 2977 |
+
},
|
| 2978 |
+
{
|
| 2979 |
+
"epoch": 2.9372545437939537,
|
| 2980 |
+
"grad_norm": 5.733661651611328,
|
| 2981 |
+
"learning_rate": 1.4003727595802152e-07,
|
| 2982 |
+
"loss": 1.9243,
|
| 2983 |
+
"step": 4020
|
| 2984 |
+
},
|
| 2985 |
+
{
|
| 2986 |
+
"epoch": 2.944561147136725,
|
| 2987 |
+
"grad_norm": 6.326545238494873,
|
| 2988 |
+
"learning_rate": 1.1002997015050476e-07,
|
| 2989 |
+
"loss": 1.8568,
|
| 2990 |
+
"step": 4030
|
| 2991 |
+
},
|
| 2992 |
+
{
|
| 2993 |
+
"epoch": 2.951867750479496,
|
| 2994 |
+
"grad_norm": 5.118636608123779,
|
| 2995 |
+
"learning_rate": 8.363303469445805e-08,
|
| 2996 |
+
"loss": 1.7683,
|
| 2997 |
+
"step": 4040
|
| 2998 |
+
},
|
| 2999 |
+
{
|
| 3000 |
+
"epoch": 2.959174353822267,
|
| 3001 |
+
"grad_norm": 5.2632317543029785,
|
| 3002 |
+
"learning_rate": 6.084837984786096e-08,
|
| 3003 |
+
"loss": 1.9612,
|
| 3004 |
+
"step": 4050
|
| 3005 |
+
},
|
| 3006 |
+
{
|
| 3007 |
+
"epoch": 2.9664809571650377,
|
| 3008 |
+
"grad_norm": 5.4591898918151855,
|
| 3009 |
+
"learning_rate": 4.167765446000393e-08,
|
| 3010 |
+
"loss": 1.8417,
|
| 3011 |
+
"step": 4060
|
| 3012 |
+
},
|
| 3013 |
+
{
|
| 3014 |
+
"epoch": 2.973787560507809,
|
| 3015 |
+
"grad_norm": 6.642773628234863,
|
| 3016 |
+
"learning_rate": 2.6122245852205906e-08,
|
| 3017 |
+
"loss": 1.9892,
|
| 3018 |
+
"step": 4070
|
| 3019 |
+
},
|
| 3020 |
+
{
|
| 3021 |
+
"epoch": 2.98109416385058,
|
| 3022 |
+
"grad_norm": 5.449717998504639,
|
| 3023 |
+
"learning_rate": 1.4183279717389087e-08,
|
| 3024 |
+
"loss": 1.8122,
|
| 3025 |
+
"step": 4080
|
| 3026 |
+
},
|
| 3027 |
+
{
|
| 3028 |
+
"epoch": 2.988400767193351,
|
| 3029 |
+
"grad_norm": 5.309086322784424,
|
| 3030 |
+
"learning_rate": 5.861620038610794e-09,
|
| 3031 |
+
"loss": 1.7895,
|
| 3032 |
+
"step": 4090
|
| 3033 |
+
},
|
| 3034 |
+
{
|
| 3035 |
+
"epoch": 2.995707370536122,
|
| 3036 |
+
"grad_norm": 6.38864803314209,
|
| 3037 |
+
"learning_rate": 1.157869026574554e-09,
|
| 3038 |
+
"loss": 1.7208,
|
| 3039 |
+
"step": 4100
|
| 3040 |
+
},
|
| 3041 |
+
{
|
| 3042 |
+
"epoch": 2.9986300118732303,
|
| 3043 |
+
"step": 4104,
|
| 3044 |
+
"total_flos": 1.3798092749438976e+17,
|
| 3045 |
+
"train_loss": 2.123500373163651,
|
| 3046 |
+
"train_runtime": 11771.2816,
|
| 3047 |
+
"train_samples_per_second": 2.79,
|
| 3048 |
+
"train_steps_per_second": 0.349
|
| 3049 |
+
}
|
| 3050 |
+
],
|
| 3051 |
+
"logging_steps": 10,
|
| 3052 |
+
"max_steps": 4104,
|
| 3053 |
+
"num_input_tokens_seen": 0,
|
| 3054 |
+
"num_train_epochs": 3,
|
| 3055 |
+
"save_steps": 500,
|
| 3056 |
+
"total_flos": 1.3798092749438976e+17,
|
| 3057 |
+
"train_batch_size": 1,
|
| 3058 |
+
"trial_name": null,
|
| 3059 |
+
"trial_params": null
|
| 3060 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|