End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1974 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: global_batchsize_512_lradjusted8_v2
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# global_batchsize_512_lradjusted8_v2
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: global_batchsize_512_lradjusted8_v2
|
|
|
|
| 16 |
|
| 17 |
# global_batchsize_512_lradjusted8_v2
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the open-thoughts/OpenThoughts-114k dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9878213802435725,
|
| 3 |
+
"total_flos": 6.629121904995205e+18,
|
| 4 |
+
"train_loss": 0.463554704102917,
|
| 5 |
+
"train_runtime": 27205.7226,
|
| 6 |
+
"train_samples_per_second": 5.214,
|
| 7 |
+
"train_steps_per_second": 0.01
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9878213802435725,
|
| 3 |
+
"total_flos": 6.629121904995205e+18,
|
| 4 |
+
"train_loss": 0.463554704102917,
|
| 5 |
+
"train_runtime": 27205.7226,
|
| 6 |
+
"train_samples_per_second": 5.214,
|
| 7 |
+
"train_steps_per_second": 0.01
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1974 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9878213802435725,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 276,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.010825439783491205,
|
| 13 |
+
"grad_norm": 5.8079091029435865,
|
| 14 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 15 |
+
"loss": 0.8567,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02165087956698241,
|
| 20 |
+
"grad_norm": 5.811672220394839,
|
| 21 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 22 |
+
"loss": 0.8604,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03247631935047361,
|
| 27 |
+
"grad_norm": 5.449507701601476,
|
| 28 |
+
"learning_rate": 8.571428571428571e-06,
|
| 29 |
+
"loss": 0.8528,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04330175913396482,
|
| 34 |
+
"grad_norm": 2.3063006110984086,
|
| 35 |
+
"learning_rate": 1.1428571428571429e-05,
|
| 36 |
+
"loss": 0.776,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05412719891745602,
|
| 41 |
+
"grad_norm": 3.9302521280440543,
|
| 42 |
+
"learning_rate": 1.4285714285714287e-05,
|
| 43 |
+
"loss": 0.7546,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06495263870094722,
|
| 48 |
+
"grad_norm": 4.207181909155043,
|
| 49 |
+
"learning_rate": 1.7142857142857142e-05,
|
| 50 |
+
"loss": 0.751,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07577807848443843,
|
| 55 |
+
"grad_norm": 4.360918963292297,
|
| 56 |
+
"learning_rate": 2e-05,
|
| 57 |
+
"loss": 0.7344,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08660351826792964,
|
| 62 |
+
"grad_norm": 2.911250822571266,
|
| 63 |
+
"learning_rate": 2.2857142857142858e-05,
|
| 64 |
+
"loss": 0.729,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09742895805142084,
|
| 69 |
+
"grad_norm": 2.7455493318657966,
|
| 70 |
+
"learning_rate": 2.5714285714285718e-05,
|
| 71 |
+
"loss": 0.672,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.10825439783491204,
|
| 76 |
+
"grad_norm": 2.1241241512314795,
|
| 77 |
+
"learning_rate": 2.8571428571428574e-05,
|
| 78 |
+
"loss": 0.6659,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11907983761840325,
|
| 83 |
+
"grad_norm": 1.5668463996929993,
|
| 84 |
+
"learning_rate": 3.142857142857143e-05,
|
| 85 |
+
"loss": 0.6379,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.12990527740189445,
|
| 90 |
+
"grad_norm": 1.3619379997378058,
|
| 91 |
+
"learning_rate": 3.4285714285714284e-05,
|
| 92 |
+
"loss": 0.6305,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.14073071718538566,
|
| 97 |
+
"grad_norm": 1.566918593012661,
|
| 98 |
+
"learning_rate": 3.714285714285715e-05,
|
| 99 |
+
"loss": 0.6142,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.15155615696887687,
|
| 104 |
+
"grad_norm": 1.2044004334843796,
|
| 105 |
+
"learning_rate": 4e-05,
|
| 106 |
+
"loss": 0.6246,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.16238159675236807,
|
| 111 |
+
"grad_norm": 1.4087070173540086,
|
| 112 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 113 |
+
"loss": 0.6075,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.17320703653585928,
|
| 118 |
+
"grad_norm": 1.203436631182031,
|
| 119 |
+
"learning_rate": 4.5714285714285716e-05,
|
| 120 |
+
"loss": 0.5986,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.18403247631935046,
|
| 125 |
+
"grad_norm": 1.722190196328099,
|
| 126 |
+
"learning_rate": 4.857142857142857e-05,
|
| 127 |
+
"loss": 0.5943,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.19485791610284167,
|
| 132 |
+
"grad_norm": 1.1305378018180656,
|
| 133 |
+
"learning_rate": 5.1428571428571436e-05,
|
| 134 |
+
"loss": 0.577,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.20568335588633288,
|
| 139 |
+
"grad_norm": 1.6031033848905385,
|
| 140 |
+
"learning_rate": 5.4285714285714295e-05,
|
| 141 |
+
"loss": 0.5697,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2165087956698241,
|
| 146 |
+
"grad_norm": 1.6591800967410824,
|
| 147 |
+
"learning_rate": 5.714285714285715e-05,
|
| 148 |
+
"loss": 0.5746,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2273342354533153,
|
| 153 |
+
"grad_norm": 1.5025149200657286,
|
| 154 |
+
"learning_rate": 6.000000000000001e-05,
|
| 155 |
+
"loss": 0.5683,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.2381596752368065,
|
| 160 |
+
"grad_norm": 1.2019447493505488,
|
| 161 |
+
"learning_rate": 6.285714285714286e-05,
|
| 162 |
+
"loss": 0.5676,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2489851150202977,
|
| 167 |
+
"grad_norm": 1.9839452102168298,
|
| 168 |
+
"learning_rate": 6.571428571428571e-05,
|
| 169 |
+
"loss": 0.5639,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.2598105548037889,
|
| 174 |
+
"grad_norm": 1.34504393296771,
|
| 175 |
+
"learning_rate": 6.857142857142857e-05,
|
| 176 |
+
"loss": 0.5667,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.2706359945872801,
|
| 181 |
+
"grad_norm": 2.227974826889621,
|
| 182 |
+
"learning_rate": 7.142857142857143e-05,
|
| 183 |
+
"loss": 0.5639,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2814614343707713,
|
| 188 |
+
"grad_norm": 1.6397454252709596,
|
| 189 |
+
"learning_rate": 7.42857142857143e-05,
|
| 190 |
+
"loss": 0.5434,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.2922868741542625,
|
| 195 |
+
"grad_norm": 1.7236920014146049,
|
| 196 |
+
"learning_rate": 7.714285714285715e-05,
|
| 197 |
+
"loss": 0.5453,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.30311231393775373,
|
| 202 |
+
"grad_norm": 1.3679886100442236,
|
| 203 |
+
"learning_rate": 8e-05,
|
| 204 |
+
"loss": 0.5518,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.31393775372124494,
|
| 209 |
+
"grad_norm": 1.7636609095058786,
|
| 210 |
+
"learning_rate": 7.999679062421315e-05,
|
| 211 |
+
"loss": 0.5488,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.32476319350473615,
|
| 216 |
+
"grad_norm": 1.8909077915573775,
|
| 217 |
+
"learning_rate": 7.998716301185722e-05,
|
| 218 |
+
"loss": 0.5434,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.33558863328822736,
|
| 223 |
+
"grad_norm": 1.4831299671419245,
|
| 224 |
+
"learning_rate": 7.997111870786354e-05,
|
| 225 |
+
"loss": 0.543,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.34641407307171856,
|
| 230 |
+
"grad_norm": 1.5685011400813527,
|
| 231 |
+
"learning_rate": 7.994866028684212e-05,
|
| 232 |
+
"loss": 0.5386,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3572395128552097,
|
| 237 |
+
"grad_norm": 1.8163117507870878,
|
| 238 |
+
"learning_rate": 7.991979135266861e-05,
|
| 239 |
+
"loss": 0.5426,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3680649526387009,
|
| 244 |
+
"grad_norm": 1.3553007909847135,
|
| 245 |
+
"learning_rate": 7.98845165379059e-05,
|
| 246 |
+
"loss": 0.5419,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.37889039242219213,
|
| 251 |
+
"grad_norm": 1.5002520273898525,
|
| 252 |
+
"learning_rate": 7.984284150306085e-05,
|
| 253 |
+
"loss": 0.525,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.38971583220568334,
|
| 258 |
+
"grad_norm": 1.5788765506619407,
|
| 259 |
+
"learning_rate": 7.97947729356758e-05,
|
| 260 |
+
"loss": 0.5161,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.40054127198917455,
|
| 265 |
+
"grad_norm": 1.0932865412546702,
|
| 266 |
+
"learning_rate": 7.974031854925562e-05,
|
| 267 |
+
"loss": 0.5241,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.41136671177266576,
|
| 272 |
+
"grad_norm": 1.1169446884746776,
|
| 273 |
+
"learning_rate": 7.967948708202972e-05,
|
| 274 |
+
"loss": 0.5135,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.42219215155615697,
|
| 279 |
+
"grad_norm": 1.7368905888186483,
|
| 280 |
+
"learning_rate": 7.961228829555003e-05,
|
| 281 |
+
"loss": 0.5232,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4330175913396482,
|
| 286 |
+
"grad_norm": 1.14825749468684,
|
| 287 |
+
"learning_rate": 7.953873297312447e-05,
|
| 288 |
+
"loss": 0.5114,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.4438430311231394,
|
| 293 |
+
"grad_norm": 1.681948559743569,
|
| 294 |
+
"learning_rate": 7.945883291808655e-05,
|
| 295 |
+
"loss": 0.5151,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.4546684709066306,
|
| 300 |
+
"grad_norm": 1.1384144753190577,
|
| 301 |
+
"learning_rate": 7.937260095190137e-05,
|
| 302 |
+
"loss": 0.5079,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.4654939106901218,
|
| 307 |
+
"grad_norm": 1.6677218265436802,
|
| 308 |
+
"learning_rate": 7.928005091210817e-05,
|
| 309 |
+
"loss": 0.5128,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.476319350473613,
|
| 314 |
+
"grad_norm": 1.4237449938878564,
|
| 315 |
+
"learning_rate": 7.918119765009979e-05,
|
| 316 |
+
"loss": 0.511,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.4871447902571042,
|
| 321 |
+
"grad_norm": 1.3134667589020321,
|
| 322 |
+
"learning_rate": 7.907605702873948e-05,
|
| 323 |
+
"loss": 0.5101,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.4979702300405954,
|
| 328 |
+
"grad_norm": 1.0853939092704676,
|
| 329 |
+
"learning_rate": 7.896464591981549e-05,
|
| 330 |
+
"loss": 0.5086,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.5087956698240866,
|
| 335 |
+
"grad_norm": 1.6016710689451,
|
| 336 |
+
"learning_rate": 7.884698220133357e-05,
|
| 337 |
+
"loss": 0.5163,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5196211096075778,
|
| 342 |
+
"grad_norm": 1.203978171530331,
|
| 343 |
+
"learning_rate": 7.872308475464818e-05,
|
| 344 |
+
"loss": 0.5086,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.530446549391069,
|
| 349 |
+
"grad_norm": 1.3897099566312143,
|
| 350 |
+
"learning_rate": 7.859297346143258e-05,
|
| 351 |
+
"loss": 0.5051,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5412719891745602,
|
| 356 |
+
"grad_norm": 1.2903904022789845,
|
| 357 |
+
"learning_rate": 7.84566692004885e-05,
|
| 358 |
+
"loss": 0.5036,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5520974289580515,
|
| 363 |
+
"grad_norm": 0.9821064573534952,
|
| 364 |
+
"learning_rate": 7.831419384439565e-05,
|
| 365 |
+
"loss": 0.4979,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5629228687415426,
|
| 370 |
+
"grad_norm": 1.1967263169199975,
|
| 371 |
+
"learning_rate": 7.816557025600196e-05,
|
| 372 |
+
"loss": 0.4954,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.5737483085250338,
|
| 377 |
+
"grad_norm": 1.5118785206523027,
|
| 378 |
+
"learning_rate": 7.80108222847547e-05,
|
| 379 |
+
"loss": 0.4935,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.584573748308525,
|
| 384 |
+
"grad_norm": 1.328515593795635,
|
| 385 |
+
"learning_rate": 7.784997476287349e-05,
|
| 386 |
+
"loss": 0.5029,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5953991880920162,
|
| 391 |
+
"grad_norm": 1.085170937371965,
|
| 392 |
+
"learning_rate": 7.76830535013654e-05,
|
| 393 |
+
"loss": 0.4922,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.6062246278755075,
|
| 398 |
+
"grad_norm": 1.1543845766613074,
|
| 399 |
+
"learning_rate": 7.751008528588322e-05,
|
| 400 |
+
"loss": 0.4954,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6170500676589986,
|
| 405 |
+
"grad_norm": 1.2151109406542857,
|
| 406 |
+
"learning_rate": 7.733109787242708e-05,
|
| 407 |
+
"loss": 0.4981,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.6278755074424899,
|
| 412 |
+
"grad_norm": 1.6809116509174178,
|
| 413 |
+
"learning_rate": 7.71461199828905e-05,
|
| 414 |
+
"loss": 0.4992,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.638700947225981,
|
| 419 |
+
"grad_norm": 1.1929403575426991,
|
| 420 |
+
"learning_rate": 7.695518130045147e-05,
|
| 421 |
+
"loss": 0.4804,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.6495263870094723,
|
| 426 |
+
"grad_norm": 1.3373806033082491,
|
| 427 |
+
"learning_rate": 7.675831246480923e-05,
|
| 428 |
+
"loss": 0.4918,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6603518267929634,
|
| 433 |
+
"grad_norm": 1.1370461825495959,
|
| 434 |
+
"learning_rate": 7.655554506726747e-05,
|
| 435 |
+
"loss": 0.5003,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6711772665764547,
|
| 440 |
+
"grad_norm": 1.5541020268182568,
|
| 441 |
+
"learning_rate": 7.6346911645665e-05,
|
| 442 |
+
"loss": 0.4872,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6820027063599459,
|
| 447 |
+
"grad_norm": 1.0196032384903368,
|
| 448 |
+
"learning_rate": 7.61324456791544e-05,
|
| 449 |
+
"loss": 0.4971,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6928281461434371,
|
| 454 |
+
"grad_norm": 1.355258802840393,
|
| 455 |
+
"learning_rate": 7.591218158282968e-05,
|
| 456 |
+
"loss": 0.4961,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.7036535859269283,
|
| 461 |
+
"grad_norm": 1.5317006352800375,
|
| 462 |
+
"learning_rate": 7.568615470220369e-05,
|
| 463 |
+
"loss": 0.4836,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.7144790257104194,
|
| 468 |
+
"grad_norm": 0.9602544899213347,
|
| 469 |
+
"learning_rate": 7.545440130753634e-05,
|
| 470 |
+
"loss": 0.4934,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7253044654939107,
|
| 475 |
+
"grad_norm": 1.3937893813198252,
|
| 476 |
+
"learning_rate": 7.52169585880143e-05,
|
| 477 |
+
"loss": 0.492,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.7361299052774019,
|
| 482 |
+
"grad_norm": 1.2170568243121482,
|
| 483 |
+
"learning_rate": 7.497386464578329e-05,
|
| 484 |
+
"loss": 0.5004,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.7469553450608931,
|
| 489 |
+
"grad_norm": 1.215088332835355,
|
| 490 |
+
"learning_rate": 7.472515848983394e-05,
|
| 491 |
+
"loss": 0.4927,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.7577807848443843,
|
| 496 |
+
"grad_norm": 1.068568427145697,
|
| 497 |
+
"learning_rate": 7.447088002974199e-05,
|
| 498 |
+
"loss": 0.4735,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7686062246278755,
|
| 503 |
+
"grad_norm": 1.2666434730845704,
|
| 504 |
+
"learning_rate": 7.421107006926408e-05,
|
| 505 |
+
"loss": 0.4776,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.7794316644113667,
|
| 510 |
+
"grad_norm": 1.0693035758098317,
|
| 511 |
+
"learning_rate": 7.394577029979004e-05,
|
| 512 |
+
"loss": 0.4878,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.790257104194858,
|
| 517 |
+
"grad_norm": 1.5502809637898465,
|
| 518 |
+
"learning_rate": 7.367502329365268e-05,
|
| 519 |
+
"loss": 0.4949,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.8010825439783491,
|
| 524 |
+
"grad_norm": 0.8873154529201425,
|
| 525 |
+
"learning_rate": 7.33988724972963e-05,
|
| 526 |
+
"loss": 0.4889,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.8119079837618404,
|
| 531 |
+
"grad_norm": 1.27058950939853,
|
| 532 |
+
"learning_rate": 7.311736222430487e-05,
|
| 533 |
+
"loss": 0.4861,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.8227334235453315,
|
| 538 |
+
"grad_norm": 0.7866888980990746,
|
| 539 |
+
"learning_rate": 7.283053764829106e-05,
|
| 540 |
+
"loss": 0.4783,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.8335588633288228,
|
| 545 |
+
"grad_norm": 1.0938731157860102,
|
| 546 |
+
"learning_rate": 7.253844479564737e-05,
|
| 547 |
+
"loss": 0.4764,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.8443843031123139,
|
| 552 |
+
"grad_norm": 1.1807281415252031,
|
| 553 |
+
"learning_rate": 7.224113053816021e-05,
|
| 554 |
+
"loss": 0.4817,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.8552097428958051,
|
| 559 |
+
"grad_norm": 1.322179202967281,
|
| 560 |
+
"learning_rate": 7.193864258548855e-05,
|
| 561 |
+
"loss": 0.4853,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.8660351826792964,
|
| 566 |
+
"grad_norm": 1.2155369734676524,
|
| 567 |
+
"learning_rate": 7.163102947750794e-05,
|
| 568 |
+
"loss": 0.4796,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8768606224627875,
|
| 573 |
+
"grad_norm": 0.588289504862737,
|
| 574 |
+
"learning_rate": 7.131834057652142e-05,
|
| 575 |
+
"loss": 0.4863,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8876860622462788,
|
| 580 |
+
"grad_norm": 1.0985408806535708,
|
| 581 |
+
"learning_rate": 7.100062605933835e-05,
|
| 582 |
+
"loss": 0.4762,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8985115020297699,
|
| 587 |
+
"grad_norm": 1.74082517425262,
|
| 588 |
+
"learning_rate": 7.067793690922268e-05,
|
| 589 |
+
"loss": 0.4853,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.9093369418132612,
|
| 594 |
+
"grad_norm": 0.7638072853814706,
|
| 595 |
+
"learning_rate": 7.035032490771165e-05,
|
| 596 |
+
"loss": 0.4794,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.9201623815967523,
|
| 601 |
+
"grad_norm": 1.2164285440187212,
|
| 602 |
+
"learning_rate": 7.001784262630652e-05,
|
| 603 |
+
"loss": 0.48,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.9309878213802436,
|
| 608 |
+
"grad_norm": 1.2499859479078839,
|
| 609 |
+
"learning_rate": 6.968054341803644e-05,
|
| 610 |
+
"loss": 0.4853,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.9418132611637348,
|
| 615 |
+
"grad_norm": 1.1495652891036365,
|
| 616 |
+
"learning_rate": 6.933848140889705e-05,
|
| 617 |
+
"loss": 0.476,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.952638700947226,
|
| 622 |
+
"grad_norm": 1.1590588610177872,
|
| 623 |
+
"learning_rate": 6.89917114891648e-05,
|
| 624 |
+
"loss": 0.4653,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.9634641407307172,
|
| 629 |
+
"grad_norm": 0.8043870196059654,
|
| 630 |
+
"learning_rate": 6.864028930458892e-05,
|
| 631 |
+
"loss": 0.4788,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.9742895805142084,
|
| 636 |
+
"grad_norm": 0.8931105851691834,
|
| 637 |
+
"learning_rate": 6.828427124746191e-05,
|
| 638 |
+
"loss": 0.4776,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9851150202976996,
|
| 643 |
+
"grad_norm": 1.2474558559687694,
|
| 644 |
+
"learning_rate": 6.792371444757037e-05,
|
| 645 |
+
"loss": 0.4803,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9959404600811907,
|
| 650 |
+
"grad_norm": 0.8560601402041761,
|
| 651 |
+
"learning_rate": 6.755867676302747e-05,
|
| 652 |
+
"loss": 0.4798,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.006765899864682,
|
| 657 |
+
"grad_norm": 2.0148609986627783,
|
| 658 |
+
"learning_rate": 6.718921677098853e-05,
|
| 659 |
+
"loss": 0.7491,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.0175913396481733,
|
| 664 |
+
"grad_norm": 0.6667855075813555,
|
| 665 |
+
"learning_rate": 6.681539375825115e-05,
|
| 666 |
+
"loss": 0.4535,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.0284167794316643,
|
| 671 |
+
"grad_norm": 0.9162841549268478,
|
| 672 |
+
"learning_rate": 6.643726771174164e-05,
|
| 673 |
+
"loss": 0.455,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.0392422192151556,
|
| 678 |
+
"grad_norm": 1.0155165692532704,
|
| 679 |
+
"learning_rate": 6.60548993088889e-05,
|
| 680 |
+
"loss": 0.454,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.0500676589986468,
|
| 685 |
+
"grad_norm": 1.078599089633044,
|
| 686 |
+
"learning_rate": 6.56683499078876e-05,
|
| 687 |
+
"loss": 0.4455,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.060893098782138,
|
| 692 |
+
"grad_norm": 1.1351986372223781,
|
| 693 |
+
"learning_rate": 6.527768153785216e-05,
|
| 694 |
+
"loss": 0.4541,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.0717185385656292,
|
| 699 |
+
"grad_norm": 0.7315670031697065,
|
| 700 |
+
"learning_rate": 6.488295688886295e-05,
|
| 701 |
+
"loss": 0.4479,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.0825439783491204,
|
| 706 |
+
"grad_norm": 1.3582463875830837,
|
| 707 |
+
"learning_rate": 6.448423930190653e-05,
|
| 708 |
+
"loss": 0.4486,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.0933694181326117,
|
| 713 |
+
"grad_norm": 0.8080345177396737,
|
| 714 |
+
"learning_rate": 6.408159275871132e-05,
|
| 715 |
+
"loss": 0.446,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.104194857916103,
|
| 720 |
+
"grad_norm": 1.160068769869394,
|
| 721 |
+
"learning_rate": 6.36750818714807e-05,
|
| 722 |
+
"loss": 0.4471,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.115020297699594,
|
| 727 |
+
"grad_norm": 0.8897562278529675,
|
| 728 |
+
"learning_rate": 6.326477187252455e-05,
|
| 729 |
+
"loss": 0.449,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.1258457374830853,
|
| 734 |
+
"grad_norm": 1.0199996165037868,
|
| 735 |
+
"learning_rate": 6.28507286037917e-05,
|
| 736 |
+
"loss": 0.4427,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.1366711772665765,
|
| 741 |
+
"grad_norm": 0.7249873315321067,
|
| 742 |
+
"learning_rate": 6.243301850630419e-05,
|
| 743 |
+
"loss": 0.4453,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.1474966170500678,
|
| 748 |
+
"grad_norm": 1.189445348682187,
|
| 749 |
+
"learning_rate": 6.201170860949565e-05,
|
| 750 |
+
"loss": 0.4515,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.1583220568335588,
|
| 755 |
+
"grad_norm": 1.1424188944548885,
|
| 756 |
+
"learning_rate": 6.15868665204552e-05,
|
| 757 |
+
"loss": 0.4452,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.16914749661705,
|
| 762 |
+
"grad_norm": 0.8857107534333528,
|
| 763 |
+
"learning_rate": 6.11585604130785e-05,
|
| 764 |
+
"loss": 0.4393,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.1799729364005414,
|
| 769 |
+
"grad_norm": 0.7495223764013124,
|
| 770 |
+
"learning_rate": 6.072685901712808e-05,
|
| 771 |
+
"loss": 0.4446,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.1907983761840324,
|
| 776 |
+
"grad_norm": 0.6658761551232241,
|
| 777 |
+
"learning_rate": 6.02918316072043e-05,
|
| 778 |
+
"loss": 0.4499,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.2016238159675237,
|
| 783 |
+
"grad_norm": 0.6905937205369403,
|
| 784 |
+
"learning_rate": 5.9853547991628967e-05,
|
| 785 |
+
"loss": 0.4477,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.212449255751015,
|
| 790 |
+
"grad_norm": 0.6381399917783538,
|
| 791 |
+
"learning_rate": 5.941207850124325e-05,
|
| 792 |
+
"loss": 0.4421,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.2232746955345062,
|
| 797 |
+
"grad_norm": 0.7541941618755881,
|
| 798 |
+
"learning_rate": 5.896749397812181e-05,
|
| 799 |
+
"loss": 0.4382,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.2341001353179972,
|
| 804 |
+
"grad_norm": 0.5923805845803851,
|
| 805 |
+
"learning_rate": 5.8519865764204834e-05,
|
| 806 |
+
"loss": 0.4459,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.2449255751014885,
|
| 811 |
+
"grad_norm": 0.6109388260249382,
|
| 812 |
+
"learning_rate": 5.8069265689849884e-05,
|
| 813 |
+
"loss": 0.4513,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.2557510148849798,
|
| 818 |
+
"grad_norm": 0.533980708840664,
|
| 819 |
+
"learning_rate": 5.761576606230538e-05,
|
| 820 |
+
"loss": 0.443,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.266576454668471,
|
| 825 |
+
"grad_norm": 0.7052629150640184,
|
| 826 |
+
"learning_rate": 5.7159439654107506e-05,
|
| 827 |
+
"loss": 0.4404,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.277401894451962,
|
| 832 |
+
"grad_norm": 0.6111064530168573,
|
| 833 |
+
"learning_rate": 5.6700359691402533e-05,
|
| 834 |
+
"loss": 0.4402,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.2882273342354533,
|
| 839 |
+
"grad_norm": 0.5508566705964758,
|
| 840 |
+
"learning_rate": 5.6238599842196285e-05,
|
| 841 |
+
"loss": 0.4449,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.2990527740189446,
|
| 846 |
+
"grad_norm": 0.6028745516198664,
|
| 847 |
+
"learning_rate": 5.5774234204532746e-05,
|
| 848 |
+
"loss": 0.4365,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.3098782138024356,
|
| 853 |
+
"grad_norm": 0.721324745203013,
|
| 854 |
+
"learning_rate": 5.5307337294603595e-05,
|
| 855 |
+
"loss": 0.4442,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.320703653585927,
|
| 860 |
+
"grad_norm": 0.9341313496474372,
|
| 861 |
+
"learning_rate": 5.483798403479072e-05,
|
| 862 |
+
"loss": 0.4403,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.3315290933694182,
|
| 867 |
+
"grad_norm": 1.0476784062313165,
|
| 868 |
+
"learning_rate": 5.436624974164349e-05,
|
| 869 |
+
"loss": 0.4434,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.3423545331529092,
|
| 874 |
+
"grad_norm": 0.917250294000449,
|
| 875 |
+
"learning_rate": 5.389221011379281e-05,
|
| 876 |
+
"loss": 0.4409,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.3531799729364005,
|
| 881 |
+
"grad_norm": 0.7818649176171457,
|
| 882 |
+
"learning_rate": 5.3415941219803895e-05,
|
| 883 |
+
"loss": 0.4405,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.3640054127198917,
|
| 888 |
+
"grad_norm": 0.5772276458708397,
|
| 889 |
+
"learning_rate": 5.2937519485969525e-05,
|
| 890 |
+
"loss": 0.4397,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.374830852503383,
|
| 895 |
+
"grad_norm": 0.4347968698038521,
|
| 896 |
+
"learning_rate": 5.245702168404616e-05,
|
| 897 |
+
"loss": 0.4325,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.3856562922868743,
|
| 902 |
+
"grad_norm": 0.49507391797694716,
|
| 903 |
+
"learning_rate": 5.1974524918934336e-05,
|
| 904 |
+
"loss": 0.4353,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.3964817320703653,
|
| 909 |
+
"grad_norm": 0.49182716841159074,
|
| 910 |
+
"learning_rate": 5.14901066163058e-05,
|
| 911 |
+
"loss": 0.4341,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.4073071718538566,
|
| 916 |
+
"grad_norm": 0.5209355239170891,
|
| 917 |
+
"learning_rate": 5.1003844510179126e-05,
|
| 918 |
+
"loss": 0.4391,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.4181326116373478,
|
| 923 |
+
"grad_norm": 0.555660282513782,
|
| 924 |
+
"learning_rate": 5.0515816630445795e-05,
|
| 925 |
+
"loss": 0.4408,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.4289580514208389,
|
| 930 |
+
"grad_norm": 0.4906371694319118,
|
| 931 |
+
"learning_rate": 5.002610129034883e-05,
|
| 932 |
+
"loss": 0.442,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.4397834912043301,
|
| 937 |
+
"grad_norm": 0.39564350538031545,
|
| 938 |
+
"learning_rate": 4.953477707391597e-05,
|
| 939 |
+
"loss": 0.4359,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.4506089309878214,
|
| 944 |
+
"grad_norm": 0.36420248532962185,
|
| 945 |
+
"learning_rate": 4.90419228233494e-05,
|
| 946 |
+
"loss": 0.4371,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.4614343707713127,
|
| 951 |
+
"grad_norm": 0.3693636826961334,
|
| 952 |
+
"learning_rate": 4.854761762637403e-05,
|
| 953 |
+
"loss": 0.4375,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.472259810554804,
|
| 958 |
+
"grad_norm": 0.37103016749538226,
|
| 959 |
+
"learning_rate": 4.805194080354641e-05,
|
| 960 |
+
"loss": 0.4367,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.483085250338295,
|
| 965 |
+
"grad_norm": 0.4634070182199332,
|
| 966 |
+
"learning_rate": 4.7554971895526175e-05,
|
| 967 |
+
"loss": 0.434,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.4939106901217862,
|
| 972 |
+
"grad_norm": 0.39637839232896416,
|
| 973 |
+
"learning_rate": 4.705679065031235e-05,
|
| 974 |
+
"loss": 0.4405,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.5047361299052775,
|
| 979 |
+
"grad_norm": 0.4167931396685767,
|
| 980 |
+
"learning_rate": 4.6557477010446206e-05,
|
| 981 |
+
"loss": 0.439,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.5155615696887685,
|
| 986 |
+
"grad_norm": 0.34078264652362333,
|
| 987 |
+
"learning_rate": 4.605711110018307e-05,
|
| 988 |
+
"loss": 0.4383,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.5263870094722598,
|
| 993 |
+
"grad_norm": 0.326633599627453,
|
| 994 |
+
"learning_rate": 4.555577321263477e-05,
|
| 995 |
+
"loss": 0.4344,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.537212449255751,
|
| 1000 |
+
"grad_norm": 0.3976277159519867,
|
| 1001 |
+
"learning_rate": 4.505354379688518e-05,
|
| 1002 |
+
"loss": 0.4354,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.548037889039242,
|
| 1007 |
+
"grad_norm": 0.4357572785065164,
|
| 1008 |
+
"learning_rate": 4.4550503445080606e-05,
|
| 1009 |
+
"loss": 0.4417,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.5588633288227334,
|
| 1014 |
+
"grad_norm": 0.3916962827302459,
|
| 1015 |
+
"learning_rate": 4.4046732879497295e-05,
|
| 1016 |
+
"loss": 0.4341,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.5696887686062246,
|
| 1021 |
+
"grad_norm": 0.30012283877463697,
|
| 1022 |
+
"learning_rate": 4.354231293958801e-05,
|
| 1023 |
+
"loss": 0.4322,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.5805142083897157,
|
| 1028 |
+
"grad_norm": 0.2885107465168404,
|
| 1029 |
+
"learning_rate": 4.3037324569009854e-05,
|
| 1030 |
+
"loss": 0.4323,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.5913396481732072,
|
| 1035 |
+
"grad_norm": 0.26624048862887445,
|
| 1036 |
+
"learning_rate": 4.2531848802635264e-05,
|
| 1037 |
+
"loss": 0.4278,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.6021650879566982,
|
| 1042 |
+
"grad_norm": 0.3506386895842403,
|
| 1043 |
+
"learning_rate": 4.202596675354851e-05,
|
| 1044 |
+
"loss": 0.4479,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.6129905277401895,
|
| 1049 |
+
"grad_norm": 0.40487113874803043,
|
| 1050 |
+
"learning_rate": 4.151975960002958e-05,
|
| 1051 |
+
"loss": 0.4302,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.6238159675236807,
|
| 1056 |
+
"grad_norm": 0.23297174487170752,
|
| 1057 |
+
"learning_rate": 4.101330857252752e-05,
|
| 1058 |
+
"loss": 0.4367,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.6346414073071718,
|
| 1063 |
+
"grad_norm": 0.30369826324347926,
|
| 1064 |
+
"learning_rate": 4.050669494062561e-05,
|
| 1065 |
+
"loss": 0.4372,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.645466847090663,
|
| 1070 |
+
"grad_norm": 0.29979293086538217,
|
| 1071 |
+
"learning_rate": 4e-05,
|
| 1072 |
+
"loss": 0.4308,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.6562922868741543,
|
| 1077 |
+
"grad_norm": 0.21998478305579725,
|
| 1078 |
+
"learning_rate": 3.9493305059374405e-05,
|
| 1079 |
+
"loss": 0.444,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.6671177266576453,
|
| 1084 |
+
"grad_norm": 0.2822111397888722,
|
| 1085 |
+
"learning_rate": 3.8986691427472496e-05,
|
| 1086 |
+
"loss": 0.4362,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.6779431664411368,
|
| 1091 |
+
"grad_norm": 0.19298090698642797,
|
| 1092 |
+
"learning_rate": 3.8480240399970436e-05,
|
| 1093 |
+
"loss": 0.4291,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.6887686062246279,
|
| 1098 |
+
"grad_norm": 0.24201705060461937,
|
| 1099 |
+
"learning_rate": 3.7974033246451496e-05,
|
| 1100 |
+
"loss": 0.4374,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.699594046008119,
|
| 1105 |
+
"grad_norm": 0.2926866609659229,
|
| 1106 |
+
"learning_rate": 3.746815119736475e-05,
|
| 1107 |
+
"loss": 0.4237,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.7104194857916104,
|
| 1112 |
+
"grad_norm": 0.22986191257503155,
|
| 1113 |
+
"learning_rate": 3.696267543099016e-05,
|
| 1114 |
+
"loss": 0.4349,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.7212449255751014,
|
| 1119 |
+
"grad_norm": 0.21762455575624456,
|
| 1120 |
+
"learning_rate": 3.6457687060412e-05,
|
| 1121 |
+
"loss": 0.4411,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.7320703653585927,
|
| 1126 |
+
"grad_norm": 0.23006031819487235,
|
| 1127 |
+
"learning_rate": 3.595326712050272e-05,
|
| 1128 |
+
"loss": 0.4352,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.742895805142084,
|
| 1133 |
+
"grad_norm": 0.23480749792102354,
|
| 1134 |
+
"learning_rate": 3.5449496554919414e-05,
|
| 1135 |
+
"loss": 0.432,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.753721244925575,
|
| 1140 |
+
"grad_norm": 0.20514925914229476,
|
| 1141 |
+
"learning_rate": 3.494645620311484e-05,
|
| 1142 |
+
"loss": 0.4374,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.7645466847090663,
|
| 1147 |
+
"grad_norm": 0.21723286316559756,
|
| 1148 |
+
"learning_rate": 3.444422678736525e-05,
|
| 1149 |
+
"loss": 0.4309,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.7753721244925575,
|
| 1154 |
+
"grad_norm": 0.19908292246838455,
|
| 1155 |
+
"learning_rate": 3.394288889981695e-05,
|
| 1156 |
+
"loss": 0.4285,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.7861975642760486,
|
| 1161 |
+
"grad_norm": 0.21848824155386468,
|
| 1162 |
+
"learning_rate": 3.34425229895538e-05,
|
| 1163 |
+
"loss": 0.4358,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.79702300405954,
|
| 1168 |
+
"grad_norm": 0.19073388047330828,
|
| 1169 |
+
"learning_rate": 3.294320934968768e-05,
|
| 1170 |
+
"loss": 0.4285,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.8078484438430311,
|
| 1175 |
+
"grad_norm": 0.20593306151831323,
|
| 1176 |
+
"learning_rate": 3.2445028104473845e-05,
|
| 1177 |
+
"loss": 0.4347,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.8186738836265224,
|
| 1182 |
+
"grad_norm": 0.21441680855669412,
|
| 1183 |
+
"learning_rate": 3.194805919645359e-05,
|
| 1184 |
+
"loss": 0.4308,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.8294993234100136,
|
| 1189 |
+
"grad_norm": 0.18002633127557865,
|
| 1190 |
+
"learning_rate": 3.145238237362596e-05,
|
| 1191 |
+
"loss": 0.4392,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.8403247631935047,
|
| 1196 |
+
"grad_norm": 0.18961769159187056,
|
| 1197 |
+
"learning_rate": 3.0958077176650606e-05,
|
| 1198 |
+
"loss": 0.43,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.851150202976996,
|
| 1203 |
+
"grad_norm": 0.16827003419417288,
|
| 1204 |
+
"learning_rate": 3.0465222926084036e-05,
|
| 1205 |
+
"loss": 0.426,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.8619756427604872,
|
| 1210 |
+
"grad_norm": 0.1766724359697219,
|
| 1211 |
+
"learning_rate": 2.997389870965118e-05,
|
| 1212 |
+
"loss": 0.4273,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.8728010825439783,
|
| 1217 |
+
"grad_norm": 0.1937265928321859,
|
| 1218 |
+
"learning_rate": 2.948418336955421e-05,
|
| 1219 |
+
"loss": 0.4216,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.8836265223274695,
|
| 1224 |
+
"grad_norm": 0.1733839063792691,
|
| 1225 |
+
"learning_rate": 2.899615548982088e-05,
|
| 1226 |
+
"loss": 0.4342,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.8944519621109608,
|
| 1231 |
+
"grad_norm": 0.1691834042941349,
|
| 1232 |
+
"learning_rate": 2.8509893383694213e-05,
|
| 1233 |
+
"loss": 0.428,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.9052774018944518,
|
| 1238 |
+
"grad_norm": 0.17556950089441328,
|
| 1239 |
+
"learning_rate": 2.8025475081065684e-05,
|
| 1240 |
+
"loss": 0.4235,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.9161028416779433,
|
| 1245 |
+
"grad_norm": 0.17292630455576483,
|
| 1246 |
+
"learning_rate": 2.754297831595385e-05,
|
| 1247 |
+
"loss": 0.4242,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.9269282814614344,
|
| 1252 |
+
"grad_norm": 0.203395158544133,
|
| 1253 |
+
"learning_rate": 2.7062480514030478e-05,
|
| 1254 |
+
"loss": 0.4346,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.9377537212449256,
|
| 1259 |
+
"grad_norm": 0.17761313930923656,
|
| 1260 |
+
"learning_rate": 2.658405878019612e-05,
|
| 1261 |
+
"loss": 0.4284,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.9485791610284169,
|
| 1266 |
+
"grad_norm": 0.1857053642993675,
|
| 1267 |
+
"learning_rate": 2.6107789886207195e-05,
|
| 1268 |
+
"loss": 0.4311,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.959404600811908,
|
| 1273 |
+
"grad_norm": 0.168623231091507,
|
| 1274 |
+
"learning_rate": 2.563375025835652e-05,
|
| 1275 |
+
"loss": 0.4296,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.9702300405953992,
|
| 1280 |
+
"grad_norm": 0.16395731003065203,
|
| 1281 |
+
"learning_rate": 2.5162015965209295e-05,
|
| 1282 |
+
"loss": 0.4263,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.9810554803788905,
|
| 1287 |
+
"grad_norm": 0.17813784128811913,
|
| 1288 |
+
"learning_rate": 2.4692662705396412e-05,
|
| 1289 |
+
"loss": 0.4337,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.9918809201623815,
|
| 1294 |
+
"grad_norm": 0.1692020847630486,
|
| 1295 |
+
"learning_rate": 2.4225765795467267e-05,
|
| 1296 |
+
"loss": 0.4345,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.002706359945873,
|
| 1301 |
+
"grad_norm": 0.29642981794747025,
|
| 1302 |
+
"learning_rate": 2.376140015780372e-05,
|
| 1303 |
+
"loss": 0.6828,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.013531799729364,
|
| 1308 |
+
"grad_norm": 0.22282508513365898,
|
| 1309 |
+
"learning_rate": 2.3299640308597487e-05,
|
| 1310 |
+
"loss": 0.4033,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.024357239512855,
|
| 1315 |
+
"grad_norm": 0.19906807218157607,
|
| 1316 |
+
"learning_rate": 2.2840560345892518e-05,
|
| 1317 |
+
"loss": 0.4028,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.0351826792963466,
|
| 1322 |
+
"grad_norm": 0.21438465882623423,
|
| 1323 |
+
"learning_rate": 2.2384233937694626e-05,
|
| 1324 |
+
"loss": 0.4036,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.0460081190798376,
|
| 1329 |
+
"grad_norm": 0.20912970610685283,
|
| 1330 |
+
"learning_rate": 2.1930734310150116e-05,
|
| 1331 |
+
"loss": 0.3975,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.0568335588633286,
|
| 1336 |
+
"grad_norm": 0.2005177294111497,
|
| 1337 |
+
"learning_rate": 2.1480134235795173e-05,
|
| 1338 |
+
"loss": 0.4022,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.06765899864682,
|
| 1343 |
+
"grad_norm": 0.19782562683985525,
|
| 1344 |
+
"learning_rate": 2.10325060218782e-05,
|
| 1345 |
+
"loss": 0.4073,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.078484438430311,
|
| 1350 |
+
"grad_norm": 0.23611436091693633,
|
| 1351 |
+
"learning_rate": 2.0587921498756768e-05,
|
| 1352 |
+
"loss": 0.4109,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.089309878213802,
|
| 1357 |
+
"grad_norm": 0.20974660868469683,
|
| 1358 |
+
"learning_rate": 2.014645200837105e-05,
|
| 1359 |
+
"loss": 0.4034,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.1001353179972937,
|
| 1364 |
+
"grad_norm": 0.19267649854044877,
|
| 1365 |
+
"learning_rate": 1.9708168392795718e-05,
|
| 1366 |
+
"loss": 0.3953,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.1109607577807847,
|
| 1371 |
+
"grad_norm": 0.22011389396639292,
|
| 1372 |
+
"learning_rate": 1.9273140982871936e-05,
|
| 1373 |
+
"loss": 0.4047,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.121786197564276,
|
| 1378 |
+
"grad_norm": 0.18180878911713266,
|
| 1379 |
+
"learning_rate": 1.8841439586921515e-05,
|
| 1380 |
+
"loss": 0.4039,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.1326116373477673,
|
| 1385 |
+
"grad_norm": 0.21012423766700125,
|
| 1386 |
+
"learning_rate": 1.841313347954482e-05,
|
| 1387 |
+
"loss": 0.3986,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.1434370771312583,
|
| 1392 |
+
"grad_norm": 0.15463640007730903,
|
| 1393 |
+
"learning_rate": 1.7988291390504348e-05,
|
| 1394 |
+
"loss": 0.3975,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.15426251691475,
|
| 1399 |
+
"grad_norm": 0.193505600575409,
|
| 1400 |
+
"learning_rate": 1.7566981493695828e-05,
|
| 1401 |
+
"loss": 0.4043,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.165087956698241,
|
| 1406 |
+
"grad_norm": 0.16774935171850072,
|
| 1407 |
+
"learning_rate": 1.71492713962083e-05,
|
| 1408 |
+
"loss": 0.4068,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.175913396481732,
|
| 1413 |
+
"grad_norm": 0.15543728413541416,
|
| 1414 |
+
"learning_rate": 1.673522812747544e-05,
|
| 1415 |
+
"loss": 0.4025,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.1867388362652234,
|
| 1420 |
+
"grad_norm": 0.1581064576242491,
|
| 1421 |
+
"learning_rate": 1.6324918128519306e-05,
|
| 1422 |
+
"loss": 0.3995,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.1975642760487144,
|
| 1427 |
+
"grad_norm": 0.1433403956609895,
|
| 1428 |
+
"learning_rate": 1.5918407241288678e-05,
|
| 1429 |
+
"loss": 0.406,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.208389715832206,
|
| 1434 |
+
"grad_norm": 0.1444921358019384,
|
| 1435 |
+
"learning_rate": 1.5515760698093485e-05,
|
| 1436 |
+
"loss": 0.4043,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.219215155615697,
|
| 1441 |
+
"grad_norm": 0.14781491117947443,
|
| 1442 |
+
"learning_rate": 1.511704311113705e-05,
|
| 1443 |
+
"loss": 0.4024,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.230040595399188,
|
| 1448 |
+
"grad_norm": 0.1394579893993049,
|
| 1449 |
+
"learning_rate": 1.4722318462147844e-05,
|
| 1450 |
+
"loss": 0.4029,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 2.2408660351826795,
|
| 1455 |
+
"grad_norm": 0.16023643196582224,
|
| 1456 |
+
"learning_rate": 1.4331650092112406e-05,
|
| 1457 |
+
"loss": 0.4066,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.2516914749661705,
|
| 1462 |
+
"grad_norm": 0.1348210060414343,
|
| 1463 |
+
"learning_rate": 1.394510069111112e-05,
|
| 1464 |
+
"loss": 0.3973,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 2.2625169147496615,
|
| 1469 |
+
"grad_norm": 0.13391825126021842,
|
| 1470 |
+
"learning_rate": 1.3562732288258377e-05,
|
| 1471 |
+
"loss": 0.4025,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.273342354533153,
|
| 1476 |
+
"grad_norm": 0.12844012269731564,
|
| 1477 |
+
"learning_rate": 1.3184606241748857e-05,
|
| 1478 |
+
"loss": 0.3978,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 2.284167794316644,
|
| 1483 |
+
"grad_norm": 0.13584285882845534,
|
| 1484 |
+
"learning_rate": 1.2810783229011486e-05,
|
| 1485 |
+
"loss": 0.3991,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.2949932341001356,
|
| 1490 |
+
"grad_norm": 0.13168511677095765,
|
| 1491 |
+
"learning_rate": 1.2441323236972536e-05,
|
| 1492 |
+
"loss": 0.3999,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 2.3058186738836266,
|
| 1497 |
+
"grad_norm": 0.12398647003241701,
|
| 1498 |
+
"learning_rate": 1.2076285552429642e-05,
|
| 1499 |
+
"loss": 0.4077,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.3166441136671176,
|
| 1504 |
+
"grad_norm": 0.13229640626265612,
|
| 1505 |
+
"learning_rate": 1.1715728752538103e-05,
|
| 1506 |
+
"loss": 0.4019,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 2.3274695534506087,
|
| 1511 |
+
"grad_norm": 0.12385054713595073,
|
| 1512 |
+
"learning_rate": 1.1359710695411086e-05,
|
| 1513 |
+
"loss": 0.4029,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 2.3382949932341,
|
| 1518 |
+
"grad_norm": 0.12591732770348577,
|
| 1519 |
+
"learning_rate": 1.100828851083521e-05,
|
| 1520 |
+
"loss": 0.4084,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 2.349120433017591,
|
| 1525 |
+
"grad_norm": 0.13251886738270005,
|
| 1526 |
+
"learning_rate": 1.0661518591102973e-05,
|
| 1527 |
+
"loss": 0.4019,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.3599458728010827,
|
| 1532 |
+
"grad_norm": 0.1231996291813589,
|
| 1533 |
+
"learning_rate": 1.0319456581963578e-05,
|
| 1534 |
+
"loss": 0.4097,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 2.3707713125845737,
|
| 1539 |
+
"grad_norm": 0.12007287837624477,
|
| 1540 |
+
"learning_rate": 9.982157373693502e-06,
|
| 1541 |
+
"loss": 0.407,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.381596752368065,
|
| 1546 |
+
"grad_norm": 0.11828968531594988,
|
| 1547 |
+
"learning_rate": 9.649675092288366e-06,
|
| 1548 |
+
"loss": 0.3942,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 2.3924221921515563,
|
| 1553 |
+
"grad_norm": 0.1266852929005505,
|
| 1554 |
+
"learning_rate": 9.322063090777331e-06,
|
| 1555 |
+
"loss": 0.402,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 2.4032476319350473,
|
| 1560 |
+
"grad_norm": 0.13131525920519638,
|
| 1561 |
+
"learning_rate": 8.99937394066165e-06,
|
| 1562 |
+
"loss": 0.4079,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 2.414073071718539,
|
| 1567 |
+
"grad_norm": 0.1276300518950864,
|
| 1568 |
+
"learning_rate": 8.681659423478587e-06,
|
| 1569 |
+
"loss": 0.4075,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.42489851150203,
|
| 1574 |
+
"grad_norm": 0.1284596882094962,
|
| 1575 |
+
"learning_rate": 8.368970522492064e-06,
|
| 1576 |
+
"loss": 0.4041,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 2.435723951285521,
|
| 1581 |
+
"grad_norm": 0.11600707488242912,
|
| 1582 |
+
"learning_rate": 8.06135741451146e-06,
|
| 1583 |
+
"loss": 0.4024,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 2.4465493910690124,
|
| 1588 |
+
"grad_norm": 0.1369413509292135,
|
| 1589 |
+
"learning_rate": 7.758869461839808e-06,
|
| 1590 |
+
"loss": 0.4054,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 2.4573748308525034,
|
| 1595 |
+
"grad_norm": 0.11693298201832159,
|
| 1596 |
+
"learning_rate": 7.461555204352655e-06,
|
| 1597 |
+
"loss": 0.3988,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 2.4682002706359945,
|
| 1602 |
+
"grad_norm": 0.11962758908233623,
|
| 1603 |
+
"learning_rate": 7.169462351708958e-06,
|
| 1604 |
+
"loss": 0.4007,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 2.479025710419486,
|
| 1609 |
+
"grad_norm": 0.12081712635114038,
|
| 1610 |
+
"learning_rate": 6.882637775695147e-06,
|
| 1611 |
+
"loss": 0.3973,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.489851150202977,
|
| 1616 |
+
"grad_norm": 0.12426870098020382,
|
| 1617 |
+
"learning_rate": 6.60112750270371e-06,
|
| 1618 |
+
"loss": 0.3985,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.500676589986468,
|
| 1623 |
+
"grad_norm": 0.11586754894597827,
|
| 1624 |
+
"learning_rate": 6.324976706347317e-06,
|
| 1625 |
+
"loss": 0.4078,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.5115020297699595,
|
| 1630 |
+
"grad_norm": 0.10751212068030966,
|
| 1631 |
+
"learning_rate": 6.054229700209959e-06,
|
| 1632 |
+
"loss": 0.4054,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.5223274695534506,
|
| 1637 |
+
"grad_norm": 0.12026857943267745,
|
| 1638 |
+
"learning_rate": 5.788929930735916e-06,
|
| 1639 |
+
"loss": 0.4,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.533152909336942,
|
| 1644 |
+
"grad_norm": 0.10350592956190559,
|
| 1645 |
+
"learning_rate": 5.529119970258014e-06,
|
| 1646 |
+
"loss": 0.4032,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 2.543978349120433,
|
| 1651 |
+
"grad_norm": 0.09850020851661254,
|
| 1652 |
+
"learning_rate": 5.274841510166062e-06,
|
| 1653 |
+
"loss": 0.3961,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.554803788903924,
|
| 1658 |
+
"grad_norm": 0.10692798146184619,
|
| 1659 |
+
"learning_rate": 5.026135354216717e-06,
|
| 1660 |
+
"loss": 0.3953,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.565629228687415,
|
| 1665 |
+
"grad_norm": 0.1026156051139519,
|
| 1666 |
+
"learning_rate": 4.783041411985716e-06,
|
| 1667 |
+
"loss": 0.4026,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.5764546684709067,
|
| 1672 |
+
"grad_norm": 0.0984956835805336,
|
| 1673 |
+
"learning_rate": 4.545598692463675e-06,
|
| 1674 |
+
"loss": 0.4012,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.5872801082543977,
|
| 1679 |
+
"grad_norm": 0.09517297299245513,
|
| 1680 |
+
"learning_rate": 4.3138452977963266e-06,
|
| 1681 |
+
"loss": 0.3986,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.598105548037889,
|
| 1686 |
+
"grad_norm": 0.1017472403907421,
|
| 1687 |
+
"learning_rate": 4.087818417170337e-06,
|
| 1688 |
+
"loss": 0.4054,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 2.60893098782138,
|
| 1693 |
+
"grad_norm": 0.10517463959416935,
|
| 1694 |
+
"learning_rate": 3.867554320845601e-06,
|
| 1695 |
+
"loss": 0.4081,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.6197564276048713,
|
| 1700 |
+
"grad_norm": 0.10283676657115687,
|
| 1701 |
+
"learning_rate": 3.6530883543350038e-06,
|
| 1702 |
+
"loss": 0.4078,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.6305818673883627,
|
| 1707 |
+
"grad_norm": 0.09668185967839843,
|
| 1708 |
+
"learning_rate": 3.4444549327325325e-06,
|
| 1709 |
+
"loss": 0.3986,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.641407307171854,
|
| 1714 |
+
"grad_norm": 0.10655772797882479,
|
| 1715 |
+
"learning_rate": 3.241687535190776e-06,
|
| 1716 |
+
"loss": 0.3971,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.6522327469553453,
|
| 1721 |
+
"grad_norm": 0.10412740990228317,
|
| 1722 |
+
"learning_rate": 3.0448186995485307e-06,
|
| 1723 |
+
"loss": 0.398,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.6630581867388363,
|
| 1728 |
+
"grad_norm": 0.10294803068291954,
|
| 1729 |
+
"learning_rate": 2.853880017109516e-06,
|
| 1730 |
+
"loss": 0.3984,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.6738836265223274,
|
| 1735 |
+
"grad_norm": 0.09605814302194737,
|
| 1736 |
+
"learning_rate": 2.6689021275729366e-06,
|
| 1737 |
+
"loss": 0.4045,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.6847090663058184,
|
| 1742 |
+
"grad_norm": 0.10355709761757277,
|
| 1743 |
+
"learning_rate": 2.489914714116788e-06,
|
| 1744 |
+
"loss": 0.3993,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.69553450608931,
|
| 1749 |
+
"grad_norm": 0.10534682127661901,
|
| 1750 |
+
"learning_rate": 2.316946498634605e-06,
|
| 1751 |
+
"loss": 0.396,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.706359945872801,
|
| 1756 |
+
"grad_norm": 0.09850383904197658,
|
| 1757 |
+
"learning_rate": 2.1500252371265253e-06,
|
| 1758 |
+
"loss": 0.4071,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.7171853856562924,
|
| 1763 |
+
"grad_norm": 0.09796274629692603,
|
| 1764 |
+
"learning_rate": 1.989177715245307e-06,
|
| 1765 |
+
"loss": 0.4029,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.7280108254397835,
|
| 1770 |
+
"grad_norm": 0.09270281667503455,
|
| 1771 |
+
"learning_rate": 1.8344297439980475e-06,
|
| 1772 |
+
"loss": 0.4067,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.7388362652232745,
|
| 1777 |
+
"grad_norm": 0.09318564757274785,
|
| 1778 |
+
"learning_rate": 1.685806155604346e-06,
|
| 1779 |
+
"loss": 0.4055,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.749661705006766,
|
| 1784 |
+
"grad_norm": 0.09361381405899726,
|
| 1785 |
+
"learning_rate": 1.5433307995115043e-06,
|
| 1786 |
+
"loss": 0.3995,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.760487144790257,
|
| 1791 |
+
"grad_norm": 0.08807467149477492,
|
| 1792 |
+
"learning_rate": 1.4070265385674176e-06,
|
| 1793 |
+
"loss": 0.3949,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.7713125845737485,
|
| 1798 |
+
"grad_norm": 0.09403613096147612,
|
| 1799 |
+
"learning_rate": 1.276915245351833e-06,
|
| 1800 |
+
"loss": 0.3936,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.7821380243572396,
|
| 1805 |
+
"grad_norm": 0.0872925419115015,
|
| 1806 |
+
"learning_rate": 1.1530177986664425e-06,
|
| 1807 |
+
"loss": 0.4077,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.7929634641407306,
|
| 1812 |
+
"grad_norm": 0.0850031850974251,
|
| 1813 |
+
"learning_rate": 1.0353540801845229e-06,
|
| 1814 |
+
"loss": 0.4003,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.803788903924222,
|
| 1819 |
+
"grad_norm": 0.08692284082323248,
|
| 1820 |
+
"learning_rate": 9.239429712605274e-07,
|
| 1821 |
+
"loss": 0.3991,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.814614343707713,
|
| 1826 |
+
"grad_norm": 0.08257477054022927,
|
| 1827 |
+
"learning_rate": 8.188023499002206e-07,
|
| 1828 |
+
"loss": 0.4046,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.825439783491204,
|
| 1833 |
+
"grad_norm": 0.08278787367986447,
|
| 1834 |
+
"learning_rate": 7.199490878918314e-07,
|
| 1835 |
+
"loss": 0.3939,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.8362652232746957,
|
| 1840 |
+
"grad_norm": 0.08297388622794061,
|
| 1841 |
+
"learning_rate": 6.273990480986314e-07,
|
| 1842 |
+
"loss": 0.3993,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.8470906630581867,
|
| 1847 |
+
"grad_norm": 0.08598446782763781,
|
| 1848 |
+
"learning_rate": 5.411670819134651e-07,
|
| 1849 |
+
"loss": 0.4034,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.8579161028416777,
|
| 1854 |
+
"grad_norm": 0.08779257361395829,
|
| 1855 |
+
"learning_rate": 4.6126702687554483e-07,
|
| 1856 |
+
"loss": 0.399,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.8687415426251692,
|
| 1861 |
+
"grad_norm": 0.08649244863382963,
|
| 1862 |
+
"learning_rate": 3.8771170444996895e-07,
|
| 1863 |
+
"loss": 0.4036,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.8795669824086603,
|
| 1868 |
+
"grad_norm": 0.08480882436189274,
|
| 1869 |
+
"learning_rate": 3.2051291797027925e-07,
|
| 1870 |
+
"loss": 0.402,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.8903924221921518,
|
| 1875 |
+
"grad_norm": 0.08449641777806831,
|
| 1876 |
+
"learning_rate": 2.59681450744389e-07,
|
| 1877 |
+
"loss": 0.3975,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.901217861975643,
|
| 1882 |
+
"grad_norm": 0.08224385199164784,
|
| 1883 |
+
"learning_rate": 2.0522706432419382e-07,
|
| 1884 |
+
"loss": 0.3989,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.912043301759134,
|
| 1889 |
+
"grad_norm": 0.08340221321505677,
|
| 1890 |
+
"learning_rate": 1.5715849693916264e-07,
|
| 1891 |
+
"loss": 0.4009,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.9228687415426253,
|
| 1896 |
+
"grad_norm": 0.08271264876685368,
|
| 1897 |
+
"learning_rate": 1.1548346209410366e-07,
|
| 1898 |
+
"loss": 0.4023,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.9336941813261164,
|
| 1903 |
+
"grad_norm": 0.08260005569786469,
|
| 1904 |
+
"learning_rate": 8.020864733140343e-08,
|
| 1905 |
+
"loss": 0.3993,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.944519621109608,
|
| 1910 |
+
"grad_norm": 0.08278469542802738,
|
| 1911 |
+
"learning_rate": 5.133971315788966e-08,
|
| 1912 |
+
"loss": 0.4053,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.955345060893099,
|
| 1917 |
+
"grad_norm": 0.08624728365001678,
|
| 1918 |
+
"learning_rate": 2.8881292136468952e-08,
|
| 1919 |
+
"loss": 0.4024,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 2.96617050067659,
|
| 1924 |
+
"grad_norm": 0.08795666384161806,
|
| 1925 |
+
"learning_rate": 1.2836988142779228e-08,
|
| 1926 |
+
"loss": 0.4015,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 2.976995940460081,
|
| 1931 |
+
"grad_norm": 0.08192972934685559,
|
| 1932 |
+
"learning_rate": 3.209375786856761e-09,
|
| 1933 |
+
"loss": 0.4094,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.9878213802435725,
|
| 1938 |
+
"grad_norm": 0.07989052292080247,
|
| 1939 |
+
"learning_rate": 0.0,
|
| 1940 |
+
"loss": 0.3967,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.9878213802435725,
|
| 1945 |
+
"step": 276,
|
| 1946 |
+
"total_flos": 6.629121904995205e+18,
|
| 1947 |
+
"train_loss": 0.463554704102917,
|
| 1948 |
+
"train_runtime": 27205.7226,
|
| 1949 |
+
"train_samples_per_second": 5.214,
|
| 1950 |
+
"train_steps_per_second": 0.01
|
| 1951 |
+
}
|
| 1952 |
+
],
|
| 1953 |
+
"logging_steps": 1.0,
|
| 1954 |
+
"max_steps": 276,
|
| 1955 |
+
"num_input_tokens_seen": 0,
|
| 1956 |
+
"num_train_epochs": 3,
|
| 1957 |
+
"save_steps": 500,
|
| 1958 |
+
"stateful_callbacks": {
|
| 1959 |
+
"TrainerControl": {
|
| 1960 |
+
"args": {
|
| 1961 |
+
"should_epoch_stop": false,
|
| 1962 |
+
"should_evaluate": false,
|
| 1963 |
+
"should_log": false,
|
| 1964 |
+
"should_save": true,
|
| 1965 |
+
"should_training_stop": true
|
| 1966 |
+
},
|
| 1967 |
+
"attributes": {}
|
| 1968 |
+
}
|
| 1969 |
+
},
|
| 1970 |
+
"total_flos": 6.629121904995205e+18,
|
| 1971 |
+
"train_batch_size": 1,
|
| 1972 |
+
"trial_name": null,
|
| 1973 |
+
"trial_params": null
|
| 1974 |
+
}
|
training_loss.png
ADDED
|