End of training
Browse files- README.md +15 -2
- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +1476 -0
README.md
CHANGED
|
@@ -1,11 +1,24 @@
|
|
| 1 |
---
|
|
|
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
|
|
|
|
|
|
| 4 |
metrics:
|
| 5 |
- accuracy
|
| 6 |
model-index:
|
| 7 |
- name: stage2_v5
|
| 8 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -13,7 +26,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# stage2_v5
|
| 15 |
|
| 16 |
-
This model
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
- Loss: 0.2448
|
| 19 |
- Accuracy: 0.9165
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model: linker_v5/stage1
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- jarod0411/linker_v3
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
| 9 |
model-index:
|
| 10 |
- name: stage2_v5
|
| 11 |
+
results:
|
| 12 |
+
- task:
|
| 13 |
+
name: Causal Language Modeling
|
| 14 |
+
type: text-generation
|
| 15 |
+
dataset:
|
| 16 |
+
name: jarod0411/linker_v3
|
| 17 |
+
type: jarod0411/linker_v3
|
| 18 |
+
metrics:
|
| 19 |
+
- name: Accuracy
|
| 20 |
+
type: accuracy
|
| 21 |
+
value: 0.916456858900755
|
| 22 |
---
|
| 23 |
|
| 24 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 26 |
|
| 27 |
# stage2_v5
|
| 28 |
|
| 29 |
+
This model is a fine-tuned version of [linker_v5/stage1](https://huggingface.co/linker_v5/stage1) on the jarod0411/linker_v3 dataset.
|
| 30 |
It achieves the following results on the evaluation set:
|
| 31 |
- Loss: 0.2448
|
| 32 |
- Accuracy: 0.9165
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.916456858900755,
|
| 4 |
+
"eval_loss": 0.24475741386413574,
|
| 5 |
+
"eval_runtime": 763.108,
|
| 6 |
+
"eval_samples": 241166,
|
| 7 |
+
"eval_samples_per_second": 316.031,
|
| 8 |
+
"eval_steps_per_second": 1.647,
|
| 9 |
+
"perplexity": 1.2773114175771296,
|
| 10 |
+
"train_loss": 0.27495618251225495,
|
| 11 |
+
"train_runtime": 96949.7276,
|
| 12 |
+
"train_samples": 2170356,
|
| 13 |
+
"train_samples_per_second": 223.864,
|
| 14 |
+
"train_steps_per_second": 1.166
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.916456858900755,
|
| 4 |
+
"eval_loss": 0.24475741386413574,
|
| 5 |
+
"eval_runtime": 763.108,
|
| 6 |
+
"eval_samples": 241166,
|
| 7 |
+
"eval_samples_per_second": 316.031,
|
| 8 |
+
"eval_steps_per_second": 1.647,
|
| 9 |
+
"perplexity": 1.2773114175771296
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"train_loss": 0.27495618251225495,
|
| 4 |
+
"train_runtime": 96949.7276,
|
| 5 |
+
"train_samples": 2170356,
|
| 6 |
+
"train_samples_per_second": 223.864,
|
| 7 |
+
"train_steps_per_second": 1.166
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1476 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 113040,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.04,
|
| 13 |
+
"learning_rate": 4.977883934890305e-05,
|
| 14 |
+
"loss": 0.4833,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.09,
|
| 19 |
+
"learning_rate": 4.9557678697806086e-05,
|
| 20 |
+
"loss": 0.4355,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.13,
|
| 25 |
+
"learning_rate": 4.933651804670913e-05,
|
| 26 |
+
"loss": 0.4192,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.18,
|
| 31 |
+
"learning_rate": 4.9115357395612176e-05,
|
| 32 |
+
"loss": 0.4073,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.22,
|
| 37 |
+
"learning_rate": 4.889419674451522e-05,
|
| 38 |
+
"loss": 0.3977,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.27,
|
| 43 |
+
"learning_rate": 4.867303609341826e-05,
|
| 44 |
+
"loss": 0.3892,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.31,
|
| 49 |
+
"learning_rate": 4.8451875442321304e-05,
|
| 50 |
+
"loss": 0.382,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.35,
|
| 55 |
+
"learning_rate": 4.823071479122435e-05,
|
| 56 |
+
"loss": 0.3749,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.4,
|
| 61 |
+
"learning_rate": 4.800955414012739e-05,
|
| 62 |
+
"loss": 0.3688,
|
| 63 |
+
"step": 4500
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.44,
|
| 67 |
+
"learning_rate": 4.778839348903044e-05,
|
| 68 |
+
"loss": 0.363,
|
| 69 |
+
"step": 5000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.49,
|
| 73 |
+
"learning_rate": 4.756723283793348e-05,
|
| 74 |
+
"loss": 0.3576,
|
| 75 |
+
"step": 5500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.53,
|
| 79 |
+
"learning_rate": 4.734607218683652e-05,
|
| 80 |
+
"loss": 0.3529,
|
| 81 |
+
"step": 6000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.58,
|
| 85 |
+
"learning_rate": 4.712491153573956e-05,
|
| 86 |
+
"loss": 0.3481,
|
| 87 |
+
"step": 6500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.62,
|
| 91 |
+
"learning_rate": 4.6903750884642606e-05,
|
| 92 |
+
"loss": 0.3437,
|
| 93 |
+
"step": 7000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.66,
|
| 97 |
+
"learning_rate": 4.668259023354565e-05,
|
| 98 |
+
"loss": 0.3394,
|
| 99 |
+
"step": 7500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.71,
|
| 103 |
+
"learning_rate": 4.6461429582448696e-05,
|
| 104 |
+
"loss": 0.3358,
|
| 105 |
+
"step": 8000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.75,
|
| 109 |
+
"learning_rate": 4.624026893135174e-05,
|
| 110 |
+
"loss": 0.3323,
|
| 111 |
+
"step": 8500
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.8,
|
| 115 |
+
"learning_rate": 4.601910828025478e-05,
|
| 116 |
+
"loss": 0.3287,
|
| 117 |
+
"step": 9000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.84,
|
| 121 |
+
"learning_rate": 4.5797947629157824e-05,
|
| 122 |
+
"loss": 0.3257,
|
| 123 |
+
"step": 9500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.88,
|
| 127 |
+
"learning_rate": 4.557678697806086e-05,
|
| 128 |
+
"loss": 0.3227,
|
| 129 |
+
"step": 10000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.93,
|
| 133 |
+
"learning_rate": 4.535562632696391e-05,
|
| 134 |
+
"loss": 0.3202,
|
| 135 |
+
"step": 10500
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.97,
|
| 139 |
+
"learning_rate": 4.513446567586695e-05,
|
| 140 |
+
"loss": 0.3173,
|
| 141 |
+
"step": 11000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 1.0,
|
| 145 |
+
"eval_accuracy": 0.9059693079835405,
|
| 146 |
+
"eval_loss": 0.2837291359901428,
|
| 147 |
+
"eval_runtime": 770.0213,
|
| 148 |
+
"eval_samples_per_second": 313.194,
|
| 149 |
+
"eval_steps_per_second": 1.632,
|
| 150 |
+
"step": 11304
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.02,
|
| 154 |
+
"learning_rate": 4.491330502477e-05,
|
| 155 |
+
"loss": 0.3145,
|
| 156 |
+
"step": 11500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.06,
|
| 160 |
+
"learning_rate": 4.469214437367304e-05,
|
| 161 |
+
"loss": 0.3116,
|
| 162 |
+
"step": 12000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.11,
|
| 166 |
+
"learning_rate": 4.447098372257608e-05,
|
| 167 |
+
"loss": 0.3098,
|
| 168 |
+
"step": 12500
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.15,
|
| 172 |
+
"learning_rate": 4.4249823071479126e-05,
|
| 173 |
+
"loss": 0.3078,
|
| 174 |
+
"step": 13000
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.19,
|
| 178 |
+
"learning_rate": 4.4028662420382164e-05,
|
| 179 |
+
"loss": 0.306,
|
| 180 |
+
"step": 13500
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 1.24,
|
| 184 |
+
"learning_rate": 4.380750176928521e-05,
|
| 185 |
+
"loss": 0.3041,
|
| 186 |
+
"step": 14000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.28,
|
| 190 |
+
"learning_rate": 4.3586341118188254e-05,
|
| 191 |
+
"loss": 0.3024,
|
| 192 |
+
"step": 14500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.33,
|
| 196 |
+
"learning_rate": 4.33651804670913e-05,
|
| 197 |
+
"loss": 0.3008,
|
| 198 |
+
"step": 15000
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.37,
|
| 202 |
+
"learning_rate": 4.3144019815994344e-05,
|
| 203 |
+
"loss": 0.2992,
|
| 204 |
+
"step": 15500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.42,
|
| 208 |
+
"learning_rate": 4.292285916489738e-05,
|
| 209 |
+
"loss": 0.2979,
|
| 210 |
+
"step": 16000
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.46,
|
| 214 |
+
"learning_rate": 4.270169851380043e-05,
|
| 215 |
+
"loss": 0.2965,
|
| 216 |
+
"step": 16500
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.5,
|
| 220 |
+
"learning_rate": 4.2480537862703466e-05,
|
| 221 |
+
"loss": 0.295,
|
| 222 |
+
"step": 17000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 1.55,
|
| 226 |
+
"learning_rate": 4.225937721160652e-05,
|
| 227 |
+
"loss": 0.2939,
|
| 228 |
+
"step": 17500
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.59,
|
| 232 |
+
"learning_rate": 4.2038216560509556e-05,
|
| 233 |
+
"loss": 0.2927,
|
| 234 |
+
"step": 18000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.64,
|
| 238 |
+
"learning_rate": 4.18170559094126e-05,
|
| 239 |
+
"loss": 0.2915,
|
| 240 |
+
"step": 18500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.68,
|
| 244 |
+
"learning_rate": 4.1595895258315646e-05,
|
| 245 |
+
"loss": 0.2906,
|
| 246 |
+
"step": 19000
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.73,
|
| 250 |
+
"learning_rate": 4.1374734607218684e-05,
|
| 251 |
+
"loss": 0.2894,
|
| 252 |
+
"step": 19500
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.77,
|
| 256 |
+
"learning_rate": 4.115357395612173e-05,
|
| 257 |
+
"loss": 0.2883,
|
| 258 |
+
"step": 20000
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 1.81,
|
| 262 |
+
"learning_rate": 4.093241330502477e-05,
|
| 263 |
+
"loss": 0.2875,
|
| 264 |
+
"step": 20500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.86,
|
| 268 |
+
"learning_rate": 4.071125265392782e-05,
|
| 269 |
+
"loss": 0.2866,
|
| 270 |
+
"step": 21000
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.9,
|
| 274 |
+
"learning_rate": 4.049009200283086e-05,
|
| 275 |
+
"loss": 0.2856,
|
| 276 |
+
"step": 21500
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.95,
|
| 280 |
+
"learning_rate": 4.02689313517339e-05,
|
| 281 |
+
"loss": 0.2849,
|
| 282 |
+
"step": 22000
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.99,
|
| 286 |
+
"learning_rate": 4.004777070063695e-05,
|
| 287 |
+
"loss": 0.284,
|
| 288 |
+
"step": 22500
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.0,
|
| 292 |
+
"eval_accuracy": 0.9123534174226251,
|
| 293 |
+
"eval_loss": 0.2612881660461426,
|
| 294 |
+
"eval_runtime": 772.6863,
|
| 295 |
+
"eval_samples_per_second": 312.114,
|
| 296 |
+
"eval_steps_per_second": 1.627,
|
| 297 |
+
"step": 22608
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.03,
|
| 301 |
+
"learning_rate": 3.9826610049539986e-05,
|
| 302 |
+
"loss": 0.283,
|
| 303 |
+
"step": 23000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.08,
|
| 307 |
+
"learning_rate": 3.960544939844303e-05,
|
| 308 |
+
"loss": 0.2821,
|
| 309 |
+
"step": 23500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.12,
|
| 313 |
+
"learning_rate": 3.9384288747346076e-05,
|
| 314 |
+
"loss": 0.2814,
|
| 315 |
+
"step": 24000
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 2.17,
|
| 319 |
+
"learning_rate": 3.916312809624912e-05,
|
| 320 |
+
"loss": 0.2808,
|
| 321 |
+
"step": 24500
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.21,
|
| 325 |
+
"learning_rate": 3.894196744515216e-05,
|
| 326 |
+
"loss": 0.2804,
|
| 327 |
+
"step": 25000
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 2.26,
|
| 331 |
+
"learning_rate": 3.8720806794055204e-05,
|
| 332 |
+
"loss": 0.2795,
|
| 333 |
+
"step": 25500
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.3,
|
| 337 |
+
"learning_rate": 3.849964614295825e-05,
|
| 338 |
+
"loss": 0.2788,
|
| 339 |
+
"step": 26000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.34,
|
| 343 |
+
"learning_rate": 3.827848549186129e-05,
|
| 344 |
+
"loss": 0.2782,
|
| 345 |
+
"step": 26500
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.39,
|
| 349 |
+
"learning_rate": 3.805732484076434e-05,
|
| 350 |
+
"loss": 0.2779,
|
| 351 |
+
"step": 27000
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 2.43,
|
| 355 |
+
"learning_rate": 3.783616418966738e-05,
|
| 356 |
+
"loss": 0.2772,
|
| 357 |
+
"step": 27500
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 2.48,
|
| 361 |
+
"learning_rate": 3.761500353857042e-05,
|
| 362 |
+
"loss": 0.2767,
|
| 363 |
+
"step": 28000
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.52,
|
| 367 |
+
"learning_rate": 3.739384288747346e-05,
|
| 368 |
+
"loss": 0.2762,
|
| 369 |
+
"step": 28500
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.57,
|
| 373 |
+
"learning_rate": 3.7172682236376506e-05,
|
| 374 |
+
"loss": 0.2759,
|
| 375 |
+
"step": 29000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.61,
|
| 379 |
+
"learning_rate": 3.6951521585279544e-05,
|
| 380 |
+
"loss": 0.2755,
|
| 381 |
+
"step": 29500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.65,
|
| 385 |
+
"learning_rate": 3.673036093418259e-05,
|
| 386 |
+
"loss": 0.2749,
|
| 387 |
+
"step": 30000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 2.7,
|
| 391 |
+
"learning_rate": 3.650920028308564e-05,
|
| 392 |
+
"loss": 0.2744,
|
| 393 |
+
"step": 30500
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 2.74,
|
| 397 |
+
"learning_rate": 3.628803963198868e-05,
|
| 398 |
+
"loss": 0.2742,
|
| 399 |
+
"step": 31000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.79,
|
| 403 |
+
"learning_rate": 3.6066878980891724e-05,
|
| 404 |
+
"loss": 0.2738,
|
| 405 |
+
"step": 31500
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.83,
|
| 409 |
+
"learning_rate": 3.584571832979476e-05,
|
| 410 |
+
"loss": 0.2732,
|
| 411 |
+
"step": 32000
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 2.88,
|
| 415 |
+
"learning_rate": 3.562455767869781e-05,
|
| 416 |
+
"loss": 0.2729,
|
| 417 |
+
"step": 32500
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 2.92,
|
| 421 |
+
"learning_rate": 3.5403397027600845e-05,
|
| 422 |
+
"loss": 0.2724,
|
| 423 |
+
"step": 33000
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.96,
|
| 427 |
+
"learning_rate": 3.51822363765039e-05,
|
| 428 |
+
"loss": 0.2719,
|
| 429 |
+
"step": 33500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 3.0,
|
| 433 |
+
"eval_accuracy": 0.9140334727013555,
|
| 434 |
+
"eval_loss": 0.254567414522171,
|
| 435 |
+
"eval_runtime": 778.5722,
|
| 436 |
+
"eval_samples_per_second": 309.754,
|
| 437 |
+
"eval_steps_per_second": 1.614,
|
| 438 |
+
"step": 33912
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 3.01,
|
| 442 |
+
"learning_rate": 3.4961075725406935e-05,
|
| 443 |
+
"loss": 0.2716,
|
| 444 |
+
"step": 34000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.05,
|
| 448 |
+
"learning_rate": 3.473991507430998e-05,
|
| 449 |
+
"loss": 0.2709,
|
| 450 |
+
"step": 34500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 3.1,
|
| 454 |
+
"learning_rate": 3.4518754423213026e-05,
|
| 455 |
+
"loss": 0.2705,
|
| 456 |
+
"step": 35000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.14,
|
| 460 |
+
"learning_rate": 3.4297593772116064e-05,
|
| 461 |
+
"loss": 0.2701,
|
| 462 |
+
"step": 35500
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 3.18,
|
| 466 |
+
"learning_rate": 3.407643312101911e-05,
|
| 467 |
+
"loss": 0.2701,
|
| 468 |
+
"step": 36000
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 3.23,
|
| 472 |
+
"learning_rate": 3.3855272469922154e-05,
|
| 473 |
+
"loss": 0.2696,
|
| 474 |
+
"step": 36500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 3.27,
|
| 478 |
+
"learning_rate": 3.36341118188252e-05,
|
| 479 |
+
"loss": 0.2694,
|
| 480 |
+
"step": 37000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 3.32,
|
| 484 |
+
"learning_rate": 3.341295116772824e-05,
|
| 485 |
+
"loss": 0.269,
|
| 486 |
+
"step": 37500
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 3.36,
|
| 490 |
+
"learning_rate": 3.319179051663128e-05,
|
| 491 |
+
"loss": 0.269,
|
| 492 |
+
"step": 38000
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 3.41,
|
| 496 |
+
"learning_rate": 3.297062986553433e-05,
|
| 497 |
+
"loss": 0.2685,
|
| 498 |
+
"step": 38500
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.45,
|
| 502 |
+
"learning_rate": 3.2749469214437365e-05,
|
| 503 |
+
"loss": 0.2683,
|
| 504 |
+
"step": 39000
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 3.49,
|
| 508 |
+
"learning_rate": 3.252830856334041e-05,
|
| 509 |
+
"loss": 0.2679,
|
| 510 |
+
"step": 39500
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 3.54,
|
| 514 |
+
"learning_rate": 3.2307147912243455e-05,
|
| 515 |
+
"loss": 0.2679,
|
| 516 |
+
"step": 40000
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 3.58,
|
| 520 |
+
"learning_rate": 3.20859872611465e-05,
|
| 521 |
+
"loss": 0.2676,
|
| 522 |
+
"step": 40500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 3.63,
|
| 526 |
+
"learning_rate": 3.186482661004954e-05,
|
| 527 |
+
"loss": 0.2673,
|
| 528 |
+
"step": 41000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 3.67,
|
| 532 |
+
"learning_rate": 3.1643665958952584e-05,
|
| 533 |
+
"loss": 0.2669,
|
| 534 |
+
"step": 41500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 3.72,
|
| 538 |
+
"learning_rate": 3.142250530785563e-05,
|
| 539 |
+
"loss": 0.267,
|
| 540 |
+
"step": 42000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 3.76,
|
| 544 |
+
"learning_rate": 3.120134465675867e-05,
|
| 545 |
+
"loss": 0.2667,
|
| 546 |
+
"step": 42500
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 3.8,
|
| 550 |
+
"learning_rate": 3.098018400566172e-05,
|
| 551 |
+
"loss": 0.2664,
|
| 552 |
+
"step": 43000
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 3.85,
|
| 556 |
+
"learning_rate": 3.075902335456476e-05,
|
| 557 |
+
"loss": 0.2662,
|
| 558 |
+
"step": 43500
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 3.89,
|
| 562 |
+
"learning_rate": 3.05378627034678e-05,
|
| 563 |
+
"loss": 0.2658,
|
| 564 |
+
"step": 44000
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 3.94,
|
| 568 |
+
"learning_rate": 3.0316702052370844e-05,
|
| 569 |
+
"loss": 0.2656,
|
| 570 |
+
"step": 44500
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 3.98,
|
| 574 |
+
"learning_rate": 3.0095541401273885e-05,
|
| 575 |
+
"loss": 0.2654,
|
| 576 |
+
"step": 45000
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 4.0,
|
| 580 |
+
"eval_accuracy": 0.9148443813733191,
|
| 581 |
+
"eval_loss": 0.25113189220428467,
|
| 582 |
+
"eval_runtime": 770.8647,
|
| 583 |
+
"eval_samples_per_second": 312.851,
|
| 584 |
+
"eval_steps_per_second": 1.631,
|
| 585 |
+
"step": 45216
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 4.03,
|
| 589 |
+
"learning_rate": 2.9874380750176927e-05,
|
| 590 |
+
"loss": 0.2652,
|
| 591 |
+
"step": 45500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.07,
|
| 595 |
+
"learning_rate": 2.9653220099079975e-05,
|
| 596 |
+
"loss": 0.2648,
|
| 597 |
+
"step": 46000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 4.11,
|
| 601 |
+
"learning_rate": 2.9432059447983017e-05,
|
| 602 |
+
"loss": 0.2645,
|
| 603 |
+
"step": 46500
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 4.16,
|
| 607 |
+
"learning_rate": 2.921089879688606e-05,
|
| 608 |
+
"loss": 0.2644,
|
| 609 |
+
"step": 47000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 4.2,
|
| 613 |
+
"learning_rate": 2.8989738145789104e-05,
|
| 614 |
+
"loss": 0.2643,
|
| 615 |
+
"step": 47500
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 4.25,
|
| 619 |
+
"learning_rate": 2.8768577494692145e-05,
|
| 620 |
+
"loss": 0.2638,
|
| 621 |
+
"step": 48000
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 4.29,
|
| 625 |
+
"learning_rate": 2.8547416843595187e-05,
|
| 626 |
+
"loss": 0.2637,
|
| 627 |
+
"step": 48500
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 4.33,
|
| 631 |
+
"learning_rate": 2.832625619249823e-05,
|
| 632 |
+
"loss": 0.2636,
|
| 633 |
+
"step": 49000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 4.38,
|
| 637 |
+
"learning_rate": 2.8105095541401277e-05,
|
| 638 |
+
"loss": 0.2634,
|
| 639 |
+
"step": 49500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 4.42,
|
| 643 |
+
"learning_rate": 2.788393489030432e-05,
|
| 644 |
+
"loss": 0.2634,
|
| 645 |
+
"step": 50000
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 4.47,
|
| 649 |
+
"learning_rate": 2.766277423920736e-05,
|
| 650 |
+
"loss": 0.2629,
|
| 651 |
+
"step": 50500
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 4.51,
|
| 655 |
+
"learning_rate": 2.7441613588110405e-05,
|
| 656 |
+
"loss": 0.2631,
|
| 657 |
+
"step": 51000
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 4.56,
|
| 661 |
+
"learning_rate": 2.7220452937013447e-05,
|
| 662 |
+
"loss": 0.2629,
|
| 663 |
+
"step": 51500
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 4.6,
|
| 667 |
+
"learning_rate": 2.699929228591649e-05,
|
| 668 |
+
"loss": 0.2627,
|
| 669 |
+
"step": 52000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 4.64,
|
| 673 |
+
"learning_rate": 2.6778131634819537e-05,
|
| 674 |
+
"loss": 0.2627,
|
| 675 |
+
"step": 52500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 4.69,
|
| 679 |
+
"learning_rate": 2.655697098372258e-05,
|
| 680 |
+
"loss": 0.2624,
|
| 681 |
+
"step": 53000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 4.73,
|
| 685 |
+
"learning_rate": 2.633581033262562e-05,
|
| 686 |
+
"loss": 0.2622,
|
| 687 |
+
"step": 53500
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 4.78,
|
| 691 |
+
"learning_rate": 2.6114649681528662e-05,
|
| 692 |
+
"loss": 0.262,
|
| 693 |
+
"step": 54000
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 4.82,
|
| 697 |
+
"learning_rate": 2.5893489030431707e-05,
|
| 698 |
+
"loss": 0.2618,
|
| 699 |
+
"step": 54500
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 4.87,
|
| 703 |
+
"learning_rate": 2.567232837933475e-05,
|
| 704 |
+
"loss": 0.262,
|
| 705 |
+
"step": 55000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 4.91,
|
| 709 |
+
"learning_rate": 2.5451167728237797e-05,
|
| 710 |
+
"loss": 0.2616,
|
| 711 |
+
"step": 55500
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 4.95,
|
| 715 |
+
"learning_rate": 2.523000707714084e-05,
|
| 716 |
+
"loss": 0.2614,
|
| 717 |
+
"step": 56000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 5.0,
|
| 721 |
+
"learning_rate": 2.500884642604388e-05,
|
| 722 |
+
"loss": 0.2613,
|
| 723 |
+
"step": 56500
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 5.0,
|
| 727 |
+
"eval_accuracy": 0.9154096796494876,
|
| 728 |
+
"eval_loss": 0.24885989725589752,
|
| 729 |
+
"eval_runtime": 775.004,
|
| 730 |
+
"eval_samples_per_second": 311.18,
|
| 731 |
+
"eval_steps_per_second": 1.622,
|
| 732 |
+
"step": 56520
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 5.04,
|
| 736 |
+
"learning_rate": 2.4787685774946922e-05,
|
| 737 |
+
"loss": 0.2609,
|
| 738 |
+
"step": 57000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 5.09,
|
| 742 |
+
"learning_rate": 2.4566525123849964e-05,
|
| 743 |
+
"loss": 0.2608,
|
| 744 |
+
"step": 57500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 5.13,
|
| 748 |
+
"learning_rate": 2.434536447275301e-05,
|
| 749 |
+
"loss": 0.2607,
|
| 750 |
+
"step": 58000
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 5.18,
|
| 754 |
+
"learning_rate": 2.4124203821656054e-05,
|
| 755 |
+
"loss": 0.2604,
|
| 756 |
+
"step": 58500
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 5.22,
|
| 760 |
+
"learning_rate": 2.3903043170559095e-05,
|
| 761 |
+
"loss": 0.2604,
|
| 762 |
+
"step": 59000
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 5.26,
|
| 766 |
+
"learning_rate": 2.368188251946214e-05,
|
| 767 |
+
"loss": 0.2603,
|
| 768 |
+
"step": 59500
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 5.31,
|
| 772 |
+
"learning_rate": 2.3460721868365182e-05,
|
| 773 |
+
"loss": 0.2602,
|
| 774 |
+
"step": 60000
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 5.35,
|
| 778 |
+
"learning_rate": 2.3239561217268224e-05,
|
| 779 |
+
"loss": 0.2602,
|
| 780 |
+
"step": 60500
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 5.4,
|
| 784 |
+
"learning_rate": 2.301840056617127e-05,
|
| 785 |
+
"loss": 0.2602,
|
| 786 |
+
"step": 61000
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 5.44,
|
| 790 |
+
"learning_rate": 2.279723991507431e-05,
|
| 791 |
+
"loss": 0.26,
|
| 792 |
+
"step": 61500
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 5.48,
|
| 796 |
+
"learning_rate": 2.2576079263977355e-05,
|
| 797 |
+
"loss": 0.2598,
|
| 798 |
+
"step": 62000
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 5.53,
|
| 802 |
+
"learning_rate": 2.2354918612880397e-05,
|
| 803 |
+
"loss": 0.2598,
|
| 804 |
+
"step": 62500
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 5.57,
|
| 808 |
+
"learning_rate": 2.2133757961783442e-05,
|
| 809 |
+
"loss": 0.2595,
|
| 810 |
+
"step": 63000
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 5.62,
|
| 814 |
+
"learning_rate": 2.1912597310686484e-05,
|
| 815 |
+
"loss": 0.2596,
|
| 816 |
+
"step": 63500
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 5.66,
|
| 820 |
+
"learning_rate": 2.1691436659589525e-05,
|
| 821 |
+
"loss": 0.2594,
|
| 822 |
+
"step": 64000
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 5.71,
|
| 826 |
+
"learning_rate": 2.147027600849257e-05,
|
| 827 |
+
"loss": 0.2592,
|
| 828 |
+
"step": 64500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 5.75,
|
| 832 |
+
"learning_rate": 2.1249115357395612e-05,
|
| 833 |
+
"loss": 0.259,
|
| 834 |
+
"step": 65000
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 5.79,
|
| 838 |
+
"learning_rate": 2.1027954706298657e-05,
|
| 839 |
+
"loss": 0.259,
|
| 840 |
+
"step": 65500
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 5.84,
|
| 844 |
+
"learning_rate": 2.0806794055201702e-05,
|
| 845 |
+
"loss": 0.2589,
|
| 846 |
+
"step": 66000
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 5.88,
|
| 850 |
+
"learning_rate": 2.0585633404104744e-05,
|
| 851 |
+
"loss": 0.2587,
|
| 852 |
+
"step": 66500
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 5.93,
|
| 856 |
+
"learning_rate": 2.0364472753007785e-05,
|
| 857 |
+
"loss": 0.2589,
|
| 858 |
+
"step": 67000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 5.97,
|
| 862 |
+
"learning_rate": 2.014331210191083e-05,
|
| 863 |
+
"loss": 0.2586,
|
| 864 |
+
"step": 67500
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 6.0,
|
| 868 |
+
"eval_accuracy": 0.9157895529502922,
|
| 869 |
+
"eval_loss": 0.24735933542251587,
|
| 870 |
+
"eval_runtime": 787.6461,
|
| 871 |
+
"eval_samples_per_second": 306.186,
|
| 872 |
+
"eval_steps_per_second": 1.596,
|
| 873 |
+
"step": 67824
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 6.02,
|
| 877 |
+
"learning_rate": 1.9922151450813872e-05,
|
| 878 |
+
"loss": 0.2585,
|
| 879 |
+
"step": 68000
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 6.06,
|
| 883 |
+
"learning_rate": 1.9700990799716914e-05,
|
| 884 |
+
"loss": 0.2581,
|
| 885 |
+
"step": 68500
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 6.1,
|
| 889 |
+
"learning_rate": 1.947983014861996e-05,
|
| 890 |
+
"loss": 0.2581,
|
| 891 |
+
"step": 69000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 6.15,
|
| 895 |
+
"learning_rate": 1.9258669497523004e-05,
|
| 896 |
+
"loss": 0.2579,
|
| 897 |
+
"step": 69500
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 6.19,
|
| 901 |
+
"learning_rate": 1.9037508846426045e-05,
|
| 902 |
+
"loss": 0.2579,
|
| 903 |
+
"step": 70000
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 6.24,
|
| 907 |
+
"learning_rate": 1.881634819532909e-05,
|
| 908 |
+
"loss": 0.2579,
|
| 909 |
+
"step": 70500
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 6.28,
|
| 913 |
+
"learning_rate": 1.8595187544232132e-05,
|
| 914 |
+
"loss": 0.2578,
|
| 915 |
+
"step": 71000
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 6.33,
|
| 919 |
+
"learning_rate": 1.8374026893135174e-05,
|
| 920 |
+
"loss": 0.2577,
|
| 921 |
+
"step": 71500
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 6.37,
|
| 925 |
+
"learning_rate": 1.8152866242038215e-05,
|
| 926 |
+
"loss": 0.2577,
|
| 927 |
+
"step": 72000
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 6.41,
|
| 931 |
+
"learning_rate": 1.793170559094126e-05,
|
| 932 |
+
"loss": 0.2576,
|
| 933 |
+
"step": 72500
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 6.46,
|
| 937 |
+
"learning_rate": 1.7710544939844302e-05,
|
| 938 |
+
"loss": 0.2575,
|
| 939 |
+
"step": 73000
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 6.5,
|
| 943 |
+
"learning_rate": 1.7489384288747347e-05,
|
| 944 |
+
"loss": 0.2572,
|
| 945 |
+
"step": 73500
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 6.55,
|
| 949 |
+
"learning_rate": 1.7268223637650392e-05,
|
| 950 |
+
"loss": 0.2575,
|
| 951 |
+
"step": 74000
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 6.59,
|
| 955 |
+
"learning_rate": 1.7047062986553434e-05,
|
| 956 |
+
"loss": 0.2571,
|
| 957 |
+
"step": 74500
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 6.63,
|
| 961 |
+
"learning_rate": 1.6825902335456475e-05,
|
| 962 |
+
"loss": 0.2573,
|
| 963 |
+
"step": 75000
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 6.68,
|
| 967 |
+
"learning_rate": 1.660474168435952e-05,
|
| 968 |
+
"loss": 0.2573,
|
| 969 |
+
"step": 75500
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 6.72,
|
| 973 |
+
"learning_rate": 1.6383581033262562e-05,
|
| 974 |
+
"loss": 0.2572,
|
| 975 |
+
"step": 76000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 6.77,
|
| 979 |
+
"learning_rate": 1.6162420382165603e-05,
|
| 980 |
+
"loss": 0.257,
|
| 981 |
+
"step": 76500
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 6.81,
|
| 985 |
+
"learning_rate": 1.594125973106865e-05,
|
| 986 |
+
"loss": 0.257,
|
| 987 |
+
"step": 77000
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 6.86,
|
| 991 |
+
"learning_rate": 1.5720099079971694e-05,
|
| 992 |
+
"loss": 0.2569,
|
| 993 |
+
"step": 77500
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 6.9,
|
| 997 |
+
"learning_rate": 1.5498938428874735e-05,
|
| 998 |
+
"loss": 0.257,
|
| 999 |
+
"step": 78000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 6.94,
|
| 1003 |
+
"learning_rate": 1.527777777777778e-05,
|
| 1004 |
+
"loss": 0.2569,
|
| 1005 |
+
"step": 78500
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 6.99,
|
| 1009 |
+
"learning_rate": 1.5056617126680822e-05,
|
| 1010 |
+
"loss": 0.2569,
|
| 1011 |
+
"step": 79000
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 7.0,
|
| 1015 |
+
"eval_accuracy": 0.9160589551288438,
|
| 1016 |
+
"eval_loss": 0.24625588953495026,
|
| 1017 |
+
"eval_runtime": 787.7182,
|
| 1018 |
+
"eval_samples_per_second": 306.158,
|
| 1019 |
+
"eval_steps_per_second": 1.596,
|
| 1020 |
+
"step": 79128
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 7.03,
|
| 1024 |
+
"learning_rate": 1.4835456475583863e-05,
|
| 1025 |
+
"loss": 0.2564,
|
| 1026 |
+
"step": 79500
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 7.08,
|
| 1030 |
+
"learning_rate": 1.461429582448691e-05,
|
| 1031 |
+
"loss": 0.2563,
|
| 1032 |
+
"step": 80000
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 7.12,
|
| 1036 |
+
"learning_rate": 1.4393135173389952e-05,
|
| 1037 |
+
"loss": 0.2562,
|
| 1038 |
+
"step": 80500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 7.17,
|
| 1042 |
+
"learning_rate": 1.4171974522292993e-05,
|
| 1043 |
+
"loss": 0.2563,
|
| 1044 |
+
"step": 81000
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 7.21,
|
| 1048 |
+
"learning_rate": 1.3950813871196037e-05,
|
| 1049 |
+
"loss": 0.2561,
|
| 1050 |
+
"step": 81500
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 7.25,
|
| 1054 |
+
"learning_rate": 1.3729653220099082e-05,
|
| 1055 |
+
"loss": 0.256,
|
| 1056 |
+
"step": 82000
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 7.3,
|
| 1060 |
+
"learning_rate": 1.3508492569002123e-05,
|
| 1061 |
+
"loss": 0.2561,
|
| 1062 |
+
"step": 82500
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 7.34,
|
| 1066 |
+
"learning_rate": 1.3287331917905165e-05,
|
| 1067 |
+
"loss": 0.2561,
|
| 1068 |
+
"step": 83000
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 7.39,
|
| 1072 |
+
"learning_rate": 1.306617126680821e-05,
|
| 1073 |
+
"loss": 0.2559,
|
| 1074 |
+
"step": 83500
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 7.43,
|
| 1078 |
+
"learning_rate": 1.2845010615711253e-05,
|
| 1079 |
+
"loss": 0.256,
|
| 1080 |
+
"step": 84000
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 7.48,
|
| 1084 |
+
"learning_rate": 1.2623849964614295e-05,
|
| 1085 |
+
"loss": 0.2558,
|
| 1086 |
+
"step": 84500
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 7.52,
|
| 1090 |
+
"learning_rate": 1.2402689313517338e-05,
|
| 1091 |
+
"loss": 0.256,
|
| 1092 |
+
"step": 85000
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 7.56,
|
| 1096 |
+
"learning_rate": 1.2181528662420383e-05,
|
| 1097 |
+
"loss": 0.2558,
|
| 1098 |
+
"step": 85500
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 7.61,
|
| 1102 |
+
"learning_rate": 1.1960368011323427e-05,
|
| 1103 |
+
"loss": 0.2559,
|
| 1104 |
+
"step": 86000
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 7.65,
|
| 1108 |
+
"learning_rate": 1.1739207360226468e-05,
|
| 1109 |
+
"loss": 0.2557,
|
| 1110 |
+
"step": 86500
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 7.7,
|
| 1114 |
+
"learning_rate": 1.1518046709129512e-05,
|
| 1115 |
+
"loss": 0.2557,
|
| 1116 |
+
"step": 87000
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 7.74,
|
| 1120 |
+
"learning_rate": 1.1296886058032555e-05,
|
| 1121 |
+
"loss": 0.2555,
|
| 1122 |
+
"step": 87500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 7.78,
|
| 1126 |
+
"learning_rate": 1.1075725406935598e-05,
|
| 1127 |
+
"loss": 0.2556,
|
| 1128 |
+
"step": 88000
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 7.83,
|
| 1132 |
+
"learning_rate": 1.0854564755838642e-05,
|
| 1133 |
+
"loss": 0.2556,
|
| 1134 |
+
"step": 88500
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 7.87,
|
| 1138 |
+
"learning_rate": 1.0633404104741685e-05,
|
| 1139 |
+
"loss": 0.2552,
|
| 1140 |
+
"step": 89000
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 7.92,
|
| 1144 |
+
"learning_rate": 1.0412243453644728e-05,
|
| 1145 |
+
"loss": 0.2554,
|
| 1146 |
+
"step": 89500
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 7.96,
|
| 1150 |
+
"learning_rate": 1.0191082802547772e-05,
|
| 1151 |
+
"loss": 0.2553,
|
| 1152 |
+
"step": 90000
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 8.0,
|
| 1156 |
+
"eval_accuracy": 0.9162647719016384,
|
| 1157 |
+
"eval_loss": 0.24548450112342834,
|
| 1158 |
+
"eval_runtime": 790.7594,
|
| 1159 |
+
"eval_samples_per_second": 304.98,
|
| 1160 |
+
"eval_steps_per_second": 1.59,
|
| 1161 |
+
"step": 90432
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 8.01,
|
| 1165 |
+
"learning_rate": 9.969922151450813e-06,
|
| 1166 |
+
"loss": 0.2553,
|
| 1167 |
+
"step": 90500
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 8.05,
|
| 1171 |
+
"learning_rate": 9.748761500353858e-06,
|
| 1172 |
+
"loss": 0.255,
|
| 1173 |
+
"step": 91000
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 8.09,
|
| 1177 |
+
"learning_rate": 9.527600849256902e-06,
|
| 1178 |
+
"loss": 0.2549,
|
| 1179 |
+
"step": 91500
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 8.14,
|
| 1183 |
+
"learning_rate": 9.306440198159943e-06,
|
| 1184 |
+
"loss": 0.255,
|
| 1185 |
+
"step": 92000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 8.18,
|
| 1189 |
+
"learning_rate": 9.085279547062987e-06,
|
| 1190 |
+
"loss": 0.2549,
|
| 1191 |
+
"step": 92500
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 8.23,
|
| 1195 |
+
"learning_rate": 8.86411889596603e-06,
|
| 1196 |
+
"loss": 0.2548,
|
| 1197 |
+
"step": 93000
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 8.27,
|
| 1201 |
+
"learning_rate": 8.642958244869073e-06,
|
| 1202 |
+
"loss": 0.2551,
|
| 1203 |
+
"step": 93500
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 8.32,
|
| 1207 |
+
"learning_rate": 8.421797593772117e-06,
|
| 1208 |
+
"loss": 0.2548,
|
| 1209 |
+
"step": 94000
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 8.36,
|
| 1213 |
+
"learning_rate": 8.200636942675158e-06,
|
| 1214 |
+
"loss": 0.2549,
|
| 1215 |
+
"step": 94500
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 8.4,
|
| 1219 |
+
"learning_rate": 7.979476291578203e-06,
|
| 1220 |
+
"loss": 0.2549,
|
| 1221 |
+
"step": 95000
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 8.45,
|
| 1225 |
+
"learning_rate": 7.758315640481247e-06,
|
| 1226 |
+
"loss": 0.2548,
|
| 1227 |
+
"step": 95500
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 8.49,
|
| 1231 |
+
"learning_rate": 7.537154989384289e-06,
|
| 1232 |
+
"loss": 0.2547,
|
| 1233 |
+
"step": 96000
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 8.54,
|
| 1237 |
+
"learning_rate": 7.3159943382873326e-06,
|
| 1238 |
+
"loss": 0.2546,
|
| 1239 |
+
"step": 96500
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 8.58,
|
| 1243 |
+
"learning_rate": 7.094833687190375e-06,
|
| 1244 |
+
"loss": 0.2548,
|
| 1245 |
+
"step": 97000
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 8.63,
|
| 1249 |
+
"learning_rate": 6.873673036093418e-06,
|
| 1250 |
+
"loss": 0.2547,
|
| 1251 |
+
"step": 97500
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 8.67,
|
| 1255 |
+
"learning_rate": 6.652512384996462e-06,
|
| 1256 |
+
"loss": 0.2547,
|
| 1257 |
+
"step": 98000
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 8.71,
|
| 1261 |
+
"learning_rate": 6.431351733899504e-06,
|
| 1262 |
+
"loss": 0.2547,
|
| 1263 |
+
"step": 98500
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 8.76,
|
| 1267 |
+
"learning_rate": 6.210191082802548e-06,
|
| 1268 |
+
"loss": 0.2545,
|
| 1269 |
+
"step": 99000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 8.8,
|
| 1273 |
+
"learning_rate": 5.989030431705591e-06,
|
| 1274 |
+
"loss": 0.2544,
|
| 1275 |
+
"step": 99500
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 8.85,
|
| 1279 |
+
"learning_rate": 5.767869780608635e-06,
|
| 1280 |
+
"loss": 0.2544,
|
| 1281 |
+
"step": 100000
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 8.89,
|
| 1285 |
+
"learning_rate": 5.5467091295116775e-06,
|
| 1286 |
+
"loss": 0.2545,
|
| 1287 |
+
"step": 100500
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 8.93,
|
| 1291 |
+
"learning_rate": 5.325548478414721e-06,
|
| 1292 |
+
"loss": 0.2545,
|
| 1293 |
+
"step": 101000
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 8.98,
|
| 1297 |
+
"learning_rate": 5.104387827317763e-06,
|
| 1298 |
+
"loss": 0.2545,
|
| 1299 |
+
"step": 101500
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 9.0,
|
| 1303 |
+
"eval_accuracy": 0.9164148337838057,
|
| 1304 |
+
"eval_loss": 0.24497151374816895,
|
| 1305 |
+
"eval_runtime": 789.0403,
|
| 1306 |
+
"eval_samples_per_second": 305.645,
|
| 1307 |
+
"eval_steps_per_second": 1.593,
|
| 1308 |
+
"step": 101736
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 9.02,
|
| 1312 |
+
"learning_rate": 4.8832271762208075e-06,
|
| 1313 |
+
"loss": 0.2543,
|
| 1314 |
+
"step": 102000
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 9.07,
|
| 1318 |
+
"learning_rate": 4.66206652512385e-06,
|
| 1319 |
+
"loss": 0.254,
|
| 1320 |
+
"step": 102500
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 9.11,
|
| 1324 |
+
"learning_rate": 4.440905874026893e-06,
|
| 1325 |
+
"loss": 0.254,
|
| 1326 |
+
"step": 103000
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 9.16,
|
| 1330 |
+
"learning_rate": 4.219745222929937e-06,
|
| 1331 |
+
"loss": 0.2542,
|
| 1332 |
+
"step": 103500
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 9.2,
|
| 1336 |
+
"learning_rate": 3.99858457183298e-06,
|
| 1337 |
+
"loss": 0.2541,
|
| 1338 |
+
"step": 104000
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 9.24,
|
| 1342 |
+
"learning_rate": 3.777423920736023e-06,
|
| 1343 |
+
"loss": 0.254,
|
| 1344 |
+
"step": 104500
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 9.29,
|
| 1348 |
+
"learning_rate": 3.556263269639066e-06,
|
| 1349 |
+
"loss": 0.2541,
|
| 1350 |
+
"step": 105000
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 9.33,
|
| 1354 |
+
"learning_rate": 3.3351026185421087e-06,
|
| 1355 |
+
"loss": 0.2541,
|
| 1356 |
+
"step": 105500
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 9.38,
|
| 1360 |
+
"learning_rate": 3.113941967445152e-06,
|
| 1361 |
+
"loss": 0.2541,
|
| 1362 |
+
"step": 106000
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 9.42,
|
| 1366 |
+
"learning_rate": 2.8927813163481954e-06,
|
| 1367 |
+
"loss": 0.2541,
|
| 1368 |
+
"step": 106500
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 9.47,
|
| 1372 |
+
"learning_rate": 2.6716206652512383e-06,
|
| 1373 |
+
"loss": 0.2539,
|
| 1374 |
+
"step": 107000
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 9.51,
|
| 1378 |
+
"learning_rate": 2.4504600141542816e-06,
|
| 1379 |
+
"loss": 0.2541,
|
| 1380 |
+
"step": 107500
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 9.55,
|
| 1384 |
+
"learning_rate": 2.229299363057325e-06,
|
| 1385 |
+
"loss": 0.2541,
|
| 1386 |
+
"step": 108000
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 9.6,
|
| 1390 |
+
"learning_rate": 2.0081387119603683e-06,
|
| 1391 |
+
"loss": 0.2541,
|
| 1392 |
+
"step": 108500
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 9.64,
|
| 1396 |
+
"learning_rate": 1.786978060863411e-06,
|
| 1397 |
+
"loss": 0.254,
|
| 1398 |
+
"step": 109000
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 9.69,
|
| 1402 |
+
"learning_rate": 1.5658174097664543e-06,
|
| 1403 |
+
"loss": 0.2541,
|
| 1404 |
+
"step": 109500
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 9.73,
|
| 1408 |
+
"learning_rate": 1.3446567586694977e-06,
|
| 1409 |
+
"loss": 0.254,
|
| 1410 |
+
"step": 110000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 9.78,
|
| 1414 |
+
"learning_rate": 1.1234961075725408e-06,
|
| 1415 |
+
"loss": 0.2539,
|
| 1416 |
+
"step": 110500
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 9.82,
|
| 1420 |
+
"learning_rate": 9.023354564755839e-07,
|
| 1421 |
+
"loss": 0.2537,
|
| 1422 |
+
"step": 111000
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 9.86,
|
| 1426 |
+
"learning_rate": 6.81174805378627e-07,
|
| 1427 |
+
"loss": 0.254,
|
| 1428 |
+
"step": 111500
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 9.91,
|
| 1432 |
+
"learning_rate": 4.6001415428167025e-07,
|
| 1433 |
+
"loss": 0.2541,
|
| 1434 |
+
"step": 112000
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 9.95,
|
| 1438 |
+
"learning_rate": 2.388535031847134e-07,
|
| 1439 |
+
"loss": 0.2539,
|
| 1440 |
+
"step": 112500
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 10.0,
|
| 1444 |
+
"learning_rate": 1.7692852087756548e-08,
|
| 1445 |
+
"loss": 0.2539,
|
| 1446 |
+
"step": 113000
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 10.0,
|
| 1450 |
+
"eval_accuracy": 0.916456858900755,
|
| 1451 |
+
"eval_loss": 0.24475741386413574,
|
| 1452 |
+
"eval_runtime": 768.7642,
|
| 1453 |
+
"eval_samples_per_second": 313.706,
|
| 1454 |
+
"eval_steps_per_second": 1.635,
|
| 1455 |
+
"step": 113040
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 10.0,
|
| 1459 |
+
"step": 113040,
|
| 1460 |
+
"total_flos": 5.670998618744553e+18,
|
| 1461 |
+
"train_loss": 0.27495618251225495,
|
| 1462 |
+
"train_runtime": 96949.7276,
|
| 1463 |
+
"train_samples_per_second": 223.864,
|
| 1464 |
+
"train_steps_per_second": 1.166
|
| 1465 |
+
}
|
| 1466 |
+
],
|
| 1467 |
+
"logging_steps": 500,
|
| 1468 |
+
"max_steps": 113040,
|
| 1469 |
+
"num_input_tokens_seen": 0,
|
| 1470 |
+
"num_train_epochs": 10,
|
| 1471 |
+
"save_steps": 500,
|
| 1472 |
+
"total_flos": 5.670998618744553e+18,
|
| 1473 |
+
"train_batch_size": 24,
|
| 1474 |
+
"trial_name": null,
|
| 1475 |
+
"trial_params": null
|
| 1476 |
+
}
|