End of training
Browse files- README.md +4 -1
- all_results.json +11 -0
- eval_results.json +7 -0
- runs/Mar20_21-36-58_c29773f36be4/events.out.tfevents.1710984671.c29773f36be4.4971.1 +3 -0
- train_results.json +7 -0
- trainer_state.json +1535 -0
README.md
CHANGED
|
@@ -1,5 +1,6 @@
|
|
| 1 |
---
|
| 2 |
tags:
|
|
|
|
| 3 |
- generated_from_trainer
|
| 4 |
datasets:
|
| 5 |
- imagefolder
|
|
@@ -13,7 +14,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# prtrnlng2
|
| 15 |
|
| 16 |
-
This model is a fine-tuned version of [](https://huggingface.co/) on the
|
|
|
|
|
|
|
| 17 |
|
| 18 |
## Model description
|
| 19 |
|
|
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
+
- masked-auto-encoding
|
| 4 |
- generated_from_trainer
|
| 5 |
datasets:
|
| 6 |
- imagefolder
|
|
|
|
| 14 |
|
| 15 |
# prtrnlng2
|
| 16 |
|
| 17 |
+
This model is a fine-tuned version of [](https://huggingface.co/) on the /home/lung/LungAll dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.5091
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_loss": 0.5090942978858948,
|
| 4 |
+
"eval_runtime": 27.4368,
|
| 5 |
+
"eval_samples_per_second": 110.946,
|
| 6 |
+
"eval_steps_per_second": 13.886,
|
| 7 |
+
"train_loss": 0.5550972292020724,
|
| 8 |
+
"train_runtime": 13976.6946,
|
| 9 |
+
"train_samples_per_second": 61.703,
|
| 10 |
+
"train_steps_per_second": 7.713
|
| 11 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_loss": 0.5090942978858948,
|
| 4 |
+
"eval_runtime": 27.4368,
|
| 5 |
+
"eval_samples_per_second": 110.946,
|
| 6 |
+
"eval_steps_per_second": 13.886
|
| 7 |
+
}
|
runs/Mar20_21-36-58_c29773f36be4/events.out.tfevents.1710984671.c29773f36be4.4971.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf68289397e098b18e7b7a669891504a518ba06583c4a1cea5e7c1b70d220cce
|
| 3 |
+
size 364
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"train_loss": 0.5550972292020724,
|
| 4 |
+
"train_runtime": 13976.6946,
|
| 5 |
+
"train_samples_per_second": 61.703,
|
| 6 |
+
"train_steps_per_second": 7.713
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1535 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 50.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 107800,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.23,
|
| 13 |
+
"grad_norm": 0.16410106420516968,
|
| 14 |
+
"learning_rate": 3.110505565862709e-05,
|
| 15 |
+
"loss": 0.8386,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.46,
|
| 20 |
+
"grad_norm": 0.42184513807296753,
|
| 21 |
+
"learning_rate": 3.0960111317254176e-05,
|
| 22 |
+
"loss": 0.8027,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.7,
|
| 27 |
+
"grad_norm": 0.24314333498477936,
|
| 28 |
+
"learning_rate": 3.0815166975881264e-05,
|
| 29 |
+
"loss": 0.782,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.93,
|
| 34 |
+
"grad_norm": 0.6049493551254272,
|
| 35 |
+
"learning_rate": 3.067022263450835e-05,
|
| 36 |
+
"loss": 0.7584,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 1.16,
|
| 41 |
+
"grad_norm": 0.6963891386985779,
|
| 42 |
+
"learning_rate": 3.0525278293135433e-05,
|
| 43 |
+
"loss": 0.7384,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.39,
|
| 48 |
+
"grad_norm": 0.4310186803340912,
|
| 49 |
+
"learning_rate": 3.038033395176252e-05,
|
| 50 |
+
"loss": 0.7254,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 1.62,
|
| 55 |
+
"grad_norm": 0.713209331035614,
|
| 56 |
+
"learning_rate": 3.0235389610389613e-05,
|
| 57 |
+
"loss": 0.7181,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.86,
|
| 62 |
+
"grad_norm": 0.6885315775871277,
|
| 63 |
+
"learning_rate": 3.0090445269016697e-05,
|
| 64 |
+
"loss": 0.703,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 2.09,
|
| 69 |
+
"grad_norm": 0.48805415630340576,
|
| 70 |
+
"learning_rate": 2.9945500927643785e-05,
|
| 71 |
+
"loss": 0.6923,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 2.32,
|
| 76 |
+
"grad_norm": 0.40450435876846313,
|
| 77 |
+
"learning_rate": 2.9800556586270873e-05,
|
| 78 |
+
"loss": 0.691,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 2.55,
|
| 83 |
+
"grad_norm": 0.5716784596443176,
|
| 84 |
+
"learning_rate": 2.965561224489796e-05,
|
| 85 |
+
"loss": 0.6831,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.78,
|
| 90 |
+
"grad_norm": 0.512924075126648,
|
| 91 |
+
"learning_rate": 2.951066790352505e-05,
|
| 92 |
+
"loss": 0.6783,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 3.01,
|
| 97 |
+
"grad_norm": 0.5809201598167419,
|
| 98 |
+
"learning_rate": 2.9365723562152134e-05,
|
| 99 |
+
"loss": 0.6682,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 3.25,
|
| 104 |
+
"grad_norm": 1.9776902198791504,
|
| 105 |
+
"learning_rate": 2.922077922077922e-05,
|
| 106 |
+
"loss": 0.6654,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 3.48,
|
| 111 |
+
"grad_norm": 0.5193782448768616,
|
| 112 |
+
"learning_rate": 2.9075834879406306e-05,
|
| 113 |
+
"loss": 0.6649,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 3.71,
|
| 118 |
+
"grad_norm": 0.5309423804283142,
|
| 119 |
+
"learning_rate": 2.8930890538033397e-05,
|
| 120 |
+
"loss": 0.6616,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 3.94,
|
| 125 |
+
"grad_norm": 0.5730230212211609,
|
| 126 |
+
"learning_rate": 2.8785946196660482e-05,
|
| 127 |
+
"loss": 0.6559,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 4.17,
|
| 132 |
+
"grad_norm": 2.4345550537109375,
|
| 133 |
+
"learning_rate": 2.864100185528757e-05,
|
| 134 |
+
"loss": 0.6535,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 4.41,
|
| 139 |
+
"grad_norm": 0.6113856434822083,
|
| 140 |
+
"learning_rate": 2.8496057513914658e-05,
|
| 141 |
+
"loss": 0.6479,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 4.64,
|
| 146 |
+
"grad_norm": 0.5557487607002258,
|
| 147 |
+
"learning_rate": 2.8351113172541746e-05,
|
| 148 |
+
"loss": 0.6416,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 4.87,
|
| 153 |
+
"grad_norm": 0.5874636173248291,
|
| 154 |
+
"learning_rate": 2.8206168831168834e-05,
|
| 155 |
+
"loss": 0.6399,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 5.1,
|
| 160 |
+
"grad_norm": 0.4734204411506653,
|
| 161 |
+
"learning_rate": 2.8061224489795918e-05,
|
| 162 |
+
"loss": 0.6375,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 5.33,
|
| 167 |
+
"grad_norm": 0.5736048221588135,
|
| 168 |
+
"learning_rate": 2.7916280148423006e-05,
|
| 169 |
+
"loss": 0.6336,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 5.57,
|
| 174 |
+
"grad_norm": 0.5358075499534607,
|
| 175 |
+
"learning_rate": 2.777133580705009e-05,
|
| 176 |
+
"loss": 0.6255,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 5.8,
|
| 181 |
+
"grad_norm": 0.39945241808891296,
|
| 182 |
+
"learning_rate": 2.7626391465677182e-05,
|
| 183 |
+
"loss": 0.6245,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 6.03,
|
| 188 |
+
"grad_norm": 0.6585483551025391,
|
| 189 |
+
"learning_rate": 2.7481447124304266e-05,
|
| 190 |
+
"loss": 0.619,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 6.26,
|
| 195 |
+
"grad_norm": 0.5212920308113098,
|
| 196 |
+
"learning_rate": 2.7336502782931354e-05,
|
| 197 |
+
"loss": 0.6195,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 6.49,
|
| 202 |
+
"grad_norm": 0.6320118308067322,
|
| 203 |
+
"learning_rate": 2.7191558441558442e-05,
|
| 204 |
+
"loss": 0.6114,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 6.73,
|
| 209 |
+
"grad_norm": 0.7715293169021606,
|
| 210 |
+
"learning_rate": 2.704661410018553e-05,
|
| 211 |
+
"loss": 0.6088,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 6.96,
|
| 216 |
+
"grad_norm": 0.571773111820221,
|
| 217 |
+
"learning_rate": 2.6901669758812618e-05,
|
| 218 |
+
"loss": 0.6103,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 7.19,
|
| 223 |
+
"grad_norm": 0.6421319246292114,
|
| 224 |
+
"learning_rate": 2.6756725417439703e-05,
|
| 225 |
+
"loss": 0.6025,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 7.42,
|
| 230 |
+
"grad_norm": 0.5680719017982483,
|
| 231 |
+
"learning_rate": 2.661178107606679e-05,
|
| 232 |
+
"loss": 0.5998,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 7.65,
|
| 237 |
+
"grad_norm": 0.5537917613983154,
|
| 238 |
+
"learning_rate": 2.646683673469388e-05,
|
| 239 |
+
"loss": 0.5981,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 7.88,
|
| 244 |
+
"grad_norm": 0.4245034158229828,
|
| 245 |
+
"learning_rate": 2.6321892393320966e-05,
|
| 246 |
+
"loss": 0.5959,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 8.12,
|
| 251 |
+
"grad_norm": 0.48047196865081787,
|
| 252 |
+
"learning_rate": 2.617694805194805e-05,
|
| 253 |
+
"loss": 0.5952,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 8.35,
|
| 258 |
+
"grad_norm": 0.4809471368789673,
|
| 259 |
+
"learning_rate": 2.603200371057514e-05,
|
| 260 |
+
"loss": 0.5923,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 8.58,
|
| 265 |
+
"grad_norm": 0.4537654519081116,
|
| 266 |
+
"learning_rate": 2.5887059369202227e-05,
|
| 267 |
+
"loss": 0.5878,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 8.81,
|
| 272 |
+
"grad_norm": 0.5159969329833984,
|
| 273 |
+
"learning_rate": 2.5742115027829315e-05,
|
| 274 |
+
"loss": 0.5933,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 9.04,
|
| 279 |
+
"grad_norm": 0.9971024990081787,
|
| 280 |
+
"learning_rate": 2.5597170686456403e-05,
|
| 281 |
+
"loss": 0.588,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 9.28,
|
| 286 |
+
"grad_norm": 0.4496227204799652,
|
| 287 |
+
"learning_rate": 2.5452226345083487e-05,
|
| 288 |
+
"loss": 0.5877,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 9.51,
|
| 293 |
+
"grad_norm": 0.5145105719566345,
|
| 294 |
+
"learning_rate": 2.5307282003710575e-05,
|
| 295 |
+
"loss": 0.5821,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 9.74,
|
| 300 |
+
"grad_norm": 0.5195399522781372,
|
| 301 |
+
"learning_rate": 2.5162337662337663e-05,
|
| 302 |
+
"loss": 0.5827,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 9.97,
|
| 307 |
+
"grad_norm": 0.5431232452392578,
|
| 308 |
+
"learning_rate": 2.501739332096475e-05,
|
| 309 |
+
"loss": 0.5845,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 10.2,
|
| 314 |
+
"grad_norm": 0.4878028333187103,
|
| 315 |
+
"learning_rate": 2.4872448979591835e-05,
|
| 316 |
+
"loss": 0.5828,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 10.44,
|
| 321 |
+
"grad_norm": 0.591066837310791,
|
| 322 |
+
"learning_rate": 2.4727504638218923e-05,
|
| 323 |
+
"loss": 0.5819,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 10.67,
|
| 328 |
+
"grad_norm": 16.133996963500977,
|
| 329 |
+
"learning_rate": 2.458256029684601e-05,
|
| 330 |
+
"loss": 0.5765,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 10.9,
|
| 335 |
+
"grad_norm": 0.5531517863273621,
|
| 336 |
+
"learning_rate": 2.44376159554731e-05,
|
| 337 |
+
"loss": 0.5781,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 11.13,
|
| 342 |
+
"grad_norm": 0.4941064417362213,
|
| 343 |
+
"learning_rate": 2.4292671614100187e-05,
|
| 344 |
+
"loss": 0.5774,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 11.36,
|
| 349 |
+
"grad_norm": 0.5101198554039001,
|
| 350 |
+
"learning_rate": 2.414772727272727e-05,
|
| 351 |
+
"loss": 0.5717,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 11.6,
|
| 356 |
+
"grad_norm": 0.40433865785598755,
|
| 357 |
+
"learning_rate": 2.400278293135436e-05,
|
| 358 |
+
"loss": 0.5729,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 11.83,
|
| 363 |
+
"grad_norm": 0.5296735763549805,
|
| 364 |
+
"learning_rate": 2.3857838589981448e-05,
|
| 365 |
+
"loss": 0.5766,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 12.06,
|
| 370 |
+
"grad_norm": 0.40421661734580994,
|
| 371 |
+
"learning_rate": 2.3712894248608535e-05,
|
| 372 |
+
"loss": 0.5674,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 12.29,
|
| 377 |
+
"grad_norm": 0.48923459649086,
|
| 378 |
+
"learning_rate": 2.356794990723562e-05,
|
| 379 |
+
"loss": 0.5673,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 12.52,
|
| 384 |
+
"grad_norm": 0.4543541371822357,
|
| 385 |
+
"learning_rate": 2.3423005565862708e-05,
|
| 386 |
+
"loss": 0.5682,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 12.76,
|
| 391 |
+
"grad_norm": 0.405658096075058,
|
| 392 |
+
"learning_rate": 2.32780612244898e-05,
|
| 393 |
+
"loss": 0.5654,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 12.99,
|
| 398 |
+
"grad_norm": 0.4983844459056854,
|
| 399 |
+
"learning_rate": 2.3133116883116884e-05,
|
| 400 |
+
"loss": 0.5657,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 13.22,
|
| 405 |
+
"grad_norm": 1.2178577184677124,
|
| 406 |
+
"learning_rate": 2.2988172541743972e-05,
|
| 407 |
+
"loss": 0.5638,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 13.45,
|
| 412 |
+
"grad_norm": 0.4428790509700775,
|
| 413 |
+
"learning_rate": 2.2843228200371056e-05,
|
| 414 |
+
"loss": 0.5658,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 13.68,
|
| 419 |
+
"grad_norm": 0.5377063155174255,
|
| 420 |
+
"learning_rate": 2.2698283858998144e-05,
|
| 421 |
+
"loss": 0.5609,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 13.91,
|
| 426 |
+
"grad_norm": 0.4518877863883972,
|
| 427 |
+
"learning_rate": 2.2553339517625232e-05,
|
| 428 |
+
"loss": 0.5629,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 14.15,
|
| 433 |
+
"grad_norm": 0.5720136761665344,
|
| 434 |
+
"learning_rate": 2.240839517625232e-05,
|
| 435 |
+
"loss": 0.5639,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 14.38,
|
| 440 |
+
"grad_norm": 0.6480860114097595,
|
| 441 |
+
"learning_rate": 2.2263450834879408e-05,
|
| 442 |
+
"loss": 0.5622,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 14.61,
|
| 447 |
+
"grad_norm": 0.40764591097831726,
|
| 448 |
+
"learning_rate": 2.2118506493506492e-05,
|
| 449 |
+
"loss": 0.5615,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 14.84,
|
| 454 |
+
"grad_norm": 0.60052889585495,
|
| 455 |
+
"learning_rate": 2.1973562152133584e-05,
|
| 456 |
+
"loss": 0.5542,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 15.07,
|
| 461 |
+
"grad_norm": 0.5227585434913635,
|
| 462 |
+
"learning_rate": 2.182861781076067e-05,
|
| 463 |
+
"loss": 0.5556,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 15.31,
|
| 468 |
+
"grad_norm": 0.5119400024414062,
|
| 469 |
+
"learning_rate": 2.1683673469387756e-05,
|
| 470 |
+
"loss": 0.556,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 15.54,
|
| 475 |
+
"grad_norm": 0.4895058870315552,
|
| 476 |
+
"learning_rate": 2.153872912801484e-05,
|
| 477 |
+
"loss": 0.5569,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 15.77,
|
| 482 |
+
"grad_norm": 0.5205515623092651,
|
| 483 |
+
"learning_rate": 2.1393784786641932e-05,
|
| 484 |
+
"loss": 0.555,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 16.0,
|
| 489 |
+
"grad_norm": 0.4375595152378082,
|
| 490 |
+
"learning_rate": 2.1248840445269017e-05,
|
| 491 |
+
"loss": 0.5544,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 16.23,
|
| 496 |
+
"grad_norm": 0.4849925637245178,
|
| 497 |
+
"learning_rate": 2.1103896103896105e-05,
|
| 498 |
+
"loss": 0.5536,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 16.47,
|
| 503 |
+
"grad_norm": 0.5083441734313965,
|
| 504 |
+
"learning_rate": 2.0958951762523192e-05,
|
| 505 |
+
"loss": 0.5522,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 16.7,
|
| 510 |
+
"grad_norm": 0.561335027217865,
|
| 511 |
+
"learning_rate": 2.0814007421150277e-05,
|
| 512 |
+
"loss": 0.5523,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 16.93,
|
| 517 |
+
"grad_norm": 0.463858962059021,
|
| 518 |
+
"learning_rate": 2.066906307977737e-05,
|
| 519 |
+
"loss": 0.5506,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 17.16,
|
| 524 |
+
"grad_norm": 0.5114089250564575,
|
| 525 |
+
"learning_rate": 2.0524118738404453e-05,
|
| 526 |
+
"loss": 0.5512,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 17.39,
|
| 531 |
+
"grad_norm": 0.9562047123908997,
|
| 532 |
+
"learning_rate": 2.037917439703154e-05,
|
| 533 |
+
"loss": 0.5506,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 17.63,
|
| 538 |
+
"grad_norm": 0.755348265171051,
|
| 539 |
+
"learning_rate": 2.0234230055658625e-05,
|
| 540 |
+
"loss": 0.5473,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 17.86,
|
| 545 |
+
"grad_norm": 0.5000788569450378,
|
| 546 |
+
"learning_rate": 2.0089285714285717e-05,
|
| 547 |
+
"loss": 0.5443,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 18.09,
|
| 552 |
+
"grad_norm": 0.5531164407730103,
|
| 553 |
+
"learning_rate": 1.99443413729128e-05,
|
| 554 |
+
"loss": 0.548,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 18.32,
|
| 559 |
+
"grad_norm": 0.49754297733306885,
|
| 560 |
+
"learning_rate": 1.979939703153989e-05,
|
| 561 |
+
"loss": 0.5475,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 18.55,
|
| 566 |
+
"grad_norm": 0.462971955537796,
|
| 567 |
+
"learning_rate": 1.9654452690166977e-05,
|
| 568 |
+
"loss": 0.5459,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 18.78,
|
| 573 |
+
"grad_norm": 0.9089665412902832,
|
| 574 |
+
"learning_rate": 1.950950834879406e-05,
|
| 575 |
+
"loss": 0.5491,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 19.02,
|
| 580 |
+
"grad_norm": 0.47444915771484375,
|
| 581 |
+
"learning_rate": 1.9364564007421153e-05,
|
| 582 |
+
"loss": 0.5433,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 19.25,
|
| 587 |
+
"grad_norm": 0.45020779967308044,
|
| 588 |
+
"learning_rate": 1.9219619666048237e-05,
|
| 589 |
+
"loss": 0.5423,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 19.48,
|
| 594 |
+
"grad_norm": 0.6115174293518066,
|
| 595 |
+
"learning_rate": 1.9074675324675325e-05,
|
| 596 |
+
"loss": 0.5455,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 19.71,
|
| 601 |
+
"grad_norm": 2.4381721019744873,
|
| 602 |
+
"learning_rate": 1.892973098330241e-05,
|
| 603 |
+
"loss": 0.5433,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 19.94,
|
| 608 |
+
"grad_norm": 0.5207477807998657,
|
| 609 |
+
"learning_rate": 1.87847866419295e-05,
|
| 610 |
+
"loss": 0.5414,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 20.18,
|
| 615 |
+
"grad_norm": 0.4906873106956482,
|
| 616 |
+
"learning_rate": 1.8639842300556586e-05,
|
| 617 |
+
"loss": 0.5428,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 20.41,
|
| 622 |
+
"grad_norm": 0.4558812081813812,
|
| 623 |
+
"learning_rate": 1.8494897959183674e-05,
|
| 624 |
+
"loss": 0.5464,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 20.64,
|
| 629 |
+
"grad_norm": 0.5274574756622314,
|
| 630 |
+
"learning_rate": 1.834995361781076e-05,
|
| 631 |
+
"loss": 0.5387,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 20.87,
|
| 636 |
+
"grad_norm": 0.46367010474205017,
|
| 637 |
+
"learning_rate": 1.820500927643785e-05,
|
| 638 |
+
"loss": 0.5407,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 21.1,
|
| 643 |
+
"grad_norm": 0.45425930619239807,
|
| 644 |
+
"learning_rate": 1.8060064935064937e-05,
|
| 645 |
+
"loss": 0.5444,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 21.34,
|
| 650 |
+
"grad_norm": 0.4426890015602112,
|
| 651 |
+
"learning_rate": 1.7915120593692022e-05,
|
| 652 |
+
"loss": 0.5409,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 21.57,
|
| 657 |
+
"grad_norm": 0.4001689851284027,
|
| 658 |
+
"learning_rate": 1.777017625231911e-05,
|
| 659 |
+
"loss": 0.5375,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 21.8,
|
| 664 |
+
"grad_norm": 0.45908644795417786,
|
| 665 |
+
"learning_rate": 1.7625231910946194e-05,
|
| 666 |
+
"loss": 0.5375,
|
| 667 |
+
"step": 47000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 22.03,
|
| 671 |
+
"grad_norm": 0.41069209575653076,
|
| 672 |
+
"learning_rate": 1.7480287569573286e-05,
|
| 673 |
+
"loss": 0.5364,
|
| 674 |
+
"step": 47500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 22.26,
|
| 678 |
+
"grad_norm": 0.5071306824684143,
|
| 679 |
+
"learning_rate": 1.733534322820037e-05,
|
| 680 |
+
"loss": 0.5392,
|
| 681 |
+
"step": 48000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 22.5,
|
| 685 |
+
"grad_norm": 0.5278964638710022,
|
| 686 |
+
"learning_rate": 1.7190398886827458e-05,
|
| 687 |
+
"loss": 0.5351,
|
| 688 |
+
"step": 48500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 22.73,
|
| 692 |
+
"grad_norm": 0.5334340929985046,
|
| 693 |
+
"learning_rate": 1.7045454545454546e-05,
|
| 694 |
+
"loss": 0.5347,
|
| 695 |
+
"step": 49000
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 22.96,
|
| 699 |
+
"grad_norm": 0.47415098547935486,
|
| 700 |
+
"learning_rate": 1.6900510204081634e-05,
|
| 701 |
+
"loss": 0.5378,
|
| 702 |
+
"step": 49500
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 23.19,
|
| 706 |
+
"grad_norm": 0.4583585858345032,
|
| 707 |
+
"learning_rate": 1.6755565862708722e-05,
|
| 708 |
+
"loss": 0.5396,
|
| 709 |
+
"step": 50000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 23.42,
|
| 713 |
+
"grad_norm": 0.42494043707847595,
|
| 714 |
+
"learning_rate": 1.6610621521335807e-05,
|
| 715 |
+
"loss": 0.5354,
|
| 716 |
+
"step": 50500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 23.65,
|
| 720 |
+
"grad_norm": 0.46691572666168213,
|
| 721 |
+
"learning_rate": 1.6465677179962894e-05,
|
| 722 |
+
"loss": 0.5383,
|
| 723 |
+
"step": 51000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 23.89,
|
| 727 |
+
"grad_norm": 0.5475151538848877,
|
| 728 |
+
"learning_rate": 1.6320732838589982e-05,
|
| 729 |
+
"loss": 0.5351,
|
| 730 |
+
"step": 51500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 24.12,
|
| 734 |
+
"grad_norm": 0.4535447061061859,
|
| 735 |
+
"learning_rate": 1.617578849721707e-05,
|
| 736 |
+
"loss": 0.5346,
|
| 737 |
+
"step": 52000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 24.35,
|
| 741 |
+
"grad_norm": 0.5584075450897217,
|
| 742 |
+
"learning_rate": 1.6030844155844155e-05,
|
| 743 |
+
"loss": 0.5345,
|
| 744 |
+
"step": 52500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 24.58,
|
| 748 |
+
"grad_norm": 0.7527787685394287,
|
| 749 |
+
"learning_rate": 1.5885899814471243e-05,
|
| 750 |
+
"loss": 0.5273,
|
| 751 |
+
"step": 53000
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 24.81,
|
| 755 |
+
"grad_norm": 0.39385178685188293,
|
| 756 |
+
"learning_rate": 1.574095547309833e-05,
|
| 757 |
+
"loss": 0.5365,
|
| 758 |
+
"step": 53500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 25.05,
|
| 762 |
+
"grad_norm": 0.5498375296592712,
|
| 763 |
+
"learning_rate": 1.559601113172542e-05,
|
| 764 |
+
"loss": 0.5375,
|
| 765 |
+
"step": 54000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 25.28,
|
| 769 |
+
"grad_norm": 0.41572290658950806,
|
| 770 |
+
"learning_rate": 1.5451066790352507e-05,
|
| 771 |
+
"loss": 0.5335,
|
| 772 |
+
"step": 54500
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 25.51,
|
| 776 |
+
"grad_norm": 0.4475056827068329,
|
| 777 |
+
"learning_rate": 1.530612244897959e-05,
|
| 778 |
+
"loss": 0.5335,
|
| 779 |
+
"step": 55000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 25.74,
|
| 783 |
+
"grad_norm": 0.48311153054237366,
|
| 784 |
+
"learning_rate": 1.516117810760668e-05,
|
| 785 |
+
"loss": 0.5329,
|
| 786 |
+
"step": 55500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 25.97,
|
| 790 |
+
"grad_norm": 0.48509538173675537,
|
| 791 |
+
"learning_rate": 1.5016233766233767e-05,
|
| 792 |
+
"loss": 0.5307,
|
| 793 |
+
"step": 56000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 26.21,
|
| 797 |
+
"grad_norm": 0.5727821588516235,
|
| 798 |
+
"learning_rate": 1.4871289424860855e-05,
|
| 799 |
+
"loss": 0.5316,
|
| 800 |
+
"step": 56500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 26.44,
|
| 804 |
+
"grad_norm": 0.46595117449760437,
|
| 805 |
+
"learning_rate": 1.4726345083487941e-05,
|
| 806 |
+
"loss": 0.5309,
|
| 807 |
+
"step": 57000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 26.67,
|
| 811 |
+
"grad_norm": 0.5022342801094055,
|
| 812 |
+
"learning_rate": 1.4581400742115027e-05,
|
| 813 |
+
"loss": 0.5354,
|
| 814 |
+
"step": 57500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 26.9,
|
| 818 |
+
"grad_norm": 0.5400568842887878,
|
| 819 |
+
"learning_rate": 1.4436456400742115e-05,
|
| 820 |
+
"loss": 0.5268,
|
| 821 |
+
"step": 58000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 27.13,
|
| 825 |
+
"grad_norm": 0.49959874153137207,
|
| 826 |
+
"learning_rate": 1.4291512059369201e-05,
|
| 827 |
+
"loss": 0.5316,
|
| 828 |
+
"step": 58500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 27.37,
|
| 832 |
+
"grad_norm": 0.5435352921485901,
|
| 833 |
+
"learning_rate": 1.414656771799629e-05,
|
| 834 |
+
"loss": 0.53,
|
| 835 |
+
"step": 59000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 27.6,
|
| 839 |
+
"grad_norm": 0.5168443918228149,
|
| 840 |
+
"learning_rate": 1.4001623376623376e-05,
|
| 841 |
+
"loss": 0.5272,
|
| 842 |
+
"step": 59500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 27.83,
|
| 846 |
+
"grad_norm": 0.60955411195755,
|
| 847 |
+
"learning_rate": 1.3856679035250465e-05,
|
| 848 |
+
"loss": 0.5283,
|
| 849 |
+
"step": 60000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 28.06,
|
| 853 |
+
"grad_norm": 0.5063374042510986,
|
| 854 |
+
"learning_rate": 1.3711734693877551e-05,
|
| 855 |
+
"loss": 0.5308,
|
| 856 |
+
"step": 60500
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 28.29,
|
| 860 |
+
"grad_norm": 0.46151864528656006,
|
| 861 |
+
"learning_rate": 1.356679035250464e-05,
|
| 862 |
+
"loss": 0.5282,
|
| 863 |
+
"step": 61000
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 28.53,
|
| 867 |
+
"grad_norm": 0.5224559903144836,
|
| 868 |
+
"learning_rate": 1.3421846011131726e-05,
|
| 869 |
+
"loss": 0.5261,
|
| 870 |
+
"step": 61500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 28.76,
|
| 874 |
+
"grad_norm": 0.43721848726272583,
|
| 875 |
+
"learning_rate": 1.3276901669758814e-05,
|
| 876 |
+
"loss": 0.525,
|
| 877 |
+
"step": 62000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 28.99,
|
| 881 |
+
"grad_norm": 0.5033512115478516,
|
| 882 |
+
"learning_rate": 1.31319573283859e-05,
|
| 883 |
+
"loss": 0.5304,
|
| 884 |
+
"step": 62500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 29.22,
|
| 888 |
+
"grad_norm": 0.5284975171089172,
|
| 889 |
+
"learning_rate": 1.2987012987012988e-05,
|
| 890 |
+
"loss": 0.5291,
|
| 891 |
+
"step": 63000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 29.45,
|
| 895 |
+
"grad_norm": 0.6417981386184692,
|
| 896 |
+
"learning_rate": 1.2842068645640074e-05,
|
| 897 |
+
"loss": 0.5267,
|
| 898 |
+
"step": 63500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 29.68,
|
| 902 |
+
"grad_norm": 0.5176807045936584,
|
| 903 |
+
"learning_rate": 1.269712430426716e-05,
|
| 904 |
+
"loss": 0.5246,
|
| 905 |
+
"step": 64000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 29.92,
|
| 909 |
+
"grad_norm": 0.46041619777679443,
|
| 910 |
+
"learning_rate": 1.255217996289425e-05,
|
| 911 |
+
"loss": 0.527,
|
| 912 |
+
"step": 64500
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 30.15,
|
| 916 |
+
"grad_norm": 0.5162688493728638,
|
| 917 |
+
"learning_rate": 1.2407235621521336e-05,
|
| 918 |
+
"loss": 0.5242,
|
| 919 |
+
"step": 65000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 30.38,
|
| 923 |
+
"grad_norm": 0.46508926153182983,
|
| 924 |
+
"learning_rate": 1.2262291280148424e-05,
|
| 925 |
+
"loss": 0.5234,
|
| 926 |
+
"step": 65500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 30.61,
|
| 930 |
+
"grad_norm": 0.4786316156387329,
|
| 931 |
+
"learning_rate": 1.211734693877551e-05,
|
| 932 |
+
"loss": 0.5287,
|
| 933 |
+
"step": 66000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 30.84,
|
| 937 |
+
"grad_norm": 0.5851900577545166,
|
| 938 |
+
"learning_rate": 1.1972402597402598e-05,
|
| 939 |
+
"loss": 0.5252,
|
| 940 |
+
"step": 66500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 31.08,
|
| 944 |
+
"grad_norm": 0.5234825611114502,
|
| 945 |
+
"learning_rate": 1.1827458256029684e-05,
|
| 946 |
+
"loss": 0.5238,
|
| 947 |
+
"step": 67000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 31.31,
|
| 951 |
+
"grad_norm": 0.4396039843559265,
|
| 952 |
+
"learning_rate": 1.1682513914656772e-05,
|
| 953 |
+
"loss": 0.523,
|
| 954 |
+
"step": 67500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 31.54,
|
| 958 |
+
"grad_norm": 0.4660816788673401,
|
| 959 |
+
"learning_rate": 1.1537569573283858e-05,
|
| 960 |
+
"loss": 0.5254,
|
| 961 |
+
"step": 68000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 31.77,
|
| 965 |
+
"grad_norm": 0.4561219811439514,
|
| 966 |
+
"learning_rate": 1.1392625231910948e-05,
|
| 967 |
+
"loss": 0.5202,
|
| 968 |
+
"step": 68500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 32.0,
|
| 972 |
+
"grad_norm": 0.4771936535835266,
|
| 973 |
+
"learning_rate": 1.1247680890538034e-05,
|
| 974 |
+
"loss": 0.526,
|
| 975 |
+
"step": 69000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 32.24,
|
| 979 |
+
"grad_norm": 0.5595078468322754,
|
| 980 |
+
"learning_rate": 1.110273654916512e-05,
|
| 981 |
+
"loss": 0.5233,
|
| 982 |
+
"step": 69500
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 32.47,
|
| 986 |
+
"grad_norm": 0.5574076175689697,
|
| 987 |
+
"learning_rate": 1.0957792207792208e-05,
|
| 988 |
+
"loss": 0.5242,
|
| 989 |
+
"step": 70000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 32.7,
|
| 993 |
+
"grad_norm": 0.5540217161178589,
|
| 994 |
+
"learning_rate": 1.0812847866419295e-05,
|
| 995 |
+
"loss": 0.5207,
|
| 996 |
+
"step": 70500
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 32.93,
|
| 1000 |
+
"grad_norm": 0.4957340359687805,
|
| 1001 |
+
"learning_rate": 1.0667903525046383e-05,
|
| 1002 |
+
"loss": 0.5234,
|
| 1003 |
+
"step": 71000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 33.16,
|
| 1007 |
+
"grad_norm": 0.5124360918998718,
|
| 1008 |
+
"learning_rate": 1.0522959183673469e-05,
|
| 1009 |
+
"loss": 0.5215,
|
| 1010 |
+
"step": 71500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 33.4,
|
| 1014 |
+
"grad_norm": 0.46945714950561523,
|
| 1015 |
+
"learning_rate": 1.0378014842300557e-05,
|
| 1016 |
+
"loss": 0.5253,
|
| 1017 |
+
"step": 72000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 33.63,
|
| 1021 |
+
"grad_norm": 0.379711389541626,
|
| 1022 |
+
"learning_rate": 1.0233070500927643e-05,
|
| 1023 |
+
"loss": 0.5217,
|
| 1024 |
+
"step": 72500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 33.86,
|
| 1028 |
+
"grad_norm": 0.6671276092529297,
|
| 1029 |
+
"learning_rate": 1.0088126159554733e-05,
|
| 1030 |
+
"loss": 0.5184,
|
| 1031 |
+
"step": 73000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 34.09,
|
| 1035 |
+
"grad_norm": 0.5381418466567993,
|
| 1036 |
+
"learning_rate": 9.943181818181819e-06,
|
| 1037 |
+
"loss": 0.5273,
|
| 1038 |
+
"step": 73500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 34.32,
|
| 1042 |
+
"grad_norm": 0.6995874643325806,
|
| 1043 |
+
"learning_rate": 9.798237476808907e-06,
|
| 1044 |
+
"loss": 0.5203,
|
| 1045 |
+
"step": 74000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 34.55,
|
| 1049 |
+
"grad_norm": 0.4824086129665375,
|
| 1050 |
+
"learning_rate": 9.653293135435993e-06,
|
| 1051 |
+
"loss": 0.5186,
|
| 1052 |
+
"step": 74500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 34.79,
|
| 1056 |
+
"grad_norm": 0.5302020311355591,
|
| 1057 |
+
"learning_rate": 9.50834879406308e-06,
|
| 1058 |
+
"loss": 0.5194,
|
| 1059 |
+
"step": 75000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 35.02,
|
| 1063 |
+
"grad_norm": 0.5452972054481506,
|
| 1064 |
+
"learning_rate": 9.363404452690167e-06,
|
| 1065 |
+
"loss": 0.5244,
|
| 1066 |
+
"step": 75500
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 35.25,
|
| 1070 |
+
"grad_norm": 0.6910521984100342,
|
| 1071 |
+
"learning_rate": 9.218460111317253e-06,
|
| 1072 |
+
"loss": 0.5185,
|
| 1073 |
+
"step": 76000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 35.48,
|
| 1077 |
+
"grad_norm": 0.5710629224777222,
|
| 1078 |
+
"learning_rate": 9.073515769944341e-06,
|
| 1079 |
+
"loss": 0.519,
|
| 1080 |
+
"step": 76500
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 35.71,
|
| 1084 |
+
"grad_norm": 0.4669000804424286,
|
| 1085 |
+
"learning_rate": 8.928571428571428e-06,
|
| 1086 |
+
"loss": 0.5171,
|
| 1087 |
+
"step": 77000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 35.95,
|
| 1091 |
+
"grad_norm": 0.508492112159729,
|
| 1092 |
+
"learning_rate": 8.783627087198517e-06,
|
| 1093 |
+
"loss": 0.5201,
|
| 1094 |
+
"step": 77500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 36.18,
|
| 1098 |
+
"grad_norm": 0.5005605816841125,
|
| 1099 |
+
"learning_rate": 8.638682745825603e-06,
|
| 1100 |
+
"loss": 0.5181,
|
| 1101 |
+
"step": 78000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 36.41,
|
| 1105 |
+
"grad_norm": 0.5115915536880493,
|
| 1106 |
+
"learning_rate": 8.493738404452691e-06,
|
| 1107 |
+
"loss": 0.517,
|
| 1108 |
+
"step": 78500
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 36.64,
|
| 1112 |
+
"grad_norm": 0.5187366604804993,
|
| 1113 |
+
"learning_rate": 8.348794063079778e-06,
|
| 1114 |
+
"loss": 0.5197,
|
| 1115 |
+
"step": 79000
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 36.87,
|
| 1119 |
+
"grad_norm": 0.5845080614089966,
|
| 1120 |
+
"learning_rate": 8.203849721706865e-06,
|
| 1121 |
+
"loss": 0.5209,
|
| 1122 |
+
"step": 79500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 37.11,
|
| 1126 |
+
"grad_norm": 0.6533055901527405,
|
| 1127 |
+
"learning_rate": 8.058905380333952e-06,
|
| 1128 |
+
"loss": 0.521,
|
| 1129 |
+
"step": 80000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 37.34,
|
| 1133 |
+
"grad_norm": 0.4986008107662201,
|
| 1134 |
+
"learning_rate": 7.913961038961038e-06,
|
| 1135 |
+
"loss": 0.5172,
|
| 1136 |
+
"step": 80500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 37.57,
|
| 1140 |
+
"grad_norm": 0.557759165763855,
|
| 1141 |
+
"learning_rate": 7.769016697588126e-06,
|
| 1142 |
+
"loss": 0.5169,
|
| 1143 |
+
"step": 81000
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 37.8,
|
| 1147 |
+
"grad_norm": 0.651156485080719,
|
| 1148 |
+
"learning_rate": 7.624072356215214e-06,
|
| 1149 |
+
"loss": 0.5179,
|
| 1150 |
+
"step": 81500
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 38.03,
|
| 1154 |
+
"grad_norm": 0.5263897180557251,
|
| 1155 |
+
"learning_rate": 7.479128014842301e-06,
|
| 1156 |
+
"loss": 0.5202,
|
| 1157 |
+
"step": 82000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 38.27,
|
| 1161 |
+
"grad_norm": 0.497855007648468,
|
| 1162 |
+
"learning_rate": 7.334183673469388e-06,
|
| 1163 |
+
"loss": 0.5169,
|
| 1164 |
+
"step": 82500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 38.5,
|
| 1168 |
+
"grad_norm": 0.5888237357139587,
|
| 1169 |
+
"learning_rate": 7.189239332096476e-06,
|
| 1170 |
+
"loss": 0.5172,
|
| 1171 |
+
"step": 83000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 38.73,
|
| 1175 |
+
"grad_norm": 0.7321680188179016,
|
| 1176 |
+
"learning_rate": 7.044294990723562e-06,
|
| 1177 |
+
"loss": 0.5175,
|
| 1178 |
+
"step": 83500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 38.96,
|
| 1182 |
+
"grad_norm": 0.5042610168457031,
|
| 1183 |
+
"learning_rate": 6.899350649350649e-06,
|
| 1184 |
+
"loss": 0.5175,
|
| 1185 |
+
"step": 84000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 39.19,
|
| 1189 |
+
"grad_norm": 0.6449227333068848,
|
| 1190 |
+
"learning_rate": 6.754406307977736e-06,
|
| 1191 |
+
"loss": 0.5185,
|
| 1192 |
+
"step": 84500
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 39.42,
|
| 1196 |
+
"grad_norm": 0.4842873215675354,
|
| 1197 |
+
"learning_rate": 6.609461966604823e-06,
|
| 1198 |
+
"loss": 0.5174,
|
| 1199 |
+
"step": 85000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 39.66,
|
| 1203 |
+
"grad_norm": 0.5853040218353271,
|
| 1204 |
+
"learning_rate": 6.464517625231911e-06,
|
| 1205 |
+
"loss": 0.5158,
|
| 1206 |
+
"step": 85500
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 39.89,
|
| 1210 |
+
"grad_norm": 0.6231125593185425,
|
| 1211 |
+
"learning_rate": 6.319573283858998e-06,
|
| 1212 |
+
"loss": 0.518,
|
| 1213 |
+
"step": 86000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 40.12,
|
| 1217 |
+
"grad_norm": 0.7321327328681946,
|
| 1218 |
+
"learning_rate": 6.174628942486085e-06,
|
| 1219 |
+
"loss": 0.5101,
|
| 1220 |
+
"step": 86500
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 40.35,
|
| 1224 |
+
"grad_norm": 0.48840710520744324,
|
| 1225 |
+
"learning_rate": 6.0296846011131725e-06,
|
| 1226 |
+
"loss": 0.5188,
|
| 1227 |
+
"step": 87000
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 40.58,
|
| 1231 |
+
"grad_norm": 0.4609430134296417,
|
| 1232 |
+
"learning_rate": 5.88474025974026e-06,
|
| 1233 |
+
"loss": 0.5156,
|
| 1234 |
+
"step": 87500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 40.82,
|
| 1238 |
+
"grad_norm": 0.7623891234397888,
|
| 1239 |
+
"learning_rate": 5.7397959183673475e-06,
|
| 1240 |
+
"loss": 0.5135,
|
| 1241 |
+
"step": 88000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 41.05,
|
| 1245 |
+
"grad_norm": 0.6465628147125244,
|
| 1246 |
+
"learning_rate": 5.5948515769944346e-06,
|
| 1247 |
+
"loss": 0.5148,
|
| 1248 |
+
"step": 88500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 41.28,
|
| 1252 |
+
"grad_norm": 0.6189027428627014,
|
| 1253 |
+
"learning_rate": 5.449907235621521e-06,
|
| 1254 |
+
"loss": 0.5162,
|
| 1255 |
+
"step": 89000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 41.51,
|
| 1259 |
+
"grad_norm": 0.48868080973625183,
|
| 1260 |
+
"learning_rate": 5.304962894248609e-06,
|
| 1261 |
+
"loss": 0.516,
|
| 1262 |
+
"step": 89500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 41.74,
|
| 1266 |
+
"grad_norm": 0.5011306405067444,
|
| 1267 |
+
"learning_rate": 5.160018552875696e-06,
|
| 1268 |
+
"loss": 0.513,
|
| 1269 |
+
"step": 90000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 41.98,
|
| 1273 |
+
"grad_norm": 0.596338152885437,
|
| 1274 |
+
"learning_rate": 5.015074211502783e-06,
|
| 1275 |
+
"loss": 0.5149,
|
| 1276 |
+
"step": 90500
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 42.21,
|
| 1280 |
+
"grad_norm": 0.681685209274292,
|
| 1281 |
+
"learning_rate": 4.87012987012987e-06,
|
| 1282 |
+
"loss": 0.5139,
|
| 1283 |
+
"step": 91000
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 42.44,
|
| 1287 |
+
"grad_norm": 0.6109485030174255,
|
| 1288 |
+
"learning_rate": 4.725185528756957e-06,
|
| 1289 |
+
"loss": 0.5145,
|
| 1290 |
+
"step": 91500
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 42.67,
|
| 1294 |
+
"grad_norm": 0.6468715071678162,
|
| 1295 |
+
"learning_rate": 4.580241187384045e-06,
|
| 1296 |
+
"loss": 0.5138,
|
| 1297 |
+
"step": 92000
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 42.9,
|
| 1301 |
+
"grad_norm": 0.47962233424186707,
|
| 1302 |
+
"learning_rate": 4.435296846011132e-06,
|
| 1303 |
+
"loss": 0.5095,
|
| 1304 |
+
"step": 92500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 43.14,
|
| 1308 |
+
"grad_norm": 0.6052845120429993,
|
| 1309 |
+
"learning_rate": 4.290352504638219e-06,
|
| 1310 |
+
"loss": 0.5129,
|
| 1311 |
+
"step": 93000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 43.37,
|
| 1315 |
+
"grad_norm": 0.49540799856185913,
|
| 1316 |
+
"learning_rate": 4.145408163265306e-06,
|
| 1317 |
+
"loss": 0.5161,
|
| 1318 |
+
"step": 93500
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 43.6,
|
| 1322 |
+
"grad_norm": 0.5836204290390015,
|
| 1323 |
+
"learning_rate": 4.000463821892394e-06,
|
| 1324 |
+
"loss": 0.509,
|
| 1325 |
+
"step": 94000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 43.83,
|
| 1329 |
+
"grad_norm": 0.42629340291023254,
|
| 1330 |
+
"learning_rate": 3.85551948051948e-06,
|
| 1331 |
+
"loss": 0.5135,
|
| 1332 |
+
"step": 94500
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 44.06,
|
| 1336 |
+
"grad_norm": 0.4765370488166809,
|
| 1337 |
+
"learning_rate": 3.710575139146568e-06,
|
| 1338 |
+
"loss": 0.5107,
|
| 1339 |
+
"step": 95000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 44.29,
|
| 1343 |
+
"grad_norm": 0.6870307326316833,
|
| 1344 |
+
"learning_rate": 3.565630797773655e-06,
|
| 1345 |
+
"loss": 0.5134,
|
| 1346 |
+
"step": 95500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 44.53,
|
| 1350 |
+
"grad_norm": 0.4927411675453186,
|
| 1351 |
+
"learning_rate": 3.4206864564007424e-06,
|
| 1352 |
+
"loss": 0.5083,
|
| 1353 |
+
"step": 96000
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 44.76,
|
| 1357 |
+
"grad_norm": 0.5409857034683228,
|
| 1358 |
+
"learning_rate": 3.275742115027829e-06,
|
| 1359 |
+
"loss": 0.5133,
|
| 1360 |
+
"step": 96500
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 44.99,
|
| 1364 |
+
"grad_norm": 0.5635824799537659,
|
| 1365 |
+
"learning_rate": 3.1307977736549166e-06,
|
| 1366 |
+
"loss": 0.5111,
|
| 1367 |
+
"step": 97000
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 45.22,
|
| 1371 |
+
"grad_norm": 0.6565483212471008,
|
| 1372 |
+
"learning_rate": 2.9858534322820037e-06,
|
| 1373 |
+
"loss": 0.5099,
|
| 1374 |
+
"step": 97500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 45.45,
|
| 1378 |
+
"grad_norm": 0.659949779510498,
|
| 1379 |
+
"learning_rate": 2.840909090909091e-06,
|
| 1380 |
+
"loss": 0.5124,
|
| 1381 |
+
"step": 98000
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 45.69,
|
| 1385 |
+
"grad_norm": 0.551925539970398,
|
| 1386 |
+
"learning_rate": 2.6959647495361782e-06,
|
| 1387 |
+
"loss": 0.5105,
|
| 1388 |
+
"step": 98500
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 45.92,
|
| 1392 |
+
"grad_norm": 0.6591493487358093,
|
| 1393 |
+
"learning_rate": 2.5510204081632653e-06,
|
| 1394 |
+
"loss": 0.5106,
|
| 1395 |
+
"step": 99000
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 46.15,
|
| 1399 |
+
"grad_norm": 0.6934979557991028,
|
| 1400 |
+
"learning_rate": 2.4060760667903524e-06,
|
| 1401 |
+
"loss": 0.5109,
|
| 1402 |
+
"step": 99500
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 46.38,
|
| 1406 |
+
"grad_norm": 0.5429938435554504,
|
| 1407 |
+
"learning_rate": 2.26113172541744e-06,
|
| 1408 |
+
"loss": 0.5081,
|
| 1409 |
+
"step": 100000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 46.61,
|
| 1413 |
+
"grad_norm": 0.5906868577003479,
|
| 1414 |
+
"learning_rate": 2.116187384044527e-06,
|
| 1415 |
+
"loss": 0.5106,
|
| 1416 |
+
"step": 100500
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 46.85,
|
| 1420 |
+
"grad_norm": 0.7781776785850525,
|
| 1421 |
+
"learning_rate": 1.971243042671614e-06,
|
| 1422 |
+
"loss": 0.5131,
|
| 1423 |
+
"step": 101000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 47.08,
|
| 1427 |
+
"grad_norm": 0.5847499966621399,
|
| 1428 |
+
"learning_rate": 1.8262987012987013e-06,
|
| 1429 |
+
"loss": 0.5101,
|
| 1430 |
+
"step": 101500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 47.31,
|
| 1434 |
+
"grad_norm": 0.49553003907203674,
|
| 1435 |
+
"learning_rate": 1.6813543599257886e-06,
|
| 1436 |
+
"loss": 0.5129,
|
| 1437 |
+
"step": 102000
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 47.54,
|
| 1441 |
+
"grad_norm": 0.581206738948822,
|
| 1442 |
+
"learning_rate": 1.5364100185528757e-06,
|
| 1443 |
+
"loss": 0.5122,
|
| 1444 |
+
"step": 102500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 47.77,
|
| 1448 |
+
"grad_norm": 0.7169119715690613,
|
| 1449 |
+
"learning_rate": 1.391465677179963e-06,
|
| 1450 |
+
"loss": 0.5075,
|
| 1451 |
+
"step": 103000
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 48.01,
|
| 1455 |
+
"grad_norm": 0.7762653231620789,
|
| 1456 |
+
"learning_rate": 1.24652133580705e-06,
|
| 1457 |
+
"loss": 0.5121,
|
| 1458 |
+
"step": 103500
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 48.24,
|
| 1462 |
+
"grad_norm": 0.6211891770362854,
|
| 1463 |
+
"learning_rate": 1.1015769944341372e-06,
|
| 1464 |
+
"loss": 0.5096,
|
| 1465 |
+
"step": 104000
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 48.47,
|
| 1469 |
+
"grad_norm": 0.6150048971176147,
|
| 1470 |
+
"learning_rate": 9.566326530612244e-07,
|
| 1471 |
+
"loss": 0.5141,
|
| 1472 |
+
"step": 104500
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 48.7,
|
| 1476 |
+
"grad_norm": 0.5785318613052368,
|
| 1477 |
+
"learning_rate": 8.116883116883117e-07,
|
| 1478 |
+
"loss": 0.5066,
|
| 1479 |
+
"step": 105000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 48.93,
|
| 1483 |
+
"grad_norm": 0.64094078540802,
|
| 1484 |
+
"learning_rate": 6.667439703153989e-07,
|
| 1485 |
+
"loss": 0.5066,
|
| 1486 |
+
"step": 105500
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 49.17,
|
| 1490 |
+
"grad_norm": 0.5458254218101501,
|
| 1491 |
+
"learning_rate": 5.217996289424861e-07,
|
| 1492 |
+
"loss": 0.5083,
|
| 1493 |
+
"step": 106000
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 49.4,
|
| 1497 |
+
"grad_norm": 0.6012634634971619,
|
| 1498 |
+
"learning_rate": 3.768552875695733e-07,
|
| 1499 |
+
"loss": 0.5112,
|
| 1500 |
+
"step": 106500
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 49.63,
|
| 1504 |
+
"grad_norm": 0.6441764235496521,
|
| 1505 |
+
"learning_rate": 2.319109461966605e-07,
|
| 1506 |
+
"loss": 0.5094,
|
| 1507 |
+
"step": 107000
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 49.86,
|
| 1511 |
+
"grad_norm": 0.5359529852867126,
|
| 1512 |
+
"learning_rate": 8.696660482374769e-08,
|
| 1513 |
+
"loss": 0.5095,
|
| 1514 |
+
"step": 107500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 50.0,
|
| 1518 |
+
"step": 107800,
|
| 1519 |
+
"total_flos": 8.716413062401229e+19,
|
| 1520 |
+
"train_loss": 0.5550972292020724,
|
| 1521 |
+
"train_runtime": 13976.6946,
|
| 1522 |
+
"train_samples_per_second": 61.703,
|
| 1523 |
+
"train_steps_per_second": 7.713
|
| 1524 |
+
}
|
| 1525 |
+
],
|
| 1526 |
+
"logging_steps": 500,
|
| 1527 |
+
"max_steps": 107800,
|
| 1528 |
+
"num_input_tokens_seen": 0,
|
| 1529 |
+
"num_train_epochs": 50,
|
| 1530 |
+
"save_steps": 1000000000,
|
| 1531 |
+
"total_flos": 8.716413062401229e+19,
|
| 1532 |
+
"train_batch_size": 8,
|
| 1533 |
+
"trial_name": null,
|
| 1534 |
+
"trial_params": null
|
| 1535 |
+
}
|