End of training
Browse files- README.md +20 -2
- all_results.json +16 -0
- eval_results.json +10 -0
- runs/Jul30_00-37-31_8ff5e8032af8/events.out.tfevents.1722307823.8ff5e8032af8.1173.1 +3 -0
- train_results.json +9 -0
- trainer_state.json +3381 -0
README.md
CHANGED
@@ -1,9 +1,24 @@
|
|
1 |
---
|
2 |
tags:
|
3 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
4 |
model-index:
|
5 |
- name: tinybert-javanese
|
6 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
---
|
8 |
|
9 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -11,7 +26,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
11 |
|
12 |
# tinybert-javanese
|
13 |
|
14 |
-
This model is a fine-tuned version of [](https://huggingface.co/) on
|
|
|
|
|
|
|
15 |
|
16 |
## Model description
|
17 |
|
|
|
1 |
---
|
2 |
tags:
|
3 |
- generated_from_trainer
|
4 |
+
datasets:
|
5 |
+
- akahana/GlotCC-V1-jav-Latn
|
6 |
+
metrics:
|
7 |
+
- accuracy
|
8 |
model-index:
|
9 |
- name: tinybert-javanese
|
10 |
+
results:
|
11 |
+
- task:
|
12 |
+
name: Masked Language Modeling
|
13 |
+
type: fill-mask
|
14 |
+
dataset:
|
15 |
+
name: akahana/GlotCC-V1-jav-Latn default
|
16 |
+
type: akahana/GlotCC-V1-jav-Latn
|
17 |
+
args: default
|
18 |
+
metrics:
|
19 |
+
- name: Accuracy
|
20 |
+
type: accuracy
|
21 |
+
value: 0.1519208618470086
|
22 |
---
|
23 |
|
24 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
26 |
|
27 |
# tinybert-javanese
|
28 |
|
29 |
+
This model is a fine-tuned version of [](https://huggingface.co/) on the akahana/GlotCC-V1-jav-Latn default dataset.
|
30 |
+
It achieves the following results on the evaluation set:
|
31 |
+
- Loss: 5.8252
|
32 |
+
- Accuracy: 0.1519
|
33 |
|
34 |
## Model description
|
35 |
|
all_results.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"eval_accuracy": 0.1519208618470086,
|
4 |
+
"eval_loss": 5.8251872062683105,
|
5 |
+
"eval_runtime": 11.1208,
|
6 |
+
"eval_samples": 3852,
|
7 |
+
"eval_samples_per_second": 346.376,
|
8 |
+
"eval_steps_per_second": 86.594,
|
9 |
+
"perplexity": 338.7245384996336,
|
10 |
+
"total_flos": 5161725447936000.0,
|
11 |
+
"train_loss": 6.140905278960581,
|
12 |
+
"train_runtime": 7883.0646,
|
13 |
+
"train_samples": 76378,
|
14 |
+
"train_samples_per_second": 484.444,
|
15 |
+
"train_steps_per_second": 30.28
|
16 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"eval_accuracy": 0.1519208618470086,
|
4 |
+
"eval_loss": 5.8251872062683105,
|
5 |
+
"eval_runtime": 11.1208,
|
6 |
+
"eval_samples": 3852,
|
7 |
+
"eval_samples_per_second": 346.376,
|
8 |
+
"eval_steps_per_second": 86.594,
|
9 |
+
"perplexity": 338.7245384996336
|
10 |
+
}
|
runs/Jul30_00-37-31_8ff5e8032af8/events.out.tfevents.1722307823.8ff5e8032af8.1173.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f6fd5cd4ca647ecdbbc92d2c3fd74bd22d3bb62df7be93d03a8546aa5a3ff644
|
3 |
+
size 417
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"total_flos": 5161725447936000.0,
|
4 |
+
"train_loss": 6.140905278960581,
|
5 |
+
"train_runtime": 7883.0646,
|
6 |
+
"train_samples": 76378,
|
7 |
+
"train_samples_per_second": 484.444,
|
8 |
+
"train_steps_per_second": 30.28
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,3381 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 50.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 238700,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.10473397570171764,
|
13 |
+
"grad_norm": 1.1384308338165283,
|
14 |
+
"learning_rate": 4.989526602429829e-05,
|
15 |
+
"loss": 8.9672,
|
16 |
+
"step": 500
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.20946795140343527,
|
20 |
+
"grad_norm": 0.9985808730125427,
|
21 |
+
"learning_rate": 4.979053204859657e-05,
|
22 |
+
"loss": 7.7253,
|
23 |
+
"step": 1000
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.31420192710515293,
|
27 |
+
"grad_norm": 1.0521348714828491,
|
28 |
+
"learning_rate": 4.968579807289485e-05,
|
29 |
+
"loss": 7.5614,
|
30 |
+
"step": 1500
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.41893590280687054,
|
34 |
+
"grad_norm": 1.0602227449417114,
|
35 |
+
"learning_rate": 4.958106409719313e-05,
|
36 |
+
"loss": 7.5037,
|
37 |
+
"step": 2000
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.5236698785085881,
|
41 |
+
"grad_norm": 1.6102268695831299,
|
42 |
+
"learning_rate": 4.9476330121491414e-05,
|
43 |
+
"loss": 7.4595,
|
44 |
+
"step": 2500
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.6284038542103059,
|
48 |
+
"grad_norm": 1.335976004600525,
|
49 |
+
"learning_rate": 4.9371596145789694e-05,
|
50 |
+
"loss": 7.4267,
|
51 |
+
"step": 3000
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.7331378299120235,
|
55 |
+
"grad_norm": 1.340728998184204,
|
56 |
+
"learning_rate": 4.926686217008798e-05,
|
57 |
+
"loss": 7.392,
|
58 |
+
"step": 3500
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.8378718056137411,
|
62 |
+
"grad_norm": 1.4520059823989868,
|
63 |
+
"learning_rate": 4.916212819438626e-05,
|
64 |
+
"loss": 7.3253,
|
65 |
+
"step": 4000
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.9426057813154587,
|
69 |
+
"grad_norm": 1.7685532569885254,
|
70 |
+
"learning_rate": 4.905760368663595e-05,
|
71 |
+
"loss": 7.3204,
|
72 |
+
"step": 4500
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 1.0473397570171763,
|
76 |
+
"grad_norm": 1.396130084991455,
|
77 |
+
"learning_rate": 4.8952869710934226e-05,
|
78 |
+
"loss": 7.2682,
|
79 |
+
"step": 5000
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 1.1520737327188941,
|
83 |
+
"grad_norm": 1.5962079763412476,
|
84 |
+
"learning_rate": 4.884813573523251e-05,
|
85 |
+
"loss": 7.2441,
|
86 |
+
"step": 5500
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 1.2568077084206117,
|
90 |
+
"grad_norm": 1.6328166723251343,
|
91 |
+
"learning_rate": 4.874340175953079e-05,
|
92 |
+
"loss": 7.2216,
|
93 |
+
"step": 6000
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 1.3615416841223293,
|
97 |
+
"grad_norm": 1.8534362316131592,
|
98 |
+
"learning_rate": 4.863887725178048e-05,
|
99 |
+
"loss": 7.1957,
|
100 |
+
"step": 6500
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 1.466275659824047,
|
104 |
+
"grad_norm": 1.4871692657470703,
|
105 |
+
"learning_rate": 4.8534143276078766e-05,
|
106 |
+
"loss": 7.1561,
|
107 |
+
"step": 7000
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 1.5710096355257646,
|
111 |
+
"grad_norm": 1.8590672016143799,
|
112 |
+
"learning_rate": 4.8429409300377045e-05,
|
113 |
+
"loss": 7.1397,
|
114 |
+
"step": 7500
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 1.6757436112274822,
|
118 |
+
"grad_norm": 1.7009446620941162,
|
119 |
+
"learning_rate": 4.8324675324675325e-05,
|
120 |
+
"loss": 7.1149,
|
121 |
+
"step": 8000
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.7804775869291998,
|
125 |
+
"grad_norm": 1.9020010232925415,
|
126 |
+
"learning_rate": 4.822015081692501e-05,
|
127 |
+
"loss": 7.1215,
|
128 |
+
"step": 8500
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 1.8852115626309174,
|
132 |
+
"grad_norm": 2.912442445755005,
|
133 |
+
"learning_rate": 4.811541684122329e-05,
|
134 |
+
"loss": 7.0616,
|
135 |
+
"step": 9000
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.989945538332635,
|
139 |
+
"grad_norm": 2.654263496398926,
|
140 |
+
"learning_rate": 4.801068286552158e-05,
|
141 |
+
"loss": 7.0508,
|
142 |
+
"step": 9500
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 2.0946795140343526,
|
146 |
+
"grad_norm": 2.1642003059387207,
|
147 |
+
"learning_rate": 4.790594888981986e-05,
|
148 |
+
"loss": 7.0251,
|
149 |
+
"step": 10000
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 2.19941348973607,
|
153 |
+
"grad_norm": 1.9420874118804932,
|
154 |
+
"learning_rate": 4.7801424382069544e-05,
|
155 |
+
"loss": 6.9806,
|
156 |
+
"step": 10500
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 2.3041474654377883,
|
160 |
+
"grad_norm": 2.2306201457977295,
|
161 |
+
"learning_rate": 4.769669040636783e-05,
|
162 |
+
"loss": 6.9721,
|
163 |
+
"step": 11000
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 2.4088814411395054,
|
167 |
+
"grad_norm": 2.8180389404296875,
|
168 |
+
"learning_rate": 4.759195643066611e-05,
|
169 |
+
"loss": 6.9582,
|
170 |
+
"step": 11500
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 2.5136154168412235,
|
174 |
+
"grad_norm": 2.387949228286743,
|
175 |
+
"learning_rate": 4.748722245496439e-05,
|
176 |
+
"loss": 6.9211,
|
177 |
+
"step": 12000
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 2.618349392542941,
|
181 |
+
"grad_norm": 3.3709394931793213,
|
182 |
+
"learning_rate": 4.7382697947214076e-05,
|
183 |
+
"loss": 6.9183,
|
184 |
+
"step": 12500
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 2.7230833682446587,
|
188 |
+
"grad_norm": 2.567798376083374,
|
189 |
+
"learning_rate": 4.727796397151236e-05,
|
190 |
+
"loss": 6.8732,
|
191 |
+
"step": 13000
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 2.8278173439463763,
|
195 |
+
"grad_norm": 2.6373414993286133,
|
196 |
+
"learning_rate": 4.717322999581064e-05,
|
197 |
+
"loss": 6.8658,
|
198 |
+
"step": 13500
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 2.932551319648094,
|
202 |
+
"grad_norm": 2.2950875759124756,
|
203 |
+
"learning_rate": 4.706849602010893e-05,
|
204 |
+
"loss": 6.8436,
|
205 |
+
"step": 14000
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 3.0372852953498115,
|
209 |
+
"grad_norm": 4.0021514892578125,
|
210 |
+
"learning_rate": 4.696397151235861e-05,
|
211 |
+
"loss": 6.8496,
|
212 |
+
"step": 14500
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 3.142019271051529,
|
216 |
+
"grad_norm": 3.289193630218506,
|
217 |
+
"learning_rate": 4.685923753665689e-05,
|
218 |
+
"loss": 6.8047,
|
219 |
+
"step": 15000
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 3.2467532467532467,
|
223 |
+
"grad_norm": 2.9973654747009277,
|
224 |
+
"learning_rate": 4.6754503560955175e-05,
|
225 |
+
"loss": 6.7651,
|
226 |
+
"step": 15500
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 3.3514872224549643,
|
230 |
+
"grad_norm": 2.9979376792907715,
|
231 |
+
"learning_rate": 4.664976958525346e-05,
|
232 |
+
"loss": 6.7768,
|
233 |
+
"step": 16000
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 3.456221198156682,
|
237 |
+
"grad_norm": 3.263784885406494,
|
238 |
+
"learning_rate": 4.654524507750315e-05,
|
239 |
+
"loss": 6.7617,
|
240 |
+
"step": 16500
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 3.5609551738583995,
|
244 |
+
"grad_norm": 3.330116033554077,
|
245 |
+
"learning_rate": 4.644051110180143e-05,
|
246 |
+
"loss": 6.7417,
|
247 |
+
"step": 17000
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 3.665689149560117,
|
251 |
+
"grad_norm": 3.224337339401245,
|
252 |
+
"learning_rate": 4.6335777126099714e-05,
|
253 |
+
"loss": 6.7028,
|
254 |
+
"step": 17500
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 3.7704231252618348,
|
258 |
+
"grad_norm": 3.21891450881958,
|
259 |
+
"learning_rate": 4.623104315039799e-05,
|
260 |
+
"loss": 6.7101,
|
261 |
+
"step": 18000
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 3.875157100963553,
|
265 |
+
"grad_norm": 2.3559324741363525,
|
266 |
+
"learning_rate": 4.6126518642647674e-05,
|
267 |
+
"loss": 6.6892,
|
268 |
+
"step": 18500
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 3.97989107666527,
|
272 |
+
"grad_norm": 3.1527633666992188,
|
273 |
+
"learning_rate": 4.602178466694596e-05,
|
274 |
+
"loss": 6.6851,
|
275 |
+
"step": 19000
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 4.084625052366988,
|
279 |
+
"grad_norm": 2.9760189056396484,
|
280 |
+
"learning_rate": 4.591705069124424e-05,
|
281 |
+
"loss": 6.6704,
|
282 |
+
"step": 19500
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 4.189359028068705,
|
286 |
+
"grad_norm": 2.8135318756103516,
|
287 |
+
"learning_rate": 4.5812316715542526e-05,
|
288 |
+
"loss": 6.6621,
|
289 |
+
"step": 20000
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 4.294093003770423,
|
293 |
+
"grad_norm": 3.060316324234009,
|
294 |
+
"learning_rate": 4.570779220779221e-05,
|
295 |
+
"loss": 6.6298,
|
296 |
+
"step": 20500
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 4.39882697947214,
|
300 |
+
"grad_norm": 2.7130279541015625,
|
301 |
+
"learning_rate": 4.560305823209049e-05,
|
302 |
+
"loss": 6.628,
|
303 |
+
"step": 21000
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 4.5035609551738585,
|
307 |
+
"grad_norm": 3.156386613845825,
|
308 |
+
"learning_rate": 4.549853372434017e-05,
|
309 |
+
"loss": 6.6423,
|
310 |
+
"step": 21500
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 4.6082949308755765,
|
314 |
+
"grad_norm": 3.039471387863159,
|
315 |
+
"learning_rate": 4.539379974863846e-05,
|
316 |
+
"loss": 6.6343,
|
317 |
+
"step": 22000
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 4.713028906577294,
|
321 |
+
"grad_norm": 3.976949453353882,
|
322 |
+
"learning_rate": 4.5289065772936745e-05,
|
323 |
+
"loss": 6.606,
|
324 |
+
"step": 22500
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 4.817762882279011,
|
328 |
+
"grad_norm": 3.310382604598999,
|
329 |
+
"learning_rate": 4.5184331797235025e-05,
|
330 |
+
"loss": 6.5956,
|
331 |
+
"step": 23000
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 4.922496857980729,
|
335 |
+
"grad_norm": 3.5924322605133057,
|
336 |
+
"learning_rate": 4.507959782153331e-05,
|
337 |
+
"loss": 6.5965,
|
338 |
+
"step": 23500
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 5.027230833682447,
|
342 |
+
"grad_norm": 2.616468667984009,
|
343 |
+
"learning_rate": 4.497486384583159e-05,
|
344 |
+
"loss": 6.5788,
|
345 |
+
"step": 24000
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 5.131964809384164,
|
349 |
+
"grad_norm": 3.3178062438964844,
|
350 |
+
"learning_rate": 4.487012987012987e-05,
|
351 |
+
"loss": 6.5684,
|
352 |
+
"step": 24500
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 5.236698785085882,
|
356 |
+
"grad_norm": 3.7108089923858643,
|
357 |
+
"learning_rate": 4.476539589442815e-05,
|
358 |
+
"loss": 6.5756,
|
359 |
+
"step": 25000
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 5.341432760787599,
|
363 |
+
"grad_norm": 3.396498918533325,
|
364 |
+
"learning_rate": 4.466087138667784e-05,
|
365 |
+
"loss": 6.5678,
|
366 |
+
"step": 25500
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 5.446166736489317,
|
370 |
+
"grad_norm": 3.7245748043060303,
|
371 |
+
"learning_rate": 4.4556137410976123e-05,
|
372 |
+
"loss": 6.5578,
|
373 |
+
"step": 26000
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 5.5509007121910345,
|
377 |
+
"grad_norm": 3.6525135040283203,
|
378 |
+
"learning_rate": 4.44514034352744e-05,
|
379 |
+
"loss": 6.5385,
|
380 |
+
"step": 26500
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 5.655634687892753,
|
384 |
+
"grad_norm": 3.4302523136138916,
|
385 |
+
"learning_rate": 4.434666945957269e-05,
|
386 |
+
"loss": 6.5143,
|
387 |
+
"step": 27000
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 5.76036866359447,
|
391 |
+
"grad_norm": 3.762871503829956,
|
392 |
+
"learning_rate": 4.4242144951822376e-05,
|
393 |
+
"loss": 6.52,
|
394 |
+
"step": 27500
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 5.865102639296188,
|
398 |
+
"grad_norm": 2.8195388317108154,
|
399 |
+
"learning_rate": 4.4137410976120656e-05,
|
400 |
+
"loss": 6.5213,
|
401 |
+
"step": 28000
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 5.969836614997905,
|
405 |
+
"grad_norm": 3.013187885284424,
|
406 |
+
"learning_rate": 4.4032677000418936e-05,
|
407 |
+
"loss": 6.5052,
|
408 |
+
"step": 28500
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 6.074570590699623,
|
412 |
+
"grad_norm": 2.9772274494171143,
|
413 |
+
"learning_rate": 4.392794302471722e-05,
|
414 |
+
"loss": 6.502,
|
415 |
+
"step": 29000
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 6.17930456640134,
|
419 |
+
"grad_norm": 3.2228713035583496,
|
420 |
+
"learning_rate": 4.38232090490155e-05,
|
421 |
+
"loss": 6.4889,
|
422 |
+
"step": 29500
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 6.284038542103058,
|
426 |
+
"grad_norm": 3.824286937713623,
|
427 |
+
"learning_rate": 4.371847507331379e-05,
|
428 |
+
"loss": 6.4792,
|
429 |
+
"step": 30000
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 6.388772517804776,
|
433 |
+
"grad_norm": 3.100308656692505,
|
434 |
+
"learning_rate": 4.361374109761207e-05,
|
435 |
+
"loss": 6.4816,
|
436 |
+
"step": 30500
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 6.4935064935064934,
|
440 |
+
"grad_norm": 3.4449245929718018,
|
441 |
+
"learning_rate": 4.350900712191035e-05,
|
442 |
+
"loss": 6.4786,
|
443 |
+
"step": 31000
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 6.5982404692082115,
|
447 |
+
"grad_norm": 3.6803085803985596,
|
448 |
+
"learning_rate": 4.3404482614160034e-05,
|
449 |
+
"loss": 6.4778,
|
450 |
+
"step": 31500
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 6.702974444909929,
|
454 |
+
"grad_norm": 3.6413722038269043,
|
455 |
+
"learning_rate": 4.329974863845832e-05,
|
456 |
+
"loss": 6.4782,
|
457 |
+
"step": 32000
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 6.807708420611647,
|
461 |
+
"grad_norm": 3.482905626296997,
|
462 |
+
"learning_rate": 4.31950146627566e-05,
|
463 |
+
"loss": 6.4719,
|
464 |
+
"step": 32500
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 6.912442396313364,
|
468 |
+
"grad_norm": 3.4605376720428467,
|
469 |
+
"learning_rate": 4.3090280687054887e-05,
|
470 |
+
"loss": 6.4458,
|
471 |
+
"step": 33000
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 7.017176372015082,
|
475 |
+
"grad_norm": 3.814375877380371,
|
476 |
+
"learning_rate": 4.2985965647255974e-05,
|
477 |
+
"loss": 6.4539,
|
478 |
+
"step": 33500
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 7.121910347716799,
|
482 |
+
"grad_norm": 4.238844871520996,
|
483 |
+
"learning_rate": 4.288123167155425e-05,
|
484 |
+
"loss": 6.4522,
|
485 |
+
"step": 34000
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 7.226644323418517,
|
489 |
+
"grad_norm": 2.8327670097351074,
|
490 |
+
"learning_rate": 4.277649769585253e-05,
|
491 |
+
"loss": 6.4383,
|
492 |
+
"step": 34500
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 7.331378299120234,
|
496 |
+
"grad_norm": 3.1451475620269775,
|
497 |
+
"learning_rate": 4.267176372015082e-05,
|
498 |
+
"loss": 6.4451,
|
499 |
+
"step": 35000
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 7.436112274821952,
|
503 |
+
"grad_norm": 3.6858575344085693,
|
504 |
+
"learning_rate": 4.2567239212400506e-05,
|
505 |
+
"loss": 6.4313,
|
506 |
+
"step": 35500
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 7.5408462505236695,
|
510 |
+
"grad_norm": 4.258295059204102,
|
511 |
+
"learning_rate": 4.2462505236698786e-05,
|
512 |
+
"loss": 6.4143,
|
513 |
+
"step": 36000
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 7.645580226225388,
|
517 |
+
"grad_norm": 4.3574676513671875,
|
518 |
+
"learning_rate": 4.235777126099707e-05,
|
519 |
+
"loss": 6.4061,
|
520 |
+
"step": 36500
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 7.750314201927106,
|
524 |
+
"grad_norm": 3.8001816272735596,
|
525 |
+
"learning_rate": 4.225303728529535e-05,
|
526 |
+
"loss": 6.3994,
|
527 |
+
"step": 37000
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 7.855048177628823,
|
531 |
+
"grad_norm": 3.487893581390381,
|
532 |
+
"learning_rate": 4.214851277754504e-05,
|
533 |
+
"loss": 6.4192,
|
534 |
+
"step": 37500
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 7.95978215333054,
|
538 |
+
"grad_norm": 3.9729723930358887,
|
539 |
+
"learning_rate": 4.204377880184332e-05,
|
540 |
+
"loss": 6.407,
|
541 |
+
"step": 38000
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 8.064516129032258,
|
545 |
+
"grad_norm": 3.4465062618255615,
|
546 |
+
"learning_rate": 4.1939044826141604e-05,
|
547 |
+
"loss": 6.3867,
|
548 |
+
"step": 38500
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 8.169250104733976,
|
552 |
+
"grad_norm": 3.706404685974121,
|
553 |
+
"learning_rate": 4.1834310850439884e-05,
|
554 |
+
"loss": 6.3877,
|
555 |
+
"step": 39000
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 8.273984080435694,
|
559 |
+
"grad_norm": 3.8204259872436523,
|
560 |
+
"learning_rate": 4.172957687473817e-05,
|
561 |
+
"loss": 6.3921,
|
562 |
+
"step": 39500
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 8.37871805613741,
|
566 |
+
"grad_norm": 3.4868948459625244,
|
567 |
+
"learning_rate": 4.162505236698786e-05,
|
568 |
+
"loss": 6.3729,
|
569 |
+
"step": 40000
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 8.483452031839128,
|
573 |
+
"grad_norm": 3.5007081031799316,
|
574 |
+
"learning_rate": 4.152031839128614e-05,
|
575 |
+
"loss": 6.3886,
|
576 |
+
"step": 40500
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 8.588186007540846,
|
580 |
+
"grad_norm": 2.937894582748413,
|
581 |
+
"learning_rate": 4.1415584415584417e-05,
|
582 |
+
"loss": 6.3814,
|
583 |
+
"step": 41000
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 8.692919983242565,
|
587 |
+
"grad_norm": 3.529237985610962,
|
588 |
+
"learning_rate": 4.1310850439882696e-05,
|
589 |
+
"loss": 6.3722,
|
590 |
+
"step": 41500
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 8.79765395894428,
|
594 |
+
"grad_norm": 3.883575677871704,
|
595 |
+
"learning_rate": 4.120611646418098e-05,
|
596 |
+
"loss": 6.3655,
|
597 |
+
"step": 42000
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 8.902387934645999,
|
601 |
+
"grad_norm": 4.439103603363037,
|
602 |
+
"learning_rate": 4.110159195643067e-05,
|
603 |
+
"loss": 6.3673,
|
604 |
+
"step": 42500
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 9.007121910347717,
|
608 |
+
"grad_norm": 4.103298664093018,
|
609 |
+
"learning_rate": 4.099685798072895e-05,
|
610 |
+
"loss": 6.3659,
|
611 |
+
"step": 43000
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 9.111855886049435,
|
615 |
+
"grad_norm": 3.491204023361206,
|
616 |
+
"learning_rate": 4.0892124005027235e-05,
|
617 |
+
"loss": 6.3744,
|
618 |
+
"step": 43500
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 9.216589861751151,
|
622 |
+
"grad_norm": 3.441976547241211,
|
623 |
+
"learning_rate": 4.0787390029325515e-05,
|
624 |
+
"loss": 6.3573,
|
625 |
+
"step": 44000
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 9.32132383745287,
|
629 |
+
"grad_norm": 3.58134126663208,
|
630 |
+
"learning_rate": 4.0682656053623795e-05,
|
631 |
+
"loss": 6.3407,
|
632 |
+
"step": 44500
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 9.426057813154587,
|
636 |
+
"grad_norm": 3.274592638015747,
|
637 |
+
"learning_rate": 4.057813154587348e-05,
|
638 |
+
"loss": 6.3373,
|
639 |
+
"step": 45000
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 9.530791788856305,
|
643 |
+
"grad_norm": 4.296390533447266,
|
644 |
+
"learning_rate": 4.047339757017177e-05,
|
645 |
+
"loss": 6.3499,
|
646 |
+
"step": 45500
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 9.635525764558023,
|
650 |
+
"grad_norm": 3.5000336170196533,
|
651 |
+
"learning_rate": 4.036866359447005e-05,
|
652 |
+
"loss": 6.3199,
|
653 |
+
"step": 46000
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 9.74025974025974,
|
657 |
+
"grad_norm": 3.4947054386138916,
|
658 |
+
"learning_rate": 4.0263929618768334e-05,
|
659 |
+
"loss": 6.3474,
|
660 |
+
"step": 46500
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 9.844993715961458,
|
664 |
+
"grad_norm": 3.3658857345581055,
|
665 |
+
"learning_rate": 4.0159195643066614e-05,
|
666 |
+
"loss": 6.3296,
|
667 |
+
"step": 47000
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 9.949727691663176,
|
671 |
+
"grad_norm": 2.9811642169952393,
|
672 |
+
"learning_rate": 4.0054671135316294e-05,
|
673 |
+
"loss": 6.345,
|
674 |
+
"step": 47500
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 10.054461667364894,
|
678 |
+
"grad_norm": 4.165875434875488,
|
679 |
+
"learning_rate": 3.994993715961458e-05,
|
680 |
+
"loss": 6.3209,
|
681 |
+
"step": 48000
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 10.15919564306661,
|
685 |
+
"grad_norm": 3.6118202209472656,
|
686 |
+
"learning_rate": 3.9845203183912866e-05,
|
687 |
+
"loss": 6.3175,
|
688 |
+
"step": 48500
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 10.263929618768328,
|
692 |
+
"grad_norm": 3.930669069290161,
|
693 |
+
"learning_rate": 3.9740469208211146e-05,
|
694 |
+
"loss": 6.3231,
|
695 |
+
"step": 49000
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 10.368663594470046,
|
699 |
+
"grad_norm": 3.1688554286956787,
|
700 |
+
"learning_rate": 3.963594470046083e-05,
|
701 |
+
"loss": 6.309,
|
702 |
+
"step": 49500
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 10.473397570171764,
|
706 |
+
"grad_norm": 3.6746394634246826,
|
707 |
+
"learning_rate": 3.953121072475911e-05,
|
708 |
+
"loss": 6.3077,
|
709 |
+
"step": 50000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 10.57813154587348,
|
713 |
+
"grad_norm": 3.5134785175323486,
|
714 |
+
"learning_rate": 3.942647674905739e-05,
|
715 |
+
"loss": 6.3299,
|
716 |
+
"step": 50500
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 10.682865521575199,
|
720 |
+
"grad_norm": 3.2903287410736084,
|
721 |
+
"learning_rate": 3.932174277335568e-05,
|
722 |
+
"loss": 6.3178,
|
723 |
+
"step": 51000
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 10.787599497276917,
|
727 |
+
"grad_norm": 3.5344769954681396,
|
728 |
+
"learning_rate": 3.921700879765396e-05,
|
729 |
+
"loss": 6.3139,
|
730 |
+
"step": 51500
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 10.892333472978635,
|
734 |
+
"grad_norm": 3.5710573196411133,
|
735 |
+
"learning_rate": 3.9112274821952245e-05,
|
736 |
+
"loss": 6.306,
|
737 |
+
"step": 52000
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 10.997067448680351,
|
741 |
+
"grad_norm": 3.4193336963653564,
|
742 |
+
"learning_rate": 3.9007540846250524e-05,
|
743 |
+
"loss": 6.3129,
|
744 |
+
"step": 52500
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 11.101801424382069,
|
748 |
+
"grad_norm": 3.683143377304077,
|
749 |
+
"learning_rate": 3.890301633850021e-05,
|
750 |
+
"loss": 6.3084,
|
751 |
+
"step": 53000
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 11.206535400083787,
|
755 |
+
"grad_norm": 3.214221239089966,
|
756 |
+
"learning_rate": 3.879828236279849e-05,
|
757 |
+
"loss": 6.302,
|
758 |
+
"step": 53500
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 11.311269375785505,
|
762 |
+
"grad_norm": 3.5691747665405273,
|
763 |
+
"learning_rate": 3.869354838709678e-05,
|
764 |
+
"loss": 6.3164,
|
765 |
+
"step": 54000
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 11.416003351487223,
|
769 |
+
"grad_norm": 3.2734036445617676,
|
770 |
+
"learning_rate": 3.858881441139506e-05,
|
771 |
+
"loss": 6.2889,
|
772 |
+
"step": 54500
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 11.52073732718894,
|
776 |
+
"grad_norm": 4.049854278564453,
|
777 |
+
"learning_rate": 3.848428990364474e-05,
|
778 |
+
"loss": 6.2957,
|
779 |
+
"step": 55000
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 11.625471302890658,
|
783 |
+
"grad_norm": 3.837921380996704,
|
784 |
+
"learning_rate": 3.837955592794303e-05,
|
785 |
+
"loss": 6.2806,
|
786 |
+
"step": 55500
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 11.730205278592376,
|
790 |
+
"grad_norm": 3.4606828689575195,
|
791 |
+
"learning_rate": 3.827482195224131e-05,
|
792 |
+
"loss": 6.2896,
|
793 |
+
"step": 56000
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 11.834939254294094,
|
797 |
+
"grad_norm": 4.859198093414307,
|
798 |
+
"learning_rate": 3.8170087976539596e-05,
|
799 |
+
"loss": 6.273,
|
800 |
+
"step": 56500
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 11.93967322999581,
|
804 |
+
"grad_norm": 4.689023494720459,
|
805 |
+
"learning_rate": 3.806535400083787e-05,
|
806 |
+
"loss": 6.2776,
|
807 |
+
"step": 57000
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 12.044407205697528,
|
811 |
+
"grad_norm": 4.234752178192139,
|
812 |
+
"learning_rate": 3.7960829493087555e-05,
|
813 |
+
"loss": 6.282,
|
814 |
+
"step": 57500
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 12.149141181399246,
|
818 |
+
"grad_norm": 3.950773239135742,
|
819 |
+
"learning_rate": 3.785609551738584e-05,
|
820 |
+
"loss": 6.283,
|
821 |
+
"step": 58000
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 12.253875157100964,
|
825 |
+
"grad_norm": 4.1780548095703125,
|
826 |
+
"learning_rate": 3.775136154168412e-05,
|
827 |
+
"loss": 6.2635,
|
828 |
+
"step": 58500
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 12.35860913280268,
|
832 |
+
"grad_norm": 3.2049672603607178,
|
833 |
+
"learning_rate": 3.764662756598241e-05,
|
834 |
+
"loss": 6.2858,
|
835 |
+
"step": 59000
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 12.463343108504398,
|
839 |
+
"grad_norm": 3.863649606704712,
|
840 |
+
"learning_rate": 3.7541893590280694e-05,
|
841 |
+
"loss": 6.2609,
|
842 |
+
"step": 59500
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 12.568077084206116,
|
846 |
+
"grad_norm": 3.881343364715576,
|
847 |
+
"learning_rate": 3.743715961457897e-05,
|
848 |
+
"loss": 6.2695,
|
849 |
+
"step": 60000
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 12.672811059907835,
|
853 |
+
"grad_norm": 3.522132635116577,
|
854 |
+
"learning_rate": 3.7332635106828654e-05,
|
855 |
+
"loss": 6.2523,
|
856 |
+
"step": 60500
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 12.777545035609553,
|
860 |
+
"grad_norm": 4.043595790863037,
|
861 |
+
"learning_rate": 3.722790113112694e-05,
|
862 |
+
"loss": 6.2546,
|
863 |
+
"step": 61000
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 12.882279011311269,
|
867 |
+
"grad_norm": 3.4860141277313232,
|
868 |
+
"learning_rate": 3.712316715542522e-05,
|
869 |
+
"loss": 6.2468,
|
870 |
+
"step": 61500
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 12.987012987012987,
|
874 |
+
"grad_norm": 3.9201574325561523,
|
875 |
+
"learning_rate": 3.7018433179723506e-05,
|
876 |
+
"loss": 6.2615,
|
877 |
+
"step": 62000
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 13.091746962714705,
|
881 |
+
"grad_norm": 3.5582118034362793,
|
882 |
+
"learning_rate": 3.6913699204021786e-05,
|
883 |
+
"loss": 6.2529,
|
884 |
+
"step": 62500
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 13.196480938416423,
|
888 |
+
"grad_norm": 3.1254093647003174,
|
889 |
+
"learning_rate": 3.680917469627147e-05,
|
890 |
+
"loss": 6.2418,
|
891 |
+
"step": 63000
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 13.30121491411814,
|
895 |
+
"grad_norm": 4.058616638183594,
|
896 |
+
"learning_rate": 3.670444072056975e-05,
|
897 |
+
"loss": 6.243,
|
898 |
+
"step": 63500
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 13.405948889819857,
|
902 |
+
"grad_norm": 3.5146963596343994,
|
903 |
+
"learning_rate": 3.659970674486803e-05,
|
904 |
+
"loss": 6.2595,
|
905 |
+
"step": 64000
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 13.510682865521575,
|
909 |
+
"grad_norm": 3.804818630218506,
|
910 |
+
"learning_rate": 3.649497276916632e-05,
|
911 |
+
"loss": 6.2438,
|
912 |
+
"step": 64500
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 13.615416841223293,
|
916 |
+
"grad_norm": 3.591214179992676,
|
917 |
+
"learning_rate": 3.6390238793464605e-05,
|
918 |
+
"loss": 6.2266,
|
919 |
+
"step": 65000
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 13.72015081692501,
|
923 |
+
"grad_norm": 4.973635196685791,
|
924 |
+
"learning_rate": 3.6285504817762885e-05,
|
925 |
+
"loss": 6.2501,
|
926 |
+
"step": 65500
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 13.824884792626728,
|
930 |
+
"grad_norm": 4.189575672149658,
|
931 |
+
"learning_rate": 3.618098031001257e-05,
|
932 |
+
"loss": 6.2341,
|
933 |
+
"step": 66000
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 13.929618768328446,
|
937 |
+
"grad_norm": 4.408186912536621,
|
938 |
+
"learning_rate": 3.607624633431085e-05,
|
939 |
+
"loss": 6.227,
|
940 |
+
"step": 66500
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 14.034352744030164,
|
944 |
+
"grad_norm": 4.066199779510498,
|
945 |
+
"learning_rate": 3.597151235860913e-05,
|
946 |
+
"loss": 6.2316,
|
947 |
+
"step": 67000
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 14.139086719731882,
|
951 |
+
"grad_norm": 3.8263683319091797,
|
952 |
+
"learning_rate": 3.586677838290742e-05,
|
953 |
+
"loss": 6.2322,
|
954 |
+
"step": 67500
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 14.243820695433598,
|
958 |
+
"grad_norm": 4.787020206451416,
|
959 |
+
"learning_rate": 3.57620444072057e-05,
|
960 |
+
"loss": 6.2099,
|
961 |
+
"step": 68000
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 14.348554671135316,
|
965 |
+
"grad_norm": 3.5196545124053955,
|
966 |
+
"learning_rate": 3.565751989945538e-05,
|
967 |
+
"loss": 6.2425,
|
968 |
+
"step": 68500
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 14.453288646837034,
|
972 |
+
"grad_norm": 4.1746439933776855,
|
973 |
+
"learning_rate": 3.555278592375367e-05,
|
974 |
+
"loss": 6.2373,
|
975 |
+
"step": 69000
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 14.558022622538752,
|
979 |
+
"grad_norm": 4.07820463180542,
|
980 |
+
"learning_rate": 3.544805194805195e-05,
|
981 |
+
"loss": 6.2099,
|
982 |
+
"step": 69500
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 14.662756598240469,
|
986 |
+
"grad_norm": 3.400038242340088,
|
987 |
+
"learning_rate": 3.534331797235023e-05,
|
988 |
+
"loss": 6.216,
|
989 |
+
"step": 70000
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 14.767490573942187,
|
993 |
+
"grad_norm": 4.578042030334473,
|
994 |
+
"learning_rate": 3.5238793464599916e-05,
|
995 |
+
"loss": 6.2091,
|
996 |
+
"step": 70500
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 14.872224549643905,
|
1000 |
+
"grad_norm": 3.6254208087921143,
|
1001 |
+
"learning_rate": 3.51340594888982e-05,
|
1002 |
+
"loss": 6.2258,
|
1003 |
+
"step": 71000
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 14.976958525345623,
|
1007 |
+
"grad_norm": 3.496166467666626,
|
1008 |
+
"learning_rate": 3.502932551319648e-05,
|
1009 |
+
"loss": 6.2138,
|
1010 |
+
"step": 71500
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 15.081692501047339,
|
1014 |
+
"grad_norm": 3.5367865562438965,
|
1015 |
+
"learning_rate": 3.492459153749477e-05,
|
1016 |
+
"loss": 6.213,
|
1017 |
+
"step": 72000
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 15.186426476749057,
|
1021 |
+
"grad_norm": 3.4754440784454346,
|
1022 |
+
"learning_rate": 3.481985756179305e-05,
|
1023 |
+
"loss": 6.2153,
|
1024 |
+
"step": 72500
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 15.291160452450775,
|
1028 |
+
"grad_norm": 4.432271957397461,
|
1029 |
+
"learning_rate": 3.4715333054042735e-05,
|
1030 |
+
"loss": 6.2007,
|
1031 |
+
"step": 73000
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 15.395894428152493,
|
1035 |
+
"grad_norm": 3.8427770137786865,
|
1036 |
+
"learning_rate": 3.4610599078341014e-05,
|
1037 |
+
"loss": 6.2071,
|
1038 |
+
"step": 73500
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 15.50062840385421,
|
1042 |
+
"grad_norm": 3.9617857933044434,
|
1043 |
+
"learning_rate": 3.4505865102639294e-05,
|
1044 |
+
"loss": 6.2142,
|
1045 |
+
"step": 74000
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 15.605362379555928,
|
1049 |
+
"grad_norm": 3.769693613052368,
|
1050 |
+
"learning_rate": 3.440113112693758e-05,
|
1051 |
+
"loss": 6.2065,
|
1052 |
+
"step": 74500
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 15.710096355257646,
|
1056 |
+
"grad_norm": 3.825507402420044,
|
1057 |
+
"learning_rate": 3.429639715123587e-05,
|
1058 |
+
"loss": 6.2072,
|
1059 |
+
"step": 75000
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 15.814830330959364,
|
1063 |
+
"grad_norm": 3.982872724533081,
|
1064 |
+
"learning_rate": 3.4191872643485554e-05,
|
1065 |
+
"loss": 6.2003,
|
1066 |
+
"step": 75500
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 15.91956430666108,
|
1070 |
+
"grad_norm": 3.9958648681640625,
|
1071 |
+
"learning_rate": 3.408713866778383e-05,
|
1072 |
+
"loss": 6.1913,
|
1073 |
+
"step": 76000
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 16.0242982823628,
|
1077 |
+
"grad_norm": 3.947957754135132,
|
1078 |
+
"learning_rate": 3.398240469208211e-05,
|
1079 |
+
"loss": 6.2019,
|
1080 |
+
"step": 76500
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 16.129032258064516,
|
1084 |
+
"grad_norm": 3.8135411739349365,
|
1085 |
+
"learning_rate": 3.387767071638039e-05,
|
1086 |
+
"loss": 6.1944,
|
1087 |
+
"step": 77000
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 16.233766233766232,
|
1091 |
+
"grad_norm": 3.940861701965332,
|
1092 |
+
"learning_rate": 3.377293674067868e-05,
|
1093 |
+
"loss": 6.1893,
|
1094 |
+
"step": 77500
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 16.338500209467952,
|
1098 |
+
"grad_norm": 5.24894905090332,
|
1099 |
+
"learning_rate": 3.3668412232928366e-05,
|
1100 |
+
"loss": 6.1984,
|
1101 |
+
"step": 78000
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 16.44323418516967,
|
1105 |
+
"grad_norm": 4.470870494842529,
|
1106 |
+
"learning_rate": 3.3563678257226645e-05,
|
1107 |
+
"loss": 6.1958,
|
1108 |
+
"step": 78500
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 16.547968160871388,
|
1112 |
+
"grad_norm": 3.699892282485962,
|
1113 |
+
"learning_rate": 3.345894428152493e-05,
|
1114 |
+
"loss": 6.1952,
|
1115 |
+
"step": 79000
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 16.652702136573104,
|
1119 |
+
"grad_norm": 4.136711120605469,
|
1120 |
+
"learning_rate": 3.335421030582321e-05,
|
1121 |
+
"loss": 6.1896,
|
1122 |
+
"step": 79500
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 16.75743611227482,
|
1126 |
+
"grad_norm": 4.904257297515869,
|
1127 |
+
"learning_rate": 3.324968579807289e-05,
|
1128 |
+
"loss": 6.1715,
|
1129 |
+
"step": 80000
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 16.86217008797654,
|
1133 |
+
"grad_norm": 4.219280242919922,
|
1134 |
+
"learning_rate": 3.314495182237118e-05,
|
1135 |
+
"loss": 6.1829,
|
1136 |
+
"step": 80500
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 16.966904063678257,
|
1140 |
+
"grad_norm": 4.426414489746094,
|
1141 |
+
"learning_rate": 3.3040217846669464e-05,
|
1142 |
+
"loss": 6.1782,
|
1143 |
+
"step": 81000
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 17.071638039379973,
|
1147 |
+
"grad_norm": 4.792020797729492,
|
1148 |
+
"learning_rate": 3.2935483870967744e-05,
|
1149 |
+
"loss": 6.1675,
|
1150 |
+
"step": 81500
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 17.176372015081693,
|
1154 |
+
"grad_norm": 3.9796903133392334,
|
1155 |
+
"learning_rate": 3.283095936321743e-05,
|
1156 |
+
"loss": 6.179,
|
1157 |
+
"step": 82000
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 17.28110599078341,
|
1161 |
+
"grad_norm": 4.554388046264648,
|
1162 |
+
"learning_rate": 3.272622538751571e-05,
|
1163 |
+
"loss": 6.1756,
|
1164 |
+
"step": 82500
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 17.38583996648513,
|
1168 |
+
"grad_norm": 4.024316787719727,
|
1169 |
+
"learning_rate": 3.262149141181399e-05,
|
1170 |
+
"loss": 6.177,
|
1171 |
+
"step": 83000
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 17.490573942186845,
|
1175 |
+
"grad_norm": 4.059772968292236,
|
1176 |
+
"learning_rate": 3.2516757436112276e-05,
|
1177 |
+
"loss": 6.1818,
|
1178 |
+
"step": 83500
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 17.59530791788856,
|
1182 |
+
"grad_norm": 4.296391487121582,
|
1183 |
+
"learning_rate": 3.241223292836196e-05,
|
1184 |
+
"loss": 6.1866,
|
1185 |
+
"step": 84000
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 17.70004189359028,
|
1189 |
+
"grad_norm": 4.008220672607422,
|
1190 |
+
"learning_rate": 3.230749895266024e-05,
|
1191 |
+
"loss": 6.172,
|
1192 |
+
"step": 84500
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 17.804775869291998,
|
1196 |
+
"grad_norm": 4.639082908630371,
|
1197 |
+
"learning_rate": 3.220276497695853e-05,
|
1198 |
+
"loss": 6.169,
|
1199 |
+
"step": 85000
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 17.909509844993718,
|
1203 |
+
"grad_norm": 4.635848522186279,
|
1204 |
+
"learning_rate": 3.209803100125681e-05,
|
1205 |
+
"loss": 6.1721,
|
1206 |
+
"step": 85500
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 18.014243820695434,
|
1210 |
+
"grad_norm": 4.662270545959473,
|
1211 |
+
"learning_rate": 3.199329702555509e-05,
|
1212 |
+
"loss": 6.1575,
|
1213 |
+
"step": 86000
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 18.11897779639715,
|
1217 |
+
"grad_norm": 4.280701637268066,
|
1218 |
+
"learning_rate": 3.1888772517804775e-05,
|
1219 |
+
"loss": 6.1482,
|
1220 |
+
"step": 86500
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 18.22371177209887,
|
1224 |
+
"grad_norm": 3.8602380752563477,
|
1225 |
+
"learning_rate": 3.178403854210306e-05,
|
1226 |
+
"loss": 6.1565,
|
1227 |
+
"step": 87000
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 18.328445747800586,
|
1231 |
+
"grad_norm": 4.634263515472412,
|
1232 |
+
"learning_rate": 3.167930456640134e-05,
|
1233 |
+
"loss": 6.1587,
|
1234 |
+
"step": 87500
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 18.433179723502302,
|
1238 |
+
"grad_norm": 4.115392208099365,
|
1239 |
+
"learning_rate": 3.157457059069963e-05,
|
1240 |
+
"loss": 6.1436,
|
1241 |
+
"step": 88000
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 18.537913699204022,
|
1245 |
+
"grad_norm": 3.6665916442871094,
|
1246 |
+
"learning_rate": 3.146983661499791e-05,
|
1247 |
+
"loss": 6.1442,
|
1248 |
+
"step": 88500
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 18.64264767490574,
|
1252 |
+
"grad_norm": 4.444345951080322,
|
1253 |
+
"learning_rate": 3.1365312107247594e-05,
|
1254 |
+
"loss": 6.1571,
|
1255 |
+
"step": 89000
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 18.74738165060746,
|
1259 |
+
"grad_norm": 3.792792558670044,
|
1260 |
+
"learning_rate": 3.1260578131545873e-05,
|
1261 |
+
"loss": 6.1535,
|
1262 |
+
"step": 89500
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 18.852115626309175,
|
1266 |
+
"grad_norm": 3.904019832611084,
|
1267 |
+
"learning_rate": 3.115584415584415e-05,
|
1268 |
+
"loss": 6.1501,
|
1269 |
+
"step": 90000
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 18.95684960201089,
|
1273 |
+
"grad_norm": 4.531284332275391,
|
1274 |
+
"learning_rate": 3.105111018014244e-05,
|
1275 |
+
"loss": 6.1592,
|
1276 |
+
"step": 90500
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 19.06158357771261,
|
1280 |
+
"grad_norm": 3.7976317405700684,
|
1281 |
+
"learning_rate": 3.0946376204440726e-05,
|
1282 |
+
"loss": 6.1474,
|
1283 |
+
"step": 91000
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 19.166317553414327,
|
1287 |
+
"grad_norm": 3.8021469116210938,
|
1288 |
+
"learning_rate": 3.084185169669041e-05,
|
1289 |
+
"loss": 6.1408,
|
1290 |
+
"step": 91500
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 19.271051529116047,
|
1294 |
+
"grad_norm": 4.194758892059326,
|
1295 |
+
"learning_rate": 3.073711772098869e-05,
|
1296 |
+
"loss": 6.1476,
|
1297 |
+
"step": 92000
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 19.375785504817763,
|
1301 |
+
"grad_norm": 4.084668159484863,
|
1302 |
+
"learning_rate": 3.063238374528697e-05,
|
1303 |
+
"loss": 6.1443,
|
1304 |
+
"step": 92500
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 19.48051948051948,
|
1308 |
+
"grad_norm": 4.383222579956055,
|
1309 |
+
"learning_rate": 3.052764976958525e-05,
|
1310 |
+
"loss": 6.1422,
|
1311 |
+
"step": 93000
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 19.5852534562212,
|
1315 |
+
"grad_norm": 4.250995635986328,
|
1316 |
+
"learning_rate": 3.042312526183494e-05,
|
1317 |
+
"loss": 6.1375,
|
1318 |
+
"step": 93500
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 19.689987431922916,
|
1322 |
+
"grad_norm": 4.78529691696167,
|
1323 |
+
"learning_rate": 3.0318391286133225e-05,
|
1324 |
+
"loss": 6.1368,
|
1325 |
+
"step": 94000
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 19.794721407624632,
|
1329 |
+
"grad_norm": 3.4997754096984863,
|
1330 |
+
"learning_rate": 3.0213657310431504e-05,
|
1331 |
+
"loss": 6.1432,
|
1332 |
+
"step": 94500
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 19.89945538332635,
|
1336 |
+
"grad_norm": 4.723648548126221,
|
1337 |
+
"learning_rate": 3.0108923334729787e-05,
|
1338 |
+
"loss": 6.1263,
|
1339 |
+
"step": 95000
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 20.004189359028068,
|
1343 |
+
"grad_norm": 3.930859088897705,
|
1344 |
+
"learning_rate": 3.0004398826979474e-05,
|
1345 |
+
"loss": 6.1387,
|
1346 |
+
"step": 95500
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 20.108923334729788,
|
1350 |
+
"grad_norm": 4.286599159240723,
|
1351 |
+
"learning_rate": 2.9899664851277754e-05,
|
1352 |
+
"loss": 6.1187,
|
1353 |
+
"step": 96000
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 20.213657310431504,
|
1357 |
+
"grad_norm": 3.8475680351257324,
|
1358 |
+
"learning_rate": 2.9794930875576037e-05,
|
1359 |
+
"loss": 6.1312,
|
1360 |
+
"step": 96500
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 20.31839128613322,
|
1364 |
+
"grad_norm": 4.844906806945801,
|
1365 |
+
"learning_rate": 2.9690196899874323e-05,
|
1366 |
+
"loss": 6.1457,
|
1367 |
+
"step": 97000
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 20.42312526183494,
|
1371 |
+
"grad_norm": 4.691315174102783,
|
1372 |
+
"learning_rate": 2.958567239212401e-05,
|
1373 |
+
"loss": 6.1351,
|
1374 |
+
"step": 97500
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 20.527859237536656,
|
1378 |
+
"grad_norm": 6.15250825881958,
|
1379 |
+
"learning_rate": 2.9480938416422286e-05,
|
1380 |
+
"loss": 6.1275,
|
1381 |
+
"step": 98000
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 20.632593213238373,
|
1385 |
+
"grad_norm": 3.8872599601745605,
|
1386 |
+
"learning_rate": 2.9376204440720573e-05,
|
1387 |
+
"loss": 6.1275,
|
1388 |
+
"step": 98500
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 20.737327188940093,
|
1392 |
+
"grad_norm": 4.541051864624023,
|
1393 |
+
"learning_rate": 2.9271470465018852e-05,
|
1394 |
+
"loss": 6.1472,
|
1395 |
+
"step": 99000
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 20.84206116464181,
|
1399 |
+
"grad_norm": 4.556408405303955,
|
1400 |
+
"learning_rate": 2.9166736489317135e-05,
|
1401 |
+
"loss": 6.1369,
|
1402 |
+
"step": 99500
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 20.94679514034353,
|
1406 |
+
"grad_norm": 4.5567498207092285,
|
1407 |
+
"learning_rate": 2.9062211981566822e-05,
|
1408 |
+
"loss": 6.1148,
|
1409 |
+
"step": 100000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 21.051529116045245,
|
1413 |
+
"grad_norm": 4.647518634796143,
|
1414 |
+
"learning_rate": 2.8957478005865102e-05,
|
1415 |
+
"loss": 6.1281,
|
1416 |
+
"step": 100500
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 21.15626309174696,
|
1420 |
+
"grad_norm": 4.372421741485596,
|
1421 |
+
"learning_rate": 2.8852744030163388e-05,
|
1422 |
+
"loss": 6.1226,
|
1423 |
+
"step": 101000
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 21.26099706744868,
|
1427 |
+
"grad_norm": 4.270533084869385,
|
1428 |
+
"learning_rate": 2.874801005446167e-05,
|
1429 |
+
"loss": 6.123,
|
1430 |
+
"step": 101500
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 21.365731043150397,
|
1434 |
+
"grad_norm": 3.5596370697021484,
|
1435 |
+
"learning_rate": 2.8643485546711358e-05,
|
1436 |
+
"loss": 6.1402,
|
1437 |
+
"step": 102000
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 21.470465018852117,
|
1441 |
+
"grad_norm": 5.230384826660156,
|
1442 |
+
"learning_rate": 2.8538751571009638e-05,
|
1443 |
+
"loss": 6.1199,
|
1444 |
+
"step": 102500
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 21.575198994553833,
|
1448 |
+
"grad_norm": 3.9881417751312256,
|
1449 |
+
"learning_rate": 2.843401759530792e-05,
|
1450 |
+
"loss": 6.1244,
|
1451 |
+
"step": 103000
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 21.67993297025555,
|
1455 |
+
"grad_norm": 4.617568016052246,
|
1456 |
+
"learning_rate": 2.83292836196062e-05,
|
1457 |
+
"loss": 6.1154,
|
1458 |
+
"step": 103500
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 21.78466694595727,
|
1462 |
+
"grad_norm": 4.641009330749512,
|
1463 |
+
"learning_rate": 2.8224759111855887e-05,
|
1464 |
+
"loss": 6.113,
|
1465 |
+
"step": 104000
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 21.889400921658986,
|
1469 |
+
"grad_norm": 4.005772113800049,
|
1470 |
+
"learning_rate": 2.812002513615417e-05,
|
1471 |
+
"loss": 6.1163,
|
1472 |
+
"step": 104500
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 21.994134897360702,
|
1476 |
+
"grad_norm": 4.2611799240112305,
|
1477 |
+
"learning_rate": 2.801529116045245e-05,
|
1478 |
+
"loss": 6.1023,
|
1479 |
+
"step": 105000
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 22.098868873062422,
|
1483 |
+
"grad_norm": 4.568357467651367,
|
1484 |
+
"learning_rate": 2.7910557184750736e-05,
|
1485 |
+
"loss": 6.1169,
|
1486 |
+
"step": 105500
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 22.203602848764138,
|
1490 |
+
"grad_norm": 4.323103427886963,
|
1491 |
+
"learning_rate": 2.780603267700042e-05,
|
1492 |
+
"loss": 6.1226,
|
1493 |
+
"step": 106000
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 22.308336824465858,
|
1497 |
+
"grad_norm": 4.507444381713867,
|
1498 |
+
"learning_rate": 2.77012987012987e-05,
|
1499 |
+
"loss": 6.1052,
|
1500 |
+
"step": 106500
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 22.413070800167574,
|
1504 |
+
"grad_norm": 4.301244735717773,
|
1505 |
+
"learning_rate": 2.7596564725596985e-05,
|
1506 |
+
"loss": 6.0944,
|
1507 |
+
"step": 107000
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 22.51780477586929,
|
1511 |
+
"grad_norm": 4.984853267669678,
|
1512 |
+
"learning_rate": 2.749183074989527e-05,
|
1513 |
+
"loss": 6.1125,
|
1514 |
+
"step": 107500
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 22.62253875157101,
|
1518 |
+
"grad_norm": 4.682931423187256,
|
1519 |
+
"learning_rate": 2.7387096774193548e-05,
|
1520 |
+
"loss": 6.1158,
|
1521 |
+
"step": 108000
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 22.727272727272727,
|
1525 |
+
"grad_norm": 4.494015693664551,
|
1526 |
+
"learning_rate": 2.7282572266443235e-05,
|
1527 |
+
"loss": 6.1035,
|
1528 |
+
"step": 108500
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 22.832006702974446,
|
1532 |
+
"grad_norm": 3.880779981613159,
|
1533 |
+
"learning_rate": 2.717783829074152e-05,
|
1534 |
+
"loss": 6.1084,
|
1535 |
+
"step": 109000
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 22.936740678676163,
|
1539 |
+
"grad_norm": 4.154653072357178,
|
1540 |
+
"learning_rate": 2.7073104315039798e-05,
|
1541 |
+
"loss": 6.0945,
|
1542 |
+
"step": 109500
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 23.04147465437788,
|
1546 |
+
"grad_norm": 5.443271160125732,
|
1547 |
+
"learning_rate": 2.6968370339338084e-05,
|
1548 |
+
"loss": 6.1016,
|
1549 |
+
"step": 110000
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 23.1462086300796,
|
1553 |
+
"grad_norm": 4.298133373260498,
|
1554 |
+
"learning_rate": 2.686384583158777e-05,
|
1555 |
+
"loss": 6.103,
|
1556 |
+
"step": 110500
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 23.250942605781315,
|
1560 |
+
"grad_norm": 4.379884243011475,
|
1561 |
+
"learning_rate": 2.675911185588605e-05,
|
1562 |
+
"loss": 6.0814,
|
1563 |
+
"step": 111000
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 23.35567658148303,
|
1567 |
+
"grad_norm": 6.175398349761963,
|
1568 |
+
"learning_rate": 2.6654377880184333e-05,
|
1569 |
+
"loss": 6.1088,
|
1570 |
+
"step": 111500
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 23.46041055718475,
|
1574 |
+
"grad_norm": 4.121715068817139,
|
1575 |
+
"learning_rate": 2.6549643904482613e-05,
|
1576 |
+
"loss": 6.0966,
|
1577 |
+
"step": 112000
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 23.565144532886467,
|
1581 |
+
"grad_norm": 5.040287494659424,
|
1582 |
+
"learning_rate": 2.6444909928780896e-05,
|
1583 |
+
"loss": 6.098,
|
1584 |
+
"step": 112500
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 23.669878508588187,
|
1588 |
+
"grad_norm": 4.766879081726074,
|
1589 |
+
"learning_rate": 2.6340175953079182e-05,
|
1590 |
+
"loss": 6.0957,
|
1591 |
+
"step": 113000
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 23.774612484289904,
|
1595 |
+
"grad_norm": 5.87930965423584,
|
1596 |
+
"learning_rate": 2.623565144532887e-05,
|
1597 |
+
"loss": 6.1089,
|
1598 |
+
"step": 113500
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 23.87934645999162,
|
1602 |
+
"grad_norm": 5.318653583526611,
|
1603 |
+
"learning_rate": 2.613091746962715e-05,
|
1604 |
+
"loss": 6.0761,
|
1605 |
+
"step": 114000
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 23.98408043569334,
|
1609 |
+
"grad_norm": 4.465319633483887,
|
1610 |
+
"learning_rate": 2.6026183493925432e-05,
|
1611 |
+
"loss": 6.0826,
|
1612 |
+
"step": 114500
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 24.088814411395056,
|
1616 |
+
"grad_norm": 4.640571594238281,
|
1617 |
+
"learning_rate": 2.592144951822371e-05,
|
1618 |
+
"loss": 6.0805,
|
1619 |
+
"step": 115000
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 24.193548387096776,
|
1623 |
+
"grad_norm": 4.252554416656494,
|
1624 |
+
"learning_rate": 2.5816925010473398e-05,
|
1625 |
+
"loss": 6.0701,
|
1626 |
+
"step": 115500
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 24.298282362798492,
|
1630 |
+
"grad_norm": 4.704644203186035,
|
1631 |
+
"learning_rate": 2.571219103477168e-05,
|
1632 |
+
"loss": 6.0936,
|
1633 |
+
"step": 116000
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 24.40301633850021,
|
1637 |
+
"grad_norm": 4.601324558258057,
|
1638 |
+
"learning_rate": 2.560745705906996e-05,
|
1639 |
+
"loss": 6.0758,
|
1640 |
+
"step": 116500
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 24.507750314201928,
|
1644 |
+
"grad_norm": 4.380444526672363,
|
1645 |
+
"learning_rate": 2.5502723083368247e-05,
|
1646 |
+
"loss": 6.0871,
|
1647 |
+
"step": 117000
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 24.612484289903644,
|
1651 |
+
"grad_norm": 4.119806289672852,
|
1652 |
+
"learning_rate": 2.5397989107666527e-05,
|
1653 |
+
"loss": 6.102,
|
1654 |
+
"step": 117500
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 24.71721826560536,
|
1658 |
+
"grad_norm": 3.9712698459625244,
|
1659 |
+
"learning_rate": 2.5293464599916217e-05,
|
1660 |
+
"loss": 6.0999,
|
1661 |
+
"step": 118000
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 24.82195224130708,
|
1665 |
+
"grad_norm": 5.146612167358398,
|
1666 |
+
"learning_rate": 2.5188730624214497e-05,
|
1667 |
+
"loss": 6.0947,
|
1668 |
+
"step": 118500
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 24.926686217008797,
|
1672 |
+
"grad_norm": 4.406741142272949,
|
1673 |
+
"learning_rate": 2.508399664851278e-05,
|
1674 |
+
"loss": 6.0829,
|
1675 |
+
"step": 119000
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 25.031420192710517,
|
1679 |
+
"grad_norm": 5.4739766120910645,
|
1680 |
+
"learning_rate": 2.497926267281106e-05,
|
1681 |
+
"loss": 6.0598,
|
1682 |
+
"step": 119500
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 25.136154168412233,
|
1686 |
+
"grad_norm": 4.6231794357299805,
|
1687 |
+
"learning_rate": 2.4874738165060746e-05,
|
1688 |
+
"loss": 6.072,
|
1689 |
+
"step": 120000
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 25.24088814411395,
|
1693 |
+
"grad_norm": 4.47750186920166,
|
1694 |
+
"learning_rate": 2.477000418935903e-05,
|
1695 |
+
"loss": 6.0664,
|
1696 |
+
"step": 120500
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 25.34562211981567,
|
1700 |
+
"grad_norm": 5.023014068603516,
|
1701 |
+
"learning_rate": 2.4665270213657312e-05,
|
1702 |
+
"loss": 6.0894,
|
1703 |
+
"step": 121000
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 25.450356095517385,
|
1707 |
+
"grad_norm": 5.687644004821777,
|
1708 |
+
"learning_rate": 2.4560536237955595e-05,
|
1709 |
+
"loss": 6.0936,
|
1710 |
+
"step": 121500
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 25.555090071219105,
|
1714 |
+
"grad_norm": 4.534958362579346,
|
1715 |
+
"learning_rate": 2.4455802262253878e-05,
|
1716 |
+
"loss": 6.0794,
|
1717 |
+
"step": 122000
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 25.65982404692082,
|
1721 |
+
"grad_norm": 5.563751697540283,
|
1722 |
+
"learning_rate": 2.435127775450356e-05,
|
1723 |
+
"loss": 6.081,
|
1724 |
+
"step": 122500
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 25.764558022622538,
|
1728 |
+
"grad_norm": 4.613626956939697,
|
1729 |
+
"learning_rate": 2.4246543778801845e-05,
|
1730 |
+
"loss": 6.0982,
|
1731 |
+
"step": 123000
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 25.869291998324258,
|
1735 |
+
"grad_norm": 4.645818710327148,
|
1736 |
+
"learning_rate": 2.4141809803100128e-05,
|
1737 |
+
"loss": 6.0665,
|
1738 |
+
"step": 123500
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 25.974025974025974,
|
1742 |
+
"grad_norm": 4.9156928062438965,
|
1743 |
+
"learning_rate": 2.4037075827398407e-05,
|
1744 |
+
"loss": 6.0674,
|
1745 |
+
"step": 124000
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 26.07875994972769,
|
1749 |
+
"grad_norm": 4.7342305183410645,
|
1750 |
+
"learning_rate": 2.3932551319648094e-05,
|
1751 |
+
"loss": 6.0741,
|
1752 |
+
"step": 124500
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 26.18349392542941,
|
1756 |
+
"grad_norm": 4.607081413269043,
|
1757 |
+
"learning_rate": 2.3827817343946377e-05,
|
1758 |
+
"loss": 6.0626,
|
1759 |
+
"step": 125000
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 26.288227901131126,
|
1763 |
+
"grad_norm": 4.820442199707031,
|
1764 |
+
"learning_rate": 2.372308336824466e-05,
|
1765 |
+
"loss": 6.0936,
|
1766 |
+
"step": 125500
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 26.392961876832846,
|
1770 |
+
"grad_norm": 4.549975395202637,
|
1771 |
+
"learning_rate": 2.3618349392542943e-05,
|
1772 |
+
"loss": 6.0804,
|
1773 |
+
"step": 126000
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 26.497695852534562,
|
1777 |
+
"grad_norm": 4.722150802612305,
|
1778 |
+
"learning_rate": 2.351382488479263e-05,
|
1779 |
+
"loss": 6.0555,
|
1780 |
+
"step": 126500
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 26.60242982823628,
|
1784 |
+
"grad_norm": 4.2948408126831055,
|
1785 |
+
"learning_rate": 2.340909090909091e-05,
|
1786 |
+
"loss": 6.0626,
|
1787 |
+
"step": 127000
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 26.707163803938,
|
1791 |
+
"grad_norm": 4.246878623962402,
|
1792 |
+
"learning_rate": 2.3304356933389193e-05,
|
1793 |
+
"loss": 6.0577,
|
1794 |
+
"step": 127500
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 26.811897779639715,
|
1798 |
+
"grad_norm": 4.165809154510498,
|
1799 |
+
"learning_rate": 2.3199622957687476e-05,
|
1800 |
+
"loss": 6.0608,
|
1801 |
+
"step": 128000
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 26.916631755341434,
|
1805 |
+
"grad_norm": 4.3159894943237305,
|
1806 |
+
"learning_rate": 2.309488898198576e-05,
|
1807 |
+
"loss": 6.0806,
|
1808 |
+
"step": 128500
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 27.02136573104315,
|
1812 |
+
"grad_norm": 4.15300989151001,
|
1813 |
+
"learning_rate": 2.2990364474235442e-05,
|
1814 |
+
"loss": 6.0654,
|
1815 |
+
"step": 129000
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 27.126099706744867,
|
1819 |
+
"grad_norm": 4.730154991149902,
|
1820 |
+
"learning_rate": 2.2885630498533725e-05,
|
1821 |
+
"loss": 6.0567,
|
1822 |
+
"step": 129500
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 27.230833682446587,
|
1826 |
+
"grad_norm": 4.300974369049072,
|
1827 |
+
"learning_rate": 2.2780896522832008e-05,
|
1828 |
+
"loss": 6.0449,
|
1829 |
+
"step": 130000
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 27.335567658148303,
|
1833 |
+
"grad_norm": 4.148283958435059,
|
1834 |
+
"learning_rate": 2.2676162547130288e-05,
|
1835 |
+
"loss": 6.0491,
|
1836 |
+
"step": 130500
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 27.44030163385002,
|
1840 |
+
"grad_norm": 4.9924421310424805,
|
1841 |
+
"learning_rate": 2.2571638039379974e-05,
|
1842 |
+
"loss": 6.0491,
|
1843 |
+
"step": 131000
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 27.54503560955174,
|
1847 |
+
"grad_norm": 5.713706016540527,
|
1848 |
+
"learning_rate": 2.246690406367826e-05,
|
1849 |
+
"loss": 6.0396,
|
1850 |
+
"step": 131500
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 27.649769585253456,
|
1854 |
+
"grad_norm": 5.007369518280029,
|
1855 |
+
"learning_rate": 2.236217008797654e-05,
|
1856 |
+
"loss": 6.0378,
|
1857 |
+
"step": 132000
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 27.754503560955175,
|
1861 |
+
"grad_norm": 4.500640392303467,
|
1862 |
+
"learning_rate": 2.2257436112274823e-05,
|
1863 |
+
"loss": 6.0313,
|
1864 |
+
"step": 132500
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 27.85923753665689,
|
1868 |
+
"grad_norm": 4.709275722503662,
|
1869 |
+
"learning_rate": 2.2152702136573107e-05,
|
1870 |
+
"loss": 6.0278,
|
1871 |
+
"step": 133000
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 27.963971512358608,
|
1875 |
+
"grad_norm": 4.891386032104492,
|
1876 |
+
"learning_rate": 2.204817762882279e-05,
|
1877 |
+
"loss": 6.0267,
|
1878 |
+
"step": 133500
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 28.068705488060328,
|
1882 |
+
"grad_norm": 4.82666540145874,
|
1883 |
+
"learning_rate": 2.1943443653121073e-05,
|
1884 |
+
"loss": 5.986,
|
1885 |
+
"step": 134000
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 28.173439463762044,
|
1889 |
+
"grad_norm": 4.489607810974121,
|
1890 |
+
"learning_rate": 2.1838709677419356e-05,
|
1891 |
+
"loss": 6.0209,
|
1892 |
+
"step": 134500
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 28.278173439463764,
|
1896 |
+
"grad_norm": 4.719301700592041,
|
1897 |
+
"learning_rate": 2.173397570171764e-05,
|
1898 |
+
"loss": 5.9998,
|
1899 |
+
"step": 135000
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 28.38290741516548,
|
1903 |
+
"grad_norm": 5.639565467834473,
|
1904 |
+
"learning_rate": 2.1629451193967322e-05,
|
1905 |
+
"loss": 5.9881,
|
1906 |
+
"step": 135500
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 28.487641390867196,
|
1910 |
+
"grad_norm": 4.745512008666992,
|
1911 |
+
"learning_rate": 2.1524717218265605e-05,
|
1912 |
+
"loss": 6.0002,
|
1913 |
+
"step": 136000
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 28.592375366568916,
|
1917 |
+
"grad_norm": 5.661725997924805,
|
1918 |
+
"learning_rate": 2.141998324256389e-05,
|
1919 |
+
"loss": 5.9967,
|
1920 |
+
"step": 136500
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 28.697109342270632,
|
1924 |
+
"grad_norm": 5.3391194343566895,
|
1925 |
+
"learning_rate": 2.131524926686217e-05,
|
1926 |
+
"loss": 6.0024,
|
1927 |
+
"step": 137000
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 28.80184331797235,
|
1931 |
+
"grad_norm": 5.1614089012146,
|
1932 |
+
"learning_rate": 2.1210724759111858e-05,
|
1933 |
+
"loss": 5.9992,
|
1934 |
+
"step": 137500
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 28.90657729367407,
|
1938 |
+
"grad_norm": 5.429248332977295,
|
1939 |
+
"learning_rate": 2.110599078341014e-05,
|
1940 |
+
"loss": 5.9929,
|
1941 |
+
"step": 138000
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 29.011311269375785,
|
1945 |
+
"grad_norm": 5.1270012855529785,
|
1946 |
+
"learning_rate": 2.100125680770842e-05,
|
1947 |
+
"loss": 5.9883,
|
1948 |
+
"step": 138500
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 29.116045245077505,
|
1952 |
+
"grad_norm": 5.027891159057617,
|
1953 |
+
"learning_rate": 2.0896522832006704e-05,
|
1954 |
+
"loss": 5.963,
|
1955 |
+
"step": 139000
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 29.22077922077922,
|
1959 |
+
"grad_norm": 5.712099552154541,
|
1960 |
+
"learning_rate": 2.079199832425639e-05,
|
1961 |
+
"loss": 5.9811,
|
1962 |
+
"step": 139500
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 29.325513196480937,
|
1966 |
+
"grad_norm": 4.954220294952393,
|
1967 |
+
"learning_rate": 2.0687264348554674e-05,
|
1968 |
+
"loss": 5.9808,
|
1969 |
+
"step": 140000
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 29.430247172182657,
|
1973 |
+
"grad_norm": 5.713419437408447,
|
1974 |
+
"learning_rate": 2.0582530372852953e-05,
|
1975 |
+
"loss": 5.9887,
|
1976 |
+
"step": 140500
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 29.534981147884373,
|
1980 |
+
"grad_norm": 4.683711528778076,
|
1981 |
+
"learning_rate": 2.0477796397151236e-05,
|
1982 |
+
"loss": 5.9612,
|
1983 |
+
"step": 141000
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 29.63971512358609,
|
1987 |
+
"grad_norm": 5.164538383483887,
|
1988 |
+
"learning_rate": 2.0373271889400923e-05,
|
1989 |
+
"loss": 5.993,
|
1990 |
+
"step": 141500
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 29.74444909928781,
|
1994 |
+
"grad_norm": 5.386078357696533,
|
1995 |
+
"learning_rate": 2.0268537913699203e-05,
|
1996 |
+
"loss": 5.9735,
|
1997 |
+
"step": 142000
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 29.849183074989526,
|
2001 |
+
"grad_norm": 4.4406418800354,
|
2002 |
+
"learning_rate": 2.016380393799749e-05,
|
2003 |
+
"loss": 5.9672,
|
2004 |
+
"step": 142500
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 29.953917050691246,
|
2008 |
+
"grad_norm": 5.029815673828125,
|
2009 |
+
"learning_rate": 2.0059069962295772e-05,
|
2010 |
+
"loss": 5.961,
|
2011 |
+
"step": 143000
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 30.058651026392962,
|
2015 |
+
"grad_norm": 4.666591167449951,
|
2016 |
+
"learning_rate": 1.9954335986594052e-05,
|
2017 |
+
"loss": 5.9505,
|
2018 |
+
"step": 143500
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 30.163385002094678,
|
2022 |
+
"grad_norm": 6.975547790527344,
|
2023 |
+
"learning_rate": 1.9849602010892335e-05,
|
2024 |
+
"loss": 5.956,
|
2025 |
+
"step": 144000
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 30.268118977796398,
|
2029 |
+
"grad_norm": 4.687684535980225,
|
2030 |
+
"learning_rate": 1.974507750314202e-05,
|
2031 |
+
"loss": 5.9475,
|
2032 |
+
"step": 144500
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 30.372852953498114,
|
2036 |
+
"grad_norm": 5.594231605529785,
|
2037 |
+
"learning_rate": 1.96403435274403e-05,
|
2038 |
+
"loss": 5.9496,
|
2039 |
+
"step": 145000
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 30.477586929199834,
|
2043 |
+
"grad_norm": 4.879722595214844,
|
2044 |
+
"learning_rate": 1.9535609551738584e-05,
|
2045 |
+
"loss": 5.9577,
|
2046 |
+
"step": 145500
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 30.58232090490155,
|
2050 |
+
"grad_norm": 5.470447540283203,
|
2051 |
+
"learning_rate": 1.9430875576036867e-05,
|
2052 |
+
"loss": 5.9672,
|
2053 |
+
"step": 146000
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 30.687054880603267,
|
2057 |
+
"grad_norm": 5.818385124206543,
|
2058 |
+
"learning_rate": 1.932614160033515e-05,
|
2059 |
+
"loss": 5.9501,
|
2060 |
+
"step": 146500
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 30.791788856304986,
|
2064 |
+
"grad_norm": 5.907487392425537,
|
2065 |
+
"learning_rate": 1.9221617092584834e-05,
|
2066 |
+
"loss": 5.9458,
|
2067 |
+
"step": 147000
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 30.896522832006703,
|
2071 |
+
"grad_norm": 4.739224433898926,
|
2072 |
+
"learning_rate": 1.911688311688312e-05,
|
2073 |
+
"loss": 5.935,
|
2074 |
+
"step": 147500
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 31.00125680770842,
|
2078 |
+
"grad_norm": 4.57131814956665,
|
2079 |
+
"learning_rate": 1.90121491411814e-05,
|
2080 |
+
"loss": 5.945,
|
2081 |
+
"step": 148000
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 31.10599078341014,
|
2085 |
+
"grad_norm": 5.128586769104004,
|
2086 |
+
"learning_rate": 1.8907415165479683e-05,
|
2087 |
+
"loss": 5.9494,
|
2088 |
+
"step": 148500
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 31.210724759111855,
|
2092 |
+
"grad_norm": 4.871676921844482,
|
2093 |
+
"learning_rate": 1.880289065772937e-05,
|
2094 |
+
"loss": 5.9415,
|
2095 |
+
"step": 149000
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 31.315458734813575,
|
2099 |
+
"grad_norm": 5.380068778991699,
|
2100 |
+
"learning_rate": 1.8698156682027652e-05,
|
2101 |
+
"loss": 5.939,
|
2102 |
+
"step": 149500
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 31.42019271051529,
|
2106 |
+
"grad_norm": 5.430812835693359,
|
2107 |
+
"learning_rate": 1.8593422706325932e-05,
|
2108 |
+
"loss": 5.9276,
|
2109 |
+
"step": 150000
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 31.524926686217007,
|
2113 |
+
"grad_norm": 4.7710442543029785,
|
2114 |
+
"learning_rate": 1.8488688730624215e-05,
|
2115 |
+
"loss": 5.9413,
|
2116 |
+
"step": 150500
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 31.629660661918727,
|
2120 |
+
"grad_norm": 5.183919906616211,
|
2121 |
+
"learning_rate": 1.8383954754922498e-05,
|
2122 |
+
"loss": 5.9257,
|
2123 |
+
"step": 151000
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 31.734394637620444,
|
2127 |
+
"grad_norm": 4.851598739624023,
|
2128 |
+
"learning_rate": 1.827943024717218e-05,
|
2129 |
+
"loss": 5.9251,
|
2130 |
+
"step": 151500
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 31.839128613322163,
|
2134 |
+
"grad_norm": 4.835882663726807,
|
2135 |
+
"learning_rate": 1.8174696271470464e-05,
|
2136 |
+
"loss": 5.92,
|
2137 |
+
"step": 152000
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 31.94386258902388,
|
2141 |
+
"grad_norm": 5.428823947906494,
|
2142 |
+
"learning_rate": 1.8069962295768748e-05,
|
2143 |
+
"loss": 5.9146,
|
2144 |
+
"step": 152500
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 32.0485965647256,
|
2148 |
+
"grad_norm": 6.11329984664917,
|
2149 |
+
"learning_rate": 1.796522832006703e-05,
|
2150 |
+
"loss": 5.9179,
|
2151 |
+
"step": 153000
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 32.15333054042731,
|
2155 |
+
"grad_norm": 4.836859226226807,
|
2156 |
+
"learning_rate": 1.7860703812316717e-05,
|
2157 |
+
"loss": 5.9189,
|
2158 |
+
"step": 153500
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 32.25806451612903,
|
2162 |
+
"grad_norm": 4.598475456237793,
|
2163 |
+
"learning_rate": 1.7755969836615e-05,
|
2164 |
+
"loss": 5.9207,
|
2165 |
+
"step": 154000
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 32.36279849183075,
|
2169 |
+
"grad_norm": 4.638394832611084,
|
2170 |
+
"learning_rate": 1.765123586091328e-05,
|
2171 |
+
"loss": 5.915,
|
2172 |
+
"step": 154500
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 32.467532467532465,
|
2176 |
+
"grad_norm": 5.637279987335205,
|
2177 |
+
"learning_rate": 1.7546501885211563e-05,
|
2178 |
+
"loss": 5.9228,
|
2179 |
+
"step": 155000
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 32.572266443234184,
|
2183 |
+
"grad_norm": 4.516068458557129,
|
2184 |
+
"learning_rate": 1.7441767909509846e-05,
|
2185 |
+
"loss": 5.9322,
|
2186 |
+
"step": 155500
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 32.677000418935904,
|
2190 |
+
"grad_norm": 4.652084827423096,
|
2191 |
+
"learning_rate": 1.7337243401759533e-05,
|
2192 |
+
"loss": 5.9395,
|
2193 |
+
"step": 156000
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 32.78173439463762,
|
2197 |
+
"grad_norm": 5.667607307434082,
|
2198 |
+
"learning_rate": 1.7232509426057812e-05,
|
2199 |
+
"loss": 5.9172,
|
2200 |
+
"step": 156500
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 32.88646837033934,
|
2204 |
+
"grad_norm": 4.980391025543213,
|
2205 |
+
"learning_rate": 1.7127775450356095e-05,
|
2206 |
+
"loss": 5.934,
|
2207 |
+
"step": 157000
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 32.99120234604106,
|
2211 |
+
"grad_norm": 4.39646053314209,
|
2212 |
+
"learning_rate": 1.702304147465438e-05,
|
2213 |
+
"loss": 5.914,
|
2214 |
+
"step": 157500
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 33.095936321742776,
|
2218 |
+
"grad_norm": 5.263533115386963,
|
2219 |
+
"learning_rate": 1.6918516966904062e-05,
|
2220 |
+
"loss": 5.9094,
|
2221 |
+
"step": 158000
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 33.20067029744449,
|
2225 |
+
"grad_norm": 4.661126136779785,
|
2226 |
+
"learning_rate": 1.6813782991202348e-05,
|
2227 |
+
"loss": 5.9189,
|
2228 |
+
"step": 158500
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 33.30540427314621,
|
2232 |
+
"grad_norm": 4.935306549072266,
|
2233 |
+
"learning_rate": 1.670904901550063e-05,
|
2234 |
+
"loss": 5.8944,
|
2235 |
+
"step": 159000
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 33.41013824884793,
|
2239 |
+
"grad_norm": 5.82065486907959,
|
2240 |
+
"learning_rate": 1.660431503979891e-05,
|
2241 |
+
"loss": 5.892,
|
2242 |
+
"step": 159500
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 33.51487222454964,
|
2246 |
+
"grad_norm": 4.6220927238464355,
|
2247 |
+
"learning_rate": 1.6499581064097194e-05,
|
2248 |
+
"loss": 5.9022,
|
2249 |
+
"step": 160000
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 33.61960620025136,
|
2253 |
+
"grad_norm": 5.109046936035156,
|
2254 |
+
"learning_rate": 1.639505655634688e-05,
|
2255 |
+
"loss": 5.8933,
|
2256 |
+
"step": 160500
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 33.72434017595308,
|
2260 |
+
"grad_norm": 5.230437278747559,
|
2261 |
+
"learning_rate": 1.6290322580645164e-05,
|
2262 |
+
"loss": 5.9107,
|
2263 |
+
"step": 161000
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 33.829074151654794,
|
2267 |
+
"grad_norm": 6.466080188751221,
|
2268 |
+
"learning_rate": 1.6185588604943443e-05,
|
2269 |
+
"loss": 5.9077,
|
2270 |
+
"step": 161500
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 33.933808127356514,
|
2274 |
+
"grad_norm": 4.655428409576416,
|
2275 |
+
"learning_rate": 1.6080854629241726e-05,
|
2276 |
+
"loss": 5.9186,
|
2277 |
+
"step": 162000
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 34.038542103058234,
|
2281 |
+
"grad_norm": 6.136354923248291,
|
2282 |
+
"learning_rate": 1.597612065354001e-05,
|
2283 |
+
"loss": 5.8815,
|
2284 |
+
"step": 162500
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 34.143276078759946,
|
2288 |
+
"grad_norm": 5.668376445770264,
|
2289 |
+
"learning_rate": 1.587138667783829e-05,
|
2290 |
+
"loss": 5.8872,
|
2291 |
+
"step": 163000
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 34.248010054461666,
|
2295 |
+
"grad_norm": 5.579314708709717,
|
2296 |
+
"learning_rate": 1.5766862170087976e-05,
|
2297 |
+
"loss": 5.8966,
|
2298 |
+
"step": 163500
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 34.352744030163386,
|
2302 |
+
"grad_norm": 5.474893569946289,
|
2303 |
+
"learning_rate": 1.5662128194386262e-05,
|
2304 |
+
"loss": 5.8865,
|
2305 |
+
"step": 164000
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 34.457478005865106,
|
2309 |
+
"grad_norm": 4.853377342224121,
|
2310 |
+
"learning_rate": 1.5557394218684542e-05,
|
2311 |
+
"loss": 5.8718,
|
2312 |
+
"step": 164500
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 34.56221198156682,
|
2316 |
+
"grad_norm": 4.8684563636779785,
|
2317 |
+
"learning_rate": 1.5452660242982825e-05,
|
2318 |
+
"loss": 5.8896,
|
2319 |
+
"step": 165000
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 34.66694595726854,
|
2323 |
+
"grad_norm": 4.851233959197998,
|
2324 |
+
"learning_rate": 1.5347926267281108e-05,
|
2325 |
+
"loss": 5.8908,
|
2326 |
+
"step": 165500
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 34.77167993297026,
|
2330 |
+
"grad_norm": 6.647628307342529,
|
2331 |
+
"learning_rate": 1.524319229157939e-05,
|
2332 |
+
"loss": 5.8972,
|
2333 |
+
"step": 166000
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 34.87641390867197,
|
2337 |
+
"grad_norm": 5.826745986938477,
|
2338 |
+
"learning_rate": 1.5138667783829074e-05,
|
2339 |
+
"loss": 5.8852,
|
2340 |
+
"step": 166500
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 34.98114788437369,
|
2344 |
+
"grad_norm": 5.109675407409668,
|
2345 |
+
"learning_rate": 1.5033933808127357e-05,
|
2346 |
+
"loss": 5.8971,
|
2347 |
+
"step": 167000
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 35.08588186007541,
|
2351 |
+
"grad_norm": 5.743017196655273,
|
2352 |
+
"learning_rate": 1.4929199832425639e-05,
|
2353 |
+
"loss": 5.8892,
|
2354 |
+
"step": 167500
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 35.19061583577712,
|
2358 |
+
"grad_norm": 4.862270355224609,
|
2359 |
+
"learning_rate": 1.482446585672392e-05,
|
2360 |
+
"loss": 5.8976,
|
2361 |
+
"step": 168000
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 35.29534981147884,
|
2365 |
+
"grad_norm": 5.532686233520508,
|
2366 |
+
"learning_rate": 1.4719941348973607e-05,
|
2367 |
+
"loss": 5.8716,
|
2368 |
+
"step": 168500
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 35.40008378718056,
|
2372 |
+
"grad_norm": 5.498019695281982,
|
2373 |
+
"learning_rate": 1.4615207373271891e-05,
|
2374 |
+
"loss": 5.8768,
|
2375 |
+
"step": 169000
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 35.504817762882276,
|
2379 |
+
"grad_norm": 4.7324042320251465,
|
2380 |
+
"learning_rate": 1.4510473397570173e-05,
|
2381 |
+
"loss": 5.8739,
|
2382 |
+
"step": 169500
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 35.609551738583995,
|
2386 |
+
"grad_norm": 4.973413944244385,
|
2387 |
+
"learning_rate": 1.4405739421868456e-05,
|
2388 |
+
"loss": 5.8801,
|
2389 |
+
"step": 170000
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 35.714285714285715,
|
2393 |
+
"grad_norm": 4.977658271789551,
|
2394 |
+
"learning_rate": 1.430121491411814e-05,
|
2395 |
+
"loss": 5.8644,
|
2396 |
+
"step": 170500
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 35.819019689987435,
|
2400 |
+
"grad_norm": 5.551715850830078,
|
2401 |
+
"learning_rate": 1.4196480938416424e-05,
|
2402 |
+
"loss": 5.8789,
|
2403 |
+
"step": 171000
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 35.92375366568915,
|
2407 |
+
"grad_norm": 5.135740756988525,
|
2408 |
+
"learning_rate": 1.4091746962714705e-05,
|
2409 |
+
"loss": 5.8862,
|
2410 |
+
"step": 171500
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 36.02848764139087,
|
2414 |
+
"grad_norm": 5.068655967712402,
|
2415 |
+
"learning_rate": 1.3987012987012987e-05,
|
2416 |
+
"loss": 5.879,
|
2417 |
+
"step": 172000
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 36.13322161709259,
|
2421 |
+
"grad_norm": 5.393857479095459,
|
2422 |
+
"learning_rate": 1.388227901131127e-05,
|
2423 |
+
"loss": 5.8544,
|
2424 |
+
"step": 172500
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 36.2379555927943,
|
2428 |
+
"grad_norm": 5.854538440704346,
|
2429 |
+
"learning_rate": 1.3777545035609551e-05,
|
2430 |
+
"loss": 5.8824,
|
2431 |
+
"step": 173000
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 36.34268956849602,
|
2435 |
+
"grad_norm": 5.566401481628418,
|
2436 |
+
"learning_rate": 1.3673020527859238e-05,
|
2437 |
+
"loss": 5.8587,
|
2438 |
+
"step": 173500
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 36.44742354419774,
|
2442 |
+
"grad_norm": 6.091250896453857,
|
2443 |
+
"learning_rate": 1.3568286552157519e-05,
|
2444 |
+
"loss": 5.8624,
|
2445 |
+
"step": 174000
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 36.55215751989945,
|
2449 |
+
"grad_norm": 4.826417922973633,
|
2450 |
+
"learning_rate": 1.3463552576455804e-05,
|
2451 |
+
"loss": 5.879,
|
2452 |
+
"step": 174500
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 36.65689149560117,
|
2456 |
+
"grad_norm": 5.28770637512207,
|
2457 |
+
"learning_rate": 1.3358818600754087e-05,
|
2458 |
+
"loss": 5.8632,
|
2459 |
+
"step": 175000
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 36.76162547130289,
|
2463 |
+
"grad_norm": 5.072086811065674,
|
2464 |
+
"learning_rate": 1.3254084625052368e-05,
|
2465 |
+
"loss": 5.8698,
|
2466 |
+
"step": 175500
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 36.866359447004605,
|
2470 |
+
"grad_norm": 6.194067001342773,
|
2471 |
+
"learning_rate": 1.3149560117302053e-05,
|
2472 |
+
"loss": 5.8701,
|
2473 |
+
"step": 176000
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 36.971093422706325,
|
2477 |
+
"grad_norm": 5.250491142272949,
|
2478 |
+
"learning_rate": 1.3044826141600336e-05,
|
2479 |
+
"loss": 5.855,
|
2480 |
+
"step": 176500
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 37.075827398408045,
|
2484 |
+
"grad_norm": 4.9726080894470215,
|
2485 |
+
"learning_rate": 1.2940092165898618e-05,
|
2486 |
+
"loss": 5.8613,
|
2487 |
+
"step": 177000
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 37.180561374109764,
|
2491 |
+
"grad_norm": 5.526548385620117,
|
2492 |
+
"learning_rate": 1.28353581901969e-05,
|
2493 |
+
"loss": 5.8519,
|
2494 |
+
"step": 177500
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 37.28529534981148,
|
2498 |
+
"grad_norm": 5.5989861488342285,
|
2499 |
+
"learning_rate": 1.2730624214495182e-05,
|
2500 |
+
"loss": 5.8642,
|
2501 |
+
"step": 178000
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 37.3900293255132,
|
2505 |
+
"grad_norm": 5.138686180114746,
|
2506 |
+
"learning_rate": 1.2625890238793465e-05,
|
2507 |
+
"loss": 5.852,
|
2508 |
+
"step": 178500
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 37.49476330121492,
|
2512 |
+
"grad_norm": 5.0514326095581055,
|
2513 |
+
"learning_rate": 1.252115626309175e-05,
|
2514 |
+
"loss": 5.8484,
|
2515 |
+
"step": 179000
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 37.59949727691663,
|
2519 |
+
"grad_norm": 4.9300360679626465,
|
2520 |
+
"learning_rate": 1.241642228739003e-05,
|
2521 |
+
"loss": 5.849,
|
2522 |
+
"step": 179500
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 37.70423125261835,
|
2526 |
+
"grad_norm": 5.487224102020264,
|
2527 |
+
"learning_rate": 1.2311897779639716e-05,
|
2528 |
+
"loss": 5.8562,
|
2529 |
+
"step": 180000
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 37.80896522832007,
|
2533 |
+
"grad_norm": 5.826539516448975,
|
2534 |
+
"learning_rate": 1.2207163803937999e-05,
|
2535 |
+
"loss": 5.8665,
|
2536 |
+
"step": 180500
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 37.91369920402178,
|
2540 |
+
"grad_norm": 5.733819961547852,
|
2541 |
+
"learning_rate": 1.2102639296187684e-05,
|
2542 |
+
"loss": 5.8569,
|
2543 |
+
"step": 181000
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 38.0184331797235,
|
2547 |
+
"grad_norm": 4.917960166931152,
|
2548 |
+
"learning_rate": 1.1997905320485967e-05,
|
2549 |
+
"loss": 5.8395,
|
2550 |
+
"step": 181500
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 38.12316715542522,
|
2554 |
+
"grad_norm": 5.337119102478027,
|
2555 |
+
"learning_rate": 1.1893171344784248e-05,
|
2556 |
+
"loss": 5.854,
|
2557 |
+
"step": 182000
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 38.227901131126934,
|
2561 |
+
"grad_norm": 5.299139022827148,
|
2562 |
+
"learning_rate": 1.178843736908253e-05,
|
2563 |
+
"loss": 5.8433,
|
2564 |
+
"step": 182500
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 38.332635106828654,
|
2568 |
+
"grad_norm": 5.900153160095215,
|
2569 |
+
"learning_rate": 1.1683703393380813e-05,
|
2570 |
+
"loss": 5.8535,
|
2571 |
+
"step": 183000
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 38.437369082530374,
|
2575 |
+
"grad_norm": 6.776584625244141,
|
2576 |
+
"learning_rate": 1.15791788856305e-05,
|
2577 |
+
"loss": 5.8454,
|
2578 |
+
"step": 183500
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 38.542103058232094,
|
2582 |
+
"grad_norm": 6.258368015289307,
|
2583 |
+
"learning_rate": 1.1474444909928783e-05,
|
2584 |
+
"loss": 5.8354,
|
2585 |
+
"step": 184000
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 38.64683703393381,
|
2589 |
+
"grad_norm": 5.288670539855957,
|
2590 |
+
"learning_rate": 1.1369710934227064e-05,
|
2591 |
+
"loss": 5.8458,
|
2592 |
+
"step": 184500
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 38.751571009635526,
|
2596 |
+
"grad_norm": 5.596650123596191,
|
2597 |
+
"learning_rate": 1.1264976958525345e-05,
|
2598 |
+
"loss": 5.8387,
|
2599 |
+
"step": 185000
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 38.856304985337246,
|
2603 |
+
"grad_norm": 5.121638774871826,
|
2604 |
+
"learning_rate": 1.1160452450775032e-05,
|
2605 |
+
"loss": 5.8268,
|
2606 |
+
"step": 185500
|
2607 |
+
},
|
2608 |
+
{
|
2609 |
+
"epoch": 38.96103896103896,
|
2610 |
+
"grad_norm": 4.5758442878723145,
|
2611 |
+
"learning_rate": 1.1055718475073313e-05,
|
2612 |
+
"loss": 5.83,
|
2613 |
+
"step": 186000
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 39.06577293674068,
|
2617 |
+
"grad_norm": 5.161282539367676,
|
2618 |
+
"learning_rate": 1.0950984499371596e-05,
|
2619 |
+
"loss": 5.8544,
|
2620 |
+
"step": 186500
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 39.1705069124424,
|
2624 |
+
"grad_norm": 4.628884315490723,
|
2625 |
+
"learning_rate": 1.084625052366988e-05,
|
2626 |
+
"loss": 5.8474,
|
2627 |
+
"step": 187000
|
2628 |
+
},
|
2629 |
+
{
|
2630 |
+
"epoch": 39.27524088814411,
|
2631 |
+
"grad_norm": 5.854598045349121,
|
2632 |
+
"learning_rate": 1.074151654796816e-05,
|
2633 |
+
"loss": 5.8501,
|
2634 |
+
"step": 187500
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 39.37997486384583,
|
2638 |
+
"grad_norm": 5.315525054931641,
|
2639 |
+
"learning_rate": 1.0636782572266444e-05,
|
2640 |
+
"loss": 5.8265,
|
2641 |
+
"step": 188000
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"epoch": 39.48470883954755,
|
2645 |
+
"grad_norm": 6.078185081481934,
|
2646 |
+
"learning_rate": 1.0532048596564727e-05,
|
2647 |
+
"loss": 5.8419,
|
2648 |
+
"step": 188500
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 39.589442815249264,
|
2652 |
+
"grad_norm": 5.223086357116699,
|
2653 |
+
"learning_rate": 1.0427524088814412e-05,
|
2654 |
+
"loss": 5.8197,
|
2655 |
+
"step": 189000
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 39.69417679095098,
|
2659 |
+
"grad_norm": 5.235757827758789,
|
2660 |
+
"learning_rate": 1.0322790113112695e-05,
|
2661 |
+
"loss": 5.8245,
|
2662 |
+
"step": 189500
|
2663 |
+
},
|
2664 |
+
{
|
2665 |
+
"epoch": 39.7989107666527,
|
2666 |
+
"grad_norm": 5.124643325805664,
|
2667 |
+
"learning_rate": 1.0218056137410976e-05,
|
2668 |
+
"loss": 5.839,
|
2669 |
+
"step": 190000
|
2670 |
+
},
|
2671 |
+
{
|
2672 |
+
"epoch": 39.90364474235442,
|
2673 |
+
"grad_norm": 5.613321304321289,
|
2674 |
+
"learning_rate": 1.011332216170926e-05,
|
2675 |
+
"loss": 5.844,
|
2676 |
+
"step": 190500
|
2677 |
+
},
|
2678 |
+
{
|
2679 |
+
"epoch": 40.008378718056136,
|
2680 |
+
"grad_norm": 5.873430252075195,
|
2681 |
+
"learning_rate": 1.0008588186007542e-05,
|
2682 |
+
"loss": 5.837,
|
2683 |
+
"step": 191000
|
2684 |
+
},
|
2685 |
+
{
|
2686 |
+
"epoch": 40.113112693757856,
|
2687 |
+
"grad_norm": 5.089309215545654,
|
2688 |
+
"learning_rate": 9.903854210305824e-06,
|
2689 |
+
"loss": 5.8384,
|
2690 |
+
"step": 191500
|
2691 |
+
},
|
2692 |
+
{
|
2693 |
+
"epoch": 40.217846669459576,
|
2694 |
+
"grad_norm": 7.3569817543029785,
|
2695 |
+
"learning_rate": 9.79932970255551e-06,
|
2696 |
+
"loss": 5.8232,
|
2697 |
+
"step": 192000
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 40.32258064516129,
|
2701 |
+
"grad_norm": 6.024489402770996,
|
2702 |
+
"learning_rate": 9.694595726853792e-06,
|
2703 |
+
"loss": 5.8292,
|
2704 |
+
"step": 192500
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"epoch": 40.42731462086301,
|
2708 |
+
"grad_norm": 5.7150983810424805,
|
2709 |
+
"learning_rate": 9.589861751152073e-06,
|
2710 |
+
"loss": 5.8614,
|
2711 |
+
"step": 193000
|
2712 |
+
},
|
2713 |
+
{
|
2714 |
+
"epoch": 40.53204859656473,
|
2715 |
+
"grad_norm": 4.717107772827148,
|
2716 |
+
"learning_rate": 9.485127775450356e-06,
|
2717 |
+
"loss": 5.8092,
|
2718 |
+
"step": 193500
|
2719 |
+
},
|
2720 |
+
{
|
2721 |
+
"epoch": 40.63678257226644,
|
2722 |
+
"grad_norm": 4.9722490310668945,
|
2723 |
+
"learning_rate": 9.380603267700043e-06,
|
2724 |
+
"loss": 5.8231,
|
2725 |
+
"step": 194000
|
2726 |
+
},
|
2727 |
+
{
|
2728 |
+
"epoch": 40.74151654796816,
|
2729 |
+
"grad_norm": 5.593094825744629,
|
2730 |
+
"learning_rate": 9.275869291998326e-06,
|
2731 |
+
"loss": 5.8339,
|
2732 |
+
"step": 194500
|
2733 |
+
},
|
2734 |
+
{
|
2735 |
+
"epoch": 40.84625052366988,
|
2736 |
+
"grad_norm": 5.731310844421387,
|
2737 |
+
"learning_rate": 9.171135316296607e-06,
|
2738 |
+
"loss": 5.8381,
|
2739 |
+
"step": 195000
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 40.95098449937159,
|
2743 |
+
"grad_norm": 5.072065353393555,
|
2744 |
+
"learning_rate": 9.066401340594889e-06,
|
2745 |
+
"loss": 5.8367,
|
2746 |
+
"step": 195500
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 41.05571847507331,
|
2750 |
+
"grad_norm": 5.219040870666504,
|
2751 |
+
"learning_rate": 8.961667364893172e-06,
|
2752 |
+
"loss": 5.8234,
|
2753 |
+
"step": 196000
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 41.16045245077503,
|
2757 |
+
"grad_norm": 5.844238758087158,
|
2758 |
+
"learning_rate": 8.856933389191455e-06,
|
2759 |
+
"loss": 5.8347,
|
2760 |
+
"step": 196500
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 41.26518642647675,
|
2764 |
+
"grad_norm": 6.088447093963623,
|
2765 |
+
"learning_rate": 8.752199413489736e-06,
|
2766 |
+
"loss": 5.8178,
|
2767 |
+
"step": 197000
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 41.369920402178465,
|
2771 |
+
"grad_norm": 5.14108943939209,
|
2772 |
+
"learning_rate": 8.647465437788019e-06,
|
2773 |
+
"loss": 5.8248,
|
2774 |
+
"step": 197500
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 41.474654377880185,
|
2778 |
+
"grad_norm": 5.424249172210693,
|
2779 |
+
"learning_rate": 8.542940930037704e-06,
|
2780 |
+
"loss": 5.8113,
|
2781 |
+
"step": 198000
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 41.579388353581905,
|
2785 |
+
"grad_norm": 4.888121604919434,
|
2786 |
+
"learning_rate": 8.43841642228739e-06,
|
2787 |
+
"loss": 5.8111,
|
2788 |
+
"step": 198500
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 41.68412232928362,
|
2792 |
+
"grad_norm": 4.9909515380859375,
|
2793 |
+
"learning_rate": 8.333682446585672e-06,
|
2794 |
+
"loss": 5.8276,
|
2795 |
+
"step": 199000
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 41.78885630498534,
|
2799 |
+
"grad_norm": 5.032175540924072,
|
2800 |
+
"learning_rate": 8.228948470883955e-06,
|
2801 |
+
"loss": 5.8332,
|
2802 |
+
"step": 199500
|
2803 |
+
},
|
2804 |
+
{
|
2805 |
+
"epoch": 41.89359028068706,
|
2806 |
+
"grad_norm": 5.116880416870117,
|
2807 |
+
"learning_rate": 8.124214495182238e-06,
|
2808 |
+
"loss": 5.8233,
|
2809 |
+
"step": 200000
|
2810 |
+
},
|
2811 |
+
{
|
2812 |
+
"epoch": 41.99832425638877,
|
2813 |
+
"grad_norm": 5.235647678375244,
|
2814 |
+
"learning_rate": 8.019689987431923e-06,
|
2815 |
+
"loss": 5.8297,
|
2816 |
+
"step": 200500
|
2817 |
+
},
|
2818 |
+
{
|
2819 |
+
"epoch": 42.10305823209049,
|
2820 |
+
"grad_norm": 5.445380210876465,
|
2821 |
+
"learning_rate": 7.914956011730206e-06,
|
2822 |
+
"loss": 5.8159,
|
2823 |
+
"step": 201000
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 42.20779220779221,
|
2827 |
+
"grad_norm": 4.979036331176758,
|
2828 |
+
"learning_rate": 7.810222036028488e-06,
|
2829 |
+
"loss": 5.809,
|
2830 |
+
"step": 201500
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"epoch": 42.31252618349392,
|
2834 |
+
"grad_norm": 5.359362602233887,
|
2835 |
+
"learning_rate": 7.70548806032677e-06,
|
2836 |
+
"loss": 5.8346,
|
2837 |
+
"step": 202000
|
2838 |
+
},
|
2839 |
+
{
|
2840 |
+
"epoch": 42.41726015919564,
|
2841 |
+
"grad_norm": 5.264519214630127,
|
2842 |
+
"learning_rate": 7.600754084625053e-06,
|
2843 |
+
"loss": 5.8089,
|
2844 |
+
"step": 202500
|
2845 |
+
},
|
2846 |
+
{
|
2847 |
+
"epoch": 42.52199413489736,
|
2848 |
+
"grad_norm": 5.985982894897461,
|
2849 |
+
"learning_rate": 7.496020108923335e-06,
|
2850 |
+
"loss": 5.8192,
|
2851 |
+
"step": 203000
|
2852 |
+
},
|
2853 |
+
{
|
2854 |
+
"epoch": 42.626728110599075,
|
2855 |
+
"grad_norm": 5.505626201629639,
|
2856 |
+
"learning_rate": 7.391286133221617e-06,
|
2857 |
+
"loss": 5.8095,
|
2858 |
+
"step": 203500
|
2859 |
+
},
|
2860 |
+
{
|
2861 |
+
"epoch": 42.731462086300795,
|
2862 |
+
"grad_norm": 5.069738388061523,
|
2863 |
+
"learning_rate": 7.286552157519899e-06,
|
2864 |
+
"loss": 5.8186,
|
2865 |
+
"step": 204000
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 42.836196062002514,
|
2869 |
+
"grad_norm": 6.004745960235596,
|
2870 |
+
"learning_rate": 7.182027649769586e-06,
|
2871 |
+
"loss": 5.8136,
|
2872 |
+
"step": 204500
|
2873 |
+
},
|
2874 |
+
{
|
2875 |
+
"epoch": 42.940930037704234,
|
2876 |
+
"grad_norm": 6.299502372741699,
|
2877 |
+
"learning_rate": 7.077293674067868e-06,
|
2878 |
+
"loss": 5.8213,
|
2879 |
+
"step": 205000
|
2880 |
+
},
|
2881 |
+
{
|
2882 |
+
"epoch": 43.04566401340595,
|
2883 |
+
"grad_norm": 6.302718162536621,
|
2884 |
+
"learning_rate": 6.97255969836615e-06,
|
2885 |
+
"loss": 5.8075,
|
2886 |
+
"step": 205500
|
2887 |
+
},
|
2888 |
+
{
|
2889 |
+
"epoch": 43.15039798910767,
|
2890 |
+
"grad_norm": 5.921250343322754,
|
2891 |
+
"learning_rate": 6.867825722664433e-06,
|
2892 |
+
"loss": 5.8276,
|
2893 |
+
"step": 206000
|
2894 |
+
},
|
2895 |
+
{
|
2896 |
+
"epoch": 43.25513196480939,
|
2897 |
+
"grad_norm": 5.123110771179199,
|
2898 |
+
"learning_rate": 6.763091746962715e-06,
|
2899 |
+
"loss": 5.7965,
|
2900 |
+
"step": 206500
|
2901 |
+
},
|
2902 |
+
{
|
2903 |
+
"epoch": 43.3598659405111,
|
2904 |
+
"grad_norm": 5.187294006347656,
|
2905 |
+
"learning_rate": 6.658357771260998e-06,
|
2906 |
+
"loss": 5.8137,
|
2907 |
+
"step": 207000
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 43.46459991621282,
|
2911 |
+
"grad_norm": 5.407510757446289,
|
2912 |
+
"learning_rate": 6.55362379555928e-06,
|
2913 |
+
"loss": 5.8305,
|
2914 |
+
"step": 207500
|
2915 |
+
},
|
2916 |
+
{
|
2917 |
+
"epoch": 43.56933389191454,
|
2918 |
+
"grad_norm": 5.892600059509277,
|
2919 |
+
"learning_rate": 6.449099287808966e-06,
|
2920 |
+
"loss": 5.8167,
|
2921 |
+
"step": 208000
|
2922 |
+
},
|
2923 |
+
{
|
2924 |
+
"epoch": 43.67406786761625,
|
2925 |
+
"grad_norm": 5.39382266998291,
|
2926 |
+
"learning_rate": 6.344365312107248e-06,
|
2927 |
+
"loss": 5.8185,
|
2928 |
+
"step": 208500
|
2929 |
+
},
|
2930 |
+
{
|
2931 |
+
"epoch": 43.77880184331797,
|
2932 |
+
"grad_norm": 5.608034133911133,
|
2933 |
+
"learning_rate": 6.23963133640553e-06,
|
2934 |
+
"loss": 5.8051,
|
2935 |
+
"step": 209000
|
2936 |
+
},
|
2937 |
+
{
|
2938 |
+
"epoch": 43.88353581901969,
|
2939 |
+
"grad_norm": 6.069722652435303,
|
2940 |
+
"learning_rate": 6.1348973607038125e-06,
|
2941 |
+
"loss": 5.8101,
|
2942 |
+
"step": 209500
|
2943 |
+
},
|
2944 |
+
{
|
2945 |
+
"epoch": 43.988269794721404,
|
2946 |
+
"grad_norm": 5.938599109649658,
|
2947 |
+
"learning_rate": 6.0301633850020955e-06,
|
2948 |
+
"loss": 5.807,
|
2949 |
+
"step": 210000
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 44.093003770423124,
|
2953 |
+
"grad_norm": 5.808456897735596,
|
2954 |
+
"learning_rate": 5.925429409300378e-06,
|
2955 |
+
"loss": 5.8246,
|
2956 |
+
"step": 210500
|
2957 |
+
},
|
2958 |
+
{
|
2959 |
+
"epoch": 44.197737746124844,
|
2960 |
+
"grad_norm": 5.229996681213379,
|
2961 |
+
"learning_rate": 5.820904901550063e-06,
|
2962 |
+
"loss": 5.8029,
|
2963 |
+
"step": 211000
|
2964 |
+
},
|
2965 |
+
{
|
2966 |
+
"epoch": 44.302471721826564,
|
2967 |
+
"grad_norm": 5.295706748962402,
|
2968 |
+
"learning_rate": 5.716170925848346e-06,
|
2969 |
+
"loss": 5.8094,
|
2970 |
+
"step": 211500
|
2971 |
+
},
|
2972 |
+
{
|
2973 |
+
"epoch": 44.407205697528276,
|
2974 |
+
"grad_norm": 5.649194240570068,
|
2975 |
+
"learning_rate": 5.611436950146628e-06,
|
2976 |
+
"loss": 5.811,
|
2977 |
+
"step": 212000
|
2978 |
+
},
|
2979 |
+
{
|
2980 |
+
"epoch": 44.511939673229996,
|
2981 |
+
"grad_norm": 6.5928521156311035,
|
2982 |
+
"learning_rate": 5.50670297444491e-06,
|
2983 |
+
"loss": 5.7974,
|
2984 |
+
"step": 212500
|
2985 |
+
},
|
2986 |
+
{
|
2987 |
+
"epoch": 44.616673648931716,
|
2988 |
+
"grad_norm": 6.246605396270752,
|
2989 |
+
"learning_rate": 5.401968998743192e-06,
|
2990 |
+
"loss": 5.8011,
|
2991 |
+
"step": 213000
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 44.72140762463343,
|
2995 |
+
"grad_norm": 5.312093734741211,
|
2996 |
+
"learning_rate": 5.2972350230414745e-06,
|
2997 |
+
"loss": 5.7819,
|
2998 |
+
"step": 213500
|
2999 |
+
},
|
3000 |
+
{
|
3001 |
+
"epoch": 44.82614160033515,
|
3002 |
+
"grad_norm": 5.348554611206055,
|
3003 |
+
"learning_rate": 5.19271051529116e-06,
|
3004 |
+
"loss": 5.8027,
|
3005 |
+
"step": 214000
|
3006 |
+
},
|
3007 |
+
{
|
3008 |
+
"epoch": 44.93087557603687,
|
3009 |
+
"grad_norm": 5.95352029800415,
|
3010 |
+
"learning_rate": 5.087976539589443e-06,
|
3011 |
+
"loss": 5.8046,
|
3012 |
+
"step": 214500
|
3013 |
+
},
|
3014 |
+
{
|
3015 |
+
"epoch": 45.03560955173858,
|
3016 |
+
"grad_norm": 5.978014945983887,
|
3017 |
+
"learning_rate": 4.983242563887726e-06,
|
3018 |
+
"loss": 5.8021,
|
3019 |
+
"step": 215000
|
3020 |
+
},
|
3021 |
+
{
|
3022 |
+
"epoch": 45.1403435274403,
|
3023 |
+
"grad_norm": 5.595849990844727,
|
3024 |
+
"learning_rate": 4.878508588186008e-06,
|
3025 |
+
"loss": 5.7996,
|
3026 |
+
"step": 215500
|
3027 |
+
},
|
3028 |
+
{
|
3029 |
+
"epoch": 45.24507750314202,
|
3030 |
+
"grad_norm": 5.570345401763916,
|
3031 |
+
"learning_rate": 4.77377461248429e-06,
|
3032 |
+
"loss": 5.7973,
|
3033 |
+
"step": 216000
|
3034 |
+
},
|
3035 |
+
{
|
3036 |
+
"epoch": 45.34981147884373,
|
3037 |
+
"grad_norm": 5.320748805999756,
|
3038 |
+
"learning_rate": 4.669040636782573e-06,
|
3039 |
+
"loss": 5.7886,
|
3040 |
+
"step": 216500
|
3041 |
+
},
|
3042 |
+
{
|
3043 |
+
"epoch": 45.45454545454545,
|
3044 |
+
"grad_norm": 4.676185607910156,
|
3045 |
+
"learning_rate": 4.564516129032258e-06,
|
3046 |
+
"loss": 5.7874,
|
3047 |
+
"step": 217000
|
3048 |
+
},
|
3049 |
+
{
|
3050 |
+
"epoch": 45.55927943024717,
|
3051 |
+
"grad_norm": 5.7768473625183105,
|
3052 |
+
"learning_rate": 4.45978215333054e-06,
|
3053 |
+
"loss": 5.7875,
|
3054 |
+
"step": 217500
|
3055 |
+
},
|
3056 |
+
{
|
3057 |
+
"epoch": 45.66401340594889,
|
3058 |
+
"grad_norm": 5.668895244598389,
|
3059 |
+
"learning_rate": 4.355048177628823e-06,
|
3060 |
+
"loss": 5.8121,
|
3061 |
+
"step": 218000
|
3062 |
+
},
|
3063 |
+
{
|
3064 |
+
"epoch": 45.768747381650606,
|
3065 |
+
"grad_norm": 5.033557891845703,
|
3066 |
+
"learning_rate": 4.2503142019271055e-06,
|
3067 |
+
"loss": 5.8137,
|
3068 |
+
"step": 218500
|
3069 |
+
},
|
3070 |
+
{
|
3071 |
+
"epoch": 45.873481357352325,
|
3072 |
+
"grad_norm": 6.15772819519043,
|
3073 |
+
"learning_rate": 4.145580226225388e-06,
|
3074 |
+
"loss": 5.8137,
|
3075 |
+
"step": 219000
|
3076 |
+
},
|
3077 |
+
{
|
3078 |
+
"epoch": 45.978215333054045,
|
3079 |
+
"grad_norm": 6.617910861968994,
|
3080 |
+
"learning_rate": 4.04084625052367e-06,
|
3081 |
+
"loss": 5.8079,
|
3082 |
+
"step": 219500
|
3083 |
+
},
|
3084 |
+
{
|
3085 |
+
"epoch": 46.08294930875576,
|
3086 |
+
"grad_norm": 5.210205554962158,
|
3087 |
+
"learning_rate": 3.936112274821952e-06,
|
3088 |
+
"loss": 5.8189,
|
3089 |
+
"step": 220000
|
3090 |
+
},
|
3091 |
+
{
|
3092 |
+
"epoch": 46.18768328445748,
|
3093 |
+
"grad_norm": 5.630945205688477,
|
3094 |
+
"learning_rate": 3.831587767071639e-06,
|
3095 |
+
"loss": 5.795,
|
3096 |
+
"step": 220500
|
3097 |
+
},
|
3098 |
+
{
|
3099 |
+
"epoch": 46.2924172601592,
|
3100 |
+
"grad_norm": 5.8690032958984375,
|
3101 |
+
"learning_rate": 3.7268537913699205e-06,
|
3102 |
+
"loss": 5.8019,
|
3103 |
+
"step": 221000
|
3104 |
+
},
|
3105 |
+
{
|
3106 |
+
"epoch": 46.39715123586091,
|
3107 |
+
"grad_norm": 5.787112712860107,
|
3108 |
+
"learning_rate": 3.6221198156682027e-06,
|
3109 |
+
"loss": 5.7901,
|
3110 |
+
"step": 221500
|
3111 |
+
},
|
3112 |
+
{
|
3113 |
+
"epoch": 46.50188521156263,
|
3114 |
+
"grad_norm": 5.568469524383545,
|
3115 |
+
"learning_rate": 3.5173858399664853e-06,
|
3116 |
+
"loss": 5.803,
|
3117 |
+
"step": 222000
|
3118 |
+
},
|
3119 |
+
{
|
3120 |
+
"epoch": 46.60661918726435,
|
3121 |
+
"grad_norm": 5.671326637268066,
|
3122 |
+
"learning_rate": 3.4126518642647675e-06,
|
3123 |
+
"loss": 5.7931,
|
3124 |
+
"step": 222500
|
3125 |
+
},
|
3126 |
+
{
|
3127 |
+
"epoch": 46.71135316296606,
|
3128 |
+
"grad_norm": 5.4085307121276855,
|
3129 |
+
"learning_rate": 3.30791788856305e-06,
|
3130 |
+
"loss": 5.7903,
|
3131 |
+
"step": 223000
|
3132 |
+
},
|
3133 |
+
{
|
3134 |
+
"epoch": 46.81608713866778,
|
3135 |
+
"grad_norm": 5.7440571784973145,
|
3136 |
+
"learning_rate": 3.203393380812736e-06,
|
3137 |
+
"loss": 5.7897,
|
3138 |
+
"step": 223500
|
3139 |
+
},
|
3140 |
+
{
|
3141 |
+
"epoch": 46.9208211143695,
|
3142 |
+
"grad_norm": 5.144542217254639,
|
3143 |
+
"learning_rate": 3.098659405111018e-06,
|
3144 |
+
"loss": 5.8145,
|
3145 |
+
"step": 224000
|
3146 |
+
},
|
3147 |
+
{
|
3148 |
+
"epoch": 47.02555509007122,
|
3149 |
+
"grad_norm": 5.842213153839111,
|
3150 |
+
"learning_rate": 2.9939254294093008e-06,
|
3151 |
+
"loss": 5.8046,
|
3152 |
+
"step": 224500
|
3153 |
+
},
|
3154 |
+
{
|
3155 |
+
"epoch": 47.130289065772935,
|
3156 |
+
"grad_norm": 6.161410331726074,
|
3157 |
+
"learning_rate": 2.8891914537075826e-06,
|
3158 |
+
"loss": 5.7965,
|
3159 |
+
"step": 225000
|
3160 |
+
},
|
3161 |
+
{
|
3162 |
+
"epoch": 47.235023041474655,
|
3163 |
+
"grad_norm": 6.173724174499512,
|
3164 |
+
"learning_rate": 2.784457478005865e-06,
|
3165 |
+
"loss": 5.7854,
|
3166 |
+
"step": 225500
|
3167 |
+
},
|
3168 |
+
{
|
3169 |
+
"epoch": 47.339757017176375,
|
3170 |
+
"grad_norm": 5.132796287536621,
|
3171 |
+
"learning_rate": 2.679932970255551e-06,
|
3172 |
+
"loss": 5.791,
|
3173 |
+
"step": 226000
|
3174 |
+
},
|
3175 |
+
{
|
3176 |
+
"epoch": 47.44449099287809,
|
3177 |
+
"grad_norm": 6.053417205810547,
|
3178 |
+
"learning_rate": 2.5751989945538332e-06,
|
3179 |
+
"loss": 5.7935,
|
3180 |
+
"step": 226500
|
3181 |
+
},
|
3182 |
+
{
|
3183 |
+
"epoch": 47.54922496857981,
|
3184 |
+
"grad_norm": 5.08466911315918,
|
3185 |
+
"learning_rate": 2.470465018852116e-06,
|
3186 |
+
"loss": 5.8006,
|
3187 |
+
"step": 227000
|
3188 |
+
},
|
3189 |
+
{
|
3190 |
+
"epoch": 47.65395894428153,
|
3191 |
+
"grad_norm": 6.060305595397949,
|
3192 |
+
"learning_rate": 2.365731043150398e-06,
|
3193 |
+
"loss": 5.7839,
|
3194 |
+
"step": 227500
|
3195 |
+
},
|
3196 |
+
{
|
3197 |
+
"epoch": 47.75869291998324,
|
3198 |
+
"grad_norm": 5.808520317077637,
|
3199 |
+
"learning_rate": 2.2609970674486806e-06,
|
3200 |
+
"loss": 5.7933,
|
3201 |
+
"step": 228000
|
3202 |
+
},
|
3203 |
+
{
|
3204 |
+
"epoch": 47.86342689568496,
|
3205 |
+
"grad_norm": 5.413971424102783,
|
3206 |
+
"learning_rate": 2.156472559698366e-06,
|
3207 |
+
"loss": 5.7985,
|
3208 |
+
"step": 228500
|
3209 |
+
},
|
3210 |
+
{
|
3211 |
+
"epoch": 47.96816087138668,
|
3212 |
+
"grad_norm": 6.4786529541015625,
|
3213 |
+
"learning_rate": 2.0517385839966487e-06,
|
3214 |
+
"loss": 5.8059,
|
3215 |
+
"step": 229000
|
3216 |
+
},
|
3217 |
+
{
|
3218 |
+
"epoch": 48.07289484708839,
|
3219 |
+
"grad_norm": 5.606147289276123,
|
3220 |
+
"learning_rate": 1.947004608294931e-06,
|
3221 |
+
"loss": 5.7934,
|
3222 |
+
"step": 229500
|
3223 |
+
},
|
3224 |
+
{
|
3225 |
+
"epoch": 48.17762882279011,
|
3226 |
+
"grad_norm": 5.827240467071533,
|
3227 |
+
"learning_rate": 1.8422706325932133e-06,
|
3228 |
+
"loss": 5.8007,
|
3229 |
+
"step": 230000
|
3230 |
+
},
|
3231 |
+
{
|
3232 |
+
"epoch": 48.28236279849183,
|
3233 |
+
"grad_norm": 5.678854465484619,
|
3234 |
+
"learning_rate": 1.7375366568914957e-06,
|
3235 |
+
"loss": 5.8038,
|
3236 |
+
"step": 230500
|
3237 |
+
},
|
3238 |
+
{
|
3239 |
+
"epoch": 48.38709677419355,
|
3240 |
+
"grad_norm": 5.42138671875,
|
3241 |
+
"learning_rate": 1.632802681189778e-06,
|
3242 |
+
"loss": 5.7976,
|
3243 |
+
"step": 231000
|
3244 |
+
},
|
3245 |
+
{
|
3246 |
+
"epoch": 48.491830749895264,
|
3247 |
+
"grad_norm": 6.03090238571167,
|
3248 |
+
"learning_rate": 1.5280687054880603e-06,
|
3249 |
+
"loss": 5.7874,
|
3250 |
+
"step": 231500
|
3251 |
+
},
|
3252 |
+
{
|
3253 |
+
"epoch": 48.596564725596984,
|
3254 |
+
"grad_norm": 4.660221099853516,
|
3255 |
+
"learning_rate": 1.4233347297863427e-06,
|
3256 |
+
"loss": 5.7742,
|
3257 |
+
"step": 232000
|
3258 |
+
},
|
3259 |
+
{
|
3260 |
+
"epoch": 48.701298701298704,
|
3261 |
+
"grad_norm": 6.032063961029053,
|
3262 |
+
"learning_rate": 1.3188102220360285e-06,
|
3263 |
+
"loss": 5.7866,
|
3264 |
+
"step": 232500
|
3265 |
+
},
|
3266 |
+
{
|
3267 |
+
"epoch": 48.80603267700042,
|
3268 |
+
"grad_norm": 6.296925067901611,
|
3269 |
+
"learning_rate": 1.2142857142857144e-06,
|
3270 |
+
"loss": 5.7922,
|
3271 |
+
"step": 233000
|
3272 |
+
},
|
3273 |
+
{
|
3274 |
+
"epoch": 48.91076665270214,
|
3275 |
+
"grad_norm": 5.2991437911987305,
|
3276 |
+
"learning_rate": 1.1095517385839968e-06,
|
3277 |
+
"loss": 5.7873,
|
3278 |
+
"step": 233500
|
3279 |
+
},
|
3280 |
+
{
|
3281 |
+
"epoch": 49.015500628403856,
|
3282 |
+
"grad_norm": 6.130777835845947,
|
3283 |
+
"learning_rate": 1.0048177628822792e-06,
|
3284 |
+
"loss": 5.8074,
|
3285 |
+
"step": 234000
|
3286 |
+
},
|
3287 |
+
{
|
3288 |
+
"epoch": 49.12023460410557,
|
3289 |
+
"grad_norm": 6.305094242095947,
|
3290 |
+
"learning_rate": 9.000837871805614e-07,
|
3291 |
+
"loss": 5.7847,
|
3292 |
+
"step": 234500
|
3293 |
+
},
|
3294 |
+
{
|
3295 |
+
"epoch": 49.22496857980729,
|
3296 |
+
"grad_norm": 6.156949996948242,
|
3297 |
+
"learning_rate": 7.953498114788438e-07,
|
3298 |
+
"loss": 5.7741,
|
3299 |
+
"step": 235000
|
3300 |
+
},
|
3301 |
+
{
|
3302 |
+
"epoch": 49.32970255550901,
|
3303 |
+
"grad_norm": 6.475966930389404,
|
3304 |
+
"learning_rate": 6.908253037285296e-07,
|
3305 |
+
"loss": 5.8083,
|
3306 |
+
"step": 235500
|
3307 |
+
},
|
3308 |
+
{
|
3309 |
+
"epoch": 49.43443653121072,
|
3310 |
+
"grad_norm": 5.687895774841309,
|
3311 |
+
"learning_rate": 5.86091328026812e-07,
|
3312 |
+
"loss": 5.7976,
|
3313 |
+
"step": 236000
|
3314 |
+
},
|
3315 |
+
{
|
3316 |
+
"epoch": 49.53917050691244,
|
3317 |
+
"grad_norm": 5.615401744842529,
|
3318 |
+
"learning_rate": 4.813573523250943e-07,
|
3319 |
+
"loss": 5.7907,
|
3320 |
+
"step": 236500
|
3321 |
+
},
|
3322 |
+
{
|
3323 |
+
"epoch": 49.64390448261416,
|
3324 |
+
"grad_norm": 6.177160263061523,
|
3325 |
+
"learning_rate": 3.7662337662337666e-07,
|
3326 |
+
"loss": 5.792,
|
3327 |
+
"step": 237000
|
3328 |
+
},
|
3329 |
+
{
|
3330 |
+
"epoch": 49.74863845831588,
|
3331 |
+
"grad_norm": 5.604287624359131,
|
3332 |
+
"learning_rate": 2.71889400921659e-07,
|
3333 |
+
"loss": 5.7837,
|
3334 |
+
"step": 237500
|
3335 |
+
},
|
3336 |
+
{
|
3337 |
+
"epoch": 49.853372434017594,
|
3338 |
+
"grad_norm": 4.826539039611816,
|
3339 |
+
"learning_rate": 1.673648931713448e-07,
|
3340 |
+
"loss": 5.7808,
|
3341 |
+
"step": 238000
|
3342 |
+
},
|
3343 |
+
{
|
3344 |
+
"epoch": 49.95810640971931,
|
3345 |
+
"grad_norm": 5.6649250984191895,
|
3346 |
+
"learning_rate": 6.263091746962715e-08,
|
3347 |
+
"loss": 5.7888,
|
3348 |
+
"step": 238500
|
3349 |
+
},
|
3350 |
+
{
|
3351 |
+
"epoch": 50.0,
|
3352 |
+
"step": 238700,
|
3353 |
+
"total_flos": 5161725447936000.0,
|
3354 |
+
"train_loss": 6.140905278960581,
|
3355 |
+
"train_runtime": 7883.0646,
|
3356 |
+
"train_samples_per_second": 484.444,
|
3357 |
+
"train_steps_per_second": 30.28
|
3358 |
+
}
|
3359 |
+
],
|
3360 |
+
"logging_steps": 500,
|
3361 |
+
"max_steps": 238700,
|
3362 |
+
"num_input_tokens_seen": 0,
|
3363 |
+
"num_train_epochs": 50,
|
3364 |
+
"save_steps": 5000,
|
3365 |
+
"stateful_callbacks": {
|
3366 |
+
"TrainerControl": {
|
3367 |
+
"args": {
|
3368 |
+
"should_epoch_stop": false,
|
3369 |
+
"should_evaluate": false,
|
3370 |
+
"should_log": false,
|
3371 |
+
"should_save": true,
|
3372 |
+
"should_training_stop": true
|
3373 |
+
},
|
3374 |
+
"attributes": {}
|
3375 |
+
}
|
3376 |
+
},
|
3377 |
+
"total_flos": 5161725447936000.0,
|
3378 |
+
"train_batch_size": 16,
|
3379 |
+
"trial_name": null,
|
3380 |
+
"trial_params": null
|
3381 |
+
}
|