End of training
Browse files- README.md +19 -2
- all_results.json +16 -0
- eval_results.json +10 -0
- runs/Feb22_21-13-07_gitspot/events.out.tfevents.1740258728.gitspot.1976043.1 +3 -0
- train_results.json +9 -0
- trainer_state.json +1890 -0
README.md
CHANGED
|
@@ -4,9 +4,23 @@ license: cc-by-sa-4.0
|
|
| 4 |
base_model: nlpaueb/legal-bert-base-uncased
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
model-index:
|
| 8 |
- name: legal-bert
|
| 9 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
---
|
| 11 |
|
| 12 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -14,7 +28,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# legal-bert
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [nlpaueb/legal-bert-base-uncased](https://huggingface.co/nlpaueb/legal-bert-base-uncased) on
|
|
|
|
|
|
|
|
|
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
|
|
|
| 4 |
base_model: nlpaueb/legal-bert-base-uncased
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
+
datasets:
|
| 8 |
+
- alayaran/bodo-monolingual-dataset
|
| 9 |
+
metrics:
|
| 10 |
+
- accuracy
|
| 11 |
model-index:
|
| 12 |
- name: legal-bert
|
| 13 |
+
results:
|
| 14 |
+
- task:
|
| 15 |
+
name: Masked Language Modeling
|
| 16 |
+
type: fill-mask
|
| 17 |
+
dataset:
|
| 18 |
+
name: alayaran/bodo-monolingual-dataset
|
| 19 |
+
type: alayaran/bodo-monolingual-dataset
|
| 20 |
+
metrics:
|
| 21 |
+
- name: Accuracy
|
| 22 |
+
type: accuracy
|
| 23 |
+
value: 0.15121982702981798
|
| 24 |
---
|
| 25 |
|
| 26 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 28 |
|
| 29 |
# legal-bert
|
| 30 |
|
| 31 |
+
This model is a fine-tuned version of [nlpaueb/legal-bert-base-uncased](https://huggingface.co/nlpaueb/legal-bert-base-uncased) on the alayaran/bodo-monolingual-dataset dataset.
|
| 32 |
+
It achieves the following results on the evaluation set:
|
| 33 |
+
- Loss: 4.3914
|
| 34 |
+
- Accuracy: 0.1512
|
| 35 |
|
| 36 |
## Model description
|
| 37 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.15121982702981798,
|
| 4 |
+
"eval_loss": 4.391414165496826,
|
| 5 |
+
"eval_runtime": 55.6914,
|
| 6 |
+
"eval_samples": 12892,
|
| 7 |
+
"eval_samples_per_second": 231.49,
|
| 8 |
+
"eval_steps_per_second": 28.945,
|
| 9 |
+
"perplexity": 80.75453855161017,
|
| 10 |
+
"total_flos": 6.962502728150323e+16,
|
| 11 |
+
"train_loss": 4.240191396162094,
|
| 12 |
+
"train_runtime": 19577.9877,
|
| 13 |
+
"train_samples": 264467,
|
| 14 |
+
"train_samples_per_second": 54.034,
|
| 15 |
+
"train_steps_per_second": 6.754
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.15121982702981798,
|
| 4 |
+
"eval_loss": 4.391414165496826,
|
| 5 |
+
"eval_runtime": 55.6914,
|
| 6 |
+
"eval_samples": 12892,
|
| 7 |
+
"eval_samples_per_second": 231.49,
|
| 8 |
+
"eval_steps_per_second": 28.945,
|
| 9 |
+
"perplexity": 80.75453855161017
|
| 10 |
+
}
|
runs/Feb22_21-13-07_gitspot/events.out.tfevents.1740258728.gitspot.1976043.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17d67f625c678e9967c33587a167db8eea9c2d9b98f3ef55b2713c2aa6456816
|
| 3 |
+
size 369
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"total_flos": 6.962502728150323e+16,
|
| 4 |
+
"train_loss": 4.240191396162094,
|
| 5 |
+
"train_runtime": 19577.9877,
|
| 6 |
+
"train_samples": 264467,
|
| 7 |
+
"train_samples_per_second": 54.034,
|
| 8 |
+
"train_steps_per_second": 6.754
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1890 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 5000.0,
|
| 6 |
+
"global_step": 132236,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.015124474424513748,
|
| 13 |
+
"grad_norm": 2.3324170112609863,
|
| 14 |
+
"learning_rate": 0.00015,
|
| 15 |
+
"loss": 5.1365,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.030248948849027495,
|
| 20 |
+
"grad_norm": 2.10058856010437,
|
| 21 |
+
"learning_rate": 0.0003,
|
| 22 |
+
"loss": 3.9899,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04537342327354124,
|
| 27 |
+
"grad_norm": 1.9626045227050781,
|
| 28 |
+
"learning_rate": 0.0002988570209393764,
|
| 29 |
+
"loss": 4.0268,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06049789769805499,
|
| 34 |
+
"grad_norm": 1.8885798454284668,
|
| 35 |
+
"learning_rate": 0.0002977140418787528,
|
| 36 |
+
"loss": 4.2117,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07562237212256874,
|
| 41 |
+
"grad_norm": 2.2193634510040283,
|
| 42 |
+
"learning_rate": 0.00029657106281812915,
|
| 43 |
+
"loss": 4.275,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09074684654708248,
|
| 48 |
+
"grad_norm": 1.4996557235717773,
|
| 49 |
+
"learning_rate": 0.0002954280837575055,
|
| 50 |
+
"loss": 4.2593,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.10587132097159624,
|
| 55 |
+
"grad_norm": 1.7600525617599487,
|
| 56 |
+
"learning_rate": 0.00029428510469688195,
|
| 57 |
+
"loss": 4.2528,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.12099579539610998,
|
| 62 |
+
"grad_norm": 1.3879656791687012,
|
| 63 |
+
"learning_rate": 0.0002931421256362583,
|
| 64 |
+
"loss": 4.2683,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.13612026982062372,
|
| 69 |
+
"grad_norm": 1.4735890626907349,
|
| 70 |
+
"learning_rate": 0.0002919991465756347,
|
| 71 |
+
"loss": 4.2656,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.15124474424513748,
|
| 76 |
+
"grad_norm": 1.8373522758483887,
|
| 77 |
+
"learning_rate": 0.00029085616751501107,
|
| 78 |
+
"loss": 4.2657,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.16636921866965124,
|
| 83 |
+
"grad_norm": 1.7759615182876587,
|
| 84 |
+
"learning_rate": 0.0002897131884543875,
|
| 85 |
+
"loss": 4.2791,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.18149369309416497,
|
| 90 |
+
"grad_norm": 0.915600061416626,
|
| 91 |
+
"learning_rate": 0.0002885702093937639,
|
| 92 |
+
"loss": 4.2707,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.19661816751867872,
|
| 97 |
+
"grad_norm": 0.9129588007926941,
|
| 98 |
+
"learning_rate": 0.0002874272303331403,
|
| 99 |
+
"loss": 4.2553,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.21174264194319248,
|
| 104 |
+
"grad_norm": 0.5569419264793396,
|
| 105 |
+
"learning_rate": 0.00028628425127251667,
|
| 106 |
+
"loss": 4.2349,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.22686711636770623,
|
| 111 |
+
"grad_norm": 0.3281975984573364,
|
| 112 |
+
"learning_rate": 0.00028514127221189304,
|
| 113 |
+
"loss": 4.2535,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.24199159079221996,
|
| 118 |
+
"grad_norm": 0.7752874493598938,
|
| 119 |
+
"learning_rate": 0.00028399829315126947,
|
| 120 |
+
"loss": 4.2396,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2571160652167337,
|
| 125 |
+
"grad_norm": 0.5675411820411682,
|
| 126 |
+
"learning_rate": 0.00028285531409064584,
|
| 127 |
+
"loss": 4.247,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.27224053964124745,
|
| 132 |
+
"grad_norm": 0.5231436491012573,
|
| 133 |
+
"learning_rate": 0.0002817123350300222,
|
| 134 |
+
"loss": 4.2569,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2873650140657612,
|
| 139 |
+
"grad_norm": 0.23958289623260498,
|
| 140 |
+
"learning_rate": 0.0002805693559693986,
|
| 141 |
+
"loss": 4.2254,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.30248948849027496,
|
| 146 |
+
"grad_norm": 0.919958233833313,
|
| 147 |
+
"learning_rate": 0.000279426376908775,
|
| 148 |
+
"loss": 4.2347,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3176139629147887,
|
| 153 |
+
"grad_norm": 0.8937901258468628,
|
| 154 |
+
"learning_rate": 0.0002782833978481514,
|
| 155 |
+
"loss": 4.2412,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3327384373393025,
|
| 160 |
+
"grad_norm": 0.6007381677627563,
|
| 161 |
+
"learning_rate": 0.00027714041878752776,
|
| 162 |
+
"loss": 4.251,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.34786291176381623,
|
| 167 |
+
"grad_norm": 0.2986898422241211,
|
| 168 |
+
"learning_rate": 0.0002759974397269042,
|
| 169 |
+
"loss": 4.2479,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.36298738618832993,
|
| 174 |
+
"grad_norm": 1.5830953121185303,
|
| 175 |
+
"learning_rate": 0.00027485446066628056,
|
| 176 |
+
"loss": 4.2368,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3781118606128437,
|
| 181 |
+
"grad_norm": 0.28822895884513855,
|
| 182 |
+
"learning_rate": 0.000273711481605657,
|
| 183 |
+
"loss": 4.2323,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.39323633503735744,
|
| 188 |
+
"grad_norm": 0.3698594570159912,
|
| 189 |
+
"learning_rate": 0.00027256850254503336,
|
| 190 |
+
"loss": 4.255,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4083608094618712,
|
| 195 |
+
"grad_norm": 0.6553381681442261,
|
| 196 |
+
"learning_rate": 0.00027142552348440973,
|
| 197 |
+
"loss": 4.2438,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.42348528388638496,
|
| 202 |
+
"grad_norm": 2.6835319995880127,
|
| 203 |
+
"learning_rate": 0.0002702825444237861,
|
| 204 |
+
"loss": 4.2409,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.4386097583108987,
|
| 209 |
+
"grad_norm": 0.6386218667030334,
|
| 210 |
+
"learning_rate": 0.00026913956536316254,
|
| 211 |
+
"loss": 4.2523,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.45373423273541247,
|
| 216 |
+
"grad_norm": 0.9448551535606384,
|
| 217 |
+
"learning_rate": 0.0002679965863025389,
|
| 218 |
+
"loss": 4.2352,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.46885870715992617,
|
| 223 |
+
"grad_norm": 1.78782057762146,
|
| 224 |
+
"learning_rate": 0.0002668536072419153,
|
| 225 |
+
"loss": 4.2642,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.4839831815844399,
|
| 230 |
+
"grad_norm": 1.835822343826294,
|
| 231 |
+
"learning_rate": 0.0002657106281812917,
|
| 232 |
+
"loss": 4.2466,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.4991076560089537,
|
| 237 |
+
"grad_norm": 0.4765397012233734,
|
| 238 |
+
"learning_rate": 0.0002645676491206681,
|
| 239 |
+
"loss": 4.2464,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5142321304334674,
|
| 244 |
+
"grad_norm": 1.7225979566574097,
|
| 245 |
+
"learning_rate": 0.0002634246700600445,
|
| 246 |
+
"loss": 4.2438,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5293566048579812,
|
| 251 |
+
"grad_norm": 0.19039462506771088,
|
| 252 |
+
"learning_rate": 0.0002622816909994209,
|
| 253 |
+
"loss": 4.2545,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5444810792824949,
|
| 258 |
+
"grad_norm": 0.16958552598953247,
|
| 259 |
+
"learning_rate": 0.00026113871193879726,
|
| 260 |
+
"loss": 4.239,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.5596055537070087,
|
| 265 |
+
"grad_norm": 0.49323347210884094,
|
| 266 |
+
"learning_rate": 0.00025999573287817363,
|
| 267 |
+
"loss": 4.2365,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.5747300281315224,
|
| 272 |
+
"grad_norm": 1.2180124521255493,
|
| 273 |
+
"learning_rate": 0.00025885275381755006,
|
| 274 |
+
"loss": 4.2318,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.5898545025560362,
|
| 279 |
+
"grad_norm": 3.211355686187744,
|
| 280 |
+
"learning_rate": 0.00025770977475692643,
|
| 281 |
+
"loss": 4.23,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6049789769805499,
|
| 286 |
+
"grad_norm": 1.5769411325454712,
|
| 287 |
+
"learning_rate": 0.0002565667956963028,
|
| 288 |
+
"loss": 4.2553,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6201034514050636,
|
| 293 |
+
"grad_norm": 1.2520736455917358,
|
| 294 |
+
"learning_rate": 0.00025542381663567923,
|
| 295 |
+
"loss": 4.2463,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.6352279258295774,
|
| 300 |
+
"grad_norm": 0.325623095035553,
|
| 301 |
+
"learning_rate": 0.0002542808375750556,
|
| 302 |
+
"loss": 4.2411,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.6503524002540911,
|
| 307 |
+
"grad_norm": 0.6519089341163635,
|
| 308 |
+
"learning_rate": 0.000253137858514432,
|
| 309 |
+
"loss": 4.2265,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.665476874678605,
|
| 314 |
+
"grad_norm": 0.885131299495697,
|
| 315 |
+
"learning_rate": 0.00025199487945380835,
|
| 316 |
+
"loss": 4.2495,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.6806013491031186,
|
| 321 |
+
"grad_norm": 1.6101298332214355,
|
| 322 |
+
"learning_rate": 0.0002508519003931848,
|
| 323 |
+
"loss": 4.2475,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.6957258235276325,
|
| 328 |
+
"grad_norm": 5.298095703125,
|
| 329 |
+
"learning_rate": 0.00024970892133256115,
|
| 330 |
+
"loss": 4.2468,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.7108502979521462,
|
| 335 |
+
"grad_norm": 0.4262215197086334,
|
| 336 |
+
"learning_rate": 0.0002485659422719376,
|
| 337 |
+
"loss": 4.2476,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.7259747723766599,
|
| 342 |
+
"grad_norm": 0.24104143679141998,
|
| 343 |
+
"learning_rate": 0.00024742296321131395,
|
| 344 |
+
"loss": 4.2513,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.7410992468011737,
|
| 349 |
+
"grad_norm": 3.1999754905700684,
|
| 350 |
+
"learning_rate": 0.0002462799841506903,
|
| 351 |
+
"loss": 4.2224,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.7562237212256874,
|
| 356 |
+
"grad_norm": 1.3713421821594238,
|
| 357 |
+
"learning_rate": 0.00024513700509006675,
|
| 358 |
+
"loss": 4.2398,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.7713481956502012,
|
| 363 |
+
"grad_norm": 0.3571838140487671,
|
| 364 |
+
"learning_rate": 0.00024399402602944312,
|
| 365 |
+
"loss": 4.2087,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.7864726700747149,
|
| 370 |
+
"grad_norm": 2.0582222938537598,
|
| 371 |
+
"learning_rate": 0.00024285104696881952,
|
| 372 |
+
"loss": 4.2377,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.8015971444992287,
|
| 377 |
+
"grad_norm": 1.5359830856323242,
|
| 378 |
+
"learning_rate": 0.0002417080679081959,
|
| 379 |
+
"loss": 4.2588,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.8167216189237424,
|
| 384 |
+
"grad_norm": 0.8071072101593018,
|
| 385 |
+
"learning_rate": 0.0002405650888475723,
|
| 386 |
+
"loss": 4.2467,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.8318460933482561,
|
| 391 |
+
"grad_norm": 3.864656686782837,
|
| 392 |
+
"learning_rate": 0.00023942210978694867,
|
| 393 |
+
"loss": 4.2408,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.8469705677727699,
|
| 398 |
+
"grad_norm": 1.8902547359466553,
|
| 399 |
+
"learning_rate": 0.00023827913072632507,
|
| 400 |
+
"loss": 4.2269,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.8620950421972836,
|
| 405 |
+
"grad_norm": 3.6469967365264893,
|
| 406 |
+
"learning_rate": 0.00023713615166570147,
|
| 407 |
+
"loss": 4.226,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.8772195166217974,
|
| 412 |
+
"grad_norm": 2.8256566524505615,
|
| 413 |
+
"learning_rate": 0.00023599317260507784,
|
| 414 |
+
"loss": 4.2356,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.8923439910463111,
|
| 419 |
+
"grad_norm": 1.8875362873077393,
|
| 420 |
+
"learning_rate": 0.00023485019354445427,
|
| 421 |
+
"loss": 4.2236,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.9074684654708249,
|
| 426 |
+
"grad_norm": 1.4305074214935303,
|
| 427 |
+
"learning_rate": 0.00023370721448383062,
|
| 428 |
+
"loss": 4.2319,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.9225929398953386,
|
| 433 |
+
"grad_norm": 1.5183324813842773,
|
| 434 |
+
"learning_rate": 0.00023256423542320704,
|
| 435 |
+
"loss": 4.2221,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.9377174143198523,
|
| 440 |
+
"grad_norm": 1.715067982673645,
|
| 441 |
+
"learning_rate": 0.00023142125636258342,
|
| 442 |
+
"loss": 4.22,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.9528418887443662,
|
| 447 |
+
"grad_norm": 1.6044344902038574,
|
| 448 |
+
"learning_rate": 0.00023027827730195982,
|
| 449 |
+
"loss": 4.222,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.9679663631688799,
|
| 454 |
+
"grad_norm": 2.6135776042938232,
|
| 455 |
+
"learning_rate": 0.0002291352982413362,
|
| 456 |
+
"loss": 4.2325,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.9830908375933937,
|
| 461 |
+
"grad_norm": 0.9000643491744995,
|
| 462 |
+
"learning_rate": 0.0002279923191807126,
|
| 463 |
+
"loss": 4.2443,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.9982153120179074,
|
| 468 |
+
"grad_norm": 1.7899664640426636,
|
| 469 |
+
"learning_rate": 0.000226849340120089,
|
| 470 |
+
"loss": 4.2452,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.0133397864424212,
|
| 475 |
+
"grad_norm": 1.7241235971450806,
|
| 476 |
+
"learning_rate": 0.00022570636105946536,
|
| 477 |
+
"loss": 4.2323,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.0284642608669348,
|
| 482 |
+
"grad_norm": 3.3212637901306152,
|
| 483 |
+
"learning_rate": 0.00022456338199884176,
|
| 484 |
+
"loss": 4.2419,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.0435887352914486,
|
| 489 |
+
"grad_norm": 5.221072673797607,
|
| 490 |
+
"learning_rate": 0.00022342040293821814,
|
| 491 |
+
"loss": 4.2402,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.0587132097159624,
|
| 496 |
+
"grad_norm": 1.0462692975997925,
|
| 497 |
+
"learning_rate": 0.00022227742387759456,
|
| 498 |
+
"loss": 4.2428,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.0738376841404762,
|
| 503 |
+
"grad_norm": 1.9516572952270508,
|
| 504 |
+
"learning_rate": 0.00022113444481697094,
|
| 505 |
+
"loss": 4.2547,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.0889621585649898,
|
| 510 |
+
"grad_norm": 4.056976318359375,
|
| 511 |
+
"learning_rate": 0.00021999146575634734,
|
| 512 |
+
"loss": 4.2332,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.1040866329895036,
|
| 517 |
+
"grad_norm": 0.6332369446754456,
|
| 518 |
+
"learning_rate": 0.0002188484866957237,
|
| 519 |
+
"loss": 4.2391,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.1192111074140174,
|
| 524 |
+
"grad_norm": 1.4901684522628784,
|
| 525 |
+
"learning_rate": 0.0002177055076351001,
|
| 526 |
+
"loss": 4.2439,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.1343355818385312,
|
| 531 |
+
"grad_norm": 1.8330302238464355,
|
| 532 |
+
"learning_rate": 0.0002165625285744765,
|
| 533 |
+
"loss": 4.2291,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.1494600562630448,
|
| 538 |
+
"grad_norm": 0.46577465534210205,
|
| 539 |
+
"learning_rate": 0.00021541954951385288,
|
| 540 |
+
"loss": 4.2441,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.1645845306875586,
|
| 545 |
+
"grad_norm": 0.9202355146408081,
|
| 546 |
+
"learning_rate": 0.00021427657045322928,
|
| 547 |
+
"loss": 4.239,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.1797090051120724,
|
| 552 |
+
"grad_norm": 1.7412337064743042,
|
| 553 |
+
"learning_rate": 0.00021313359139260566,
|
| 554 |
+
"loss": 4.2266,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.194833479536586,
|
| 559 |
+
"grad_norm": 2.4705281257629395,
|
| 560 |
+
"learning_rate": 0.00021199061233198206,
|
| 561 |
+
"loss": 4.2278,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.2099579539610998,
|
| 566 |
+
"grad_norm": 2.2124886512756348,
|
| 567 |
+
"learning_rate": 0.00021084763327135843,
|
| 568 |
+
"loss": 4.2315,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.2250824283856137,
|
| 573 |
+
"grad_norm": 2.529521942138672,
|
| 574 |
+
"learning_rate": 0.00020970465421073486,
|
| 575 |
+
"loss": 4.2197,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.2402069028101272,
|
| 580 |
+
"grad_norm": 1.8712756633758545,
|
| 581 |
+
"learning_rate": 0.00020856167515011123,
|
| 582 |
+
"loss": 4.2369,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.255331377234641,
|
| 587 |
+
"grad_norm": 2.6887283325195312,
|
| 588 |
+
"learning_rate": 0.00020741869608948763,
|
| 589 |
+
"loss": 4.2505,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.2704558516591549,
|
| 594 |
+
"grad_norm": 3.8678267002105713,
|
| 595 |
+
"learning_rate": 0.00020627571702886403,
|
| 596 |
+
"loss": 4.2362,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.2855803260836685,
|
| 601 |
+
"grad_norm": 2.4472031593322754,
|
| 602 |
+
"learning_rate": 0.0002051327379682404,
|
| 603 |
+
"loss": 4.2348,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.3007048005081823,
|
| 608 |
+
"grad_norm": 1.4369362592697144,
|
| 609 |
+
"learning_rate": 0.0002039897589076168,
|
| 610 |
+
"loss": 4.2266,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.315829274932696,
|
| 615 |
+
"grad_norm": 1.2380897998809814,
|
| 616 |
+
"learning_rate": 0.00020284677984699318,
|
| 617 |
+
"loss": 4.252,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.33095374935721,
|
| 622 |
+
"grad_norm": 2.7404069900512695,
|
| 623 |
+
"learning_rate": 0.00020170380078636958,
|
| 624 |
+
"loss": 4.2388,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.3460782237817237,
|
| 629 |
+
"grad_norm": 1.010291576385498,
|
| 630 |
+
"learning_rate": 0.00020056082172574595,
|
| 631 |
+
"loss": 4.2341,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.3612026982062373,
|
| 636 |
+
"grad_norm": 3.9233856201171875,
|
| 637 |
+
"learning_rate": 0.00019941784266512235,
|
| 638 |
+
"loss": 4.2447,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.376327172630751,
|
| 643 |
+
"grad_norm": 0.4064319431781769,
|
| 644 |
+
"learning_rate": 0.00019827486360449872,
|
| 645 |
+
"loss": 4.2319,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.391451647055265,
|
| 650 |
+
"grad_norm": 2.5723445415496826,
|
| 651 |
+
"learning_rate": 0.00019713188454387515,
|
| 652 |
+
"loss": 4.2262,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.4065761214797785,
|
| 657 |
+
"grad_norm": 10.290382385253906,
|
| 658 |
+
"learning_rate": 0.00019598890548325155,
|
| 659 |
+
"loss": 4.2245,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.4217005959042923,
|
| 664 |
+
"grad_norm": 1.3730300664901733,
|
| 665 |
+
"learning_rate": 0.00019484592642262792,
|
| 666 |
+
"loss": 4.2528,
|
| 667 |
+
"step": 47000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.4368250703288061,
|
| 671 |
+
"grad_norm": 4.968954563140869,
|
| 672 |
+
"learning_rate": 0.00019370294736200432,
|
| 673 |
+
"loss": 4.2352,
|
| 674 |
+
"step": 47500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.4519495447533197,
|
| 678 |
+
"grad_norm": 4.111246585845947,
|
| 679 |
+
"learning_rate": 0.0001925599683013807,
|
| 680 |
+
"loss": 4.2399,
|
| 681 |
+
"step": 48000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.4670740191778335,
|
| 685 |
+
"grad_norm": 1.1323810815811157,
|
| 686 |
+
"learning_rate": 0.0001914169892407571,
|
| 687 |
+
"loss": 4.2593,
|
| 688 |
+
"step": 48500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.4821984936023473,
|
| 692 |
+
"grad_norm": 0.8210377097129822,
|
| 693 |
+
"learning_rate": 0.00019027401018013347,
|
| 694 |
+
"loss": 4.2292,
|
| 695 |
+
"step": 49000
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.497322968026861,
|
| 699 |
+
"grad_norm": 1.7300924062728882,
|
| 700 |
+
"learning_rate": 0.00018913103111950987,
|
| 701 |
+
"loss": 4.2337,
|
| 702 |
+
"step": 49500
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.512447442451375,
|
| 706 |
+
"grad_norm": 1.0358113050460815,
|
| 707 |
+
"learning_rate": 0.00018798805205888624,
|
| 708 |
+
"loss": 4.2362,
|
| 709 |
+
"step": 50000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.5275719168758886,
|
| 713 |
+
"grad_norm": 3.6714601516723633,
|
| 714 |
+
"learning_rate": 0.00018684507299826264,
|
| 715 |
+
"loss": 4.2454,
|
| 716 |
+
"step": 50500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.5426963913004021,
|
| 720 |
+
"grad_norm": 5.806258201599121,
|
| 721 |
+
"learning_rate": 0.00018570209393763907,
|
| 722 |
+
"loss": 4.2293,
|
| 723 |
+
"step": 51000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.5578208657249162,
|
| 727 |
+
"grad_norm": 1.5212393999099731,
|
| 728 |
+
"learning_rate": 0.00018455911487701544,
|
| 729 |
+
"loss": 4.2286,
|
| 730 |
+
"step": 51500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.5729453401494298,
|
| 734 |
+
"grad_norm": 2.4763267040252686,
|
| 735 |
+
"learning_rate": 0.00018341613581639185,
|
| 736 |
+
"loss": 4.2263,
|
| 737 |
+
"step": 52000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.5880698145739436,
|
| 741 |
+
"grad_norm": 3.5433552265167236,
|
| 742 |
+
"learning_rate": 0.00018227315675576822,
|
| 743 |
+
"loss": 4.2369,
|
| 744 |
+
"step": 52500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.6031942889984574,
|
| 748 |
+
"grad_norm": 2.173997402191162,
|
| 749 |
+
"learning_rate": 0.00018113017769514462,
|
| 750 |
+
"loss": 4.2303,
|
| 751 |
+
"step": 53000
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.618318763422971,
|
| 755 |
+
"grad_norm": 1.0033237934112549,
|
| 756 |
+
"learning_rate": 0.000179987198634521,
|
| 757 |
+
"loss": 4.2287,
|
| 758 |
+
"step": 53500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.6334432378474848,
|
| 762 |
+
"grad_norm": 2.0896666049957275,
|
| 763 |
+
"learning_rate": 0.0001788442195738974,
|
| 764 |
+
"loss": 4.2314,
|
| 765 |
+
"step": 54000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.6485677122719986,
|
| 769 |
+
"grad_norm": 4.681240558624268,
|
| 770 |
+
"learning_rate": 0.00017770124051327376,
|
| 771 |
+
"loss": 4.2317,
|
| 772 |
+
"step": 54500
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.6636921866965122,
|
| 776 |
+
"grad_norm": 0.8361565470695496,
|
| 777 |
+
"learning_rate": 0.00017655826145265016,
|
| 778 |
+
"loss": 4.2267,
|
| 779 |
+
"step": 55000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.678816661121026,
|
| 783 |
+
"grad_norm": 1.8475239276885986,
|
| 784 |
+
"learning_rate": 0.0001754152823920266,
|
| 785 |
+
"loss": 4.2515,
|
| 786 |
+
"step": 55500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.6939411355455398,
|
| 790 |
+
"grad_norm": 1.012220025062561,
|
| 791 |
+
"learning_rate": 0.00017427230333140294,
|
| 792 |
+
"loss": 4.2489,
|
| 793 |
+
"step": 56000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.7090656099700534,
|
| 797 |
+
"grad_norm": 2.960779905319214,
|
| 798 |
+
"learning_rate": 0.00017312932427077937,
|
| 799 |
+
"loss": 4.2336,
|
| 800 |
+
"step": 56500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.7241900843945674,
|
| 804 |
+
"grad_norm": 0.4395509362220764,
|
| 805 |
+
"learning_rate": 0.00017198634521015574,
|
| 806 |
+
"loss": 4.2275,
|
| 807 |
+
"step": 57000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.739314558819081,
|
| 811 |
+
"grad_norm": 3.430058479309082,
|
| 812 |
+
"learning_rate": 0.00017084336614953214,
|
| 813 |
+
"loss": 4.2436,
|
| 814 |
+
"step": 57500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.7544390332435946,
|
| 818 |
+
"grad_norm": 1.85948646068573,
|
| 819 |
+
"learning_rate": 0.0001697003870889085,
|
| 820 |
+
"loss": 4.2245,
|
| 821 |
+
"step": 58000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.7695635076681087,
|
| 825 |
+
"grad_norm": 2.0236196517944336,
|
| 826 |
+
"learning_rate": 0.0001685574080282849,
|
| 827 |
+
"loss": 4.2429,
|
| 828 |
+
"step": 58500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.7846879820926222,
|
| 832 |
+
"grad_norm": 2.449800491333008,
|
| 833 |
+
"learning_rate": 0.00016741442896766129,
|
| 834 |
+
"loss": 4.2419,
|
| 835 |
+
"step": 59000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.799812456517136,
|
| 839 |
+
"grad_norm": 2.323369026184082,
|
| 840 |
+
"learning_rate": 0.00016627144990703769,
|
| 841 |
+
"loss": 4.2273,
|
| 842 |
+
"step": 59500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.8149369309416499,
|
| 846 |
+
"grad_norm": 3.488776445388794,
|
| 847 |
+
"learning_rate": 0.00016512847084641406,
|
| 848 |
+
"loss": 4.2253,
|
| 849 |
+
"step": 60000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.8300614053661635,
|
| 853 |
+
"grad_norm": 1.216118335723877,
|
| 854 |
+
"learning_rate": 0.00016398549178579046,
|
| 855 |
+
"loss": 4.223,
|
| 856 |
+
"step": 60500
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.8451858797906773,
|
| 860 |
+
"grad_norm": 1.0968749523162842,
|
| 861 |
+
"learning_rate": 0.00016284251272516689,
|
| 862 |
+
"loss": 4.2317,
|
| 863 |
+
"step": 61000
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.860310354215191,
|
| 867 |
+
"grad_norm": 0.250358909368515,
|
| 868 |
+
"learning_rate": 0.00016169953366454326,
|
| 869 |
+
"loss": 4.2296,
|
| 870 |
+
"step": 61500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.8754348286397047,
|
| 874 |
+
"grad_norm": 0.9802926778793335,
|
| 875 |
+
"learning_rate": 0.00016055655460391966,
|
| 876 |
+
"loss": 4.2458,
|
| 877 |
+
"step": 62000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.8905593030642185,
|
| 881 |
+
"grad_norm": 2.8225202560424805,
|
| 882 |
+
"learning_rate": 0.00015941357554329603,
|
| 883 |
+
"loss": 4.2419,
|
| 884 |
+
"step": 62500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.9056837774887323,
|
| 888 |
+
"grad_norm": 1.5534439086914062,
|
| 889 |
+
"learning_rate": 0.00015827059648267243,
|
| 890 |
+
"loss": 4.2373,
|
| 891 |
+
"step": 63000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.920808251913246,
|
| 895 |
+
"grad_norm": 2.5185372829437256,
|
| 896 |
+
"learning_rate": 0.0001571276174220488,
|
| 897 |
+
"loss": 4.2209,
|
| 898 |
+
"step": 63500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.93593272633776,
|
| 902 |
+
"grad_norm": 2.2351784706115723,
|
| 903 |
+
"learning_rate": 0.0001559846383614252,
|
| 904 |
+
"loss": 4.2387,
|
| 905 |
+
"step": 64000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.9510572007622735,
|
| 909 |
+
"grad_norm": 2.1786580085754395,
|
| 910 |
+
"learning_rate": 0.00015484165930080158,
|
| 911 |
+
"loss": 4.2345,
|
| 912 |
+
"step": 64500
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.966181675186787,
|
| 916 |
+
"grad_norm": 0.5652241110801697,
|
| 917 |
+
"learning_rate": 0.00015369868024017798,
|
| 918 |
+
"loss": 4.2231,
|
| 919 |
+
"step": 65000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.9813061496113011,
|
| 923 |
+
"grad_norm": 1.5961374044418335,
|
| 924 |
+
"learning_rate": 0.00015255570117955438,
|
| 925 |
+
"loss": 4.2318,
|
| 926 |
+
"step": 65500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.9964306240358147,
|
| 930 |
+
"grad_norm": 4.318066120147705,
|
| 931 |
+
"learning_rate": 0.00015141272211893075,
|
| 932 |
+
"loss": 4.2304,
|
| 933 |
+
"step": 66000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.0115550984603283,
|
| 937 |
+
"grad_norm": 0.36664879322052,
|
| 938 |
+
"learning_rate": 0.00015026974305830718,
|
| 939 |
+
"loss": 4.2285,
|
| 940 |
+
"step": 66500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.0266795728848424,
|
| 944 |
+
"grad_norm": 3.6058526039123535,
|
| 945 |
+
"learning_rate": 0.00014912676399768355,
|
| 946 |
+
"loss": 4.241,
|
| 947 |
+
"step": 67000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.041804047309356,
|
| 951 |
+
"grad_norm": 3.449469566345215,
|
| 952 |
+
"learning_rate": 0.00014798378493705995,
|
| 953 |
+
"loss": 4.2419,
|
| 954 |
+
"step": 67500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.0569285217338695,
|
| 958 |
+
"grad_norm": 3.3256688117980957,
|
| 959 |
+
"learning_rate": 0.00014684080587643633,
|
| 960 |
+
"loss": 4.2448,
|
| 961 |
+
"step": 68000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.0720529961583836,
|
| 965 |
+
"grad_norm": 0.7478310465812683,
|
| 966 |
+
"learning_rate": 0.00014569782681581273,
|
| 967 |
+
"loss": 4.2331,
|
| 968 |
+
"step": 68500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.087177470582897,
|
| 972 |
+
"grad_norm": 0.9061185717582703,
|
| 973 |
+
"learning_rate": 0.0001445548477551891,
|
| 974 |
+
"loss": 4.2386,
|
| 975 |
+
"step": 69000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.102301945007411,
|
| 979 |
+
"grad_norm": 0.9583618640899658,
|
| 980 |
+
"learning_rate": 0.0001434118686945655,
|
| 981 |
+
"loss": 4.2457,
|
| 982 |
+
"step": 69500
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.1174264194319248,
|
| 986 |
+
"grad_norm": 2.831644058227539,
|
| 987 |
+
"learning_rate": 0.0001422688896339419,
|
| 988 |
+
"loss": 4.2299,
|
| 989 |
+
"step": 70000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.1325508938564384,
|
| 993 |
+
"grad_norm": 2.579869031906128,
|
| 994 |
+
"learning_rate": 0.00014112591057331827,
|
| 995 |
+
"loss": 4.2449,
|
| 996 |
+
"step": 70500
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.1476753682809524,
|
| 1000 |
+
"grad_norm": 1.278503179550171,
|
| 1001 |
+
"learning_rate": 0.00013998293151269467,
|
| 1002 |
+
"loss": 4.2469,
|
| 1003 |
+
"step": 71000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.162799842705466,
|
| 1007 |
+
"grad_norm": 2.678473711013794,
|
| 1008 |
+
"learning_rate": 0.00013883995245207107,
|
| 1009 |
+
"loss": 4.2255,
|
| 1010 |
+
"step": 71500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.1779243171299796,
|
| 1014 |
+
"grad_norm": 0.805012583732605,
|
| 1015 |
+
"learning_rate": 0.00013769697339144747,
|
| 1016 |
+
"loss": 4.2419,
|
| 1017 |
+
"step": 72000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.1930487915544936,
|
| 1021 |
+
"grad_norm": 0.4923524856567383,
|
| 1022 |
+
"learning_rate": 0.00013655399433082385,
|
| 1023 |
+
"loss": 4.248,
|
| 1024 |
+
"step": 72500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.208173265979007,
|
| 1028 |
+
"grad_norm": 0.9246107339859009,
|
| 1029 |
+
"learning_rate": 0.00013541101527020025,
|
| 1030 |
+
"loss": 4.241,
|
| 1031 |
+
"step": 73000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.223297740403521,
|
| 1035 |
+
"grad_norm": 2.786689519882202,
|
| 1036 |
+
"learning_rate": 0.00013426803620957662,
|
| 1037 |
+
"loss": 4.2293,
|
| 1038 |
+
"step": 73500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.238422214828035,
|
| 1042 |
+
"grad_norm": 1.1377443075180054,
|
| 1043 |
+
"learning_rate": 0.00013312505714895302,
|
| 1044 |
+
"loss": 4.2338,
|
| 1045 |
+
"step": 74000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.2535466892525484,
|
| 1049 |
+
"grad_norm": 1.9654011726379395,
|
| 1050 |
+
"learning_rate": 0.00013198207808832942,
|
| 1051 |
+
"loss": 4.2252,
|
| 1052 |
+
"step": 74500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.2686711636770625,
|
| 1056 |
+
"grad_norm": 3.373481035232544,
|
| 1057 |
+
"learning_rate": 0.0001308390990277058,
|
| 1058 |
+
"loss": 4.2329,
|
| 1059 |
+
"step": 75000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.283795638101576,
|
| 1063 |
+
"grad_norm": 3.874128818511963,
|
| 1064 |
+
"learning_rate": 0.0001296961199670822,
|
| 1065 |
+
"loss": 4.2216,
|
| 1066 |
+
"step": 75500
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.2989201125260896,
|
| 1070 |
+
"grad_norm": 0.6507995128631592,
|
| 1071 |
+
"learning_rate": 0.0001285531409064586,
|
| 1072 |
+
"loss": 4.2456,
|
| 1073 |
+
"step": 76000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.3140445869506037,
|
| 1077 |
+
"grad_norm": 0.5525983572006226,
|
| 1078 |
+
"learning_rate": 0.00012741016184583497,
|
| 1079 |
+
"loss": 4.2202,
|
| 1080 |
+
"step": 76500
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.3291690613751173,
|
| 1084 |
+
"grad_norm": 1.062055230140686,
|
| 1085 |
+
"learning_rate": 0.00012626718278521137,
|
| 1086 |
+
"loss": 4.2504,
|
| 1087 |
+
"step": 77000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.344293535799631,
|
| 1091 |
+
"grad_norm": 2.6598470211029053,
|
| 1092 |
+
"learning_rate": 0.00012512420372458777,
|
| 1093 |
+
"loss": 4.2382,
|
| 1094 |
+
"step": 77500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.359418010224145,
|
| 1098 |
+
"grad_norm": 8.42147159576416,
|
| 1099 |
+
"learning_rate": 0.00012398122466396414,
|
| 1100 |
+
"loss": 4.2427,
|
| 1101 |
+
"step": 78000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.3745424846486585,
|
| 1105 |
+
"grad_norm": 0.7806519865989685,
|
| 1106 |
+
"learning_rate": 0.00012283824560334054,
|
| 1107 |
+
"loss": 4.2387,
|
| 1108 |
+
"step": 78500
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.389666959073172,
|
| 1112 |
+
"grad_norm": 1.3354737758636475,
|
| 1113 |
+
"learning_rate": 0.00012169526654271693,
|
| 1114 |
+
"loss": 4.2482,
|
| 1115 |
+
"step": 79000
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.404791433497686,
|
| 1119 |
+
"grad_norm": 1.9078588485717773,
|
| 1120 |
+
"learning_rate": 0.00012055228748209331,
|
| 1121 |
+
"loss": 4.2273,
|
| 1122 |
+
"step": 79500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.4199159079221997,
|
| 1126 |
+
"grad_norm": 5.637246608734131,
|
| 1127 |
+
"learning_rate": 0.0001194093084214697,
|
| 1128 |
+
"loss": 4.2516,
|
| 1129 |
+
"step": 80000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.4350403823467133,
|
| 1133 |
+
"grad_norm": 6.461385250091553,
|
| 1134 |
+
"learning_rate": 0.0001182663293608461,
|
| 1135 |
+
"loss": 4.2398,
|
| 1136 |
+
"step": 80500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.4501648567712273,
|
| 1140 |
+
"grad_norm": 0.5900661945343018,
|
| 1141 |
+
"learning_rate": 0.0001171233503002225,
|
| 1142 |
+
"loss": 4.2323,
|
| 1143 |
+
"step": 81000
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.465289331195741,
|
| 1147 |
+
"grad_norm": 2.4329442977905273,
|
| 1148 |
+
"learning_rate": 0.00011598037123959889,
|
| 1149 |
+
"loss": 4.2464,
|
| 1150 |
+
"step": 81500
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.4804138056202545,
|
| 1154 |
+
"grad_norm": 1.6276785135269165,
|
| 1155 |
+
"learning_rate": 0.00011483739217897527,
|
| 1156 |
+
"loss": 4.226,
|
| 1157 |
+
"step": 82000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.4955382800447685,
|
| 1161 |
+
"grad_norm": 2.940244436264038,
|
| 1162 |
+
"learning_rate": 0.00011369441311835166,
|
| 1163 |
+
"loss": 4.229,
|
| 1164 |
+
"step": 82500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.510662754469282,
|
| 1168 |
+
"grad_norm": 1.7740508317947388,
|
| 1169 |
+
"learning_rate": 0.00011255143405772805,
|
| 1170 |
+
"loss": 4.2534,
|
| 1171 |
+
"step": 83000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.5257872288937957,
|
| 1175 |
+
"grad_norm": 1.6791937351226807,
|
| 1176 |
+
"learning_rate": 0.00011140845499710443,
|
| 1177 |
+
"loss": 4.2287,
|
| 1178 |
+
"step": 83500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.5409117033183097,
|
| 1182 |
+
"grad_norm": 1.4454323053359985,
|
| 1183 |
+
"learning_rate": 0.00011026547593648083,
|
| 1184 |
+
"loss": 4.2301,
|
| 1185 |
+
"step": 84000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.5560361777428233,
|
| 1189 |
+
"grad_norm": 5.167677402496338,
|
| 1190 |
+
"learning_rate": 0.00010912249687585722,
|
| 1191 |
+
"loss": 4.2514,
|
| 1192 |
+
"step": 84500
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.571160652167337,
|
| 1196 |
+
"grad_norm": 2.0292153358459473,
|
| 1197 |
+
"learning_rate": 0.00010797951781523362,
|
| 1198 |
+
"loss": 4.234,
|
| 1199 |
+
"step": 85000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.586285126591851,
|
| 1203 |
+
"grad_norm": 0.5824300646781921,
|
| 1204 |
+
"learning_rate": 0.00010683653875461001,
|
| 1205 |
+
"loss": 4.2346,
|
| 1206 |
+
"step": 85500
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.6014096010163645,
|
| 1210 |
+
"grad_norm": 0.37599918246269226,
|
| 1211 |
+
"learning_rate": 0.00010569355969398641,
|
| 1212 |
+
"loss": 4.2406,
|
| 1213 |
+
"step": 86000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.6165340754408786,
|
| 1217 |
+
"grad_norm": 1.8334258794784546,
|
| 1218 |
+
"learning_rate": 0.0001045505806333628,
|
| 1219 |
+
"loss": 4.2246,
|
| 1220 |
+
"step": 86500
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.631658549865392,
|
| 1224 |
+
"grad_norm": 2.094484329223633,
|
| 1225 |
+
"learning_rate": 0.00010340760157273918,
|
| 1226 |
+
"loss": 4.227,
|
| 1227 |
+
"step": 87000
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.646783024289906,
|
| 1231 |
+
"grad_norm": 0.40919598937034607,
|
| 1232 |
+
"learning_rate": 0.00010226462251211557,
|
| 1233 |
+
"loss": 4.2397,
|
| 1234 |
+
"step": 87500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.66190749871442,
|
| 1238 |
+
"grad_norm": 0.6805312037467957,
|
| 1239 |
+
"learning_rate": 0.00010112164345149195,
|
| 1240 |
+
"loss": 4.2492,
|
| 1241 |
+
"step": 88000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.6770319731389334,
|
| 1245 |
+
"grad_norm": 0.4282049834728241,
|
| 1246 |
+
"learning_rate": 9.997866439086834e-05,
|
| 1247 |
+
"loss": 4.2382,
|
| 1248 |
+
"step": 88500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.6921564475634474,
|
| 1252 |
+
"grad_norm": 0.6906024217605591,
|
| 1253 |
+
"learning_rate": 9.883568533024474e-05,
|
| 1254 |
+
"loss": 4.2409,
|
| 1255 |
+
"step": 89000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.707280921987961,
|
| 1259 |
+
"grad_norm": 1.0855090618133545,
|
| 1260 |
+
"learning_rate": 9.769270626962113e-05,
|
| 1261 |
+
"loss": 4.2297,
|
| 1262 |
+
"step": 89500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.7224053964124746,
|
| 1266 |
+
"grad_norm": 2.2810654640197754,
|
| 1267 |
+
"learning_rate": 9.654972720899753e-05,
|
| 1268 |
+
"loss": 4.2333,
|
| 1269 |
+
"step": 90000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.7375298708369886,
|
| 1273 |
+
"grad_norm": 2.0734617710113525,
|
| 1274 |
+
"learning_rate": 9.540674814837391e-05,
|
| 1275 |
+
"loss": 4.2284,
|
| 1276 |
+
"step": 90500
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.752654345261502,
|
| 1280 |
+
"grad_norm": 3.4077634811401367,
|
| 1281 |
+
"learning_rate": 9.42637690877503e-05,
|
| 1282 |
+
"loss": 4.2075,
|
| 1283 |
+
"step": 91000
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.767778819686016,
|
| 1287 |
+
"grad_norm": 2.311880588531494,
|
| 1288 |
+
"learning_rate": 9.31207900271267e-05,
|
| 1289 |
+
"loss": 4.2438,
|
| 1290 |
+
"step": 91500
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.78290329411053,
|
| 1294 |
+
"grad_norm": 1.2504137754440308,
|
| 1295 |
+
"learning_rate": 9.197781096650309e-05,
|
| 1296 |
+
"loss": 4.2429,
|
| 1297 |
+
"step": 92000
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.7980277685350434,
|
| 1301 |
+
"grad_norm": 3.5147488117218018,
|
| 1302 |
+
"learning_rate": 9.083483190587947e-05,
|
| 1303 |
+
"loss": 4.2518,
|
| 1304 |
+
"step": 92500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.813152242959557,
|
| 1308 |
+
"grad_norm": 0.31045272946357727,
|
| 1309 |
+
"learning_rate": 8.969185284525586e-05,
|
| 1310 |
+
"loss": 4.2343,
|
| 1311 |
+
"step": 93000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.828276717384071,
|
| 1315 |
+
"grad_norm": 0.6451097726821899,
|
| 1316 |
+
"learning_rate": 8.854887378463225e-05,
|
| 1317 |
+
"loss": 4.2367,
|
| 1318 |
+
"step": 93500
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.8434011918085846,
|
| 1322 |
+
"grad_norm": 1.0421180725097656,
|
| 1323 |
+
"learning_rate": 8.740589472400863e-05,
|
| 1324 |
+
"loss": 4.2477,
|
| 1325 |
+
"step": 94000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.8585256662330982,
|
| 1329 |
+
"grad_norm": 3.758476734161377,
|
| 1330 |
+
"learning_rate": 8.626291566338505e-05,
|
| 1331 |
+
"loss": 4.2592,
|
| 1332 |
+
"step": 94500
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.8736501406576123,
|
| 1336 |
+
"grad_norm": 1.188808798789978,
|
| 1337 |
+
"learning_rate": 8.511993660276144e-05,
|
| 1338 |
+
"loss": 4.229,
|
| 1339 |
+
"step": 95000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.888774615082126,
|
| 1343 |
+
"grad_norm": 0.6678575277328491,
|
| 1344 |
+
"learning_rate": 8.397695754213782e-05,
|
| 1345 |
+
"loss": 4.232,
|
| 1346 |
+
"step": 95500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.9038990895066394,
|
| 1350 |
+
"grad_norm": 0.5926124453544617,
|
| 1351 |
+
"learning_rate": 8.283397848151421e-05,
|
| 1352 |
+
"loss": 4.2447,
|
| 1353 |
+
"step": 96000
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.9190235639311535,
|
| 1357 |
+
"grad_norm": 2.189749002456665,
|
| 1358 |
+
"learning_rate": 8.16909994208906e-05,
|
| 1359 |
+
"loss": 4.2276,
|
| 1360 |
+
"step": 96500
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.934148038355667,
|
| 1364 |
+
"grad_norm": 1.5343018770217896,
|
| 1365 |
+
"learning_rate": 8.0548020360267e-05,
|
| 1366 |
+
"loss": 4.2479,
|
| 1367 |
+
"step": 97000
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.9492725127801807,
|
| 1371 |
+
"grad_norm": 1.3845748901367188,
|
| 1372 |
+
"learning_rate": 7.940504129964338e-05,
|
| 1373 |
+
"loss": 4.2189,
|
| 1374 |
+
"step": 97500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.9643969872046947,
|
| 1378 |
+
"grad_norm": 1.2643513679504395,
|
| 1379 |
+
"learning_rate": 7.826206223901977e-05,
|
| 1380 |
+
"loss": 4.239,
|
| 1381 |
+
"step": 98000
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.9795214616292083,
|
| 1385 |
+
"grad_norm": 0.6748919486999512,
|
| 1386 |
+
"learning_rate": 7.711908317839616e-05,
|
| 1387 |
+
"loss": 4.2356,
|
| 1388 |
+
"step": 98500
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.994645936053722,
|
| 1392 |
+
"grad_norm": 2.6603946685791016,
|
| 1393 |
+
"learning_rate": 7.597610411777257e-05,
|
| 1394 |
+
"loss": 4.2306,
|
| 1395 |
+
"step": 99000
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.009770410478236,
|
| 1399 |
+
"grad_norm": 3.638791799545288,
|
| 1400 |
+
"learning_rate": 7.483312505714894e-05,
|
| 1401 |
+
"loss": 4.2389,
|
| 1402 |
+
"step": 99500
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 3.0248948849027495,
|
| 1406 |
+
"grad_norm": 5.572286128997803,
|
| 1407 |
+
"learning_rate": 7.369014599652534e-05,
|
| 1408 |
+
"loss": 4.2362,
|
| 1409 |
+
"step": 100000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.0400193593272635,
|
| 1413 |
+
"grad_norm": 1.5848674774169922,
|
| 1414 |
+
"learning_rate": 7.254716693590173e-05,
|
| 1415 |
+
"loss": 4.2477,
|
| 1416 |
+
"step": 100500
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.055143833751777,
|
| 1420 |
+
"grad_norm": 2.54856014251709,
|
| 1421 |
+
"learning_rate": 7.140418787527812e-05,
|
| 1422 |
+
"loss": 4.2418,
|
| 1423 |
+
"step": 101000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 3.0702683081762907,
|
| 1427 |
+
"grad_norm": 1.2424782514572144,
|
| 1428 |
+
"learning_rate": 7.02612088146545e-05,
|
| 1429 |
+
"loss": 4.2551,
|
| 1430 |
+
"step": 101500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.0853927826008047,
|
| 1434 |
+
"grad_norm": 4.676440715789795,
|
| 1435 |
+
"learning_rate": 6.91182297540309e-05,
|
| 1436 |
+
"loss": 4.2418,
|
| 1437 |
+
"step": 102000
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 3.1005172570253183,
|
| 1441 |
+
"grad_norm": 0.246148020029068,
|
| 1442 |
+
"learning_rate": 6.797525069340729e-05,
|
| 1443 |
+
"loss": 4.2604,
|
| 1444 |
+
"step": 102500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 3.115641731449832,
|
| 1448 |
+
"grad_norm": 0.8470159769058228,
|
| 1449 |
+
"learning_rate": 6.683227163278368e-05,
|
| 1450 |
+
"loss": 4.2266,
|
| 1451 |
+
"step": 103000
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.130766205874346,
|
| 1455 |
+
"grad_norm": 3.184859037399292,
|
| 1456 |
+
"learning_rate": 6.568929257216008e-05,
|
| 1457 |
+
"loss": 4.2668,
|
| 1458 |
+
"step": 103500
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 3.1458906802988595,
|
| 1462 |
+
"grad_norm": 3.391896963119507,
|
| 1463 |
+
"learning_rate": 6.454631351153646e-05,
|
| 1464 |
+
"loss": 4.243,
|
| 1465 |
+
"step": 104000
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 3.1610151547233736,
|
| 1469 |
+
"grad_norm": 4.900694847106934,
|
| 1470 |
+
"learning_rate": 6.340333445091286e-05,
|
| 1471 |
+
"loss": 4.2495,
|
| 1472 |
+
"step": 104500
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.176139629147887,
|
| 1476 |
+
"grad_norm": 2.035632610321045,
|
| 1477 |
+
"learning_rate": 6.226035539028925e-05,
|
| 1478 |
+
"loss": 4.2393,
|
| 1479 |
+
"step": 105000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 3.1912641035724008,
|
| 1483 |
+
"grad_norm": 0.5731152892112732,
|
| 1484 |
+
"learning_rate": 6.111737632966564e-05,
|
| 1485 |
+
"loss": 4.2261,
|
| 1486 |
+
"step": 105500
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 3.206388577996915,
|
| 1490 |
+
"grad_norm": 1.0584532022476196,
|
| 1491 |
+
"learning_rate": 5.997439726904202e-05,
|
| 1492 |
+
"loss": 4.2219,
|
| 1493 |
+
"step": 106000
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 3.2215130524214284,
|
| 1497 |
+
"grad_norm": 0.9379057288169861,
|
| 1498 |
+
"learning_rate": 5.883141820841841e-05,
|
| 1499 |
+
"loss": 4.2371,
|
| 1500 |
+
"step": 106500
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 3.236637526845942,
|
| 1504 |
+
"grad_norm": 1.500435471534729,
|
| 1505 |
+
"learning_rate": 5.768843914779481e-05,
|
| 1506 |
+
"loss": 4.2357,
|
| 1507 |
+
"step": 107000
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 3.251762001270456,
|
| 1511 |
+
"grad_norm": 2.035886526107788,
|
| 1512 |
+
"learning_rate": 5.6545460087171196e-05,
|
| 1513 |
+
"loss": 4.2477,
|
| 1514 |
+
"step": 107500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 3.2668864756949696,
|
| 1518 |
+
"grad_norm": 1.6449440717697144,
|
| 1519 |
+
"learning_rate": 5.540248102654759e-05,
|
| 1520 |
+
"loss": 4.2428,
|
| 1521 |
+
"step": 108000
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 3.282010950119483,
|
| 1525 |
+
"grad_norm": 0.8462217450141907,
|
| 1526 |
+
"learning_rate": 5.4259501965923976e-05,
|
| 1527 |
+
"loss": 4.2521,
|
| 1528 |
+
"step": 108500
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 3.297135424543997,
|
| 1532 |
+
"grad_norm": 2.7994236946105957,
|
| 1533 |
+
"learning_rate": 5.3116522905300376e-05,
|
| 1534 |
+
"loss": 4.2423,
|
| 1535 |
+
"step": 109000
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 3.312259898968511,
|
| 1539 |
+
"grad_norm": 4.005298614501953,
|
| 1540 |
+
"learning_rate": 5.197354384467676e-05,
|
| 1541 |
+
"loss": 4.2385,
|
| 1542 |
+
"step": 109500
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 3.3273843733930244,
|
| 1546 |
+
"grad_norm": 2.996264696121216,
|
| 1547 |
+
"learning_rate": 5.083056478405315e-05,
|
| 1548 |
+
"loss": 4.2422,
|
| 1549 |
+
"step": 110000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 3.3425088478175384,
|
| 1553 |
+
"grad_norm": 3.2069146633148193,
|
| 1554 |
+
"learning_rate": 4.968758572342954e-05,
|
| 1555 |
+
"loss": 4.2503,
|
| 1556 |
+
"step": 110500
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 3.357633322242052,
|
| 1560 |
+
"grad_norm": 1.230765461921692,
|
| 1561 |
+
"learning_rate": 4.854460666280593e-05,
|
| 1562 |
+
"loss": 4.2219,
|
| 1563 |
+
"step": 111000
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 3.3727577966665656,
|
| 1567 |
+
"grad_norm": 1.5857611894607544,
|
| 1568 |
+
"learning_rate": 4.740162760218232e-05,
|
| 1569 |
+
"loss": 4.2336,
|
| 1570 |
+
"step": 111500
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 3.3878822710910796,
|
| 1574 |
+
"grad_norm": 6.604252815246582,
|
| 1575 |
+
"learning_rate": 4.625864854155872e-05,
|
| 1576 |
+
"loss": 4.2456,
|
| 1577 |
+
"step": 112000
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 3.4030067455155932,
|
| 1581 |
+
"grad_norm": 2.8378994464874268,
|
| 1582 |
+
"learning_rate": 4.51156694809351e-05,
|
| 1583 |
+
"loss": 4.2364,
|
| 1584 |
+
"step": 112500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 3.4181312199401073,
|
| 1588 |
+
"grad_norm": 0.8336028456687927,
|
| 1589 |
+
"learning_rate": 4.397269042031149e-05,
|
| 1590 |
+
"loss": 4.2313,
|
| 1591 |
+
"step": 113000
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 3.433255694364621,
|
| 1595 |
+
"grad_norm": 3.1367087364196777,
|
| 1596 |
+
"learning_rate": 4.282971135968789e-05,
|
| 1597 |
+
"loss": 4.2133,
|
| 1598 |
+
"step": 113500
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 3.4483801687891344,
|
| 1602 |
+
"grad_norm": 2.0495407581329346,
|
| 1603 |
+
"learning_rate": 4.168673229906428e-05,
|
| 1604 |
+
"loss": 4.2316,
|
| 1605 |
+
"step": 114000
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 3.4635046432136485,
|
| 1609 |
+
"grad_norm": 1.0248985290527344,
|
| 1610 |
+
"learning_rate": 4.054375323844067e-05,
|
| 1611 |
+
"loss": 4.2344,
|
| 1612 |
+
"step": 114500
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 3.478629117638162,
|
| 1616 |
+
"grad_norm": 7.0710930824279785,
|
| 1617 |
+
"learning_rate": 3.940077417781706e-05,
|
| 1618 |
+
"loss": 4.2359,
|
| 1619 |
+
"step": 115000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 3.4937535920626757,
|
| 1623 |
+
"grad_norm": 0.7784177660942078,
|
| 1624 |
+
"learning_rate": 3.8257795117193444e-05,
|
| 1625 |
+
"loss": 4.2439,
|
| 1626 |
+
"step": 115500
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 3.5088780664871897,
|
| 1630 |
+
"grad_norm": 3.4285647869110107,
|
| 1631 |
+
"learning_rate": 3.711481605656984e-05,
|
| 1632 |
+
"loss": 4.2608,
|
| 1633 |
+
"step": 116000
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 3.5240025409117033,
|
| 1637 |
+
"grad_norm": 0.867929220199585,
|
| 1638 |
+
"learning_rate": 3.597183699594623e-05,
|
| 1639 |
+
"loss": 4.2394,
|
| 1640 |
+
"step": 116500
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 3.5391270153362173,
|
| 1644 |
+
"grad_norm": 1.1300420761108398,
|
| 1645 |
+
"learning_rate": 3.4828857935322624e-05,
|
| 1646 |
+
"loss": 4.247,
|
| 1647 |
+
"step": 117000
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 3.554251489760731,
|
| 1651 |
+
"grad_norm": 2.1013739109039307,
|
| 1652 |
+
"learning_rate": 3.368587887469902e-05,
|
| 1653 |
+
"loss": 4.2243,
|
| 1654 |
+
"step": 117500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 3.5693759641852445,
|
| 1658 |
+
"grad_norm": 2.511087656021118,
|
| 1659 |
+
"learning_rate": 3.2542899814075404e-05,
|
| 1660 |
+
"loss": 4.234,
|
| 1661 |
+
"step": 118000
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 3.5845004386097585,
|
| 1665 |
+
"grad_norm": 1.7033735513687134,
|
| 1666 |
+
"learning_rate": 3.13999207534518e-05,
|
| 1667 |
+
"loss": 4.2333,
|
| 1668 |
+
"step": 118500
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 3.599624913034272,
|
| 1672 |
+
"grad_norm": 2.8376271724700928,
|
| 1673 |
+
"learning_rate": 3.0256941692828184e-05,
|
| 1674 |
+
"loss": 4.2464,
|
| 1675 |
+
"step": 119000
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 3.6147493874587857,
|
| 1679 |
+
"grad_norm": 1.8306654691696167,
|
| 1680 |
+
"learning_rate": 2.9113962632204574e-05,
|
| 1681 |
+
"loss": 4.2422,
|
| 1682 |
+
"step": 119500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 3.6298738618832997,
|
| 1686 |
+
"grad_norm": 0.9379622936248779,
|
| 1687 |
+
"learning_rate": 2.7970983571580967e-05,
|
| 1688 |
+
"loss": 4.2411,
|
| 1689 |
+
"step": 120000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 3.6449983363078133,
|
| 1693 |
+
"grad_norm": 1.011892557144165,
|
| 1694 |
+
"learning_rate": 2.6828004510957357e-05,
|
| 1695 |
+
"loss": 4.2383,
|
| 1696 |
+
"step": 120500
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 3.660122810732327,
|
| 1700 |
+
"grad_norm": 2.614025354385376,
|
| 1701 |
+
"learning_rate": 2.5685025450333747e-05,
|
| 1702 |
+
"loss": 4.2438,
|
| 1703 |
+
"step": 121000
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 3.675247285156841,
|
| 1707 |
+
"grad_norm": 0.6671600341796875,
|
| 1708 |
+
"learning_rate": 2.4542046389710137e-05,
|
| 1709 |
+
"loss": 4.2384,
|
| 1710 |
+
"step": 121500
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 3.6903717595813545,
|
| 1714 |
+
"grad_norm": 0.979447603225708,
|
| 1715 |
+
"learning_rate": 2.339906732908653e-05,
|
| 1716 |
+
"loss": 4.2395,
|
| 1717 |
+
"step": 122000
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 3.705496234005868,
|
| 1721 |
+
"grad_norm": 1.3064838647842407,
|
| 1722 |
+
"learning_rate": 2.225608826846292e-05,
|
| 1723 |
+
"loss": 4.2364,
|
| 1724 |
+
"step": 122500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 3.720620708430382,
|
| 1728 |
+
"grad_norm": 0.5672982335090637,
|
| 1729 |
+
"learning_rate": 2.1113109207839308e-05,
|
| 1730 |
+
"loss": 4.2382,
|
| 1731 |
+
"step": 123000
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 3.7357451828548958,
|
| 1735 |
+
"grad_norm": 1.065990686416626,
|
| 1736 |
+
"learning_rate": 1.99701301472157e-05,
|
| 1737 |
+
"loss": 4.2398,
|
| 1738 |
+
"step": 123500
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 3.7508696572794094,
|
| 1742 |
+
"grad_norm": 0.629031240940094,
|
| 1743 |
+
"learning_rate": 1.882715108659209e-05,
|
| 1744 |
+
"loss": 4.2492,
|
| 1745 |
+
"step": 124000
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 3.7659941317039234,
|
| 1749 |
+
"grad_norm": 3.2597365379333496,
|
| 1750 |
+
"learning_rate": 1.7684172025968484e-05,
|
| 1751 |
+
"loss": 4.248,
|
| 1752 |
+
"step": 124500
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 3.781118606128437,
|
| 1756 |
+
"grad_norm": 1.9842573404312134,
|
| 1757 |
+
"learning_rate": 1.654119296534487e-05,
|
| 1758 |
+
"loss": 4.2413,
|
| 1759 |
+
"step": 125000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 3.7962430805529506,
|
| 1763 |
+
"grad_norm": 4.207962512969971,
|
| 1764 |
+
"learning_rate": 1.5398213904721264e-05,
|
| 1765 |
+
"loss": 4.2354,
|
| 1766 |
+
"step": 125500
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 3.8113675549774646,
|
| 1770 |
+
"grad_norm": 1.403330683708191,
|
| 1771 |
+
"learning_rate": 1.4255234844097655e-05,
|
| 1772 |
+
"loss": 4.2356,
|
| 1773 |
+
"step": 126000
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 3.826492029401978,
|
| 1777 |
+
"grad_norm": 1.0082250833511353,
|
| 1778 |
+
"learning_rate": 1.3112255783474045e-05,
|
| 1779 |
+
"loss": 4.2337,
|
| 1780 |
+
"step": 126500
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 3.841616503826492,
|
| 1784 |
+
"grad_norm": 1.6467324495315552,
|
| 1785 |
+
"learning_rate": 1.1969276722850436e-05,
|
| 1786 |
+
"loss": 4.2311,
|
| 1787 |
+
"step": 127000
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 3.856740978251006,
|
| 1791 |
+
"grad_norm": 2.5783376693725586,
|
| 1792 |
+
"learning_rate": 1.0826297662226828e-05,
|
| 1793 |
+
"loss": 4.2397,
|
| 1794 |
+
"step": 127500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 3.8718654526755194,
|
| 1798 |
+
"grad_norm": 3.2470388412475586,
|
| 1799 |
+
"learning_rate": 9.683318601603218e-06,
|
| 1800 |
+
"loss": 4.2255,
|
| 1801 |
+
"step": 128000
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 3.8869899271000334,
|
| 1805 |
+
"grad_norm": 1.2348912954330444,
|
| 1806 |
+
"learning_rate": 8.540339540979608e-06,
|
| 1807 |
+
"loss": 4.2199,
|
| 1808 |
+
"step": 128500
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 3.902114401524547,
|
| 1812 |
+
"grad_norm": 1.769829273223877,
|
| 1813 |
+
"learning_rate": 7.397360480355999e-06,
|
| 1814 |
+
"loss": 4.2265,
|
| 1815 |
+
"step": 129000
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 3.9172388759490606,
|
| 1819 |
+
"grad_norm": 6.194337844848633,
|
| 1820 |
+
"learning_rate": 6.25438141973239e-06,
|
| 1821 |
+
"loss": 4.2207,
|
| 1822 |
+
"step": 129500
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 3.9323633503735747,
|
| 1826 |
+
"grad_norm": 3.9080488681793213,
|
| 1827 |
+
"learning_rate": 5.111402359108781e-06,
|
| 1828 |
+
"loss": 4.2272,
|
| 1829 |
+
"step": 130000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 3.9474878247980882,
|
| 1833 |
+
"grad_norm": 1.4277572631835938,
|
| 1834 |
+
"learning_rate": 3.968423298485172e-06,
|
| 1835 |
+
"loss": 4.2415,
|
| 1836 |
+
"step": 130500
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 3.9626122992226023,
|
| 1840 |
+
"grad_norm": 2.8284835815429688,
|
| 1841 |
+
"learning_rate": 2.825444237861562e-06,
|
| 1842 |
+
"loss": 4.2406,
|
| 1843 |
+
"step": 131000
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 3.977736773647116,
|
| 1847 |
+
"grad_norm": 3.0596628189086914,
|
| 1848 |
+
"learning_rate": 1.6824651772379527e-06,
|
| 1849 |
+
"loss": 4.2342,
|
| 1850 |
+
"step": 131500
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 3.9928612480716295,
|
| 1854 |
+
"grad_norm": 2.9372382164001465,
|
| 1855 |
+
"learning_rate": 5.394861166143436e-07,
|
| 1856 |
+
"loss": 4.24,
|
| 1857 |
+
"step": 132000
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 4.0,
|
| 1861 |
+
"step": 132236,
|
| 1862 |
+
"total_flos": 6.962502728150323e+16,
|
| 1863 |
+
"train_loss": 4.240191396162094,
|
| 1864 |
+
"train_runtime": 19577.9877,
|
| 1865 |
+
"train_samples_per_second": 54.034,
|
| 1866 |
+
"train_steps_per_second": 6.754
|
| 1867 |
+
}
|
| 1868 |
+
],
|
| 1869 |
+
"logging_steps": 500,
|
| 1870 |
+
"max_steps": 132236,
|
| 1871 |
+
"num_input_tokens_seen": 0,
|
| 1872 |
+
"num_train_epochs": 4,
|
| 1873 |
+
"save_steps": 5000,
|
| 1874 |
+
"stateful_callbacks": {
|
| 1875 |
+
"TrainerControl": {
|
| 1876 |
+
"args": {
|
| 1877 |
+
"should_epoch_stop": false,
|
| 1878 |
+
"should_evaluate": false,
|
| 1879 |
+
"should_log": false,
|
| 1880 |
+
"should_save": true,
|
| 1881 |
+
"should_training_stop": true
|
| 1882 |
+
},
|
| 1883 |
+
"attributes": {}
|
| 1884 |
+
}
|
| 1885 |
+
},
|
| 1886 |
+
"total_flos": 6.962502728150323e+16,
|
| 1887 |
+
"train_batch_size": 8,
|
| 1888 |
+
"trial_name": null,
|
| 1889 |
+
"trial_params": null
|
| 1890 |
+
}
|