Commit
•
b1b0269
1
Parent(s):
caffaf3
End of training
Browse files- README.md +19 -2
- all_results.json +16 -0
- eval_results.json +10 -0
- logs/events.out.tfevents.1720019630.s_003_m.242741.1 +3 -0
- train_results.json +9 -0
- trainer_state.json +2002 -0
README.md
CHANGED
|
@@ -3,9 +3,23 @@ license: mit
|
|
| 3 |
base_model: openai-community/gpt2
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
model-index:
|
| 7 |
- name: gpt_train_12_384_new
|
| 8 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -13,7 +27,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# gpt_train_12_384_new
|
| 15 |
|
| 16 |
-
This model is a fine-tuned version of [openai-community/gpt2](https://huggingface.co/openai-community/gpt2) on
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
## Model description
|
| 19 |
|
|
|
|
| 3 |
base_model: openai-community/gpt2
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
| 6 |
+
datasets:
|
| 7 |
+
- gokuls/wiki_book_corpus_raw_dataset_tiny
|
| 8 |
+
metrics:
|
| 9 |
+
- accuracy
|
| 10 |
model-index:
|
| 11 |
- name: gpt_train_12_384_new
|
| 12 |
+
results:
|
| 13 |
+
- task:
|
| 14 |
+
name: Causal Language Modeling
|
| 15 |
+
type: text-generation
|
| 16 |
+
dataset:
|
| 17 |
+
name: gokuls/wiki_book_corpus_raw_dataset_tiny
|
| 18 |
+
type: gokuls/wiki_book_corpus_raw_dataset_tiny
|
| 19 |
+
metrics:
|
| 20 |
+
- name: Accuracy
|
| 21 |
+
type: accuracy
|
| 22 |
+
value: 0.364555259244386
|
| 23 |
---
|
| 24 |
|
| 25 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 27 |
|
| 28 |
# gpt_train_12_384_new
|
| 29 |
|
| 30 |
+
This model is a fine-tuned version of [openai-community/gpt2](https://huggingface.co/openai-community/gpt2) on the gokuls/wiki_book_corpus_raw_dataset_tiny dataset.
|
| 31 |
+
It achieves the following results on the evaluation set:
|
| 32 |
+
- Loss: 3.7116
|
| 33 |
+
- Accuracy: 0.3646
|
| 34 |
|
| 35 |
## Model description
|
| 36 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.7809745229100065,
|
| 3 |
+
"eval_accuracy": 0.364555259244386,
|
| 4 |
+
"eval_loss": 3.7116007804870605,
|
| 5 |
+
"eval_runtime": 273.6863,
|
| 6 |
+
"eval_samples": 33767,
|
| 7 |
+
"eval_samples_per_second": 123.378,
|
| 8 |
+
"eval_steps_per_second": 3.858,
|
| 9 |
+
"perplexity": 40.919256875144555,
|
| 10 |
+
"total_flos": 5.861064073050849e+17,
|
| 11 |
+
"train_loss": 4.399306026785714,
|
| 12 |
+
"train_runtime": 93598.1238,
|
| 13 |
+
"train_samples": 660643,
|
| 14 |
+
"train_samples_per_second": 705.829,
|
| 15 |
+
"train_steps_per_second": 22.058
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.7809745229100065,
|
| 3 |
+
"eval_accuracy": 0.364555259244386,
|
| 4 |
+
"eval_loss": 3.7116007804870605,
|
| 5 |
+
"eval_runtime": 273.6863,
|
| 6 |
+
"eval_samples": 33767,
|
| 7 |
+
"eval_samples_per_second": 123.378,
|
| 8 |
+
"eval_steps_per_second": 3.858,
|
| 9 |
+
"perplexity": 40.919256875144555
|
| 10 |
+
}
|
logs/events.out.tfevents.1720019630.s_003_m.242741.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d2df5f8332f6cf191e60dfe4c970c6d30af2573c75671af57c6b66b586178e5
|
| 3 |
+
size 369
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.7809745229100065,
|
| 3 |
+
"total_flos": 5.861064073050849e+17,
|
| 4 |
+
"train_loss": 4.399306026785714,
|
| 5 |
+
"train_runtime": 93598.1238,
|
| 6 |
+
"train_samples": 660643,
|
| 7 |
+
"train_samples_per_second": 705.829,
|
| 8 |
+
"train_steps_per_second": 22.058
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2002 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.7809745229100065,
|
| 5 |
+
"eval_steps": 1000000,
|
| 6 |
+
"global_step": 140000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.024217766153250025,
|
| 13 |
+
"grad_norm": 1.516142725944519,
|
| 14 |
+
"learning_rate": 9.997578223384676e-06,
|
| 15 |
+
"loss": 9.1668,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04843553230650005,
|
| 20 |
+
"grad_norm": 1.0418092012405396,
|
| 21 |
+
"learning_rate": 9.995156446769351e-06,
|
| 22 |
+
"loss": 7.7891,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.07265329845975008,
|
| 27 |
+
"grad_norm": 0.9371763467788696,
|
| 28 |
+
"learning_rate": 9.992734670154026e-06,
|
| 29 |
+
"loss": 7.1939,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0968710646130001,
|
| 34 |
+
"grad_norm": 1.1670751571655273,
|
| 35 |
+
"learning_rate": 9.990312893538701e-06,
|
| 36 |
+
"loss": 6.9189,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.12108883076625013,
|
| 41 |
+
"grad_norm": 1.3695101737976074,
|
| 42 |
+
"learning_rate": 9.987891116923376e-06,
|
| 43 |
+
"loss": 6.7278,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.14530659691950015,
|
| 48 |
+
"grad_norm": 1.796486735343933,
|
| 49 |
+
"learning_rate": 9.985469340308051e-06,
|
| 50 |
+
"loss": 6.5867,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.16952436307275018,
|
| 55 |
+
"grad_norm": 1.509717583656311,
|
| 56 |
+
"learning_rate": 9.983047563692726e-06,
|
| 57 |
+
"loss": 6.4508,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.1937421292260002,
|
| 62 |
+
"grad_norm": 1.8329906463623047,
|
| 63 |
+
"learning_rate": 9.9806257870774e-06,
|
| 64 |
+
"loss": 6.3551,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.21795989537925023,
|
| 69 |
+
"grad_norm": 1.5139986276626587,
|
| 70 |
+
"learning_rate": 9.978204010462076e-06,
|
| 71 |
+
"loss": 6.2743,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.24217766153250025,
|
| 76 |
+
"grad_norm": 2.2407052516937256,
|
| 77 |
+
"learning_rate": 9.97578223384675e-06,
|
| 78 |
+
"loss": 6.2001,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.26639542768575025,
|
| 83 |
+
"grad_norm": 2.087357521057129,
|
| 84 |
+
"learning_rate": 9.973360457231426e-06,
|
| 85 |
+
"loss": 6.1334,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2906131938390003,
|
| 90 |
+
"grad_norm": 2.0182762145996094,
|
| 91 |
+
"learning_rate": 9.970938680616102e-06,
|
| 92 |
+
"loss": 6.0656,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3148309599922503,
|
| 97 |
+
"grad_norm": 1.9544531106948853,
|
| 98 |
+
"learning_rate": 9.968516904000775e-06,
|
| 99 |
+
"loss": 6.0134,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.33904872614550036,
|
| 104 |
+
"grad_norm": 2.3156166076660156,
|
| 105 |
+
"learning_rate": 9.966095127385452e-06,
|
| 106 |
+
"loss": 5.9546,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.36326649229875035,
|
| 111 |
+
"grad_norm": 2.5564098358154297,
|
| 112 |
+
"learning_rate": 9.963673350770125e-06,
|
| 113 |
+
"loss": 5.9063,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3874842584520004,
|
| 118 |
+
"grad_norm": 2.191112518310547,
|
| 119 |
+
"learning_rate": 9.961251574154802e-06,
|
| 120 |
+
"loss": 5.8564,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.4117020246052504,
|
| 125 |
+
"grad_norm": 2.1813371181488037,
|
| 126 |
+
"learning_rate": 9.958829797539475e-06,
|
| 127 |
+
"loss": 5.8053,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.43591979075850046,
|
| 132 |
+
"grad_norm": 1.9756942987442017,
|
| 133 |
+
"learning_rate": 9.95640802092415e-06,
|
| 134 |
+
"loss": 5.7669,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.46013755691175046,
|
| 139 |
+
"grad_norm": 2.2932822704315186,
|
| 140 |
+
"learning_rate": 9.953986244308825e-06,
|
| 141 |
+
"loss": 5.7218,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.4843553230650005,
|
| 146 |
+
"grad_norm": 2.218536376953125,
|
| 147 |
+
"learning_rate": 9.9515644676935e-06,
|
| 148 |
+
"loss": 5.6818,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5085730892182505,
|
| 153 |
+
"grad_norm": 2.3896877765655518,
|
| 154 |
+
"learning_rate": 9.949142691078175e-06,
|
| 155 |
+
"loss": 5.6481,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5327908553715005,
|
| 160 |
+
"grad_norm": 2.5433712005615234,
|
| 161 |
+
"learning_rate": 9.94672091446285e-06,
|
| 162 |
+
"loss": 5.6124,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5570086215247505,
|
| 167 |
+
"grad_norm": 2.5442490577697754,
|
| 168 |
+
"learning_rate": 9.944299137847525e-06,
|
| 169 |
+
"loss": 5.5728,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.5812263876780006,
|
| 174 |
+
"grad_norm": 2.327425241470337,
|
| 175 |
+
"learning_rate": 9.9418773612322e-06,
|
| 176 |
+
"loss": 5.5406,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6054441538312506,
|
| 181 |
+
"grad_norm": 2.290090799331665,
|
| 182 |
+
"learning_rate": 9.939455584616876e-06,
|
| 183 |
+
"loss": 5.5121,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6296619199845006,
|
| 188 |
+
"grad_norm": 3.161325216293335,
|
| 189 |
+
"learning_rate": 9.93703380800155e-06,
|
| 190 |
+
"loss": 5.4739,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6538796861377506,
|
| 195 |
+
"grad_norm": 2.6134533882141113,
|
| 196 |
+
"learning_rate": 9.934612031386226e-06,
|
| 197 |
+
"loss": 5.4384,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.6780974522910007,
|
| 202 |
+
"grad_norm": 2.674760580062866,
|
| 203 |
+
"learning_rate": 9.9321902547709e-06,
|
| 204 |
+
"loss": 5.413,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7023152184442507,
|
| 209 |
+
"grad_norm": 2.431614398956299,
|
| 210 |
+
"learning_rate": 9.929768478155576e-06,
|
| 211 |
+
"loss": 5.3903,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7265329845975007,
|
| 216 |
+
"grad_norm": 2.4028687477111816,
|
| 217 |
+
"learning_rate": 9.927346701540251e-06,
|
| 218 |
+
"loss": 5.3637,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7507507507507507,
|
| 223 |
+
"grad_norm": 2.4807944297790527,
|
| 224 |
+
"learning_rate": 9.924924924924926e-06,
|
| 225 |
+
"loss": 5.3279,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7749685169040008,
|
| 230 |
+
"grad_norm": 2.9065611362457275,
|
| 231 |
+
"learning_rate": 9.922503148309601e-06,
|
| 232 |
+
"loss": 5.3098,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.7991862830572508,
|
| 237 |
+
"grad_norm": 2.359736204147339,
|
| 238 |
+
"learning_rate": 9.920081371694276e-06,
|
| 239 |
+
"loss": 5.2858,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8234040492105008,
|
| 244 |
+
"grad_norm": 2.642854690551758,
|
| 245 |
+
"learning_rate": 9.917659595078951e-06,
|
| 246 |
+
"loss": 5.2518,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.8476218153637508,
|
| 251 |
+
"grad_norm": 3.2326414585113525,
|
| 252 |
+
"learning_rate": 9.915237818463626e-06,
|
| 253 |
+
"loss": 5.2354,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.8718395815170009,
|
| 258 |
+
"grad_norm": 2.285203218460083,
|
| 259 |
+
"learning_rate": 9.912816041848301e-06,
|
| 260 |
+
"loss": 5.2117,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.8960573476702509,
|
| 265 |
+
"grad_norm": 2.551164388656616,
|
| 266 |
+
"learning_rate": 9.910394265232976e-06,
|
| 267 |
+
"loss": 5.1941,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9202751138235009,
|
| 272 |
+
"grad_norm": 2.678759813308716,
|
| 273 |
+
"learning_rate": 9.907972488617651e-06,
|
| 274 |
+
"loss": 5.1706,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.9444928799767509,
|
| 279 |
+
"grad_norm": 2.6895062923431396,
|
| 280 |
+
"learning_rate": 9.905550712002325e-06,
|
| 281 |
+
"loss": 5.1499,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.968710646130001,
|
| 286 |
+
"grad_norm": 2.554659128189087,
|
| 287 |
+
"learning_rate": 9.903128935387001e-06,
|
| 288 |
+
"loss": 5.1276,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.992928412283251,
|
| 293 |
+
"grad_norm": 2.785282850265503,
|
| 294 |
+
"learning_rate": 9.900707158771675e-06,
|
| 295 |
+
"loss": 5.1079,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.017146178436501,
|
| 300 |
+
"grad_norm": 2.7283270359039307,
|
| 301 |
+
"learning_rate": 9.89828538215635e-06,
|
| 302 |
+
"loss": 5.0726,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.0413639445897511,
|
| 307 |
+
"grad_norm": 2.654245615005493,
|
| 308 |
+
"learning_rate": 9.895863605541027e-06,
|
| 309 |
+
"loss": 5.0654,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.065581710743001,
|
| 314 |
+
"grad_norm": 2.563713550567627,
|
| 315 |
+
"learning_rate": 9.8934418289257e-06,
|
| 316 |
+
"loss": 5.0436,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.0897994768962511,
|
| 321 |
+
"grad_norm": 2.6896631717681885,
|
| 322 |
+
"learning_rate": 9.891020052310377e-06,
|
| 323 |
+
"loss": 5.0161,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.114017243049501,
|
| 328 |
+
"grad_norm": 2.8477983474731445,
|
| 329 |
+
"learning_rate": 9.88859827569505e-06,
|
| 330 |
+
"loss": 5.008,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.1382350092027511,
|
| 335 |
+
"grad_norm": 2.6253600120544434,
|
| 336 |
+
"learning_rate": 9.886176499079725e-06,
|
| 337 |
+
"loss": 4.987,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.1624527753560012,
|
| 342 |
+
"grad_norm": 2.7618229389190674,
|
| 343 |
+
"learning_rate": 9.8837547224644e-06,
|
| 344 |
+
"loss": 4.9655,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.186670541509251,
|
| 349 |
+
"grad_norm": 2.7631571292877197,
|
| 350 |
+
"learning_rate": 9.881332945849075e-06,
|
| 351 |
+
"loss": 4.9426,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.2108883076625012,
|
| 356 |
+
"grad_norm": 3.108574390411377,
|
| 357 |
+
"learning_rate": 9.87891116923375e-06,
|
| 358 |
+
"loss": 4.9264,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.2351060738157513,
|
| 363 |
+
"grad_norm": 2.5930752754211426,
|
| 364 |
+
"learning_rate": 9.876489392618425e-06,
|
| 365 |
+
"loss": 4.9068,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.2593238399690012,
|
| 370 |
+
"grad_norm": 2.4590559005737305,
|
| 371 |
+
"learning_rate": 9.8740676160031e-06,
|
| 372 |
+
"loss": 4.8908,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.2835416061222513,
|
| 377 |
+
"grad_norm": 2.7004990577697754,
|
| 378 |
+
"learning_rate": 9.871645839387776e-06,
|
| 379 |
+
"loss": 4.8767,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.3077593722755014,
|
| 384 |
+
"grad_norm": 2.5023412704467773,
|
| 385 |
+
"learning_rate": 9.86922406277245e-06,
|
| 386 |
+
"loss": 4.8543,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.3319771384287513,
|
| 391 |
+
"grad_norm": 3.338123083114624,
|
| 392 |
+
"learning_rate": 9.866802286157126e-06,
|
| 393 |
+
"loss": 4.8324,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.3561949045820014,
|
| 398 |
+
"grad_norm": 2.871856689453125,
|
| 399 |
+
"learning_rate": 9.8643805095418e-06,
|
| 400 |
+
"loss": 4.8138,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.3804126707352513,
|
| 405 |
+
"grad_norm": 3.148714303970337,
|
| 406 |
+
"learning_rate": 9.861958732926476e-06,
|
| 407 |
+
"loss": 4.7991,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.4046304368885014,
|
| 412 |
+
"grad_norm": 2.986448287963867,
|
| 413 |
+
"learning_rate": 9.85953695631115e-06,
|
| 414 |
+
"loss": 4.781,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.4288482030417513,
|
| 419 |
+
"grad_norm": 2.5939040184020996,
|
| 420 |
+
"learning_rate": 9.857115179695826e-06,
|
| 421 |
+
"loss": 4.7634,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.4530659691950014,
|
| 426 |
+
"grad_norm": 2.674027442932129,
|
| 427 |
+
"learning_rate": 9.854693403080501e-06,
|
| 428 |
+
"loss": 4.7446,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.4772837353482515,
|
| 433 |
+
"grad_norm": 3.018937826156616,
|
| 434 |
+
"learning_rate": 9.852271626465176e-06,
|
| 435 |
+
"loss": 4.7332,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.5015015015015014,
|
| 440 |
+
"grad_norm": 2.862410306930542,
|
| 441 |
+
"learning_rate": 9.849849849849851e-06,
|
| 442 |
+
"loss": 4.7151,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.5257192676547515,
|
| 447 |
+
"grad_norm": 2.9605488777160645,
|
| 448 |
+
"learning_rate": 9.847428073234524e-06,
|
| 449 |
+
"loss": 4.7023,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.5499370338080016,
|
| 454 |
+
"grad_norm": 3.116225242614746,
|
| 455 |
+
"learning_rate": 9.845006296619201e-06,
|
| 456 |
+
"loss": 4.6834,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.5741547999612515,
|
| 461 |
+
"grad_norm": 3.074164390563965,
|
| 462 |
+
"learning_rate": 9.842584520003876e-06,
|
| 463 |
+
"loss": 4.6676,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.5983725661145016,
|
| 468 |
+
"grad_norm": 2.677706003189087,
|
| 469 |
+
"learning_rate": 9.840162743388551e-06,
|
| 470 |
+
"loss": 4.6547,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.6225903322677517,
|
| 475 |
+
"grad_norm": 2.832223653793335,
|
| 476 |
+
"learning_rate": 9.837740966773226e-06,
|
| 477 |
+
"loss": 4.6402,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.6468080984210016,
|
| 482 |
+
"grad_norm": 3.1041297912597656,
|
| 483 |
+
"learning_rate": 9.8353191901579e-06,
|
| 484 |
+
"loss": 4.6271,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.6710258645742515,
|
| 489 |
+
"grad_norm": 2.883216381072998,
|
| 490 |
+
"learning_rate": 9.832897413542576e-06,
|
| 491 |
+
"loss": 4.6141,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.6952436307275018,
|
| 496 |
+
"grad_norm": 2.894000291824341,
|
| 497 |
+
"learning_rate": 9.83047563692725e-06,
|
| 498 |
+
"loss": 4.6031,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.7194613968807517,
|
| 503 |
+
"grad_norm": 2.9335453510284424,
|
| 504 |
+
"learning_rate": 9.828053860311927e-06,
|
| 505 |
+
"loss": 4.5911,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.7436791630340016,
|
| 510 |
+
"grad_norm": 2.7511613368988037,
|
| 511 |
+
"learning_rate": 9.8256320836966e-06,
|
| 512 |
+
"loss": 4.5824,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.7678969291872517,
|
| 517 |
+
"grad_norm": 2.8148419857025146,
|
| 518 |
+
"learning_rate": 9.823210307081275e-06,
|
| 519 |
+
"loss": 4.5693,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.7921146953405018,
|
| 524 |
+
"grad_norm": 2.8832480907440186,
|
| 525 |
+
"learning_rate": 9.820788530465952e-06,
|
| 526 |
+
"loss": 4.5622,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.8163324614937517,
|
| 531 |
+
"grad_norm": 2.9674079418182373,
|
| 532 |
+
"learning_rate": 9.818366753850625e-06,
|
| 533 |
+
"loss": 4.5473,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.8405502276470018,
|
| 538 |
+
"grad_norm": 2.971090793609619,
|
| 539 |
+
"learning_rate": 9.815944977235302e-06,
|
| 540 |
+
"loss": 4.538,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.864767993800252,
|
| 545 |
+
"grad_norm": 2.785881996154785,
|
| 546 |
+
"learning_rate": 9.813523200619975e-06,
|
| 547 |
+
"loss": 4.5327,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.8889857599535018,
|
| 552 |
+
"grad_norm": 2.9853248596191406,
|
| 553 |
+
"learning_rate": 9.81110142400465e-06,
|
| 554 |
+
"loss": 4.5078,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.913203526106752,
|
| 559 |
+
"grad_norm": 2.899179697036743,
|
| 560 |
+
"learning_rate": 9.808679647389325e-06,
|
| 561 |
+
"loss": 4.5002,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.937421292260002,
|
| 566 |
+
"grad_norm": 2.5843992233276367,
|
| 567 |
+
"learning_rate": 9.806257870774e-06,
|
| 568 |
+
"loss": 4.4928,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.961639058413252,
|
| 573 |
+
"grad_norm": 2.8425755500793457,
|
| 574 |
+
"learning_rate": 9.803836094158675e-06,
|
| 575 |
+
"loss": 4.4881,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.985856824566502,
|
| 580 |
+
"grad_norm": 2.87211275100708,
|
| 581 |
+
"learning_rate": 9.80141431754335e-06,
|
| 582 |
+
"loss": 4.4822,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.010074590719752,
|
| 587 |
+
"grad_norm": 3.0297703742980957,
|
| 588 |
+
"learning_rate": 9.798992540928026e-06,
|
| 589 |
+
"loss": 4.4619,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.034292356873002,
|
| 594 |
+
"grad_norm": 2.9869863986968994,
|
| 595 |
+
"learning_rate": 9.7965707643127e-06,
|
| 596 |
+
"loss": 4.4526,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.058510123026252,
|
| 601 |
+
"grad_norm": 2.777209520339966,
|
| 602 |
+
"learning_rate": 9.794148987697376e-06,
|
| 603 |
+
"loss": 4.432,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.0827278891795022,
|
| 608 |
+
"grad_norm": 3.0258235931396484,
|
| 609 |
+
"learning_rate": 9.79172721108205e-06,
|
| 610 |
+
"loss": 4.4386,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.106945655332752,
|
| 615 |
+
"grad_norm": 2.8184220790863037,
|
| 616 |
+
"learning_rate": 9.789305434466726e-06,
|
| 617 |
+
"loss": 4.4254,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.131163421486002,
|
| 622 |
+
"grad_norm": 2.9428908824920654,
|
| 623 |
+
"learning_rate": 9.7868836578514e-06,
|
| 624 |
+
"loss": 4.4172,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.1553811876392523,
|
| 629 |
+
"grad_norm": 3.1215102672576904,
|
| 630 |
+
"learning_rate": 9.784461881236076e-06,
|
| 631 |
+
"loss": 4.4078,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.1795989537925022,
|
| 636 |
+
"grad_norm": 3.032611846923828,
|
| 637 |
+
"learning_rate": 9.782040104620751e-06,
|
| 638 |
+
"loss": 4.4036,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.203816719945752,
|
| 643 |
+
"grad_norm": 2.9431488513946533,
|
| 644 |
+
"learning_rate": 9.779618328005426e-06,
|
| 645 |
+
"loss": 4.3997,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.228034486099002,
|
| 650 |
+
"grad_norm": 2.9058682918548584,
|
| 651 |
+
"learning_rate": 9.7771965513901e-06,
|
| 652 |
+
"loss": 4.389,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.2522522522522523,
|
| 657 |
+
"grad_norm": 2.703967809677124,
|
| 658 |
+
"learning_rate": 9.774774774774776e-06,
|
| 659 |
+
"loss": 4.3753,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.2764700184055022,
|
| 664 |
+
"grad_norm": 2.764721155166626,
|
| 665 |
+
"learning_rate": 9.77235299815945e-06,
|
| 666 |
+
"loss": 4.3658,
|
| 667 |
+
"step": 47000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.300687784558752,
|
| 671 |
+
"grad_norm": 2.834578514099121,
|
| 672 |
+
"learning_rate": 9.769931221544126e-06,
|
| 673 |
+
"loss": 4.3577,
|
| 674 |
+
"step": 47500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.3249055507120024,
|
| 678 |
+
"grad_norm": 2.9823198318481445,
|
| 679 |
+
"learning_rate": 9.767509444928801e-06,
|
| 680 |
+
"loss": 4.3531,
|
| 681 |
+
"step": 48000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.3491233168652523,
|
| 685 |
+
"grad_norm": 2.8373069763183594,
|
| 686 |
+
"learning_rate": 9.765087668313475e-06,
|
| 687 |
+
"loss": 4.3515,
|
| 688 |
+
"step": 48500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.373341083018502,
|
| 692 |
+
"grad_norm": 2.6971516609191895,
|
| 693 |
+
"learning_rate": 9.762665891698151e-06,
|
| 694 |
+
"loss": 4.3367,
|
| 695 |
+
"step": 49000
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.3975588491717525,
|
| 699 |
+
"grad_norm": 2.8022115230560303,
|
| 700 |
+
"learning_rate": 9.760244115082825e-06,
|
| 701 |
+
"loss": 4.3302,
|
| 702 |
+
"step": 49500
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.4217766153250024,
|
| 706 |
+
"grad_norm": 2.9047532081604004,
|
| 707 |
+
"learning_rate": 9.757822338467502e-06,
|
| 708 |
+
"loss": 4.3202,
|
| 709 |
+
"step": 50000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.4459943814782523,
|
| 713 |
+
"grad_norm": 2.81803297996521,
|
| 714 |
+
"learning_rate": 9.755400561852175e-06,
|
| 715 |
+
"loss": 4.3184,
|
| 716 |
+
"step": 50500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.4702121476315027,
|
| 720 |
+
"grad_norm": 2.9668848514556885,
|
| 721 |
+
"learning_rate": 9.75297878523685e-06,
|
| 722 |
+
"loss": 4.3093,
|
| 723 |
+
"step": 51000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.4944299137847525,
|
| 727 |
+
"grad_norm": 3.0008721351623535,
|
| 728 |
+
"learning_rate": 9.750557008621525e-06,
|
| 729 |
+
"loss": 4.3089,
|
| 730 |
+
"step": 51500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.5186476799380024,
|
| 734 |
+
"grad_norm": 2.76766300201416,
|
| 735 |
+
"learning_rate": 9.7481352320062e-06,
|
| 736 |
+
"loss": 4.2961,
|
| 737 |
+
"step": 52000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.5428654460912528,
|
| 741 |
+
"grad_norm": 2.961453914642334,
|
| 742 |
+
"learning_rate": 9.745713455390875e-06,
|
| 743 |
+
"loss": 4.284,
|
| 744 |
+
"step": 52500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.5670832122445026,
|
| 748 |
+
"grad_norm": 3.030158758163452,
|
| 749 |
+
"learning_rate": 9.74329167877555e-06,
|
| 750 |
+
"loss": 4.2849,
|
| 751 |
+
"step": 53000
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.5913009783977525,
|
| 755 |
+
"grad_norm": 2.9656057357788086,
|
| 756 |
+
"learning_rate": 9.740869902160225e-06,
|
| 757 |
+
"loss": 4.2712,
|
| 758 |
+
"step": 53500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.615518744551003,
|
| 762 |
+
"grad_norm": 3.3482959270477295,
|
| 763 |
+
"learning_rate": 9.7384481255449e-06,
|
| 764 |
+
"loss": 4.2833,
|
| 765 |
+
"step": 54000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.6397365107042527,
|
| 769 |
+
"grad_norm": 2.8142096996307373,
|
| 770 |
+
"learning_rate": 9.736026348929575e-06,
|
| 771 |
+
"loss": 4.2652,
|
| 772 |
+
"step": 54500
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.6639542768575026,
|
| 776 |
+
"grad_norm": 2.776679277420044,
|
| 777 |
+
"learning_rate": 9.73360457231425e-06,
|
| 778 |
+
"loss": 4.2653,
|
| 779 |
+
"step": 55000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.688172043010753,
|
| 783 |
+
"grad_norm": 2.7612788677215576,
|
| 784 |
+
"learning_rate": 9.731182795698925e-06,
|
| 785 |
+
"loss": 4.2562,
|
| 786 |
+
"step": 55500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.712389809164003,
|
| 790 |
+
"grad_norm": 2.959991931915283,
|
| 791 |
+
"learning_rate": 9.7287610190836e-06,
|
| 792 |
+
"loss": 4.2515,
|
| 793 |
+
"step": 56000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.7366075753172527,
|
| 797 |
+
"grad_norm": 2.969061851501465,
|
| 798 |
+
"learning_rate": 9.726339242468276e-06,
|
| 799 |
+
"loss": 4.2378,
|
| 800 |
+
"step": 56500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.7608253414705026,
|
| 804 |
+
"grad_norm": 3.1710784435272217,
|
| 805 |
+
"learning_rate": 9.72391746585295e-06,
|
| 806 |
+
"loss": 4.2408,
|
| 807 |
+
"step": 57000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.7850431076237525,
|
| 811 |
+
"grad_norm": 2.9343762397766113,
|
| 812 |
+
"learning_rate": 9.721495689237626e-06,
|
| 813 |
+
"loss": 4.2316,
|
| 814 |
+
"step": 57500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.809260873777003,
|
| 818 |
+
"grad_norm": 2.98744535446167,
|
| 819 |
+
"learning_rate": 9.7190739126223e-06,
|
| 820 |
+
"loss": 4.2302,
|
| 821 |
+
"step": 58000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.8334786399302527,
|
| 825 |
+
"grad_norm": 2.8376593589782715,
|
| 826 |
+
"learning_rate": 9.716652136006976e-06,
|
| 827 |
+
"loss": 4.2229,
|
| 828 |
+
"step": 58500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.8576964060835026,
|
| 832 |
+
"grad_norm": 2.7830283641815186,
|
| 833 |
+
"learning_rate": 9.714230359391651e-06,
|
| 834 |
+
"loss": 4.2138,
|
| 835 |
+
"step": 59000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.881914172236753,
|
| 839 |
+
"grad_norm": 2.824352741241455,
|
| 840 |
+
"learning_rate": 9.711808582776326e-06,
|
| 841 |
+
"loss": 4.2039,
|
| 842 |
+
"step": 59500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.906131938390003,
|
| 846 |
+
"grad_norm": 2.8537116050720215,
|
| 847 |
+
"learning_rate": 9.709386806161001e-06,
|
| 848 |
+
"loss": 4.2063,
|
| 849 |
+
"step": 60000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.9303497045432527,
|
| 853 |
+
"grad_norm": 3.004157543182373,
|
| 854 |
+
"learning_rate": 9.706965029545674e-06,
|
| 855 |
+
"loss": 4.1983,
|
| 856 |
+
"step": 60500
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.954567470696503,
|
| 860 |
+
"grad_norm": 2.8163509368896484,
|
| 861 |
+
"learning_rate": 9.704543252930351e-06,
|
| 862 |
+
"loss": 4.1938,
|
| 863 |
+
"step": 61000
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.978785236849753,
|
| 867 |
+
"grad_norm": 2.8276596069335938,
|
| 868 |
+
"learning_rate": 9.702121476315024e-06,
|
| 869 |
+
"loss": 4.1915,
|
| 870 |
+
"step": 61500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.003003003003003,
|
| 874 |
+
"grad_norm": 2.7849977016448975,
|
| 875 |
+
"learning_rate": 9.699699699699701e-06,
|
| 876 |
+
"loss": 4.1942,
|
| 877 |
+
"step": 62000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.027220769156253,
|
| 881 |
+
"grad_norm": 2.782846212387085,
|
| 882 |
+
"learning_rate": 9.697277923084375e-06,
|
| 883 |
+
"loss": 4.1741,
|
| 884 |
+
"step": 62500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.051438535309503,
|
| 888 |
+
"grad_norm": 2.906552314758301,
|
| 889 |
+
"learning_rate": 9.69485614646905e-06,
|
| 890 |
+
"loss": 4.1767,
|
| 891 |
+
"step": 63000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.075656301462753,
|
| 895 |
+
"grad_norm": 3.0256595611572266,
|
| 896 |
+
"learning_rate": 9.692434369853726e-06,
|
| 897 |
+
"loss": 4.1665,
|
| 898 |
+
"step": 63500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.0998740676160033,
|
| 902 |
+
"grad_norm": 2.847698450088501,
|
| 903 |
+
"learning_rate": 9.6900125932384e-06,
|
| 904 |
+
"loss": 4.1642,
|
| 905 |
+
"step": 64000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.124091833769253,
|
| 909 |
+
"grad_norm": 2.8021674156188965,
|
| 910 |
+
"learning_rate": 9.687590816623077e-06,
|
| 911 |
+
"loss": 4.1663,
|
| 912 |
+
"step": 64500
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.148309599922503,
|
| 916 |
+
"grad_norm": 2.784911632537842,
|
| 917 |
+
"learning_rate": 9.68516904000775e-06,
|
| 918 |
+
"loss": 4.1546,
|
| 919 |
+
"step": 65000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.1725273660757534,
|
| 923 |
+
"grad_norm": 3.019435167312622,
|
| 924 |
+
"learning_rate": 9.682747263392425e-06,
|
| 925 |
+
"loss": 4.1443,
|
| 926 |
+
"step": 65500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.1967451322290033,
|
| 930 |
+
"grad_norm": 2.60965895652771,
|
| 931 |
+
"learning_rate": 9.6803254867771e-06,
|
| 932 |
+
"loss": 4.1465,
|
| 933 |
+
"step": 66000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.220962898382253,
|
| 937 |
+
"grad_norm": 2.740164041519165,
|
| 938 |
+
"learning_rate": 9.677903710161775e-06,
|
| 939 |
+
"loss": 4.1345,
|
| 940 |
+
"step": 66500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.2451806645355035,
|
| 944 |
+
"grad_norm": 2.862274646759033,
|
| 945 |
+
"learning_rate": 9.67548193354645e-06,
|
| 946 |
+
"loss": 4.1461,
|
| 947 |
+
"step": 67000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.2693984306887534,
|
| 951 |
+
"grad_norm": 2.8547213077545166,
|
| 952 |
+
"learning_rate": 9.673060156931125e-06,
|
| 953 |
+
"loss": 4.137,
|
| 954 |
+
"step": 67500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.2936161968420032,
|
| 958 |
+
"grad_norm": 3.0033137798309326,
|
| 959 |
+
"learning_rate": 9.6706383803158e-06,
|
| 960 |
+
"loss": 4.1253,
|
| 961 |
+
"step": 68000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.317833962995253,
|
| 965 |
+
"grad_norm": 2.795989513397217,
|
| 966 |
+
"learning_rate": 9.668216603700475e-06,
|
| 967 |
+
"loss": 4.1232,
|
| 968 |
+
"step": 68500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.3420517291485035,
|
| 972 |
+
"grad_norm": 2.8020830154418945,
|
| 973 |
+
"learning_rate": 9.66579482708515e-06,
|
| 974 |
+
"loss": 4.1238,
|
| 975 |
+
"step": 69000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.3662694953017533,
|
| 979 |
+
"grad_norm": 2.808565855026245,
|
| 980 |
+
"learning_rate": 9.663373050469825e-06,
|
| 981 |
+
"loss": 4.1155,
|
| 982 |
+
"step": 69500
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.3904872614550032,
|
| 986 |
+
"grad_norm": 2.7904319763183594,
|
| 987 |
+
"learning_rate": 9.6609512738545e-06,
|
| 988 |
+
"loss": 4.1143,
|
| 989 |
+
"step": 70000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.4147050276082536,
|
| 993 |
+
"grad_norm": 2.7850215435028076,
|
| 994 |
+
"learning_rate": 9.658529497239176e-06,
|
| 995 |
+
"loss": 4.1102,
|
| 996 |
+
"step": 70500
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 3.4389227937615034,
|
| 1000 |
+
"grad_norm": 2.6868176460266113,
|
| 1001 |
+
"learning_rate": 9.65610772062385e-06,
|
| 1002 |
+
"loss": 4.0994,
|
| 1003 |
+
"step": 71000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 3.4631405599147533,
|
| 1007 |
+
"grad_norm": 2.862273931503296,
|
| 1008 |
+
"learning_rate": 9.653685944008526e-06,
|
| 1009 |
+
"loss": 4.1054,
|
| 1010 |
+
"step": 71500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.4873583260680037,
|
| 1014 |
+
"grad_norm": 3.01948881149292,
|
| 1015 |
+
"learning_rate": 9.6512641673932e-06,
|
| 1016 |
+
"loss": 4.0975,
|
| 1017 |
+
"step": 72000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.5115760922212536,
|
| 1021 |
+
"grad_norm": 2.945227861404419,
|
| 1022 |
+
"learning_rate": 9.648842390777876e-06,
|
| 1023 |
+
"loss": 4.0941,
|
| 1024 |
+
"step": 72500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 3.5357938583745034,
|
| 1028 |
+
"grad_norm": 3.265650987625122,
|
| 1029 |
+
"learning_rate": 9.64642061416255e-06,
|
| 1030 |
+
"loss": 4.0854,
|
| 1031 |
+
"step": 73000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.5600116245277533,
|
| 1035 |
+
"grad_norm": 2.839852809906006,
|
| 1036 |
+
"learning_rate": 9.643998837547224e-06,
|
| 1037 |
+
"loss": 4.0889,
|
| 1038 |
+
"step": 73500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 3.5842293906810037,
|
| 1042 |
+
"grad_norm": 3.0958175659179688,
|
| 1043 |
+
"learning_rate": 9.641577060931901e-06,
|
| 1044 |
+
"loss": 4.0793,
|
| 1045 |
+
"step": 74000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.6084471568342535,
|
| 1049 |
+
"grad_norm": 2.957026481628418,
|
| 1050 |
+
"learning_rate": 9.639155284316576e-06,
|
| 1051 |
+
"loss": 4.0764,
|
| 1052 |
+
"step": 74500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.6326649229875034,
|
| 1056 |
+
"grad_norm": 3.0738115310668945,
|
| 1057 |
+
"learning_rate": 9.636733507701251e-06,
|
| 1058 |
+
"loss": 4.0742,
|
| 1059 |
+
"step": 75000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 3.6568826891407538,
|
| 1063 |
+
"grad_norm": 2.877403736114502,
|
| 1064 |
+
"learning_rate": 9.634311731085926e-06,
|
| 1065 |
+
"loss": 4.0744,
|
| 1066 |
+
"step": 75500
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.6811004552940036,
|
| 1070 |
+
"grad_norm": 3.0667495727539062,
|
| 1071 |
+
"learning_rate": 9.6318899544706e-06,
|
| 1072 |
+
"loss": 4.0715,
|
| 1073 |
+
"step": 76000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.7053182214472535,
|
| 1077 |
+
"grad_norm": 2.8147807121276855,
|
| 1078 |
+
"learning_rate": 9.629468177855276e-06,
|
| 1079 |
+
"loss": 4.0666,
|
| 1080 |
+
"step": 76500
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 3.729535987600504,
|
| 1084 |
+
"grad_norm": 2.8717801570892334,
|
| 1085 |
+
"learning_rate": 9.62704640123995e-06,
|
| 1086 |
+
"loss": 4.064,
|
| 1087 |
+
"step": 77000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 3.7537537537537538,
|
| 1091 |
+
"grad_norm": 2.7591042518615723,
|
| 1092 |
+
"learning_rate": 9.624624624624626e-06,
|
| 1093 |
+
"loss": 4.0557,
|
| 1094 |
+
"step": 77500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 3.7779715199070036,
|
| 1098 |
+
"grad_norm": 2.843806743621826,
|
| 1099 |
+
"learning_rate": 9.6222028480093e-06,
|
| 1100 |
+
"loss": 4.0579,
|
| 1101 |
+
"step": 78000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 3.802189286060254,
|
| 1105 |
+
"grad_norm": 2.869080066680908,
|
| 1106 |
+
"learning_rate": 9.619781071393975e-06,
|
| 1107 |
+
"loss": 4.0537,
|
| 1108 |
+
"step": 78500
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 3.826407052213504,
|
| 1112 |
+
"grad_norm": 2.792863607406616,
|
| 1113 |
+
"learning_rate": 9.617359294778652e-06,
|
| 1114 |
+
"loss": 4.0484,
|
| 1115 |
+
"step": 79000
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.8506248183667537,
|
| 1119 |
+
"grad_norm": 2.991138458251953,
|
| 1120 |
+
"learning_rate": 9.614937518163325e-06,
|
| 1121 |
+
"loss": 4.039,
|
| 1122 |
+
"step": 79500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 3.874842584520004,
|
| 1126 |
+
"grad_norm": 2.7616770267486572,
|
| 1127 |
+
"learning_rate": 9.612515741548002e-06,
|
| 1128 |
+
"loss": 4.0438,
|
| 1129 |
+
"step": 80000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 3.899060350673254,
|
| 1133 |
+
"grad_norm": 2.718642234802246,
|
| 1134 |
+
"learning_rate": 9.610093964932675e-06,
|
| 1135 |
+
"loss": 4.0333,
|
| 1136 |
+
"step": 80500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 3.923278116826504,
|
| 1140 |
+
"grad_norm": 2.8432154655456543,
|
| 1141 |
+
"learning_rate": 9.60767218831735e-06,
|
| 1142 |
+
"loss": 4.0419,
|
| 1143 |
+
"step": 81000
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.947495882979754,
|
| 1147 |
+
"grad_norm": 3.018446683883667,
|
| 1148 |
+
"learning_rate": 9.605250411702025e-06,
|
| 1149 |
+
"loss": 4.0374,
|
| 1150 |
+
"step": 81500
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 3.971713649133004,
|
| 1154 |
+
"grad_norm": 2.909247636795044,
|
| 1155 |
+
"learning_rate": 9.6028286350867e-06,
|
| 1156 |
+
"loss": 4.0299,
|
| 1157 |
+
"step": 82000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.995931415286254,
|
| 1161 |
+
"grad_norm": 3.047041654586792,
|
| 1162 |
+
"learning_rate": 9.600406858471375e-06,
|
| 1163 |
+
"loss": 4.0195,
|
| 1164 |
+
"step": 82500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 4.020149181439504,
|
| 1168 |
+
"grad_norm": 2.8578057289123535,
|
| 1169 |
+
"learning_rate": 9.59798508185605e-06,
|
| 1170 |
+
"loss": 4.0212,
|
| 1171 |
+
"step": 83000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 4.044366947592754,
|
| 1175 |
+
"grad_norm": 2.8038136959075928,
|
| 1176 |
+
"learning_rate": 9.595563305240725e-06,
|
| 1177 |
+
"loss": 4.0196,
|
| 1178 |
+
"step": 83500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 4.068584713746004,
|
| 1182 |
+
"grad_norm": 2.879891872406006,
|
| 1183 |
+
"learning_rate": 9.5931415286254e-06,
|
| 1184 |
+
"loss": 4.0127,
|
| 1185 |
+
"step": 84000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 4.092802479899254,
|
| 1189 |
+
"grad_norm": 2.875603437423706,
|
| 1190 |
+
"learning_rate": 9.590719752010075e-06,
|
| 1191 |
+
"loss": 4.0142,
|
| 1192 |
+
"step": 84500
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 4.117020246052504,
|
| 1196 |
+
"grad_norm": 2.975302219390869,
|
| 1197 |
+
"learning_rate": 9.58829797539475e-06,
|
| 1198 |
+
"loss": 4.0002,
|
| 1199 |
+
"step": 85000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 4.141238012205754,
|
| 1203 |
+
"grad_norm": 2.9974005222320557,
|
| 1204 |
+
"learning_rate": 9.585876198779426e-06,
|
| 1205 |
+
"loss": 4.0038,
|
| 1206 |
+
"step": 85500
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 4.1654557783590045,
|
| 1210 |
+
"grad_norm": 2.8580379486083984,
|
| 1211 |
+
"learning_rate": 9.5834544221641e-06,
|
| 1212 |
+
"loss": 4.0003,
|
| 1213 |
+
"step": 86000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 4.189673544512254,
|
| 1217 |
+
"grad_norm": 2.987436056137085,
|
| 1218 |
+
"learning_rate": 9.581032645548776e-06,
|
| 1219 |
+
"loss": 4.005,
|
| 1220 |
+
"step": 86500
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.213891310665504,
|
| 1224 |
+
"grad_norm": 2.6872076988220215,
|
| 1225 |
+
"learning_rate": 9.57861086893345e-06,
|
| 1226 |
+
"loss": 3.9909,
|
| 1227 |
+
"step": 87000
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 4.238109076818755,
|
| 1231 |
+
"grad_norm": 2.991762638092041,
|
| 1232 |
+
"learning_rate": 9.576189092318126e-06,
|
| 1233 |
+
"loss": 3.9897,
|
| 1234 |
+
"step": 87500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 4.262326842972004,
|
| 1238 |
+
"grad_norm": 2.8275723457336426,
|
| 1239 |
+
"learning_rate": 9.5737673157028e-06,
|
| 1240 |
+
"loss": 3.996,
|
| 1241 |
+
"step": 88000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 4.286544609125254,
|
| 1245 |
+
"grad_norm": 2.892839193344116,
|
| 1246 |
+
"learning_rate": 9.571345539087476e-06,
|
| 1247 |
+
"loss": 3.9946,
|
| 1248 |
+
"step": 88500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 4.310762375278505,
|
| 1252 |
+
"grad_norm": 2.8410208225250244,
|
| 1253 |
+
"learning_rate": 9.56892376247215e-06,
|
| 1254 |
+
"loss": 3.9862,
|
| 1255 |
+
"step": 89000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 4.334980141431754,
|
| 1259 |
+
"grad_norm": 2.797422409057617,
|
| 1260 |
+
"learning_rate": 9.566501985856826e-06,
|
| 1261 |
+
"loss": 3.9843,
|
| 1262 |
+
"step": 89500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.3591979075850045,
|
| 1266 |
+
"grad_norm": 2.855832099914551,
|
| 1267 |
+
"learning_rate": 9.564080209241501e-06,
|
| 1268 |
+
"loss": 3.9796,
|
| 1269 |
+
"step": 90000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 4.383415673738254,
|
| 1273 |
+
"grad_norm": 3.0120160579681396,
|
| 1274 |
+
"learning_rate": 9.561658432626174e-06,
|
| 1275 |
+
"loss": 3.9768,
|
| 1276 |
+
"step": 90500
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 4.407633439891504,
|
| 1280 |
+
"grad_norm": 2.7952980995178223,
|
| 1281 |
+
"learning_rate": 9.559236656010851e-06,
|
| 1282 |
+
"loss": 3.9742,
|
| 1283 |
+
"step": 91000
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 4.431851206044755,
|
| 1287 |
+
"grad_norm": 2.8430566787719727,
|
| 1288 |
+
"learning_rate": 9.556814879395525e-06,
|
| 1289 |
+
"loss": 3.9741,
|
| 1290 |
+
"step": 91500
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 4.456068972198004,
|
| 1294 |
+
"grad_norm": 2.9674031734466553,
|
| 1295 |
+
"learning_rate": 9.554393102780201e-06,
|
| 1296 |
+
"loss": 3.9697,
|
| 1297 |
+
"step": 92000
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 4.480286738351254,
|
| 1301 |
+
"grad_norm": 3.0408644676208496,
|
| 1302 |
+
"learning_rate": 9.551971326164875e-06,
|
| 1303 |
+
"loss": 3.9624,
|
| 1304 |
+
"step": 92500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.504504504504505,
|
| 1308 |
+
"grad_norm": 2.9981327056884766,
|
| 1309 |
+
"learning_rate": 9.54954954954955e-06,
|
| 1310 |
+
"loss": 3.9652,
|
| 1311 |
+
"step": 93000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 4.528722270657754,
|
| 1315 |
+
"grad_norm": 2.7843706607818604,
|
| 1316 |
+
"learning_rate": 9.547127772934225e-06,
|
| 1317 |
+
"loss": 3.9722,
|
| 1318 |
+
"step": 93500
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 4.5529400368110045,
|
| 1322 |
+
"grad_norm": 2.7166874408721924,
|
| 1323 |
+
"learning_rate": 9.5447059963189e-06,
|
| 1324 |
+
"loss": 3.9628,
|
| 1325 |
+
"step": 94000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 4.577157802964255,
|
| 1329 |
+
"grad_norm": 2.923854351043701,
|
| 1330 |
+
"learning_rate": 9.542284219703575e-06,
|
| 1331 |
+
"loss": 3.9594,
|
| 1332 |
+
"step": 94500
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 4.601375569117504,
|
| 1336 |
+
"grad_norm": 2.915800094604492,
|
| 1337 |
+
"learning_rate": 9.53986244308825e-06,
|
| 1338 |
+
"loss": 3.9609,
|
| 1339 |
+
"step": 95000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 4.625593335270755,
|
| 1343 |
+
"grad_norm": 2.9524765014648438,
|
| 1344 |
+
"learning_rate": 9.537440666472925e-06,
|
| 1345 |
+
"loss": 3.9587,
|
| 1346 |
+
"step": 95500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 4.649811101424005,
|
| 1350 |
+
"grad_norm": 2.898005723953247,
|
| 1351 |
+
"learning_rate": 9.5350188898576e-06,
|
| 1352 |
+
"loss": 3.9498,
|
| 1353 |
+
"step": 96000
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 4.674028867577254,
|
| 1357 |
+
"grad_norm": 2.9840903282165527,
|
| 1358 |
+
"learning_rate": 9.532597113242275e-06,
|
| 1359 |
+
"loss": 3.9508,
|
| 1360 |
+
"step": 96500
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 4.698246633730505,
|
| 1364 |
+
"grad_norm": 2.7765541076660156,
|
| 1365 |
+
"learning_rate": 9.53017533662695e-06,
|
| 1366 |
+
"loss": 3.9481,
|
| 1367 |
+
"step": 97000
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 4.722464399883755,
|
| 1371 |
+
"grad_norm": 2.8900692462921143,
|
| 1372 |
+
"learning_rate": 9.527753560011625e-06,
|
| 1373 |
+
"loss": 3.9365,
|
| 1374 |
+
"step": 97500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 4.746682166037004,
|
| 1378 |
+
"grad_norm": 2.8892781734466553,
|
| 1379 |
+
"learning_rate": 9.5253317833963e-06,
|
| 1380 |
+
"loss": 3.9492,
|
| 1381 |
+
"step": 98000
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 4.770899932190255,
|
| 1385 |
+
"grad_norm": 2.960374355316162,
|
| 1386 |
+
"learning_rate": 9.522910006780975e-06,
|
| 1387 |
+
"loss": 3.942,
|
| 1388 |
+
"step": 98500
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 4.795117698343505,
|
| 1392 |
+
"grad_norm": 2.7404415607452393,
|
| 1393 |
+
"learning_rate": 9.52048823016565e-06,
|
| 1394 |
+
"loss": 3.938,
|
| 1395 |
+
"step": 99000
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 4.8193354644967545,
|
| 1399 |
+
"grad_norm": 3.024486780166626,
|
| 1400 |
+
"learning_rate": 9.518066453550326e-06,
|
| 1401 |
+
"loss": 3.9296,
|
| 1402 |
+
"step": 99500
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 4.843553230650005,
|
| 1406 |
+
"grad_norm": 2.8316361904144287,
|
| 1407 |
+
"learning_rate": 9.515644676935e-06,
|
| 1408 |
+
"loss": 3.9351,
|
| 1409 |
+
"step": 100000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 4.867770996803255,
|
| 1413 |
+
"grad_norm": 2.8669049739837646,
|
| 1414 |
+
"learning_rate": 9.513222900319676e-06,
|
| 1415 |
+
"loss": 3.9373,
|
| 1416 |
+
"step": 100500
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 4.891988762956505,
|
| 1420 |
+
"grad_norm": 2.7152950763702393,
|
| 1421 |
+
"learning_rate": 9.51080112370435e-06,
|
| 1422 |
+
"loss": 3.9163,
|
| 1423 |
+
"step": 101000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 4.916206529109755,
|
| 1427 |
+
"grad_norm": 2.7430613040924072,
|
| 1428 |
+
"learning_rate": 9.508379347089026e-06,
|
| 1429 |
+
"loss": 3.9273,
|
| 1430 |
+
"step": 101500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 4.940424295263005,
|
| 1434 |
+
"grad_norm": 3.0171566009521484,
|
| 1435 |
+
"learning_rate": 9.5059575704737e-06,
|
| 1436 |
+
"loss": 3.9247,
|
| 1437 |
+
"step": 102000
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 4.964642061416255,
|
| 1441 |
+
"grad_norm": 2.833829164505005,
|
| 1442 |
+
"learning_rate": 9.503535793858374e-06,
|
| 1443 |
+
"loss": 3.9307,
|
| 1444 |
+
"step": 102500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 4.988859827569505,
|
| 1448 |
+
"grad_norm": 2.7739973068237305,
|
| 1449 |
+
"learning_rate": 9.501114017243051e-06,
|
| 1450 |
+
"loss": 3.9195,
|
| 1451 |
+
"step": 103000
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 5.013077593722755,
|
| 1455 |
+
"grad_norm": 2.774411201477051,
|
| 1456 |
+
"learning_rate": 9.498692240627724e-06,
|
| 1457 |
+
"loss": 3.9116,
|
| 1458 |
+
"step": 103500
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 5.037295359876005,
|
| 1462 |
+
"grad_norm": 2.851175546646118,
|
| 1463 |
+
"learning_rate": 9.496270464012401e-06,
|
| 1464 |
+
"loss": 3.9113,
|
| 1465 |
+
"step": 104000
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.061513126029255,
|
| 1469 |
+
"grad_norm": 2.8700265884399414,
|
| 1470 |
+
"learning_rate": 9.493848687397074e-06,
|
| 1471 |
+
"loss": 3.9058,
|
| 1472 |
+
"step": 104500
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 5.0857308921825055,
|
| 1476 |
+
"grad_norm": 2.8087737560272217,
|
| 1477 |
+
"learning_rate": 9.49142691078175e-06,
|
| 1478 |
+
"loss": 3.9087,
|
| 1479 |
+
"step": 105000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 5.109948658335755,
|
| 1483 |
+
"grad_norm": 2.882826328277588,
|
| 1484 |
+
"learning_rate": 9.489005134166426e-06,
|
| 1485 |
+
"loss": 3.907,
|
| 1486 |
+
"step": 105500
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 5.134166424489005,
|
| 1490 |
+
"grad_norm": 2.900575637817383,
|
| 1491 |
+
"learning_rate": 9.4865833575511e-06,
|
| 1492 |
+
"loss": 3.9022,
|
| 1493 |
+
"step": 106000
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 5.158384190642255,
|
| 1497 |
+
"grad_norm": 2.7019128799438477,
|
| 1498 |
+
"learning_rate": 9.484161580935776e-06,
|
| 1499 |
+
"loss": 3.9017,
|
| 1500 |
+
"step": 106500
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 5.182601956795505,
|
| 1504 |
+
"grad_norm": 2.8361051082611084,
|
| 1505 |
+
"learning_rate": 9.48173980432045e-06,
|
| 1506 |
+
"loss": 3.9108,
|
| 1507 |
+
"step": 107000
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 5.206819722948755,
|
| 1511 |
+
"grad_norm": 2.741563558578491,
|
| 1512 |
+
"learning_rate": 9.479318027705125e-06,
|
| 1513 |
+
"loss": 3.8919,
|
| 1514 |
+
"step": 107500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 5.231037489102005,
|
| 1518 |
+
"grad_norm": 2.967627763748169,
|
| 1519 |
+
"learning_rate": 9.4768962510898e-06,
|
| 1520 |
+
"loss": 3.8967,
|
| 1521 |
+
"step": 108000
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 5.255255255255255,
|
| 1525 |
+
"grad_norm": 2.8605451583862305,
|
| 1526 |
+
"learning_rate": 9.474474474474475e-06,
|
| 1527 |
+
"loss": 3.897,
|
| 1528 |
+
"step": 108500
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 5.2794730214085055,
|
| 1532 |
+
"grad_norm": 2.7184574604034424,
|
| 1533 |
+
"learning_rate": 9.47205269785915e-06,
|
| 1534 |
+
"loss": 3.8975,
|
| 1535 |
+
"step": 109000
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 5.303690787561755,
|
| 1539 |
+
"grad_norm": 2.7433297634124756,
|
| 1540 |
+
"learning_rate": 9.469630921243825e-06,
|
| 1541 |
+
"loss": 3.8954,
|
| 1542 |
+
"step": 109500
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 5.327908553715005,
|
| 1546 |
+
"grad_norm": 2.7750015258789062,
|
| 1547 |
+
"learning_rate": 9.4672091446285e-06,
|
| 1548 |
+
"loss": 3.8913,
|
| 1549 |
+
"step": 110000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 5.352126319868256,
|
| 1553 |
+
"grad_norm": 2.9533851146698,
|
| 1554 |
+
"learning_rate": 9.464787368013175e-06,
|
| 1555 |
+
"loss": 3.8844,
|
| 1556 |
+
"step": 110500
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 5.376344086021505,
|
| 1560 |
+
"grad_norm": 2.8131632804870605,
|
| 1561 |
+
"learning_rate": 9.46236559139785e-06,
|
| 1562 |
+
"loss": 3.8809,
|
| 1563 |
+
"step": 111000
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 5.400561852174755,
|
| 1567 |
+
"grad_norm": 2.791193723678589,
|
| 1568 |
+
"learning_rate": 9.459943814782525e-06,
|
| 1569 |
+
"loss": 3.885,
|
| 1570 |
+
"step": 111500
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 5.424779618328006,
|
| 1574 |
+
"grad_norm": 2.869932174682617,
|
| 1575 |
+
"learning_rate": 9.4575220381672e-06,
|
| 1576 |
+
"loss": 3.8839,
|
| 1577 |
+
"step": 112000
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 5.448997384481255,
|
| 1581 |
+
"grad_norm": 2.906806707382202,
|
| 1582 |
+
"learning_rate": 9.455100261551875e-06,
|
| 1583 |
+
"loss": 3.8816,
|
| 1584 |
+
"step": 112500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 5.4732151506345055,
|
| 1588 |
+
"grad_norm": 2.6837105751037598,
|
| 1589 |
+
"learning_rate": 9.45267848493655e-06,
|
| 1590 |
+
"loss": 3.88,
|
| 1591 |
+
"step": 113000
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 5.497432916787756,
|
| 1595 |
+
"grad_norm": 2.9571547508239746,
|
| 1596 |
+
"learning_rate": 9.450256708321225e-06,
|
| 1597 |
+
"loss": 3.877,
|
| 1598 |
+
"step": 113500
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 5.521650682941005,
|
| 1602 |
+
"grad_norm": 2.706204891204834,
|
| 1603 |
+
"learning_rate": 9.4478349317059e-06,
|
| 1604 |
+
"loss": 3.8772,
|
| 1605 |
+
"step": 114000
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 5.545868449094256,
|
| 1609 |
+
"grad_norm": 2.7605583667755127,
|
| 1610 |
+
"learning_rate": 9.445413155090576e-06,
|
| 1611 |
+
"loss": 3.875,
|
| 1612 |
+
"step": 114500
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 5.570086215247506,
|
| 1616 |
+
"grad_norm": 2.711221933364868,
|
| 1617 |
+
"learning_rate": 9.44299137847525e-06,
|
| 1618 |
+
"loss": 3.8712,
|
| 1619 |
+
"step": 115000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 5.594303981400755,
|
| 1623 |
+
"grad_norm": 2.9056496620178223,
|
| 1624 |
+
"learning_rate": 9.440569601859924e-06,
|
| 1625 |
+
"loss": 3.8639,
|
| 1626 |
+
"step": 115500
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 5.618521747554006,
|
| 1630 |
+
"grad_norm": 2.7061548233032227,
|
| 1631 |
+
"learning_rate": 9.4381478252446e-06,
|
| 1632 |
+
"loss": 3.8677,
|
| 1633 |
+
"step": 116000
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 5.642739513707256,
|
| 1637 |
+
"grad_norm": 2.9951186180114746,
|
| 1638 |
+
"learning_rate": 9.435726048629276e-06,
|
| 1639 |
+
"loss": 3.8667,
|
| 1640 |
+
"step": 116500
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 5.6669572798605055,
|
| 1644 |
+
"grad_norm": 2.753833293914795,
|
| 1645 |
+
"learning_rate": 9.43330427201395e-06,
|
| 1646 |
+
"loss": 3.8669,
|
| 1647 |
+
"step": 117000
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 5.691175046013756,
|
| 1651 |
+
"grad_norm": 2.8989222049713135,
|
| 1652 |
+
"learning_rate": 9.430882495398626e-06,
|
| 1653 |
+
"loss": 3.8751,
|
| 1654 |
+
"step": 117500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 5.715392812167005,
|
| 1658 |
+
"grad_norm": 3.0137453079223633,
|
| 1659 |
+
"learning_rate": 9.4284607187833e-06,
|
| 1660 |
+
"loss": 3.8706,
|
| 1661 |
+
"step": 118000
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 5.739610578320256,
|
| 1665 |
+
"grad_norm": 2.7698261737823486,
|
| 1666 |
+
"learning_rate": 9.426038942167976e-06,
|
| 1667 |
+
"loss": 3.857,
|
| 1668 |
+
"step": 118500
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 5.763828344473506,
|
| 1672 |
+
"grad_norm": 2.877211332321167,
|
| 1673 |
+
"learning_rate": 9.42361716555265e-06,
|
| 1674 |
+
"loss": 3.8541,
|
| 1675 |
+
"step": 119000
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 5.788046110626755,
|
| 1679 |
+
"grad_norm": 2.8494150638580322,
|
| 1680 |
+
"learning_rate": 9.421195388937326e-06,
|
| 1681 |
+
"loss": 3.8594,
|
| 1682 |
+
"step": 119500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 5.812263876780006,
|
| 1686 |
+
"grad_norm": 2.72268009185791,
|
| 1687 |
+
"learning_rate": 9.418773612322e-06,
|
| 1688 |
+
"loss": 3.8526,
|
| 1689 |
+
"step": 120000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 5.836481642933256,
|
| 1693 |
+
"grad_norm": 3.0423946380615234,
|
| 1694 |
+
"learning_rate": 9.416351835706675e-06,
|
| 1695 |
+
"loss": 3.8558,
|
| 1696 |
+
"step": 120500
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 5.860699409086505,
|
| 1700 |
+
"grad_norm": 2.7056820392608643,
|
| 1701 |
+
"learning_rate": 9.413930059091351e-06,
|
| 1702 |
+
"loss": 3.8494,
|
| 1703 |
+
"step": 121000
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 5.884917175239756,
|
| 1707 |
+
"grad_norm": 2.7295594215393066,
|
| 1708 |
+
"learning_rate": 9.411508282476025e-06,
|
| 1709 |
+
"loss": 3.8557,
|
| 1710 |
+
"step": 121500
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 5.909134941393006,
|
| 1714 |
+
"grad_norm": 2.8661701679229736,
|
| 1715 |
+
"learning_rate": 9.409086505860701e-06,
|
| 1716 |
+
"loss": 3.8528,
|
| 1717 |
+
"step": 122000
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 5.9333527075462555,
|
| 1721 |
+
"grad_norm": 2.8183608055114746,
|
| 1722 |
+
"learning_rate": 9.406664729245375e-06,
|
| 1723 |
+
"loss": 3.8511,
|
| 1724 |
+
"step": 122500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 5.957570473699506,
|
| 1728 |
+
"grad_norm": 2.974858283996582,
|
| 1729 |
+
"learning_rate": 9.40424295263005e-06,
|
| 1730 |
+
"loss": 3.8438,
|
| 1731 |
+
"step": 123000
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 5.981788239852756,
|
| 1735 |
+
"grad_norm": 2.8071188926696777,
|
| 1736 |
+
"learning_rate": 9.401821176014725e-06,
|
| 1737 |
+
"loss": 3.8338,
|
| 1738 |
+
"step": 123500
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 6.006006006006006,
|
| 1742 |
+
"grad_norm": 2.679610252380371,
|
| 1743 |
+
"learning_rate": 9.3993993993994e-06,
|
| 1744 |
+
"loss": 3.8432,
|
| 1745 |
+
"step": 124000
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 6.030223772159256,
|
| 1749 |
+
"grad_norm": 2.7918217182159424,
|
| 1750 |
+
"learning_rate": 9.396977622784075e-06,
|
| 1751 |
+
"loss": 3.8359,
|
| 1752 |
+
"step": 124500
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 6.054441538312506,
|
| 1756 |
+
"grad_norm": 2.9353878498077393,
|
| 1757 |
+
"learning_rate": 9.39455584616875e-06,
|
| 1758 |
+
"loss": 3.8423,
|
| 1759 |
+
"step": 125000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 6.078659304465756,
|
| 1763 |
+
"grad_norm": 2.7717785835266113,
|
| 1764 |
+
"learning_rate": 9.392134069553425e-06,
|
| 1765 |
+
"loss": 3.8362,
|
| 1766 |
+
"step": 125500
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 6.102877070619006,
|
| 1770 |
+
"grad_norm": 2.8372817039489746,
|
| 1771 |
+
"learning_rate": 9.3897122929381e-06,
|
| 1772 |
+
"loss": 3.8308,
|
| 1773 |
+
"step": 126000
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 6.127094836772256,
|
| 1777 |
+
"grad_norm": 2.823821544647217,
|
| 1778 |
+
"learning_rate": 9.387290516322775e-06,
|
| 1779 |
+
"loss": 3.8314,
|
| 1780 |
+
"step": 126500
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 6.151312602925506,
|
| 1784 |
+
"grad_norm": 2.7613956928253174,
|
| 1785 |
+
"learning_rate": 9.38486873970745e-06,
|
| 1786 |
+
"loss": 3.8317,
|
| 1787 |
+
"step": 127000
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 6.175530369078756,
|
| 1791 |
+
"grad_norm": 2.745297431945801,
|
| 1792 |
+
"learning_rate": 9.382446963092125e-06,
|
| 1793 |
+
"loss": 3.8274,
|
| 1794 |
+
"step": 127500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 6.1997481352320065,
|
| 1798 |
+
"grad_norm": 2.7873809337615967,
|
| 1799 |
+
"learning_rate": 9.3800251864768e-06,
|
| 1800 |
+
"loss": 3.8203,
|
| 1801 |
+
"step": 128000
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 6.223965901385256,
|
| 1805 |
+
"grad_norm": 2.871760606765747,
|
| 1806 |
+
"learning_rate": 9.377603409861475e-06,
|
| 1807 |
+
"loss": 3.8356,
|
| 1808 |
+
"step": 128500
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 6.248183667538506,
|
| 1812 |
+
"grad_norm": 2.788484811782837,
|
| 1813 |
+
"learning_rate": 9.37518163324615e-06,
|
| 1814 |
+
"loss": 3.8239,
|
| 1815 |
+
"step": 129000
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 6.272401433691757,
|
| 1819 |
+
"grad_norm": 2.7170634269714355,
|
| 1820 |
+
"learning_rate": 9.372759856630826e-06,
|
| 1821 |
+
"loss": 3.8208,
|
| 1822 |
+
"step": 129500
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 6.296619199845006,
|
| 1826 |
+
"grad_norm": 2.8259615898132324,
|
| 1827 |
+
"learning_rate": 9.370338080015499e-06,
|
| 1828 |
+
"loss": 3.8241,
|
| 1829 |
+
"step": 130000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 6.320836965998256,
|
| 1833 |
+
"grad_norm": 2.9876561164855957,
|
| 1834 |
+
"learning_rate": 9.367916303400176e-06,
|
| 1835 |
+
"loss": 3.8259,
|
| 1836 |
+
"step": 130500
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 6.345054732151507,
|
| 1840 |
+
"grad_norm": 2.844414710998535,
|
| 1841 |
+
"learning_rate": 9.365494526784849e-06,
|
| 1842 |
+
"loss": 3.8173,
|
| 1843 |
+
"step": 131000
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 6.369272498304756,
|
| 1847 |
+
"grad_norm": 2.8100152015686035,
|
| 1848 |
+
"learning_rate": 9.363072750169526e-06,
|
| 1849 |
+
"loss": 3.8245,
|
| 1850 |
+
"step": 131500
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 6.3934902644580065,
|
| 1854 |
+
"grad_norm": 2.8489105701446533,
|
| 1855 |
+
"learning_rate": 9.360650973554201e-06,
|
| 1856 |
+
"loss": 3.8147,
|
| 1857 |
+
"step": 132000
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 6.417708030611257,
|
| 1861 |
+
"grad_norm": 2.8502120971679688,
|
| 1862 |
+
"learning_rate": 9.358229196938874e-06,
|
| 1863 |
+
"loss": 3.8122,
|
| 1864 |
+
"step": 132500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 6.441925796764506,
|
| 1868 |
+
"grad_norm": 2.9556784629821777,
|
| 1869 |
+
"learning_rate": 9.355807420323551e-06,
|
| 1870 |
+
"loss": 3.8131,
|
| 1871 |
+
"step": 133000
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 6.466143562917757,
|
| 1875 |
+
"grad_norm": 2.762270212173462,
|
| 1876 |
+
"learning_rate": 9.353385643708224e-06,
|
| 1877 |
+
"loss": 3.807,
|
| 1878 |
+
"step": 133500
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 6.490361329071007,
|
| 1882 |
+
"grad_norm": 2.7611629962921143,
|
| 1883 |
+
"learning_rate": 9.350963867092901e-06,
|
| 1884 |
+
"loss": 3.8113,
|
| 1885 |
+
"step": 134000
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 6.514579095224256,
|
| 1889 |
+
"grad_norm": 2.738227605819702,
|
| 1890 |
+
"learning_rate": 9.348542090477574e-06,
|
| 1891 |
+
"loss": 3.8137,
|
| 1892 |
+
"step": 134500
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 6.538796861377507,
|
| 1896 |
+
"grad_norm": 2.822857618331909,
|
| 1897 |
+
"learning_rate": 9.34612031386225e-06,
|
| 1898 |
+
"loss": 3.7986,
|
| 1899 |
+
"step": 135000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 6.563014627530757,
|
| 1903 |
+
"grad_norm": 2.731264591217041,
|
| 1904 |
+
"learning_rate": 9.343698537246925e-06,
|
| 1905 |
+
"loss": 3.808,
|
| 1906 |
+
"step": 135500
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 6.5872323936840065,
|
| 1910 |
+
"grad_norm": 2.699312448501587,
|
| 1911 |
+
"learning_rate": 9.3412767606316e-06,
|
| 1912 |
+
"loss": 3.803,
|
| 1913 |
+
"step": 136000
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 6.611450159837257,
|
| 1917 |
+
"grad_norm": 2.8332924842834473,
|
| 1918 |
+
"learning_rate": 9.338854984016275e-06,
|
| 1919 |
+
"loss": 3.8153,
|
| 1920 |
+
"step": 136500
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 6.635667925990506,
|
| 1924 |
+
"grad_norm": 2.8680782318115234,
|
| 1925 |
+
"learning_rate": 9.33643320740095e-06,
|
| 1926 |
+
"loss": 3.7969,
|
| 1927 |
+
"step": 137000
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 6.659885692143757,
|
| 1931 |
+
"grad_norm": 2.844148635864258,
|
| 1932 |
+
"learning_rate": 9.334011430785625e-06,
|
| 1933 |
+
"loss": 3.8032,
|
| 1934 |
+
"step": 137500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 6.684103458297007,
|
| 1938 |
+
"grad_norm": 2.7583229541778564,
|
| 1939 |
+
"learning_rate": 9.3315896541703e-06,
|
| 1940 |
+
"loss": 3.8099,
|
| 1941 |
+
"step": 138000
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 6.708321224450256,
|
| 1945 |
+
"grad_norm": 2.8120036125183105,
|
| 1946 |
+
"learning_rate": 9.329167877554975e-06,
|
| 1947 |
+
"loss": 3.799,
|
| 1948 |
+
"step": 138500
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 6.732538990603507,
|
| 1952 |
+
"grad_norm": 2.8804004192352295,
|
| 1953 |
+
"learning_rate": 9.32674610093965e-06,
|
| 1954 |
+
"loss": 3.7951,
|
| 1955 |
+
"step": 139000
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 6.756756756756757,
|
| 1959 |
+
"grad_norm": 2.8031482696533203,
|
| 1960 |
+
"learning_rate": 9.324324324324325e-06,
|
| 1961 |
+
"loss": 3.7966,
|
| 1962 |
+
"step": 139500
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 6.7809745229100065,
|
| 1966 |
+
"grad_norm": 2.7140092849731445,
|
| 1967 |
+
"learning_rate": 9.321902547709e-06,
|
| 1968 |
+
"loss": 3.7959,
|
| 1969 |
+
"step": 140000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 6.7809745229100065,
|
| 1973 |
+
"step": 140000,
|
| 1974 |
+
"total_flos": 5.861064073050849e+17,
|
| 1975 |
+
"train_loss": 4.399306026785714,
|
| 1976 |
+
"train_runtime": 93598.1238,
|
| 1977 |
+
"train_samples_per_second": 705.829,
|
| 1978 |
+
"train_steps_per_second": 22.058
|
| 1979 |
+
}
|
| 1980 |
+
],
|
| 1981 |
+
"logging_steps": 500,
|
| 1982 |
+
"max_steps": 2064600,
|
| 1983 |
+
"num_input_tokens_seen": 0,
|
| 1984 |
+
"num_train_epochs": 100,
|
| 1985 |
+
"save_steps": 1000000,
|
| 1986 |
+
"stateful_callbacks": {
|
| 1987 |
+
"TrainerControl": {
|
| 1988 |
+
"args": {
|
| 1989 |
+
"should_epoch_stop": false,
|
| 1990 |
+
"should_evaluate": false,
|
| 1991 |
+
"should_log": false,
|
| 1992 |
+
"should_save": false,
|
| 1993 |
+
"should_training_stop": false
|
| 1994 |
+
},
|
| 1995 |
+
"attributes": {}
|
| 1996 |
+
}
|
| 1997 |
+
},
|
| 1998 |
+
"total_flos": 5.861064073050849e+17,
|
| 1999 |
+
"train_batch_size": 32,
|
| 2000 |
+
"trial_name": null,
|
| 2001 |
+
"trial_params": null
|
| 2002 |
+
}
|