Model save
Browse files- README.md +60 -0
- all_results.json +9 -0
- generation_config.json +6 -0
- runs/Sep27_13-50-34_COE-CS-sv003/events.out.tfevents.1727445412.COE-CS-sv003.923912.0 +2 -2
- train_results.json +9 -0
- trainer_state.json +3444 -0
README.md
ADDED
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: transformers
|
3 |
+
license: apache-2.0
|
4 |
+
base_model: mistralai/Mistral-7B-Instruct-v0.2
|
5 |
+
tags:
|
6 |
+
- trl
|
7 |
+
- sft
|
8 |
+
- generated_from_trainer
|
9 |
+
model-index:
|
10 |
+
- name: preference_p0.1_seed42_level2_stylecleanbatch16_reverse
|
11 |
+
results: []
|
12 |
+
---
|
13 |
+
|
14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
15 |
+
should probably proofread and complete it, then remove this comment. -->
|
16 |
+
|
17 |
+
# preference_p0.1_seed42_level2_stylecleanbatch16_reverse
|
18 |
+
|
19 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the None dataset.
|
20 |
+
|
21 |
+
## Model description
|
22 |
+
|
23 |
+
More information needed
|
24 |
+
|
25 |
+
## Intended uses & limitations
|
26 |
+
|
27 |
+
More information needed
|
28 |
+
|
29 |
+
## Training and evaluation data
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Training procedure
|
34 |
+
|
35 |
+
### Training hyperparameters
|
36 |
+
|
37 |
+
The following hyperparameters were used during training:
|
38 |
+
- learning_rate: 1e-05
|
39 |
+
- train_batch_size: 4
|
40 |
+
- eval_batch_size: 1
|
41 |
+
- seed: 42
|
42 |
+
- distributed_type: multi-GPU
|
43 |
+
- num_devices: 4
|
44 |
+
- total_train_batch_size: 16
|
45 |
+
- total_eval_batch_size: 4
|
46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
+
- lr_scheduler_type: cosine
|
48 |
+
- lr_scheduler_warmup_ratio: 0.1
|
49 |
+
- num_epochs: 1
|
50 |
+
|
51 |
+
### Training results
|
52 |
+
|
53 |
+
|
54 |
+
|
55 |
+
### Framework versions
|
56 |
+
|
57 |
+
- Transformers 4.45.1
|
58 |
+
- Pytorch 2.4.1+cu121
|
59 |
+
- Datasets 3.0.1
|
60 |
+
- Tokenizers 0.20.0
|
all_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"total_flos": 253768142684160.0,
|
4 |
+
"train_loss": 0.5300087753695624,
|
5 |
+
"train_runtime": 21368.6318,
|
6 |
+
"train_samples": 98881,
|
7 |
+
"train_samples_per_second": 1.814,
|
8 |
+
"train_steps_per_second": 0.113
|
9 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_from_model_config": true,
|
3 |
+
"bos_token_id": 1,
|
4 |
+
"eos_token_id": 2,
|
5 |
+
"transformers_version": "4.45.1"
|
6 |
+
}
|
runs/Sep27_13-50-34_COE-CS-sv003/events.out.tfevents.1727445412.COE-CS-sv003.923912.0
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f4a6ab66225d932b0f9a6c8fa8f5d1a4d106c1166486c83ce85a2dc3f9c0e60a
|
3 |
+
size 108645
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"total_flos": 253768142684160.0,
|
4 |
+
"train_loss": 0.5300087753695624,
|
5 |
+
"train_runtime": 21368.6318,
|
6 |
+
"train_samples": 98881,
|
7 |
+
"train_samples_per_second": 1.814,
|
8 |
+
"train_steps_per_second": 0.113
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,3444 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 1.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 2424,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.00041254125412541255,
|
13 |
+
"grad_norm": 25.173291317801098,
|
14 |
+
"learning_rate": 4.1152263374485605e-08,
|
15 |
+
"loss": 1.4315,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.0020627062706270625,
|
20 |
+
"grad_norm": 23.244172144374918,
|
21 |
+
"learning_rate": 2.05761316872428e-07,
|
22 |
+
"loss": 1.4238,
|
23 |
+
"step": 5
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.004125412541254125,
|
27 |
+
"grad_norm": 15.407649804349377,
|
28 |
+
"learning_rate": 4.11522633744856e-07,
|
29 |
+
"loss": 1.4148,
|
30 |
+
"step": 10
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.006188118811881188,
|
34 |
+
"grad_norm": 8.402022752616448,
|
35 |
+
"learning_rate": 6.17283950617284e-07,
|
36 |
+
"loss": 1.278,
|
37 |
+
"step": 15
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.00825082508250825,
|
41 |
+
"grad_norm": 11.034069123055462,
|
42 |
+
"learning_rate": 8.23045267489712e-07,
|
43 |
+
"loss": 1.1772,
|
44 |
+
"step": 20
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.010313531353135313,
|
48 |
+
"grad_norm": 4.332735330250593,
|
49 |
+
"learning_rate": 1.02880658436214e-06,
|
50 |
+
"loss": 1.0531,
|
51 |
+
"step": 25
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.012376237623762377,
|
55 |
+
"grad_norm": 3.55215196003044,
|
56 |
+
"learning_rate": 1.234567901234568e-06,
|
57 |
+
"loss": 0.9881,
|
58 |
+
"step": 30
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.014438943894389438,
|
62 |
+
"grad_norm": 3.3491262120808307,
|
63 |
+
"learning_rate": 1.440329218106996e-06,
|
64 |
+
"loss": 0.9604,
|
65 |
+
"step": 35
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.0165016501650165,
|
69 |
+
"grad_norm": 3.022805529228879,
|
70 |
+
"learning_rate": 1.646090534979424e-06,
|
71 |
+
"loss": 0.9382,
|
72 |
+
"step": 40
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.018564356435643563,
|
76 |
+
"grad_norm": 3.068991080972076,
|
77 |
+
"learning_rate": 1.8518518518518519e-06,
|
78 |
+
"loss": 0.9276,
|
79 |
+
"step": 45
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.020627062706270627,
|
83 |
+
"grad_norm": 3.1689853717520595,
|
84 |
+
"learning_rate": 2.05761316872428e-06,
|
85 |
+
"loss": 0.9118,
|
86 |
+
"step": 50
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.02268976897689769,
|
90 |
+
"grad_norm": 3.095012272596912,
|
91 |
+
"learning_rate": 2.263374485596708e-06,
|
92 |
+
"loss": 0.9069,
|
93 |
+
"step": 55
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.024752475247524754,
|
97 |
+
"grad_norm": 3.1980712687602972,
|
98 |
+
"learning_rate": 2.469135802469136e-06,
|
99 |
+
"loss": 0.9114,
|
100 |
+
"step": 60
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.026815181518151814,
|
104 |
+
"grad_norm": 3.0048714944711734,
|
105 |
+
"learning_rate": 2.674897119341564e-06,
|
106 |
+
"loss": 0.8915,
|
107 |
+
"step": 65
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.028877887788778877,
|
111 |
+
"grad_norm": 3.2438404533032807,
|
112 |
+
"learning_rate": 2.880658436213992e-06,
|
113 |
+
"loss": 0.9011,
|
114 |
+
"step": 70
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.03094059405940594,
|
118 |
+
"grad_norm": 3.0946505652248786,
|
119 |
+
"learning_rate": 3.08641975308642e-06,
|
120 |
+
"loss": 0.8911,
|
121 |
+
"step": 75
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.033003300330033,
|
125 |
+
"grad_norm": 2.976460498240933,
|
126 |
+
"learning_rate": 3.292181069958848e-06,
|
127 |
+
"loss": 0.8905,
|
128 |
+
"step": 80
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.03506600660066007,
|
132 |
+
"grad_norm": 2.964486117776111,
|
133 |
+
"learning_rate": 3.4979423868312762e-06,
|
134 |
+
"loss": 0.8711,
|
135 |
+
"step": 85
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.03712871287128713,
|
139 |
+
"grad_norm": 3.1472293237072018,
|
140 |
+
"learning_rate": 3.7037037037037037e-06,
|
141 |
+
"loss": 0.8776,
|
142 |
+
"step": 90
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.039191419141914194,
|
146 |
+
"grad_norm": 2.9491584240395547,
|
147 |
+
"learning_rate": 3.909465020576132e-06,
|
148 |
+
"loss": 0.8731,
|
149 |
+
"step": 95
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.041254125412541254,
|
153 |
+
"grad_norm": 3.1222334377001575,
|
154 |
+
"learning_rate": 4.11522633744856e-06,
|
155 |
+
"loss": 0.8673,
|
156 |
+
"step": 100
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.043316831683168314,
|
160 |
+
"grad_norm": 3.1224801634532744,
|
161 |
+
"learning_rate": 4.3209876543209875e-06,
|
162 |
+
"loss": 0.8568,
|
163 |
+
"step": 105
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.04537953795379538,
|
167 |
+
"grad_norm": 3.1837554638007255,
|
168 |
+
"learning_rate": 4.526748971193416e-06,
|
169 |
+
"loss": 0.8641,
|
170 |
+
"step": 110
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.04744224422442244,
|
174 |
+
"grad_norm": 3.1279379452438336,
|
175 |
+
"learning_rate": 4.732510288065844e-06,
|
176 |
+
"loss": 0.8654,
|
177 |
+
"step": 115
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.04950495049504951,
|
181 |
+
"grad_norm": 3.165713359110299,
|
182 |
+
"learning_rate": 4.938271604938272e-06,
|
183 |
+
"loss": 0.8593,
|
184 |
+
"step": 120
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.05156765676567657,
|
188 |
+
"grad_norm": 2.9608774490517282,
|
189 |
+
"learning_rate": 5.1440329218107e-06,
|
190 |
+
"loss": 0.8548,
|
191 |
+
"step": 125
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.05363036303630363,
|
195 |
+
"grad_norm": 2.9604320712666485,
|
196 |
+
"learning_rate": 5.349794238683128e-06,
|
197 |
+
"loss": 0.8506,
|
198 |
+
"step": 130
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.055693069306930694,
|
202 |
+
"grad_norm": 3.0433993058342765,
|
203 |
+
"learning_rate": 5.555555555555557e-06,
|
204 |
+
"loss": 0.8439,
|
205 |
+
"step": 135
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.057755775577557754,
|
209 |
+
"grad_norm": 3.3372432183258036,
|
210 |
+
"learning_rate": 5.761316872427984e-06,
|
211 |
+
"loss": 0.8352,
|
212 |
+
"step": 140
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.05981848184818482,
|
216 |
+
"grad_norm": 3.166742755838131,
|
217 |
+
"learning_rate": 5.967078189300412e-06,
|
218 |
+
"loss": 0.8584,
|
219 |
+
"step": 145
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.06188118811881188,
|
223 |
+
"grad_norm": 3.2086952524615326,
|
224 |
+
"learning_rate": 6.17283950617284e-06,
|
225 |
+
"loss": 0.825,
|
226 |
+
"step": 150
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.06394389438943894,
|
230 |
+
"grad_norm": 3.4577671140527557,
|
231 |
+
"learning_rate": 6.3786008230452675e-06,
|
232 |
+
"loss": 0.824,
|
233 |
+
"step": 155
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.066006600660066,
|
237 |
+
"grad_norm": 3.3939747315335613,
|
238 |
+
"learning_rate": 6.584362139917696e-06,
|
239 |
+
"loss": 0.8412,
|
240 |
+
"step": 160
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.06806930693069307,
|
244 |
+
"grad_norm": 3.1642303616218195,
|
245 |
+
"learning_rate": 6.790123456790124e-06,
|
246 |
+
"loss": 0.8372,
|
247 |
+
"step": 165
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.07013201320132013,
|
251 |
+
"grad_norm": 3.9078331437297975,
|
252 |
+
"learning_rate": 6.9958847736625525e-06,
|
253 |
+
"loss": 0.8393,
|
254 |
+
"step": 170
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.0721947194719472,
|
258 |
+
"grad_norm": 3.086012476362717,
|
259 |
+
"learning_rate": 7.201646090534981e-06,
|
260 |
+
"loss": 0.8549,
|
261 |
+
"step": 175
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.07425742574257425,
|
265 |
+
"grad_norm": 3.163457682817171,
|
266 |
+
"learning_rate": 7.4074074074074075e-06,
|
267 |
+
"loss": 0.8319,
|
268 |
+
"step": 180
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.07632013201320131,
|
272 |
+
"grad_norm": 3.1200124817434607,
|
273 |
+
"learning_rate": 7.613168724279836e-06,
|
274 |
+
"loss": 0.8229,
|
275 |
+
"step": 185
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.07838283828382839,
|
279 |
+
"grad_norm": 3.010419511128134,
|
280 |
+
"learning_rate": 7.818930041152263e-06,
|
281 |
+
"loss": 0.8329,
|
282 |
+
"step": 190
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.08044554455445545,
|
286 |
+
"grad_norm": 3.0362752294117317,
|
287 |
+
"learning_rate": 8.024691358024692e-06,
|
288 |
+
"loss": 0.8288,
|
289 |
+
"step": 195
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.08250825082508251,
|
293 |
+
"grad_norm": 3.1939826585064997,
|
294 |
+
"learning_rate": 8.23045267489712e-06,
|
295 |
+
"loss": 0.8134,
|
296 |
+
"step": 200
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.08457095709570957,
|
300 |
+
"grad_norm": 2.9811252237712056,
|
301 |
+
"learning_rate": 8.43621399176955e-06,
|
302 |
+
"loss": 0.8293,
|
303 |
+
"step": 205
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.08663366336633663,
|
307 |
+
"grad_norm": 3.2289018691242264,
|
308 |
+
"learning_rate": 8.641975308641975e-06,
|
309 |
+
"loss": 0.8108,
|
310 |
+
"step": 210
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.0886963696369637,
|
314 |
+
"grad_norm": 2.920715386036233,
|
315 |
+
"learning_rate": 8.847736625514404e-06,
|
316 |
+
"loss": 0.8199,
|
317 |
+
"step": 215
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.09075907590759076,
|
321 |
+
"grad_norm": 2.8888271523586835,
|
322 |
+
"learning_rate": 9.053497942386832e-06,
|
323 |
+
"loss": 0.831,
|
324 |
+
"step": 220
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.09282178217821782,
|
328 |
+
"grad_norm": 2.8414889865764312,
|
329 |
+
"learning_rate": 9.25925925925926e-06,
|
330 |
+
"loss": 0.8188,
|
331 |
+
"step": 225
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.09488448844884488,
|
335 |
+
"grad_norm": 3.3784549124194743,
|
336 |
+
"learning_rate": 9.465020576131688e-06,
|
337 |
+
"loss": 0.8083,
|
338 |
+
"step": 230
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.09694719471947194,
|
342 |
+
"grad_norm": 3.2323221908817996,
|
343 |
+
"learning_rate": 9.670781893004116e-06,
|
344 |
+
"loss": 0.8186,
|
345 |
+
"step": 235
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.09900990099009901,
|
349 |
+
"grad_norm": 2.8554149805575846,
|
350 |
+
"learning_rate": 9.876543209876543e-06,
|
351 |
+
"loss": 0.8035,
|
352 |
+
"step": 240
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.10107260726072607,
|
356 |
+
"grad_norm": 3.027573880753027,
|
357 |
+
"learning_rate": 9.9999792514327e-06,
|
358 |
+
"loss": 0.8091,
|
359 |
+
"step": 245
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.10313531353135313,
|
363 |
+
"grad_norm": 3.0804491286944145,
|
364 |
+
"learning_rate": 9.999745832028163e-06,
|
365 |
+
"loss": 0.7989,
|
366 |
+
"step": 250
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.1051980198019802,
|
370 |
+
"grad_norm": 3.2290033265195848,
|
371 |
+
"learning_rate": 9.999253069658074e-06,
|
372 |
+
"loss": 0.8178,
|
373 |
+
"step": 255
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.10726072607260725,
|
377 |
+
"grad_norm": 3.4736769564268233,
|
378 |
+
"learning_rate": 9.998500989882627e-06,
|
379 |
+
"loss": 0.8178,
|
380 |
+
"step": 260
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.10932343234323433,
|
384 |
+
"grad_norm": 3.0428405529671263,
|
385 |
+
"learning_rate": 9.997489631713117e-06,
|
386 |
+
"loss": 0.8006,
|
387 |
+
"step": 265
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.11138613861386139,
|
391 |
+
"grad_norm": 2.8688354965301426,
|
392 |
+
"learning_rate": 9.996219047609943e-06,
|
393 |
+
"loss": 0.7866,
|
394 |
+
"step": 270
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.11344884488448845,
|
398 |
+
"grad_norm": 3.0556606109594644,
|
399 |
+
"learning_rate": 9.99468930347986e-06,
|
400 |
+
"loss": 0.8174,
|
401 |
+
"step": 275
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.11551155115511551,
|
405 |
+
"grad_norm": 3.0647994656665603,
|
406 |
+
"learning_rate": 9.99290047867258e-06,
|
407 |
+
"loss": 0.8044,
|
408 |
+
"step": 280
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.11757425742574257,
|
412 |
+
"grad_norm": 2.856082139020889,
|
413 |
+
"learning_rate": 9.990852665976648e-06,
|
414 |
+
"loss": 0.7871,
|
415 |
+
"step": 285
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.11963696369636964,
|
419 |
+
"grad_norm": 2.894869532549818,
|
420 |
+
"learning_rate": 9.98854597161462e-06,
|
421 |
+
"loss": 0.7967,
|
422 |
+
"step": 290
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.1216996699669967,
|
426 |
+
"grad_norm": 2.9795368246449483,
|
427 |
+
"learning_rate": 9.98598051523758e-06,
|
428 |
+
"loss": 0.7987,
|
429 |
+
"step": 295
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.12376237623762376,
|
433 |
+
"grad_norm": 3.16205487552237,
|
434 |
+
"learning_rate": 9.983156429918895e-06,
|
435 |
+
"loss": 0.7675,
|
436 |
+
"step": 300
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.12582508250825084,
|
440 |
+
"grad_norm": 3.245576749246031,
|
441 |
+
"learning_rate": 9.98007386214735e-06,
|
442 |
+
"loss": 0.7915,
|
443 |
+
"step": 305
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.12788778877887788,
|
447 |
+
"grad_norm": 3.025661649853723,
|
448 |
+
"learning_rate": 9.976732971819526e-06,
|
449 |
+
"loss": 0.7895,
|
450 |
+
"step": 310
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.12995049504950495,
|
454 |
+
"grad_norm": 2.843144690280715,
|
455 |
+
"learning_rate": 9.973133932231514e-06,
|
456 |
+
"loss": 0.7782,
|
457 |
+
"step": 315
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.132013201320132,
|
461 |
+
"grad_norm": 3.105956078671983,
|
462 |
+
"learning_rate": 9.96927693006992e-06,
|
463 |
+
"loss": 0.7754,
|
464 |
+
"step": 320
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.13407590759075907,
|
468 |
+
"grad_norm": 2.7999853476012877,
|
469 |
+
"learning_rate": 9.965162165402194e-06,
|
470 |
+
"loss": 0.7652,
|
471 |
+
"step": 325
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.13613861386138615,
|
475 |
+
"grad_norm": 2.8609256056169987,
|
476 |
+
"learning_rate": 9.960789851666237e-06,
|
477 |
+
"loss": 0.7786,
|
478 |
+
"step": 330
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.1382013201320132,
|
482 |
+
"grad_norm": 3.1630409852297454,
|
483 |
+
"learning_rate": 9.956160215659342e-06,
|
484 |
+
"loss": 0.7828,
|
485 |
+
"step": 335
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.14026402640264027,
|
489 |
+
"grad_norm": 2.795363964969132,
|
490 |
+
"learning_rate": 9.951273497526423e-06,
|
491 |
+
"loss": 0.7686,
|
492 |
+
"step": 340
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.14232673267326731,
|
496 |
+
"grad_norm": 2.9053928333152723,
|
497 |
+
"learning_rate": 9.94612995074756e-06,
|
498 |
+
"loss": 0.7635,
|
499 |
+
"step": 345
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.1443894389438944,
|
503 |
+
"grad_norm": 2.7319984337656478,
|
504 |
+
"learning_rate": 9.94072984212485e-06,
|
505 |
+
"loss": 0.7396,
|
506 |
+
"step": 350
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.14645214521452146,
|
510 |
+
"grad_norm": 2.90922485756749,
|
511 |
+
"learning_rate": 9.935073451768567e-06,
|
512 |
+
"loss": 0.7675,
|
513 |
+
"step": 355
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.1485148514851485,
|
517 |
+
"grad_norm": 2.7539644778310852,
|
518 |
+
"learning_rate": 9.929161073082636e-06,
|
519 |
+
"loss": 0.7559,
|
520 |
+
"step": 360
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.15057755775577558,
|
524 |
+
"grad_norm": 2.8959145263782644,
|
525 |
+
"learning_rate": 9.922993012749413e-06,
|
526 |
+
"loss": 0.748,
|
527 |
+
"step": 365
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.15264026402640263,
|
531 |
+
"grad_norm": 2.8562652818653516,
|
532 |
+
"learning_rate": 9.916569590713775e-06,
|
533 |
+
"loss": 0.7412,
|
534 |
+
"step": 370
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.1547029702970297,
|
538 |
+
"grad_norm": 2.860687978591301,
|
539 |
+
"learning_rate": 9.90989114016652e-06,
|
540 |
+
"loss": 0.7648,
|
541 |
+
"step": 375
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.15676567656765678,
|
545 |
+
"grad_norm": 2.7743547846726995,
|
546 |
+
"learning_rate": 9.902958007527092e-06,
|
547 |
+
"loss": 0.7469,
|
548 |
+
"step": 380
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.15882838283828382,
|
552 |
+
"grad_norm": 2.933130436544695,
|
553 |
+
"learning_rate": 9.89577055242561e-06,
|
554 |
+
"loss": 0.7415,
|
555 |
+
"step": 385
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.1608910891089109,
|
559 |
+
"grad_norm": 2.777976255262721,
|
560 |
+
"learning_rate": 9.88832914768421e-06,
|
561 |
+
"loss": 0.719,
|
562 |
+
"step": 390
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.16295379537953794,
|
566 |
+
"grad_norm": 2.8659385000525983,
|
567 |
+
"learning_rate": 9.880634179297706e-06,
|
568 |
+
"loss": 0.742,
|
569 |
+
"step": 395
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.16501650165016502,
|
573 |
+
"grad_norm": 2.8897535873983418,
|
574 |
+
"learning_rate": 9.872686046413575e-06,
|
575 |
+
"loss": 0.7346,
|
576 |
+
"step": 400
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.1670792079207921,
|
580 |
+
"grad_norm": 2.8576246139096892,
|
581 |
+
"learning_rate": 9.864485161311242e-06,
|
582 |
+
"loss": 0.7325,
|
583 |
+
"step": 405
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.16914191419141913,
|
587 |
+
"grad_norm": 3.1030879684908346,
|
588 |
+
"learning_rate": 9.856031949380707e-06,
|
589 |
+
"loss": 0.7606,
|
590 |
+
"step": 410
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.1712046204620462,
|
594 |
+
"grad_norm": 2.713632760138165,
|
595 |
+
"learning_rate": 9.847326849100467e-06,
|
596 |
+
"loss": 0.7379,
|
597 |
+
"step": 415
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.17326732673267325,
|
601 |
+
"grad_norm": 2.80346541082337,
|
602 |
+
"learning_rate": 9.838370312014783e-06,
|
603 |
+
"loss": 0.7381,
|
604 |
+
"step": 420
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.17533003300330033,
|
608 |
+
"grad_norm": 3.200752732606346,
|
609 |
+
"learning_rate": 9.829162802710246e-06,
|
610 |
+
"loss": 0.7328,
|
611 |
+
"step": 425
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.1773927392739274,
|
615 |
+
"grad_norm": 3.1792821336770567,
|
616 |
+
"learning_rate": 9.819704798791691e-06,
|
617 |
+
"loss": 0.7305,
|
618 |
+
"step": 430
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.17945544554455445,
|
622 |
+
"grad_norm": 2.8356276160053424,
|
623 |
+
"learning_rate": 9.80999679085741e-06,
|
624 |
+
"loss": 0.714,
|
625 |
+
"step": 435
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.18151815181518152,
|
629 |
+
"grad_norm": 2.6413041173544043,
|
630 |
+
"learning_rate": 9.800039282473719e-06,
|
631 |
+
"loss": 0.735,
|
632 |
+
"step": 440
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.18358085808580857,
|
636 |
+
"grad_norm": 2.9983638727309856,
|
637 |
+
"learning_rate": 9.789832790148822e-06,
|
638 |
+
"loss": 0.7148,
|
639 |
+
"step": 445
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.18564356435643564,
|
643 |
+
"grad_norm": 2.995170348181315,
|
644 |
+
"learning_rate": 9.77937784330603e-06,
|
645 |
+
"loss": 0.7137,
|
646 |
+
"step": 450
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.18770627062706272,
|
650 |
+
"grad_norm": 3.0031426949210456,
|
651 |
+
"learning_rate": 9.768674984256292e-06,
|
652 |
+
"loss": 0.7334,
|
653 |
+
"step": 455
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 0.18976897689768976,
|
657 |
+
"grad_norm": 2.789041969236007,
|
658 |
+
"learning_rate": 9.757724768170074e-06,
|
659 |
+
"loss": 0.7161,
|
660 |
+
"step": 460
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 0.19183168316831684,
|
664 |
+
"grad_norm": 2.65991213841404,
|
665 |
+
"learning_rate": 9.74652776304855e-06,
|
666 |
+
"loss": 0.709,
|
667 |
+
"step": 465
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 0.19389438943894388,
|
671 |
+
"grad_norm": 2.7707492384242007,
|
672 |
+
"learning_rate": 9.73508454969415e-06,
|
673 |
+
"loss": 0.7162,
|
674 |
+
"step": 470
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 0.19595709570957096,
|
678 |
+
"grad_norm": 2.6804566008309654,
|
679 |
+
"learning_rate": 9.723395721680418e-06,
|
680 |
+
"loss": 0.7173,
|
681 |
+
"step": 475
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.19801980198019803,
|
685 |
+
"grad_norm": 2.8279805454985545,
|
686 |
+
"learning_rate": 9.711461885321247e-06,
|
687 |
+
"loss": 0.6969,
|
688 |
+
"step": 480
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 0.20008250825082508,
|
692 |
+
"grad_norm": 3.1208245898107045,
|
693 |
+
"learning_rate": 9.699283659639402e-06,
|
694 |
+
"loss": 0.7091,
|
695 |
+
"step": 485
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 0.20214521452145215,
|
699 |
+
"grad_norm": 2.6589644032154363,
|
700 |
+
"learning_rate": 9.68686167633443e-06,
|
701 |
+
"loss": 0.6952,
|
702 |
+
"step": 490
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 0.2042079207920792,
|
706 |
+
"grad_norm": 2.9084930227852137,
|
707 |
+
"learning_rate": 9.67419657974988e-06,
|
708 |
+
"loss": 0.7134,
|
709 |
+
"step": 495
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 0.20627062706270627,
|
713 |
+
"grad_norm": 2.8303925657999853,
|
714 |
+
"learning_rate": 9.661289026839889e-06,
|
715 |
+
"loss": 0.6975,
|
716 |
+
"step": 500
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 0.20833333333333334,
|
720 |
+
"grad_norm": 2.784037458254395,
|
721 |
+
"learning_rate": 9.648139687135106e-06,
|
722 |
+
"loss": 0.6979,
|
723 |
+
"step": 505
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 0.2103960396039604,
|
727 |
+
"grad_norm": 2.7696913715189897,
|
728 |
+
"learning_rate": 9.634749242707948e-06,
|
729 |
+
"loss": 0.6733,
|
730 |
+
"step": 510
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 0.21245874587458746,
|
734 |
+
"grad_norm": 2.9247340254888545,
|
735 |
+
"learning_rate": 9.62111838813724e-06,
|
736 |
+
"loss": 0.6945,
|
737 |
+
"step": 515
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 0.2145214521452145,
|
741 |
+
"grad_norm": 2.719893437871808,
|
742 |
+
"learning_rate": 9.607247830472174e-06,
|
743 |
+
"loss": 0.6761,
|
744 |
+
"step": 520
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 0.21658415841584158,
|
748 |
+
"grad_norm": 2.8726720120792386,
|
749 |
+
"learning_rate": 9.593138289195634e-06,
|
750 |
+
"loss": 0.6913,
|
751 |
+
"step": 525
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 0.21864686468646866,
|
755 |
+
"grad_norm": 2.781250153342701,
|
756 |
+
"learning_rate": 9.578790496186879e-06,
|
757 |
+
"loss": 0.6903,
|
758 |
+
"step": 530
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 0.2207095709570957,
|
762 |
+
"grad_norm": 2.7701813874151244,
|
763 |
+
"learning_rate": 9.56420519568358e-06,
|
764 |
+
"loss": 0.6831,
|
765 |
+
"step": 535
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 0.22277227722772278,
|
769 |
+
"grad_norm": 2.7057701780207983,
|
770 |
+
"learning_rate": 9.549383144243213e-06,
|
771 |
+
"loss": 0.6956,
|
772 |
+
"step": 540
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 0.22483498349834982,
|
776 |
+
"grad_norm": 2.7687957358970112,
|
777 |
+
"learning_rate": 9.534325110703814e-06,
|
778 |
+
"loss": 0.6709,
|
779 |
+
"step": 545
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 0.2268976897689769,
|
783 |
+
"grad_norm": 2.7158766270699983,
|
784 |
+
"learning_rate": 9.519031876144106e-06,
|
785 |
+
"loss": 0.6608,
|
786 |
+
"step": 550
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 0.22896039603960397,
|
790 |
+
"grad_norm": 2.7406561041756095,
|
791 |
+
"learning_rate": 9.503504233842973e-06,
|
792 |
+
"loss": 0.6857,
|
793 |
+
"step": 555
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 0.23102310231023102,
|
797 |
+
"grad_norm": 2.7133954058150525,
|
798 |
+
"learning_rate": 9.487742989238318e-06,
|
799 |
+
"loss": 0.6723,
|
800 |
+
"step": 560
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 0.2330858085808581,
|
804 |
+
"grad_norm": 2.7907877653122797,
|
805 |
+
"learning_rate": 9.471748959885284e-06,
|
806 |
+
"loss": 0.6809,
|
807 |
+
"step": 565
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 0.23514851485148514,
|
811 |
+
"grad_norm": 2.8110823806166922,
|
812 |
+
"learning_rate": 9.455522975413846e-06,
|
813 |
+
"loss": 0.6695,
|
814 |
+
"step": 570
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 0.2372112211221122,
|
818 |
+
"grad_norm": 2.9142519407957543,
|
819 |
+
"learning_rate": 9.439065877485774e-06,
|
820 |
+
"loss": 0.6506,
|
821 |
+
"step": 575
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 0.23927392739273928,
|
825 |
+
"grad_norm": 2.7148544728300195,
|
826 |
+
"learning_rate": 9.422378519750978e-06,
|
827 |
+
"loss": 0.6499,
|
828 |
+
"step": 580
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 0.24133663366336633,
|
832 |
+
"grad_norm": 2.713997914652474,
|
833 |
+
"learning_rate": 9.40546176780323e-06,
|
834 |
+
"loss": 0.6525,
|
835 |
+
"step": 585
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 0.2433993399339934,
|
839 |
+
"grad_norm": 2.692751552453192,
|
840 |
+
"learning_rate": 9.388316499135259e-06,
|
841 |
+
"loss": 0.6496,
|
842 |
+
"step": 590
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 0.24546204620462045,
|
846 |
+
"grad_norm": 2.6802330864864854,
|
847 |
+
"learning_rate": 9.370943603093235e-06,
|
848 |
+
"loss": 0.6417,
|
849 |
+
"step": 595
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 0.24752475247524752,
|
853 |
+
"grad_norm": 2.8231163075907437,
|
854 |
+
"learning_rate": 9.353343980830644e-06,
|
855 |
+
"loss": 0.652,
|
856 |
+
"step": 600
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 0.2495874587458746,
|
860 |
+
"grad_norm": 2.667983921285145,
|
861 |
+
"learning_rate": 9.33551854526154e-06,
|
862 |
+
"loss": 0.6516,
|
863 |
+
"step": 605
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 0.25165016501650167,
|
867 |
+
"grad_norm": 2.7364796827272038,
|
868 |
+
"learning_rate": 9.31746822101319e-06,
|
869 |
+
"loss": 0.6402,
|
870 |
+
"step": 610
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 0.2537128712871287,
|
874 |
+
"grad_norm": 2.6289768861588256,
|
875 |
+
"learning_rate": 9.299193944378112e-06,
|
876 |
+
"loss": 0.651,
|
877 |
+
"step": 615
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 0.25577557755775576,
|
881 |
+
"grad_norm": 3.115748743305885,
|
882 |
+
"learning_rate": 9.280696663265512e-06,
|
883 |
+
"loss": 0.6409,
|
884 |
+
"step": 620
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 0.25783828382838286,
|
888 |
+
"grad_norm": 2.7801532039806727,
|
889 |
+
"learning_rate": 9.261977337152107e-06,
|
890 |
+
"loss": 0.6232,
|
891 |
+
"step": 625
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 0.2599009900990099,
|
895 |
+
"grad_norm": 2.8268570410512415,
|
896 |
+
"learning_rate": 9.243036937032373e-06,
|
897 |
+
"loss": 0.652,
|
898 |
+
"step": 630
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 0.26196369636963696,
|
902 |
+
"grad_norm": 2.734252985059684,
|
903 |
+
"learning_rate": 9.223876445368153e-06,
|
904 |
+
"loss": 0.6408,
|
905 |
+
"step": 635
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 0.264026402640264,
|
909 |
+
"grad_norm": 2.8594124607696862,
|
910 |
+
"learning_rate": 9.204496856037718e-06,
|
911 |
+
"loss": 0.6608,
|
912 |
+
"step": 640
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 0.2660891089108911,
|
916 |
+
"grad_norm": 2.7632847146041937,
|
917 |
+
"learning_rate": 9.184899174284201e-06,
|
918 |
+
"loss": 0.6435,
|
919 |
+
"step": 645
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 0.26815181518151815,
|
923 |
+
"grad_norm": 2.7892742600413056,
|
924 |
+
"learning_rate": 9.16508441666346e-06,
|
925 |
+
"loss": 0.6382,
|
926 |
+
"step": 650
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 0.2702145214521452,
|
930 |
+
"grad_norm": 2.9224506916430926,
|
931 |
+
"learning_rate": 9.14505361099134e-06,
|
932 |
+
"loss": 0.6249,
|
933 |
+
"step": 655
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 0.2722772277227723,
|
937 |
+
"grad_norm": 2.916964064023966,
|
938 |
+
"learning_rate": 9.124807796290366e-06,
|
939 |
+
"loss": 0.6284,
|
940 |
+
"step": 660
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 0.27433993399339934,
|
944 |
+
"grad_norm": 2.7731836971623953,
|
945 |
+
"learning_rate": 9.104348022735853e-06,
|
946 |
+
"loss": 0.633,
|
947 |
+
"step": 665
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 0.2764026402640264,
|
951 |
+
"grad_norm": 2.7535471272900196,
|
952 |
+
"learning_rate": 9.083675351601417e-06,
|
953 |
+
"loss": 0.6194,
|
954 |
+
"step": 670
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 0.2784653465346535,
|
958 |
+
"grad_norm": 2.856389390689046,
|
959 |
+
"learning_rate": 9.062790855203932e-06,
|
960 |
+
"loss": 0.6376,
|
961 |
+
"step": 675
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 0.28052805280528054,
|
965 |
+
"grad_norm": 2.9038867407258224,
|
966 |
+
"learning_rate": 9.041695616847915e-06,
|
967 |
+
"loss": 0.6263,
|
968 |
+
"step": 680
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 0.2825907590759076,
|
972 |
+
"grad_norm": 2.5804402002125415,
|
973 |
+
"learning_rate": 9.020390730769324e-06,
|
974 |
+
"loss": 0.6393,
|
975 |
+
"step": 685
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 0.28465346534653463,
|
979 |
+
"grad_norm": 2.632496010078948,
|
980 |
+
"learning_rate": 8.998877302078803e-06,
|
981 |
+
"loss": 0.6076,
|
982 |
+
"step": 690
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 0.28671617161716173,
|
986 |
+
"grad_norm": 4.356850125546882,
|
987 |
+
"learning_rate": 8.97715644670436e-06,
|
988 |
+
"loss": 0.6133,
|
989 |
+
"step": 695
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 0.2887788778877888,
|
993 |
+
"grad_norm": 2.559920441209793,
|
994 |
+
"learning_rate": 8.955229291333473e-06,
|
995 |
+
"loss": 0.6167,
|
996 |
+
"step": 700
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 0.2908415841584158,
|
1000 |
+
"grad_norm": 2.7757209022740823,
|
1001 |
+
"learning_rate": 8.933096973354665e-06,
|
1002 |
+
"loss": 0.6171,
|
1003 |
+
"step": 705
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 0.2929042904290429,
|
1007 |
+
"grad_norm": 2.673673023680112,
|
1008 |
+
"learning_rate": 8.910760640798487e-06,
|
1009 |
+
"loss": 0.6142,
|
1010 |
+
"step": 710
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 0.29496699669966997,
|
1014 |
+
"grad_norm": 2.8111377522291856,
|
1015 |
+
"learning_rate": 8.88822145227798e-06,
|
1016 |
+
"loss": 0.609,
|
1017 |
+
"step": 715
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 0.297029702970297,
|
1021 |
+
"grad_norm": 2.6340814150599017,
|
1022 |
+
"learning_rate": 8.865480576928578e-06,
|
1023 |
+
"loss": 0.6205,
|
1024 |
+
"step": 720
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 0.2990924092409241,
|
1028 |
+
"grad_norm": 2.6824532363700566,
|
1029 |
+
"learning_rate": 8.842539194347448e-06,
|
1030 |
+
"loss": 0.5866,
|
1031 |
+
"step": 725
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 0.30115511551155116,
|
1035 |
+
"grad_norm": 2.619983825517638,
|
1036 |
+
"learning_rate": 8.819398494532328e-06,
|
1037 |
+
"loss": 0.6133,
|
1038 |
+
"step": 730
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 0.3032178217821782,
|
1042 |
+
"grad_norm": 2.686298593112456,
|
1043 |
+
"learning_rate": 8.796059677819773e-06,
|
1044 |
+
"loss": 0.6011,
|
1045 |
+
"step": 735
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 0.30528052805280526,
|
1049 |
+
"grad_norm": 2.9363501744882132,
|
1050 |
+
"learning_rate": 8.77252395482291e-06,
|
1051 |
+
"loss": 0.6084,
|
1052 |
+
"step": 740
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 0.30734323432343236,
|
1056 |
+
"grad_norm": 2.752287473506256,
|
1057 |
+
"learning_rate": 8.748792546368641e-06,
|
1058 |
+
"loss": 0.5966,
|
1059 |
+
"step": 745
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 0.3094059405940594,
|
1063 |
+
"grad_norm": 2.71844410518512,
|
1064 |
+
"learning_rate": 8.72486668343431e-06,
|
1065 |
+
"loss": 0.614,
|
1066 |
+
"step": 750
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 0.31146864686468645,
|
1070 |
+
"grad_norm": 3.145817966285669,
|
1071 |
+
"learning_rate": 8.700747607083851e-06,
|
1072 |
+
"loss": 0.5957,
|
1073 |
+
"step": 755
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 0.31353135313531355,
|
1077 |
+
"grad_norm": 3.0310831048054303,
|
1078 |
+
"learning_rate": 8.676436568403422e-06,
|
1079 |
+
"loss": 0.5978,
|
1080 |
+
"step": 760
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 0.3155940594059406,
|
1084 |
+
"grad_norm": 2.701739599075158,
|
1085 |
+
"learning_rate": 8.651934828436497e-06,
|
1086 |
+
"loss": 0.5923,
|
1087 |
+
"step": 765
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 0.31765676567656764,
|
1091 |
+
"grad_norm": 2.8448862694962944,
|
1092 |
+
"learning_rate": 8.627243658118466e-06,
|
1093 |
+
"loss": 0.5981,
|
1094 |
+
"step": 770
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 0.31971947194719474,
|
1098 |
+
"grad_norm": 2.719726931310169,
|
1099 |
+
"learning_rate": 8.602364338210699e-06,
|
1100 |
+
"loss": 0.5937,
|
1101 |
+
"step": 775
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 0.3217821782178218,
|
1105 |
+
"grad_norm": 2.8019550606602652,
|
1106 |
+
"learning_rate": 8.57729815923412e-06,
|
1107 |
+
"loss": 0.6029,
|
1108 |
+
"step": 780
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 0.32384488448844884,
|
1112 |
+
"grad_norm": 2.503107027258152,
|
1113 |
+
"learning_rate": 8.55204642140226e-06,
|
1114 |
+
"loss": 0.5952,
|
1115 |
+
"step": 785
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 0.3259075907590759,
|
1119 |
+
"grad_norm": 2.6677149194472762,
|
1120 |
+
"learning_rate": 8.52661043455382e-06,
|
1121 |
+
"loss": 0.5917,
|
1122 |
+
"step": 790
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 0.327970297029703,
|
1126 |
+
"grad_norm": 2.746672046408305,
|
1127 |
+
"learning_rate": 8.50099151808472e-06,
|
1128 |
+
"loss": 0.5973,
|
1129 |
+
"step": 795
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 0.33003300330033003,
|
1133 |
+
"grad_norm": 2.6263940638611554,
|
1134 |
+
"learning_rate": 8.47519100087967e-06,
|
1135 |
+
"loss": 0.5898,
|
1136 |
+
"step": 800
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 0.3320957095709571,
|
1140 |
+
"grad_norm": 2.503813501518845,
|
1141 |
+
"learning_rate": 8.449210221243225e-06,
|
1142 |
+
"loss": 0.5991,
|
1143 |
+
"step": 805
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 0.3341584158415842,
|
1147 |
+
"grad_norm": 2.4979356769923573,
|
1148 |
+
"learning_rate": 8.42305052683038e-06,
|
1149 |
+
"loss": 0.5912,
|
1150 |
+
"step": 810
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 0.3362211221122112,
|
1154 |
+
"grad_norm": 2.847043832115963,
|
1155 |
+
"learning_rate": 8.39671327457666e-06,
|
1156 |
+
"loss": 0.5799,
|
1157 |
+
"step": 815
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 0.33828382838283827,
|
1161 |
+
"grad_norm": 2.7801453739069633,
|
1162 |
+
"learning_rate": 8.370199830627732e-06,
|
1163 |
+
"loss": 0.5823,
|
1164 |
+
"step": 820
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 0.34034653465346537,
|
1168 |
+
"grad_norm": 2.627346750326928,
|
1169 |
+
"learning_rate": 8.343511570268541e-06,
|
1170 |
+
"loss": 0.5865,
|
1171 |
+
"step": 825
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 0.3424092409240924,
|
1175 |
+
"grad_norm": 2.8090361015173246,
|
1176 |
+
"learning_rate": 8.316649877851977e-06,
|
1177 |
+
"loss": 0.568,
|
1178 |
+
"step": 830
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 0.34447194719471946,
|
1182 |
+
"grad_norm": 2.6857687788227107,
|
1183 |
+
"learning_rate": 8.289616146727062e-06,
|
1184 |
+
"loss": 0.5811,
|
1185 |
+
"step": 835
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 0.3465346534653465,
|
1189 |
+
"grad_norm": 2.7827938400617964,
|
1190 |
+
"learning_rate": 8.262411779166681e-06,
|
1191 |
+
"loss": 0.5608,
|
1192 |
+
"step": 840
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 0.3485973597359736,
|
1196 |
+
"grad_norm": 2.7509505960907084,
|
1197 |
+
"learning_rate": 8.235038186294836e-06,
|
1198 |
+
"loss": 0.5888,
|
1199 |
+
"step": 845
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 0.35066006600660066,
|
1203 |
+
"grad_norm": 2.6638572156656584,
|
1204 |
+
"learning_rate": 8.207496788013456e-06,
|
1205 |
+
"loss": 0.562,
|
1206 |
+
"step": 850
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 0.3527227722772277,
|
1210 |
+
"grad_norm": 2.591282044159344,
|
1211 |
+
"learning_rate": 8.179789012928747e-06,
|
1212 |
+
"loss": 0.5712,
|
1213 |
+
"step": 855
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 0.3547854785478548,
|
1217 |
+
"grad_norm": 2.5109540424205576,
|
1218 |
+
"learning_rate": 8.151916298277078e-06,
|
1219 |
+
"loss": 0.553,
|
1220 |
+
"step": 860
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 0.35684818481848185,
|
1224 |
+
"grad_norm": 2.892814594553888,
|
1225 |
+
"learning_rate": 8.123880089850438e-06,
|
1226 |
+
"loss": 0.5686,
|
1227 |
+
"step": 865
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 0.3589108910891089,
|
1231 |
+
"grad_norm": 2.5359593127713898,
|
1232 |
+
"learning_rate": 8.095681841921441e-06,
|
1233 |
+
"loss": 0.5454,
|
1234 |
+
"step": 870
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 0.360973597359736,
|
1238 |
+
"grad_norm": 2.6515995361569202,
|
1239 |
+
"learning_rate": 8.06732301716789e-06,
|
1240 |
+
"loss": 0.562,
|
1241 |
+
"step": 875
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 0.36303630363036304,
|
1245 |
+
"grad_norm": 2.485706938051772,
|
1246 |
+
"learning_rate": 8.038805086596903e-06,
|
1247 |
+
"loss": 0.5532,
|
1248 |
+
"step": 880
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 0.3650990099009901,
|
1252 |
+
"grad_norm": 2.5183163477684576,
|
1253 |
+
"learning_rate": 8.010129529468614e-06,
|
1254 |
+
"loss": 0.5666,
|
1255 |
+
"step": 885
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 0.36716171617161714,
|
1259 |
+
"grad_norm": 2.675472695073968,
|
1260 |
+
"learning_rate": 7.981297833219435e-06,
|
1261 |
+
"loss": 0.5631,
|
1262 |
+
"step": 890
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 0.36922442244224424,
|
1266 |
+
"grad_norm": 2.6653612064334045,
|
1267 |
+
"learning_rate": 7.952311493384916e-06,
|
1268 |
+
"loss": 0.5681,
|
1269 |
+
"step": 895
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 0.3712871287128713,
|
1273 |
+
"grad_norm": 2.57514354803963,
|
1274 |
+
"learning_rate": 7.923172013522153e-06,
|
1275 |
+
"loss": 0.5668,
|
1276 |
+
"step": 900
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 0.37334983498349833,
|
1280 |
+
"grad_norm": 2.707762540885758,
|
1281 |
+
"learning_rate": 7.893880905131807e-06,
|
1282 |
+
"loss": 0.558,
|
1283 |
+
"step": 905
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 0.37541254125412543,
|
1287 |
+
"grad_norm": 2.6980529523902472,
|
1288 |
+
"learning_rate": 7.864439687579695e-06,
|
1289 |
+
"loss": 0.5447,
|
1290 |
+
"step": 910
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 0.3774752475247525,
|
1294 |
+
"grad_norm": 2.7666172686862267,
|
1295 |
+
"learning_rate": 7.834849888017979e-06,
|
1296 |
+
"loss": 0.555,
|
1297 |
+
"step": 915
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 0.3795379537953795,
|
1301 |
+
"grad_norm": 2.8852303996201876,
|
1302 |
+
"learning_rate": 7.805113041305958e-06,
|
1303 |
+
"loss": 0.5381,
|
1304 |
+
"step": 920
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 0.3816006600660066,
|
1308 |
+
"grad_norm": 2.583967025080346,
|
1309 |
+
"learning_rate": 7.775230689930445e-06,
|
1310 |
+
"loss": 0.5465,
|
1311 |
+
"step": 925
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 0.38366336633663367,
|
1315 |
+
"grad_norm": 2.612939793985183,
|
1316 |
+
"learning_rate": 7.745204383925753e-06,
|
1317 |
+
"loss": 0.5542,
|
1318 |
+
"step": 930
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 0.3857260726072607,
|
1322 |
+
"grad_norm": 2.779912992420415,
|
1323 |
+
"learning_rate": 7.715035680793311e-06,
|
1324 |
+
"loss": 0.5527,
|
1325 |
+
"step": 935
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 0.38778877887788776,
|
1329 |
+
"grad_norm": 2.644547318860223,
|
1330 |
+
"learning_rate": 7.684726145420853e-06,
|
1331 |
+
"loss": 0.5405,
|
1332 |
+
"step": 940
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 0.38985148514851486,
|
1336 |
+
"grad_norm": 2.5007394214042433,
|
1337 |
+
"learning_rate": 7.654277350001255e-06,
|
1338 |
+
"loss": 0.5635,
|
1339 |
+
"step": 945
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 0.3919141914191419,
|
1343 |
+
"grad_norm": 2.806021602965884,
|
1344 |
+
"learning_rate": 7.623690873950988e-06,
|
1345 |
+
"loss": 0.5273,
|
1346 |
+
"step": 950
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 0.39397689768976896,
|
1350 |
+
"grad_norm": 2.5596611924464754,
|
1351 |
+
"learning_rate": 7.592968303828181e-06,
|
1352 |
+
"loss": 0.5275,
|
1353 |
+
"step": 955
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 0.39603960396039606,
|
1357 |
+
"grad_norm": 2.6100310716117994,
|
1358 |
+
"learning_rate": 7.5621112332503325e-06,
|
1359 |
+
"loss": 0.5296,
|
1360 |
+
"step": 960
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 0.3981023102310231,
|
1364 |
+
"grad_norm": 2.4652574522767017,
|
1365 |
+
"learning_rate": 7.531121262811645e-06,
|
1366 |
+
"loss": 0.5327,
|
1367 |
+
"step": 965
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 0.40016501650165015,
|
1371 |
+
"grad_norm": 2.5835355525531054,
|
1372 |
+
"learning_rate": 7.500000000000001e-06,
|
1373 |
+
"loss": 0.5313,
|
1374 |
+
"step": 970
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 0.40222772277227725,
|
1378 |
+
"grad_norm": 2.556870458540024,
|
1379 |
+
"learning_rate": 7.468749059113578e-06,
|
1380 |
+
"loss": 0.5219,
|
1381 |
+
"step": 975
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 0.4042904290429043,
|
1385 |
+
"grad_norm": 2.6560183770094166,
|
1386 |
+
"learning_rate": 7.437370061177116e-06,
|
1387 |
+
"loss": 0.529,
|
1388 |
+
"step": 980
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 0.40635313531353134,
|
1392 |
+
"grad_norm": 2.6201898522482683,
|
1393 |
+
"learning_rate": 7.40586463385783e-06,
|
1394 |
+
"loss": 0.534,
|
1395 |
+
"step": 985
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 0.4084158415841584,
|
1399 |
+
"grad_norm": 2.6803677560952712,
|
1400 |
+
"learning_rate": 7.374234411380987e-06,
|
1401 |
+
"loss": 0.5249,
|
1402 |
+
"step": 990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 0.4104785478547855,
|
1406 |
+
"grad_norm": 2.5303749301843586,
|
1407 |
+
"learning_rate": 7.342481034445127e-06,
|
1408 |
+
"loss": 0.5318,
|
1409 |
+
"step": 995
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 0.41254125412541254,
|
1413 |
+
"grad_norm": 2.5847848779090694,
|
1414 |
+
"learning_rate": 7.310606150136965e-06,
|
1415 |
+
"loss": 0.5137,
|
1416 |
+
"step": 1000
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 0.4146039603960396,
|
1420 |
+
"grad_norm": 2.624837840218052,
|
1421 |
+
"learning_rate": 7.2786114118459564e-06,
|
1422 |
+
"loss": 0.5275,
|
1423 |
+
"step": 1005
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 0.4166666666666667,
|
1427 |
+
"grad_norm": 2.7154629758532574,
|
1428 |
+
"learning_rate": 7.246498479178523e-06,
|
1429 |
+
"loss": 0.5311,
|
1430 |
+
"step": 1010
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 0.41872937293729373,
|
1434 |
+
"grad_norm": 2.7217754848899234,
|
1435 |
+
"learning_rate": 7.214269017871981e-06,
|
1436 |
+
"loss": 0.5388,
|
1437 |
+
"step": 1015
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 0.4207920792079208,
|
1441 |
+
"grad_norm": 2.8374602937448956,
|
1442 |
+
"learning_rate": 7.181924699708127e-06,
|
1443 |
+
"loss": 0.5114,
|
1444 |
+
"step": 1020
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 0.4228547854785479,
|
1448 |
+
"grad_norm": 2.669653807310483,
|
1449 |
+
"learning_rate": 7.149467202426525e-06,
|
1450 |
+
"loss": 0.5162,
|
1451 |
+
"step": 1025
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 0.4249174917491749,
|
1455 |
+
"grad_norm": 2.8259143043104853,
|
1456 |
+
"learning_rate": 7.116898209637478e-06,
|
1457 |
+
"loss": 0.5207,
|
1458 |
+
"step": 1030
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 0.42698019801980197,
|
1462 |
+
"grad_norm": 2.4487631528416034,
|
1463 |
+
"learning_rate": 7.084219410734701e-06,
|
1464 |
+
"loss": 0.5239,
|
1465 |
+
"step": 1035
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 0.429042904290429,
|
1469 |
+
"grad_norm": 2.5467769274691157,
|
1470 |
+
"learning_rate": 7.051432500807682e-06,
|
1471 |
+
"loss": 0.5143,
|
1472 |
+
"step": 1040
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 0.4311056105610561,
|
1476 |
+
"grad_norm": 2.674655727993759,
|
1477 |
+
"learning_rate": 7.018539180553768e-06,
|
1478 |
+
"loss": 0.5179,
|
1479 |
+
"step": 1045
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 0.43316831683168316,
|
1483 |
+
"grad_norm": 2.5887768807891107,
|
1484 |
+
"learning_rate": 6.985541156189932e-06,
|
1485 |
+
"loss": 0.5084,
|
1486 |
+
"step": 1050
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 0.4352310231023102,
|
1490 |
+
"grad_norm": 2.7440060414213394,
|
1491 |
+
"learning_rate": 6.952440139364286e-06,
|
1492 |
+
"loss": 0.512,
|
1493 |
+
"step": 1055
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 0.4372937293729373,
|
1497 |
+
"grad_norm": 2.645722871973474,
|
1498 |
+
"learning_rate": 6.919237847067282e-06,
|
1499 |
+
"loss": 0.5199,
|
1500 |
+
"step": 1060
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 0.43935643564356436,
|
1504 |
+
"grad_norm": 2.611780386705664,
|
1505 |
+
"learning_rate": 6.885936001542658e-06,
|
1506 |
+
"loss": 0.5078,
|
1507 |
+
"step": 1065
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 0.4414191419141914,
|
1511 |
+
"grad_norm": 2.5086528986935703,
|
1512 |
+
"learning_rate": 6.852536330198099e-06,
|
1513 |
+
"loss": 0.5201,
|
1514 |
+
"step": 1070
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 0.4434818481848185,
|
1518 |
+
"grad_norm": 2.4576048338159486,
|
1519 |
+
"learning_rate": 6.819040565515636e-06,
|
1520 |
+
"loss": 0.5126,
|
1521 |
+
"step": 1075
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 0.44554455445544555,
|
1525 |
+
"grad_norm": 2.7304284435440542,
|
1526 |
+
"learning_rate": 6.785450444961783e-06,
|
1527 |
+
"loss": 0.5145,
|
1528 |
+
"step": 1080
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 0.4476072607260726,
|
1532 |
+
"grad_norm": 2.6166324259831435,
|
1533 |
+
"learning_rate": 6.751767710897404e-06,
|
1534 |
+
"loss": 0.5141,
|
1535 |
+
"step": 1085
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 0.44966996699669964,
|
1539 |
+
"grad_norm": 2.528382351784196,
|
1540 |
+
"learning_rate": 6.71799411048734e-06,
|
1541 |
+
"loss": 0.4858,
|
1542 |
+
"step": 1090
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 0.45173267326732675,
|
1546 |
+
"grad_norm": 2.633514406168111,
|
1547 |
+
"learning_rate": 6.684131395609784e-06,
|
1548 |
+
"loss": 0.5084,
|
1549 |
+
"step": 1095
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 0.4537953795379538,
|
1553 |
+
"grad_norm": 2.534482871115365,
|
1554 |
+
"learning_rate": 6.650181322765407e-06,
|
1555 |
+
"loss": 0.4981,
|
1556 |
+
"step": 1100
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 0.45585808580858084,
|
1560 |
+
"grad_norm": 2.4508414675719505,
|
1561 |
+
"learning_rate": 6.61614565298624e-06,
|
1562 |
+
"loss": 0.5008,
|
1563 |
+
"step": 1105
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 0.45792079207920794,
|
1567 |
+
"grad_norm": 2.4274494034766203,
|
1568 |
+
"learning_rate": 6.5820261517443365e-06,
|
1569 |
+
"loss": 0.4981,
|
1570 |
+
"step": 1110
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 0.459983498349835,
|
1574 |
+
"grad_norm": 2.538858548171987,
|
1575 |
+
"learning_rate": 6.54782458886019e-06,
|
1576 |
+
"loss": 0.4891,
|
1577 |
+
"step": 1115
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 0.46204620462046203,
|
1581 |
+
"grad_norm": 2.5483259090729575,
|
1582 |
+
"learning_rate": 6.5135427384109315e-06,
|
1583 |
+
"loss": 0.4999,
|
1584 |
+
"step": 1120
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 0.46410891089108913,
|
1588 |
+
"grad_norm": 2.5232721697679166,
|
1589 |
+
"learning_rate": 6.479182378638308e-06,
|
1590 |
+
"loss": 0.5024,
|
1591 |
+
"step": 1125
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 0.4661716171617162,
|
1595 |
+
"grad_norm": 2.591445497366531,
|
1596 |
+
"learning_rate": 6.444745291856442e-06,
|
1597 |
+
"loss": 0.4964,
|
1598 |
+
"step": 1130
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 0.4682343234323432,
|
1602 |
+
"grad_norm": 2.568897490439017,
|
1603 |
+
"learning_rate": 6.410233264359379e-06,
|
1604 |
+
"loss": 0.4978,
|
1605 |
+
"step": 1135
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 0.47029702970297027,
|
1609 |
+
"grad_norm": 2.494168999146032,
|
1610 |
+
"learning_rate": 6.375648086328431e-06,
|
1611 |
+
"loss": 0.5058,
|
1612 |
+
"step": 1140
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 0.47235973597359737,
|
1616 |
+
"grad_norm": 2.4409245464037967,
|
1617 |
+
"learning_rate": 6.340991551739319e-06,
|
1618 |
+
"loss": 0.4885,
|
1619 |
+
"step": 1145
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 0.4744224422442244,
|
1623 |
+
"grad_norm": 2.603683438326256,
|
1624 |
+
"learning_rate": 6.3062654582691175e-06,
|
1625 |
+
"loss": 0.4983,
|
1626 |
+
"step": 1150
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 0.47648514851485146,
|
1630 |
+
"grad_norm": 2.5017059333187404,
|
1631 |
+
"learning_rate": 6.271471607203006e-06,
|
1632 |
+
"loss": 0.4868,
|
1633 |
+
"step": 1155
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 0.47854785478547857,
|
1637 |
+
"grad_norm": 2.664734878049621,
|
1638 |
+
"learning_rate": 6.236611803340829e-06,
|
1639 |
+
"loss": 0.4946,
|
1640 |
+
"step": 1160
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 0.4806105610561056,
|
1644 |
+
"grad_norm": 2.6730486310091037,
|
1645 |
+
"learning_rate": 6.201687854903492e-06,
|
1646 |
+
"loss": 0.4811,
|
1647 |
+
"step": 1165
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 0.48267326732673266,
|
1651 |
+
"grad_norm": 2.8065530727044927,
|
1652 |
+
"learning_rate": 6.16670157343915e-06,
|
1653 |
+
"loss": 0.4974,
|
1654 |
+
"step": 1170
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 0.48473597359735976,
|
1658 |
+
"grad_norm": 2.660140416974183,
|
1659 |
+
"learning_rate": 6.131654773729255e-06,
|
1660 |
+
"loss": 0.485,
|
1661 |
+
"step": 1175
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 0.4867986798679868,
|
1665 |
+
"grad_norm": 2.494718317338255,
|
1666 |
+
"learning_rate": 6.096549273694411e-06,
|
1667 |
+
"loss": 0.4832,
|
1668 |
+
"step": 1180
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 0.48886138613861385,
|
1672 |
+
"grad_norm": 2.6067794556409427,
|
1673 |
+
"learning_rate": 6.061386894300082e-06,
|
1674 |
+
"loss": 0.4822,
|
1675 |
+
"step": 1185
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 0.4909240924092409,
|
1679 |
+
"grad_norm": 2.411365711902098,
|
1680 |
+
"learning_rate": 6.026169459462132e-06,
|
1681 |
+
"loss": 0.488,
|
1682 |
+
"step": 1190
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 0.492986798679868,
|
1686 |
+
"grad_norm": 2.5304361708398253,
|
1687 |
+
"learning_rate": 5.990898795952225e-06,
|
1688 |
+
"loss": 0.4808,
|
1689 |
+
"step": 1195
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 0.49504950495049505,
|
1693 |
+
"grad_norm": 2.542256557791119,
|
1694 |
+
"learning_rate": 5.955576733303053e-06,
|
1695 |
+
"loss": 0.4922,
|
1696 |
+
"step": 1200
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 0.4971122112211221,
|
1700 |
+
"grad_norm": 2.416811649448376,
|
1701 |
+
"learning_rate": 5.920205103713449e-06,
|
1702 |
+
"loss": 0.4857,
|
1703 |
+
"step": 1205
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 0.4991749174917492,
|
1707 |
+
"grad_norm": 2.4894521032789023,
|
1708 |
+
"learning_rate": 5.884785741953345e-06,
|
1709 |
+
"loss": 0.4742,
|
1710 |
+
"step": 1210
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 0.5012376237623762,
|
1714 |
+
"grad_norm": 2.4053368303294995,
|
1715 |
+
"learning_rate": 5.849320485268597e-06,
|
1716 |
+
"loss": 0.4691,
|
1717 |
+
"step": 1215
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 0.5033003300330033,
|
1721 |
+
"grad_norm": 2.440300312110612,
|
1722 |
+
"learning_rate": 5.8138111732856906e-06,
|
1723 |
+
"loss": 0.4794,
|
1724 |
+
"step": 1220
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 0.5053630363036303,
|
1728 |
+
"grad_norm": 2.507917784275635,
|
1729 |
+
"learning_rate": 5.778259647916309e-06,
|
1730 |
+
"loss": 0.4807,
|
1731 |
+
"step": 1225
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 0.5074257425742574,
|
1735 |
+
"grad_norm": 2.564227750985733,
|
1736 |
+
"learning_rate": 5.7426677532618e-06,
|
1737 |
+
"loss": 0.4623,
|
1738 |
+
"step": 1230
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 0.5094884488448845,
|
1742 |
+
"grad_norm": 2.5559157486419037,
|
1743 |
+
"learning_rate": 5.707037335517514e-06,
|
1744 |
+
"loss": 0.4599,
|
1745 |
+
"step": 1235
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 0.5115511551155115,
|
1749 |
+
"grad_norm": 2.4886710886810666,
|
1750 |
+
"learning_rate": 5.67137024287704e-06,
|
1751 |
+
"loss": 0.4539,
|
1752 |
+
"step": 1240
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 0.5136138613861386,
|
1756 |
+
"grad_norm": 2.533851130168081,
|
1757 |
+
"learning_rate": 5.635668325436343e-06,
|
1758 |
+
"loss": 0.4762,
|
1759 |
+
"step": 1245
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 0.5156765676567657,
|
1763 |
+
"grad_norm": 2.5682484148678593,
|
1764 |
+
"learning_rate": 5.599933435097791e-06,
|
1765 |
+
"loss": 0.461,
|
1766 |
+
"step": 1250
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 0.5177392739273927,
|
1770 |
+
"grad_norm": 2.7619029888240334,
|
1771 |
+
"learning_rate": 5.564167425474093e-06,
|
1772 |
+
"loss": 0.4708,
|
1773 |
+
"step": 1255
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 0.5198019801980198,
|
1777 |
+
"grad_norm": 2.65748720318593,
|
1778 |
+
"learning_rate": 5.528372151792161e-06,
|
1779 |
+
"loss": 0.4835,
|
1780 |
+
"step": 1260
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 0.5218646864686468,
|
1784 |
+
"grad_norm": 2.343200657687914,
|
1785 |
+
"learning_rate": 5.492549470796865e-06,
|
1786 |
+
"loss": 0.4562,
|
1787 |
+
"step": 1265
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 0.5239273927392739,
|
1791 |
+
"grad_norm": 2.398515835994456,
|
1792 |
+
"learning_rate": 5.456701240654726e-06,
|
1793 |
+
"loss": 0.4448,
|
1794 |
+
"step": 1270
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 0.525990099009901,
|
1798 |
+
"grad_norm": 2.480816874843841,
|
1799 |
+
"learning_rate": 5.420829320857532e-06,
|
1800 |
+
"loss": 0.4599,
|
1801 |
+
"step": 1275
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 0.528052805280528,
|
1805 |
+
"grad_norm": 2.5823441471809643,
|
1806 |
+
"learning_rate": 5.384935572125882e-06,
|
1807 |
+
"loss": 0.4352,
|
1808 |
+
"step": 1280
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 0.5301155115511551,
|
1812 |
+
"grad_norm": 2.4908465249480853,
|
1813 |
+
"learning_rate": 5.349021856312669e-06,
|
1814 |
+
"loss": 0.4594,
|
1815 |
+
"step": 1285
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 0.5321782178217822,
|
1819 |
+
"grad_norm": 2.399600740783908,
|
1820 |
+
"learning_rate": 5.3130900363065055e-06,
|
1821 |
+
"loss": 0.4466,
|
1822 |
+
"step": 1290
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 0.5342409240924092,
|
1826 |
+
"grad_norm": 2.454620597807439,
|
1827 |
+
"learning_rate": 5.277141975935083e-06,
|
1828 |
+
"loss": 0.4585,
|
1829 |
+
"step": 1295
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 0.5363036303630363,
|
1833 |
+
"grad_norm": 2.372249644843968,
|
1834 |
+
"learning_rate": 5.24117953986851e-06,
|
1835 |
+
"loss": 0.4512,
|
1836 |
+
"step": 1300
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 0.5383663366336634,
|
1840 |
+
"grad_norm": 2.3367801961389407,
|
1841 |
+
"learning_rate": 5.2052045935225725e-06,
|
1842 |
+
"loss": 0.4568,
|
1843 |
+
"step": 1305
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 0.5404290429042904,
|
1847 |
+
"grad_norm": 2.4464214105982856,
|
1848 |
+
"learning_rate": 5.169219002961987e-06,
|
1849 |
+
"loss": 0.4561,
|
1850 |
+
"step": 1310
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 0.5424917491749175,
|
1854 |
+
"grad_norm": 10.09033249652104,
|
1855 |
+
"learning_rate": 5.133224634803594e-06,
|
1856 |
+
"loss": 0.4506,
|
1857 |
+
"step": 1315
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 0.5445544554455446,
|
1861 |
+
"grad_norm": 2.525626316753546,
|
1862 |
+
"learning_rate": 5.097223356119538e-06,
|
1863 |
+
"loss": 0.4651,
|
1864 |
+
"step": 1320
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 0.5466171617161716,
|
1868 |
+
"grad_norm": 2.521794410914058,
|
1869 |
+
"learning_rate": 5.061217034340426e-06,
|
1870 |
+
"loss": 0.4729,
|
1871 |
+
"step": 1325
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 0.5486798679867987,
|
1875 |
+
"grad_norm": 2.5260338421626445,
|
1876 |
+
"learning_rate": 5.02520753715845e-06,
|
1877 |
+
"loss": 0.4525,
|
1878 |
+
"step": 1330
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 0.5507425742574258,
|
1882 |
+
"grad_norm": 2.446840796366364,
|
1883 |
+
"learning_rate": 4.989196732430518e-06,
|
1884 |
+
"loss": 0.4567,
|
1885 |
+
"step": 1335
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 0.5528052805280528,
|
1889 |
+
"grad_norm": 2.3859413099257893,
|
1890 |
+
"learning_rate": 4.953186488081362e-06,
|
1891 |
+
"loss": 0.4471,
|
1892 |
+
"step": 1340
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 0.5548679867986799,
|
1896 |
+
"grad_norm": 2.5520983341553105,
|
1897 |
+
"learning_rate": 4.9171786720066465e-06,
|
1898 |
+
"loss": 0.4523,
|
1899 |
+
"step": 1345
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 0.556930693069307,
|
1903 |
+
"grad_norm": 2.470761509016717,
|
1904 |
+
"learning_rate": 4.881175151976075e-06,
|
1905 |
+
"loss": 0.4589,
|
1906 |
+
"step": 1350
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 0.558993399339934,
|
1910 |
+
"grad_norm": 2.6345528475282847,
|
1911 |
+
"learning_rate": 4.845177795536516e-06,
|
1912 |
+
"loss": 0.4429,
|
1913 |
+
"step": 1355
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 0.5610561056105611,
|
1917 |
+
"grad_norm": 2.4651932232525366,
|
1918 |
+
"learning_rate": 4.809188469915121e-06,
|
1919 |
+
"loss": 0.4341,
|
1920 |
+
"step": 1360
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 0.5631188118811881,
|
1924 |
+
"grad_norm": 2.423072698536209,
|
1925 |
+
"learning_rate": 4.773209041922472e-06,
|
1926 |
+
"loss": 0.4416,
|
1927 |
+
"step": 1365
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 0.5651815181518152,
|
1931 |
+
"grad_norm": 2.453471069490191,
|
1932 |
+
"learning_rate": 4.737241377855751e-06,
|
1933 |
+
"loss": 0.4307,
|
1934 |
+
"step": 1370
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 0.5672442244224423,
|
1938 |
+
"grad_norm": 2.373055640370215,
|
1939 |
+
"learning_rate": 4.7012873434019296e-06,
|
1940 |
+
"loss": 0.4355,
|
1941 |
+
"step": 1375
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 0.5693069306930693,
|
1945 |
+
"grad_norm": 2.5437533790570646,
|
1946 |
+
"learning_rate": 4.6653488035409975e-06,
|
1947 |
+
"loss": 0.4415,
|
1948 |
+
"step": 1380
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 0.5713696369636964,
|
1952 |
+
"grad_norm": 2.391498166308705,
|
1953 |
+
"learning_rate": 4.629427622449217e-06,
|
1954 |
+
"loss": 0.4313,
|
1955 |
+
"step": 1385
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 0.5734323432343235,
|
1959 |
+
"grad_norm": 2.643038273859464,
|
1960 |
+
"learning_rate": 4.59352566340243e-06,
|
1961 |
+
"loss": 0.4531,
|
1962 |
+
"step": 1390
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 0.5754950495049505,
|
1966 |
+
"grad_norm": 2.515338122652954,
|
1967 |
+
"learning_rate": 4.557644788679413e-06,
|
1968 |
+
"loss": 0.4372,
|
1969 |
+
"step": 1395
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 0.5775577557755776,
|
1973 |
+
"grad_norm": 2.5228479375899355,
|
1974 |
+
"learning_rate": 4.521786859465263e-06,
|
1975 |
+
"loss": 0.4343,
|
1976 |
+
"step": 1400
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 0.5796204620462047,
|
1980 |
+
"grad_norm": 2.526725915460239,
|
1981 |
+
"learning_rate": 4.485953735754872e-06,
|
1982 |
+
"loss": 0.4236,
|
1983 |
+
"step": 1405
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 0.5816831683168316,
|
1987 |
+
"grad_norm": 2.3164427894674224,
|
1988 |
+
"learning_rate": 4.450147276256439e-06,
|
1989 |
+
"loss": 0.4164,
|
1990 |
+
"step": 1410
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 0.5837458745874587,
|
1994 |
+
"grad_norm": 2.3425181209777963,
|
1995 |
+
"learning_rate": 4.414369338295056e-06,
|
1996 |
+
"loss": 0.4427,
|
1997 |
+
"step": 1415
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 0.5858085808580858,
|
2001 |
+
"grad_norm": 2.450541853696667,
|
2002 |
+
"learning_rate": 4.37862177771637e-06,
|
2003 |
+
"loss": 0.4406,
|
2004 |
+
"step": 1420
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 0.5878712871287128,
|
2008 |
+
"grad_norm": 2.420195141307401,
|
2009 |
+
"learning_rate": 4.342906448790315e-06,
|
2010 |
+
"loss": 0.4475,
|
2011 |
+
"step": 1425
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 0.5899339933993399,
|
2015 |
+
"grad_norm": 2.4136105308611624,
|
2016 |
+
"learning_rate": 4.307225204114927e-06,
|
2017 |
+
"loss": 0.435,
|
2018 |
+
"step": 1430
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 0.591996699669967,
|
2022 |
+
"grad_norm": 2.4087084238655962,
|
2023 |
+
"learning_rate": 4.271579894520254e-06,
|
2024 |
+
"loss": 0.44,
|
2025 |
+
"step": 1435
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 0.594059405940594,
|
2029 |
+
"grad_norm": 2.457797382505865,
|
2030 |
+
"learning_rate": 4.235972368972343e-06,
|
2031 |
+
"loss": 0.4368,
|
2032 |
+
"step": 1440
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 0.5961221122112211,
|
2036 |
+
"grad_norm": 2.3957561493714827,
|
2037 |
+
"learning_rate": 4.200404474477341e-06,
|
2038 |
+
"loss": 0.4163,
|
2039 |
+
"step": 1445
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 0.5981848184818482,
|
2043 |
+
"grad_norm": 2.3457771252145587,
|
2044 |
+
"learning_rate": 4.16487805598568e-06,
|
2045 |
+
"loss": 0.4094,
|
2046 |
+
"step": 1450
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 0.6002475247524752,
|
2050 |
+
"grad_norm": 2.573777931737885,
|
2051 |
+
"learning_rate": 4.12939495629638e-06,
|
2052 |
+
"loss": 0.4162,
|
2053 |
+
"step": 1455
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 0.6023102310231023,
|
2057 |
+
"grad_norm": 2.348703417322529,
|
2058 |
+
"learning_rate": 4.093957015961465e-06,
|
2059 |
+
"loss": 0.4208,
|
2060 |
+
"step": 1460
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 0.6043729372937293,
|
2064 |
+
"grad_norm": 2.6136735137117997,
|
2065 |
+
"learning_rate": 4.0585660731904855e-06,
|
2066 |
+
"loss": 0.4357,
|
2067 |
+
"step": 1465
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 0.6064356435643564,
|
2071 |
+
"grad_norm": 2.3717596586411633,
|
2072 |
+
"learning_rate": 4.023223963755168e-06,
|
2073 |
+
"loss": 0.4111,
|
2074 |
+
"step": 1470
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 0.6084983498349835,
|
2078 |
+
"grad_norm": 2.3925003782413117,
|
2079 |
+
"learning_rate": 3.987932520894201e-06,
|
2080 |
+
"loss": 0.422,
|
2081 |
+
"step": 1475
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 0.6105610561056105,
|
2085 |
+
"grad_norm": 2.4066751622931086,
|
2086 |
+
"learning_rate": 3.9526935752181275e-06,
|
2087 |
+
"loss": 0.4262,
|
2088 |
+
"step": 1480
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 0.6126237623762376,
|
2092 |
+
"grad_norm": 2.2489948265452435,
|
2093 |
+
"learning_rate": 3.917508954614401e-06,
|
2094 |
+
"loss": 0.4137,
|
2095 |
+
"step": 1485
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 0.6146864686468647,
|
2099 |
+
"grad_norm": 2.3814132782379636,
|
2100 |
+
"learning_rate": 3.882380484152567e-06,
|
2101 |
+
"loss": 0.4114,
|
2102 |
+
"step": 1490
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 0.6167491749174917,
|
2106 |
+
"grad_norm": 2.503060690023786,
|
2107 |
+
"learning_rate": 3.847309985989593e-06,
|
2108 |
+
"loss": 0.4222,
|
2109 |
+
"step": 1495
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 0.6188118811881188,
|
2113 |
+
"grad_norm": 2.485844712788725,
|
2114 |
+
"learning_rate": 3.8122992792753534e-06,
|
2115 |
+
"loss": 0.4212,
|
2116 |
+
"step": 1500
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 0.6208745874587459,
|
2120 |
+
"grad_norm": 2.428549383590827,
|
2121 |
+
"learning_rate": 3.777350180058264e-06,
|
2122 |
+
"loss": 0.4187,
|
2123 |
+
"step": 1505
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 0.6229372937293729,
|
2127 |
+
"grad_norm": 2.6875741250581533,
|
2128 |
+
"learning_rate": 3.7424645011910847e-06,
|
2129 |
+
"loss": 0.4123,
|
2130 |
+
"step": 1510
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 0.625,
|
2134 |
+
"grad_norm": 2.425981196955932,
|
2135 |
+
"learning_rate": 3.707644052236887e-06,
|
2136 |
+
"loss": 0.4096,
|
2137 |
+
"step": 1515
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 0.6270627062706271,
|
2141 |
+
"grad_norm": 2.7328115668919155,
|
2142 |
+
"learning_rate": 3.672890639375184e-06,
|
2143 |
+
"loss": 0.4206,
|
2144 |
+
"step": 1520
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 0.6291254125412541,
|
2148 |
+
"grad_norm": 2.3245488388421593,
|
2149 |
+
"learning_rate": 3.6382060653082434e-06,
|
2150 |
+
"loss": 0.4052,
|
2151 |
+
"step": 1525
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 0.6311881188118812,
|
2155 |
+
"grad_norm": 2.384682866362113,
|
2156 |
+
"learning_rate": 3.6035921291675815e-06,
|
2157 |
+
"loss": 0.4156,
|
2158 |
+
"step": 1530
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 0.6332508250825083,
|
2162 |
+
"grad_norm": 2.4129965482745486,
|
2163 |
+
"learning_rate": 3.569050626420636e-06,
|
2164 |
+
"loss": 0.4093,
|
2165 |
+
"step": 1535
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 0.6353135313531353,
|
2169 |
+
"grad_norm": 2.4547665669069776,
|
2170 |
+
"learning_rate": 3.5345833487776404e-06,
|
2171 |
+
"loss": 0.425,
|
2172 |
+
"step": 1540
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 0.6373762376237624,
|
2176 |
+
"grad_norm": 2.489381449919041,
|
2177 |
+
"learning_rate": 3.500192084098677e-06,
|
2178 |
+
"loss": 0.4215,
|
2179 |
+
"step": 1545
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 0.6394389438943895,
|
2183 |
+
"grad_norm": 2.382148749916793,
|
2184 |
+
"learning_rate": 3.4658786163009416e-06,
|
2185 |
+
"loss": 0.4152,
|
2186 |
+
"step": 1550
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 0.6415016501650165,
|
2190 |
+
"grad_norm": 2.3828897057876746,
|
2191 |
+
"learning_rate": 3.4316447252662142e-06,
|
2192 |
+
"loss": 0.4084,
|
2193 |
+
"step": 1555
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 0.6435643564356436,
|
2197 |
+
"grad_norm": 2.446574644486715,
|
2198 |
+
"learning_rate": 3.3974921867485238e-06,
|
2199 |
+
"loss": 0.407,
|
2200 |
+
"step": 1560
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 0.6456270627062707,
|
2204 |
+
"grad_norm": 2.3536881611551035,
|
2205 |
+
"learning_rate": 3.3634227722820496e-06,
|
2206 |
+
"loss": 0.4007,
|
2207 |
+
"step": 1565
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 0.6476897689768977,
|
2211 |
+
"grad_norm": 2.48572252066864,
|
2212 |
+
"learning_rate": 3.3294382490892226e-06,
|
2213 |
+
"loss": 0.401,
|
2214 |
+
"step": 1570
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 0.6497524752475248,
|
2218 |
+
"grad_norm": 2.4835672414143604,
|
2219 |
+
"learning_rate": 3.2955403799890567e-06,
|
2220 |
+
"loss": 0.4005,
|
2221 |
+
"step": 1575
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 0.6518151815181518,
|
2225 |
+
"grad_norm": 2.3799907386582224,
|
2226 |
+
"learning_rate": 3.261730923305717e-06,
|
2227 |
+
"loss": 0.4026,
|
2228 |
+
"step": 1580
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 0.6538778877887789,
|
2232 |
+
"grad_norm": 2.486466648614617,
|
2233 |
+
"learning_rate": 3.2280116327773028e-06,
|
2234 |
+
"loss": 0.4048,
|
2235 |
+
"step": 1585
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 0.655940594059406,
|
2239 |
+
"grad_norm": 2.4042700940181523,
|
2240 |
+
"learning_rate": 3.194384257464884e-06,
|
2241 |
+
"loss": 0.3945,
|
2242 |
+
"step": 1590
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 0.658003300330033,
|
2246 |
+
"grad_norm": 2.327134482890906,
|
2247 |
+
"learning_rate": 3.160850541661779e-06,
|
2248 |
+
"loss": 0.4047,
|
2249 |
+
"step": 1595
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 0.6600660066006601,
|
2253 |
+
"grad_norm": 2.4535001026550844,
|
2254 |
+
"learning_rate": 3.127412224803068e-06,
|
2255 |
+
"loss": 0.3885,
|
2256 |
+
"step": 1600
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 0.6621287128712872,
|
2260 |
+
"grad_norm": 2.6395091605070786,
|
2261 |
+
"learning_rate": 3.094071041375375e-06,
|
2262 |
+
"loss": 0.4085,
|
2263 |
+
"step": 1605
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 0.6641914191419142,
|
2267 |
+
"grad_norm": 2.45831697505499,
|
2268 |
+
"learning_rate": 3.060828720826889e-06,
|
2269 |
+
"loss": 0.4063,
|
2270 |
+
"step": 1610
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 0.6662541254125413,
|
2274 |
+
"grad_norm": 2.25651867695093,
|
2275 |
+
"learning_rate": 3.0276869874776632e-06,
|
2276 |
+
"loss": 0.3991,
|
2277 |
+
"step": 1615
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 0.6683168316831684,
|
2281 |
+
"grad_norm": 2.36719278095781,
|
2282 |
+
"learning_rate": 2.994647560430167e-06,
|
2283 |
+
"loss": 0.3943,
|
2284 |
+
"step": 1620
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 0.6703795379537953,
|
2288 |
+
"grad_norm": 2.4049216832251035,
|
2289 |
+
"learning_rate": 2.961712153480118e-06,
|
2290 |
+
"loss": 0.4029,
|
2291 |
+
"step": 1625
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 0.6724422442244224,
|
2295 |
+
"grad_norm": 2.559007669700738,
|
2296 |
+
"learning_rate": 2.9288824750275803e-06,
|
2297 |
+
"loss": 0.4032,
|
2298 |
+
"step": 1630
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 0.6745049504950495,
|
2302 |
+
"grad_norm": 2.310036214506543,
|
2303 |
+
"learning_rate": 2.896160227988357e-06,
|
2304 |
+
"loss": 0.4005,
|
2305 |
+
"step": 1635
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 0.6765676567656765,
|
2309 |
+
"grad_norm": 2.43914641154883,
|
2310 |
+
"learning_rate": 2.8635471097056423e-06,
|
2311 |
+
"loss": 0.3943,
|
2312 |
+
"step": 1640
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 0.6786303630363036,
|
2316 |
+
"grad_norm": 2.443819723387033,
|
2317 |
+
"learning_rate": 2.8310448118619967e-06,
|
2318 |
+
"loss": 0.407,
|
2319 |
+
"step": 1645
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 0.6806930693069307,
|
2323 |
+
"grad_norm": 2.4768276955867794,
|
2324 |
+
"learning_rate": 2.7986550203915807e-06,
|
2325 |
+
"loss": 0.3909,
|
2326 |
+
"step": 1650
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 0.6827557755775577,
|
2330 |
+
"grad_norm": 2.3876332218325365,
|
2331 |
+
"learning_rate": 2.7663794153927165e-06,
|
2332 |
+
"loss": 0.3989,
|
2333 |
+
"step": 1655
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 0.6848184818481848,
|
2337 |
+
"grad_norm": 2.373869717825122,
|
2338 |
+
"learning_rate": 2.7342196710407337e-06,
|
2339 |
+
"loss": 0.3848,
|
2340 |
+
"step": 1660
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 0.6868811881188119,
|
2344 |
+
"grad_norm": 2.2944477847209703,
|
2345 |
+
"learning_rate": 2.7021774555011214e-06,
|
2346 |
+
"loss": 0.3952,
|
2347 |
+
"step": 1665
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 0.6889438943894389,
|
2351 |
+
"grad_norm": 2.4458499611784044,
|
2352 |
+
"learning_rate": 2.6702544308430122e-06,
|
2353 |
+
"loss": 0.3841,
|
2354 |
+
"step": 1670
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 0.691006600660066,
|
2358 |
+
"grad_norm": 2.7433908365448443,
|
2359 |
+
"learning_rate": 2.6384522529529542e-06,
|
2360 |
+
"loss": 0.383,
|
2361 |
+
"step": 1675
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 0.693069306930693,
|
2365 |
+
"grad_norm": 2.4164071285356052,
|
2366 |
+
"learning_rate": 2.6067725714490307e-06,
|
2367 |
+
"loss": 0.3841,
|
2368 |
+
"step": 1680
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 0.6951320132013201,
|
2372 |
+
"grad_norm": 2.5329201914328077,
|
2373 |
+
"learning_rate": 2.5752170295952856e-06,
|
2374 |
+
"loss": 0.4024,
|
2375 |
+
"step": 1685
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 0.6971947194719472,
|
2379 |
+
"grad_norm": 2.362743279232679,
|
2380 |
+
"learning_rate": 2.5437872642164818e-06,
|
2381 |
+
"loss": 0.3965,
|
2382 |
+
"step": 1690
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 0.6992574257425742,
|
2386 |
+
"grad_norm": 2.3597073251243064,
|
2387 |
+
"learning_rate": 2.5124849056132094e-06,
|
2388 |
+
"loss": 0.3658,
|
2389 |
+
"step": 1695
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 0.7013201320132013,
|
2393 |
+
"grad_norm": 2.645876828690864,
|
2394 |
+
"learning_rate": 2.4813115774773046e-06,
|
2395 |
+
"loss": 0.3884,
|
2396 |
+
"step": 1700
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 0.7033828382838284,
|
2400 |
+
"grad_norm": 2.392821189181109,
|
2401 |
+
"learning_rate": 2.4502688968076416e-06,
|
2402 |
+
"loss": 0.3827,
|
2403 |
+
"step": 1705
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 0.7054455445544554,
|
2407 |
+
"grad_norm": 2.4614918453264565,
|
2408 |
+
"learning_rate": 2.4193584738262426e-06,
|
2409 |
+
"loss": 0.3825,
|
2410 |
+
"step": 1710
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 0.7075082508250825,
|
2414 |
+
"grad_norm": 2.3794205872949252,
|
2415 |
+
"learning_rate": 2.388581911894767e-06,
|
2416 |
+
"loss": 0.3825,
|
2417 |
+
"step": 1715
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 0.7095709570957096,
|
2421 |
+
"grad_norm": 2.4464619731962634,
|
2422 |
+
"learning_rate": 2.357940807431339e-06,
|
2423 |
+
"loss": 0.3785,
|
2424 |
+
"step": 1720
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 0.7116336633663366,
|
2428 |
+
"grad_norm": 2.3852404958442013,
|
2429 |
+
"learning_rate": 2.3274367498277246e-06,
|
2430 |
+
"loss": 0.3773,
|
2431 |
+
"step": 1725
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 0.7136963696369637,
|
2435 |
+
"grad_norm": 2.376518354617449,
|
2436 |
+
"learning_rate": 2.2970713213669127e-06,
|
2437 |
+
"loss": 0.3831,
|
2438 |
+
"step": 1730
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 0.7157590759075908,
|
2442 |
+
"grad_norm": 2.6079917176596448,
|
2443 |
+
"learning_rate": 2.266846097141026e-06,
|
2444 |
+
"loss": 0.3857,
|
2445 |
+
"step": 1735
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 0.7178217821782178,
|
2449 |
+
"grad_norm": 2.3458483796617418,
|
2450 |
+
"learning_rate": 2.2367626449696168e-06,
|
2451 |
+
"loss": 0.3751,
|
2452 |
+
"step": 1740
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 0.7198844884488449,
|
2456 |
+
"grad_norm": 2.270229411474977,
|
2457 |
+
"learning_rate": 2.206822525318352e-06,
|
2458 |
+
"loss": 0.3784,
|
2459 |
+
"step": 1745
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 0.721947194719472,
|
2463 |
+
"grad_norm": 2.3820977460708077,
|
2464 |
+
"learning_rate": 2.1770272912180577e-06,
|
2465 |
+
"loss": 0.3726,
|
2466 |
+
"step": 1750
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 0.724009900990099,
|
2470 |
+
"grad_norm": 2.3917376427271746,
|
2471 |
+
"learning_rate": 2.1473784881841753e-06,
|
2472 |
+
"loss": 0.3749,
|
2473 |
+
"step": 1755
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 0.7260726072607261,
|
2477 |
+
"grad_norm": 2.430451456865234,
|
2478 |
+
"learning_rate": 2.117877654136584e-06,
|
2479 |
+
"loss": 0.3621,
|
2480 |
+
"step": 1760
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 0.7281353135313532,
|
2484 |
+
"grad_norm": 2.3121061406028107,
|
2485 |
+
"learning_rate": 2.088526319319827e-06,
|
2486 |
+
"loss": 0.3672,
|
2487 |
+
"step": 1765
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 0.7301980198019802,
|
2491 |
+
"grad_norm": 2.360101764240328,
|
2492 |
+
"learning_rate": 2.059326006223743e-06,
|
2493 |
+
"loss": 0.3717,
|
2494 |
+
"step": 1770
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 0.7322607260726073,
|
2498 |
+
"grad_norm": 2.573193311702909,
|
2499 |
+
"learning_rate": 2.030278229504484e-06,
|
2500 |
+
"loss": 0.379,
|
2501 |
+
"step": 1775
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 0.7343234323432343,
|
2505 |
+
"grad_norm": 2.3389637944355943,
|
2506 |
+
"learning_rate": 2.001384495905954e-06,
|
2507 |
+
"loss": 0.3752,
|
2508 |
+
"step": 1780
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 0.7363861386138614,
|
2512 |
+
"grad_norm": 2.488613631746104,
|
2513 |
+
"learning_rate": 1.972646304181656e-06,
|
2514 |
+
"loss": 0.3782,
|
2515 |
+
"step": 1785
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 0.7384488448844885,
|
2519 |
+
"grad_norm": 2.4022170179819096,
|
2520 |
+
"learning_rate": 1.944065145016935e-06,
|
2521 |
+
"loss": 0.3726,
|
2522 |
+
"step": 1790
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 0.7405115511551155,
|
2526 |
+
"grad_norm": 2.256385150261715,
|
2527 |
+
"learning_rate": 1.9156425009516736e-06,
|
2528 |
+
"loss": 0.3688,
|
2529 |
+
"step": 1795
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 0.7425742574257426,
|
2533 |
+
"grad_norm": 2.4167248024578813,
|
2534 |
+
"learning_rate": 1.8873798463033742e-06,
|
2535 |
+
"loss": 0.3684,
|
2536 |
+
"step": 1800
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 0.7446369636963697,
|
2540 |
+
"grad_norm": 2.451766042702295,
|
2541 |
+
"learning_rate": 1.8592786470906932e-06,
|
2542 |
+
"loss": 0.375,
|
2543 |
+
"step": 1805
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 0.7466996699669967,
|
2547 |
+
"grad_norm": 2.383004059186879,
|
2548 |
+
"learning_rate": 1.8313403609573976e-06,
|
2549 |
+
"loss": 0.3759,
|
2550 |
+
"step": 1810
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 0.7487623762376238,
|
2554 |
+
"grad_norm": 2.12270311435573,
|
2555 |
+
"learning_rate": 1.8035664370967493e-06,
|
2556 |
+
"loss": 0.3763,
|
2557 |
+
"step": 1815
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 0.7508250825082509,
|
2561 |
+
"grad_norm": 2.4415551404560847,
|
2562 |
+
"learning_rate": 1.775958316176339e-06,
|
2563 |
+
"loss": 0.3676,
|
2564 |
+
"step": 1820
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 0.7528877887788779,
|
2568 |
+
"grad_norm": 2.381919666851577,
|
2569 |
+
"learning_rate": 1.7485174302633557e-06,
|
2570 |
+
"loss": 0.3707,
|
2571 |
+
"step": 1825
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 0.754950495049505,
|
2575 |
+
"grad_norm": 2.468112725751951,
|
2576 |
+
"learning_rate": 1.721245202750299e-06,
|
2577 |
+
"loss": 0.3743,
|
2578 |
+
"step": 1830
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 0.7570132013201321,
|
2582 |
+
"grad_norm": 2.4412319646605707,
|
2583 |
+
"learning_rate": 1.694143048281156e-06,
|
2584 |
+
"loss": 0.3697,
|
2585 |
+
"step": 1835
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 0.759075907590759,
|
2589 |
+
"grad_norm": 2.3933092088835592,
|
2590 |
+
"learning_rate": 1.6672123726780083e-06,
|
2591 |
+
"loss": 0.362,
|
2592 |
+
"step": 1840
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 0.7611386138613861,
|
2596 |
+
"grad_norm": 2.3599050961874357,
|
2597 |
+
"learning_rate": 1.6404545728681232e-06,
|
2598 |
+
"loss": 0.3621,
|
2599 |
+
"step": 1845
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 0.7632013201320133,
|
2603 |
+
"grad_norm": 2.3969987242185407,
|
2604 |
+
"learning_rate": 1.613871036811489e-06,
|
2605 |
+
"loss": 0.3631,
|
2606 |
+
"step": 1850
|
2607 |
+
},
|
2608 |
+
{
|
2609 |
+
"epoch": 0.7652640264026402,
|
2610 |
+
"grad_norm": 2.3728886726584144,
|
2611 |
+
"learning_rate": 1.5874631434288128e-06,
|
2612 |
+
"loss": 0.3747,
|
2613 |
+
"step": 1855
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 0.7673267326732673,
|
2617 |
+
"grad_norm": 2.334377590582883,
|
2618 |
+
"learning_rate": 1.5612322625300064e-06,
|
2619 |
+
"loss": 0.3647,
|
2620 |
+
"step": 1860
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 0.7693894389438944,
|
2624 |
+
"grad_norm": 2.3804587872682865,
|
2625 |
+
"learning_rate": 1.5351797547431212e-06,
|
2626 |
+
"loss": 0.3694,
|
2627 |
+
"step": 1865
|
2628 |
+
},
|
2629 |
+
{
|
2630 |
+
"epoch": 0.7714521452145214,
|
2631 |
+
"grad_norm": 2.3600418503523444,
|
2632 |
+
"learning_rate": 1.5093069714437803e-06,
|
2633 |
+
"loss": 0.3636,
|
2634 |
+
"step": 1870
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 0.7735148514851485,
|
2638 |
+
"grad_norm": 2.553398892828675,
|
2639 |
+
"learning_rate": 1.483615254685075e-06,
|
2640 |
+
"loss": 0.3611,
|
2641 |
+
"step": 1875
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"epoch": 0.7755775577557755,
|
2645 |
+
"grad_norm": 2.3499524569210157,
|
2646 |
+
"learning_rate": 1.4581059371279516e-06,
|
2647 |
+
"loss": 0.3674,
|
2648 |
+
"step": 1880
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 0.7776402640264026,
|
2652 |
+
"grad_norm": 2.4025948209352848,
|
2653 |
+
"learning_rate": 1.4327803419720836e-06,
|
2654 |
+
"loss": 0.3609,
|
2655 |
+
"step": 1885
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 0.7797029702970297,
|
2659 |
+
"grad_norm": 2.44960933153857,
|
2660 |
+
"learning_rate": 1.4076397828872441e-06,
|
2661 |
+
"loss": 0.3546,
|
2662 |
+
"step": 1890
|
2663 |
+
},
|
2664 |
+
{
|
2665 |
+
"epoch": 0.7817656765676567,
|
2666 |
+
"grad_norm": 2.453171947739372,
|
2667 |
+
"learning_rate": 1.3826855639451492e-06,
|
2668 |
+
"loss": 0.3675,
|
2669 |
+
"step": 1895
|
2670 |
+
},
|
2671 |
+
{
|
2672 |
+
"epoch": 0.7838283828382838,
|
2673 |
+
"grad_norm": 2.243857634486722,
|
2674 |
+
"learning_rate": 1.357918979551831e-06,
|
2675 |
+
"loss": 0.3655,
|
2676 |
+
"step": 1900
|
2677 |
+
},
|
2678 |
+
{
|
2679 |
+
"epoch": 0.7858910891089109,
|
2680 |
+
"grad_norm": 2.334627109038789,
|
2681 |
+
"learning_rate": 1.333341314380479e-06,
|
2682 |
+
"loss": 0.3528,
|
2683 |
+
"step": 1905
|
2684 |
+
},
|
2685 |
+
{
|
2686 |
+
"epoch": 0.7879537953795379,
|
2687 |
+
"grad_norm": 2.4450603903304367,
|
2688 |
+
"learning_rate": 1.308953843304816e-06,
|
2689 |
+
"loss": 0.3696,
|
2690 |
+
"step": 1910
|
2691 |
+
},
|
2692 |
+
{
|
2693 |
+
"epoch": 0.790016501650165,
|
2694 |
+
"grad_norm": 2.469101543725218,
|
2695 |
+
"learning_rate": 1.2847578313329623e-06,
|
2696 |
+
"loss": 0.3722,
|
2697 |
+
"step": 1915
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 0.7920792079207921,
|
2701 |
+
"grad_norm": 2.347298229927237,
|
2702 |
+
"learning_rate": 1.2607545335418154e-06,
|
2703 |
+
"loss": 0.364,
|
2704 |
+
"step": 1920
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"epoch": 0.7941419141914191,
|
2708 |
+
"grad_norm": 2.478215187683226,
|
2709 |
+
"learning_rate": 1.2369451950119553e-06,
|
2710 |
+
"loss": 0.3634,
|
2711 |
+
"step": 1925
|
2712 |
+
},
|
2713 |
+
{
|
2714 |
+
"epoch": 0.7962046204620462,
|
2715 |
+
"grad_norm": 2.321071149582765,
|
2716 |
+
"learning_rate": 1.2133310507630535e-06,
|
2717 |
+
"loss": 0.3639,
|
2718 |
+
"step": 1930
|
2719 |
+
},
|
2720 |
+
{
|
2721 |
+
"epoch": 0.7982673267326733,
|
2722 |
+
"grad_norm": 2.363187519551145,
|
2723 |
+
"learning_rate": 1.189913325689816e-06,
|
2724 |
+
"loss": 0.3694,
|
2725 |
+
"step": 1935
|
2726 |
+
},
|
2727 |
+
{
|
2728 |
+
"epoch": 0.8003300330033003,
|
2729 |
+
"grad_norm": 2.1565024791128615,
|
2730 |
+
"learning_rate": 1.166693234498446e-06,
|
2731 |
+
"loss": 0.3444,
|
2732 |
+
"step": 1940
|
2733 |
+
},
|
2734 |
+
{
|
2735 |
+
"epoch": 0.8023927392739274,
|
2736 |
+
"grad_norm": 2.238362330023504,
|
2737 |
+
"learning_rate": 1.1436719816436293e-06,
|
2738 |
+
"loss": 0.3535,
|
2739 |
+
"step": 1945
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 0.8044554455445545,
|
2743 |
+
"grad_norm": 2.682507072878571,
|
2744 |
+
"learning_rate": 1.120850761266068e-06,
|
2745 |
+
"loss": 0.3622,
|
2746 |
+
"step": 1950
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 0.8065181518151815,
|
2750 |
+
"grad_norm": 2.4616448463288974,
|
2751 |
+
"learning_rate": 1.098230757130529e-06,
|
2752 |
+
"loss": 0.3521,
|
2753 |
+
"step": 1955
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 0.8085808580858086,
|
2757 |
+
"grad_norm": 2.518356814157187,
|
2758 |
+
"learning_rate": 1.075813142564448e-06,
|
2759 |
+
"loss": 0.3621,
|
2760 |
+
"step": 1960
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 0.8106435643564357,
|
2764 |
+
"grad_norm": 2.41717183499352,
|
2765 |
+
"learning_rate": 1.053599080397068e-06,
|
2766 |
+
"loss": 0.3534,
|
2767 |
+
"step": 1965
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 0.8127062706270627,
|
2771 |
+
"grad_norm": 2.3274563555835357,
|
2772 |
+
"learning_rate": 1.031589722899109e-06,
|
2773 |
+
"loss": 0.3497,
|
2774 |
+
"step": 1970
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 0.8147689768976898,
|
2778 |
+
"grad_norm": 2.41058923252155,
|
2779 |
+
"learning_rate": 1.0097862117230162e-06,
|
2780 |
+
"loss": 0.3453,
|
2781 |
+
"step": 1975
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 0.8168316831683168,
|
2785 |
+
"grad_norm": 2.441792027931522,
|
2786 |
+
"learning_rate": 9.881896778437328e-07,
|
2787 |
+
"loss": 0.3444,
|
2788 |
+
"step": 1980
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 0.8188943894389439,
|
2792 |
+
"grad_norm": 2.4018876695962383,
|
2793 |
+
"learning_rate": 9.6680124150003e-07,
|
2794 |
+
"loss": 0.3654,
|
2795 |
+
"step": 1985
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 0.820957095709571,
|
2799 |
+
"grad_norm": 2.3575482669225902,
|
2800 |
+
"learning_rate": 9.456220121364091e-07,
|
2801 |
+
"loss": 0.348,
|
2802 |
+
"step": 1990
|
2803 |
+
},
|
2804 |
+
{
|
2805 |
+
"epoch": 0.823019801980198,
|
2806 |
+
"grad_norm": 2.525821025154249,
|
2807 |
+
"learning_rate": 9.24653088345544e-07,
|
2808 |
+
"loss": 0.3523,
|
2809 |
+
"step": 1995
|
2810 |
+
},
|
2811 |
+
{
|
2812 |
+
"epoch": 0.8250825082508251,
|
2813 |
+
"grad_norm": 2.7783807540038294,
|
2814 |
+
"learning_rate": 9.038955578113018e-07,
|
2815 |
+
"loss": 0.3582,
|
2816 |
+
"step": 2000
|
2817 |
+
},
|
2818 |
+
{
|
2819 |
+
"epoch": 0.8271452145214522,
|
2820 |
+
"grad_norm": 2.50572956814031,
|
2821 |
+
"learning_rate": 8.833504972523238e-07,
|
2822 |
+
"loss": 0.3593,
|
2823 |
+
"step": 2005
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 0.8292079207920792,
|
2827 |
+
"grad_norm": 2.252557918607386,
|
2828 |
+
"learning_rate": 8.630189723661663e-07,
|
2829 |
+
"loss": 0.3536,
|
2830 |
+
"step": 2010
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"epoch": 0.8312706270627063,
|
2834 |
+
"grad_norm": 2.4651130321437758,
|
2835 |
+
"learning_rate": 8.429020377740338e-07,
|
2836 |
+
"loss": 0.3535,
|
2837 |
+
"step": 2015
|
2838 |
+
},
|
2839 |
+
{
|
2840 |
+
"epoch": 0.8333333333333334,
|
2841 |
+
"grad_norm": 2.400123594291822,
|
2842 |
+
"learning_rate": 8.230007369660636e-07,
|
2843 |
+
"loss": 0.3543,
|
2844 |
+
"step": 2020
|
2845 |
+
},
|
2846 |
+
{
|
2847 |
+
"epoch": 0.8353960396039604,
|
2848 |
+
"grad_norm": 2.2893414346006806,
|
2849 |
+
"learning_rate": 8.033161022472063e-07,
|
2850 |
+
"loss": 0.3633,
|
2851 |
+
"step": 2025
|
2852 |
+
},
|
2853 |
+
{
|
2854 |
+
"epoch": 0.8374587458745875,
|
2855 |
+
"grad_norm": 2.387886294048468,
|
2856 |
+
"learning_rate": 7.838491546836763e-07,
|
2857 |
+
"loss": 0.3501,
|
2858 |
+
"step": 2030
|
2859 |
+
},
|
2860 |
+
{
|
2861 |
+
"epoch": 0.8395214521452146,
|
2862 |
+
"grad_norm": 2.380305733427394,
|
2863 |
+
"learning_rate": 7.646009040499846e-07,
|
2864 |
+
"loss": 0.3396,
|
2865 |
+
"step": 2035
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 0.8415841584158416,
|
2869 |
+
"grad_norm": 2.6014881264846164,
|
2870 |
+
"learning_rate": 7.455723487765664e-07,
|
2871 |
+
"loss": 0.3586,
|
2872 |
+
"step": 2040
|
2873 |
+
},
|
2874 |
+
{
|
2875 |
+
"epoch": 0.8436468646864687,
|
2876 |
+
"grad_norm": 2.4907278984166106,
|
2877 |
+
"learning_rate": 7.267644758979869e-07,
|
2878 |
+
"loss": 0.3468,
|
2879 |
+
"step": 2045
|
2880 |
+
},
|
2881 |
+
{
|
2882 |
+
"epoch": 0.8457095709570958,
|
2883 |
+
"grad_norm": 2.34019203542543,
|
2884 |
+
"learning_rate": 7.08178261001743e-07,
|
2885 |
+
"loss": 0.3514,
|
2886 |
+
"step": 2050
|
2887 |
+
},
|
2888 |
+
{
|
2889 |
+
"epoch": 0.8477722772277227,
|
2890 |
+
"grad_norm": 2.3922323702113637,
|
2891 |
+
"learning_rate": 6.898146681776629e-07,
|
2892 |
+
"loss": 0.3486,
|
2893 |
+
"step": 2055
|
2894 |
+
},
|
2895 |
+
{
|
2896 |
+
"epoch": 0.8498349834983498,
|
2897 |
+
"grad_norm": 2.3826898973626127,
|
2898 |
+
"learning_rate": 6.7167464996789e-07,
|
2899 |
+
"loss": 0.3621,
|
2900 |
+
"step": 2060
|
2901 |
+
},
|
2902 |
+
{
|
2903 |
+
"epoch": 0.851897689768977,
|
2904 |
+
"grad_norm": 2.3165319710343324,
|
2905 |
+
"learning_rate": 6.537591473174814e-07,
|
2906 |
+
"loss": 0.3413,
|
2907 |
+
"step": 2065
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 0.8539603960396039,
|
2911 |
+
"grad_norm": 2.5692025408049064,
|
2912 |
+
"learning_rate": 6.360690895255916e-07,
|
2913 |
+
"loss": 0.3461,
|
2914 |
+
"step": 2070
|
2915 |
+
},
|
2916 |
+
{
|
2917 |
+
"epoch": 0.856023102310231,
|
2918 |
+
"grad_norm": 2.272769119087517,
|
2919 |
+
"learning_rate": 6.186053941972775e-07,
|
2920 |
+
"loss": 0.348,
|
2921 |
+
"step": 2075
|
2922 |
+
},
|
2923 |
+
{
|
2924 |
+
"epoch": 0.858085808580858,
|
2925 |
+
"grad_norm": 2.4550162262909447,
|
2926 |
+
"learning_rate": 6.013689671958944e-07,
|
2927 |
+
"loss": 0.3531,
|
2928 |
+
"step": 2080
|
2929 |
+
},
|
2930 |
+
{
|
2931 |
+
"epoch": 0.8601485148514851,
|
2932 |
+
"grad_norm": 2.388763908670921,
|
2933 |
+
"learning_rate": 5.84360702596109e-07,
|
2934 |
+
"loss": 0.3608,
|
2935 |
+
"step": 2085
|
2936 |
+
},
|
2937 |
+
{
|
2938 |
+
"epoch": 0.8622112211221122,
|
2939 |
+
"grad_norm": 2.5005604744641174,
|
2940 |
+
"learning_rate": 5.67581482637527e-07,
|
2941 |
+
"loss": 0.3444,
|
2942 |
+
"step": 2090
|
2943 |
+
},
|
2944 |
+
{
|
2945 |
+
"epoch": 0.8642739273927392,
|
2946 |
+
"grad_norm": 2.4560197080279234,
|
2947 |
+
"learning_rate": 5.510321776789213e-07,
|
2948 |
+
"loss": 0.3463,
|
2949 |
+
"step": 2095
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 0.8663366336633663,
|
2953 |
+
"grad_norm": 2.365882695902703,
|
2954 |
+
"learning_rate": 5.347136461530966e-07,
|
2955 |
+
"loss": 0.3366,
|
2956 |
+
"step": 2100
|
2957 |
+
},
|
2958 |
+
{
|
2959 |
+
"epoch": 0.8683993399339934,
|
2960 |
+
"grad_norm": 2.4380355384519925,
|
2961 |
+
"learning_rate": 5.186267345223539e-07,
|
2962 |
+
"loss": 0.3385,
|
2963 |
+
"step": 2105
|
2964 |
+
},
|
2965 |
+
{
|
2966 |
+
"epoch": 0.8704620462046204,
|
2967 |
+
"grad_norm": 2.399986551951994,
|
2968 |
+
"learning_rate": 5.027722772345828e-07,
|
2969 |
+
"loss": 0.3448,
|
2970 |
+
"step": 2110
|
2971 |
+
},
|
2972 |
+
{
|
2973 |
+
"epoch": 0.8725247524752475,
|
2974 |
+
"grad_norm": 2.454557551769108,
|
2975 |
+
"learning_rate": 4.871510966799847e-07,
|
2976 |
+
"loss": 0.3507,
|
2977 |
+
"step": 2115
|
2978 |
+
},
|
2979 |
+
{
|
2980 |
+
"epoch": 0.8745874587458746,
|
2981 |
+
"grad_norm": 2.456742477580607,
|
2982 |
+
"learning_rate": 4.717640031484055e-07,
|
2983 |
+
"loss": 0.3525,
|
2984 |
+
"step": 2120
|
2985 |
+
},
|
2986 |
+
{
|
2987 |
+
"epoch": 0.8766501650165016,
|
2988 |
+
"grad_norm": 2.41826728768274,
|
2989 |
+
"learning_rate": 4.566117947873139e-07,
|
2990 |
+
"loss": 0.3526,
|
2991 |
+
"step": 2125
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 0.8787128712871287,
|
2995 |
+
"grad_norm": 2.4671547541589183,
|
2996 |
+
"learning_rate": 4.4169525756039164e-07,
|
2997 |
+
"loss": 0.3375,
|
2998 |
+
"step": 2130
|
2999 |
+
},
|
3000 |
+
{
|
3001 |
+
"epoch": 0.8807755775577558,
|
3002 |
+
"grad_norm": 2.3142164149499793,
|
3003 |
+
"learning_rate": 4.2701516520677054e-07,
|
3004 |
+
"loss": 0.3428,
|
3005 |
+
"step": 2135
|
3006 |
+
},
|
3007 |
+
{
|
3008 |
+
"epoch": 0.8828382838283828,
|
3009 |
+
"grad_norm": 2.282591060676799,
|
3010 |
+
"learning_rate": 4.1257227920089684e-07,
|
3011 |
+
"loss": 0.315,
|
3012 |
+
"step": 2140
|
3013 |
+
},
|
3014 |
+
{
|
3015 |
+
"epoch": 0.8849009900990099,
|
3016 |
+
"grad_norm": 2.2945870317486445,
|
3017 |
+
"learning_rate": 3.983673487130313e-07,
|
3018 |
+
"loss": 0.3438,
|
3019 |
+
"step": 2145
|
3020 |
+
},
|
3021 |
+
{
|
3022 |
+
"epoch": 0.886963696369637,
|
3023 |
+
"grad_norm": 2.4023456680454003,
|
3024 |
+
"learning_rate": 3.8440111057038874e-07,
|
3025 |
+
"loss": 0.3466,
|
3026 |
+
"step": 2150
|
3027 |
+
},
|
3028 |
+
{
|
3029 |
+
"epoch": 0.889026402640264,
|
3030 |
+
"grad_norm": 2.4628827167617278,
|
3031 |
+
"learning_rate": 3.706742892189197e-07,
|
3032 |
+
"loss": 0.3578,
|
3033 |
+
"step": 2155
|
3034 |
+
},
|
3035 |
+
{
|
3036 |
+
"epoch": 0.8910891089108911,
|
3037 |
+
"grad_norm": 2.518821614068097,
|
3038 |
+
"learning_rate": 3.5718759668572913e-07,
|
3039 |
+
"loss": 0.3484,
|
3040 |
+
"step": 2160
|
3041 |
+
},
|
3042 |
+
{
|
3043 |
+
"epoch": 0.8931518151815182,
|
3044 |
+
"grad_norm": 2.4044153870357734,
|
3045 |
+
"learning_rate": 3.439417325421468e-07,
|
3046 |
+
"loss": 0.3331,
|
3047 |
+
"step": 2165
|
3048 |
+
},
|
3049 |
+
{
|
3050 |
+
"epoch": 0.8952145214521452,
|
3051 |
+
"grad_norm": 2.263400317365007,
|
3052 |
+
"learning_rate": 3.3093738386743734e-07,
|
3053 |
+
"loss": 0.3359,
|
3054 |
+
"step": 2170
|
3055 |
+
},
|
3056 |
+
{
|
3057 |
+
"epoch": 0.8972772277227723,
|
3058 |
+
"grad_norm": 2.4013680975289162,
|
3059 |
+
"learning_rate": 3.1817522521316034e-07,
|
3060 |
+
"loss": 0.3509,
|
3061 |
+
"step": 2175
|
3062 |
+
},
|
3063 |
+
{
|
3064 |
+
"epoch": 0.8993399339933993,
|
3065 |
+
"grad_norm": 2.4793048291891324,
|
3066 |
+
"learning_rate": 3.0565591856818236e-07,
|
3067 |
+
"loss": 0.3499,
|
3068 |
+
"step": 2180
|
3069 |
+
},
|
3070 |
+
{
|
3071 |
+
"epoch": 0.9014026402640264,
|
3072 |
+
"grad_norm": 2.524550570591509,
|
3073 |
+
"learning_rate": 2.9338011332433525e-07,
|
3074 |
+
"loss": 0.3314,
|
3075 |
+
"step": 2185
|
3076 |
+
},
|
3077 |
+
{
|
3078 |
+
"epoch": 0.9034653465346535,
|
3079 |
+
"grad_norm": 2.4135168075751294,
|
3080 |
+
"learning_rate": 2.813484462427357e-07,
|
3081 |
+
"loss": 0.3359,
|
3082 |
+
"step": 2190
|
3083 |
+
},
|
3084 |
+
{
|
3085 |
+
"epoch": 0.9055280528052805,
|
3086 |
+
"grad_norm": 2.4118495326051996,
|
3087 |
+
"learning_rate": 2.695615414207542e-07,
|
3088 |
+
"loss": 0.3499,
|
3089 |
+
"step": 2195
|
3090 |
+
},
|
3091 |
+
{
|
3092 |
+
"epoch": 0.9075907590759076,
|
3093 |
+
"grad_norm": 2.392533982344082,
|
3094 |
+
"learning_rate": 2.5802001025963917e-07,
|
3095 |
+
"loss": 0.3417,
|
3096 |
+
"step": 2200
|
3097 |
+
},
|
3098 |
+
{
|
3099 |
+
"epoch": 0.9096534653465347,
|
3100 |
+
"grad_norm": 2.2880229451881418,
|
3101 |
+
"learning_rate": 2.467244514328082e-07,
|
3102 |
+
"loss": 0.3493,
|
3103 |
+
"step": 2205
|
3104 |
+
},
|
3105 |
+
{
|
3106 |
+
"epoch": 0.9117161716171617,
|
3107 |
+
"grad_norm": 2.3091247857735504,
|
3108 |
+
"learning_rate": 2.3567545085478983e-07,
|
3109 |
+
"loss": 0.3383,
|
3110 |
+
"step": 2210
|
3111 |
+
},
|
3112 |
+
{
|
3113 |
+
"epoch": 0.9137788778877888,
|
3114 |
+
"grad_norm": 2.4901336136450136,
|
3115 |
+
"learning_rate": 2.248735816508324e-07,
|
3116 |
+
"loss": 0.3398,
|
3117 |
+
"step": 2215
|
3118 |
+
},
|
3119 |
+
{
|
3120 |
+
"epoch": 0.9158415841584159,
|
3121 |
+
"grad_norm": 2.3142795015199114,
|
3122 |
+
"learning_rate": 2.1431940412717843e-07,
|
3123 |
+
"loss": 0.3427,
|
3124 |
+
"step": 2220
|
3125 |
+
},
|
3126 |
+
{
|
3127 |
+
"epoch": 0.9179042904290429,
|
3128 |
+
"grad_norm": 2.4277647469057744,
|
3129 |
+
"learning_rate": 2.040134657419951e-07,
|
3130 |
+
"loss": 0.3387,
|
3131 |
+
"step": 2225
|
3132 |
+
},
|
3133 |
+
{
|
3134 |
+
"epoch": 0.91996699669967,
|
3135 |
+
"grad_norm": 2.3617745475614034,
|
3136 |
+
"learning_rate": 1.9395630107698293e-07,
|
3137 |
+
"loss": 0.3392,
|
3138 |
+
"step": 2230
|
3139 |
+
},
|
3140 |
+
{
|
3141 |
+
"epoch": 0.9220297029702971,
|
3142 |
+
"grad_norm": 2.5202449484112615,
|
3143 |
+
"learning_rate": 1.8414843180964316e-07,
|
3144 |
+
"loss": 0.3451,
|
3145 |
+
"step": 2235
|
3146 |
+
},
|
3147 |
+
{
|
3148 |
+
"epoch": 0.9240924092409241,
|
3149 |
+
"grad_norm": 2.6186027806455727,
|
3150 |
+
"learning_rate": 1.7459036668621586e-07,
|
3151 |
+
"loss": 0.345,
|
3152 |
+
"step": 2240
|
3153 |
+
},
|
3154 |
+
{
|
3155 |
+
"epoch": 0.9261551155115512,
|
3156 |
+
"grad_norm": 2.357402116923623,
|
3157 |
+
"learning_rate": 1.6528260149529573e-07,
|
3158 |
+
"loss": 0.3426,
|
3159 |
+
"step": 2245
|
3160 |
+
},
|
3161 |
+
{
|
3162 |
+
"epoch": 0.9282178217821783,
|
3163 |
+
"grad_norm": 2.7714941079884365,
|
3164 |
+
"learning_rate": 1.562256190421102e-07,
|
3165 |
+
"loss": 0.3444,
|
3166 |
+
"step": 2250
|
3167 |
+
},
|
3168 |
+
{
|
3169 |
+
"epoch": 0.9302805280528053,
|
3170 |
+
"grad_norm": 2.384372509379296,
|
3171 |
+
"learning_rate": 1.4741988912347848e-07,
|
3172 |
+
"loss": 0.349,
|
3173 |
+
"step": 2255
|
3174 |
+
},
|
3175 |
+
{
|
3176 |
+
"epoch": 0.9323432343234324,
|
3177 |
+
"grad_norm": 2.423642340690504,
|
3178 |
+
"learning_rate": 1.3886586850344276e-07,
|
3179 |
+
"loss": 0.3411,
|
3180 |
+
"step": 2260
|
3181 |
+
},
|
3182 |
+
{
|
3183 |
+
"epoch": 0.9344059405940595,
|
3184 |
+
"grad_norm": 2.4976812347623762,
|
3185 |
+
"learning_rate": 1.30564000889572e-07,
|
3186 |
+
"loss": 0.3427,
|
3187 |
+
"step": 2265
|
3188 |
+
},
|
3189 |
+
{
|
3190 |
+
"epoch": 0.9364686468646864,
|
3191 |
+
"grad_norm": 2.4409748992106244,
|
3192 |
+
"learning_rate": 1.225147169099511e-07,
|
3193 |
+
"loss": 0.3344,
|
3194 |
+
"step": 2270
|
3195 |
+
},
|
3196 |
+
{
|
3197 |
+
"epoch": 0.9385313531353136,
|
3198 |
+
"grad_norm": 2.4594177758611226,
|
3199 |
+
"learning_rate": 1.147184340908386e-07,
|
3200 |
+
"loss": 0.3466,
|
3201 |
+
"step": 2275
|
3202 |
+
},
|
3203 |
+
{
|
3204 |
+
"epoch": 0.9405940594059405,
|
3205 |
+
"grad_norm": 2.3563533473196054,
|
3206 |
+
"learning_rate": 1.0717555683501413e-07,
|
3207 |
+
"loss": 0.3437,
|
3208 |
+
"step": 2280
|
3209 |
+
},
|
3210 |
+
{
|
3211 |
+
"epoch": 0.9426567656765676,
|
3212 |
+
"grad_norm": 2.438141654343306,
|
3213 |
+
"learning_rate": 9.988647640079785e-08,
|
3214 |
+
"loss": 0.3481,
|
3215 |
+
"step": 2285
|
3216 |
+
},
|
3217 |
+
{
|
3218 |
+
"epoch": 0.9447194719471947,
|
3219 |
+
"grad_norm": 2.3653534995422465,
|
3220 |
+
"learning_rate": 9.285157088175678e-08,
|
3221 |
+
"loss": 0.3305,
|
3222 |
+
"step": 2290
|
3223 |
+
},
|
3224 |
+
{
|
3225 |
+
"epoch": 0.9467821782178217,
|
3226 |
+
"grad_norm": 2.5595084581881893,
|
3227 |
+
"learning_rate": 8.607120518709156e-08,
|
3228 |
+
"loss": 0.3447,
|
3229 |
+
"step": 2295
|
3230 |
+
},
|
3231 |
+
{
|
3232 |
+
"epoch": 0.9488448844884488,
|
3233 |
+
"grad_norm": 2.298821712031628,
|
3234 |
+
"learning_rate": 7.954573102271157e-08,
|
3235 |
+
"loss": 0.3427,
|
3236 |
+
"step": 2300
|
3237 |
+
},
|
3238 |
+
{
|
3239 |
+
"epoch": 0.9509075907590759,
|
3240 |
+
"grad_norm": 2.3999698309349786,
|
3241 |
+
"learning_rate": 7.327548687298625e-08,
|
3242 |
+
"loss": 0.3387,
|
3243 |
+
"step": 2305
|
3244 |
+
},
|
3245 |
+
{
|
3246 |
+
"epoch": 0.9529702970297029,
|
3247 |
+
"grad_norm": 2.2525707814009617,
|
3248 |
+
"learning_rate": 6.726079798319185e-08,
|
3249 |
+
"loss": 0.3561,
|
3250 |
+
"step": 2310
|
3251 |
+
},
|
3252 |
+
{
|
3253 |
+
"epoch": 0.95503300330033,
|
3254 |
+
"grad_norm": 2.3217448485216248,
|
3255 |
+
"learning_rate": 6.150197634263888e-08,
|
3256 |
+
"loss": 0.3469,
|
3257 |
+
"step": 2315
|
3258 |
+
},
|
3259 |
+
{
|
3260 |
+
"epoch": 0.9570957095709571,
|
3261 |
+
"grad_norm": 2.365101163841893,
|
3262 |
+
"learning_rate": 5.599932066848834e-08,
|
3263 |
+
"loss": 0.3367,
|
3264 |
+
"step": 2320
|
3265 |
+
},
|
3266 |
+
{
|
3267 |
+
"epoch": 0.9591584158415841,
|
3268 |
+
"grad_norm": 2.425841031506538,
|
3269 |
+
"learning_rate": 5.0753116390258594e-08,
|
3270 |
+
"loss": 0.3401,
|
3271 |
+
"step": 2325
|
3272 |
+
},
|
3273 |
+
{
|
3274 |
+
"epoch": 0.9612211221122112,
|
3275 |
+
"grad_norm": 2.563836501970768,
|
3276 |
+
"learning_rate": 4.576363563501718e-08,
|
3277 |
+
"loss": 0.3409,
|
3278 |
+
"step": 2330
|
3279 |
+
},
|
3280 |
+
{
|
3281 |
+
"epoch": 0.9632838283828383,
|
3282 |
+
"grad_norm": 2.34358481427922,
|
3283 |
+
"learning_rate": 4.103113721326768e-08,
|
3284 |
+
"loss": 0.3484,
|
3285 |
+
"step": 2335
|
3286 |
+
},
|
3287 |
+
{
|
3288 |
+
"epoch": 0.9653465346534653,
|
3289 |
+
"grad_norm": 2.402352860387557,
|
3290 |
+
"learning_rate": 3.655586660552324e-08,
|
3291 |
+
"loss": 0.3419,
|
3292 |
+
"step": 2340
|
3293 |
+
},
|
3294 |
+
{
|
3295 |
+
"epoch": 0.9674092409240924,
|
3296 |
+
"grad_norm": 2.412967620073454,
|
3297 |
+
"learning_rate": 3.233805594957506e-08,
|
3298 |
+
"loss": 0.3272,
|
3299 |
+
"step": 2345
|
3300 |
+
},
|
3301 |
+
{
|
3302 |
+
"epoch": 0.9694719471947195,
|
3303 |
+
"grad_norm": 2.3840732180526487,
|
3304 |
+
"learning_rate": 2.8377924028449855e-08,
|
3305 |
+
"loss": 0.3383,
|
3306 |
+
"step": 2350
|
3307 |
+
},
|
3308 |
+
{
|
3309 |
+
"epoch": 0.9715346534653465,
|
3310 |
+
"grad_norm": 2.508928823648118,
|
3311 |
+
"learning_rate": 2.4675676259059976e-08,
|
3312 |
+
"loss": 0.3447,
|
3313 |
+
"step": 2355
|
3314 |
+
},
|
3315 |
+
{
|
3316 |
+
"epoch": 0.9735973597359736,
|
3317 |
+
"grad_norm": 2.1993358612568317,
|
3318 |
+
"learning_rate": 2.123150468155144e-08,
|
3319 |
+
"loss": 0.3294,
|
3320 |
+
"step": 2360
|
3321 |
+
},
|
3322 |
+
{
|
3323 |
+
"epoch": 0.9756600660066007,
|
3324 |
+
"grad_norm": 2.4190680800537567,
|
3325 |
+
"learning_rate": 1.8045587949339637e-08,
|
3326 |
+
"loss": 0.3351,
|
3327 |
+
"step": 2365
|
3328 |
+
},
|
3329 |
+
{
|
3330 |
+
"epoch": 0.9777227722772277,
|
3331 |
+
"grad_norm": 2.2890929930955037,
|
3332 |
+
"learning_rate": 1.5118091319843985e-08,
|
3333 |
+
"loss": 0.34,
|
3334 |
+
"step": 2370
|
3335 |
+
},
|
3336 |
+
{
|
3337 |
+
"epoch": 0.9797854785478548,
|
3338 |
+
"grad_norm": 2.5688702019838883,
|
3339 |
+
"learning_rate": 1.2449166645915333e-08,
|
3340 |
+
"loss": 0.3304,
|
3341 |
+
"step": 2375
|
3342 |
+
},
|
3343 |
+
{
|
3344 |
+
"epoch": 0.9818481848184818,
|
3345 |
+
"grad_norm": 2.483690216637316,
|
3346 |
+
"learning_rate": 1.0038952367958376e-08,
|
3347 |
+
"loss": 0.341,
|
3348 |
+
"step": 2380
|
3349 |
+
},
|
3350 |
+
{
|
3351 |
+
"epoch": 0.9839108910891089,
|
3352 |
+
"grad_norm": 2.179239333588943,
|
3353 |
+
"learning_rate": 7.887573506752954e-09,
|
3354 |
+
"loss": 0.3384,
|
3355 |
+
"step": 2385
|
3356 |
+
},
|
3357 |
+
{
|
3358 |
+
"epoch": 0.985973597359736,
|
3359 |
+
"grad_norm": 2.512101769493497,
|
3360 |
+
"learning_rate": 5.9951416569659085e-09,
|
3361 |
+
"loss": 0.3348,
|
3362 |
+
"step": 2390
|
3363 |
+
},
|
3364 |
+
{
|
3365 |
+
"epoch": 0.988036303630363,
|
3366 |
+
"grad_norm": 2.4449313798457597,
|
3367 |
+
"learning_rate": 4.361754981365152e-09,
|
3368 |
+
"loss": 0.3339,
|
3369 |
+
"step": 2395
|
3370 |
+
},
|
3371 |
+
{
|
3372 |
+
"epoch": 0.9900990099009901,
|
3373 |
+
"grad_norm": 2.2966537435115444,
|
3374 |
+
"learning_rate": 2.98749820572708e-09,
|
3375 |
+
"loss": 0.3263,
|
3376 |
+
"step": 2400
|
3377 |
+
},
|
3378 |
+
{
|
3379 |
+
"epoch": 0.9921617161716172,
|
3380 |
+
"grad_norm": 2.5187709917417727,
|
3381 |
+
"learning_rate": 1.8724426144395293e-09,
|
3382 |
+
"loss": 0.3357,
|
3383 |
+
"step": 2405
|
3384 |
+
},
|
3385 |
+
{
|
3386 |
+
"epoch": 0.9942244224422442,
|
3387 |
+
"grad_norm": 2.340909896310804,
|
3388 |
+
"learning_rate": 1.0166460468080674e-09,
|
3389 |
+
"loss": 0.3426,
|
3390 |
+
"step": 2410
|
3391 |
+
},
|
3392 |
+
{
|
3393 |
+
"epoch": 0.9962871287128713,
|
3394 |
+
"grad_norm": 2.3871751186126207,
|
3395 |
+
"learning_rate": 4.2015289405339386e-10,
|
3396 |
+
"loss": 0.3369,
|
3397 |
+
"step": 2415
|
3398 |
+
},
|
3399 |
+
{
|
3400 |
+
"epoch": 0.9983498349834984,
|
3401 |
+
"grad_norm": 2.377304997896102,
|
3402 |
+
"learning_rate": 8.29940970092924e-11,
|
3403 |
+
"loss": 0.3308,
|
3404 |
+
"step": 2420
|
3405 |
+
},
|
3406 |
+
{
|
3407 |
+
"epoch": 1.0,
|
3408 |
+
"eval_runtime": 3.3729,
|
3409 |
+
"eval_samples_per_second": 2.965,
|
3410 |
+
"eval_steps_per_second": 0.889,
|
3411 |
+
"step": 2424
|
3412 |
+
},
|
3413 |
+
{
|
3414 |
+
"epoch": 1.0,
|
3415 |
+
"step": 2424,
|
3416 |
+
"total_flos": 253768142684160.0,
|
3417 |
+
"train_loss": 0.5300087753695624,
|
3418 |
+
"train_runtime": 21368.6318,
|
3419 |
+
"train_samples_per_second": 1.814,
|
3420 |
+
"train_steps_per_second": 0.113
|
3421 |
+
}
|
3422 |
+
],
|
3423 |
+
"logging_steps": 5,
|
3424 |
+
"max_steps": 2424,
|
3425 |
+
"num_input_tokens_seen": 0,
|
3426 |
+
"num_train_epochs": 1,
|
3427 |
+
"save_steps": 100,
|
3428 |
+
"stateful_callbacks": {
|
3429 |
+
"TrainerControl": {
|
3430 |
+
"args": {
|
3431 |
+
"should_epoch_stop": false,
|
3432 |
+
"should_evaluate": false,
|
3433 |
+
"should_log": false,
|
3434 |
+
"should_save": true,
|
3435 |
+
"should_training_stop": true
|
3436 |
+
},
|
3437 |
+
"attributes": {}
|
3438 |
+
}
|
3439 |
+
},
|
3440 |
+
"total_flos": 253768142684160.0,
|
3441 |
+
"train_batch_size": 4,
|
3442 |
+
"trial_name": null,
|
3443 |
+
"trial_params": null
|
3444 |
+
}
|