Model save
Browse files- README.md +60 -0
- all_results.json +9 -0
- generation_config.json +6 -0
- runs/Sep20_09-49-58_COE-CS-sv003/events.out.tfevents.1726826250.COE-CS-sv003.479391.0 +2 -2
- train_results.json +9 -0
- trainer_state.json +3430 -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_syntaxdirtybatch16
|
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_syntaxdirtybatch16
|
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 4.44.2
|
58 |
+
- Pytorch 2.3.1+cu121
|
59 |
+
- Datasets 2.19.1
|
60 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"total_flos": 252302485094400.0,
|
4 |
+
"train_loss": 0.5277333631555075,
|
5 |
+
"train_runtime": 21339.31,
|
6 |
+
"train_samples": 98873,
|
7 |
+
"train_samples_per_second": 1.807,
|
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": "4.44.2"
|
6 |
+
}
|
runs/Sep20_09-49-58_COE-CS-sv003/events.out.tfevents.1726826250.COE-CS-sv003.479391.0
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 sha256:09f158f2d25013371f0da94d5daf6d821338e7372c9c2e95d7aecd259bfc14f7
|
3 |
+
size 107668
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"total_flos": 252302485094400.0,
|
4 |
+
"train_loss": 0.5277333631555075,
|
5 |
+
"train_runtime": 21339.31,
|
6 |
+
"train_samples": 98873,
|
7 |
+
"train_samples_per_second": 1.807,
|
8 |
+
"train_steps_per_second": 0.113
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,3430 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 1.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 2410,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.0004149377593360996,
|
13 |
+
"grad_norm": 26.93341562765044,
|
14 |
+
"learning_rate": 4.1493775933609963e-08,
|
15 |
+
"loss": 1.4725,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.002074688796680498,
|
20 |
+
"grad_norm": 25.297316212761707,
|
21 |
+
"learning_rate": 2.074688796680498e-07,
|
22 |
+
"loss": 1.4346,
|
23 |
+
"step": 5
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.004149377593360996,
|
27 |
+
"grad_norm": 16.240075815340298,
|
28 |
+
"learning_rate": 4.149377593360996e-07,
|
29 |
+
"loss": 1.3793,
|
30 |
+
"step": 10
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.006224066390041493,
|
34 |
+
"grad_norm": 8.414236179504735,
|
35 |
+
"learning_rate": 6.224066390041494e-07,
|
36 |
+
"loss": 1.274,
|
37 |
+
"step": 15
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.008298755186721992,
|
41 |
+
"grad_norm": 10.510339353255072,
|
42 |
+
"learning_rate": 8.298755186721992e-07,
|
43 |
+
"loss": 1.1601,
|
44 |
+
"step": 20
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.01037344398340249,
|
48 |
+
"grad_norm": 4.437200517809562,
|
49 |
+
"learning_rate": 1.037344398340249e-06,
|
50 |
+
"loss": 1.0375,
|
51 |
+
"step": 25
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.012448132780082987,
|
55 |
+
"grad_norm": 3.4287551512904613,
|
56 |
+
"learning_rate": 1.2448132780082988e-06,
|
57 |
+
"loss": 0.9974,
|
58 |
+
"step": 30
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.014522821576763486,
|
62 |
+
"grad_norm": 3.4049944593465016,
|
63 |
+
"learning_rate": 1.4522821576763488e-06,
|
64 |
+
"loss": 0.9596,
|
65 |
+
"step": 35
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.016597510373443983,
|
69 |
+
"grad_norm": 3.0107447624660275,
|
70 |
+
"learning_rate": 1.6597510373443984e-06,
|
71 |
+
"loss": 0.938,
|
72 |
+
"step": 40
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.01867219917012448,
|
76 |
+
"grad_norm": 3.0215484563628694,
|
77 |
+
"learning_rate": 1.8672199170124482e-06,
|
78 |
+
"loss": 0.9083,
|
79 |
+
"step": 45
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.02074688796680498,
|
83 |
+
"grad_norm": 2.973401033499103,
|
84 |
+
"learning_rate": 2.074688796680498e-06,
|
85 |
+
"loss": 0.9175,
|
86 |
+
"step": 50
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.022821576763485476,
|
90 |
+
"grad_norm": 3.1166975342093104,
|
91 |
+
"learning_rate": 2.282157676348548e-06,
|
92 |
+
"loss": 0.9051,
|
93 |
+
"step": 55
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.024896265560165973,
|
97 |
+
"grad_norm": 3.1493583912363667,
|
98 |
+
"learning_rate": 2.4896265560165977e-06,
|
99 |
+
"loss": 0.8867,
|
100 |
+
"step": 60
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.026970954356846474,
|
104 |
+
"grad_norm": 2.971810139692584,
|
105 |
+
"learning_rate": 2.6970954356846475e-06,
|
106 |
+
"loss": 0.9024,
|
107 |
+
"step": 65
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.029045643153526972,
|
111 |
+
"grad_norm": 2.9673454675974535,
|
112 |
+
"learning_rate": 2.9045643153526977e-06,
|
113 |
+
"loss": 0.8855,
|
114 |
+
"step": 70
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.03112033195020747,
|
118 |
+
"grad_norm": 3.195994891124804,
|
119 |
+
"learning_rate": 3.112033195020747e-06,
|
120 |
+
"loss": 0.8873,
|
121 |
+
"step": 75
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.03319502074688797,
|
125 |
+
"grad_norm": 3.0341908061093603,
|
126 |
+
"learning_rate": 3.319502074688797e-06,
|
127 |
+
"loss": 0.8785,
|
128 |
+
"step": 80
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.035269709543568464,
|
132 |
+
"grad_norm": 3.0888190331536958,
|
133 |
+
"learning_rate": 3.526970954356847e-06,
|
134 |
+
"loss": 0.8686,
|
135 |
+
"step": 85
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.03734439834024896,
|
139 |
+
"grad_norm": 3.069834595940876,
|
140 |
+
"learning_rate": 3.7344398340248965e-06,
|
141 |
+
"loss": 0.8795,
|
142 |
+
"step": 90
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.03941908713692946,
|
146 |
+
"grad_norm": 3.07342880372721,
|
147 |
+
"learning_rate": 3.941908713692946e-06,
|
148 |
+
"loss": 0.8543,
|
149 |
+
"step": 95
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.04149377593360996,
|
153 |
+
"grad_norm": 3.149595655831201,
|
154 |
+
"learning_rate": 4.149377593360996e-06,
|
155 |
+
"loss": 0.8516,
|
156 |
+
"step": 100
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.043568464730290454,
|
160 |
+
"grad_norm": 3.1588136250157293,
|
161 |
+
"learning_rate": 4.356846473029046e-06,
|
162 |
+
"loss": 0.861,
|
163 |
+
"step": 105
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.04564315352697095,
|
167 |
+
"grad_norm": 3.2472498773631298,
|
168 |
+
"learning_rate": 4.564315352697096e-06,
|
169 |
+
"loss": 0.8626,
|
170 |
+
"step": 110
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.04771784232365145,
|
174 |
+
"grad_norm": 3.0627212516038003,
|
175 |
+
"learning_rate": 4.771784232365146e-06,
|
176 |
+
"loss": 0.8585,
|
177 |
+
"step": 115
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.04979253112033195,
|
181 |
+
"grad_norm": 3.3866567736302415,
|
182 |
+
"learning_rate": 4.979253112033195e-06,
|
183 |
+
"loss": 0.8546,
|
184 |
+
"step": 120
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.05186721991701245,
|
188 |
+
"grad_norm": 3.045817807536523,
|
189 |
+
"learning_rate": 5.1867219917012455e-06,
|
190 |
+
"loss": 0.8579,
|
191 |
+
"step": 125
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.05394190871369295,
|
195 |
+
"grad_norm": 2.9392093861675868,
|
196 |
+
"learning_rate": 5.394190871369295e-06,
|
197 |
+
"loss": 0.8282,
|
198 |
+
"step": 130
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.056016597510373446,
|
202 |
+
"grad_norm": 3.310150868761337,
|
203 |
+
"learning_rate": 5.601659751037345e-06,
|
204 |
+
"loss": 0.846,
|
205 |
+
"step": 135
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.058091286307053944,
|
209 |
+
"grad_norm": 3.03717223043934,
|
210 |
+
"learning_rate": 5.809128630705395e-06,
|
211 |
+
"loss": 0.8459,
|
212 |
+
"step": 140
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.06016597510373444,
|
216 |
+
"grad_norm": 3.028995697830279,
|
217 |
+
"learning_rate": 6.016597510373444e-06,
|
218 |
+
"loss": 0.8488,
|
219 |
+
"step": 145
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.06224066390041494,
|
223 |
+
"grad_norm": 3.2031060057481935,
|
224 |
+
"learning_rate": 6.224066390041494e-06,
|
225 |
+
"loss": 0.8323,
|
226 |
+
"step": 150
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.06431535269709543,
|
230 |
+
"grad_norm": 3.3375024554537545,
|
231 |
+
"learning_rate": 6.431535269709544e-06,
|
232 |
+
"loss": 0.8365,
|
233 |
+
"step": 155
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.06639004149377593,
|
237 |
+
"grad_norm": 3.348800008787665,
|
238 |
+
"learning_rate": 6.639004149377594e-06,
|
239 |
+
"loss": 0.8337,
|
240 |
+
"step": 160
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.06846473029045644,
|
244 |
+
"grad_norm": 3.052961732749455,
|
245 |
+
"learning_rate": 6.846473029045644e-06,
|
246 |
+
"loss": 0.8315,
|
247 |
+
"step": 165
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.07053941908713693,
|
251 |
+
"grad_norm": 3.1121690085579607,
|
252 |
+
"learning_rate": 7.053941908713694e-06,
|
253 |
+
"loss": 0.8401,
|
254 |
+
"step": 170
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.07261410788381743,
|
258 |
+
"grad_norm": 3.242023378743259,
|
259 |
+
"learning_rate": 7.261410788381743e-06,
|
260 |
+
"loss": 0.8158,
|
261 |
+
"step": 175
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.07468879668049792,
|
265 |
+
"grad_norm": 3.075162049471092,
|
266 |
+
"learning_rate": 7.468879668049793e-06,
|
267 |
+
"loss": 0.8347,
|
268 |
+
"step": 180
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.07676348547717843,
|
272 |
+
"grad_norm": 2.9264920323994255,
|
273 |
+
"learning_rate": 7.676348547717844e-06,
|
274 |
+
"loss": 0.8088,
|
275 |
+
"step": 185
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.07883817427385892,
|
279 |
+
"grad_norm": 2.948569896440745,
|
280 |
+
"learning_rate": 7.883817427385892e-06,
|
281 |
+
"loss": 0.8218,
|
282 |
+
"step": 190
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.08091286307053942,
|
286 |
+
"grad_norm": 3.3517457188955224,
|
287 |
+
"learning_rate": 8.091286307053943e-06,
|
288 |
+
"loss": 0.8147,
|
289 |
+
"step": 195
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.08298755186721991,
|
293 |
+
"grad_norm": 3.220462454017356,
|
294 |
+
"learning_rate": 8.298755186721992e-06,
|
295 |
+
"loss": 0.8349,
|
296 |
+
"step": 200
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.08506224066390042,
|
300 |
+
"grad_norm": 3.1111446024937006,
|
301 |
+
"learning_rate": 8.506224066390042e-06,
|
302 |
+
"loss": 0.8057,
|
303 |
+
"step": 205
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.08713692946058091,
|
307 |
+
"grad_norm": 2.892565595009811,
|
308 |
+
"learning_rate": 8.713692946058093e-06,
|
309 |
+
"loss": 0.8294,
|
310 |
+
"step": 210
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.08921161825726141,
|
314 |
+
"grad_norm": 3.235067221557126,
|
315 |
+
"learning_rate": 8.921161825726142e-06,
|
316 |
+
"loss": 0.7982,
|
317 |
+
"step": 215
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.0912863070539419,
|
321 |
+
"grad_norm": 3.1287542161799013,
|
322 |
+
"learning_rate": 9.128630705394191e-06,
|
323 |
+
"loss": 0.7973,
|
324 |
+
"step": 220
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.09336099585062241,
|
328 |
+
"grad_norm": 2.9450638727706315,
|
329 |
+
"learning_rate": 9.33609958506224e-06,
|
330 |
+
"loss": 0.8147,
|
331 |
+
"step": 225
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.0954356846473029,
|
335 |
+
"grad_norm": 3.170060713175617,
|
336 |
+
"learning_rate": 9.543568464730292e-06,
|
337 |
+
"loss": 0.8161,
|
338 |
+
"step": 230
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.0975103734439834,
|
342 |
+
"grad_norm": 3.0907201923890146,
|
343 |
+
"learning_rate": 9.751037344398341e-06,
|
344 |
+
"loss": 0.8163,
|
345 |
+
"step": 235
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.0995850622406639,
|
349 |
+
"grad_norm": 3.102859260824346,
|
350 |
+
"learning_rate": 9.95850622406639e-06,
|
351 |
+
"loss": 0.7918,
|
352 |
+
"step": 240
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.1016597510373444,
|
356 |
+
"grad_norm": 2.9868921661921517,
|
357 |
+
"learning_rate": 9.999916085034977e-06,
|
358 |
+
"loss": 0.8164,
|
359 |
+
"step": 245
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.1037344398340249,
|
363 |
+
"grad_norm": 2.8504855656123937,
|
364 |
+
"learning_rate": 9.999575185316994e-06,
|
365 |
+
"loss": 0.8105,
|
366 |
+
"step": 250
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.10580912863070539,
|
370 |
+
"grad_norm": 2.969587333276916,
|
371 |
+
"learning_rate": 9.998972074026074e-06,
|
372 |
+
"loss": 0.8055,
|
373 |
+
"step": 255
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.1078838174273859,
|
377 |
+
"grad_norm": 2.9939024309770157,
|
378 |
+
"learning_rate": 9.998106782793455e-06,
|
379 |
+
"loss": 0.7973,
|
380 |
+
"step": 260
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.10995850622406639,
|
384 |
+
"grad_norm": 2.997723701099069,
|
385 |
+
"learning_rate": 9.996979357000869e-06,
|
386 |
+
"loss": 0.7941,
|
387 |
+
"step": 265
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.11203319502074689,
|
391 |
+
"grad_norm": 2.979776887568956,
|
392 |
+
"learning_rate": 9.995589855778159e-06,
|
393 |
+
"loss": 0.7897,
|
394 |
+
"step": 270
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.11410788381742738,
|
398 |
+
"grad_norm": 2.7677213383011443,
|
399 |
+
"learning_rate": 9.993938352000174e-06,
|
400 |
+
"loss": 0.7962,
|
401 |
+
"step": 275
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.11618257261410789,
|
405 |
+
"grad_norm": 2.936072860527269,
|
406 |
+
"learning_rate": 9.992024932282955e-06,
|
407 |
+
"loss": 0.7934,
|
408 |
+
"step": 280
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.11825726141078838,
|
412 |
+
"grad_norm": 2.7901975755448003,
|
413 |
+
"learning_rate": 9.989849696979188e-06,
|
414 |
+
"loss": 0.8062,
|
415 |
+
"step": 285
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.12033195020746888,
|
419 |
+
"grad_norm": 2.8991854152652974,
|
420 |
+
"learning_rate": 9.987412760172939e-06,
|
421 |
+
"loss": 0.8007,
|
422 |
+
"step": 290
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.12240663900414937,
|
426 |
+
"grad_norm": 3.0373899188891267,
|
427 |
+
"learning_rate": 9.984714249673676e-06,
|
428 |
+
"loss": 0.7941,
|
429 |
+
"step": 295
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.12448132780082988,
|
433 |
+
"grad_norm": 2.8040271211308596,
|
434 |
+
"learning_rate": 9.981754307009556e-06,
|
435 |
+
"loss": 0.8004,
|
436 |
+
"step": 300
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.12655601659751037,
|
440 |
+
"grad_norm": 2.8585104734793263,
|
441 |
+
"learning_rate": 9.978533087420015e-06,
|
442 |
+
"loss": 0.8029,
|
443 |
+
"step": 305
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.12863070539419086,
|
447 |
+
"grad_norm": 2.87456284995761,
|
448 |
+
"learning_rate": 9.97505075984762e-06,
|
449 |
+
"loss": 0.7807,
|
450 |
+
"step": 310
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.13070539419087138,
|
454 |
+
"grad_norm": 3.056578905955286,
|
455 |
+
"learning_rate": 9.971307506929202e-06,
|
456 |
+
"loss": 0.7795,
|
457 |
+
"step": 315
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.13278008298755187,
|
461 |
+
"grad_norm": 2.722917451156625,
|
462 |
+
"learning_rate": 9.967303524986294e-06,
|
463 |
+
"loss": 0.7873,
|
464 |
+
"step": 320
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.13485477178423236,
|
468 |
+
"grad_norm": 2.9180827249941013,
|
469 |
+
"learning_rate": 9.963039024014811e-06,
|
470 |
+
"loss": 0.7861,
|
471 |
+
"step": 325
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.13692946058091288,
|
475 |
+
"grad_norm": 2.73431405256196,
|
476 |
+
"learning_rate": 9.958514227674065e-06,
|
477 |
+
"loss": 0.766,
|
478 |
+
"step": 330
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.13900414937759337,
|
482 |
+
"grad_norm": 2.9085046091623137,
|
483 |
+
"learning_rate": 9.953729373275008e-06,
|
484 |
+
"loss": 0.7731,
|
485 |
+
"step": 335
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.14107883817427386,
|
489 |
+
"grad_norm": 2.8708373658306523,
|
490 |
+
"learning_rate": 9.9486847117678e-06,
|
491 |
+
"loss": 0.7594,
|
492 |
+
"step": 340
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.14315352697095435,
|
496 |
+
"grad_norm": 3.0046135483136696,
|
497 |
+
"learning_rate": 9.943380507728647e-06,
|
498 |
+
"loss": 0.7618,
|
499 |
+
"step": 345
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.14522821576763487,
|
503 |
+
"grad_norm": 2.8484743596508486,
|
504 |
+
"learning_rate": 9.937817039345924e-06,
|
505 |
+
"loss": 0.7604,
|
506 |
+
"step": 350
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.14730290456431536,
|
510 |
+
"grad_norm": 2.72605173131386,
|
511 |
+
"learning_rate": 9.931994598405576e-06,
|
512 |
+
"loss": 0.7564,
|
513 |
+
"step": 355
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.14937759336099585,
|
517 |
+
"grad_norm": 2.909791550347346,
|
518 |
+
"learning_rate": 9.925913490275834e-06,
|
519 |
+
"loss": 0.7512,
|
520 |
+
"step": 360
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.15145228215767634,
|
524 |
+
"grad_norm": 2.819547837925891,
|
525 |
+
"learning_rate": 9.919574033891175e-06,
|
526 |
+
"loss": 0.7457,
|
527 |
+
"step": 365
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.15352697095435686,
|
531 |
+
"grad_norm": 2.7924198276169854,
|
532 |
+
"learning_rate": 9.912976561735617e-06,
|
533 |
+
"loss": 0.7414,
|
534 |
+
"step": 370
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.15560165975103735,
|
538 |
+
"grad_norm": 2.832091750839748,
|
539 |
+
"learning_rate": 9.906121419825269e-06,
|
540 |
+
"loss": 0.7363,
|
541 |
+
"step": 375
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.15767634854771784,
|
545 |
+
"grad_norm": 2.829243284643042,
|
546 |
+
"learning_rate": 9.899008967690185e-06,
|
547 |
+
"loss": 0.7326,
|
548 |
+
"step": 380
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.15975103734439833,
|
552 |
+
"grad_norm": 2.826535782313376,
|
553 |
+
"learning_rate": 9.891639578355511e-06,
|
554 |
+
"loss": 0.7538,
|
555 |
+
"step": 385
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.16182572614107885,
|
559 |
+
"grad_norm": 2.882936305924773,
|
560 |
+
"learning_rate": 9.88401363832192e-06,
|
561 |
+
"loss": 0.7302,
|
562 |
+
"step": 390
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.16390041493775934,
|
566 |
+
"grad_norm": 2.906899687666024,
|
567 |
+
"learning_rate": 9.87613154754534e-06,
|
568 |
+
"loss": 0.7129,
|
569 |
+
"step": 395
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.16597510373443983,
|
573 |
+
"grad_norm": 2.7967270291757944,
|
574 |
+
"learning_rate": 9.867993719415974e-06,
|
575 |
+
"loss": 0.7416,
|
576 |
+
"step": 400
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.16804979253112035,
|
580 |
+
"grad_norm": 2.8084997358792023,
|
581 |
+
"learning_rate": 9.859600580736632e-06,
|
582 |
+
"loss": 0.7353,
|
583 |
+
"step": 405
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.17012448132780084,
|
587 |
+
"grad_norm": 2.680436779833713,
|
588 |
+
"learning_rate": 9.850952571700332e-06,
|
589 |
+
"loss": 0.7376,
|
590 |
+
"step": 410
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.17219917012448133,
|
594 |
+
"grad_norm": 2.8865976677763907,
|
595 |
+
"learning_rate": 9.842050145867219e-06,
|
596 |
+
"loss": 0.7262,
|
597 |
+
"step": 415
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.17427385892116182,
|
601 |
+
"grad_norm": 2.8914725126052416,
|
602 |
+
"learning_rate": 9.832893770140778e-06,
|
603 |
+
"loss": 0.7161,
|
604 |
+
"step": 420
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.17634854771784234,
|
608 |
+
"grad_norm": 2.7982408695630725,
|
609 |
+
"learning_rate": 9.823483924743348e-06,
|
610 |
+
"loss": 0.7283,
|
611 |
+
"step": 425
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.17842323651452283,
|
615 |
+
"grad_norm": 2.6688029764570604,
|
616 |
+
"learning_rate": 9.813821103190932e-06,
|
617 |
+
"loss": 0.7192,
|
618 |
+
"step": 430
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.18049792531120332,
|
622 |
+
"grad_norm": 2.8287346081752327,
|
623 |
+
"learning_rate": 9.803905812267317e-06,
|
624 |
+
"loss": 0.724,
|
625 |
+
"step": 435
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.1825726141078838,
|
629 |
+
"grad_norm": 2.744322820321148,
|
630 |
+
"learning_rate": 9.793738571997488e-06,
|
631 |
+
"loss": 0.7204,
|
632 |
+
"step": 440
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.18464730290456433,
|
636 |
+
"grad_norm": 2.736399026435038,
|
637 |
+
"learning_rate": 9.783319915620365e-06,
|
638 |
+
"loss": 0.7265,
|
639 |
+
"step": 445
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.18672199170124482,
|
643 |
+
"grad_norm": 2.8039795008875887,
|
644 |
+
"learning_rate": 9.772650389560829e-06,
|
645 |
+
"loss": 0.7217,
|
646 |
+
"step": 450
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.1887966804979253,
|
650 |
+
"grad_norm": 2.6831409383313893,
|
651 |
+
"learning_rate": 9.761730553401067e-06,
|
652 |
+
"loss": 0.7016,
|
653 |
+
"step": 455
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 0.1908713692946058,
|
657 |
+
"grad_norm": 2.7717760950966492,
|
658 |
+
"learning_rate": 9.750560979851222e-06,
|
659 |
+
"loss": 0.728,
|
660 |
+
"step": 460
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 0.19294605809128632,
|
664 |
+
"grad_norm": 2.614242945562297,
|
665 |
+
"learning_rate": 9.739142254719351e-06,
|
666 |
+
"loss": 0.7025,
|
667 |
+
"step": 465
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 0.1950207468879668,
|
671 |
+
"grad_norm": 2.649844888470243,
|
672 |
+
"learning_rate": 9.727474976880718e-06,
|
673 |
+
"loss": 0.7069,
|
674 |
+
"step": 470
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 0.1970954356846473,
|
678 |
+
"grad_norm": 2.800760446265028,
|
679 |
+
"learning_rate": 9.715559758246363e-06,
|
680 |
+
"loss": 0.6901,
|
681 |
+
"step": 475
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.1991701244813278,
|
685 |
+
"grad_norm": 2.7922242993251265,
|
686 |
+
"learning_rate": 9.703397223731028e-06,
|
687 |
+
"loss": 0.6913,
|
688 |
+
"step": 480
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 0.2012448132780083,
|
692 |
+
"grad_norm": 2.907407187432472,
|
693 |
+
"learning_rate": 9.690988011220367e-06,
|
694 |
+
"loss": 0.7121,
|
695 |
+
"step": 485
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 0.2033195020746888,
|
699 |
+
"grad_norm": 2.626218495773325,
|
700 |
+
"learning_rate": 9.678332771537506e-06,
|
701 |
+
"loss": 0.7026,
|
702 |
+
"step": 490
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 0.2053941908713693,
|
706 |
+
"grad_norm": 2.8404176936123338,
|
707 |
+
"learning_rate": 9.665432168408895e-06,
|
708 |
+
"loss": 0.7113,
|
709 |
+
"step": 495
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 0.2074688796680498,
|
713 |
+
"grad_norm": 2.786382253827723,
|
714 |
+
"learning_rate": 9.652286878429508e-06,
|
715 |
+
"loss": 0.6852,
|
716 |
+
"step": 500
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 0.2095435684647303,
|
720 |
+
"grad_norm": 2.7800940175136515,
|
721 |
+
"learning_rate": 9.638897591027355e-06,
|
722 |
+
"loss": 0.6917,
|
723 |
+
"step": 505
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 0.21161825726141079,
|
727 |
+
"grad_norm": 2.710182741542463,
|
728 |
+
"learning_rate": 9.625265008427317e-06,
|
729 |
+
"loss": 0.6903,
|
730 |
+
"step": 510
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 0.21369294605809128,
|
734 |
+
"grad_norm": 3.0976476780998827,
|
735 |
+
"learning_rate": 9.61138984561433e-06,
|
736 |
+
"loss": 0.6776,
|
737 |
+
"step": 515
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 0.2157676348547718,
|
741 |
+
"grad_norm": 2.6986195603682015,
|
742 |
+
"learning_rate": 9.597272830295877e-06,
|
743 |
+
"loss": 0.6759,
|
744 |
+
"step": 520
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 0.21784232365145229,
|
748 |
+
"grad_norm": 2.8093612594263493,
|
749 |
+
"learning_rate": 9.582914702863816e-06,
|
750 |
+
"loss": 0.691,
|
751 |
+
"step": 525
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 0.21991701244813278,
|
755 |
+
"grad_norm": 2.668102059117624,
|
756 |
+
"learning_rate": 9.568316216355569e-06,
|
757 |
+
"loss": 0.6658,
|
758 |
+
"step": 530
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 0.22199170124481327,
|
762 |
+
"grad_norm": 2.968496552447308,
|
763 |
+
"learning_rate": 9.553478136414606e-06,
|
764 |
+
"loss": 0.6585,
|
765 |
+
"step": 535
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 0.22406639004149378,
|
769 |
+
"grad_norm": 2.788495237514456,
|
770 |
+
"learning_rate": 9.538401241250302e-06,
|
771 |
+
"loss": 0.6761,
|
772 |
+
"step": 540
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 0.22614107883817428,
|
776 |
+
"grad_norm": 2.692620475621199,
|
777 |
+
"learning_rate": 9.523086321597123e-06,
|
778 |
+
"loss": 0.6862,
|
779 |
+
"step": 545
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 0.22821576763485477,
|
783 |
+
"grad_norm": 2.8519734853628274,
|
784 |
+
"learning_rate": 9.507534180673142e-06,
|
785 |
+
"loss": 0.6765,
|
786 |
+
"step": 550
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 0.23029045643153526,
|
790 |
+
"grad_norm": 2.6424845030364934,
|
791 |
+
"learning_rate": 9.49174563413793e-06,
|
792 |
+
"loss": 0.693,
|
793 |
+
"step": 555
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 0.23236514522821577,
|
797 |
+
"grad_norm": 2.68422130623634,
|
798 |
+
"learning_rate": 9.475721510049765e-06,
|
799 |
+
"loss": 0.6549,
|
800 |
+
"step": 560
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 0.23443983402489627,
|
804 |
+
"grad_norm": 2.745364732974406,
|
805 |
+
"learning_rate": 9.459462648822209e-06,
|
806 |
+
"loss": 0.6762,
|
807 |
+
"step": 565
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 0.23651452282157676,
|
811 |
+
"grad_norm": 2.7138771122564496,
|
812 |
+
"learning_rate": 9.442969903180021e-06,
|
813 |
+
"loss": 0.6617,
|
814 |
+
"step": 570
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 0.23858921161825727,
|
818 |
+
"grad_norm": 2.594975020618961,
|
819 |
+
"learning_rate": 9.426244138114456e-06,
|
820 |
+
"loss": 0.6644,
|
821 |
+
"step": 575
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 0.24066390041493776,
|
825 |
+
"grad_norm": 2.6711485289021217,
|
826 |
+
"learning_rate": 9.409286230837876e-06,
|
827 |
+
"loss": 0.6515,
|
828 |
+
"step": 580
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 0.24273858921161826,
|
832 |
+
"grad_norm": 3.157835949894035,
|
833 |
+
"learning_rate": 9.39209707073775e-06,
|
834 |
+
"loss": 0.6552,
|
835 |
+
"step": 585
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 0.24481327800829875,
|
839 |
+
"grad_norm": 2.7728392273294427,
|
840 |
+
"learning_rate": 9.37467755933002e-06,
|
841 |
+
"loss": 0.6547,
|
842 |
+
"step": 590
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 0.24688796680497926,
|
846 |
+
"grad_norm": 2.6326098618284726,
|
847 |
+
"learning_rate": 9.357028610211802e-06,
|
848 |
+
"loss": 0.6581,
|
849 |
+
"step": 595
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 0.24896265560165975,
|
853 |
+
"grad_norm": 2.645526952797483,
|
854 |
+
"learning_rate": 9.339151149013483e-06,
|
855 |
+
"loss": 0.66,
|
856 |
+
"step": 600
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 0.25103734439834025,
|
860 |
+
"grad_norm": 2.695303011335832,
|
861 |
+
"learning_rate": 9.32104611335017e-06,
|
862 |
+
"loss": 0.6494,
|
863 |
+
"step": 605
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 0.25311203319502074,
|
867 |
+
"grad_norm": 2.595062725441063,
|
868 |
+
"learning_rate": 9.302714452772515e-06,
|
869 |
+
"loss": 0.6443,
|
870 |
+
"step": 610
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 0.2551867219917012,
|
874 |
+
"grad_norm": 2.6458460885695168,
|
875 |
+
"learning_rate": 9.284157128716916e-06,
|
876 |
+
"loss": 0.6514,
|
877 |
+
"step": 615
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 0.2572614107883817,
|
881 |
+
"grad_norm": 2.5786064674401583,
|
882 |
+
"learning_rate": 9.265375114455091e-06,
|
883 |
+
"loss": 0.6405,
|
884 |
+
"step": 620
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 0.25933609958506226,
|
888 |
+
"grad_norm": 2.6907850585932027,
|
889 |
+
"learning_rate": 9.246369395043033e-06,
|
890 |
+
"loss": 0.6476,
|
891 |
+
"step": 625
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 0.26141078838174275,
|
895 |
+
"grad_norm": 2.6953268245247557,
|
896 |
+
"learning_rate": 9.227140967269348e-06,
|
897 |
+
"loss": 0.6414,
|
898 |
+
"step": 630
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 0.26348547717842324,
|
902 |
+
"grad_norm": 2.6657263650658427,
|
903 |
+
"learning_rate": 9.20769083960298e-06,
|
904 |
+
"loss": 0.6514,
|
905 |
+
"step": 635
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 0.26556016597510373,
|
909 |
+
"grad_norm": 2.5956487125743615,
|
910 |
+
"learning_rate": 9.188020032140308e-06,
|
911 |
+
"loss": 0.6494,
|
912 |
+
"step": 640
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 0.2676348547717842,
|
916 |
+
"grad_norm": 2.6461882359542095,
|
917 |
+
"learning_rate": 9.168129576551665e-06,
|
918 |
+
"loss": 0.6388,
|
919 |
+
"step": 645
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 0.2697095435684647,
|
923 |
+
"grad_norm": 2.734469941118851,
|
924 |
+
"learning_rate": 9.148020516027207e-06,
|
925 |
+
"loss": 0.639,
|
926 |
+
"step": 650
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 0.2717842323651452,
|
930 |
+
"grad_norm": 2.85311061995995,
|
931 |
+
"learning_rate": 9.127693905222223e-06,
|
932 |
+
"loss": 0.6239,
|
933 |
+
"step": 655
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 0.27385892116182575,
|
937 |
+
"grad_norm": 2.905457532313008,
|
938 |
+
"learning_rate": 9.107150810201805e-06,
|
939 |
+
"loss": 0.6332,
|
940 |
+
"step": 660
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 0.27593360995850624,
|
944 |
+
"grad_norm": 2.866240537811393,
|
945 |
+
"learning_rate": 9.086392308384946e-06,
|
946 |
+
"loss": 0.6402,
|
947 |
+
"step": 665
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 0.27800829875518673,
|
951 |
+
"grad_norm": 2.6848783944743717,
|
952 |
+
"learning_rate": 9.065419488488029e-06,
|
953 |
+
"loss": 0.6198,
|
954 |
+
"step": 670
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 0.2800829875518672,
|
958 |
+
"grad_norm": 2.806882576661356,
|
959 |
+
"learning_rate": 9.044233450467728e-06,
|
960 |
+
"loss": 0.6329,
|
961 |
+
"step": 675
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 0.2821576763485477,
|
965 |
+
"grad_norm": 2.6697354108572457,
|
966 |
+
"learning_rate": 9.022835305463322e-06,
|
967 |
+
"loss": 0.628,
|
968 |
+
"step": 680
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 0.2842323651452282,
|
972 |
+
"grad_norm": 2.692739103753555,
|
973 |
+
"learning_rate": 9.001226175738409e-06,
|
974 |
+
"loss": 0.6285,
|
975 |
+
"step": 685
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 0.2863070539419087,
|
979 |
+
"grad_norm": 2.6356169772477887,
|
980 |
+
"learning_rate": 8.979407194622062e-06,
|
981 |
+
"loss": 0.6029,
|
982 |
+
"step": 690
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 0.2883817427385892,
|
986 |
+
"grad_norm": 2.684304580451599,
|
987 |
+
"learning_rate": 8.95737950644938e-06,
|
988 |
+
"loss": 0.6163,
|
989 |
+
"step": 695
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 0.29045643153526973,
|
993 |
+
"grad_norm": 2.665591903369743,
|
994 |
+
"learning_rate": 8.93514426650147e-06,
|
995 |
+
"loss": 0.6036,
|
996 |
+
"step": 700
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 0.2925311203319502,
|
1000 |
+
"grad_norm": 2.5849646219326394,
|
1001 |
+
"learning_rate": 8.912702640944862e-06,
|
1002 |
+
"loss": 0.63,
|
1003 |
+
"step": 705
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 0.2946058091286307,
|
1007 |
+
"grad_norm": 2.642637231105898,
|
1008 |
+
"learning_rate": 8.89005580677034e-06,
|
1009 |
+
"loss": 0.6142,
|
1010 |
+
"step": 710
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 0.2966804979253112,
|
1014 |
+
"grad_norm": 2.8131632244224543,
|
1015 |
+
"learning_rate": 8.867204951731227e-06,
|
1016 |
+
"loss": 0.6177,
|
1017 |
+
"step": 715
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 0.2987551867219917,
|
1021 |
+
"grad_norm": 2.7483245438817794,
|
1022 |
+
"learning_rate": 8.844151274281074e-06,
|
1023 |
+
"loss": 0.6194,
|
1024 |
+
"step": 720
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 0.3008298755186722,
|
1028 |
+
"grad_norm": 3.0664428917760675,
|
1029 |
+
"learning_rate": 8.820895983510813e-06,
|
1030 |
+
"loss": 0.6186,
|
1031 |
+
"step": 725
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 0.3029045643153527,
|
1035 |
+
"grad_norm": 2.688656442588083,
|
1036 |
+
"learning_rate": 8.797440299085344e-06,
|
1037 |
+
"loss": 0.5988,
|
1038 |
+
"step": 730
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 0.3049792531120332,
|
1042 |
+
"grad_norm": 2.7050533397545258,
|
1043 |
+
"learning_rate": 8.77378545117957e-06,
|
1044 |
+
"loss": 0.616,
|
1045 |
+
"step": 735
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 0.3070539419087137,
|
1049 |
+
"grad_norm": 2.693473576801362,
|
1050 |
+
"learning_rate": 8.74993268041387e-06,
|
1051 |
+
"loss": 0.5999,
|
1052 |
+
"step": 740
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 0.3091286307053942,
|
1056 |
+
"grad_norm": 2.6799339600570766,
|
1057 |
+
"learning_rate": 8.725883237789046e-06,
|
1058 |
+
"loss": 0.5903,
|
1059 |
+
"step": 745
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 0.3112033195020747,
|
1063 |
+
"grad_norm": 2.5430481996060528,
|
1064 |
+
"learning_rate": 8.701638384620694e-06,
|
1065 |
+
"loss": 0.5856,
|
1066 |
+
"step": 750
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 0.3132780082987552,
|
1070 |
+
"grad_norm": 2.672587466894312,
|
1071 |
+
"learning_rate": 8.677199392473068e-06,
|
1072 |
+
"loss": 0.6042,
|
1073 |
+
"step": 755
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 0.3153526970954357,
|
1077 |
+
"grad_norm": 2.7790737972229484,
|
1078 |
+
"learning_rate": 8.652567543092385e-06,
|
1079 |
+
"loss": 0.5917,
|
1080 |
+
"step": 760
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 0.31742738589211617,
|
1084 |
+
"grad_norm": 2.7954825670577677,
|
1085 |
+
"learning_rate": 8.627744128339599e-06,
|
1086 |
+
"loss": 0.5939,
|
1087 |
+
"step": 765
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 0.31950207468879666,
|
1091 |
+
"grad_norm": 2.662658318367537,
|
1092 |
+
"learning_rate": 8.602730450122648e-06,
|
1093 |
+
"loss": 0.6011,
|
1094 |
+
"step": 770
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 0.3215767634854772,
|
1098 |
+
"grad_norm": 2.6270972350688884,
|
1099 |
+
"learning_rate": 8.577527820328176e-06,
|
1100 |
+
"loss": 0.5859,
|
1101 |
+
"step": 775
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 0.3236514522821577,
|
1105 |
+
"grad_norm": 2.5997336791640144,
|
1106 |
+
"learning_rate": 8.552137560752728e-06,
|
1107 |
+
"loss": 0.5695,
|
1108 |
+
"step": 780
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 0.3257261410788382,
|
1112 |
+
"grad_norm": 2.6129156491647096,
|
1113 |
+
"learning_rate": 8.526561003033424e-06,
|
1114 |
+
"loss": 0.5991,
|
1115 |
+
"step": 785
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 0.3278008298755187,
|
1119 |
+
"grad_norm": 2.641434925698528,
|
1120 |
+
"learning_rate": 8.50079948857812e-06,
|
1121 |
+
"loss": 0.5756,
|
1122 |
+
"step": 790
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 0.32987551867219916,
|
1126 |
+
"grad_norm": 2.744754568258112,
|
1127 |
+
"learning_rate": 8.474854368495055e-06,
|
1128 |
+
"loss": 0.5916,
|
1129 |
+
"step": 795
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 0.33195020746887965,
|
1133 |
+
"grad_norm": 2.8742812662542767,
|
1134 |
+
"learning_rate": 8.44872700352199e-06,
|
1135 |
+
"loss": 0.5858,
|
1136 |
+
"step": 800
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 0.33402489626556015,
|
1140 |
+
"grad_norm": 2.675688772837824,
|
1141 |
+
"learning_rate": 8.422418763954841e-06,
|
1142 |
+
"loss": 0.5858,
|
1143 |
+
"step": 805
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 0.3360995850622407,
|
1147 |
+
"grad_norm": 2.6454009935375202,
|
1148 |
+
"learning_rate": 8.395931029575817e-06,
|
1149 |
+
"loss": 0.5793,
|
1150 |
+
"step": 810
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 0.3381742738589212,
|
1154 |
+
"grad_norm": 2.5799174291219855,
|
1155 |
+
"learning_rate": 8.369265189581048e-06,
|
1156 |
+
"loss": 0.5577,
|
1157 |
+
"step": 815
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 0.34024896265560167,
|
1161 |
+
"grad_norm": 2.6266021968279025,
|
1162 |
+
"learning_rate": 8.342422642507727e-06,
|
1163 |
+
"loss": 0.5908,
|
1164 |
+
"step": 820
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 0.34232365145228216,
|
1168 |
+
"grad_norm": 2.6270567170702104,
|
1169 |
+
"learning_rate": 8.31540479616076e-06,
|
1170 |
+
"loss": 0.5583,
|
1171 |
+
"step": 825
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 0.34439834024896265,
|
1175 |
+
"grad_norm": 2.6605567528471283,
|
1176 |
+
"learning_rate": 8.288213067538936e-06,
|
1177 |
+
"loss": 0.5792,
|
1178 |
+
"step": 830
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 0.34647302904564314,
|
1182 |
+
"grad_norm": 2.6648839547096936,
|
1183 |
+
"learning_rate": 8.260848882760616e-06,
|
1184 |
+
"loss": 0.5758,
|
1185 |
+
"step": 835
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 0.34854771784232363,
|
1189 |
+
"grad_norm": 2.4985116541737487,
|
1190 |
+
"learning_rate": 8.233313676988917e-06,
|
1191 |
+
"loss": 0.5589,
|
1192 |
+
"step": 840
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 0.3506224066390041,
|
1196 |
+
"grad_norm": 2.5295015470513014,
|
1197 |
+
"learning_rate": 8.205608894356461e-06,
|
1198 |
+
"loss": 0.5747,
|
1199 |
+
"step": 845
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 0.35269709543568467,
|
1203 |
+
"grad_norm": 2.6313000253018317,
|
1204 |
+
"learning_rate": 8.177735987889628e-06,
|
1205 |
+
"loss": 0.5794,
|
1206 |
+
"step": 850
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 0.35477178423236516,
|
1210 |
+
"grad_norm": 2.841463850249757,
|
1211 |
+
"learning_rate": 8.149696419432352e-06,
|
1212 |
+
"loss": 0.5662,
|
1213 |
+
"step": 855
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 0.35684647302904565,
|
1217 |
+
"grad_norm": 2.499217630200596,
|
1218 |
+
"learning_rate": 8.121491659569442e-06,
|
1219 |
+
"loss": 0.5489,
|
1220 |
+
"step": 860
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 0.35892116182572614,
|
1224 |
+
"grad_norm": 2.624500775254657,
|
1225 |
+
"learning_rate": 8.093123187549475e-06,
|
1226 |
+
"loss": 0.5529,
|
1227 |
+
"step": 865
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 0.36099585062240663,
|
1231 |
+
"grad_norm": 2.891141879312641,
|
1232 |
+
"learning_rate": 8.064592491207193e-06,
|
1233 |
+
"loss": 0.5615,
|
1234 |
+
"step": 870
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 0.3630705394190871,
|
1238 |
+
"grad_norm": 2.985654722661823,
|
1239 |
+
"learning_rate": 8.035901066885486e-06,
|
1240 |
+
"loss": 0.5727,
|
1241 |
+
"step": 875
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 0.3651452282157676,
|
1245 |
+
"grad_norm": 2.7907723136875995,
|
1246 |
+
"learning_rate": 8.007050419356898e-06,
|
1247 |
+
"loss": 0.5608,
|
1248 |
+
"step": 880
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 0.36721991701244816,
|
1252 |
+
"grad_norm": 2.6416506986010813,
|
1253 |
+
"learning_rate": 7.978042061744728e-06,
|
1254 |
+
"loss": 0.5557,
|
1255 |
+
"step": 885
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 0.36929460580912865,
|
1259 |
+
"grad_norm": 2.5258498715580067,
|
1260 |
+
"learning_rate": 7.94887751544365e-06,
|
1261 |
+
"loss": 0.5534,
|
1262 |
+
"step": 890
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 0.37136929460580914,
|
1266 |
+
"grad_norm": 2.561386861258123,
|
1267 |
+
"learning_rate": 7.919558310039937e-06,
|
1268 |
+
"loss": 0.556,
|
1269 |
+
"step": 895
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 0.37344398340248963,
|
1273 |
+
"grad_norm": 2.727627068843935,
|
1274 |
+
"learning_rate": 7.890085983231225e-06,
|
1275 |
+
"loss": 0.5462,
|
1276 |
+
"step": 900
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 0.3755186721991701,
|
1280 |
+
"grad_norm": 2.5718069262603533,
|
1281 |
+
"learning_rate": 7.860462080745884e-06,
|
1282 |
+
"loss": 0.5399,
|
1283 |
+
"step": 905
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 0.3775933609958506,
|
1287 |
+
"grad_norm": 2.5380387267872746,
|
1288 |
+
"learning_rate": 7.830688156261927e-06,
|
1289 |
+
"loss": 0.5502,
|
1290 |
+
"step": 910
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 0.3796680497925311,
|
1294 |
+
"grad_norm": 2.6214419026204565,
|
1295 |
+
"learning_rate": 7.800765771325546e-06,
|
1296 |
+
"loss": 0.5417,
|
1297 |
+
"step": 915
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 0.3817427385892116,
|
1301 |
+
"grad_norm": 2.712261230434538,
|
1302 |
+
"learning_rate": 7.7706964952692e-06,
|
1303 |
+
"loss": 0.545,
|
1304 |
+
"step": 920
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 0.38381742738589214,
|
1308 |
+
"grad_norm": 2.5420503220632797,
|
1309 |
+
"learning_rate": 7.740481905129307e-06,
|
1310 |
+
"loss": 0.5333,
|
1311 |
+
"step": 925
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 0.38589211618257263,
|
1315 |
+
"grad_norm": 2.6001135316476893,
|
1316 |
+
"learning_rate": 7.710123585563552e-06,
|
1317 |
+
"loss": 0.532,
|
1318 |
+
"step": 930
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 0.3879668049792531,
|
1322 |
+
"grad_norm": 2.720274250658824,
|
1323 |
+
"learning_rate": 7.679623128767754e-06,
|
1324 |
+
"loss": 0.5466,
|
1325 |
+
"step": 935
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 0.3900414937759336,
|
1329 |
+
"grad_norm": 2.751465451077315,
|
1330 |
+
"learning_rate": 7.648982134392378e-06,
|
1331 |
+
"loss": 0.5411,
|
1332 |
+
"step": 940
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 0.3921161825726141,
|
1336 |
+
"grad_norm": 2.5435730720837864,
|
1337 |
+
"learning_rate": 7.618202209458623e-06,
|
1338 |
+
"loss": 0.5225,
|
1339 |
+
"step": 945
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 0.3941908713692946,
|
1343 |
+
"grad_norm": 2.5155581478521456,
|
1344 |
+
"learning_rate": 7.587284968274155e-06,
|
1345 |
+
"loss": 0.5361,
|
1346 |
+
"step": 950
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 0.3962655601659751,
|
1350 |
+
"grad_norm": 2.6500904333286384,
|
1351 |
+
"learning_rate": 7.556232032348429e-06,
|
1352 |
+
"loss": 0.5323,
|
1353 |
+
"step": 955
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 0.3983402489626556,
|
1357 |
+
"grad_norm": 2.6832927074173947,
|
1358 |
+
"learning_rate": 7.5250450303076526e-06,
|
1359 |
+
"loss": 0.5324,
|
1360 |
+
"step": 960
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 0.4004149377593361,
|
1364 |
+
"grad_norm": 2.5794687904843934,
|
1365 |
+
"learning_rate": 7.49372559780937e-06,
|
1366 |
+
"loss": 0.5401,
|
1367 |
+
"step": 965
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 0.4024896265560166,
|
1371 |
+
"grad_norm": 2.779973776708556,
|
1372 |
+
"learning_rate": 7.462275377456671e-06,
|
1373 |
+
"loss": 0.5353,
|
1374 |
+
"step": 970
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 0.4045643153526971,
|
1378 |
+
"grad_norm": 2.510096649942157,
|
1379 |
+
"learning_rate": 7.430696018712049e-06,
|
1380 |
+
"loss": 0.5178,
|
1381 |
+
"step": 975
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 0.4066390041493776,
|
1385 |
+
"grad_norm": 2.544609714596366,
|
1386 |
+
"learning_rate": 7.398989177810889e-06,
|
1387 |
+
"loss": 0.5305,
|
1388 |
+
"step": 980
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 0.4087136929460581,
|
1392 |
+
"grad_norm": 2.6358450407419394,
|
1393 |
+
"learning_rate": 7.3671565176746025e-06,
|
1394 |
+
"loss": 0.5083,
|
1395 |
+
"step": 985
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 0.4107883817427386,
|
1399 |
+
"grad_norm": 2.5104688131658115,
|
1400 |
+
"learning_rate": 7.335199707823415e-06,
|
1401 |
+
"loss": 0.5213,
|
1402 |
+
"step": 990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 0.41286307053941906,
|
1406 |
+
"grad_norm": 2.534767895856582,
|
1407 |
+
"learning_rate": 7.3031204242888e-06,
|
1408 |
+
"loss": 0.5003,
|
1409 |
+
"step": 995
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 0.4149377593360996,
|
1413 |
+
"grad_norm": 2.734749407560919,
|
1414 |
+
"learning_rate": 7.270920349525584e-06,
|
1415 |
+
"loss": 0.5179,
|
1416 |
+
"step": 1000
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 0.4170124481327801,
|
1420 |
+
"grad_norm": 2.404837852320902,
|
1421 |
+
"learning_rate": 7.238601172323701e-06,
|
1422 |
+
"loss": 0.5231,
|
1423 |
+
"step": 1005
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 0.4190871369294606,
|
1427 |
+
"grad_norm": 2.655040755780098,
|
1428 |
+
"learning_rate": 7.206164587719627e-06,
|
1429 |
+
"loss": 0.5191,
|
1430 |
+
"step": 1010
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 0.4211618257261411,
|
1434 |
+
"grad_norm": 2.537938752408207,
|
1435 |
+
"learning_rate": 7.173612296907473e-06,
|
1436 |
+
"loss": 0.5206,
|
1437 |
+
"step": 1015
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 0.42323651452282157,
|
1441 |
+
"grad_norm": 2.6276182145014135,
|
1442 |
+
"learning_rate": 7.1409460071497675e-06,
|
1443 |
+
"loss": 0.5294,
|
1444 |
+
"step": 1020
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 0.42531120331950206,
|
1448 |
+
"grad_norm": 2.543796942084407,
|
1449 |
+
"learning_rate": 7.108167431687917e-06,
|
1450 |
+
"loss": 0.5166,
|
1451 |
+
"step": 1025
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 0.42738589211618255,
|
1455 |
+
"grad_norm": 2.6672869217839468,
|
1456 |
+
"learning_rate": 7.075278289652349e-06,
|
1457 |
+
"loss": 0.5225,
|
1458 |
+
"step": 1030
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 0.42946058091286304,
|
1462 |
+
"grad_norm": 2.531083972850949,
|
1463 |
+
"learning_rate": 7.042280305972354e-06,
|
1464 |
+
"loss": 0.5142,
|
1465 |
+
"step": 1035
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 0.4315352697095436,
|
1469 |
+
"grad_norm": 2.4747366529454693,
|
1470 |
+
"learning_rate": 7.009175211285611e-06,
|
1471 |
+
"loss": 0.5141,
|
1472 |
+
"step": 1040
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 0.4336099585062241,
|
1476 |
+
"grad_norm": 2.493745172465378,
|
1477 |
+
"learning_rate": 6.975964741847427e-06,
|
1478 |
+
"loss": 0.5241,
|
1479 |
+
"step": 1045
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 0.43568464730290457,
|
1483 |
+
"grad_norm": 2.59288402554468,
|
1484 |
+
"learning_rate": 6.942650639439678e-06,
|
1485 |
+
"loss": 0.5127,
|
1486 |
+
"step": 1050
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 0.43775933609958506,
|
1490 |
+
"grad_norm": 2.5743234047985952,
|
1491 |
+
"learning_rate": 6.9092346512794475e-06,
|
1492 |
+
"loss": 0.5112,
|
1493 |
+
"step": 1055
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 0.43983402489626555,
|
1497 |
+
"grad_norm": 2.46036118970127,
|
1498 |
+
"learning_rate": 6.875718529927404e-06,
|
1499 |
+
"loss": 0.4881,
|
1500 |
+
"step": 1060
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 0.44190871369294604,
|
1504 |
+
"grad_norm": 2.582868746430813,
|
1505 |
+
"learning_rate": 6.8421040331958745e-06,
|
1506 |
+
"loss": 0.5072,
|
1507 |
+
"step": 1065
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 0.44398340248962653,
|
1511 |
+
"grad_norm": 2.5161140611104056,
|
1512 |
+
"learning_rate": 6.808392924056659e-06,
|
1513 |
+
"loss": 0.5068,
|
1514 |
+
"step": 1070
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 0.4460580912863071,
|
1518 |
+
"grad_norm": 2.616608707294859,
|
1519 |
+
"learning_rate": 6.774586970548567e-06,
|
1520 |
+
"loss": 0.5132,
|
1521 |
+
"step": 1075
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 0.44813278008298757,
|
1525 |
+
"grad_norm": 2.4704454065037424,
|
1526 |
+
"learning_rate": 6.7406879456846875e-06,
|
1527 |
+
"loss": 0.4942,
|
1528 |
+
"step": 1080
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 0.45020746887966806,
|
1532 |
+
"grad_norm": 2.5654508809058805,
|
1533 |
+
"learning_rate": 6.7066976273594e-06,
|
1534 |
+
"loss": 0.4914,
|
1535 |
+
"step": 1085
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 0.45228215767634855,
|
1539 |
+
"grad_norm": 2.5642213246435017,
|
1540 |
+
"learning_rate": 6.672617798255135e-06,
|
1541 |
+
"loss": 0.5043,
|
1542 |
+
"step": 1090
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 0.45435684647302904,
|
1546 |
+
"grad_norm": 2.3759068788938538,
|
1547 |
+
"learning_rate": 6.63845024574887e-06,
|
1548 |
+
"loss": 0.4744,
|
1549 |
+
"step": 1095
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 0.45643153526970953,
|
1553 |
+
"grad_norm": 2.4485238164491,
|
1554 |
+
"learning_rate": 6.604196761818395e-06,
|
1555 |
+
"loss": 0.4992,
|
1556 |
+
"step": 1100
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 0.45850622406639,
|
1560 |
+
"grad_norm": 2.481384678496845,
|
1561 |
+
"learning_rate": 6.5698591429483286e-06,
|
1562 |
+
"loss": 0.5016,
|
1563 |
+
"step": 1105
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 0.4605809128630705,
|
1567 |
+
"grad_norm": 2.571902891821277,
|
1568 |
+
"learning_rate": 6.535439190035884e-06,
|
1569 |
+
"loss": 0.4893,
|
1570 |
+
"step": 1110
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 0.46265560165975106,
|
1574 |
+
"grad_norm": 2.559158463935003,
|
1575 |
+
"learning_rate": 6.5009387082964405e-06,
|
1576 |
+
"loss": 0.4991,
|
1577 |
+
"step": 1115
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 0.46473029045643155,
|
1581 |
+
"grad_norm": 2.538459567637443,
|
1582 |
+
"learning_rate": 6.466359507168849e-06,
|
1583 |
+
"loss": 0.4953,
|
1584 |
+
"step": 1120
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 0.46680497925311204,
|
1588 |
+
"grad_norm": 2.5267222175872623,
|
1589 |
+
"learning_rate": 6.431703400220541e-06,
|
1590 |
+
"loss": 0.482,
|
1591 |
+
"step": 1125
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 0.46887966804979253,
|
1595 |
+
"grad_norm": 2.38064786704353,
|
1596 |
+
"learning_rate": 6.396972205052407e-06,
|
1597 |
+
"loss": 0.4839,
|
1598 |
+
"step": 1130
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 0.470954356846473,
|
1602 |
+
"grad_norm": 2.5624442145243935,
|
1603 |
+
"learning_rate": 6.362167743203474e-06,
|
1604 |
+
"loss": 0.4975,
|
1605 |
+
"step": 1135
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 0.4730290456431535,
|
1609 |
+
"grad_norm": 2.4427415841202014,
|
1610 |
+
"learning_rate": 6.327291840055365e-06,
|
1611 |
+
"loss": 0.4759,
|
1612 |
+
"step": 1140
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 0.475103734439834,
|
1616 |
+
"grad_norm": 2.5750114799610095,
|
1617 |
+
"learning_rate": 6.292346324736578e-06,
|
1618 |
+
"loss": 0.4931,
|
1619 |
+
"step": 1145
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 0.47717842323651455,
|
1623 |
+
"grad_norm": 2.368089990602912,
|
1624 |
+
"learning_rate": 6.2573330300265375e-06,
|
1625 |
+
"loss": 0.4943,
|
1626 |
+
"step": 1150
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 0.47925311203319504,
|
1630 |
+
"grad_norm": 2.430038445579993,
|
1631 |
+
"learning_rate": 6.22225379225948e-06,
|
1632 |
+
"loss": 0.4776,
|
1633 |
+
"step": 1155
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 0.48132780082987553,
|
1637 |
+
"grad_norm": 2.655610966122537,
|
1638 |
+
"learning_rate": 6.18711045122814e-06,
|
1639 |
+
"loss": 0.4931,
|
1640 |
+
"step": 1160
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 0.483402489626556,
|
1644 |
+
"grad_norm": 2.561229036359029,
|
1645 |
+
"learning_rate": 6.151904850087265e-06,
|
1646 |
+
"loss": 0.4853,
|
1647 |
+
"step": 1165
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 0.4854771784232365,
|
1651 |
+
"grad_norm": 2.5716333011695487,
|
1652 |
+
"learning_rate": 6.116638835256943e-06,
|
1653 |
+
"loss": 0.4662,
|
1654 |
+
"step": 1170
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 0.487551867219917,
|
1658 |
+
"grad_norm": 2.6260675883477123,
|
1659 |
+
"learning_rate": 6.081314256325762e-06,
|
1660 |
+
"loss": 0.4771,
|
1661 |
+
"step": 1175
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 0.4896265560165975,
|
1665 |
+
"grad_norm": 2.464680850201926,
|
1666 |
+
"learning_rate": 6.045932965953813e-06,
|
1667 |
+
"loss": 0.4819,
|
1668 |
+
"step": 1180
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 0.491701244813278,
|
1672 |
+
"grad_norm": 2.4612043275661852,
|
1673 |
+
"learning_rate": 6.010496819775518e-06,
|
1674 |
+
"loss": 0.4744,
|
1675 |
+
"step": 1185
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 0.49377593360995853,
|
1679 |
+
"grad_norm": 2.4847373767230994,
|
1680 |
+
"learning_rate": 5.9750076763023025e-06,
|
1681 |
+
"loss": 0.4895,
|
1682 |
+
"step": 1190
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 0.495850622406639,
|
1686 |
+
"grad_norm": 2.324351303455581,
|
1687 |
+
"learning_rate": 5.939467396825137e-06,
|
1688 |
+
"loss": 0.4771,
|
1689 |
+
"step": 1195
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 0.4979253112033195,
|
1693 |
+
"grad_norm": 2.4411119886082546,
|
1694 |
+
"learning_rate": 5.903877845316906e-06,
|
1695 |
+
"loss": 0.4698,
|
1696 |
+
"step": 1200
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 0.5,
|
1700 |
+
"grad_norm": 2.5240319324749887,
|
1701 |
+
"learning_rate": 5.8682408883346535e-06,
|
1702 |
+
"loss": 0.4705,
|
1703 |
+
"step": 1205
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 0.5020746887966805,
|
1707 |
+
"grad_norm": 2.6132898743051123,
|
1708 |
+
"learning_rate": 5.832558394921688e-06,
|
1709 |
+
"loss": 0.4557,
|
1710 |
+
"step": 1210
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 0.504149377593361,
|
1714 |
+
"grad_norm": 2.484591361229376,
|
1715 |
+
"learning_rate": 5.796832236509556e-06,
|
1716 |
+
"loss": 0.4753,
|
1717 |
+
"step": 1215
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 0.5062240663900415,
|
1721 |
+
"grad_norm": 2.392319441698671,
|
1722 |
+
"learning_rate": 5.761064286819895e-06,
|
1723 |
+
"loss": 0.458,
|
1724 |
+
"step": 1220
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 0.508298755186722,
|
1728 |
+
"grad_norm": 2.6251874596416966,
|
1729 |
+
"learning_rate": 5.725256421766158e-06,
|
1730 |
+
"loss": 0.4809,
|
1731 |
+
"step": 1225
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 0.5103734439834025,
|
1735 |
+
"grad_norm": 2.4584477311068125,
|
1736 |
+
"learning_rate": 5.689410519355226e-06,
|
1737 |
+
"loss": 0.4709,
|
1738 |
+
"step": 1230
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 0.5124481327800829,
|
1742 |
+
"grad_norm": 2.3560580563552858,
|
1743 |
+
"learning_rate": 5.653528459588925e-06,
|
1744 |
+
"loss": 0.4659,
|
1745 |
+
"step": 1235
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 0.5145228215767634,
|
1749 |
+
"grad_norm": 2.485258794734814,
|
1750 |
+
"learning_rate": 5.617612124365411e-06,
|
1751 |
+
"loss": 0.4647,
|
1752 |
+
"step": 1240
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 0.516597510373444,
|
1756 |
+
"grad_norm": 2.4250973675636938,
|
1757 |
+
"learning_rate": 5.5816633973804766e-06,
|
1758 |
+
"loss": 0.4721,
|
1759 |
+
"step": 1245
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 0.5186721991701245,
|
1763 |
+
"grad_norm": 2.4772730090302515,
|
1764 |
+
"learning_rate": 5.545684164028764e-06,
|
1765 |
+
"loss": 0.4558,
|
1766 |
+
"step": 1250
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 0.520746887966805,
|
1770 |
+
"grad_norm": 2.6817558433388053,
|
1771 |
+
"learning_rate": 5.509676311304869e-06,
|
1772 |
+
"loss": 0.4632,
|
1773 |
+
"step": 1255
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 0.5228215767634855,
|
1777 |
+
"grad_norm": 2.482278609637873,
|
1778 |
+
"learning_rate": 5.4736417277043865e-06,
|
1779 |
+
"loss": 0.4662,
|
1780 |
+
"step": 1260
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 0.524896265560166,
|
1784 |
+
"grad_norm": 2.4538360481145873,
|
1785 |
+
"learning_rate": 5.4375823031248545e-06,
|
1786 |
+
"loss": 0.4542,
|
1787 |
+
"step": 1265
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 0.5269709543568465,
|
1791 |
+
"grad_norm": 2.3974115486101693,
|
1792 |
+
"learning_rate": 5.401499928766644e-06,
|
1793 |
+
"loss": 0.4688,
|
1794 |
+
"step": 1270
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 0.529045643153527,
|
1798 |
+
"grad_norm": 2.35868507934765,
|
1799 |
+
"learning_rate": 5.365396497033763e-06,
|
1800 |
+
"loss": 0.4528,
|
1801 |
+
"step": 1275
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 0.5311203319502075,
|
1805 |
+
"grad_norm": 2.350069783189846,
|
1806 |
+
"learning_rate": 5.32927390143462e-06,
|
1807 |
+
"loss": 0.4606,
|
1808 |
+
"step": 1280
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 0.533195020746888,
|
1812 |
+
"grad_norm": 2.48472920397625,
|
1813 |
+
"learning_rate": 5.293134036482697e-06,
|
1814 |
+
"loss": 0.443,
|
1815 |
+
"step": 1285
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 0.5352697095435685,
|
1819 |
+
"grad_norm": 2.4861989102755,
|
1820 |
+
"learning_rate": 5.256978797597202e-06,
|
1821 |
+
"loss": 0.4639,
|
1822 |
+
"step": 1290
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 0.5373443983402489,
|
1826 |
+
"grad_norm": 2.4356508616340116,
|
1827 |
+
"learning_rate": 5.220810081003656e-06,
|
1828 |
+
"loss": 0.4453,
|
1829 |
+
"step": 1295
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 0.5394190871369294,
|
1833 |
+
"grad_norm": 2.7333647570132213,
|
1834 |
+
"learning_rate": 5.184629783634441e-06,
|
1835 |
+
"loss": 0.4575,
|
1836 |
+
"step": 1300
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 0.5414937759336099,
|
1840 |
+
"grad_norm": 2.4190653155415967,
|
1841 |
+
"learning_rate": 5.1484398030293135e-06,
|
1842 |
+
"loss": 0.4522,
|
1843 |
+
"step": 1305
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 0.5435684647302904,
|
1847 |
+
"grad_norm": 2.4062268029705636,
|
1848 |
+
"learning_rate": 5.112242037235885e-06,
|
1849 |
+
"loss": 0.4517,
|
1850 |
+
"step": 1310
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 0.5456431535269709,
|
1854 |
+
"grad_norm": 2.474917237517535,
|
1855 |
+
"learning_rate": 5.076038384710077e-06,
|
1856 |
+
"loss": 0.456,
|
1857 |
+
"step": 1315
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 0.5477178423236515,
|
1861 |
+
"grad_norm": 2.5820885338481077,
|
1862 |
+
"learning_rate": 5.039830744216548e-06,
|
1863 |
+
"loss": 0.4575,
|
1864 |
+
"step": 1320
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 0.549792531120332,
|
1868 |
+
"grad_norm": 2.5020192978566365,
|
1869 |
+
"learning_rate": 5.003621014729113e-06,
|
1870 |
+
"loss": 0.4468,
|
1871 |
+
"step": 1325
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 0.5518672199170125,
|
1875 |
+
"grad_norm": 2.4844474809833224,
|
1876 |
+
"learning_rate": 4.967411095331149e-06,
|
1877 |
+
"loss": 0.4426,
|
1878 |
+
"step": 1330
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 0.553941908713693,
|
1882 |
+
"grad_norm": 2.567949384188064,
|
1883 |
+
"learning_rate": 4.931202885115994e-06,
|
1884 |
+
"loss": 0.4491,
|
1885 |
+
"step": 1335
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 0.5560165975103735,
|
1889 |
+
"grad_norm": 2.3971814860863128,
|
1890 |
+
"learning_rate": 4.894998283087341e-06,
|
1891 |
+
"loss": 0.4521,
|
1892 |
+
"step": 1340
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 0.558091286307054,
|
1896 |
+
"grad_norm": 2.6050382832392502,
|
1897 |
+
"learning_rate": 4.858799188059651e-06,
|
1898 |
+
"loss": 0.4402,
|
1899 |
+
"step": 1345
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 0.5601659751037344,
|
1903 |
+
"grad_norm": 2.504067986804623,
|
1904 |
+
"learning_rate": 4.822607498558555e-06,
|
1905 |
+
"loss": 0.4309,
|
1906 |
+
"step": 1350
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 0.5622406639004149,
|
1910 |
+
"grad_norm": 2.436795227940985,
|
1911 |
+
"learning_rate": 4.786425112721293e-06,
|
1912 |
+
"loss": 0.4432,
|
1913 |
+
"step": 1355
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 0.5643153526970954,
|
1917 |
+
"grad_norm": 2.463086006536084,
|
1918 |
+
"learning_rate": 4.75025392819715e-06,
|
1919 |
+
"loss": 0.4547,
|
1920 |
+
"step": 1360
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 0.5663900414937759,
|
1924 |
+
"grad_norm": 2.526031011478867,
|
1925 |
+
"learning_rate": 4.714095842047952e-06,
|
1926 |
+
"loss": 0.4382,
|
1927 |
+
"step": 1365
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 0.5684647302904564,
|
1931 |
+
"grad_norm": 2.484312790550504,
|
1932 |
+
"learning_rate": 4.677952750648544e-06,
|
1933 |
+
"loss": 0.4395,
|
1934 |
+
"step": 1370
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 0.5705394190871369,
|
1938 |
+
"grad_norm": 2.51470713708296,
|
1939 |
+
"learning_rate": 4.641826549587352e-06,
|
1940 |
+
"loss": 0.4324,
|
1941 |
+
"step": 1375
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 0.5726141078838174,
|
1945 |
+
"grad_norm": 2.328972994785194,
|
1946 |
+
"learning_rate": 4.605719133566955e-06,
|
1947 |
+
"loss": 0.4409,
|
1948 |
+
"step": 1380
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 0.5746887966804979,
|
1952 |
+
"grad_norm": 2.467684453460364,
|
1953 |
+
"learning_rate": 4.56963239630472e-06,
|
1954 |
+
"loss": 0.4502,
|
1955 |
+
"step": 1385
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 0.5767634854771784,
|
1959 |
+
"grad_norm": 2.5795480839970093,
|
1960 |
+
"learning_rate": 4.533568230433477e-06,
|
1961 |
+
"loss": 0.4417,
|
1962 |
+
"step": 1390
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 0.578838174273859,
|
1966 |
+
"grad_norm": 2.3301068759580907,
|
1967 |
+
"learning_rate": 4.497528527402262e-06,
|
1968 |
+
"loss": 0.4361,
|
1969 |
+
"step": 1395
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 0.5809128630705395,
|
1973 |
+
"grad_norm": 2.581867988962586,
|
1974 |
+
"learning_rate": 4.461515177377113e-06,
|
1975 |
+
"loss": 0.4398,
|
1976 |
+
"step": 1400
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 0.58298755186722,
|
1980 |
+
"grad_norm": 2.437392550100252,
|
1981 |
+
"learning_rate": 4.42553006914194e-06,
|
1982 |
+
"loss": 0.4234,
|
1983 |
+
"step": 1405
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 0.5850622406639004,
|
1987 |
+
"grad_norm": 2.43460113329211,
|
1988 |
+
"learning_rate": 4.3895750899994566e-06,
|
1989 |
+
"loss": 0.4299,
|
1990 |
+
"step": 1410
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 0.5871369294605809,
|
1994 |
+
"grad_norm": 2.4540600474475016,
|
1995 |
+
"learning_rate": 4.353652125672208e-06,
|
1996 |
+
"loss": 0.429,
|
1997 |
+
"step": 1415
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 0.5892116182572614,
|
2001 |
+
"grad_norm": 2.402133768218254,
|
2002 |
+
"learning_rate": 4.317763060203665e-06,
|
2003 |
+
"loss": 0.4195,
|
2004 |
+
"step": 1420
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 0.5912863070539419,
|
2008 |
+
"grad_norm": 2.3040455325782103,
|
2009 |
+
"learning_rate": 4.281909775859411e-06,
|
2010 |
+
"loss": 0.4142,
|
2011 |
+
"step": 1425
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 0.5933609958506224,
|
2015 |
+
"grad_norm": 2.437892690340681,
|
2016 |
+
"learning_rate": 4.246094153028426e-06,
|
2017 |
+
"loss": 0.4376,
|
2018 |
+
"step": 1430
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 0.5954356846473029,
|
2022 |
+
"grad_norm": 2.4883366883550067,
|
2023 |
+
"learning_rate": 4.210318070124465e-06,
|
2024 |
+
"loss": 0.4343,
|
2025 |
+
"step": 1435
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 0.5975103734439834,
|
2029 |
+
"grad_norm": 2.623855396529678,
|
2030 |
+
"learning_rate": 4.1745834034875435e-06,
|
2031 |
+
"loss": 0.4316,
|
2032 |
+
"step": 1440
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 0.5995850622406639,
|
2036 |
+
"grad_norm": 2.4818254566920186,
|
2037 |
+
"learning_rate": 4.138892027285525e-06,
|
2038 |
+
"loss": 0.4201,
|
2039 |
+
"step": 1445
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 0.6016597510373444,
|
2043 |
+
"grad_norm": 2.487523042972382,
|
2044 |
+
"learning_rate": 4.10324581341583e-06,
|
2045 |
+
"loss": 0.4159,
|
2046 |
+
"step": 1450
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 0.6037344398340249,
|
2050 |
+
"grad_norm": 2.4142763091450186,
|
2051 |
+
"learning_rate": 4.067646631407259e-06,
|
2052 |
+
"loss": 0.4171,
|
2053 |
+
"step": 1455
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 0.6058091286307054,
|
2057 |
+
"grad_norm": 2.535593504431872,
|
2058 |
+
"learning_rate": 4.0320963483219485e-06,
|
2059 |
+
"loss": 0.4239,
|
2060 |
+
"step": 1460
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 0.6078838174273858,
|
2064 |
+
"grad_norm": 2.3293103348557094,
|
2065 |
+
"learning_rate": 3.996596828657437e-06,
|
2066 |
+
"loss": 0.423,
|
2067 |
+
"step": 1465
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 0.6099585062240664,
|
2071 |
+
"grad_norm": 2.3875069760256045,
|
2072 |
+
"learning_rate": 3.961149934248893e-06,
|
2073 |
+
"loss": 0.426,
|
2074 |
+
"step": 1470
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 0.6120331950207469,
|
2078 |
+
"grad_norm": 2.473197291626687,
|
2079 |
+
"learning_rate": 3.925757524171455e-06,
|
2080 |
+
"loss": 0.427,
|
2081 |
+
"step": 1475
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 0.6141078838174274,
|
2085 |
+
"grad_norm": 2.7741410864382754,
|
2086 |
+
"learning_rate": 3.8904214546427355e-06,
|
2087 |
+
"loss": 0.4122,
|
2088 |
+
"step": 1480
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 0.6161825726141079,
|
2092 |
+
"grad_norm": 2.4047571270961345,
|
2093 |
+
"learning_rate": 3.855143578925468e-06,
|
2094 |
+
"loss": 0.4131,
|
2095 |
+
"step": 1485
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 0.6182572614107884,
|
2099 |
+
"grad_norm": 2.423122390796353,
|
2100 |
+
"learning_rate": 3.819925747230309e-06,
|
2101 |
+
"loss": 0.4234,
|
2102 |
+
"step": 1490
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 0.6203319502074689,
|
2106 |
+
"grad_norm": 2.625870115353325,
|
2107 |
+
"learning_rate": 3.7847698066187975e-06,
|
2108 |
+
"loss": 0.4169,
|
2109 |
+
"step": 1495
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 0.6224066390041494,
|
2113 |
+
"grad_norm": 2.3339760463576047,
|
2114 |
+
"learning_rate": 3.749677600906489e-06,
|
2115 |
+
"loss": 0.4167,
|
2116 |
+
"step": 1500
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 0.6244813278008299,
|
2120 |
+
"grad_norm": 2.381759807149871,
|
2121 |
+
"learning_rate": 3.7146509705662453e-06,
|
2122 |
+
"loss": 0.4127,
|
2123 |
+
"step": 1505
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 0.6265560165975104,
|
2127 |
+
"grad_norm": 2.49037216210766,
|
2128 |
+
"learning_rate": 3.6796917526317153e-06,
|
2129 |
+
"loss": 0.4196,
|
2130 |
+
"step": 1510
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 0.6286307053941909,
|
2134 |
+
"grad_norm": 2.3307430355853813,
|
2135 |
+
"learning_rate": 3.6448017806009804e-06,
|
2136 |
+
"loss": 0.4032,
|
2137 |
+
"step": 1515
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 0.6307053941908713,
|
2141 |
+
"grad_norm": 2.459076699312393,
|
2142 |
+
"learning_rate": 3.609982884340402e-06,
|
2143 |
+
"loss": 0.4191,
|
2144 |
+
"step": 1520
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 0.6327800829875518,
|
2148 |
+
"grad_norm": 2.349992533584889,
|
2149 |
+
"learning_rate": 3.575236889988646e-06,
|
2150 |
+
"loss": 0.4096,
|
2151 |
+
"step": 1525
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 0.6348547717842323,
|
2155 |
+
"grad_norm": 2.505421167742011,
|
2156 |
+
"learning_rate": 3.540565619860906e-06,
|
2157 |
+
"loss": 0.4149,
|
2158 |
+
"step": 1530
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 0.6369294605809128,
|
2162 |
+
"grad_norm": 2.3662854321031994,
|
2163 |
+
"learning_rate": 3.5059708923533354e-06,
|
2164 |
+
"loss": 0.4092,
|
2165 |
+
"step": 1535
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 0.6390041493775933,
|
2169 |
+
"grad_norm": 2.33810078459921,
|
2170 |
+
"learning_rate": 3.4714545218476727e-06,
|
2171 |
+
"loss": 0.4022,
|
2172 |
+
"step": 1540
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 0.6410788381742739,
|
2176 |
+
"grad_norm": 2.292723665647368,
|
2177 |
+
"learning_rate": 3.437018318616084e-06,
|
2178 |
+
"loss": 0.3924,
|
2179 |
+
"step": 1545
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 0.6431535269709544,
|
2183 |
+
"grad_norm": 2.3059685019935343,
|
2184 |
+
"learning_rate": 3.4026640887262196e-06,
|
2185 |
+
"loss": 0.4062,
|
2186 |
+
"step": 1550
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 0.6452282157676349,
|
2190 |
+
"grad_norm": 2.4576342501679966,
|
2191 |
+
"learning_rate": 3.3683936339464957e-06,
|
2192 |
+
"loss": 0.408,
|
2193 |
+
"step": 1555
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 0.6473029045643154,
|
2197 |
+
"grad_norm": 2.2902363586226624,
|
2198 |
+
"learning_rate": 3.334208751651593e-06,
|
2199 |
+
"loss": 0.4123,
|
2200 |
+
"step": 1560
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 0.6493775933609959,
|
2204 |
+
"grad_norm": 2.2743147574529328,
|
2205 |
+
"learning_rate": 3.300111234728191e-06,
|
2206 |
+
"loss": 0.4097,
|
2207 |
+
"step": 1565
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 0.6514522821576764,
|
2211 |
+
"grad_norm": 2.4569598537485327,
|
2212 |
+
"learning_rate": 3.2661028714809405e-06,
|
2213 |
+
"loss": 0.4012,
|
2214 |
+
"step": 1570
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 0.6535269709543569,
|
2218 |
+
"grad_norm": 2.28227869578094,
|
2219 |
+
"learning_rate": 3.2321854455386657e-06,
|
2220 |
+
"loss": 0.4008,
|
2221 |
+
"step": 1575
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 0.6556016597510373,
|
2225 |
+
"grad_norm": 2.4708407304872533,
|
2226 |
+
"learning_rate": 3.198360735760827e-06,
|
2227 |
+
"loss": 0.3873,
|
2228 |
+
"step": 1580
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 0.6576763485477178,
|
2232 |
+
"grad_norm": 2.2707165698258103,
|
2233 |
+
"learning_rate": 3.1646305161442183e-06,
|
2234 |
+
"loss": 0.4059,
|
2235 |
+
"step": 1585
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 0.6597510373443983,
|
2239 |
+
"grad_norm": 2.2036975566545993,
|
2240 |
+
"learning_rate": 3.1309965557299303e-06,
|
2241 |
+
"loss": 0.3878,
|
2242 |
+
"step": 1590
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 0.6618257261410788,
|
2246 |
+
"grad_norm": 2.4960520314367627,
|
2247 |
+
"learning_rate": 3.097460618510571e-06,
|
2248 |
+
"loss": 0.3885,
|
2249 |
+
"step": 1595
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 0.6639004149377593,
|
2253 |
+
"grad_norm": 2.2804809511371755,
|
2254 |
+
"learning_rate": 3.064024463337747e-06,
|
2255 |
+
"loss": 0.4002,
|
2256 |
+
"step": 1600
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 0.6659751037344398,
|
2260 |
+
"grad_norm": 2.6325297095462767,
|
2261 |
+
"learning_rate": 3.0306898438298184e-06,
|
2262 |
+
"loss": 0.4003,
|
2263 |
+
"step": 1605
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 0.6680497925311203,
|
2267 |
+
"grad_norm": 2.366691569211348,
|
2268 |
+
"learning_rate": 2.997458508279928e-06,
|
2269 |
+
"loss": 0.3941,
|
2270 |
+
"step": 1610
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 0.6701244813278008,
|
2274 |
+
"grad_norm": 2.2056640704322494,
|
2275 |
+
"learning_rate": 2.964332199564309e-06,
|
2276 |
+
"loss": 0.3966,
|
2277 |
+
"step": 1615
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 0.6721991701244814,
|
2281 |
+
"grad_norm": 2.423642247816294,
|
2282 |
+
"learning_rate": 2.9313126550508762e-06,
|
2283 |
+
"loss": 0.396,
|
2284 |
+
"step": 1620
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 0.6742738589211619,
|
2288 |
+
"grad_norm": 2.433390053325066,
|
2289 |
+
"learning_rate": 2.8984016065081073e-06,
|
2290 |
+
"loss": 0.4023,
|
2291 |
+
"step": 1625
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 0.6763485477178424,
|
2295 |
+
"grad_norm": 2.357419717448147,
|
2296 |
+
"learning_rate": 2.865600780014216e-06,
|
2297 |
+
"loss": 0.4088,
|
2298 |
+
"step": 1630
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 0.6784232365145229,
|
2302 |
+
"grad_norm": 2.245544606902691,
|
2303 |
+
"learning_rate": 2.8329118958666236e-06,
|
2304 |
+
"loss": 0.3858,
|
2305 |
+
"step": 1635
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 0.6804979253112033,
|
2309 |
+
"grad_norm": 2.439298195310241,
|
2310 |
+
"learning_rate": 2.80033666849174e-06,
|
2311 |
+
"loss": 0.3986,
|
2312 |
+
"step": 1640
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 0.6825726141078838,
|
2316 |
+
"grad_norm": 2.4100345874144518,
|
2317 |
+
"learning_rate": 2.7678768063550454e-06,
|
2318 |
+
"loss": 0.3946,
|
2319 |
+
"step": 1645
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 0.6846473029045643,
|
2323 |
+
"grad_norm": 2.3584382118069973,
|
2324 |
+
"learning_rate": 2.735534011871479e-06,
|
2325 |
+
"loss": 0.378,
|
2326 |
+
"step": 1650
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 0.6867219917012448,
|
2330 |
+
"grad_norm": 2.5751863977070903,
|
2331 |
+
"learning_rate": 2.7033099813161696e-06,
|
2332 |
+
"loss": 0.3901,
|
2333 |
+
"step": 1655
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 0.6887966804979253,
|
2337 |
+
"grad_norm": 2.2159484402883827,
|
2338 |
+
"learning_rate": 2.6712064047354515e-06,
|
2339 |
+
"loss": 0.393,
|
2340 |
+
"step": 1660
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 0.6908713692946058,
|
2344 |
+
"grad_norm": 2.3569350609642257,
|
2345 |
+
"learning_rate": 2.6392249658582454e-06,
|
2346 |
+
"loss": 0.3818,
|
2347 |
+
"step": 1665
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 0.6929460580912863,
|
2351 |
+
"grad_norm": 2.476100739673621,
|
2352 |
+
"learning_rate": 2.607367342007738e-06,
|
2353 |
+
"loss": 0.387,
|
2354 |
+
"step": 1670
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 0.6950207468879668,
|
2358 |
+
"grad_norm": 2.394265629557479,
|
2359 |
+
"learning_rate": 2.5756352040134193e-06,
|
2360 |
+
"loss": 0.3893,
|
2361 |
+
"step": 1675
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 0.6970954356846473,
|
2365 |
+
"grad_norm": 2.4572385167590234,
|
2366 |
+
"learning_rate": 2.5440302161234542e-06,
|
2367 |
+
"loss": 0.3962,
|
2368 |
+
"step": 1680
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 0.6991701244813278,
|
2372 |
+
"grad_norm": 2.3166108869604387,
|
2373 |
+
"learning_rate": 2.5125540359173893e-06,
|
2374 |
+
"loss": 0.3935,
|
2375 |
+
"step": 1685
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 0.7012448132780082,
|
2379 |
+
"grad_norm": 2.446525027779214,
|
2380 |
+
"learning_rate": 2.481208314219233e-06,
|
2381 |
+
"loss": 0.3882,
|
2382 |
+
"step": 1690
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 0.7033195020746889,
|
2386 |
+
"grad_norm": 2.284958027056727,
|
2387 |
+
"learning_rate": 2.449994695010856e-06,
|
2388 |
+
"loss": 0.3758,
|
2389 |
+
"step": 1695
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 0.7053941908713693,
|
2393 |
+
"grad_norm": 2.516170194826004,
|
2394 |
+
"learning_rate": 2.4189148153457875e-06,
|
2395 |
+
"loss": 0.3882,
|
2396 |
+
"step": 1700
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 0.7074688796680498,
|
2400 |
+
"grad_norm": 2.3397843827362914,
|
2401 |
+
"learning_rate": 2.387970305263349e-06,
|
2402 |
+
"loss": 0.3887,
|
2403 |
+
"step": 1705
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 0.7095435684647303,
|
2407 |
+
"grad_norm": 2.3903105786705123,
|
2408 |
+
"learning_rate": 2.3571627877031596e-06,
|
2409 |
+
"loss": 0.3785,
|
2410 |
+
"step": 1710
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 0.7116182572614108,
|
2414 |
+
"grad_norm": 2.4388910192687523,
|
2415 |
+
"learning_rate": 2.326493878420028e-06,
|
2416 |
+
"loss": 0.3704,
|
2417 |
+
"step": 1715
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 0.7136929460580913,
|
2421 |
+
"grad_norm": 2.3725455021448245,
|
2422 |
+
"learning_rate": 2.295965185899205e-06,
|
2423 |
+
"loss": 0.3778,
|
2424 |
+
"step": 1720
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 0.7157676348547718,
|
2428 |
+
"grad_norm": 2.3541520314694946,
|
2429 |
+
"learning_rate": 2.265578311272021e-06,
|
2430 |
+
"loss": 0.3827,
|
2431 |
+
"step": 1725
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 0.7178423236514523,
|
2435 |
+
"grad_norm": 2.318059450682999,
|
2436 |
+
"learning_rate": 2.2353348482319233e-06,
|
2437 |
+
"loss": 0.3641,
|
2438 |
+
"step": 1730
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 0.7199170124481328,
|
2442 |
+
"grad_norm": 2.303956633945153,
|
2443 |
+
"learning_rate": 2.2052363829508776e-06,
|
2444 |
+
"loss": 0.3804,
|
2445 |
+
"step": 1735
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 0.7219917012448133,
|
2449 |
+
"grad_norm": 2.3145838554705547,
|
2450 |
+
"learning_rate": 2.1752844939961926e-06,
|
2451 |
+
"loss": 0.3747,
|
2452 |
+
"step": 1740
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 0.7240663900414938,
|
2456 |
+
"grad_norm": 2.2586817786943603,
|
2457 |
+
"learning_rate": 2.1454807522477128e-06,
|
2458 |
+
"loss": 0.3722,
|
2459 |
+
"step": 1745
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 0.7261410788381742,
|
2463 |
+
"grad_norm": 2.4038449021090256,
|
2464 |
+
"learning_rate": 2.1158267208154497e-06,
|
2465 |
+
"loss": 0.3807,
|
2466 |
+
"step": 1750
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 0.7282157676348547,
|
2470 |
+
"grad_norm": 2.2154594010619166,
|
2471 |
+
"learning_rate": 2.0863239549575865e-06,
|
2472 |
+
"loss": 0.3763,
|
2473 |
+
"step": 1755
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 0.7302904564315352,
|
2477 |
+
"grad_norm": 2.5107957004825976,
|
2478 |
+
"learning_rate": 2.0569740019989136e-06,
|
2479 |
+
"loss": 0.3752,
|
2480 |
+
"step": 1760
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 0.7323651452282157,
|
2484 |
+
"grad_norm": 2.350354806134385,
|
2485 |
+
"learning_rate": 2.0277784012496865e-06,
|
2486 |
+
"loss": 0.3819,
|
2487 |
+
"step": 1765
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 0.7344398340248963,
|
2491 |
+
"grad_norm": 2.4464933220936125,
|
2492 |
+
"learning_rate": 1.998738683924875e-06,
|
2493 |
+
"loss": 0.3628,
|
2494 |
+
"step": 1770
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 0.7365145228215768,
|
2498 |
+
"grad_norm": 2.5270841137858064,
|
2499 |
+
"learning_rate": 1.9698563730638794e-06,
|
2500 |
+
"loss": 0.3754,
|
2501 |
+
"step": 1775
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 0.7385892116182573,
|
2505 |
+
"grad_norm": 2.250774099933747,
|
2506 |
+
"learning_rate": 1.9411329834506286e-06,
|
2507 |
+
"loss": 0.3667,
|
2508 |
+
"step": 1780
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 0.7406639004149378,
|
2512 |
+
"grad_norm": 2.3407235003652804,
|
2513 |
+
"learning_rate": 1.9125700215341476e-06,
|
2514 |
+
"loss": 0.3735,
|
2515 |
+
"step": 1785
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 0.7427385892116183,
|
2519 |
+
"grad_norm": 2.3446167395429347,
|
2520 |
+
"learning_rate": 1.8841689853495516e-06,
|
2521 |
+
"loss": 0.3723,
|
2522 |
+
"step": 1790
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 0.7448132780082988,
|
2526 |
+
"grad_norm": 2.3350133034877656,
|
2527 |
+
"learning_rate": 1.8559313644394677e-06,
|
2528 |
+
"loss": 0.3785,
|
2529 |
+
"step": 1795
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 0.7468879668049793,
|
2533 |
+
"grad_norm": 2.37500798404668,
|
2534 |
+
"learning_rate": 1.827858639775925e-06,
|
2535 |
+
"loss": 0.3786,
|
2536 |
+
"step": 1800
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 0.7489626556016598,
|
2540 |
+
"grad_norm": 2.4392836582028945,
|
2541 |
+
"learning_rate": 1.7999522836826744e-06,
|
2542 |
+
"loss": 0.3736,
|
2543 |
+
"step": 1805
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 0.7510373443983402,
|
2547 |
+
"grad_norm": 2.408718217237952,
|
2548 |
+
"learning_rate": 1.7722137597579698e-06,
|
2549 |
+
"loss": 0.3648,
|
2550 |
+
"step": 1810
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 0.7531120331950207,
|
2554 |
+
"grad_norm": 2.4393541244678216,
|
2555 |
+
"learning_rate": 1.744644522797817e-06,
|
2556 |
+
"loss": 0.3556,
|
2557 |
+
"step": 1815
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 0.7551867219917012,
|
2561 |
+
"grad_norm": 2.360665467065917,
|
2562 |
+
"learning_rate": 1.7172460187196588e-06,
|
2563 |
+
"loss": 0.3754,
|
2564 |
+
"step": 1820
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 0.7572614107883817,
|
2568 |
+
"grad_norm": 2.3486896399291,
|
2569 |
+
"learning_rate": 1.6900196844865575e-06,
|
2570 |
+
"loss": 0.3615,
|
2571 |
+
"step": 1825
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 0.7593360995850622,
|
2575 |
+
"grad_norm": 2.330649947928342,
|
2576 |
+
"learning_rate": 1.6629669480318166e-06,
|
2577 |
+
"loss": 0.3763,
|
2578 |
+
"step": 1830
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 0.7614107883817427,
|
2582 |
+
"grad_norm": 2.3447118121017287,
|
2583 |
+
"learning_rate": 1.6360892281841007e-06,
|
2584 |
+
"loss": 0.3639,
|
2585 |
+
"step": 1835
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 0.7634854771784232,
|
2589 |
+
"grad_norm": 2.507491155826583,
|
2590 |
+
"learning_rate": 1.609387934593019e-06,
|
2591 |
+
"loss": 0.3779,
|
2592 |
+
"step": 1840
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 0.7655601659751037,
|
2596 |
+
"grad_norm": 2.288411757086649,
|
2597 |
+
"learning_rate": 1.5828644676551892e-06,
|
2598 |
+
"loss": 0.366,
|
2599 |
+
"step": 1845
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 0.7676348547717843,
|
2603 |
+
"grad_norm": 2.3840080652247826,
|
2604 |
+
"learning_rate": 1.5565202184408e-06,
|
2605 |
+
"loss": 0.3589,
|
2606 |
+
"step": 1850
|
2607 |
+
},
|
2608 |
+
{
|
2609 |
+
"epoch": 0.7697095435684648,
|
2610 |
+
"grad_norm": 2.4900677150816364,
|
2611 |
+
"learning_rate": 1.5303565686206452e-06,
|
2612 |
+
"loss": 0.3683,
|
2613 |
+
"step": 1855
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 0.7717842323651453,
|
2617 |
+
"grad_norm": 2.5276222084732067,
|
2618 |
+
"learning_rate": 1.5043748903936672e-06,
|
2619 |
+
"loss": 0.3725,
|
2620 |
+
"step": 1860
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 0.7738589211618258,
|
2624 |
+
"grad_norm": 2.406387332540463,
|
2625 |
+
"learning_rate": 1.4785765464149836e-06,
|
2626 |
+
"loss": 0.3511,
|
2627 |
+
"step": 1865
|
2628 |
+
},
|
2629 |
+
{
|
2630 |
+
"epoch": 0.7759336099585062,
|
2631 |
+
"grad_norm": 2.409678473099456,
|
2632 |
+
"learning_rate": 1.4529628897244214e-06,
|
2633 |
+
"loss": 0.3637,
|
2634 |
+
"step": 1870
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 0.7780082987551867,
|
2638 |
+
"grad_norm": 2.372236015081896,
|
2639 |
+
"learning_rate": 1.4275352636755613e-06,
|
2640 |
+
"loss": 0.3693,
|
2641 |
+
"step": 1875
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"epoch": 0.7800829875518672,
|
2645 |
+
"grad_norm": 2.3018141960934013,
|
2646 |
+
"learning_rate": 1.4022950018652703e-06,
|
2647 |
+
"loss": 0.3557,
|
2648 |
+
"step": 1880
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 0.7821576763485477,
|
2652 |
+
"grad_norm": 2.3440334623732375,
|
2653 |
+
"learning_rate": 1.3772434280637737e-06,
|
2654 |
+
"loss": 0.3669,
|
2655 |
+
"step": 1885
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 0.7842323651452282,
|
2659 |
+
"grad_norm": 2.4523078746497755,
|
2660 |
+
"learning_rate": 1.3523818561452145e-06,
|
2661 |
+
"loss": 0.3646,
|
2662 |
+
"step": 1890
|
2663 |
+
},
|
2664 |
+
{
|
2665 |
+
"epoch": 0.7863070539419087,
|
2666 |
+
"grad_norm": 2.4219074003538346,
|
2667 |
+
"learning_rate": 1.3277115900187515e-06,
|
2668 |
+
"loss": 0.3577,
|
2669 |
+
"step": 1895
|
2670 |
+
},
|
2671 |
+
{
|
2672 |
+
"epoch": 0.7883817427385892,
|
2673 |
+
"grad_norm": 2.8682882199437842,
|
2674 |
+
"learning_rate": 1.3032339235601749e-06,
|
2675 |
+
"loss": 0.3474,
|
2676 |
+
"step": 1900
|
2677 |
+
},
|
2678 |
+
{
|
2679 |
+
"epoch": 0.7904564315352697,
|
2680 |
+
"grad_norm": 2.329298210972444,
|
2681 |
+
"learning_rate": 1.2789501405440397e-06,
|
2682 |
+
"loss": 0.3725,
|
2683 |
+
"step": 1905
|
2684 |
+
},
|
2685 |
+
{
|
2686 |
+
"epoch": 0.7925311203319502,
|
2687 |
+
"grad_norm": 2.365427473253902,
|
2688 |
+
"learning_rate": 1.254861514576348e-06,
|
2689 |
+
"loss": 0.3705,
|
2690 |
+
"step": 1910
|
2691 |
+
},
|
2692 |
+
{
|
2693 |
+
"epoch": 0.7946058091286307,
|
2694 |
+
"grad_norm": 2.449442819984942,
|
2695 |
+
"learning_rate": 1.2309693090277392e-06,
|
2696 |
+
"loss": 0.3498,
|
2697 |
+
"step": 1915
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 0.7966804979253111,
|
2701 |
+
"grad_norm": 2.2309766163591362,
|
2702 |
+
"learning_rate": 1.207274776967235e-06,
|
2703 |
+
"loss": 0.3497,
|
2704 |
+
"step": 1920
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"epoch": 0.7987551867219918,
|
2708 |
+
"grad_norm": 2.5332149638146606,
|
2709 |
+
"learning_rate": 1.183779161096526e-06,
|
2710 |
+
"loss": 0.3533,
|
2711 |
+
"step": 1925
|
2712 |
+
},
|
2713 |
+
{
|
2714 |
+
"epoch": 0.8008298755186722,
|
2715 |
+
"grad_norm": 2.464873395759963,
|
2716 |
+
"learning_rate": 1.1604836936847852e-06,
|
2717 |
+
"loss": 0.3669,
|
2718 |
+
"step": 1930
|
2719 |
+
},
|
2720 |
+
{
|
2721 |
+
"epoch": 0.8029045643153527,
|
2722 |
+
"grad_norm": 2.353844278985767,
|
2723 |
+
"learning_rate": 1.1373895965040522e-06,
|
2724 |
+
"loss": 0.3531,
|
2725 |
+
"step": 1935
|
2726 |
+
},
|
2727 |
+
{
|
2728 |
+
"epoch": 0.8049792531120332,
|
2729 |
+
"grad_norm": 2.478326080734339,
|
2730 |
+
"learning_rate": 1.1144980807651413e-06,
|
2731 |
+
"loss": 0.3469,
|
2732 |
+
"step": 1940
|
2733 |
+
},
|
2734 |
+
{
|
2735 |
+
"epoch": 0.8070539419087137,
|
2736 |
+
"grad_norm": 2.3657798765264837,
|
2737 |
+
"learning_rate": 1.0918103470541297e-06,
|
2738 |
+
"loss": 0.3434,
|
2739 |
+
"step": 1945
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 0.8091286307053942,
|
2743 |
+
"grad_norm": 2.4345052899894517,
|
2744 |
+
"learning_rate": 1.069327585269383e-06,
|
2745 |
+
"loss": 0.3575,
|
2746 |
+
"step": 1950
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 0.8112033195020747,
|
2750 |
+
"grad_norm": 2.335213511709306,
|
2751 |
+
"learning_rate": 1.047050974559149e-06,
|
2752 |
+
"loss": 0.3542,
|
2753 |
+
"step": 1955
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 0.8132780082987552,
|
2757 |
+
"grad_norm": 2.3666035829182164,
|
2758 |
+
"learning_rate": 1.024981683259723e-06,
|
2759 |
+
"loss": 0.3484,
|
2760 |
+
"step": 1960
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 0.8153526970954357,
|
2764 |
+
"grad_norm": 2.461171014362592,
|
2765 |
+
"learning_rate": 1.0031208688341609e-06,
|
2766 |
+
"loss": 0.3484,
|
2767 |
+
"step": 1965
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 0.8174273858921162,
|
2771 |
+
"grad_norm": 2.5462843981961223,
|
2772 |
+
"learning_rate": 9.814696778115834e-07,
|
2773 |
+
"loss": 0.3466,
|
2774 |
+
"step": 1970
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 0.8195020746887967,
|
2778 |
+
"grad_norm": 2.3821519258481705,
|
2779 |
+
"learning_rate": 9.600292457270416e-07,
|
2780 |
+
"loss": 0.3502,
|
2781 |
+
"step": 1975
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 0.8215767634854771,
|
2785 |
+
"grad_norm": 2.397078021612525,
|
2786 |
+
"learning_rate": 9.388006970619557e-07,
|
2787 |
+
"loss": 0.3518,
|
2788 |
+
"step": 1980
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 0.8236514522821576,
|
2792 |
+
"grad_norm": 2.3345715923241945,
|
2793 |
+
"learning_rate": 9.177851451851511e-07,
|
2794 |
+
"loss": 0.3483,
|
2795 |
+
"step": 1985
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 0.8257261410788381,
|
2799 |
+
"grad_norm": 2.363653449360418,
|
2800 |
+
"learning_rate": 8.969836922944536e-07,
|
2801 |
+
"loss": 0.3414,
|
2802 |
+
"step": 1990
|
2803 |
+
},
|
2804 |
+
{
|
2805 |
+
"epoch": 0.8278008298755186,
|
2806 |
+
"grad_norm": 2.3624129572192616,
|
2807 |
+
"learning_rate": 8.76397429358895e-07,
|
2808 |
+
"loss": 0.359,
|
2809 |
+
"step": 1995
|
2810 |
+
},
|
2811 |
+
{
|
2812 |
+
"epoch": 0.8298755186721992,
|
2813 |
+
"grad_norm": 2.314328813522686,
|
2814 |
+
"learning_rate": 8.560274360614829e-07,
|
2815 |
+
"loss": 0.3385,
|
2816 |
+
"step": 2000
|
2817 |
+
},
|
2818 |
+
{
|
2819 |
+
"epoch": 0.8319502074688797,
|
2820 |
+
"grad_norm": 2.199340475549847,
|
2821 |
+
"learning_rate": 8.358747807425827e-07,
|
2822 |
+
"loss": 0.3539,
|
2823 |
+
"step": 2005
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 0.8340248962655602,
|
2827 |
+
"grad_norm": 2.4399680516899633,
|
2828 |
+
"learning_rate": 8.159405203438875e-07,
|
2829 |
+
"loss": 0.3518,
|
2830 |
+
"step": 2010
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"epoch": 0.8360995850622407,
|
2834 |
+
"grad_norm": 2.4262479785068733,
|
2835 |
+
"learning_rate": 7.962257003529777e-07,
|
2836 |
+
"loss": 0.344,
|
2837 |
+
"step": 2015
|
2838 |
+
},
|
2839 |
+
{
|
2840 |
+
"epoch": 0.8381742738589212,
|
2841 |
+
"grad_norm": 2.3545302681837987,
|
2842 |
+
"learning_rate": 7.767313547484979e-07,
|
2843 |
+
"loss": 0.3487,
|
2844 |
+
"step": 2020
|
2845 |
+
},
|
2846 |
+
{
|
2847 |
+
"epoch": 0.8402489626556017,
|
2848 |
+
"grad_norm": 2.3064171274116747,
|
2849 |
+
"learning_rate": 7.574585059459171e-07,
|
2850 |
+
"loss": 0.3447,
|
2851 |
+
"step": 2025
|
2852 |
+
},
|
2853 |
+
{
|
2854 |
+
"epoch": 0.8423236514522822,
|
2855 |
+
"grad_norm": 2.5487816431519477,
|
2856 |
+
"learning_rate": 7.384081647439184e-07,
|
2857 |
+
"loss": 0.356,
|
2858 |
+
"step": 2030
|
2859 |
+
},
|
2860 |
+
{
|
2861 |
+
"epoch": 0.8443983402489627,
|
2862 |
+
"grad_norm": 2.459116013365039,
|
2863 |
+
"learning_rate": 7.195813302713761e-07,
|
2864 |
+
"loss": 0.3375,
|
2865 |
+
"step": 2035
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 0.8464730290456431,
|
2869 |
+
"grad_norm": 2.4240198038099905,
|
2870 |
+
"learning_rate": 7.009789899349567e-07,
|
2871 |
+
"loss": 0.3427,
|
2872 |
+
"step": 2040
|
2873 |
+
},
|
2874 |
+
{
|
2875 |
+
"epoch": 0.8485477178423236,
|
2876 |
+
"grad_norm": 2.339613272043201,
|
2877 |
+
"learning_rate": 6.826021193673416e-07,
|
2878 |
+
"loss": 0.3481,
|
2879 |
+
"step": 2045
|
2880 |
+
},
|
2881 |
+
{
|
2882 |
+
"epoch": 0.8506224066390041,
|
2883 |
+
"grad_norm": 2.4864055871608484,
|
2884 |
+
"learning_rate": 6.644516823760439e-07,
|
2885 |
+
"loss": 0.3488,
|
2886 |
+
"step": 2050
|
2887 |
+
},
|
2888 |
+
{
|
2889 |
+
"epoch": 0.8526970954356846,
|
2890 |
+
"grad_norm": 2.3920126614464254,
|
2891 |
+
"learning_rate": 6.465286308928676e-07,
|
2892 |
+
"loss": 0.3427,
|
2893 |
+
"step": 2055
|
2894 |
+
},
|
2895 |
+
{
|
2896 |
+
"epoch": 0.8547717842323651,
|
2897 |
+
"grad_norm": 2.432956486348477,
|
2898 |
+
"learning_rate": 6.28833904923985e-07,
|
2899 |
+
"loss": 0.3522,
|
2900 |
+
"step": 2060
|
2901 |
+
},
|
2902 |
+
{
|
2903 |
+
"epoch": 0.8568464730290456,
|
2904 |
+
"grad_norm": 2.4012672614414243,
|
2905 |
+
"learning_rate": 6.113684325006286e-07,
|
2906 |
+
"loss": 0.3426,
|
2907 |
+
"step": 2065
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 0.8589211618257261,
|
2911 |
+
"grad_norm": 2.4925509857922656,
|
2912 |
+
"learning_rate": 5.941331296304254e-07,
|
2913 |
+
"loss": 0.3428,
|
2914 |
+
"step": 2070
|
2915 |
+
},
|
2916 |
+
{
|
2917 |
+
"epoch": 0.8609958506224067,
|
2918 |
+
"grad_norm": 2.2832723407803606,
|
2919 |
+
"learning_rate": 5.771289002493502e-07,
|
2920 |
+
"loss": 0.3472,
|
2921 |
+
"step": 2075
|
2922 |
+
},
|
2923 |
+
{
|
2924 |
+
"epoch": 0.8630705394190872,
|
2925 |
+
"grad_norm": 2.3874716687155564,
|
2926 |
+
"learning_rate": 5.603566361743229e-07,
|
2927 |
+
"loss": 0.3413,
|
2928 |
+
"step": 2080
|
2929 |
+
},
|
2930 |
+
{
|
2931 |
+
"epoch": 0.8651452282157677,
|
2932 |
+
"grad_norm": 2.351570855990709,
|
2933 |
+
"learning_rate": 5.438172170564288e-07,
|
2934 |
+
"loss": 0.3322,
|
2935 |
+
"step": 2085
|
2936 |
+
},
|
2937 |
+
{
|
2938 |
+
"epoch": 0.8672199170124482,
|
2939 |
+
"grad_norm": 2.5008832392383207,
|
2940 |
+
"learning_rate": 5.275115103347884e-07,
|
2941 |
+
"loss": 0.3477,
|
2942 |
+
"step": 2090
|
2943 |
+
},
|
2944 |
+
{
|
2945 |
+
"epoch": 0.8692946058091287,
|
2946 |
+
"grad_norm": 2.3216651986422088,
|
2947 |
+
"learning_rate": 5.114403711910631e-07,
|
2948 |
+
"loss": 0.336,
|
2949 |
+
"step": 2095
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 0.8713692946058091,
|
2953 |
+
"grad_norm": 2.326035363037846,
|
2954 |
+
"learning_rate": 4.956046425045991e-07,
|
2955 |
+
"loss": 0.3439,
|
2956 |
+
"step": 2100
|
2957 |
+
},
|
2958 |
+
{
|
2959 |
+
"epoch": 0.8734439834024896,
|
2960 |
+
"grad_norm": 2.4613685736167557,
|
2961 |
+
"learning_rate": 4.80005154808228e-07,
|
2962 |
+
"loss": 0.3468,
|
2963 |
+
"step": 2105
|
2964 |
+
},
|
2965 |
+
{
|
2966 |
+
"epoch": 0.8755186721991701,
|
2967 |
+
"grad_norm": 2.3791168252802537,
|
2968 |
+
"learning_rate": 4.646427262447034e-07,
|
2969 |
+
"loss": 0.3572,
|
2970 |
+
"step": 2110
|
2971 |
+
},
|
2972 |
+
{
|
2973 |
+
"epoch": 0.8775933609958506,
|
2974 |
+
"grad_norm": 2.437202584127699,
|
2975 |
+
"learning_rate": 4.49518162523791e-07,
|
2976 |
+
"loss": 0.3462,
|
2977 |
+
"step": 2115
|
2978 |
+
},
|
2979 |
+
{
|
2980 |
+
"epoch": 0.8796680497925311,
|
2981 |
+
"grad_norm": 2.4764470732059096,
|
2982 |
+
"learning_rate": 4.34632256880016e-07,
|
2983 |
+
"loss": 0.3445,
|
2984 |
+
"step": 2120
|
2985 |
+
},
|
2986 |
+
{
|
2987 |
+
"epoch": 0.8817427385892116,
|
2988 |
+
"grad_norm": 2.3787585521250945,
|
2989 |
+
"learning_rate": 4.1998579003105553e-07,
|
2990 |
+
"loss": 0.3401,
|
2991 |
+
"step": 2125
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 0.8838174273858921,
|
2995 |
+
"grad_norm": 2.3948308594804346,
|
2996 |
+
"learning_rate": 4.0557953013679873e-07,
|
2997 |
+
"loss": 0.3536,
|
2998 |
+
"step": 2130
|
2999 |
+
},
|
3000 |
+
{
|
3001 |
+
"epoch": 0.8858921161825726,
|
3002 |
+
"grad_norm": 2.3397934739178363,
|
3003 |
+
"learning_rate": 3.914142327590531e-07,
|
3004 |
+
"loss": 0.3334,
|
3005 |
+
"step": 2135
|
3006 |
+
},
|
3007 |
+
{
|
3008 |
+
"epoch": 0.8879668049792531,
|
3009 |
+
"grad_norm": 2.45922644953553,
|
3010 |
+
"learning_rate": 3.7749064082191976e-07,
|
3011 |
+
"loss": 0.3349,
|
3012 |
+
"step": 2140
|
3013 |
+
},
|
3014 |
+
{
|
3015 |
+
"epoch": 0.8900414937759336,
|
3016 |
+
"grad_norm": 2.413840835433163,
|
3017 |
+
"learning_rate": 3.6380948457283293e-07,
|
3018 |
+
"loss": 0.3315,
|
3019 |
+
"step": 2145
|
3020 |
+
},
|
3021 |
+
{
|
3022 |
+
"epoch": 0.8921161825726142,
|
3023 |
+
"grad_norm": 2.347524977176241,
|
3024 |
+
"learning_rate": 3.5037148154425494e-07,
|
3025 |
+
"loss": 0.346,
|
3026 |
+
"step": 2150
|
3027 |
+
},
|
3028 |
+
{
|
3029 |
+
"epoch": 0.8941908713692946,
|
3030 |
+
"grad_norm": 2.47677069324449,
|
3031 |
+
"learning_rate": 3.3717733651604967e-07,
|
3032 |
+
"loss": 0.3415,
|
3033 |
+
"step": 2155
|
3034 |
+
},
|
3035 |
+
{
|
3036 |
+
"epoch": 0.8962655601659751,
|
3037 |
+
"grad_norm": 2.397745011270569,
|
3038 |
+
"learning_rate": 3.242277414785128e-07,
|
3039 |
+
"loss": 0.3384,
|
3040 |
+
"step": 2160
|
3041 |
+
},
|
3042 |
+
{
|
3043 |
+
"epoch": 0.8983402489626556,
|
3044 |
+
"grad_norm": 2.415367296859196,
|
3045 |
+
"learning_rate": 3.1152337559608725e-07,
|
3046 |
+
"loss": 0.3387,
|
3047 |
+
"step": 2165
|
3048 |
+
},
|
3049 |
+
{
|
3050 |
+
"epoch": 0.9004149377593361,
|
3051 |
+
"grad_norm": 2.23801113438236,
|
3052 |
+
"learning_rate": 2.990649051717348e-07,
|
3053 |
+
"loss": 0.3458,
|
3054 |
+
"step": 2170
|
3055 |
+
},
|
3056 |
+
{
|
3057 |
+
"epoch": 0.9024896265560166,
|
3058 |
+
"grad_norm": 2.4178418849940413,
|
3059 |
+
"learning_rate": 2.868529836119954e-07,
|
3060 |
+
"loss": 0.3347,
|
3061 |
+
"step": 2175
|
3062 |
+
},
|
3063 |
+
{
|
3064 |
+
"epoch": 0.9045643153526971,
|
3065 |
+
"grad_norm": 2.5082956753743204,
|
3066 |
+
"learning_rate": 2.7488825139272037e-07,
|
3067 |
+
"loss": 0.3368,
|
3068 |
+
"step": 2180
|
3069 |
+
},
|
3070 |
+
{
|
3071 |
+
"epoch": 0.9066390041493776,
|
3072 |
+
"grad_norm": 2.2379502275244194,
|
3073 |
+
"learning_rate": 2.631713360254734e-07,
|
3074 |
+
"loss": 0.3436,
|
3075 |
+
"step": 2185
|
3076 |
+
},
|
3077 |
+
{
|
3078 |
+
"epoch": 0.9087136929460581,
|
3079 |
+
"grad_norm": 2.424932946521832,
|
3080 |
+
"learning_rate": 2.5170285202462895e-07,
|
3081 |
+
"loss": 0.3446,
|
3082 |
+
"step": 2190
|
3083 |
+
},
|
3084 |
+
{
|
3085 |
+
"epoch": 0.9107883817427386,
|
3086 |
+
"grad_norm": 2.4571108685084107,
|
3087 |
+
"learning_rate": 2.4048340087513665e-07,
|
3088 |
+
"loss": 0.336,
|
3089 |
+
"step": 2195
|
3090 |
+
},
|
3091 |
+
{
|
3092 |
+
"epoch": 0.9128630705394191,
|
3093 |
+
"grad_norm": 2.3387656990162395,
|
3094 |
+
"learning_rate": 2.295135710009755e-07,
|
3095 |
+
"loss": 0.3414,
|
3096 |
+
"step": 2200
|
3097 |
+
},
|
3098 |
+
{
|
3099 |
+
"epoch": 0.9149377593360996,
|
3100 |
+
"grad_norm": 2.3291759526847113,
|
3101 |
+
"learning_rate": 2.1879393773429903e-07,
|
3102 |
+
"loss": 0.3442,
|
3103 |
+
"step": 2205
|
3104 |
+
},
|
3105 |
+
{
|
3106 |
+
"epoch": 0.91701244813278,
|
3107 |
+
"grad_norm": 2.365416524223355,
|
3108 |
+
"learning_rate": 2.0832506328525237e-07,
|
3109 |
+
"loss": 0.3317,
|
3110 |
+
"step": 2210
|
3111 |
+
},
|
3112 |
+
{
|
3113 |
+
"epoch": 0.9190871369294605,
|
3114 |
+
"grad_norm": 2.384106710657764,
|
3115 |
+
"learning_rate": 1.9810749671249353e-07,
|
3116 |
+
"loss": 0.347,
|
3117 |
+
"step": 2215
|
3118 |
+
},
|
3119 |
+
{
|
3120 |
+
"epoch": 0.921161825726141,
|
3121 |
+
"grad_norm": 2.380728030255305,
|
3122 |
+
"learning_rate": 1.8814177389439304e-07,
|
3123 |
+
"loss": 0.3466,
|
3124 |
+
"step": 2220
|
3125 |
+
},
|
3126 |
+
{
|
3127 |
+
"epoch": 0.9232365145228216,
|
3128 |
+
"grad_norm": 2.297466874753593,
|
3129 |
+
"learning_rate": 1.784284175009282e-07,
|
3130 |
+
"loss": 0.3368,
|
3131 |
+
"step": 2225
|
3132 |
+
},
|
3133 |
+
{
|
3134 |
+
"epoch": 0.9253112033195021,
|
3135 |
+
"grad_norm": 2.4935459102531685,
|
3136 |
+
"learning_rate": 1.68967936966275e-07,
|
3137 |
+
"loss": 0.3404,
|
3138 |
+
"step": 2230
|
3139 |
+
},
|
3140 |
+
{
|
3141 |
+
"epoch": 0.9273858921161826,
|
3142 |
+
"grad_norm": 2.36033858236717,
|
3143 |
+
"learning_rate": 1.5976082846208552e-07,
|
3144 |
+
"loss": 0.3499,
|
3145 |
+
"step": 2235
|
3146 |
+
},
|
3147 |
+
{
|
3148 |
+
"epoch": 0.9294605809128631,
|
3149 |
+
"grad_norm": 2.4730950951949353,
|
3150 |
+
"learning_rate": 1.5080757487146823e-07,
|
3151 |
+
"loss": 0.3327,
|
3152 |
+
"step": 2240
|
3153 |
+
},
|
3154 |
+
{
|
3155 |
+
"epoch": 0.9315352697095436,
|
3156 |
+
"grad_norm": 2.4704731647631255,
|
3157 |
+
"learning_rate": 1.4210864576365891e-07,
|
3158 |
+
"loss": 0.338,
|
3159 |
+
"step": 2245
|
3160 |
+
},
|
3161 |
+
{
|
3162 |
+
"epoch": 0.9336099585062241,
|
3163 |
+
"grad_norm": 2.3890389217669723,
|
3164 |
+
"learning_rate": 1.3366449736939845e-07,
|
3165 |
+
"loss": 0.3375,
|
3166 |
+
"step": 2250
|
3167 |
+
},
|
3168 |
+
{
|
3169 |
+
"epoch": 0.9356846473029046,
|
3170 |
+
"grad_norm": 2.6765870984768663,
|
3171 |
+
"learning_rate": 1.254755725570006e-07,
|
3172 |
+
"loss": 0.3504,
|
3173 |
+
"step": 2255
|
3174 |
+
},
|
3175 |
+
{
|
3176 |
+
"epoch": 0.9377593360995851,
|
3177 |
+
"grad_norm": 2.3147038405622333,
|
3178 |
+
"learning_rate": 1.1754230080912588e-07,
|
3179 |
+
"loss": 0.3267,
|
3180 |
+
"step": 2260
|
3181 |
+
},
|
3182 |
+
{
|
3183 |
+
"epoch": 0.9398340248962656,
|
3184 |
+
"grad_norm": 2.559635037749535,
|
3185 |
+
"learning_rate": 1.0986509820025915e-07,
|
3186 |
+
"loss": 0.3502,
|
3187 |
+
"step": 2265
|
3188 |
+
},
|
3189 |
+
{
|
3190 |
+
"epoch": 0.941908713692946,
|
3191 |
+
"grad_norm": 2.378006403702019,
|
3192 |
+
"learning_rate": 1.0244436737488373e-07,
|
3193 |
+
"loss": 0.3496,
|
3194 |
+
"step": 2270
|
3195 |
+
},
|
3196 |
+
{
|
3197 |
+
"epoch": 0.9439834024896265,
|
3198 |
+
"grad_norm": 2.4406461268285216,
|
3199 |
+
"learning_rate": 9.528049752636714e-08,
|
3200 |
+
"loss": 0.3292,
|
3201 |
+
"step": 2275
|
3202 |
+
},
|
3203 |
+
{
|
3204 |
+
"epoch": 0.946058091286307,
|
3205 |
+
"grad_norm": 2.3994648318804632,
|
3206 |
+
"learning_rate": 8.837386437654861e-08,
|
3207 |
+
"loss": 0.3369,
|
3208 |
+
"step": 2280
|
3209 |
+
},
|
3210 |
+
{
|
3211 |
+
"epoch": 0.9481327800829875,
|
3212 |
+
"grad_norm": 2.3387687484550583,
|
3213 |
+
"learning_rate": 8.172483015603139e-08,
|
3214 |
+
"loss": 0.3445,
|
3215 |
+
"step": 2285
|
3216 |
+
},
|
3217 |
+
{
|
3218 |
+
"epoch": 0.950207468879668,
|
3219 |
+
"grad_norm": 2.3329266437152243,
|
3220 |
+
"learning_rate": 7.533374358518974e-08,
|
3221 |
+
"loss": 0.3374,
|
3222 |
+
"step": 2290
|
3223 |
+
},
|
3224 |
+
{
|
3225 |
+
"epoch": 0.9522821576763485,
|
3226 |
+
"grad_norm": 2.344823680655971,
|
3227 |
+
"learning_rate": 6.920093985587351e-08,
|
3228 |
+
"loss": 0.3369,
|
3229 |
+
"step": 2295
|
3230 |
+
},
|
3231 |
+
{
|
3232 |
+
"epoch": 0.9543568464730291,
|
3233 |
+
"grad_norm": 2.297823156508915,
|
3234 |
+
"learning_rate": 6.332674061383492e-08,
|
3235 |
+
"loss": 0.3387,
|
3236 |
+
"step": 2300
|
3237 |
+
},
|
3238 |
+
{
|
3239 |
+
"epoch": 0.9564315352697096,
|
3240 |
+
"grad_norm": 2.436614135753504,
|
3241 |
+
"learning_rate": 5.7711453941852736e-08,
|
3242 |
+
"loss": 0.334,
|
3243 |
+
"step": 2305
|
3244 |
+
},
|
3245 |
+
{
|
3246 |
+
"epoch": 0.9585062240663901,
|
3247 |
+
"grad_norm": 2.2863118140565564,
|
3248 |
+
"learning_rate": 5.23553743435784e-08,
|
3249 |
+
"loss": 0.3284,
|
3250 |
+
"step": 2310
|
3251 |
+
},
|
3252 |
+
{
|
3253 |
+
"epoch": 0.9605809128630706,
|
3254 |
+
"grad_norm": 2.1820719536802504,
|
3255 |
+
"learning_rate": 4.7258782728092366e-08,
|
3256 |
+
"loss": 0.3293,
|
3257 |
+
"step": 2315
|
3258 |
+
},
|
3259 |
+
{
|
3260 |
+
"epoch": 0.9626556016597511,
|
3261 |
+
"grad_norm": 2.333916190818195,
|
3262 |
+
"learning_rate": 4.2421946395164174e-08,
|
3263 |
+
"loss": 0.33,
|
3264 |
+
"step": 2320
|
3265 |
+
},
|
3266 |
+
{
|
3267 |
+
"epoch": 0.9647302904564315,
|
3268 |
+
"grad_norm": 2.4479781364843096,
|
3269 |
+
"learning_rate": 3.784511902124033e-08,
|
3270 |
+
"loss": 0.3491,
|
3271 |
+
"step": 2325
|
3272 |
+
},
|
3273 |
+
{
|
3274 |
+
"epoch": 0.966804979253112,
|
3275 |
+
"grad_norm": 2.4249579257495566,
|
3276 |
+
"learning_rate": 3.352854064613553e-08,
|
3277 |
+
"loss": 0.345,
|
3278 |
+
"step": 2330
|
3279 |
+
},
|
3280 |
+
{
|
3281 |
+
"epoch": 0.9688796680497925,
|
3282 |
+
"grad_norm": 2.2853865711169785,
|
3283 |
+
"learning_rate": 2.9472437660446605e-08,
|
3284 |
+
"loss": 0.3358,
|
3285 |
+
"step": 2335
|
3286 |
+
},
|
3287 |
+
{
|
3288 |
+
"epoch": 0.970954356846473,
|
3289 |
+
"grad_norm": 2.4792048595371647,
|
3290 |
+
"learning_rate": 2.56770227936759e-08,
|
3291 |
+
"loss": 0.3365,
|
3292 |
+
"step": 2340
|
3293 |
+
},
|
3294 |
+
{
|
3295 |
+
"epoch": 0.9730290456431535,
|
3296 |
+
"grad_norm": 2.38571031272437,
|
3297 |
+
"learning_rate": 2.2142495103075757e-08,
|
3298 |
+
"loss": 0.3426,
|
3299 |
+
"step": 2345
|
3300 |
+
},
|
3301 |
+
{
|
3302 |
+
"epoch": 0.975103734439834,
|
3303 |
+
"grad_norm": 2.3533072019330095,
|
3304 |
+
"learning_rate": 1.8869039963210766e-08,
|
3305 |
+
"loss": 0.3384,
|
3306 |
+
"step": 2350
|
3307 |
+
},
|
3308 |
+
{
|
3309 |
+
"epoch": 0.9771784232365145,
|
3310 |
+
"grad_norm": 2.42964006589875,
|
3311 |
+
"learning_rate": 1.585682905623054e-08,
|
3312 |
+
"loss": 0.3406,
|
3313 |
+
"step": 2355
|
3314 |
+
},
|
3315 |
+
{
|
3316 |
+
"epoch": 0.979253112033195,
|
3317 |
+
"grad_norm": 2.2589268313436968,
|
3318 |
+
"learning_rate": 1.3106020362869675e-08,
|
3319 |
+
"loss": 0.3379,
|
3320 |
+
"step": 2360
|
3321 |
+
},
|
3322 |
+
{
|
3323 |
+
"epoch": 0.9813278008298755,
|
3324 |
+
"grad_norm": 2.4389590802442203,
|
3325 |
+
"learning_rate": 1.0616758154161633e-08,
|
3326 |
+
"loss": 0.342,
|
3327 |
+
"step": 2365
|
3328 |
+
},
|
3329 |
+
{
|
3330 |
+
"epoch": 0.983402489626556,
|
3331 |
+
"grad_norm": 2.387697736294216,
|
3332 |
+
"learning_rate": 8.38917298386921e-09,
|
3333 |
+
"loss": 0.3414,
|
3334 |
+
"step": 2370
|
3335 |
+
},
|
3336 |
+
{
|
3337 |
+
"epoch": 0.9854771784232366,
|
3338 |
+
"grad_norm": 2.2923054120317836,
|
3339 |
+
"learning_rate": 6.423381681641694e-09,
|
3340 |
+
"loss": 0.3416,
|
3341 |
+
"step": 2375
|
3342 |
+
},
|
3343 |
+
{
|
3344 |
+
"epoch": 0.9875518672199171,
|
3345 |
+
"grad_norm": 2.341070250308901,
|
3346 |
+
"learning_rate": 4.719487346884211e-09,
|
3347 |
+
"loss": 0.3353,
|
3348 |
+
"step": 2380
|
3349 |
+
},
|
3350 |
+
{
|
3351 |
+
"epoch": 0.9896265560165975,
|
3352 |
+
"grad_norm": 2.390677320345825,
|
3353 |
+
"learning_rate": 3.277579343352599e-09,
|
3354 |
+
"loss": 0.3273,
|
3355 |
+
"step": 2385
|
3356 |
+
},
|
3357 |
+
{
|
3358 |
+
"epoch": 0.991701244813278,
|
3359 |
+
"grad_norm": 2.3301588985719905,
|
3360 |
+
"learning_rate": 2.097733294464943e-09,
|
3361 |
+
"loss": 0.355,
|
3362 |
+
"step": 2390
|
3363 |
+
},
|
3364 |
+
{
|
3365 |
+
"epoch": 0.9937759336099585,
|
3366 |
+
"grad_norm": 2.423326896617218,
|
3367 |
+
"learning_rate": 1.1800110793358521e-09,
|
3368 |
+
"loss": 0.3402,
|
3369 |
+
"step": 2395
|
3370 |
+
},
|
3371 |
+
{
|
3372 |
+
"epoch": 0.995850622406639,
|
3373 |
+
"grad_norm": 2.3567177848437986,
|
3374 |
+
"learning_rate": 5.244608295323916e-10,
|
3375 |
+
"loss": 0.3361,
|
3376 |
+
"step": 2400
|
3377 |
+
},
|
3378 |
+
{
|
3379 |
+
"epoch": 0.9979253112033195,
|
3380 |
+
"grad_norm": 2.4643918613730165,
|
3381 |
+
"learning_rate": 1.3111692654776963e-10,
|
3382 |
+
"loss": 0.3175,
|
3383 |
+
"step": 2405
|
3384 |
+
},
|
3385 |
+
{
|
3386 |
+
"epoch": 1.0,
|
3387 |
+
"grad_norm": 2.3790989221416656,
|
3388 |
+
"learning_rate": 0.0,
|
3389 |
+
"loss": 0.3387,
|
3390 |
+
"step": 2410
|
3391 |
+
},
|
3392 |
+
{
|
3393 |
+
"epoch": 1.0,
|
3394 |
+
"eval_runtime": 3.4735,
|
3395 |
+
"eval_samples_per_second": 2.879,
|
3396 |
+
"eval_steps_per_second": 0.864,
|
3397 |
+
"step": 2410
|
3398 |
+
},
|
3399 |
+
{
|
3400 |
+
"epoch": 1.0,
|
3401 |
+
"step": 2410,
|
3402 |
+
"total_flos": 252302485094400.0,
|
3403 |
+
"train_loss": 0.5277333631555075,
|
3404 |
+
"train_runtime": 21339.31,
|
3405 |
+
"train_samples_per_second": 1.807,
|
3406 |
+
"train_steps_per_second": 0.113
|
3407 |
+
}
|
3408 |
+
],
|
3409 |
+
"logging_steps": 5,
|
3410 |
+
"max_steps": 2410,
|
3411 |
+
"num_input_tokens_seen": 0,
|
3412 |
+
"num_train_epochs": 1,
|
3413 |
+
"save_steps": 100,
|
3414 |
+
"stateful_callbacks": {
|
3415 |
+
"TrainerControl": {
|
3416 |
+
"args": {
|
3417 |
+
"should_epoch_stop": false,
|
3418 |
+
"should_evaluate": false,
|
3419 |
+
"should_log": false,
|
3420 |
+
"should_save": true,
|
3421 |
+
"should_training_stop": true
|
3422 |
+
},
|
3423 |
+
"attributes": {}
|
3424 |
+
}
|
3425 |
+
},
|
3426 |
+
"total_flos": 252302485094400.0,
|
3427 |
+
"train_batch_size": 4,
|
3428 |
+
"trial_name": null,
|
3429 |
+
"trial_params": null
|
3430 |
+
}
|