Model save
Browse files- README.md +70 -0
- adapter_model.safetensors +1 -1
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +2682 -0
README.md
ADDED
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: llama3
|
3 |
+
library_name: peft
|
4 |
+
tags:
|
5 |
+
- trl
|
6 |
+
- sft
|
7 |
+
- generated_from_trainer
|
8 |
+
base_model: meta-llama/Meta-Llama-3-8B-Instruct
|
9 |
+
datasets:
|
10 |
+
- generator
|
11 |
+
model-index:
|
12 |
+
- name: Meta-Llama-3-8B-Instruct
|
13 |
+
results: []
|
14 |
+
---
|
15 |
+
|
16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
+
should probably proofread and complete it, then remove this comment. -->
|
18 |
+
|
19 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/statking/huggingface/runs/1d81qcom)
|
20 |
+
# Meta-Llama-3-8B-Instruct
|
21 |
+
|
22 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the generator dataset.
|
23 |
+
It achieves the following results on the evaluation set:
|
24 |
+
- Loss: 1.1189
|
25 |
+
|
26 |
+
## Model description
|
27 |
+
|
28 |
+
More information needed
|
29 |
+
|
30 |
+
## Intended uses & limitations
|
31 |
+
|
32 |
+
More information needed
|
33 |
+
|
34 |
+
## Training and evaluation data
|
35 |
+
|
36 |
+
More information needed
|
37 |
+
|
38 |
+
## Training procedure
|
39 |
+
|
40 |
+
### Training hyperparameters
|
41 |
+
|
42 |
+
The following hyperparameters were used during training:
|
43 |
+
- learning_rate: 0.0002
|
44 |
+
- train_batch_size: 8
|
45 |
+
- eval_batch_size: 8
|
46 |
+
- seed: 42
|
47 |
+
- distributed_type: multi-GPU
|
48 |
+
- num_devices: 2
|
49 |
+
- gradient_accumulation_steps: 4
|
50 |
+
- total_train_batch_size: 64
|
51 |
+
- total_eval_batch_size: 16
|
52 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
53 |
+
- lr_scheduler_type: cosine
|
54 |
+
- lr_scheduler_warmup_ratio: 0.1
|
55 |
+
- num_epochs: 1
|
56 |
+
|
57 |
+
### Training results
|
58 |
+
|
59 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
60 |
+
|:-------------:|:------:|:----:|:---------------:|
|
61 |
+
| 1.1206 | 0.9997 | 1877 | 1.1189 |
|
62 |
+
|
63 |
+
|
64 |
+
### Framework versions
|
65 |
+
|
66 |
+
- PEFT 0.10.0
|
67 |
+
- Transformers 4.41.0.dev0
|
68 |
+
- Pytorch 2.3.0+cu121
|
69 |
+
- Datasets 2.19.1
|
70 |
+
- Tokenizers 0.19.1
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 2185327392
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:38de2d88d470a960573d4f858dc17b44a1a293b92f99ddf486879bf39164c766
|
3 |
size 2185327392
|
all_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 0.9997336884154461,
|
3 |
+
"total_flos": 2.979798729936077e+16,
|
4 |
+
"train_loss": 1.1429596533467938,
|
5 |
+
"train_runtime": 55739.5179,
|
6 |
+
"train_samples": 207864,
|
7 |
+
"train_samples_per_second": 2.156,
|
8 |
+
"train_steps_per_second": 0.034
|
9 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 0.9997336884154461,
|
3 |
+
"total_flos": 2.979798729936077e+16,
|
4 |
+
"train_loss": 1.1429596533467938,
|
5 |
+
"train_runtime": 55739.5179,
|
6 |
+
"train_samples": 207864,
|
7 |
+
"train_samples_per_second": 2.156,
|
8 |
+
"train_steps_per_second": 0.034
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,2682 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.9997336884154461,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1877,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.0005326231691078562,
|
13 |
+
"grad_norm": 0.5528136747380069,
|
14 |
+
"learning_rate": 1.0638297872340427e-06,
|
15 |
+
"loss": 1.7189,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.002663115845539281,
|
20 |
+
"grad_norm": 0.5731888996610293,
|
21 |
+
"learning_rate": 5.319148936170213e-06,
|
22 |
+
"loss": 1.7582,
|
23 |
+
"step": 5
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.005326231691078562,
|
27 |
+
"grad_norm": 0.5597088455341374,
|
28 |
+
"learning_rate": 1.0638297872340426e-05,
|
29 |
+
"loss": 1.7331,
|
30 |
+
"step": 10
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.007989347536617843,
|
34 |
+
"grad_norm": 0.6373028746703387,
|
35 |
+
"learning_rate": 1.595744680851064e-05,
|
36 |
+
"loss": 1.7083,
|
37 |
+
"step": 15
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.010652463382157125,
|
41 |
+
"grad_norm": 0.6190956100247592,
|
42 |
+
"learning_rate": 2.1276595744680852e-05,
|
43 |
+
"loss": 1.6453,
|
44 |
+
"step": 20
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.013315579227696404,
|
48 |
+
"grad_norm": 0.5346124338518666,
|
49 |
+
"learning_rate": 2.6595744680851064e-05,
|
50 |
+
"loss": 1.5715,
|
51 |
+
"step": 25
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.015978695073235686,
|
55 |
+
"grad_norm": 0.25993843509864667,
|
56 |
+
"learning_rate": 3.191489361702128e-05,
|
57 |
+
"loss": 1.4359,
|
58 |
+
"step": 30
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.018641810918774968,
|
62 |
+
"grad_norm": 0.45126098872878045,
|
63 |
+
"learning_rate": 3.723404255319149e-05,
|
64 |
+
"loss": 1.4122,
|
65 |
+
"step": 35
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.02130492676431425,
|
69 |
+
"grad_norm": 0.16268004246261306,
|
70 |
+
"learning_rate": 4.2553191489361704e-05,
|
71 |
+
"loss": 1.3378,
|
72 |
+
"step": 40
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.023968042609853527,
|
76 |
+
"grad_norm": 0.16755798993478008,
|
77 |
+
"learning_rate": 4.787234042553192e-05,
|
78 |
+
"loss": 1.3315,
|
79 |
+
"step": 45
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.02663115845539281,
|
83 |
+
"grad_norm": 0.14666977328343594,
|
84 |
+
"learning_rate": 5.319148936170213e-05,
|
85 |
+
"loss": 1.3259,
|
86 |
+
"step": 50
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.02929427430093209,
|
90 |
+
"grad_norm": 0.10629659674680719,
|
91 |
+
"learning_rate": 5.851063829787234e-05,
|
92 |
+
"loss": 1.288,
|
93 |
+
"step": 55
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.03195739014647137,
|
97 |
+
"grad_norm": 0.11263809733941188,
|
98 |
+
"learning_rate": 6.382978723404256e-05,
|
99 |
+
"loss": 1.2763,
|
100 |
+
"step": 60
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.03462050599201065,
|
104 |
+
"grad_norm": 0.14445654795383106,
|
105 |
+
"learning_rate": 6.914893617021277e-05,
|
106 |
+
"loss": 1.256,
|
107 |
+
"step": 65
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.037283621837549935,
|
111 |
+
"grad_norm": 0.07679775841947081,
|
112 |
+
"learning_rate": 7.446808510638298e-05,
|
113 |
+
"loss": 1.2481,
|
114 |
+
"step": 70
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.03994673768308921,
|
118 |
+
"grad_norm": 0.07097172061630208,
|
119 |
+
"learning_rate": 7.978723404255319e-05,
|
120 |
+
"loss": 1.2417,
|
121 |
+
"step": 75
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.0426098535286285,
|
125 |
+
"grad_norm": 0.07545648577598915,
|
126 |
+
"learning_rate": 8.510638297872341e-05,
|
127 |
+
"loss": 1.2048,
|
128 |
+
"step": 80
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.045272969374167776,
|
132 |
+
"grad_norm": 0.06657701317469632,
|
133 |
+
"learning_rate": 9.042553191489363e-05,
|
134 |
+
"loss": 1.2138,
|
135 |
+
"step": 85
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.047936085219707054,
|
139 |
+
"grad_norm": 0.08276972517588223,
|
140 |
+
"learning_rate": 9.574468085106384e-05,
|
141 |
+
"loss": 1.2267,
|
142 |
+
"step": 90
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.05059920106524634,
|
146 |
+
"grad_norm": 0.14086227463755532,
|
147 |
+
"learning_rate": 0.00010106382978723406,
|
148 |
+
"loss": 1.2185,
|
149 |
+
"step": 95
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.05326231691078562,
|
153 |
+
"grad_norm": 0.08222057468849275,
|
154 |
+
"learning_rate": 0.00010638297872340425,
|
155 |
+
"loss": 1.2021,
|
156 |
+
"step": 100
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.0559254327563249,
|
160 |
+
"grad_norm": 0.06741247738810993,
|
161 |
+
"learning_rate": 0.00011170212765957446,
|
162 |
+
"loss": 1.1957,
|
163 |
+
"step": 105
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.05858854860186418,
|
167 |
+
"grad_norm": 0.07692927491859038,
|
168 |
+
"learning_rate": 0.00011702127659574468,
|
169 |
+
"loss": 1.1901,
|
170 |
+
"step": 110
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.06125166444740346,
|
174 |
+
"grad_norm": 0.08953054401601632,
|
175 |
+
"learning_rate": 0.0001223404255319149,
|
176 |
+
"loss": 1.2002,
|
177 |
+
"step": 115
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.06391478029294274,
|
181 |
+
"grad_norm": 0.066045987418387,
|
182 |
+
"learning_rate": 0.00012765957446808513,
|
183 |
+
"loss": 1.2086,
|
184 |
+
"step": 120
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.06657789613848203,
|
188 |
+
"grad_norm": 0.06476207146640194,
|
189 |
+
"learning_rate": 0.00013297872340425532,
|
190 |
+
"loss": 1.215,
|
191 |
+
"step": 125
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.0692410119840213,
|
195 |
+
"grad_norm": 0.07334327374343644,
|
196 |
+
"learning_rate": 0.00013829787234042554,
|
197 |
+
"loss": 1.181,
|
198 |
+
"step": 130
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.07190412782956059,
|
202 |
+
"grad_norm": 0.071815699820189,
|
203 |
+
"learning_rate": 0.00014361702127659576,
|
204 |
+
"loss": 1.1795,
|
205 |
+
"step": 135
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.07456724367509987,
|
209 |
+
"grad_norm": 0.08268574123602224,
|
210 |
+
"learning_rate": 0.00014893617021276596,
|
211 |
+
"loss": 1.1892,
|
212 |
+
"step": 140
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.07723035952063914,
|
216 |
+
"grad_norm": 0.07585606469879155,
|
217 |
+
"learning_rate": 0.00015425531914893618,
|
218 |
+
"loss": 1.173,
|
219 |
+
"step": 145
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.07989347536617843,
|
223 |
+
"grad_norm": 0.06900850276332868,
|
224 |
+
"learning_rate": 0.00015957446808510637,
|
225 |
+
"loss": 1.1889,
|
226 |
+
"step": 150
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.08255659121171771,
|
230 |
+
"grad_norm": 0.07702070923317432,
|
231 |
+
"learning_rate": 0.00016489361702127662,
|
232 |
+
"loss": 1.1705,
|
233 |
+
"step": 155
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.085219707057257,
|
237 |
+
"grad_norm": 0.0724505634260966,
|
238 |
+
"learning_rate": 0.00017021276595744682,
|
239 |
+
"loss": 1.1801,
|
240 |
+
"step": 160
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.08788282290279627,
|
244 |
+
"grad_norm": 0.07762044850846143,
|
245 |
+
"learning_rate": 0.000175531914893617,
|
246 |
+
"loss": 1.1863,
|
247 |
+
"step": 165
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.09054593874833555,
|
251 |
+
"grad_norm": 0.0825988729050522,
|
252 |
+
"learning_rate": 0.00018085106382978726,
|
253 |
+
"loss": 1.1687,
|
254 |
+
"step": 170
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.09320905459387484,
|
258 |
+
"grad_norm": 0.08362625552402488,
|
259 |
+
"learning_rate": 0.00018617021276595746,
|
260 |
+
"loss": 1.1875,
|
261 |
+
"step": 175
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.09587217043941411,
|
265 |
+
"grad_norm": 0.07968629276225715,
|
266 |
+
"learning_rate": 0.00019148936170212768,
|
267 |
+
"loss": 1.1629,
|
268 |
+
"step": 180
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.0985352862849534,
|
272 |
+
"grad_norm": 0.09010747938874886,
|
273 |
+
"learning_rate": 0.00019680851063829787,
|
274 |
+
"loss": 1.1682,
|
275 |
+
"step": 185
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.10119840213049268,
|
279 |
+
"grad_norm": 0.10276104499280464,
|
280 |
+
"learning_rate": 0.00019999930805760402,
|
281 |
+
"loss": 1.1618,
|
282 |
+
"step": 190
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.10386151797603196,
|
286 |
+
"grad_norm": 0.07684956533429005,
|
287 |
+
"learning_rate": 0.00019999152381561955,
|
288 |
+
"loss": 1.1902,
|
289 |
+
"step": 195
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.10652463382157124,
|
293 |
+
"grad_norm": 0.09365634557072464,
|
294 |
+
"learning_rate": 0.0001999750910791767,
|
295 |
+
"loss": 1.1673,
|
296 |
+
"step": 200
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.10918774966711052,
|
300 |
+
"grad_norm": 0.08320439951882774,
|
301 |
+
"learning_rate": 0.00019995001126958025,
|
302 |
+
"loss": 1.1845,
|
303 |
+
"step": 205
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.1118508655126498,
|
307 |
+
"grad_norm": 0.08595161594886752,
|
308 |
+
"learning_rate": 0.00019991628655604003,
|
309 |
+
"loss": 1.1444,
|
310 |
+
"step": 210
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.11451398135818908,
|
314 |
+
"grad_norm": 0.07678856693871118,
|
315 |
+
"learning_rate": 0.00019987391985548328,
|
316 |
+
"loss": 1.1724,
|
317 |
+
"step": 215
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.11717709720372836,
|
321 |
+
"grad_norm": 0.07433277571881601,
|
322 |
+
"learning_rate": 0.0001998229148323023,
|
323 |
+
"loss": 1.1469,
|
324 |
+
"step": 220
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.11984021304926765,
|
328 |
+
"grad_norm": 0.11001095605643386,
|
329 |
+
"learning_rate": 0.00019976327589803767,
|
330 |
+
"loss": 1.1383,
|
331 |
+
"step": 225
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.12250332889480692,
|
335 |
+
"grad_norm": 0.0784178760835021,
|
336 |
+
"learning_rate": 0.0001996950082109965,
|
337 |
+
"loss": 1.1818,
|
338 |
+
"step": 230
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.12516644474034622,
|
342 |
+
"grad_norm": 0.08047194112395492,
|
343 |
+
"learning_rate": 0.00019961811767580648,
|
344 |
+
"loss": 1.1445,
|
345 |
+
"step": 235
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.1278295605858855,
|
349 |
+
"grad_norm": 0.0670667235371544,
|
350 |
+
"learning_rate": 0.0001995326109429049,
|
351 |
+
"loss": 1.1741,
|
352 |
+
"step": 240
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.13049267643142476,
|
356 |
+
"grad_norm": 0.07072589446768075,
|
357 |
+
"learning_rate": 0.00019943849540796375,
|
358 |
+
"loss": 1.157,
|
359 |
+
"step": 245
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.13315579227696406,
|
363 |
+
"grad_norm": 0.07466892570841129,
|
364 |
+
"learning_rate": 0.0001993357792112498,
|
365 |
+
"loss": 1.125,
|
366 |
+
"step": 250
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.13581890812250333,
|
370 |
+
"grad_norm": 0.07302104613788317,
|
371 |
+
"learning_rate": 0.0001992244712369207,
|
372 |
+
"loss": 1.1615,
|
373 |
+
"step": 255
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.1384820239680426,
|
377 |
+
"grad_norm": 0.07211635352591637,
|
378 |
+
"learning_rate": 0.00019910458111225646,
|
379 |
+
"loss": 1.1441,
|
380 |
+
"step": 260
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.1411451398135819,
|
384 |
+
"grad_norm": 0.07103357444221702,
|
385 |
+
"learning_rate": 0.00019897611920682677,
|
386 |
+
"loss": 1.1493,
|
387 |
+
"step": 265
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.14380825565912117,
|
391 |
+
"grad_norm": 0.0698227187710226,
|
392 |
+
"learning_rate": 0.00019883909663159424,
|
393 |
+
"loss": 1.1568,
|
394 |
+
"step": 270
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.14647137150466044,
|
398 |
+
"grad_norm": 0.07137557168765225,
|
399 |
+
"learning_rate": 0.0001986935252379532,
|
400 |
+
"loss": 1.171,
|
401 |
+
"step": 275
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.14913448735019974,
|
405 |
+
"grad_norm": 0.07605080544337586,
|
406 |
+
"learning_rate": 0.00019853941761670483,
|
407 |
+
"loss": 1.1623,
|
408 |
+
"step": 280
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.151797603195739,
|
412 |
+
"grad_norm": 0.09532848101140429,
|
413 |
+
"learning_rate": 0.00019837678709696798,
|
414 |
+
"loss": 1.1888,
|
415 |
+
"step": 285
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.15446071904127828,
|
419 |
+
"grad_norm": 0.07485256895909924,
|
420 |
+
"learning_rate": 0.00019820564774502644,
|
421 |
+
"loss": 1.1483,
|
422 |
+
"step": 290
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.15712383488681758,
|
426 |
+
"grad_norm": 0.07483378156117482,
|
427 |
+
"learning_rate": 0.0001980260143631122,
|
428 |
+
"loss": 1.1375,
|
429 |
+
"step": 295
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.15978695073235685,
|
433 |
+
"grad_norm": 0.07954155407009747,
|
434 |
+
"learning_rate": 0.00019783790248812533,
|
435 |
+
"loss": 1.1696,
|
436 |
+
"step": 300
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.16245006657789615,
|
440 |
+
"grad_norm": 0.08502452471103343,
|
441 |
+
"learning_rate": 0.00019764132839029,
|
442 |
+
"loss": 1.168,
|
443 |
+
"step": 305
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.16511318242343542,
|
447 |
+
"grad_norm": 0.08384910068571033,
|
448 |
+
"learning_rate": 0.00019743630907174725,
|
449 |
+
"loss": 1.1659,
|
450 |
+
"step": 310
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.1677762982689747,
|
454 |
+
"grad_norm": 0.06905463640642404,
|
455 |
+
"learning_rate": 0.0001972228622650846,
|
456 |
+
"loss": 1.1612,
|
457 |
+
"step": 315
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.170439414114514,
|
461 |
+
"grad_norm": 0.19257912301232658,
|
462 |
+
"learning_rate": 0.0001970010064318021,
|
463 |
+
"loss": 1.1517,
|
464 |
+
"step": 320
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.17310252996005326,
|
468 |
+
"grad_norm": 0.0793114626498931,
|
469 |
+
"learning_rate": 0.00019677076076071566,
|
470 |
+
"loss": 1.1385,
|
471 |
+
"step": 325
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.17576564580559254,
|
475 |
+
"grad_norm": 0.07393026070000318,
|
476 |
+
"learning_rate": 0.00019653214516629735,
|
477 |
+
"loss": 1.1426,
|
478 |
+
"step": 330
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.17842876165113183,
|
482 |
+
"grad_norm": 0.08179432509124362,
|
483 |
+
"learning_rate": 0.00019628518028695307,
|
484 |
+
"loss": 1.1104,
|
485 |
+
"step": 335
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.1810918774966711,
|
489 |
+
"grad_norm": 0.09735291608528279,
|
490 |
+
"learning_rate": 0.00019602988748323717,
|
491 |
+
"loss": 1.1563,
|
492 |
+
"step": 340
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.18375499334221038,
|
496 |
+
"grad_norm": 0.06743724715009518,
|
497 |
+
"learning_rate": 0.00019576628883600535,
|
498 |
+
"loss": 1.1406,
|
499 |
+
"step": 345
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.18641810918774968,
|
503 |
+
"grad_norm": 0.075326384879952,
|
504 |
+
"learning_rate": 0.00019549440714450444,
|
505 |
+
"loss": 1.1572,
|
506 |
+
"step": 350
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.18908122503328895,
|
510 |
+
"grad_norm": 0.07438689728031705,
|
511 |
+
"learning_rate": 0.00019521426592440072,
|
512 |
+
"loss": 1.1479,
|
513 |
+
"step": 355
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.19174434087882822,
|
517 |
+
"grad_norm": 0.07277611336304127,
|
518 |
+
"learning_rate": 0.00019492588940574586,
|
519 |
+
"loss": 1.1549,
|
520 |
+
"step": 360
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.19440745672436752,
|
524 |
+
"grad_norm": 0.0695324135241875,
|
525 |
+
"learning_rate": 0.0001946293025308813,
|
526 |
+
"loss": 1.1435,
|
527 |
+
"step": 365
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.1970705725699068,
|
531 |
+
"grad_norm": 0.06685927618032904,
|
532 |
+
"learning_rate": 0.00019432453095228076,
|
533 |
+
"loss": 1.1641,
|
534 |
+
"step": 370
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.19973368841544606,
|
538 |
+
"grad_norm": 0.0680367135740568,
|
539 |
+
"learning_rate": 0.00019401160103033174,
|
540 |
+
"loss": 1.1261,
|
541 |
+
"step": 375
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.20239680426098536,
|
545 |
+
"grad_norm": 0.08027336453756874,
|
546 |
+
"learning_rate": 0.00019369053983105532,
|
547 |
+
"loss": 1.1368,
|
548 |
+
"step": 380
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.20505992010652463,
|
552 |
+
"grad_norm": 0.0707161713953054,
|
553 |
+
"learning_rate": 0.00019336137512376532,
|
554 |
+
"loss": 1.1588,
|
555 |
+
"step": 385
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.20772303595206393,
|
559 |
+
"grad_norm": 0.07189527593634382,
|
560 |
+
"learning_rate": 0.00019302413537866642,
|
561 |
+
"loss": 1.1552,
|
562 |
+
"step": 390
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.2103861517976032,
|
566 |
+
"grad_norm": 0.0716934364253126,
|
567 |
+
"learning_rate": 0.0001926788497643916,
|
568 |
+
"loss": 1.1577,
|
569 |
+
"step": 395
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.21304926764314247,
|
573 |
+
"grad_norm": 0.065943892133018,
|
574 |
+
"learning_rate": 0.00019232554814547953,
|
575 |
+
"loss": 1.1203,
|
576 |
+
"step": 400
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.21571238348868177,
|
580 |
+
"grad_norm": 0.07352621386091099,
|
581 |
+
"learning_rate": 0.00019196426107979128,
|
582 |
+
"loss": 1.1266,
|
583 |
+
"step": 405
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.21837549933422104,
|
587 |
+
"grad_norm": 0.07441803674470306,
|
588 |
+
"learning_rate": 0.00019159501981586737,
|
589 |
+
"loss": 1.1432,
|
590 |
+
"step": 410
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.2210386151797603,
|
594 |
+
"grad_norm": 0.07291702193187057,
|
595 |
+
"learning_rate": 0.00019121785629022501,
|
596 |
+
"loss": 1.1344,
|
597 |
+
"step": 415
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.2237017310252996,
|
601 |
+
"grad_norm": 0.07094925179230635,
|
602 |
+
"learning_rate": 0.00019083280312459593,
|
603 |
+
"loss": 1.1137,
|
604 |
+
"step": 420
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.22636484687083888,
|
608 |
+
"grad_norm": 0.07399044805064979,
|
609 |
+
"learning_rate": 0.0001904398936231047,
|
610 |
+
"loss": 1.1533,
|
611 |
+
"step": 425
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.22902796271637815,
|
615 |
+
"grad_norm": 0.07782197426798759,
|
616 |
+
"learning_rate": 0.00019003916176938836,
|
617 |
+
"loss": 1.1458,
|
618 |
+
"step": 430
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.23169107856191745,
|
622 |
+
"grad_norm": 0.06822071830212563,
|
623 |
+
"learning_rate": 0.00018963064222365694,
|
624 |
+
"loss": 1.1448,
|
625 |
+
"step": 435
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.23435419440745672,
|
629 |
+
"grad_norm": 0.06944246120343146,
|
630 |
+
"learning_rate": 0.00018921437031969558,
|
631 |
+
"loss": 1.1577,
|
632 |
+
"step": 440
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.237017310252996,
|
636 |
+
"grad_norm": 0.07108688216307608,
|
637 |
+
"learning_rate": 0.0001887903820618087,
|
638 |
+
"loss": 1.1526,
|
639 |
+
"step": 445
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.2396804260985353,
|
643 |
+
"grad_norm": 0.08455610060485116,
|
644 |
+
"learning_rate": 0.00018835871412170563,
|
645 |
+
"loss": 1.1517,
|
646 |
+
"step": 450
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.24234354194407456,
|
650 |
+
"grad_norm": 0.06664786445884358,
|
651 |
+
"learning_rate": 0.0001879194038353289,
|
652 |
+
"loss": 1.1537,
|
653 |
+
"step": 455
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 0.24500665778961384,
|
657 |
+
"grad_norm": 0.07089581724112333,
|
658 |
+
"learning_rate": 0.00018747248919962498,
|
659 |
+
"loss": 1.1409,
|
660 |
+
"step": 460
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 0.24766977363515313,
|
664 |
+
"grad_norm": 0.07242825833109466,
|
665 |
+
"learning_rate": 0.00018701800886925782,
|
666 |
+
"loss": 1.1303,
|
667 |
+
"step": 465
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 0.25033288948069243,
|
671 |
+
"grad_norm": 0.06598593287452807,
|
672 |
+
"learning_rate": 0.00018655600215326546,
|
673 |
+
"loss": 1.1401,
|
674 |
+
"step": 470
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 0.2529960053262317,
|
678 |
+
"grad_norm": 0.07020789015379635,
|
679 |
+
"learning_rate": 0.00018608650901166032,
|
680 |
+
"loss": 1.1542,
|
681 |
+
"step": 475
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.255659121171771,
|
685 |
+
"grad_norm": 0.06441793150662321,
|
686 |
+
"learning_rate": 0.0001856095700519726,
|
687 |
+
"loss": 1.1276,
|
688 |
+
"step": 480
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 0.2583222370173103,
|
692 |
+
"grad_norm": 0.07254719498292789,
|
693 |
+
"learning_rate": 0.0001851252265257384,
|
694 |
+
"loss": 1.1212,
|
695 |
+
"step": 485
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 0.2609853528628495,
|
699 |
+
"grad_norm": 0.06917909155716108,
|
700 |
+
"learning_rate": 0.0001846335203249316,
|
701 |
+
"loss": 1.1298,
|
702 |
+
"step": 490
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 0.2636484687083888,
|
706 |
+
"grad_norm": 0.07470942417701212,
|
707 |
+
"learning_rate": 0.00018413449397834051,
|
708 |
+
"loss": 1.1456,
|
709 |
+
"step": 495
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 0.2663115845539281,
|
713 |
+
"grad_norm": 0.0693858861935873,
|
714 |
+
"learning_rate": 0.00018362819064788956,
|
715 |
+
"loss": 1.1327,
|
716 |
+
"step": 500
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 0.26897470039946736,
|
720 |
+
"grad_norm": 0.07182079092553902,
|
721 |
+
"learning_rate": 0.00018311465412490608,
|
722 |
+
"loss": 1.1628,
|
723 |
+
"step": 505
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 0.27163781624500666,
|
727 |
+
"grad_norm": 0.06682954118119949,
|
728 |
+
"learning_rate": 0.00018259392882633265,
|
729 |
+
"loss": 1.1528,
|
730 |
+
"step": 510
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 0.27430093209054596,
|
734 |
+
"grad_norm": 0.07248673749669132,
|
735 |
+
"learning_rate": 0.00018206605979088542,
|
736 |
+
"loss": 1.156,
|
737 |
+
"step": 515
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 0.2769640479360852,
|
741 |
+
"grad_norm": 0.06950216959497392,
|
742 |
+
"learning_rate": 0.0001815310926751586,
|
743 |
+
"loss": 1.119,
|
744 |
+
"step": 520
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 0.2796271637816245,
|
748 |
+
"grad_norm": 0.07067673018407011,
|
749 |
+
"learning_rate": 0.00018098907374967555,
|
750 |
+
"loss": 1.1211,
|
751 |
+
"step": 525
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 0.2822902796271638,
|
755 |
+
"grad_norm": 0.06820842392733384,
|
756 |
+
"learning_rate": 0.00018044004989488664,
|
757 |
+
"loss": 1.1281,
|
758 |
+
"step": 530
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 0.28495339547270304,
|
762 |
+
"grad_norm": 0.07418230217074875,
|
763 |
+
"learning_rate": 0.00017988406859711456,
|
764 |
+
"loss": 1.1409,
|
765 |
+
"step": 535
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 0.28761651131824234,
|
769 |
+
"grad_norm": 0.07009876259688716,
|
770 |
+
"learning_rate": 0.00017932117794444713,
|
771 |
+
"loss": 1.1381,
|
772 |
+
"step": 540
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 0.29027962716378164,
|
776 |
+
"grad_norm": 0.07129309605598672,
|
777 |
+
"learning_rate": 0.00017875142662257786,
|
778 |
+
"loss": 1.1387,
|
779 |
+
"step": 545
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 0.2929427430093209,
|
783 |
+
"grad_norm": 0.07830622678131702,
|
784 |
+
"learning_rate": 0.00017817486391059532,
|
785 |
+
"loss": 1.1165,
|
786 |
+
"step": 550
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 0.2956058588548602,
|
790 |
+
"grad_norm": 0.0709756673443606,
|
791 |
+
"learning_rate": 0.0001775915396767205,
|
792 |
+
"loss": 1.129,
|
793 |
+
"step": 555
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 0.2982689747003995,
|
797 |
+
"grad_norm": 0.06710174636010342,
|
798 |
+
"learning_rate": 0.00017700150437399405,
|
799 |
+
"loss": 1.1183,
|
800 |
+
"step": 560
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 0.3009320905459387,
|
804 |
+
"grad_norm": 0.07321620332053846,
|
805 |
+
"learning_rate": 0.0001764048090359121,
|
806 |
+
"loss": 1.1502,
|
807 |
+
"step": 565
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 0.303595206391478,
|
811 |
+
"grad_norm": 0.07613131980579707,
|
812 |
+
"learning_rate": 0.00017580150527201241,
|
813 |
+
"loss": 1.1322,
|
814 |
+
"step": 570
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 0.3062583222370173,
|
818 |
+
"grad_norm": 0.07480921806539248,
|
819 |
+
"learning_rate": 0.0001751916452634105,
|
820 |
+
"loss": 1.1269,
|
821 |
+
"step": 575
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 0.30892143808255657,
|
825 |
+
"grad_norm": 0.07386122393966031,
|
826 |
+
"learning_rate": 0.0001745752817582865,
|
827 |
+
"loss": 1.1528,
|
828 |
+
"step": 580
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 0.31158455392809586,
|
832 |
+
"grad_norm": 0.07032971968151802,
|
833 |
+
"learning_rate": 0.00017395246806732267,
|
834 |
+
"loss": 1.1642,
|
835 |
+
"step": 585
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 0.31424766977363516,
|
839 |
+
"grad_norm": 0.07910944378906298,
|
840 |
+
"learning_rate": 0.00017332325805909256,
|
841 |
+
"loss": 1.1328,
|
842 |
+
"step": 590
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 0.3169107856191744,
|
846 |
+
"grad_norm": 0.06775943214366806,
|
847 |
+
"learning_rate": 0.00017268770615540177,
|
848 |
+
"loss": 1.1142,
|
849 |
+
"step": 595
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 0.3195739014647137,
|
853 |
+
"grad_norm": 0.0858014191359942,
|
854 |
+
"learning_rate": 0.00017204586732658087,
|
855 |
+
"loss": 1.1393,
|
856 |
+
"step": 600
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 0.322237017310253,
|
860 |
+
"grad_norm": 0.06968407560583738,
|
861 |
+
"learning_rate": 0.00017139779708673085,
|
862 |
+
"loss": 1.1428,
|
863 |
+
"step": 605
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 0.3249001331557923,
|
867 |
+
"grad_norm": 0.06812443512073688,
|
868 |
+
"learning_rate": 0.00017074355148892167,
|
869 |
+
"loss": 1.1592,
|
870 |
+
"step": 610
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 0.32756324900133155,
|
874 |
+
"grad_norm": 0.07150170839574509,
|
875 |
+
"learning_rate": 0.00017008318712034403,
|
876 |
+
"loss": 1.1018,
|
877 |
+
"step": 615
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 0.33022636484687085,
|
881 |
+
"grad_norm": 0.06906369302490485,
|
882 |
+
"learning_rate": 0.00016941676109741508,
|
883 |
+
"loss": 1.1442,
|
884 |
+
"step": 620
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 0.33288948069241014,
|
888 |
+
"grad_norm": 0.07869503084625909,
|
889 |
+
"learning_rate": 0.00016874433106083814,
|
890 |
+
"loss": 1.1132,
|
891 |
+
"step": 625
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 0.3355525965379494,
|
895 |
+
"grad_norm": 0.07767900677929127,
|
896 |
+
"learning_rate": 0.00016806595517061744,
|
897 |
+
"loss": 1.1362,
|
898 |
+
"step": 630
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 0.3382157123834887,
|
902 |
+
"grad_norm": 0.06780573276986938,
|
903 |
+
"learning_rate": 0.00016738169210102764,
|
904 |
+
"loss": 1.1382,
|
905 |
+
"step": 635
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 0.340878828229028,
|
909 |
+
"grad_norm": 0.07855904717914698,
|
910 |
+
"learning_rate": 0.00016669160103553884,
|
911 |
+
"loss": 1.1146,
|
912 |
+
"step": 640
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 0.34354194407456723,
|
916 |
+
"grad_norm": 0.06976051466447154,
|
917 |
+
"learning_rate": 0.00016599574166169782,
|
918 |
+
"loss": 1.1156,
|
919 |
+
"step": 645
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 0.34620505992010653,
|
923 |
+
"grad_norm": 0.0659729198246215,
|
924 |
+
"learning_rate": 0.0001652941741659655,
|
925 |
+
"loss": 1.1636,
|
926 |
+
"step": 650
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 0.3488681757656458,
|
930 |
+
"grad_norm": 0.06820327345322129,
|
931 |
+
"learning_rate": 0.00016458695922851125,
|
932 |
+
"loss": 1.1272,
|
933 |
+
"step": 655
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 0.35153129161118507,
|
937 |
+
"grad_norm": 0.0706423611601847,
|
938 |
+
"learning_rate": 0.0001638741580179645,
|
939 |
+
"loss": 1.15,
|
940 |
+
"step": 660
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 0.35419440745672437,
|
944 |
+
"grad_norm": 0.07009862917994238,
|
945 |
+
"learning_rate": 0.0001631558321861241,
|
946 |
+
"loss": 1.1133,
|
947 |
+
"step": 665
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 0.35685752330226367,
|
951 |
+
"grad_norm": 0.10252097837226529,
|
952 |
+
"learning_rate": 0.00016243204386262616,
|
953 |
+
"loss": 1.1275,
|
954 |
+
"step": 670
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 0.3595206391478029,
|
958 |
+
"grad_norm": 0.0677270369109815,
|
959 |
+
"learning_rate": 0.0001617028556495699,
|
960 |
+
"loss": 1.1463,
|
961 |
+
"step": 675
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 0.3621837549933422,
|
965 |
+
"grad_norm": 0.07081566637081647,
|
966 |
+
"learning_rate": 0.00016096833061610336,
|
967 |
+
"loss": 1.1557,
|
968 |
+
"step": 680
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 0.3648468708388815,
|
972 |
+
"grad_norm": 0.07606309640077409,
|
973 |
+
"learning_rate": 0.0001602285322929684,
|
974 |
+
"loss": 1.1279,
|
975 |
+
"step": 685
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 0.36750998668442075,
|
979 |
+
"grad_norm": 0.06926585358652293,
|
980 |
+
"learning_rate": 0.00015948352466700562,
|
981 |
+
"loss": 1.1058,
|
982 |
+
"step": 690
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 0.37017310252996005,
|
986 |
+
"grad_norm": 0.0768394516058797,
|
987 |
+
"learning_rate": 0.00015873337217562012,
|
988 |
+
"loss": 1.1451,
|
989 |
+
"step": 695
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 0.37283621837549935,
|
993 |
+
"grad_norm": 0.07574776045146851,
|
994 |
+
"learning_rate": 0.00015797813970120806,
|
995 |
+
"loss": 1.1529,
|
996 |
+
"step": 700
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 0.3754993342210386,
|
1000 |
+
"grad_norm": 0.08825811667362324,
|
1001 |
+
"learning_rate": 0.00015721789256554493,
|
1002 |
+
"loss": 1.1427,
|
1003 |
+
"step": 705
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 0.3781624500665779,
|
1007 |
+
"grad_norm": 0.07195501325596203,
|
1008 |
+
"learning_rate": 0.00015645269652413572,
|
1009 |
+
"loss": 1.1348,
|
1010 |
+
"step": 710
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 0.3808255659121172,
|
1014 |
+
"grad_norm": 0.07470988656942844,
|
1015 |
+
"learning_rate": 0.00015568261776052747,
|
1016 |
+
"loss": 1.1389,
|
1017 |
+
"step": 715
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 0.38348868175765644,
|
1021 |
+
"grad_norm": 0.07234292608880714,
|
1022 |
+
"learning_rate": 0.0001549077228805851,
|
1023 |
+
"loss": 1.1265,
|
1024 |
+
"step": 720
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 0.38615179760319573,
|
1028 |
+
"grad_norm": 0.07603813138240277,
|
1029 |
+
"learning_rate": 0.00015412807890673012,
|
1030 |
+
"loss": 1.0975,
|
1031 |
+
"step": 725
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 0.38881491344873503,
|
1035 |
+
"grad_norm": 0.06914740850103318,
|
1036 |
+
"learning_rate": 0.00015334375327214435,
|
1037 |
+
"loss": 1.1656,
|
1038 |
+
"step": 730
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 0.3914780292942743,
|
1042 |
+
"grad_norm": 0.0723373355088882,
|
1043 |
+
"learning_rate": 0.00015255481381493686,
|
1044 |
+
"loss": 1.1235,
|
1045 |
+
"step": 735
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 0.3941411451398136,
|
1049 |
+
"grad_norm": 0.07469762097501292,
|
1050 |
+
"learning_rate": 0.00015176132877227672,
|
1051 |
+
"loss": 1.1401,
|
1052 |
+
"step": 740
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 0.3968042609853529,
|
1056 |
+
"grad_norm": 0.06845354027517625,
|
1057 |
+
"learning_rate": 0.00015096336677449123,
|
1058 |
+
"loss": 1.1299,
|
1059 |
+
"step": 745
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 0.3994673768308921,
|
1063 |
+
"grad_norm": 0.07857096344059177,
|
1064 |
+
"learning_rate": 0.0001501609968391295,
|
1065 |
+
"loss": 1.1362,
|
1066 |
+
"step": 750
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 0.4021304926764314,
|
1070 |
+
"grad_norm": 0.07079465135436822,
|
1071 |
+
"learning_rate": 0.00014935428836499332,
|
1072 |
+
"loss": 1.1268,
|
1073 |
+
"step": 755
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 0.4047936085219707,
|
1077 |
+
"grad_norm": 0.07113035589654983,
|
1078 |
+
"learning_rate": 0.0001485433111261346,
|
1079 |
+
"loss": 1.1357,
|
1080 |
+
"step": 760
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 0.40745672436750996,
|
1084 |
+
"grad_norm": 0.0703269232774503,
|
1085 |
+
"learning_rate": 0.0001477281352658203,
|
1086 |
+
"loss": 1.1239,
|
1087 |
+
"step": 765
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 0.41011984021304926,
|
1091 |
+
"grad_norm": 0.07059355929742223,
|
1092 |
+
"learning_rate": 0.00014690883129046584,
|
1093 |
+
"loss": 1.1442,
|
1094 |
+
"step": 770
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 0.41278295605858856,
|
1098 |
+
"grad_norm": 0.07289277380542494,
|
1099 |
+
"learning_rate": 0.0001460854700635366,
|
1100 |
+
"loss": 1.1267,
|
1101 |
+
"step": 775
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 0.41544607190412786,
|
1105 |
+
"grad_norm": 0.06984202720886337,
|
1106 |
+
"learning_rate": 0.00014525812279941896,
|
1107 |
+
"loss": 1.1258,
|
1108 |
+
"step": 780
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 0.4181091877496671,
|
1112 |
+
"grad_norm": 0.07422048925652114,
|
1113 |
+
"learning_rate": 0.00014442686105726067,
|
1114 |
+
"loss": 1.1193,
|
1115 |
+
"step": 785
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 0.4207723035952064,
|
1119 |
+
"grad_norm": 0.0718716738772194,
|
1120 |
+
"learning_rate": 0.00014359175673478162,
|
1121 |
+
"loss": 1.133,
|
1122 |
+
"step": 790
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 0.4234354194407457,
|
1126 |
+
"grad_norm": 0.07204344165616748,
|
1127 |
+
"learning_rate": 0.00014275288206205524,
|
1128 |
+
"loss": 1.0967,
|
1129 |
+
"step": 795
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 0.42609853528628494,
|
1133 |
+
"grad_norm": 0.07353413673583019,
|
1134 |
+
"learning_rate": 0.00014191030959526105,
|
1135 |
+
"loss": 1.1261,
|
1136 |
+
"step": 800
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 0.42876165113182424,
|
1140 |
+
"grad_norm": 0.0707040442967912,
|
1141 |
+
"learning_rate": 0.00014106411221040933,
|
1142 |
+
"loss": 1.128,
|
1143 |
+
"step": 805
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 0.43142476697736354,
|
1147 |
+
"grad_norm": 0.07086259967904554,
|
1148 |
+
"learning_rate": 0.00014021436309703765,
|
1149 |
+
"loss": 1.107,
|
1150 |
+
"step": 810
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 0.4340878828229028,
|
1154 |
+
"grad_norm": 0.06994136097058291,
|
1155 |
+
"learning_rate": 0.00013936113575188075,
|
1156 |
+
"loss": 1.1221,
|
1157 |
+
"step": 815
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 0.4367509986684421,
|
1161 |
+
"grad_norm": 0.06961342073957084,
|
1162 |
+
"learning_rate": 0.00013850450397251345,
|
1163 |
+
"loss": 1.1208,
|
1164 |
+
"step": 820
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 0.4394141145139814,
|
1168 |
+
"grad_norm": 0.07040932769118938,
|
1169 |
+
"learning_rate": 0.0001376445418509679,
|
1170 |
+
"loss": 1.1208,
|
1171 |
+
"step": 825
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 0.4420772303595206,
|
1175 |
+
"grad_norm": 0.07187314857901307,
|
1176 |
+
"learning_rate": 0.00013678132376732517,
|
1177 |
+
"loss": 1.1267,
|
1178 |
+
"step": 830
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 0.4447403462050599,
|
1182 |
+
"grad_norm": 0.07002729221567015,
|
1183 |
+
"learning_rate": 0.00013591492438328183,
|
1184 |
+
"loss": 1.1421,
|
1185 |
+
"step": 835
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 0.4474034620505992,
|
1189 |
+
"grad_norm": 0.07235067324392022,
|
1190 |
+
"learning_rate": 0.0001350454186356924,
|
1191 |
+
"loss": 1.1191,
|
1192 |
+
"step": 840
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 0.45006657789613846,
|
1196 |
+
"grad_norm": 0.07568410556158327,
|
1197 |
+
"learning_rate": 0.00013417288173008776,
|
1198 |
+
"loss": 1.1123,
|
1199 |
+
"step": 845
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 0.45272969374167776,
|
1203 |
+
"grad_norm": 0.07613155957113646,
|
1204 |
+
"learning_rate": 0.00013329738913417068,
|
1205 |
+
"loss": 1.1137,
|
1206 |
+
"step": 850
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 0.45539280958721706,
|
1210 |
+
"grad_norm": 0.06854447943505289,
|
1211 |
+
"learning_rate": 0.00013241901657128825,
|
1212 |
+
"loss": 1.132,
|
1213 |
+
"step": 855
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 0.4580559254327563,
|
1217 |
+
"grad_norm": 0.06884442803824642,
|
1218 |
+
"learning_rate": 0.00013153784001388247,
|
1219 |
+
"loss": 1.1352,
|
1220 |
+
"step": 860
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 0.4607190412782956,
|
1224 |
+
"grad_norm": 0.0818107600119684,
|
1225 |
+
"learning_rate": 0.00013065393567691913,
|
1226 |
+
"loss": 1.101,
|
1227 |
+
"step": 865
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 0.4633821571238349,
|
1231 |
+
"grad_norm": 0.07690425325074156,
|
1232 |
+
"learning_rate": 0.00012976738001129606,
|
1233 |
+
"loss": 1.1052,
|
1234 |
+
"step": 870
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 0.46604527296937415,
|
1238 |
+
"grad_norm": 0.07952080928038242,
|
1239 |
+
"learning_rate": 0.00012887824969723034,
|
1240 |
+
"loss": 1.1172,
|
1241 |
+
"step": 875
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 0.46870838881491345,
|
1245 |
+
"grad_norm": 0.06637081804522467,
|
1246 |
+
"learning_rate": 0.00012798662163762635,
|
1247 |
+
"loss": 1.1236,
|
1248 |
+
"step": 880
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 0.47137150466045274,
|
1252 |
+
"grad_norm": 0.06753657536020181,
|
1253 |
+
"learning_rate": 0.00012709257295142422,
|
1254 |
+
"loss": 1.1304,
|
1255 |
+
"step": 885
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 0.474034620505992,
|
1259 |
+
"grad_norm": 0.07408199495580661,
|
1260 |
+
"learning_rate": 0.00012619618096692943,
|
1261 |
+
"loss": 1.1523,
|
1262 |
+
"step": 890
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 0.4766977363515313,
|
1266 |
+
"grad_norm": 0.07508862680813526,
|
1267 |
+
"learning_rate": 0.0001252975232151248,
|
1268 |
+
"loss": 1.1158,
|
1269 |
+
"step": 895
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 0.4793608521970706,
|
1273 |
+
"grad_norm": 0.07064406721668945,
|
1274 |
+
"learning_rate": 0.0001243966774229645,
|
1275 |
+
"loss": 1.1334,
|
1276 |
+
"step": 900
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 0.48202396804260983,
|
1280 |
+
"grad_norm": 0.0709335857697155,
|
1281 |
+
"learning_rate": 0.00012349372150665118,
|
1282 |
+
"loss": 1.1104,
|
1283 |
+
"step": 905
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 0.48468708388814913,
|
1287 |
+
"grad_norm": 0.06779392856513489,
|
1288 |
+
"learning_rate": 0.00012258873356489714,
|
1289 |
+
"loss": 1.1299,
|
1290 |
+
"step": 910
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 0.4873501997336884,
|
1294 |
+
"grad_norm": 0.07922531663031743,
|
1295 |
+
"learning_rate": 0.00012168179187216893,
|
1296 |
+
"loss": 1.13,
|
1297 |
+
"step": 915
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 0.49001331557922767,
|
1301 |
+
"grad_norm": 0.07185035702343927,
|
1302 |
+
"learning_rate": 0.0001207729748719177,
|
1303 |
+
"loss": 1.1402,
|
1304 |
+
"step": 920
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 0.49267643142476697,
|
1308 |
+
"grad_norm": 0.07162588557593508,
|
1309 |
+
"learning_rate": 0.00011986236116979406,
|
1310 |
+
"loss": 1.1308,
|
1311 |
+
"step": 925
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 0.49533954727030627,
|
1315 |
+
"grad_norm": 0.07242326329471309,
|
1316 |
+
"learning_rate": 0.0001189500295268495,
|
1317 |
+
"loss": 1.106,
|
1318 |
+
"step": 930
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 0.4980026631158455,
|
1322 |
+
"grad_norm": 0.07434286419305244,
|
1323 |
+
"learning_rate": 0.0001180360588527242,
|
1324 |
+
"loss": 1.119,
|
1325 |
+
"step": 935
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 0.5006657789613849,
|
1329 |
+
"grad_norm": 0.07304306593688702,
|
1330 |
+
"learning_rate": 0.00011712052819882171,
|
1331 |
+
"loss": 1.1503,
|
1332 |
+
"step": 940
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 0.5033288948069241,
|
1336 |
+
"grad_norm": 0.07214685709611712,
|
1337 |
+
"learning_rate": 0.00011620351675147195,
|
1338 |
+
"loss": 1.1095,
|
1339 |
+
"step": 945
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 0.5059920106524634,
|
1343 |
+
"grad_norm": 0.07192838129765652,
|
1344 |
+
"learning_rate": 0.0001152851038250819,
|
1345 |
+
"loss": 1.1451,
|
1346 |
+
"step": 950
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 0.5086551264980027,
|
1350 |
+
"grad_norm": 0.06935787206272043,
|
1351 |
+
"learning_rate": 0.00011436536885527576,
|
1352 |
+
"loss": 1.1251,
|
1353 |
+
"step": 955
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 0.511318242343542,
|
1357 |
+
"grad_norm": 0.06801727149157547,
|
1358 |
+
"learning_rate": 0.00011344439139202421,
|
1359 |
+
"loss": 1.1084,
|
1360 |
+
"step": 960
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 0.5139813581890812,
|
1364 |
+
"grad_norm": 0.07024555806497951,
|
1365 |
+
"learning_rate": 0.00011252225109276404,
|
1366 |
+
"loss": 1.1278,
|
1367 |
+
"step": 965
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 0.5166444740346205,
|
1371 |
+
"grad_norm": 0.06796418511383114,
|
1372 |
+
"learning_rate": 0.00011159902771550837,
|
1373 |
+
"loss": 1.1092,
|
1374 |
+
"step": 970
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 0.5193075898801598,
|
1378 |
+
"grad_norm": 0.07219790467805971,
|
1379 |
+
"learning_rate": 0.00011067480111194817,
|
1380 |
+
"loss": 1.1286,
|
1381 |
+
"step": 975
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 0.521970705725699,
|
1385 |
+
"grad_norm": 0.06944360419194191,
|
1386 |
+
"learning_rate": 0.00010974965122054579,
|
1387 |
+
"loss": 1.1184,
|
1388 |
+
"step": 980
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 0.5246338215712384,
|
1392 |
+
"grad_norm": 0.07229326850745169,
|
1393 |
+
"learning_rate": 0.00010882365805962083,
|
1394 |
+
"loss": 1.1212,
|
1395 |
+
"step": 985
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 0.5272969374167776,
|
1399 |
+
"grad_norm": 0.07181125732929394,
|
1400 |
+
"learning_rate": 0.00010789690172042912,
|
1401 |
+
"loss": 1.1137,
|
1402 |
+
"step": 990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 0.5299600532623169,
|
1406 |
+
"grad_norm": 0.07145294348037948,
|
1407 |
+
"learning_rate": 0.00010696946236023567,
|
1408 |
+
"loss": 1.1365,
|
1409 |
+
"step": 995
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 0.5326231691078562,
|
1413 |
+
"grad_norm": 0.07057067837966788,
|
1414 |
+
"learning_rate": 0.00010604142019538135,
|
1415 |
+
"loss": 1.1176,
|
1416 |
+
"step": 1000
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 0.5352862849533955,
|
1420 |
+
"grad_norm": 0.07461665274600122,
|
1421 |
+
"learning_rate": 0.00010511285549434509,
|
1422 |
+
"loss": 1.1152,
|
1423 |
+
"step": 1005
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 0.5379494007989347,
|
1427 |
+
"grad_norm": 0.07009722058482384,
|
1428 |
+
"learning_rate": 0.00010418384857080117,
|
1429 |
+
"loss": 1.1117,
|
1430 |
+
"step": 1010
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 0.5406125166444741,
|
1434 |
+
"grad_norm": 0.07081680661261375,
|
1435 |
+
"learning_rate": 0.00010325447977667263,
|
1436 |
+
"loss": 1.1328,
|
1437 |
+
"step": 1015
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 0.5432756324900133,
|
1441 |
+
"grad_norm": 0.06980388274242631,
|
1442 |
+
"learning_rate": 0.00010232482949518156,
|
1443 |
+
"loss": 1.1404,
|
1444 |
+
"step": 1020
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 0.5459387483355526,
|
1448 |
+
"grad_norm": 0.06946364728493262,
|
1449 |
+
"learning_rate": 0.00010139497813389654,
|
1450 |
+
"loss": 1.1127,
|
1451 |
+
"step": 1025
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 0.5486018641810919,
|
1455 |
+
"grad_norm": 0.06947172503952885,
|
1456 |
+
"learning_rate": 0.00010046500611777798,
|
1457 |
+
"loss": 1.0937,
|
1458 |
+
"step": 1030
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 0.5512649800266312,
|
1462 |
+
"grad_norm": 0.07414647895551518,
|
1463 |
+
"learning_rate": 9.953499388222202e-05,
|
1464 |
+
"loss": 1.132,
|
1465 |
+
"step": 1035
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 0.5539280958721704,
|
1469 |
+
"grad_norm": 0.07085672498663681,
|
1470 |
+
"learning_rate": 9.860502186610349e-05,
|
1471 |
+
"loss": 1.0998,
|
1472 |
+
"step": 1040
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 0.5565912117177098,
|
1476 |
+
"grad_norm": 0.07198312270884867,
|
1477 |
+
"learning_rate": 9.767517050481846e-05,
|
1478 |
+
"loss": 1.1263,
|
1479 |
+
"step": 1045
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 0.559254327563249,
|
1483 |
+
"grad_norm": 0.07070349541708286,
|
1484 |
+
"learning_rate": 9.67455202233274e-05,
|
1485 |
+
"loss": 1.1143,
|
1486 |
+
"step": 1050
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 0.5619174434087882,
|
1490 |
+
"grad_norm": 0.06990981328605791,
|
1491 |
+
"learning_rate": 9.581615142919887e-05,
|
1492 |
+
"loss": 1.1168,
|
1493 |
+
"step": 1055
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 0.5645805592543276,
|
1497 |
+
"grad_norm": 0.07221018297233557,
|
1498 |
+
"learning_rate": 9.488714450565491e-05,
|
1499 |
+
"loss": 1.1123,
|
1500 |
+
"step": 1060
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 0.5672436750998668,
|
1504 |
+
"grad_norm": 0.06895775963564511,
|
1505 |
+
"learning_rate": 9.395857980461867e-05,
|
1506 |
+
"loss": 1.1294,
|
1507 |
+
"step": 1065
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 0.5699067909454061,
|
1511 |
+
"grad_norm": 0.06904508970279108,
|
1512 |
+
"learning_rate": 9.303053763976434e-05,
|
1513 |
+
"loss": 1.1179,
|
1514 |
+
"step": 1070
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 0.5725699067909454,
|
1518 |
+
"grad_norm": 0.07131791944898686,
|
1519 |
+
"learning_rate": 9.210309827957089e-05,
|
1520 |
+
"loss": 1.1297,
|
1521 |
+
"step": 1075
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 0.5752330226364847,
|
1525 |
+
"grad_norm": 0.07117429268373339,
|
1526 |
+
"learning_rate": 9.117634194037922e-05,
|
1527 |
+
"loss": 1.1285,
|
1528 |
+
"step": 1080
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 0.5778961384820239,
|
1532 |
+
"grad_norm": 0.0720403827517469,
|
1533 |
+
"learning_rate": 9.025034877945422e-05,
|
1534 |
+
"loss": 1.1418,
|
1535 |
+
"step": 1085
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 0.5805592543275633,
|
1539 |
+
"grad_norm": 0.07399424819774852,
|
1540 |
+
"learning_rate": 8.932519888805185e-05,
|
1541 |
+
"loss": 1.1521,
|
1542 |
+
"step": 1090
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 0.5832223701731025,
|
1546 |
+
"grad_norm": 0.06757782738616369,
|
1547 |
+
"learning_rate": 8.840097228449165e-05,
|
1548 |
+
"loss": 1.1468,
|
1549 |
+
"step": 1095
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 0.5858854860186418,
|
1553 |
+
"grad_norm": 0.07513462470349377,
|
1554 |
+
"learning_rate": 8.747774890723599e-05,
|
1555 |
+
"loss": 1.1008,
|
1556 |
+
"step": 1100
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 0.5885486018641811,
|
1560 |
+
"grad_norm": 0.07193663724723076,
|
1561 |
+
"learning_rate": 8.655560860797582e-05,
|
1562 |
+
"loss": 1.1364,
|
1563 |
+
"step": 1105
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 0.5912117177097204,
|
1567 |
+
"grad_norm": 0.07451485853238139,
|
1568 |
+
"learning_rate": 8.563463114472425e-05,
|
1569 |
+
"loss": 1.1077,
|
1570 |
+
"step": 1110
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 0.5938748335552596,
|
1574 |
+
"grad_norm": 0.07208026414944517,
|
1575 |
+
"learning_rate": 8.471489617491812e-05,
|
1576 |
+
"loss": 1.0828,
|
1577 |
+
"step": 1115
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 0.596537949400799,
|
1581 |
+
"grad_norm": 0.06968442188475359,
|
1582 |
+
"learning_rate": 8.379648324852808e-05,
|
1583 |
+
"loss": 1.0975,
|
1584 |
+
"step": 1120
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 0.5992010652463382,
|
1588 |
+
"grad_norm": 0.07690321076998301,
|
1589 |
+
"learning_rate": 8.287947180117832e-05,
|
1590 |
+
"loss": 1.1149,
|
1591 |
+
"step": 1125
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 0.6018641810918774,
|
1595 |
+
"grad_norm": 0.07088556456471248,
|
1596 |
+
"learning_rate": 8.196394114727585e-05,
|
1597 |
+
"loss": 1.1193,
|
1598 |
+
"step": 1130
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 0.6045272969374168,
|
1602 |
+
"grad_norm": 0.07306182012233282,
|
1603 |
+
"learning_rate": 8.104997047315048e-05,
|
1604 |
+
"loss": 1.1222,
|
1605 |
+
"step": 1135
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 0.607190412782956,
|
1609 |
+
"grad_norm": 0.0735136578466246,
|
1610 |
+
"learning_rate": 8.013763883020596e-05,
|
1611 |
+
"loss": 1.1326,
|
1612 |
+
"step": 1140
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 0.6098535286284953,
|
1616 |
+
"grad_norm": 0.0792004084312011,
|
1617 |
+
"learning_rate": 7.92270251280823e-05,
|
1618 |
+
"loss": 1.1125,
|
1619 |
+
"step": 1145
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 0.6125166444740346,
|
1623 |
+
"grad_norm": 0.06995636578434544,
|
1624 |
+
"learning_rate": 7.831820812783108e-05,
|
1625 |
+
"loss": 1.1397,
|
1626 |
+
"step": 1150
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 0.6151797603195739,
|
1630 |
+
"grad_norm": 0.07156672219633958,
|
1631 |
+
"learning_rate": 7.741126643510292e-05,
|
1632 |
+
"loss": 1.1047,
|
1633 |
+
"step": 1155
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 0.6178428761651131,
|
1637 |
+
"grad_norm": 0.06990042451203095,
|
1638 |
+
"learning_rate": 7.650627849334881e-05,
|
1639 |
+
"loss": 1.0991,
|
1640 |
+
"step": 1160
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 0.6205059920106525,
|
1644 |
+
"grad_norm": 0.07175460373797926,
|
1645 |
+
"learning_rate": 7.560332257703555e-05,
|
1646 |
+
"loss": 1.1179,
|
1647 |
+
"step": 1165
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 0.6231691078561917,
|
1651 |
+
"grad_norm": 0.07079860289283964,
|
1652 |
+
"learning_rate": 7.470247678487522e-05,
|
1653 |
+
"loss": 1.1179,
|
1654 |
+
"step": 1170
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 0.625832223701731,
|
1658 |
+
"grad_norm": 0.06955306779443195,
|
1659 |
+
"learning_rate": 7.380381903307061e-05,
|
1660 |
+
"loss": 1.1261,
|
1661 |
+
"step": 1175
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 0.6284953395472703,
|
1665 |
+
"grad_norm": 0.07040121930899221,
|
1666 |
+
"learning_rate": 7.290742704857585e-05,
|
1667 |
+
"loss": 1.128,
|
1668 |
+
"step": 1180
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 0.6311584553928096,
|
1672 |
+
"grad_norm": 0.0696134794381735,
|
1673 |
+
"learning_rate": 7.201337836237365e-05,
|
1674 |
+
"loss": 1.1006,
|
1675 |
+
"step": 1185
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 0.6338215712383488,
|
1679 |
+
"grad_norm": 0.0731972119965864,
|
1680 |
+
"learning_rate": 7.112175030276969e-05,
|
1681 |
+
"loss": 1.122,
|
1682 |
+
"step": 1190
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 0.6364846870838882,
|
1686 |
+
"grad_norm": 0.07600883906312528,
|
1687 |
+
"learning_rate": 7.023261998870395e-05,
|
1688 |
+
"loss": 1.1054,
|
1689 |
+
"step": 1195
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 0.6391478029294274,
|
1693 |
+
"grad_norm": 0.06927208348922254,
|
1694 |
+
"learning_rate": 6.934606432308086e-05,
|
1695 |
+
"loss": 1.1128,
|
1696 |
+
"step": 1200
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 0.6418109187749668,
|
1700 |
+
"grad_norm": 0.06936985978073082,
|
1701 |
+
"learning_rate": 6.846215998611757e-05,
|
1702 |
+
"loss": 1.118,
|
1703 |
+
"step": 1205
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 0.644474034620506,
|
1707 |
+
"grad_norm": 0.07077273319280314,
|
1708 |
+
"learning_rate": 6.758098342871174e-05,
|
1709 |
+
"loss": 1.1093,
|
1710 |
+
"step": 1210
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 0.6471371504660453,
|
1714 |
+
"grad_norm": 0.07079727575348424,
|
1715 |
+
"learning_rate": 6.670261086582933e-05,
|
1716 |
+
"loss": 1.1231,
|
1717 |
+
"step": 1215
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 0.6498002663115846,
|
1721 |
+
"grad_norm": 0.07181688909396321,
|
1722 |
+
"learning_rate": 6.582711826991226e-05,
|
1723 |
+
"loss": 1.1042,
|
1724 |
+
"step": 1220
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 0.6524633821571239,
|
1728 |
+
"grad_norm": 0.07337324643783341,
|
1729 |
+
"learning_rate": 6.495458136430765e-05,
|
1730 |
+
"loss": 1.1042,
|
1731 |
+
"step": 1225
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 0.6551264980026631,
|
1735 |
+
"grad_norm": 0.07212862237245772,
|
1736 |
+
"learning_rate": 6.408507561671819e-05,
|
1737 |
+
"loss": 1.1509,
|
1738 |
+
"step": 1230
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 0.6577896138482024,
|
1742 |
+
"grad_norm": 0.07047120035609092,
|
1743 |
+
"learning_rate": 6.321867623267481e-05,
|
1744 |
+
"loss": 1.1355,
|
1745 |
+
"step": 1235
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 0.6604527296937417,
|
1749 |
+
"grad_norm": 0.07459478385718604,
|
1750 |
+
"learning_rate": 6.23554581490321e-05,
|
1751 |
+
"loss": 1.1178,
|
1752 |
+
"step": 1240
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 0.6631158455392809,
|
1756 |
+
"grad_norm": 0.06825726344400981,
|
1757 |
+
"learning_rate": 6.149549602748656e-05,
|
1758 |
+
"loss": 1.0862,
|
1759 |
+
"step": 1245
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 0.6657789613848203,
|
1763 |
+
"grad_norm": 0.07240193352920372,
|
1764 |
+
"learning_rate": 6.063886424811929e-05,
|
1765 |
+
"loss": 1.1292,
|
1766 |
+
"step": 1250
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 0.6684420772303595,
|
1770 |
+
"grad_norm": 0.07197359550672061,
|
1771 |
+
"learning_rate": 5.9785636902962374e-05,
|
1772 |
+
"loss": 1.1306,
|
1773 |
+
"step": 1255
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 0.6711051930758988,
|
1777 |
+
"grad_norm": 0.07343322507680913,
|
1778 |
+
"learning_rate": 5.893588778959067e-05,
|
1779 |
+
"loss": 1.1365,
|
1780 |
+
"step": 1260
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 0.6737683089214381,
|
1784 |
+
"grad_norm": 0.07842185463764602,
|
1785 |
+
"learning_rate": 5.8089690404738925e-05,
|
1786 |
+
"loss": 1.1395,
|
1787 |
+
"step": 1265
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 0.6764314247669774,
|
1791 |
+
"grad_norm": 0.07270053681642126,
|
1792 |
+
"learning_rate": 5.7247117937944786e-05,
|
1793 |
+
"loss": 1.1035,
|
1794 |
+
"step": 1270
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 0.6790945406125166,
|
1798 |
+
"grad_norm": 0.07124055104083139,
|
1799 |
+
"learning_rate": 5.640824326521841e-05,
|
1800 |
+
"loss": 1.1121,
|
1801 |
+
"step": 1275
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 0.681757656458056,
|
1805 |
+
"grad_norm": 0.07031788560347749,
|
1806 |
+
"learning_rate": 5.5573138942739365e-05,
|
1807 |
+
"loss": 1.1192,
|
1808 |
+
"step": 1280
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 0.6844207723035952,
|
1812 |
+
"grad_norm": 0.07253058219251593,
|
1813 |
+
"learning_rate": 5.4741877200581057e-05,
|
1814 |
+
"loss": 1.1324,
|
1815 |
+
"step": 1285
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 0.6870838881491345,
|
1819 |
+
"grad_norm": 0.0725305584439251,
|
1820 |
+
"learning_rate": 5.391452993646342e-05,
|
1821 |
+
"loss": 1.1387,
|
1822 |
+
"step": 1290
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 0.6897470039946738,
|
1826 |
+
"grad_norm": 0.07021354602161774,
|
1827 |
+
"learning_rate": 5.30911687095342e-05,
|
1828 |
+
"loss": 1.126,
|
1829 |
+
"step": 1295
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 0.6924101198402131,
|
1833 |
+
"grad_norm": 0.07012940672098344,
|
1834 |
+
"learning_rate": 5.227186473417971e-05,
|
1835 |
+
"loss": 1.1486,
|
1836 |
+
"step": 1300
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 0.6950732356857523,
|
1840 |
+
"grad_norm": 0.07232533360594767,
|
1841 |
+
"learning_rate": 5.145668887386543e-05,
|
1842 |
+
"loss": 1.1111,
|
1843 |
+
"step": 1305
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 0.6977363515312917,
|
1847 |
+
"grad_norm": 0.07242821758103567,
|
1848 |
+
"learning_rate": 5.064571163500667e-05,
|
1849 |
+
"loss": 1.1181,
|
1850 |
+
"step": 1310
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 0.7003994673768309,
|
1854 |
+
"grad_norm": 0.07148030530795384,
|
1855 |
+
"learning_rate": 4.983900316087051e-05,
|
1856 |
+
"loss": 1.0922,
|
1857 |
+
"step": 1315
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 0.7030625832223701,
|
1861 |
+
"grad_norm": 0.07277623674879777,
|
1862 |
+
"learning_rate": 4.90366332255088e-05,
|
1863 |
+
"loss": 1.0985,
|
1864 |
+
"step": 1320
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 0.7057256990679095,
|
1868 |
+
"grad_norm": 0.07450043090731064,
|
1869 |
+
"learning_rate": 4.823867122772329e-05,
|
1870 |
+
"loss": 1.1177,
|
1871 |
+
"step": 1325
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 0.7083888149134487,
|
1875 |
+
"grad_norm": 0.07438166449706331,
|
1876 |
+
"learning_rate": 4.744518618506319e-05,
|
1877 |
+
"loss": 1.1225,
|
1878 |
+
"step": 1330
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 0.711051930758988,
|
1882 |
+
"grad_norm": 0.07157294185481793,
|
1883 |
+
"learning_rate": 4.665624672785566e-05,
|
1884 |
+
"loss": 1.1291,
|
1885 |
+
"step": 1335
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 0.7137150466045273,
|
1889 |
+
"grad_norm": 0.07520781395221099,
|
1890 |
+
"learning_rate": 4.5871921093269875e-05,
|
1891 |
+
"loss": 1.1082,
|
1892 |
+
"step": 1340
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 0.7163781624500666,
|
1896 |
+
"grad_norm": 0.07203181974145231,
|
1897 |
+
"learning_rate": 4.5092277119414975e-05,
|
1898 |
+
"loss": 1.1333,
|
1899 |
+
"step": 1345
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 0.7190412782956058,
|
1903 |
+
"grad_norm": 0.07130465564504203,
|
1904 |
+
"learning_rate": 4.431738223947252e-05,
|
1905 |
+
"loss": 1.0951,
|
1906 |
+
"step": 1350
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 0.7217043941411452,
|
1910 |
+
"grad_norm": 0.075489827909183,
|
1911 |
+
"learning_rate": 4.35473034758643e-05,
|
1912 |
+
"loss": 1.1223,
|
1913 |
+
"step": 1355
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 0.7243675099866844,
|
1917 |
+
"grad_norm": 0.07030155738463333,
|
1918 |
+
"learning_rate": 4.2782107434455054e-05,
|
1919 |
+
"loss": 1.1222,
|
1920 |
+
"step": 1360
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 0.7270306258322237,
|
1924 |
+
"grad_norm": 0.07218531177873333,
|
1925 |
+
"learning_rate": 4.202186029879195e-05,
|
1926 |
+
"loss": 1.1135,
|
1927 |
+
"step": 1365
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 0.729693741677763,
|
1931 |
+
"grad_norm": 0.07513760712282253,
|
1932 |
+
"learning_rate": 4.12666278243799e-05,
|
1933 |
+
"loss": 1.1181,
|
1934 |
+
"step": 1370
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 0.7323568575233023,
|
1938 |
+
"grad_norm": 0.07435576640285345,
|
1939 |
+
"learning_rate": 4.0516475332994383e-05,
|
1940 |
+
"loss": 1.119,
|
1941 |
+
"step": 1375
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 0.7350199733688415,
|
1945 |
+
"grad_norm": 0.07125188708419379,
|
1946 |
+
"learning_rate": 3.9771467707031615e-05,
|
1947 |
+
"loss": 1.1201,
|
1948 |
+
"step": 1380
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 0.7376830892143809,
|
1952 |
+
"grad_norm": 0.07352772721979937,
|
1953 |
+
"learning_rate": 3.903166938389664e-05,
|
1954 |
+
"loss": 1.112,
|
1955 |
+
"step": 1385
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 0.7403462050599201,
|
1959 |
+
"grad_norm": 0.07055845215545062,
|
1960 |
+
"learning_rate": 3.8297144350430144e-05,
|
1961 |
+
"loss": 1.1046,
|
1962 |
+
"step": 1390
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 0.7430093209054593,
|
1966 |
+
"grad_norm": 0.07019837722638089,
|
1967 |
+
"learning_rate": 3.756795613737388e-05,
|
1968 |
+
"loss": 1.1306,
|
1969 |
+
"step": 1395
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 0.7456724367509987,
|
1973 |
+
"grad_norm": 0.07239297940522006,
|
1974 |
+
"learning_rate": 3.684416781387589e-05,
|
1975 |
+
"loss": 1.1184,
|
1976 |
+
"step": 1400
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 0.748335552596538,
|
1980 |
+
"grad_norm": 0.07225126596889433,
|
1981 |
+
"learning_rate": 3.6125841982035536e-05,
|
1982 |
+
"loss": 1.0843,
|
1983 |
+
"step": 1405
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 0.7509986684420772,
|
1987 |
+
"grad_norm": 0.07431737331558284,
|
1988 |
+
"learning_rate": 3.5413040771488746e-05,
|
1989 |
+
"loss": 1.1145,
|
1990 |
+
"step": 1410
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 0.7536617842876165,
|
1994 |
+
"grad_norm": 0.07502128782750854,
|
1995 |
+
"learning_rate": 3.47058258340345e-05,
|
1996 |
+
"loss": 1.1114,
|
1997 |
+
"step": 1415
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 0.7563249001331558,
|
2001 |
+
"grad_norm": 0.07207529195587527,
|
2002 |
+
"learning_rate": 3.4004258338302195e-05,
|
2003 |
+
"loss": 1.116,
|
2004 |
+
"step": 1420
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 0.758988015978695,
|
2008 |
+
"grad_norm": 0.07002467859956689,
|
2009 |
+
"learning_rate": 3.3308398964461206e-05,
|
2010 |
+
"loss": 1.1198,
|
2011 |
+
"step": 1425
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 0.7616511318242344,
|
2015 |
+
"grad_norm": 0.07224058742693344,
|
2016 |
+
"learning_rate": 3.261830789897241e-05,
|
2017 |
+
"loss": 1.1367,
|
2018 |
+
"step": 1430
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 0.7643142476697736,
|
2022 |
+
"grad_norm": 0.07150872607987452,
|
2023 |
+
"learning_rate": 3.193404482938256e-05,
|
2024 |
+
"loss": 1.0982,
|
2025 |
+
"step": 1435
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 0.7669773635153129,
|
2029 |
+
"grad_norm": 0.07178998194161153,
|
2030 |
+
"learning_rate": 3.1255668939161894e-05,
|
2031 |
+
"loss": 1.1301,
|
2032 |
+
"step": 1440
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 0.7696404793608522,
|
2036 |
+
"grad_norm": 0.07537933178179766,
|
2037 |
+
"learning_rate": 3.058323890258498e-05,
|
2038 |
+
"loss": 1.0962,
|
2039 |
+
"step": 1445
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 0.7723035952063915,
|
2043 |
+
"grad_norm": 0.07189617812023931,
|
2044 |
+
"learning_rate": 2.9916812879655975e-05,
|
2045 |
+
"loss": 1.1299,
|
2046 |
+
"step": 1450
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 0.7749667110519307,
|
2050 |
+
"grad_norm": 0.07135524342299995,
|
2051 |
+
"learning_rate": 2.925644851107835e-05,
|
2052 |
+
"loss": 1.1189,
|
2053 |
+
"step": 1455
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 0.7776298268974701,
|
2057 |
+
"grad_norm": 0.06926273022163672,
|
2058 |
+
"learning_rate": 2.860220291326915e-05,
|
2059 |
+
"loss": 1.1068,
|
2060 |
+
"step": 1460
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 0.7802929427430093,
|
2064 |
+
"grad_norm": 0.07236268331467403,
|
2065 |
+
"learning_rate": 2.7954132673419143e-05,
|
2066 |
+
"loss": 1.0981,
|
2067 |
+
"step": 1465
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 0.7829560585885486,
|
2071 |
+
"grad_norm": 0.07137745473045948,
|
2072 |
+
"learning_rate": 2.7312293844598246e-05,
|
2073 |
+
"loss": 1.1045,
|
2074 |
+
"step": 1470
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 0.7856191744340879,
|
2078 |
+
"grad_norm": 0.07319916214034358,
|
2079 |
+
"learning_rate": 2.6676741940907478e-05,
|
2080 |
+
"loss": 1.1281,
|
2081 |
+
"step": 1475
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 0.7882822902796272,
|
2085 |
+
"grad_norm": 0.07414585611908868,
|
2086 |
+
"learning_rate": 2.6047531932677383e-05,
|
2087 |
+
"loss": 1.1225,
|
2088 |
+
"step": 1480
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 0.7909454061251664,
|
2092 |
+
"grad_norm": 0.07066978423343756,
|
2093 |
+
"learning_rate": 2.542471824171353e-05,
|
2094 |
+
"loss": 1.1356,
|
2095 |
+
"step": 1485
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 0.7936085219707057,
|
2099 |
+
"grad_norm": 0.07124448208473912,
|
2100 |
+
"learning_rate": 2.4808354736589523e-05,
|
2101 |
+
"loss": 1.1323,
|
2102 |
+
"step": 1490
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 0.796271637816245,
|
2106 |
+
"grad_norm": 0.07192509266254882,
|
2107 |
+
"learning_rate": 2.419849472798761e-05,
|
2108 |
+
"loss": 1.1386,
|
2109 |
+
"step": 1495
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 0.7989347536617842,
|
2113 |
+
"grad_norm": 0.07470536521159465,
|
2114 |
+
"learning_rate": 2.359519096408791e-05,
|
2115 |
+
"loss": 1.1103,
|
2116 |
+
"step": 1500
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 0.8015978695073236,
|
2120 |
+
"grad_norm": 0.07210949951260932,
|
2121 |
+
"learning_rate": 2.2998495626005957e-05,
|
2122 |
+
"loss": 1.1108,
|
2123 |
+
"step": 1505
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 0.8042609853528628,
|
2127 |
+
"grad_norm": 0.07516415250373631,
|
2128 |
+
"learning_rate": 2.240846032327949e-05,
|
2129 |
+
"loss": 1.1404,
|
2130 |
+
"step": 1510
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 0.8069241011984021,
|
2134 |
+
"grad_norm": 0.07560562529629619,
|
2135 |
+
"learning_rate": 2.1825136089404718e-05,
|
2136 |
+
"loss": 1.0935,
|
2137 |
+
"step": 1515
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 0.8095872170439414,
|
2141 |
+
"grad_norm": 0.07195974938474745,
|
2142 |
+
"learning_rate": 2.1248573377422155e-05,
|
2143 |
+
"loss": 1.1182,
|
2144 |
+
"step": 1520
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 0.8122503328894807,
|
2148 |
+
"grad_norm": 0.07250882969384367,
|
2149 |
+
"learning_rate": 2.0678822055552906e-05,
|
2150 |
+
"loss": 1.1189,
|
2151 |
+
"step": 1525
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 0.8149134487350199,
|
2155 |
+
"grad_norm": 0.0721751215640965,
|
2156 |
+
"learning_rate": 2.0115931402885458e-05,
|
2157 |
+
"loss": 1.1115,
|
2158 |
+
"step": 1530
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 0.8175765645805593,
|
2162 |
+
"grad_norm": 0.0753848259347461,
|
2163 |
+
"learning_rate": 1.955995010511338e-05,
|
2164 |
+
"loss": 1.1348,
|
2165 |
+
"step": 1535
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 0.8202396804260985,
|
2169 |
+
"grad_norm": 0.0719207373284397,
|
2170 |
+
"learning_rate": 1.901092625032448e-05,
|
2171 |
+
"loss": 1.1042,
|
2172 |
+
"step": 1540
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 0.8229027962716379,
|
2176 |
+
"grad_norm": 0.07032664869488064,
|
2177 |
+
"learning_rate": 1.84689073248414e-05,
|
2178 |
+
"loss": 1.1009,
|
2179 |
+
"step": 1545
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 0.8255659121171771,
|
2183 |
+
"grad_norm": 0.0700654057925292,
|
2184 |
+
"learning_rate": 1.7933940209114597e-05,
|
2185 |
+
"loss": 1.1269,
|
2186 |
+
"step": 1550
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 0.8282290279627164,
|
2190 |
+
"grad_norm": 0.07325193867745135,
|
2191 |
+
"learning_rate": 1.7406071173667372e-05,
|
2192 |
+
"loss": 1.1138,
|
2193 |
+
"step": 1555
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 0.8308921438082557,
|
2197 |
+
"grad_norm": 0.07059680065497263,
|
2198 |
+
"learning_rate": 1.6885345875093918e-05,
|
2199 |
+
"loss": 1.1202,
|
2200 |
+
"step": 1560
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 0.833555259653795,
|
2204 |
+
"grad_norm": 0.06973843219886788,
|
2205 |
+
"learning_rate": 1.6371809352110447e-05,
|
2206 |
+
"loss": 1.109,
|
2207 |
+
"step": 1565
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 0.8362183754993342,
|
2211 |
+
"grad_norm": 0.07028429615451927,
|
2212 |
+
"learning_rate": 1.5865506021659516e-05,
|
2213 |
+
"loss": 1.1422,
|
2214 |
+
"step": 1570
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 0.8388814913448736,
|
2218 |
+
"grad_norm": 0.07070004444035057,
|
2219 |
+
"learning_rate": 1.5366479675068435e-05,
|
2220 |
+
"loss": 1.1139,
|
2221 |
+
"step": 1575
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 0.8415446071904128,
|
2225 |
+
"grad_norm": 0.06902941716806588,
|
2226 |
+
"learning_rate": 1.4874773474261638e-05,
|
2227 |
+
"loss": 1.1179,
|
2228 |
+
"step": 1580
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 0.844207723035952,
|
2232 |
+
"grad_norm": 0.07227429481260465,
|
2233 |
+
"learning_rate": 1.4390429948027428e-05,
|
2234 |
+
"loss": 1.1156,
|
2235 |
+
"step": 1585
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 0.8468708388814914,
|
2239 |
+
"grad_norm": 0.07067313252269349,
|
2240 |
+
"learning_rate": 1.3913490988339718e-05,
|
2241 |
+
"loss": 1.1209,
|
2242 |
+
"step": 1590
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 0.8495339547270306,
|
2246 |
+
"grad_norm": 0.07249077076436629,
|
2247 |
+
"learning_rate": 1.3443997846734535e-05,
|
2248 |
+
"loss": 1.1303,
|
2249 |
+
"step": 1595
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 0.8521970705725699,
|
2253 |
+
"grad_norm": 0.07121930519374212,
|
2254 |
+
"learning_rate": 1.2981991130742211e-05,
|
2255 |
+
"loss": 1.1069,
|
2256 |
+
"step": 1600
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 0.8548601864181092,
|
2260 |
+
"grad_norm": 0.06967637890151408,
|
2261 |
+
"learning_rate": 1.2527510800375043e-05,
|
2262 |
+
"loss": 1.1007,
|
2263 |
+
"step": 1605
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 0.8575233022636485,
|
2267 |
+
"grad_norm": 0.07205061200000021,
|
2268 |
+
"learning_rate": 1.20805961646711e-05,
|
2269 |
+
"loss": 1.1199,
|
2270 |
+
"step": 1610
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 0.8601864181091877,
|
2274 |
+
"grad_norm": 0.07129006954483187,
|
2275 |
+
"learning_rate": 1.1641285878294372e-05,
|
2276 |
+
"loss": 1.1054,
|
2277 |
+
"step": 1615
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 0.8628495339547271,
|
2281 |
+
"grad_norm": 0.07540152645446788,
|
2282 |
+
"learning_rate": 1.1209617938191307e-05,
|
2283 |
+
"loss": 1.1032,
|
2284 |
+
"step": 1620
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 0.8655126498002663,
|
2288 |
+
"grad_norm": 0.07238275344561401,
|
2289 |
+
"learning_rate": 1.0785629680304432e-05,
|
2290 |
+
"loss": 1.1246,
|
2291 |
+
"step": 1625
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 0.8681757656458056,
|
2295 |
+
"grad_norm": 0.07120538294411066,
|
2296 |
+
"learning_rate": 1.0369357776343103e-05,
|
2297 |
+
"loss": 1.0932,
|
2298 |
+
"step": 1630
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 0.8708388814913449,
|
2302 |
+
"grad_norm": 0.07125849884630578,
|
2303 |
+
"learning_rate": 9.960838230611635e-06,
|
2304 |
+
"loss": 1.0728,
|
2305 |
+
"step": 1635
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 0.8735019973368842,
|
2309 |
+
"grad_norm": 0.07387109327159767,
|
2310 |
+
"learning_rate": 9.560106376895306e-06,
|
2311 |
+
"loss": 1.1275,
|
2312 |
+
"step": 1640
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 0.8761651131824234,
|
2316 |
+
"grad_norm": 0.08867831950654811,
|
2317 |
+
"learning_rate": 9.167196875404094e-06,
|
2318 |
+
"loss": 1.1134,
|
2319 |
+
"step": 1645
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 0.8788282290279628,
|
2323 |
+
"grad_norm": 0.07206488784580595,
|
2324 |
+
"learning_rate": 8.782143709775015e-06,
|
2325 |
+
"loss": 1.109,
|
2326 |
+
"step": 1650
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 0.881491344873502,
|
2330 |
+
"grad_norm": 0.07037650282884113,
|
2331 |
+
"learning_rate": 8.40498018413266e-06,
|
2332 |
+
"loss": 1.0862,
|
2333 |
+
"step": 1655
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 0.8841544607190412,
|
2337 |
+
"grad_norm": 0.07329640323219759,
|
2338 |
+
"learning_rate": 8.035738920208714e-06,
|
2339 |
+
"loss": 1.1539,
|
2340 |
+
"step": 1660
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 0.8868175765645806,
|
2344 |
+
"grad_norm": 0.07238224996992595,
|
2345 |
+
"learning_rate": 7.67445185452046e-06,
|
2346 |
+
"loss": 1.14,
|
2347 |
+
"step": 1665
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 0.8894806924101198,
|
2351 |
+
"grad_norm": 0.0709614207657161,
|
2352 |
+
"learning_rate": 7.321150235608399e-06,
|
2353 |
+
"loss": 1.1084,
|
2354 |
+
"step": 1670
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 0.8921438082556591,
|
2358 |
+
"grad_norm": 0.0726275702018448,
|
2359 |
+
"learning_rate": 6.9758646213336165e-06,
|
2360 |
+
"loss": 1.1227,
|
2361 |
+
"step": 1675
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 0.8948069241011984,
|
2365 |
+
"grad_norm": 0.07467788565919785,
|
2366 |
+
"learning_rate": 6.6386248762347004e-06,
|
2367 |
+
"loss": 1.1135,
|
2368 |
+
"step": 1680
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 0.8974700399467377,
|
2372 |
+
"grad_norm": 0.07203922681266979,
|
2373 |
+
"learning_rate": 6.309460168944692e-06,
|
2374 |
+
"loss": 1.1071,
|
2375 |
+
"step": 1685
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 0.9001331557922769,
|
2379 |
+
"grad_norm": 0.07114745434226347,
|
2380 |
+
"learning_rate": 5.988398969668285e-06,
|
2381 |
+
"loss": 1.1248,
|
2382 |
+
"step": 1690
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 0.9027962716378163,
|
2386 |
+
"grad_norm": 0.07111413913413335,
|
2387 |
+
"learning_rate": 5.6754690477192396e-06,
|
2388 |
+
"loss": 1.0872,
|
2389 |
+
"step": 1695
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 0.9054593874833555,
|
2393 |
+
"grad_norm": 0.07072306177993802,
|
2394 |
+
"learning_rate": 5.370697469118713e-06,
|
2395 |
+
"loss": 1.0824,
|
2396 |
+
"step": 1700
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 0.9081225033288948,
|
2400 |
+
"grad_norm": 0.07185114603062198,
|
2401 |
+
"learning_rate": 5.074110594254133e-06,
|
2402 |
+
"loss": 1.107,
|
2403 |
+
"step": 1705
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 0.9107856191744341,
|
2407 |
+
"grad_norm": 0.07285137048040193,
|
2408 |
+
"learning_rate": 4.78573407559928e-06,
|
2409 |
+
"loss": 1.1173,
|
2410 |
+
"step": 1710
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 0.9134487350199734,
|
2414 |
+
"grad_norm": 0.07013700832744958,
|
2415 |
+
"learning_rate": 4.5055928554955665e-06,
|
2416 |
+
"loss": 1.116,
|
2417 |
+
"step": 1715
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 0.9161118508655126,
|
2421 |
+
"grad_norm": 0.06933918410316768,
|
2422 |
+
"learning_rate": 4.233711163994669e-06,
|
2423 |
+
"loss": 1.1038,
|
2424 |
+
"step": 1720
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 0.918774966711052,
|
2428 |
+
"grad_norm": 0.072978505928099,
|
2429 |
+
"learning_rate": 3.970112516762825e-06,
|
2430 |
+
"loss": 1.104,
|
2431 |
+
"step": 1725
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 0.9214380825565912,
|
2435 |
+
"grad_norm": 0.07096666064833597,
|
2436 |
+
"learning_rate": 3.7148197130469576e-06,
|
2437 |
+
"loss": 1.1169,
|
2438 |
+
"step": 1730
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 0.9241011984021305,
|
2442 |
+
"grad_norm": 0.07024364472390462,
|
2443 |
+
"learning_rate": 3.467854833702644e-06,
|
2444 |
+
"loss": 1.1051,
|
2445 |
+
"step": 1735
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 0.9267643142476698,
|
2449 |
+
"grad_norm": 0.0725948329149083,
|
2450 |
+
"learning_rate": 3.229239239284354e-06,
|
2451 |
+
"loss": 1.1257,
|
2452 |
+
"step": 1740
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 0.929427430093209,
|
2456 |
+
"grad_norm": 0.07215970561357568,
|
2457 |
+
"learning_rate": 2.9989935681979164e-06,
|
2458 |
+
"loss": 1.1262,
|
2459 |
+
"step": 1745
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 0.9320905459387483,
|
2463 |
+
"grad_norm": 0.0737761172543898,
|
2464 |
+
"learning_rate": 2.777137734915403e-06,
|
2465 |
+
"loss": 1.1091,
|
2466 |
+
"step": 1750
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 0.9347536617842876,
|
2470 |
+
"grad_norm": 0.0714550851543958,
|
2471 |
+
"learning_rate": 2.563690928252749e-06,
|
2472 |
+
"loss": 1.1283,
|
2473 |
+
"step": 1755
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 0.9374167776298269,
|
2477 |
+
"grad_norm": 0.07157410935337963,
|
2478 |
+
"learning_rate": 2.358671609710017e-06,
|
2479 |
+
"loss": 1.1239,
|
2480 |
+
"step": 1760
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 0.9400798934753661,
|
2484 |
+
"grad_norm": 0.0715459262435765,
|
2485 |
+
"learning_rate": 2.1620975118746835e-06,
|
2486 |
+
"loss": 1.1283,
|
2487 |
+
"step": 1765
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 0.9427430093209055,
|
2491 |
+
"grad_norm": 0.07286793622379441,
|
2492 |
+
"learning_rate": 1.9739856368878096e-06,
|
2493 |
+
"loss": 1.1443,
|
2494 |
+
"step": 1770
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 0.9454061251664447,
|
2498 |
+
"grad_norm": 0.07254981755374063,
|
2499 |
+
"learning_rate": 1.794352254973597e-06,
|
2500 |
+
"loss": 1.0752,
|
2501 |
+
"step": 1775
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 0.948069241011984,
|
2505 |
+
"grad_norm": 0.07583062173231125,
|
2506 |
+
"learning_rate": 1.6232129030320453e-06,
|
2507 |
+
"loss": 1.1011,
|
2508 |
+
"step": 1780
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 0.9507323568575233,
|
2512 |
+
"grad_norm": 0.07237886602221254,
|
2513 |
+
"learning_rate": 1.4605823832951948e-06,
|
2514 |
+
"loss": 1.1063,
|
2515 |
+
"step": 1785
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 0.9533954727030626,
|
2519 |
+
"grad_norm": 0.07194863306691765,
|
2520 |
+
"learning_rate": 1.3064747620468054e-06,
|
2521 |
+
"loss": 1.0914,
|
2522 |
+
"step": 1790
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 0.9560585885486018,
|
2526 |
+
"grad_norm": 0.07189535177125876,
|
2527 |
+
"learning_rate": 1.1609033684057857e-06,
|
2528 |
+
"loss": 1.1048,
|
2529 |
+
"step": 1795
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 0.9587217043941412,
|
2533 |
+
"grad_norm": 0.07110020697306084,
|
2534 |
+
"learning_rate": 1.0238807931732487e-06,
|
2535 |
+
"loss": 1.1219,
|
2536 |
+
"step": 1800
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 0.9613848202396804,
|
2540 |
+
"grad_norm": 0.07370460890021728,
|
2541 |
+
"learning_rate": 8.95418887743571e-07,
|
2542 |
+
"loss": 1.1317,
|
2543 |
+
"step": 1805
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 0.9640479360852197,
|
2547 |
+
"grad_norm": 0.07207691237434337,
|
2548 |
+
"learning_rate": 7.75528763079314e-07,
|
2549 |
+
"loss": 1.1126,
|
2550 |
+
"step": 1810
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 0.966711051930759,
|
2554 |
+
"grad_norm": 0.07440378502380679,
|
2555 |
+
"learning_rate": 6.642207887502027e-07,
|
2556 |
+
"loss": 1.1262,
|
2557 |
+
"step": 1815
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 0.9693741677762983,
|
2561 |
+
"grad_norm": 0.06937782526053558,
|
2562 |
+
"learning_rate": 5.615045920362549e-07,
|
2563 |
+
"loss": 1.1191,
|
2564 |
+
"step": 1820
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 0.9720372836218375,
|
2568 |
+
"grad_norm": 0.07098237504174736,
|
2569 |
+
"learning_rate": 4.673890570951023e-07,
|
2570 |
+
"loss": 1.1218,
|
2571 |
+
"step": 1825
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 0.9747003994673769,
|
2575 |
+
"grad_norm": 0.07181367245928705,
|
2576 |
+
"learning_rate": 3.8188232419352764e-07,
|
2577 |
+
"loss": 1.1514,
|
2578 |
+
"step": 1830
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 0.9773635153129161,
|
2582 |
+
"grad_norm": 0.07128549274991237,
|
2583 |
+
"learning_rate": 3.049917890034837e-07,
|
2584 |
+
"loss": 1.0945,
|
2585 |
+
"step": 1835
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 0.9800266311584553,
|
2589 |
+
"grad_norm": 0.07161530878264005,
|
2590 |
+
"learning_rate": 2.3672410196232675e-07,
|
2591 |
+
"loss": 1.1056,
|
2592 |
+
"step": 1840
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 0.9826897470039947,
|
2596 |
+
"grad_norm": 0.07090950092701657,
|
2597 |
+
"learning_rate": 1.7708516769769924e-07,
|
2598 |
+
"loss": 1.1109,
|
2599 |
+
"step": 1845
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 0.9853528628495339,
|
2603 |
+
"grad_norm": 0.07140303003446424,
|
2604 |
+
"learning_rate": 1.2608014451672702e-07,
|
2605 |
+
"loss": 1.1252,
|
2606 |
+
"step": 1850
|
2607 |
+
},
|
2608 |
+
{
|
2609 |
+
"epoch": 0.9880159786950732,
|
2610 |
+
"grad_norm": 0.07103068810909154,
|
2611 |
+
"learning_rate": 8.371344395996516e-08,
|
2612 |
+
"loss": 1.1255,
|
2613 |
+
"step": 1855
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 0.9906790945406125,
|
2617 |
+
"grad_norm": 0.07153307782175972,
|
2618 |
+
"learning_rate": 4.998873041975882e-08,
|
2619 |
+
"loss": 1.1365,
|
2620 |
+
"step": 1860
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 0.9933422103861518,
|
2624 |
+
"grad_norm": 0.06996817906726589,
|
2625 |
+
"learning_rate": 2.490892082331886e-08,
|
2626 |
+
"loss": 1.1142,
|
2627 |
+
"step": 1865
|
2628 |
+
},
|
2629 |
+
{
|
2630 |
+
"epoch": 0.996005326231691,
|
2631 |
+
"grad_norm": 0.07126895143317796,
|
2632 |
+
"learning_rate": 8.476184380468155e-09,
|
2633 |
+
"loss": 1.1091,
|
2634 |
+
"step": 1870
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 0.9986684420772304,
|
2638 |
+
"grad_norm": 0.0706408297239244,
|
2639 |
+
"learning_rate": 6.919423959805826e-10,
|
2640 |
+
"loss": 1.1206,
|
2641 |
+
"step": 1875
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"epoch": 0.9997336884154461,
|
2645 |
+
"eval_loss": 1.118857979774475,
|
2646 |
+
"eval_runtime": 1652.5253,
|
2647 |
+
"eval_samples_per_second": 8.045,
|
2648 |
+
"eval_steps_per_second": 0.503,
|
2649 |
+
"step": 1877
|
2650 |
+
},
|
2651 |
+
{
|
2652 |
+
"epoch": 0.9997336884154461,
|
2653 |
+
"step": 1877,
|
2654 |
+
"total_flos": 2.979798729936077e+16,
|
2655 |
+
"train_loss": 1.1429596533467938,
|
2656 |
+
"train_runtime": 55739.5179,
|
2657 |
+
"train_samples_per_second": 2.156,
|
2658 |
+
"train_steps_per_second": 0.034
|
2659 |
+
}
|
2660 |
+
],
|
2661 |
+
"logging_steps": 5,
|
2662 |
+
"max_steps": 1877,
|
2663 |
+
"num_input_tokens_seen": 0,
|
2664 |
+
"num_train_epochs": 1,
|
2665 |
+
"save_steps": 100,
|
2666 |
+
"stateful_callbacks": {
|
2667 |
+
"TrainerControl": {
|
2668 |
+
"args": {
|
2669 |
+
"should_epoch_stop": false,
|
2670 |
+
"should_evaluate": false,
|
2671 |
+
"should_log": false,
|
2672 |
+
"should_save": true,
|
2673 |
+
"should_training_stop": false
|
2674 |
+
},
|
2675 |
+
"attributes": {}
|
2676 |
+
}
|
2677 |
+
},
|
2678 |
+
"total_flos": 2.979798729936077e+16,
|
2679 |
+
"train_batch_size": 8,
|
2680 |
+
"trial_name": null,
|
2681 |
+
"trial_params": null
|
2682 |
+
}
|