File size: 1,874 Bytes
306a855
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
---
tags:
- generated_from_trainer
model-index:
- name: myBit-Llama2-jp-127M-test-20
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# myBit-Llama2-jp-127M-test-20

This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.4890

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.0024
- train_batch_size: 96
- eval_batch_size: 96
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: polynomial
- lr_scheduler_warmup_steps: 500
- num_epochs: 1

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 6.9256        | 0.07  | 200  | 4.9236          |
| 4.6058        | 0.15  | 400  | 4.4149          |
| 4.3066        | 0.22  | 600  | 4.2187          |
| 4.126         | 0.29  | 800  | 4.0780          |
| 4.0117        | 0.37  | 1000 | 3.9963          |
| 3.9295        | 0.44  | 1200 | 3.9215          |
| 3.8601        | 0.51  | 1400 | 3.8477          |
| 3.7821        | 0.59  | 1600 | 3.7748          |
| 3.7065        | 0.66  | 1800 | 3.7078          |
| 3.6407        | 0.73  | 2000 | 3.6518          |
| 3.5957        | 0.8   | 2200 | 3.5923          |
| 3.5302        | 0.88  | 2400 | 3.5368          |
| 3.482         | 0.95  | 2600 | 3.4890          |


### Framework versions

- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2