File size: 3,148 Bytes
10b206f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
---
license: other
base_model: yahma/llama-7b-hf
tags:
- generated_from_trainer
model-index:
- name: V0305B1
  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. -->

# V0305B1

This model is a fine-tuned version of [yahma/llama-7b-hf](https://huggingface.co/yahma/llama-7b-hf) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0747

## 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: 0.0003
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 32
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine_with_restarts
- lr_scheduler_warmup_steps: 20
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 2.1466        | 0.09  | 10   | 1.4267          |
| 0.6179        | 0.17  | 20   | 0.1910          |
| 0.1703        | 0.26  | 30   | 0.1573          |
| 0.1533        | 0.34  | 40   | 0.1496          |
| 0.1508        | 0.43  | 50   | 0.1499          |
| 0.1562        | 0.51  | 60   | 0.1490          |
| 0.1515        | 0.6   | 70   | 0.1506          |
| 0.1531        | 0.68  | 80   | 0.1493          |
| 0.1481        | 0.77  | 90   | 0.1490          |
| 0.1502        | 0.85  | 100  | 0.1408          |
| 0.1455        | 0.94  | 110  | 0.1351          |
| 0.1343        | 1.02  | 120  | 0.1112          |
| 0.129         | 1.11  | 130  | 0.1147          |
| 0.1165        | 1.19  | 140  | 0.1065          |
| 0.1152        | 1.28  | 150  | 0.0977          |
| 0.1095        | 1.37  | 160  | 0.0959          |
| 0.11          | 1.45  | 170  | 0.0942          |
| 0.1006        | 1.54  | 180  | 0.0933          |
| 0.1017        | 1.62  | 190  | 0.0892          |
| 0.1023        | 1.71  | 200  | 0.0885          |
| 0.1009        | 1.79  | 210  | 0.0840          |
| 0.0979        | 1.88  | 220  | 0.0806          |
| 0.0932        | 1.96  | 230  | 0.0830          |
| 0.0877        | 2.05  | 240  | 0.0818          |
| 0.0839        | 2.13  | 250  | 0.0821          |
| 0.0815        | 2.22  | 260  | 0.0783          |
| 0.0783        | 2.3   | 270  | 0.0803          |
| 0.0846        | 2.39  | 280  | 0.0773          |
| 0.0862        | 2.47  | 290  | 0.0753          |
| 0.0856        | 2.56  | 300  | 0.0752          |
| 0.0893        | 2.65  | 310  | 0.0755          |
| 0.0821        | 2.73  | 320  | 0.0750          |
| 0.0821        | 2.82  | 330  | 0.0747          |
| 0.0791        | 2.9   | 340  | 0.0747          |
| 0.0796        | 2.99  | 350  | 0.0747          |


### Framework versions

- Transformers 4.36.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.14.1