File size: 3,140 Bytes
4e5ac49
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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: mit
base_model: microsoft/phi-2
tags:
- generated_from_trainer
model-index:
- name: V0309B1
  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. -->

# V0309B1

This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0618

## 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 |
|:-------------:|:-----:|:----:|:---------------:|
| 2.5503        | 0.09  | 10   | 1.8060          |
| 0.926         | 0.17  | 20   | 0.1557          |
| 0.1416        | 0.26  | 30   | 0.0878          |
| 0.1055        | 0.34  | 40   | 0.0739          |
| 0.1001        | 0.43  | 50   | 0.0704          |
| 0.0863        | 0.51  | 60   | 0.0660          |
| 0.0819        | 0.6   | 70   | 0.0676          |
| 0.0838        | 0.68  | 80   | 0.0638          |
| 0.0736        | 0.77  | 90   | 0.0636          |
| 0.0766        | 0.85  | 100  | 0.0610          |
| 0.0787        | 0.94  | 110  | 0.0607          |
| 0.076         | 1.02  | 120  | 0.0604          |
| 0.0738        | 1.11  | 130  | 0.0619          |
| 0.0711        | 1.19  | 140  | 0.0583          |
| 0.068         | 1.28  | 150  | 0.0573          |
| 0.0696        | 1.37  | 160  | 0.0606          |
| 0.068         | 1.45  | 170  | 0.0610          |
| 0.0637        | 1.54  | 180  | 0.0596          |
| 0.0678        | 1.62  | 190  | 0.0583          |
| 0.066         | 1.71  | 200  | 0.0594          |
| 0.0679        | 1.79  | 210  | 0.0586          |
| 0.0632        | 1.88  | 220  | 0.0605          |
| 0.0606        | 1.96  | 230  | 0.0606          |
| 0.0622        | 2.05  | 240  | 0.0611          |
| 0.0578        | 2.13  | 250  | 0.0610          |
| 0.0562        | 2.22  | 260  | 0.0627          |
| 0.0507        | 2.3   | 270  | 0.0659          |
| 0.0615        | 2.39  | 280  | 0.0642          |
| 0.06          | 2.47  | 290  | 0.0627          |
| 0.0588        | 2.56  | 300  | 0.0619          |
| 0.0626        | 2.65  | 310  | 0.0614          |
| 0.053         | 2.73  | 320  | 0.0618          |
| 0.0567        | 2.82  | 330  | 0.0616          |
| 0.0525        | 2.9   | 340  | 0.0619          |
| 0.057         | 2.99  | 350  | 0.0618          |


### Framework versions

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