File size: 3,099 Bytes
bca69fb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5d10862
bca69fb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5d10862
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bca69fb
 
 
 
 
 
 
 
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
---
license: mit
base_model: microsoft/Phi-3-mini-4k-instruct
tags:
- generated_from_trainer
model-index:
- name: Phi0503HMA18
  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. -->

# Phi0503HMA18

This model is a fine-tuned version of [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0803

## 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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- 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: 60
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 3.9264        | 0.09  | 10   | 0.4060          |
| 0.5196        | 0.18  | 20   | 2.6047          |
| 0.5297        | 0.27  | 30   | 0.2263          |
| 0.2143        | 0.36  | 40   | 0.1990          |
| 0.2823        | 0.45  | 50   | 0.2488          |
| 0.2513        | 0.54  | 60   | 0.1911          |
| 0.1606        | 0.63  | 70   | 0.1463          |
| 0.1446        | 0.73  | 80   | 0.1406          |
| 0.1202        | 0.82  | 90   | 0.1288          |
| 0.1229        | 0.91  | 100  | 0.1081          |
| 0.1123        | 1.0   | 110  | 0.1439          |
| 0.123         | 1.09  | 120  | 0.1062          |
| 0.0765        | 1.18  | 130  | 0.0812          |
| 0.0736        | 1.27  | 140  | 0.0723          |
| 0.0629        | 1.36  | 150  | 0.0730          |
| 0.0554        | 1.45  | 160  | 0.0738          |
| 0.0532        | 1.54  | 170  | 0.0671          |
| 0.0595        | 1.63  | 180  | 0.0657          |
| 0.0594        | 1.72  | 190  | 0.0681          |
| 0.0613        | 1.81  | 200  | 0.0624          |
| 0.0488        | 1.9   | 210  | 0.0623          |
| 0.0576        | 1.99  | 220  | 0.0607          |
| 0.0284        | 2.08  | 230  | 0.0712          |
| 0.0171        | 2.18  | 240  | 0.1021          |
| 0.0287        | 2.27  | 250  | 0.0831          |
| 0.0209        | 2.36  | 260  | 0.0753          |
| 0.0229        | 2.45  | 270  | 0.0752          |
| 0.0209        | 2.54  | 280  | 0.0759          |
| 0.0206        | 2.63  | 290  | 0.0773          |
| 0.0199        | 2.72  | 300  | 0.0788          |
| 0.0162        | 2.81  | 310  | 0.0796          |
| 0.0181        | 2.9   | 320  | 0.0802          |
| 0.0213        | 2.99  | 330  | 0.0803          |


### Framework versions

- Transformers 4.36.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.14.0