File size: 3,104 Bytes
55dc937
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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: PHI30512HMAB7H
  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. -->

# PHI30512HMAB7H

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: 0.1660

## 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: 100
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 4.9249        | 0.09  | 10   | 2.4664          |
| 1.3773        | 0.18  | 20   | 0.4052          |
| 0.394         | 0.27  | 30   | 0.3430          |
| 2.5727        | 0.36  | 40   | 0.3864          |
| 0.2708        | 0.45  | 50   | 0.1622          |
| 0.1648        | 0.54  | 60   | 0.1491          |
| 0.1254        | 0.63  | 70   | 0.1389          |
| 0.1202        | 0.73  | 80   | 0.1068          |
| 0.092         | 0.82  | 90   | 0.0929          |
| 0.097         | 0.91  | 100  | 0.0817          |
| 0.0815        | 1.0   | 110  | 0.0795          |
| 0.0971        | 1.09  | 120  | 0.1372          |
| 0.3829        | 1.18  | 130  | 0.2014          |
| 0.2626        | 1.27  | 140  | 0.1422          |
| 0.1206        | 1.36  | 150  | 0.1053          |
| 2.8589        | 1.45  | 160  | 2.3060          |
| 1.7749        | 1.54  | 170  | 1.1543          |
| 0.8021        | 1.63  | 180  | 0.5702          |
| 0.464         | 1.72  | 190  | 0.3593          |
| 0.3491        | 1.81  | 200  | 0.3201          |
| 0.3161        | 1.9   | 210  | 0.3053          |
| 0.2851        | 1.99  | 220  | 0.2623          |
| 0.2537        | 2.08  | 230  | 0.2722          |
| 0.244         | 2.18  | 240  | 0.1909          |
| 0.1926        | 2.27  | 250  | 0.1829          |
| 0.1805        | 2.36  | 260  | 0.1712          |
| 0.1712        | 2.45  | 270  | 0.1778          |
| 0.1665        | 2.54  | 280  | 0.1669          |
| 0.1733        | 2.63  | 290  | 0.1705          |
| 0.173         | 2.72  | 300  | 0.1668          |
| 0.1687        | 2.81  | 310  | 0.1676          |
| 0.1673        | 2.9   | 320  | 0.1662          |
| 0.1684        | 2.99  | 330  | 0.1660          |


### Framework versions

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