File size: 3,043 Bytes
30cd257
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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-2
tags:
- generated_from_trainer
model-index:
- name: V0424HMA9
  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. -->

# V0424HMA9

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: 0.0624

## 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 |
|:-------------:|:-----:|:----:|:---------------:|
| 1.7629        | 0.09  | 10   | 0.3668          |
| 0.1867        | 0.18  | 20   | 0.1122          |
| 0.1113        | 0.27  | 30   | 0.0923          |
| 0.1065        | 0.36  | 40   | 0.0843          |
| 0.081         | 0.45  | 50   | 0.0724          |
| 0.1068        | 0.54  | 60   | 0.0807          |
| 0.0797        | 0.63  | 70   | 0.0752          |
| 0.0773        | 0.73  | 80   | 0.0826          |
| 0.0898        | 0.82  | 90   | 0.0796          |
| 0.0923        | 0.91  | 100  | 0.0766          |
| 0.0803        | 1.0   | 110  | 0.0688          |
| 0.0663        | 1.09  | 120  | 0.0683          |
| 0.0629        | 1.18  | 130  | 0.0847          |
| 0.073         | 1.27  | 140  | 0.0767          |
| 0.0691        | 1.36  | 150  | 0.0683          |
| 0.0769        | 1.45  | 160  | 0.0649          |
| 0.0648        | 1.54  | 170  | 0.0673          |
| 0.0697        | 1.63  | 180  | 0.0685          |
| 0.0622        | 1.72  | 190  | 0.0604          |
| 0.0677        | 1.81  | 200  | 0.0656          |
| 0.0571        | 1.9   | 210  | 0.0620          |
| 0.0534        | 1.99  | 220  | 0.0579          |
| 0.0382        | 2.08  | 230  | 0.0640          |
| 0.036         | 2.18  | 240  | 0.0711          |
| 0.0345        | 2.27  | 250  | 0.0664          |
| 0.0303        | 2.36  | 260  | 0.0660          |
| 0.0354        | 2.45  | 270  | 0.0670          |
| 0.0336        | 2.54  | 280  | 0.0653          |
| 0.0318        | 2.63  | 290  | 0.0620          |
| 0.0322        | 2.72  | 300  | 0.0622          |
| 0.035         | 2.81  | 310  | 0.0627          |
| 0.0332        | 2.9   | 320  | 0.0626          |
| 0.0344        | 2.99  | 330  | 0.0624          |


### Framework versions

- Transformers 4.36.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.18.0
- Tokenizers 0.14.1