File size: 3,097 Bytes
81af1d2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bff0904
81af1d2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bff0904
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
81af1d2
 
 
 
 
 
 
 
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: Phi0503HMA4
  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. -->

# Phi0503HMA4

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: 0.0153

## 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: 80
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 4.327         | 0.09  | 10   | 0.8261          |
| 0.4359        | 0.18  | 20   | 0.2664          |
| 0.2914        | 0.27  | 30   | 0.2506          |
| 0.2542        | 0.36  | 40   | 0.2504          |
| 0.2976        | 0.45  | 50   | 0.3288          |
| 0.3986        | 0.54  | 60   | 0.2542          |
| 2.3932        | 0.63  | 70   | 0.2711          |
| 1.9638        | 0.73  | 80   | 4.8527          |
| 3.6131        | 0.82  | 90   | 1.5739          |
| 1.1269        | 0.91  | 100  | 0.7721          |
| 0.4633        | 1.0   | 110  | 0.3521          |
| 0.2947        | 1.09  | 120  | 0.2266          |
| 0.2156        | 1.18  | 130  | 0.1790          |
| 0.2026        | 1.27  | 140  | 0.1381          |
| 0.1618        | 1.36  | 150  | 0.2401          |
| 0.1723        | 1.45  | 160  | 0.1317          |
| 0.1256        | 1.54  | 170  | 0.0996          |
| 0.1171        | 1.63  | 180  | 0.0833          |
| 0.0767        | 1.72  | 190  | 0.0579          |
| 0.0578        | 1.81  | 200  | 0.0514          |
| 0.0497        | 1.9   | 210  | 0.0414          |
| 0.0456        | 1.99  | 220  | 0.0376          |
| 0.042         | 2.08  | 230  | 0.0374          |
| 0.0435        | 2.18  | 240  | 0.0295          |
| 0.0429        | 2.27  | 250  | 0.0304          |
| 0.0396        | 2.36  | 260  | 0.0243          |
| 0.0305        | 2.45  | 270  | 0.0214          |
| 0.0277        | 2.54  | 280  | 0.0191          |
| 0.0205        | 2.63  | 290  | 0.0186          |
| 0.0228        | 2.72  | 300  | 0.0165          |
| 0.0202        | 2.81  | 310  | 0.0157          |
| 0.0236        | 2.9   | 320  | 0.0155          |
| 0.0196        | 2.99  | 330  | 0.0153          |


### Framework versions

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