File size: 2,643 Bytes
8e90cf4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
language:
- ar
library_name: peft
tags:
- generated_from_trainer
datasets:
- dalyaa/darebah2400
base_model: microsoft/phi-2
model-index:
- name: phi-2
  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. -->

# phi-2

This model is a fine-tuned version of [microsoftl](https://huggingface.co/microsoftl) on the dalyaa/darebah2400 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8345

## 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: 2.5e-05
- train_batch_size: 2
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 5
- training_steps: 2500

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.1815        | 0.4   | 100  | 1.1010          |
| 0.984         | 0.8   | 200  | 0.9893          |
| 0.8706        | 1.2   | 300  | 0.9545          |
| 0.8095        | 1.6   | 400  | 0.9203          |
| 0.767         | 2.0   | 500  | 0.9030          |
| 0.7439        | 2.4   | 600  | 0.8878          |
| 0.7163        | 2.8   | 700  | 0.8716          |
| 0.671         | 3.2   | 800  | 0.8679          |
| 0.679         | 3.6   | 900  | 0.8587          |
| 0.6585        | 4.0   | 1000 | 0.8586          |
| 0.6528        | 4.4   | 1100 | 0.8535          |
| 0.6238        | 4.8   | 1200 | 0.8501          |
| 0.6041        | 5.2   | 1300 | 0.8434          |
| 0.6011        | 5.6   | 1400 | 0.8471          |
| 0.6386        | 6.0   | 1500 | 0.8379          |
| 0.6288        | 6.4   | 1600 | 0.8374          |
| 0.5823        | 6.8   | 1700 | 0.8409          |
| 0.6074        | 7.2   | 1800 | 0.8363          |
| 0.5944        | 7.6   | 1900 | 0.8367          |
| 0.5918        | 8.0   | 2000 | 0.8412          |
| 0.5931        | 8.4   | 2100 | 0.8343          |
| 0.5624        | 8.8   | 2200 | 0.8363          |
| 0.5709        | 9.2   | 2300 | 0.8354          |
| 0.5783        | 9.6   | 2400 | 0.8354          |
| 0.5933        | 10.0  | 2500 | 0.8345          |


### Framework versions

- PEFT 0.8.2
- Transformers 4.38.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1