File size: 3,040 Bytes
576e590
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d6bd9d8
576e590
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d6bd9d8
576e590
 
 
 
 
 
d6bd9d8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
576e590
 
 
 
 
 
 
 
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: gemma
base_model: google/gemma-2b
tags:
- generated_from_trainer
model-index:
- name: G0428B2
  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. -->

# G0428B2

This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1291

## 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: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 32
- 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 |
|:-------------:|:-----:|:----:|:---------------:|
| 2.2513        | 0.09  | 10   | 1.9189          |
| 1.9252        | 0.18  | 20   | 1.9019          |
| 1.8761        | 0.27  | 30   | 1.7972          |
| 1.7045        | 0.36  | 40   | 1.5332          |
| 1.348         | 0.45  | 50   | 1.0846          |
| 0.9036        | 0.54  | 60   | 0.4970          |
| 0.3466        | 0.63  | 70   | 0.2054          |
| 0.1888        | 0.73  | 80   | 0.1562          |
| 0.1458        | 0.82  | 90   | 0.1490          |
| 0.1531        | 0.91  | 100  | 0.1478          |
| 0.1561        | 1.0   | 110  | 0.1477          |
| 0.142         | 1.09  | 120  | 0.1474          |
| 0.1687        | 1.18  | 130  | 0.1463          |
| 0.1426        | 1.27  | 140  | 0.1451          |
| 0.1577        | 1.36  | 150  | 0.1434          |
| 0.1386        | 1.45  | 160  | 0.1419          |
| 0.136         | 1.54  | 170  | 0.1397          |
| 0.135         | 1.63  | 180  | 0.1385          |
| 0.1489        | 1.72  | 190  | 0.1377          |
| 0.146         | 1.81  | 200  | 0.1349          |
| 0.1367        | 1.9   | 210  | 0.1340          |
| 0.1347        | 1.99  | 220  | 0.1338          |
| 0.1317        | 2.08  | 230  | 0.1318          |
| 0.1554        | 2.18  | 240  | 0.1309          |
| 0.1285        | 2.27  | 250  | 0.1308          |
| 0.1328        | 2.36  | 260  | 0.1310          |
| 0.1354        | 2.45  | 270  | 0.1305          |
| 0.1324        | 2.54  | 280  | 0.1301          |
| 0.1362        | 2.63  | 290  | 0.1297          |
| 0.1257        | 2.72  | 300  | 0.1293          |
| 0.1274        | 2.81  | 310  | 0.1291          |
| 0.1472        | 2.9   | 320  | 0.1291          |
| 0.1405        | 2.99  | 330  | 0.1291          |


### Framework versions

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