File size: 1,940 Bytes
7695d97
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0fc6783
7695d97
0fc6783
 
7695d97
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4c943ea
7695d97
 
 
 
 
0fc6783
 
 
 
 
 
 
 
 
7695d97
 
 
 
4c943ea
7695d97
 
4c943ea
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
---
license: mit
base_model: microsoft/git-base
tags:
- generated_from_trainer
model-index:
- name: git-base-pokemon
  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. -->

# git-base-pokemon

This model is a fine-tuned version of [microsoft/git-base](https://huggingface.co/microsoft/git-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2324
- Wer Score: 5.2456

## 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: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer Score |
|:-------------:|:-----:|:----:|:---------------:|:---------:|
| 7.3165        | 5.26  | 50   | 4.6355          | 4.8031    |
| 2.6531        | 10.53 | 100  | 0.8536          | 4.2762    |
| 0.4           | 15.79 | 150  | 0.2527          | 4.6248    |
| 0.1634        | 21.05 | 200  | 0.2272          | 4.4392    |
| 0.125         | 26.32 | 250  | 0.2262          | 4.3356    |
| 0.1054        | 31.58 | 300  | 0.2286          | 6.6022    |
| 0.0932        | 36.84 | 350  | 0.2305          | 4.6967    |
| 0.0856        | 42.11 | 400  | 0.2320          | 4.4171    |
| 0.0811        | 47.37 | 450  | 0.2324          | 5.2456    |


### Framework versions

- Transformers 4.36.2
- Pytorch 2.1.0+cu121
- Datasets 2.14.7
- Tokenizers 0.15.0