File size: 3,463 Bytes
94e2b85
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
aeb1b92
 
 
94e2b85
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
00d5ab8
 
94e2b85
 
 
aeb1b92
94e2b85
 
 
 
baa68fb
 
aeb1b92
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
94e2b85
 
 
 
 
 
 
 
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
---
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: English2AkuapemTwi
  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. -->

# English2AkuapemTwi

This model is a fine-tuned version of [allenai/unifiedqa-t5-small](https://huggingface.co/allenai/unifiedqa-t5-small) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3098
- Bleu: 8.8875
- Gen Len: 17.5114

## 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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Bleu   | Gen Len |
|:-------------:|:-----:|:-----:|:---------------:|:------:|:-------:|
| 2.8304        | 1.0   | 2542  | 2.5363          | 0.2404 | 18.9072 |
| 2.5079        | 2.0   | 5084  | 2.2533          | 0.5763 | 18.6814 |
| 2.3362        | 3.0   | 7626  | 2.0838          | 0.8836 | 18.4972 |
| 2.2155        | 4.0   | 10168 | 1.9649          | 1.3748 | 18.2368 |
| 2.1066        | 5.0   | 12710 | 1.8767          | 2.0954 | 18.0588 |
| 2.0368        | 6.0   | 15252 | 1.8057          | 2.6294 | 17.9451 |
| 1.9641        | 7.0   | 17794 | 1.7436          | 3.0267 | 17.8342 |
| 1.9194        | 8.0   | 20336 | 1.6907          | 3.4947 | 18.0199 |
| 1.8474        | 9.0   | 22878 | 1.6438          | 4.0431 | 17.7758 |
| 1.8261        | 10.0  | 25420 | 1.6008          | 4.4608 | 17.7919 |
| 1.7833        | 11.0  | 27962 | 1.5642          | 4.9271 | 17.7868 |
| 1.7255        | 12.0  | 30504 | 1.5322          | 5.3659 | 17.7962 |
| 1.6913        | 13.0  | 33046 | 1.5020          | 5.9051 | 17.7301 |
| 1.6616        | 14.0  | 35588 | 1.4780          | 6.0824 | 17.7217 |
| 1.6522        | 15.0  | 38130 | 1.4550          | 6.6326 | 17.6227 |
| 1.6039        | 16.0  | 40672 | 1.4327          | 6.7955 | 17.6564 |
| 1.5926        | 17.0  | 43214 | 1.4119          | 7.0477 | 17.6876 |
| 1.5802        | 18.0  | 45756 | 1.3960          | 7.4328 | 17.6233 |
| 1.5438        | 19.0  | 48298 | 1.3816          | 7.4937 | 17.6359 |
| 1.5452        | 20.0  | 50840 | 1.3695          | 7.7696 | 17.5515 |
| 1.5423        | 21.0  | 53382 | 1.3569          | 7.9677 | 17.5596 |
| 1.5076        | 22.0  | 55924 | 1.3465          | 8.2829 | 17.5193 |
| 1.5183        | 23.0  | 58466 | 1.3386          | 8.4304 | 17.5486 |
| 1.4919        | 24.0  | 61008 | 1.3296          | 8.536  | 17.5714 |
| 1.4923        | 25.0  | 63550 | 1.3242          | 8.6446 | 17.4343 |
| 1.488         | 26.0  | 66092 | 1.3177          | 8.7099 | 17.5126 |
| 1.477         | 27.0  | 68634 | 1.3144          | 8.8431 | 17.5035 |
| 1.4724        | 28.0  | 71176 | 1.3116          | 8.8588 | 17.5277 |
| 1.4488        | 29.0  | 73718 | 1.3105          | 8.8989 | 17.5024 |
| 1.4837        | 30.0  | 76260 | 1.3098          | 8.8875 | 17.5114 |


### Framework versions

- Transformers 4.27.4
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.3