File size: 2,582 Bytes
e932913
 
 
 
 
 
dcfd813
e932913
 
 
 
 
 
 
 
 
 
 
 
dcfd813
 
 
 
 
e932913
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
dcfd813
 
e932913
 
 
dcfd813
e932913
 
 
dcfd813
 
 
 
 
 
 
 
 
 
 
 
 
 
e932913
 
 
 
 
dcfd813
e932913
 
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
---
license: apache-2.0
base_model: google/mt5-small
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: mt5-small-task3-dataset3
  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. -->

# mt5-small-task3-dataset3

This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.4093
- Accuracy: 0.128
- Mse: 1.5841
- Log-distance: 0.6809
- S Score: 0.4800

## 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: 5.6e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 12

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | Mse    | Log-distance | S Score |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:------------:|:-------:|
| 2.1991        | 1.0   | 250  | 1.5275          | 0.106    | 1.7120 | 0.7824       | 0.4048  |
| 2.1115        | 2.0   | 500  | 1.5009          | 0.106    | 1.7469 | 0.8062       | 0.3844  |
| 1.8295        | 3.0   | 750  | 1.4483          | 0.108    | 1.7239 | 0.7902       | 0.3972  |
| 1.7033        | 4.0   | 1000 | 1.4335          | 0.112    | 1.7052 | 0.7759       | 0.4088  |
| 1.6426        | 5.0   | 1250 | 1.4224          | 0.12     | 1.5337 | 0.6427       | 0.5112  |
| 1.5923        | 6.0   | 1500 | 1.4236          | 0.126    | 1.6061 | 0.7015       | 0.4628  |
| 1.5529        | 7.0   | 1750 | 1.4284          | 0.122    | 1.5984 | 0.6967       | 0.4676  |
| 1.546         | 8.0   | 2000 | 1.4132          | 0.124    | 1.6032 | 0.6948       | 0.4704  |
| 1.5364        | 9.0   | 2250 | 1.4306          | 0.116    | 1.6403 | 0.7282       | 0.4460  |
| 1.5365        | 10.0  | 2500 | 1.4107          | 0.118    | 1.5702 | 0.6681       | 0.4948  |
| 1.5145        | 11.0  | 2750 | 1.4182          | 0.118    | 1.6041 | 0.7063       | 0.4596  |
| 1.5103        | 12.0  | 3000 | 1.4093          | 0.128    | 1.5841 | 0.6809       | 0.4800  |


### Framework versions

- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0