File size: 2,049 Bytes
d51843e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
base_model: csebuetnlp/mT5_multilingual_XLSum
tags:
- generated_from_trainer
model-index:
- name: type2_pos_1
  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. -->

# type2_pos_1

This model is a fine-tuned version of [csebuetnlp/mT5_multilingual_XLSum](https://huggingface.co./csebuetnlp/mT5_multilingual_XLSum) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.9319

## 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: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 10

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 2.7093        | 0.63  | 500  | 2.9195          |
| 3.2641        | 1.25  | 1000 | 2.7276          |
| 2.3791        | 1.88  | 1500 | 2.6590          |
| 2.4823        | 2.51  | 2000 | 2.6422          |
| 2.3649        | 3.14  | 2500 | 2.6791          |
| 1.8109        | 3.76  | 3000 | 2.6492          |
| 1.7853        | 4.39  | 3500 | 2.6922          |
| 1.5986        | 5.02  | 4000 | 2.7080          |
| 1.71          | 5.65  | 4500 | 2.7670          |
| 1.6342        | 6.27  | 5000 | 2.8096          |
| 1.6734        | 6.9   | 5500 | 2.8439          |
| 1.4521        | 7.53  | 6000 | 2.8824          |
| 1.3753        | 8.16  | 6500 | 2.8878          |
| 1.8335        | 8.78  | 7000 | 2.9046          |
| 1.276         | 9.41  | 7500 | 2.9319          |


### Framework versions

- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2