File size: 1,711 Bytes
d967808
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
base_model: microsoft/infoxlm-base
tags:
- generated_from_trainer
metrics:
- f1
model-index:
- name: MRC_ER_infoXLM-base_syl_DSC
  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. -->

# MRC_ER_infoXLM-base_syl_DSC

This model is a fine-tuned version of [microsoft/infoxlm-base](https://huggingface.co/microsoft/infoxlm-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8322
- Exact Match: 0.671
- F1: 0.8432

## 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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Exact Match | F1     |
|:-------------:|:-----:|:-----:|:---------------:|:-----------:|:------:|
| 0.867         | 1.0   | 6364  | 0.7899          | 0.593       | 0.7980 |
| 0.7092        | 2.0   | 12728 | 0.7218          | 0.641       | 0.8262 |
| 0.6151        | 3.0   | 19092 | 0.6985          | 0.623       | 0.8371 |
| 0.4986        | 4.0   | 25456 | 0.7597          | 0.657       | 0.8500 |
| 0.416         | 5.0   | 31820 | 0.8322          | 0.671       | 0.8432 |


### Framework versions

- Transformers 4.44.0
- Pytorch 2.4.0
- Datasets 2.21.0
- Tokenizers 0.19.1