File size: 1,771 Bytes
1e45555
 
 
 
8f761cf
 
1e45555
 
 
 
 
 
 
 
 
 
 
8f761cf
 
 
 
 
1e45555
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8f761cf
 
 
 
 
 
 
 
 
 
 
1e45555
 
 
 
 
 
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
---
base_model: samzirbo/gendered
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: debiased
  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. -->

# debiased

This model is a fine-tuned version of [samzirbo/gendered](https://huggingface.co/samzirbo/gendered) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.2305
- Bleu: 43.086
- Meteor: 0.6836
- Chrf++: 61.9553

## 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: 0.0001
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 1000
- training_steps: 10000

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Bleu    | Meteor | Chrf++  |
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:------:|:-------:|
| 1.3358        | 3.2   | 2000  | 1.1980          | 43.4198 | 0.6874 | 62.3558 |
| 1.2174        | 6.4   | 4000  | 1.2181          | 43.3333 | 0.6844 | 62.1229 |
| 1.1529        | 9.6   | 6000  | 1.2246          | 43.222  | 0.6846 | 62.054  |
| 1.1161        | 12.8  | 8000  | 1.2291          | 43.1386 | 0.6836 | 61.9676 |
| 1.1026        | 16.0  | 10000 | 1.2305          | 43.086  | 0.6836 | 61.9553 |


### Framework versions

- Transformers 4.38.0
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.15.2