File size: 1,823 Bytes
4b3d157
 
 
 
81f677f
4b3d157
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: mit
tags:
- generated_from_trainer
base_model: ai4bharat/indic-bert
metrics:
- accuracy
- precision
- recall
model-index:
- name: IndicBERT_Finetuned_Final
  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. -->

# IndicBERT_Finetuned_Final

This model is a fine-tuned version of [ai4bharat/indic-bert](https://huggingface.co/ai4bharat/indic-bert) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6539
- Accuracy: 0.7227
- Precision: 0.7377
- Recall: 0.7227

## 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: 64
- eval_batch_size: 16
- 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 | Accuracy | Precision | Recall |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|
| 0.8979        | 1.0   | 190  | 0.9064          | 0.5493   | 0.3712    | 0.5493 |
| 0.807         | 2.0   | 380  | 0.7564          | 0.65     | 0.6417    | 0.65   |
| 0.6731        | 3.0   | 570  | 0.6962          | 0.6833   | 0.7411    | 0.6833 |
| 0.6579        | 4.0   | 760  | 0.6723          | 0.6987   | 0.7213    | 0.6987 |
| 0.5946        | 5.0   | 950  | 0.6539          | 0.7227   | 0.7377    | 0.7227 |


### Framework versions

- Transformers 4.41.2
- Pytorch 2.1.2
- Datasets 2.19.2
- Tokenizers 0.19.1