File size: 4,650 Bytes
2184b50
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
---
base_model: csebuetnlp/banglat5
tags:
- generated_from_trainer
model-index:
- name: banglat5-bcoqa
  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. -->

# banglat5-bcoqa

This model is a fine-tuned version of [csebuetnlp/banglat5](https://huggingface.co/csebuetnlp/banglat5) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.4135

## 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: 6
- eval_batch_size: 6
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2

### Training results

| Training Loss | Epoch | Step  | Validation Loss |
|:-------------:|:-----:|:-----:|:---------------:|
| 6.2253        | 0.03  | 700   | 2.8997          |
| 3.5361        | 0.06  | 1400  | 2.5143          |
| 2.8585        | 0.09  | 2100  | 2.3899          |
| 2.7763        | 0.12  | 2800  | 2.3435          |
| 2.6044        | 0.15  | 3500  | 2.3001          |
| 2.6166        | 0.18  | 4200  | 2.2498          |
| 2.5002        | 0.21  | 4900  | 2.1958          |
| 2.4498        | 0.24  | 5600  | 2.1454          |
| 2.4349        | 0.27  | 6300  | 2.1049          |
| 2.3176        | 0.3   | 7000  | 2.0382          |
| 2.2667        | 0.33  | 7700  | 1.9124          |
| 2.2382        | 0.36  | 8400  | 1.7847          |
| 2.1296        | 0.39  | 9100  | 1.6963          |
| 2.0856        | 0.42  | 9800  | 1.6489          |
| 2.0527        | 0.45  | 10500 | 1.6299          |
| 2.0363        | 0.48  | 11200 | 1.6085          |
| 1.9999        | 0.51  | 11900 | 1.5947          |
| 1.9888        | 0.54  | 12600 | 1.5661          |
| 1.9438        | 0.58  | 13300 | 1.5666          |
| 1.9365        | 0.61  | 14000 | 1.5636          |
| 1.9311        | 0.64  | 14700 | 1.5502          |
| 1.9649        | 0.67  | 15400 | 1.5419          |
| 1.9782        | 0.7   | 16100 | 1.5309          |
| 1.8764        | 0.73  | 16800 | 1.5147          |
| 1.9236        | 0.76  | 17500 | 1.5066          |
| 1.8818        | 0.79  | 18200 | 1.4963          |
| 1.9031        | 0.82  | 18900 | 1.4939          |
| 1.8583        | 0.85  | 19600 | 1.4923          |
| 1.8436        | 0.88  | 20300 | 1.4948          |
| 1.8258        | 0.91  | 21000 | 1.4784          |
| 1.8701        | 0.94  | 21700 | 1.4642          |
| 1.8413        | 0.97  | 22400 | 1.4807          |
| 1.8417        | 1.0   | 23100 | 1.4654          |
| 1.7898        | 1.03  | 23800 | 1.4711          |
| 1.7661        | 1.06  | 24500 | 1.4632          |
| 1.7223        | 1.09  | 25200 | 1.4514          |
| 1.7461        | 1.12  | 25900 | 1.4568          |
| 1.7457        | 1.15  | 26600 | 1.4492          |
| 1.7588        | 1.18  | 27300 | 1.4500          |
| 1.6475        | 1.21  | 28000 | 1.4515          |
| 1.7428        | 1.24  | 28700 | 1.4377          |
| 1.782         | 1.27  | 29400 | 1.4456          |
| 1.6906        | 1.3   | 30100 | 1.4435          |
| 1.6865        | 1.33  | 30800 | 1.4378          |
| 1.7806        | 1.36  | 31500 | 1.4327          |
| 1.7444        | 1.39  | 32200 | 1.4372          |
| 1.7136        | 1.42  | 32900 | 1.4293          |
| 1.7252        | 1.45  | 33600 | 1.4246          |
| 1.7209        | 1.48  | 34300 | 1.4218          |
| 1.7523        | 1.51  | 35000 | 1.4283          |
| 1.6808        | 1.54  | 35700 | 1.4216          |
| 1.7167        | 1.57  | 36400 | 1.4246          |
| 1.7246        | 1.6   | 37100 | 1.4171          |
| 1.7614        | 1.63  | 37800 | 1.4204          |
| 1.6704        | 1.66  | 38500 | 1.4116          |
| 1.6823        | 1.7   | 39200 | 1.4213          |
| 1.6744        | 1.73  | 39900 | 1.4236          |
| 1.7086        | 1.76  | 40600 | 1.4197          |
| 1.7179        | 1.79  | 41300 | 1.4156          |
| 1.6223        | 1.82  | 42000 | 1.4205          |
| 1.6817        | 1.85  | 42700 | 1.4159          |
| 1.6786        | 1.88  | 43400 | 1.4131          |
| 1.7163        | 1.91  | 44100 | 1.4147          |
| 1.6381        | 1.94  | 44800 | 1.4131          |
| 1.6961        | 1.97  | 45500 | 1.4134          |
| 1.6247        | 2.0   | 46200 | 1.4135          |


### Framework versions

- Transformers 4.39.0.dev0
- Pytorch 2.2.1+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1