metadata

license: mit
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: indic-bert-finetuned-code-mixed-DS
    results: []

indic-bert-finetuned-code-mixed-DS

This model is a fine-tuned version of ai4bharat/indic-bert on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.8981
Accuracy: 0.5594
Precision: 0.3838
Recall: 0.5263
F1: 0.4118

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: 1e-06
train_batch_size: 16
eval_batch_size: 16
seed: 43
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.0941	2.0	497	1.0845	0.3441	0.3615	0.4180	0.2706
1.0379	3.99	994	0.9775	0.5412	0.3779	0.5128	0.4003
0.9509	5.99	1491	0.9271	0.5513	0.3752	0.5144	0.4043
0.9152	7.98	1988	0.9047	0.5614	0.3852	0.5275	0.4131
0.8953	9.98	2485	0.8981	0.5594	0.3838	0.5263	0.4118

Framework versions

Transformers 4.20.1
Pytorch 1.10.1+cu111
Datasets 2.3.2
Tokenizers 0.12.1