README.md

metadata

language: ne
license: mit
tags:
  - generated_from_trainer
  - gpt2
  - ne
datasets: Oscar
widget:
  - text: गर्मि मौसममा चिसो खाने

gpt2-medium-ne

This model is a fine-tuned version of gpt2 on Oscar Dataset.

Model description

This model is trained on Oscar Nepali Dataset.

How to use

You can use this model directly with a pipeline for text generation.

>>> from transformers import pipeline, set_seed
>>> generator = pipeline('text-generation', model='Someman/gpt2-medium-ne')
>>> set_seed(42)
>>> generator("उच्च अदालतले बिहीबार दिएको आदेशले", max_length=30, num_return_sequences=5)

[{'generated_text': 'उच्च अदालतले बिहीबार दिएको आदेशले महिनात्रि'},
 {'generated_text': 'उच्च अदालतले बिहीबार दिएको आदेशले बिहानैदे'},
 {'generated_text': 'उच्च अदालतले बिहीबार दिएको आदेशले गिरिजाली'},
 {'generated_text': 'उच्च अदालतले बिहीबार दिएको आदेशले गरेको प्रथम त'},
 {'generated_text': 'उच्च अदालतले बिहीबार दिएको आदेशले कुनै साथी'}]

Here is how to use this model to get the features of a given text in PyTorch:

from transformers import GPT2Tokenizer, GPT2Model
tokenizer = GPT2Tokenizer.from_pretrained('Someman/gpt2-medium-ne')
model = GPT2Model.from_pretrained('Someman/gpt2-medium-ne')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)

and in TensorFlow:

from transformers import GPT2Tokenizer, TFGPT2Model
tokenizer = GPT2Tokenizer.from_pretrained('Someman/gpt2-medium-ne')
model = TFGPT2Model.from_pretrained('Someman/gpt2-medium-ne')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)

More information needed

Training and evaluation data

Training data contains 197k Nepali sentences.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 1

Training results

Framework versions

• Transformers 4.21.1
• Pytorch 1.12.0+cu116
• Datasets 2.4.0
• Tokenizers 0.12.1