OpenAI GPT-2 model was proposed in Language Models are Unsupervised Multitask Learners paper .Original GPT2 model was a causal (unidirectional) transformer pretrained using language modeling on a very large corpus of ~40 GB of text data. This model has same configuration but has been pretrained on bengali corpus of mC4(multilingual C4) dataset. The code for training the model has all been open-sourced here.
Eval loss : 1.45, Eval Perplexity : 3.141
Train Steps: 250k steps
link 🤗 flax-community/gpt2-bengali
For using the model there are multiple options available. For example using the pipeline directly we can try to generate sentences.
from transformers import pipeline gpt2_bengali = pipeline('text-generation',model="flax-community/gpt2-bengali", tokenizer='flax-community/gpt2-bengali')
Similarly for using the finetuned model on bangla songs we can use following.
from transformers import pipeline singer = pipeline('text-generation',model="khalidsaifullaah/bengali-lyricist-gpt2", tokenizer='khalidsaifullaah/bengali-lyricist-gpt2')
For using on other tasks the model needs to be fine-tuned on custom datasets. Details can be found in huggingface documentation
- Khalid Saifullah
- Tasmiah Tahsin Mayeesha
- Ritobrata Ghosh
- Ibrahim Musa
- M Saiful Bari
- Downloads last month