metadata
license: mit
language:
- cs
Czech Poetry GPT
GPT2 finetuned on Czech poetry from github project by
Institute of Czech Literature, Czech Academy of Sciences.
https://github.com/versotym/corpusCzechVerse
Usage
Use as any other GPT2 style model
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
tokenizer = AutoTokenizer.from_pretrained("gpt")
model = AutoModelForCausalLM.from_pretrained("jinymusim/dialogmodel")
# Input Poet Start
poet_start = "Byl pozdní večer první máj."
poet_start = poet_start.strip()
tokenized_poet_start = tokenizer.encode(poet_start + tokenizer.eos_token, return_tensors='pt')
# generated a continuation to it
out = model.generate(tokenized_poet_start,
max_length=1000,
num_beams=2,
no_repeat_ngram_size=2,
early_stopping=True,
pad_token_id=self.tokenizer.eos_token_id)
# Decode Poet
decoded_cont = self.tokenizer.decode(out[0], skip_special_tokens=True)
print(decoded_cont)