This model was finetuned from the original
DialoGPT-medium model on subtitles from Spanish movies and telenovelas from the awesome
OpenSubtitle dataset.
Example code for trying out the model (taken directly from the
DialoGPT-medium model card):
1from transformers import AutoModelWithLMHead, AutoTokenizer
2import torch
3
4
5tokenizer = AutoTokenizer.from_pretrained("emre/spanish-dialoGPT")
6model = AutoModelWithLMHead.from_pretrained("emre/spanish-dialoGPT")
7
8# Let's chat for 5 lines
9for step in range(5):
10 # encode the new user input, add the eos_token and return a tensor in Pytorch
11 new_user_input_ids = tokenizer.encode(input(">> User:") + tokenizer.eos_token, return_tensors='pt')
12
13 # append the new user input tokens to the chat history
14 bot_input_ids = torch.cat([chat_history_ids, new_user_input_ids], dim=-1) if step > 0 else new_user_input_ids
15
16 # generated a response while limiting the total chat history to 1000 tokens,
17 chat_history_ids = model.generate(bot_input_ids, max_length=1000, pad_token_id=tokenizer.eos_token_id)
18
19 # pretty print last ouput tokens from bot
20 print("DialoGPT: {}".format(tokenizer.decode(chat_history_ids[:, bot_input_ids.shape[-1]:][0], skip_special_tokens=True)))