README.md

metadata

language: zh
widget:
  - text: 用户：写一首歌，歌手“毛毛”，歌曲“多少年后”，以“谎言，伤痛，心爱，掩饰，退一步，寻找”为主题。\n小L：
license: apache-2.0
pipeline_tag: text2text-generation
tags:
  - music
  - art

Chinese T5

Model description

Controllable lyrics creation generation model. Chinese lyrics can be generated based on the provided singer name, song name and several song theme keywords.

The lyric creation robot is called “小L”. So at the end of the input question, you need to add "\n小L："

How to use

You can use this model directly with a pipeline for text2text generation:

>>> from transformers import T5Tokenizer, T5ForConditionalGeneration, Text2TextGenerationPipeline
>>> tokenizer = T5Tokenizer.from_pretrained("souljoy/t5-chinese-lyric")
>>> model = T5ForConditionalGeneration.from_pretrained("souljoy/t5-chinese-lyric")
>>> text2text_generator = Text2TextGenerationPipeline(model, tokenizer)  
>>> text2text_generator("用户：写一首歌，歌手“毛毛”，歌曲“多少年后”，以“谎言，伤痛，心爱，掩饰，退一步，寻找”为主题。\\n小L：", max_length=512, do_sample=True)