songcomposer_sft

<p align="center"> <b><font size="6">[ACL 2025] SongComposer</font></b> <p> <div align="center">

💻Github Repo

📖Paper

</div>

SongComposer is a language large model (LLM) based on InternLM2 for lyric and melody composition in song generation.

We release SongComposer series in two versions:

• SongComposer_pretrain: The pretrained SongComposer with InternLM2 as the initialization of the LLM, gains basic knowledge on lyric and melody.
• SongComposer_sft: The finetuned SongComposer for instruction-following song generation including lyric to melody, melody to lyric, song continuation, text to song.

Import from Transformers

To load the SongComposer_sft model using Transformers, use the following code:

from transformers import AutoTokenizer, AutoModel
ckpt_path = "Mar2Ding/songcomposer_sft"
tokenizer = AutoTokenizer.from_pretrained(ckpt_path, trust_remote_code=True)
model = AutoModel.from_pretrained(ckpt_path, trust_remote_code=True).cuda().half()
prompt = 'Create a song on brave and sacrificing with a rapid pace.'
model.inference(prompt, tokenizer)

通过 Transformers 加载

通过以下的代码加载 SongComposer_sft 模型

from transformers import AutoTokenizer, AutoModel
ckpt_path = "Mar2Ding/songcomposer_sft"
tokenizer = AutoTokenizer.from_pretrained(ckpt_path, trust_remote_code=True)
model = AutoModel.from_pretrained(ckpt_path, trust_remote_code=True).cuda().half()
prompt = 'Create a song on brave and sacrificing with a rapid pace.'
model.inference(prompt, tokenizer)

Open Source License

The code is licensed under Apache-2.0, while model weights are fully open for academic research and also allow free commercial usage.