Instructions to use DeepBeepMeep/TTS with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusion Single File
How to use DeepBeepMeep/TTS with Diffusion Single File:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Download sheetsage2/README.md from DeepBeepMeep/TTS: direct link, hf CLI and curl.
- Browser
- Download file 1.19 kB
-
https://huggingface.co/DeepBeepMeep/TTS/resolve/main/sheetsage2/README.md
- Command line
-
hf download hf://DeepBeepMeep/TTS/sheetsage2/README.md
-
curl -L -o README.md https://huggingface.co/DeepBeepMeep/TTS/resolve/main/sheetsage2/README.md
SheetSage2 / MERT2 for WanGP
Merged BF16 music-transcription checkpoint prepared for WanGP YuE2 source-song scoring. It extracts a melody or melody and chords as ABC notation; lyrics must be supplied separately.
Sources and attribution: Multimodal Art Projection (MAP) Team and collaborators.
Weights are subject to CC BY-NC 4.0 (non-commercial use); see the accompanying upstream license files.
SheetSage2 attention adapters are merged into MERT2 in FP32 before BF16 conversion. Four audio frontend buffers remain FP32. Shared embeddings are recorded using MMGP tied-weight metadata. The conversion manifest records source revisions and tensor coverage.
Place SheetSage2_MERT2_bf16.safetensors in ckpts/sheetsage2/. WanGP loads it only when source scoring is requested and releases the model after transcription. This checkpoint uses WanGP's local SheetSage2 implementation.
Checkpoint SHA-256: b4b0d83c2fa0c6e912479866078a2c22f4a0ff4e4e63ed33b780b941cc2bad47.