Spaces:
Running on Zero
Running on Zero
Commit History
-
Remove the LLM compose step (Inference Providers cost) Running
Lyrics + structured caption are now editable reference nodes wired directly into generate_song, with the official-guide defaults. Drops the openai dep, the MiniMax-M3 prompts, OAuth token plumbing, and the inference-api scope.cffc867 -
make_video: resolve audio port value robustly
The executor strips 'path' when re-serializing the audio dict between nodes; accept path/name keys, /gradio_api/file= URLs, or download http(s) URLs to a temp file.8da1cb7 -
Note optional sign-in for custom lyrics in the workflow description
The canvas already provides the LoginButton; no extra button needed.4c1e11d -
Skip the LLM compose step for logged-out visitors
compose_song falls back to the built-in demo lyrics + structured caption when no OAuth token and no HF_TOKEN secret are available, so anonymous users can still run the generation pipeline.6885314 -
compose_song: use visitor OAuth token for the LLM step
- token: gr.OAuthToken param injected by gr.Workflow (excluded from canvas ports); bills the signed-in visitor's own inference quota. - Falls back to the Space's HF_TOKEN secret for anonymous/local runs, with a clear gr.Error when neither is available. - README: hf_oauth_scopes: [inference-api].1b8ccaf -
Convert to gr.Workflow canvas app on ZeroGPU
- app.py: gr.Workflow(graph=workflow.json) with three bound fn operators: compose_song (MiniMax-M3 LLM lyrics + structured caption), generate_song (@spaces.GPU xlarge worker, wav file dict + seed + stats), make_video (ffmpeg visualizer). All model machinery (AoTI kernels, StaticCache ladder, windowed DiT decode) unchanged. - workflow.json: references -> compose_song -> generate_song -> subjects, with audio fan-out into make_video -> Share Video. - app_blocks.py: original streaming Blocks app preserved. - README: hf_oauth: true for owner canvas editing.de1b9e6 -
Update app.py
Verified signature1c4da36 -
Load weights from MiniMaxAI/MiniMax-Music3 (#1)
- Load weights from MiniMaxAI/MiniMax-Music3 (5272e579dbcc4b947e9021bbf13c10c28cfa8ab0) - Rebase on main (theme-aware logo), keep only the MiniMaxAI/MiniMax-Music3 swap (bc8fe79830f9f35182f9f2cdbda38953219cd761) - Rebuild on current main (57e8935): keep only the MiniMaxAI/MiniMax-Music3 swap (30702c273ba7eb4d5152157d91f2f1f5178a4182)ab5ad1e -
Update README.md
Verified signature57e8935 -
Update app.py
Verified signaturec9a390f -
Update app.py
Verified signature001ee71 -
Update app.py
Verified signature7670b1b -
Replace H1 title with theme-aware logo (600px, centered)
c225762 -
Update README.md
Verified signatureaa18265 -
MiniMax Music 3 demo (modular diffusers pipeline + AoTI)
Verified signature1688c15 -
initial commit
Verified signature70f16e1