cb9247e537
- base_precision="mxfp8": torchao MX block-scaled float8 compute on the frozen base linears (Blackwell sm100+, cuBLAS kernels). Applied after add_adapter like fp8, never fatal, weights stay bf16 in memory. Measured 1.16x over compiled bf16 on Z-Image at 1024px batch 4 (16k tokens/step); a wash at small token counts, so it stays an explicit opt-in and auto never picks it. - SDXL: regionally compile the U-Net's BasicTransformerBlocks through the same never-fatal wrapper the DiT trainer uses. 1.35x steady state at 1024px batch 4 with same-seed loss parity (~1e-5 per step) and unchanged peak VRAM; ~30 s one-time warmup. Steady-state samples/sec now excludes step 1, matching the DiT trainer. - /info: mxfp8 advertised only on sm100+; supports_compile now true for sdxl. - NVFP4 training: not available in torchao 0.16 (no autograd path, no training recipe), so NVFP4 stays an inference-only quant for now. 193 diffusion backend tests green; frontend build clean. |
||
|---|---|---|
| .. | ||
| public | ||
| src | ||
| .gitignore | ||
| .gitkeep | ||
| .npmrc | ||
| biome.json | ||
| components.json | ||
| data-designer.openapi (1).yaml | ||
| eslint.config.js | ||
| index.html | ||
| package-lock.json | ||
| package.json | ||
| tsconfig.app.json | ||
| tsconfig.json | ||
| tsconfig.node.json | ||
| vite.config.ts | ||