odysseus-dev/odysseus · error · HTTPException

Boogu MLX generation failed: {e}

Error message

Boogu MLX generation failed: {e}

What it means

Raised by _generate_boogu in scripts/mlx_image_server.py with HTTP 500 wrapping any exception from BooguImagePipeline.from_pretrained or pipe.generate (load, VLM fetch, generation, or Image.fromarray(...).save). The chained exception text {e} carries the underlying cause; it is a runtime failure, not a client-input error.

Source

Thrown at scripts/mlx_image_server.py:320

            422,
            "This MLX image pipeline requires a companion vision-language model. "
            "Relaunch with --vlm-model <repo_or_path> or set ODYSSEUS_MLX_IMAGE_VLM_MODEL.",
        )
    try:
        pipe = BooguImagePipeline.from_pretrained(
            str(model_path),
            vlm_model,
        )
        img = pipe.generate(
            prompt,
            height=height,
            width=width,
            steps=steps,
            guidance=3.5,
        )
        Image.fromarray(img).save(out_path)
    except Exception as e:
        raise HTTPException(500, f"Boogu MLX generation failed: {e}") from e


@app.get("/v1/models")
def list_models():
    return {"data": [{"id": _args.model, "object": "model", "owned_by": "local"}]}


@app.post("/v1/images/generations")
def generate(req: ImageRequest):
    model = req.model or _args.model
    width, height = _size(req.size)
    out_images = []
    count = max(1, min(int(req.n or 1), 4))
    for _ in range(count):
        with tempfile.TemporaryDirectory(prefix="odysseus-mlx-image-") as td:
            out_path = Path(td) / "image.png"
            if _is_hidream(model):
                _generate_hidream(model, req.prompt, out_path, width, height, _steps(req.quality))

View on GitHub (pinned to f9235ebbf1)

Solutions

1. Read {e} in the response to identify the failing stage (from_pretrained vs generate)
2. Verify the VLM repo id/path resolves and re-download if the error mentions missing files
3. Pin/upgrade mlx and boogu-image-mlx to a known-compatible pair (reinstall the git package)
4. Lower height/width/steps if the error indicates memory pressure

from pathlib import Path
assert Path(vlm_model).exists() or '/' in vlm_model, 'vlm_model must be a hub id or local path'

except HTTPException as e:
    if e.status_code == 500 and 'Boogu MLX generation failed' in e.detail:
        log(e.detail)  # contains chained cause
        if 'No such file' in e.detail: refresh_vlm_snapshot(); retry_once()
        else: raise

When it happens

Trigger: Boogu generation where pipeline construction or generate raises: bad/missing VLM path, corrupted weights, MLX version incompatibility, unsupported height/width/steps causing an internal assertion, or out-of-memory during generate.

Common situations: vlm-model pointing at a wrong repo id; mlx package version mismatch with boogu-image-mlx; first-run weight download interrupted; very large resolution requests exhausting unified memory.

Related errors

• No safetensors weights found for {model} in {snap}
• detail[-4000:]
• HiDream generator script not found in snapshot: {script}
• Boogu MLX serving requires boogu-image-mlx in the launch Pyt
• This MLX image pipeline requires a companion vision-language

AI-assisted analysis of odysseus-dev/odysseus@f9235ebbf1 (2026-08-14). Data as JSON: /api/errors/2cfda2e59ac3b1c8. Report an issue: GitHub.