Full multimodal vision-language model; the vision tower is preserved.
The checkpoint's embedded Lightning MTP head is preserved.
8-bit affine weights with group size 64.
Activation-weighted clipping used 128 multilingual/code calibration samples of 512 tokens and 504 collected importance entries.
Layer sensitivity was measured on a RAM-safe proxy rather than assigned uniformly.

oMLX usage

bash

hf download scaryrawr/Qwen3.8-27B-oQ8e-mtp \
  --local-dir ~/.models/scaryrawr/Qwen3.8-27B-oQ8e-mtp
omlx serve --model-dir ~/.models/scaryrawr

Enable mtp_enabled with mtp_num_draft_tokens: 3 in the model's oMLX settings. Text-only requests use native MTP draft/verify. Requests containing images or video retain multimodal support and conservatively use standard decode.

The affine 8-bit path can use oMLX's guarded custom MTP verify QMM kernel when eligible; unsupported shapes automatically fall back to the standard path.

Local validation

Measured on an M4 Max with 128 GB unified memory using adaptive MTP depth 3:

Runtime model memory: 28.37 GB
Warm predictable-code generation: 30.1 tokens/s
Open-ended prose generation: 24.5 tokens/s

The generated LIS implementation passed 207/207 correctness cases. Focused checks also passed arithmetic, strict structured JSON, French instruction following, and image-color recognition. This is the highest-fidelity practical affine variant in this set. These checks are useful smoke tests, not a comprehensive benchmark suite.

License

Apache 2.0, following the base model.