Downloads · 30 days
71
7% of all-time downloads
groxaxo/qwen36-reap-2k-mlx-q8
qwen36-reap-2k-mlx-q8 is a text generation model from groxaxo. Use it when you need the model to write or continue text. It is set up for mlx. The card lists the license as other.
qwen36-reap-2k-mlx-q8 is an MLX-formatted checkpoint optimized for Apple silicon, published by groxaxo. It is intended for open-source evaluation, reproducible experimentation, and compatible local or hosted inference…
Downloads · 30 days
71
7% of all-time downloads
All-time downloads
1.1K
Public
Parameters
19B
21.6 GB on disk
Likes
1
Public
Click a slice to open those files.
.safetensors21.6 GB · 100%
How the weights are stored.
U3217.5B · 92%
From the Hugging Face model README
qwen36-reap-2k-mlx-q8 is an MLX-formatted checkpoint optimized for Apple silicon, published by groxaxo.
It is intended for open-source evaluation, reproducible experimentation, and compatible local or
hosted inference workflows. The wording below is deliberately limited to what can be verified
from this repository's metadata and artifacts.
| Field | Details |
|---|---|
| Format | MLX |
| Source / base | RangerX-Qwen3.6-35B-REAP-Pruned-ratio-0.5 |
| Intended task | image-text-to-text |
| License | other |
*.safetensors (5 files)config.jsongeneration_config.jsontokenizer.jsontokenizer_config.jsonprocessor_config.jsonchat_template.jinjaUse an up-to-date MLX-compatible runtime on Apple silicon and point it at this repository:
mlx_lm.generate --model groxaxo/qwen36-reap-2k-mlx-q8 --prompt "Write a concise technical summary."
Embedding and audio repositories may require the task-specific MLX package documented by the upstream project.
Quantization or conversion changes numerical behavior, memory use, and throughput relative to the source checkpoint; validate quality on your own workload.
Generated outputs may be inaccurate or unsuitable for a given use case. Users are responsible for testing behavior, applying appropriate safeguards, and complying with applicable licenses and laws.
<!-- polished-overview:end -->The same model from the story — the one that grew up on the filtered public internet, pruned but whole, with one chapter of structured-reasoning traces folded into its final layer — here pressed into a lighter form.
Where the bf16 release carries every weight at full precision, this is the 8-bit reading: the text and mixture-of-experts weights quantized to 8-bit affine (group size 64), the embeddings and output projection kept at full bf16 so the vocabulary stays sharp, and the vision tower preserved exactly as in the source — bit for bit, untouched by the compression. It thinks the same thoughts; it just carries them in a smaller satchel.
Built by fusing the rank-four LoRA into the base and quantizing in a single pass, so no precision was lost to an intermediate round-trip. Higher fidelity than the 4-bit release, lighter than full precision — for when you want the model's best foot forward without paying for the whole library.
RangerX-Qwen3.6-35B-REAP-Pruned-ratio-0.5 — Qwen3.6 35B MoE (128 experts, 8 active), REAP-pruned at ratio 0.5, multimodal (text + vision)embed_tokens and lm_head kept at bf16; vision tower bf16 (preserved bit-identical from source)mlx_lm fuse(), then quantized in the same pass; vision tower re-attached verbatimfull_attention, rest linear_attention), hidden size 2048, 16 attention heads, MRoPEmlx_lm 0.31.3 with zero missing or unexpected weights (full multimodal: text + vision)Generation was not verified on the build machine (24 GB unified memory): the Q8 weights (~20 GB) plus vision tower exceed available GPU memory at runtime. The model strict-loads cleanly and the weights are produced by the same proven fuse + quantize path as the Q4 (which generates correctly). Verify generation on a host with ≥ 32 GB unified memory.