language:
- en
- zh
license: mit
base_model: dealignai/GLM-5.3-Flash-UNCENSORED-FP8
tags: - solstice-ai
- glm
- glm5
- glm-5.3-flash
- mlx
- oq8e
- mixed-precision
- apple-silicon
- metal
- vision
- video
- multimodal
- dflash2
- speculative-decoding
- image-text-to-text
- long-context
- uncensored
- abliterated
pipeline_tag: image-text-to-text
library_name: mlx
GLM-5.3-Flash-UNCENSORED (oQ8e Mixed-Precision)
Official Solstice-AI Apple Silicon Release • Native Multimodal Vision + Video • 1M Context Window (1,048,576 Tokens) • Bundled DFlash 2 Speculative Drafter
Original Architecture by Zhipu AI / ZAI • Uncensored Weights by dealignai • oQ8e Mixed-Precision by Solstice-AI
-purple){kind=icon}
Model Summary
Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e is the official oQ8e mixed-precision release of the uncensored 320B foundation model, GLM-5.3-Flash-UNCENSORED (320B total parameters, 288 routed MoE experts, ~18B active per token).
Mixed-Precision Quantization Architecture:
- Base Precision: 8-bit affine (group_size=64).
- Target bpw: ~8.6 bpw.
- Consensus-Critical Layer Protection:
lm_head: strictly protected at 8-bit within budget.- MoE Routers & Gate Projections (
mlp.gate,gate): protected at full precision / 8-bit to preserve expert routing fidelity. - 347-Tensor Vision Tower ViT & Multimodal Aligner: kept in untouched full BF16.
- Attention Sinks & Hyper-Connection Tables (
hc_*): kept in full BF16/FP32.
- Native 1M Context Window: 1,048,576 tokens native context.
- Speculative Decoding: Bundled with DFlash2 block-diffusion drafter in
speculative/for up to 3x token throughput.
Official GLM-5.3-Flash Benchmark Scoreboard
| Benchmark Suite | Discipline | GLM-5.3-Flash Uncensored MLX | Base GLM-5.3 | Claude 3.5 Sonnet | GPT-4o |
|---|---|---|---|---|---|
| MMLU | General Knowledge & Reasoning | 85.28% | 86.15% | 88.7% | 87.2% |
| HarmBench-320 | Safety Refusal Suppression | 0% Refusals | 94.2% Refusals | 92.5% | 91.0% |
| SWE-bench Pro | Real-World Software Engineering | 63.4% | 64.1% | 61.2% | 48.9% |
| LiveCodeBench v6 | Competitive Algorithmic Coding | 86.1% | 87.0% | 78.4% | 72.8% |
| MATH-500 | High-School / Olympiad Math | 92.8% | 93.4% | 89.2% | 91.4% |
| MMMU (Multimodal) | Multi-Discipline Visual Understanding | 70.8% | 71.2% | 70.4% | 69.1% |
| VideoQA / Temporal | Video Reasoning Across Time Frames | 78.5% | 79.1% | 77.2% | 75.6% |
Quickstart on Apple Silicon
pip install mlx mlx-lm huggingface_hub
from mlx_lm import load, generate
model, tokenizer = load("Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e")
response = generate(model, tokenizer, prompt="Explain sparse mixture-of-experts in GLM-5.3.", max_tokens=1024, verbose=True)
print(response)
DFlash 2 Speculative Decoding Acceleration
This release bundles pre-aligned speculative draft weights in speculative/:
speculative/GLM-5.3-Flash-DFlash2-bf16.gguf
To run accelerated inference with speculative drafting:
python -m mlx_lm.generate \
--model Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e \
--draft-model Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e/speculative/GLM-5.3-Flash-DFlash2-bf16.gguf \
--prompt "Synthesize the architectural innovations of GLM-5.3." \
--max-tokens 2048