Instructions to use philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX

Run Hermes

hermes

OpenClaw new

How to use philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

MLX LM

How to use philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "philipjohnbasile/GLM-5.2-Demolition-q4a4-soul-MLX",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

GLM-5.2-Demolition-q4a4-soul-MLX / README.md

philipjohnbasile

Upload README.md with huggingface_hub

a88adac verified 15 days ago

12.8 kB

license: mit
base_model: zai-org/GLM-5.2
library_name: mlx
pipeline_tag: text-generation
language:
  - en
tags:
  - mlx
  - moe
  - code
  - agentic
  - glm
  - pruned
  - quantized
  - verified-decoding
  - apple-silicon
  - local-agent

GLM-5.2-Demolition — a 743B frontier MoE, demolished to run on a 128 GB Mac

One line: we took zai-org/GLM-5.2 (743B-parameter Mixture-of-Experts, ~381 GB at 4-bit) and demolished it to 99 GB so it runs fully on-device on a MacBook Pro M5 Max (128 GB) — then healed it and wrapped it in a 51-tool local agent that does things a cloud model structurally cannot: the compiler steers every line it writes, it can't fake a passing test or leak a secret, and it can be fine-tuned on your private repo so it writes in your style.

A niche specialist, not a general model — tuned to beat a frontier model in one lane (agentic coding + design for TS/JS/Python/Rust/Go/HTML/CSS + Postgres) by out-verifying it, not out-knowing it.

My AI-Engineer / Full-Stack / Data-Science / ML build

This is the version I run wearing all four hats — one on-device model, no cloud key, tooled for the whole stack of those roles (strongest in the coding/agentic lane, deliberately so):

AI Engineer — builds and ships agentic AI locally: the 51-tool ReAct agent, verified + constrained decoding, grammar-constrained tool I/O, MLX-native serving + the speed/stability work (prompt-cache, continuous batching, frontier-grade serving). The model that makes AI products.
Full-Stack — front-to-back in TS/JS/Python/Rust/Go/HTML/CSS + Postgres, the compiler steering every line, a design soul (render-and-see critic: WCAG / type-scale / OKLCH) for the UI, and SQL-on-a-real-schema for the API — plus edit→test→fix agentic loops on your repo.
Data Science — stateful REPL, SymPy / pandas / numpy / sklearn, arXiv-RAG, competition-grade math (GSM8K-style), and code-rendered figures (matplotlib / manim / TikZ).
Machine Learning — it is applied ML end-to-end: REAP expert-pruning (256→77), mixed-precision quantization, LoRA healing, distillation, MTP self-speculation, GRPO/RLVR experiments — the build itself is a working reference.

…and the hats that fall straight out of "verify-everything":

Security / DevSecOps — secret-scanning (16 providers: AWS/GitHub/OpenAI/Anthropic/HuggingFace/Slack/Stripe/Google/DB-URLs/JWT/PEM…), prompt-injection guard, test-tamper + fabrication-proof done, slopsquat/typosquat guard, risk-gated tools. It structurally can't leak a key or fake a green test.
Formal-Methods / Verification Engineer — a local Lean-4 prover (premise selection, expert-iteration, self-correction from the real Lean error) → correct-by-construction math, not vibes.
MLOps / Inference — the serving spine: prompt-cache, continuous batching, watchdog + circuit-breaker + memory-ceiling — frontier-grade stability for hours-long local runs on one box.
Multimodal / CV — reads images + video (VLM), palette-steered image-gen, code-rendered video/figures (manim/TikZ) — all MLX.
Design Engineer — a render-and-see critic enforcing WCAG contrast, modular type scale, 8 px grid, OKLCH harmony (not just "looks fine").

One model, fully local, verify-everything — every hat above, on a MacBook.

How it was made

Pruned the MoE experts 256 → 77 by router-weighted saliency (REAP = router_weight × activation_norm, padding-masked), streaming layer-by-layer (~5 GB working set — it never fits in RAM).
Quantized mixed-precision (MLX): experts 3-bit, attention/embeddings/lm_head 4-bit → 99 GB.
Healed with LoRA SFT (--no-mask-prompt, grad-checkpointed). The current v4 rebuild uses a code-first balanced calibration (so the math super-experts survive the prune — v3's coding-only calibration collapsed math) + heal/distill on R1 long-CoT reasoning traces. Router-KD / expert-wise Logit-KD are research-validated recovery stages (optional). (GRPO/RLVR was tried and regressed → SFT.)

What makes it different (built + selftested)

Verified decoding (compiler-steered): generates line-by-line while the real type-checker runs in the loop; a line that adds an error is backtracked. TS 0.3 ms · Python ~0 ms · Rust 34 ms per check. Practical only on Apple Silicon — unified memory lets the model (GPU) and compiler (CPU) share RAM.
The verifier mesh: every output meets its real tool — compile+run+idiomatic lint (clippy/ruff/ gofmt/prettier) for 5 langs, SQL (sqlite), math (SymPy), proofs (Lean 4), design (render+see).
A 51-tool agent with five defense layers the frontier lacks out of the box: trust (checkpoint/rollback, secret-scan, prompt-injection guard, audit, risk-gate), reliability (constraint-pinning vs context-rot, false-success guard, flaky-test re-run, onboarding map), self-improvement (skill library, large-output pointers, clarify-before-assuming), integrity (test-tamper guard, fabrication-proof done, scope enforcement, slopsquat guard), plus a humanizer (kills AI-slop, matches your voice).
Own your repo: scripts/64_own_your_repo.py fine-tunes the model on your private codebase so it writes in your style — a cloud flagship can't be tuned on your private code.
Design soul (render-and-measure critic: WCAG/type-scale/OKLCH), CallSieve zero-token retrieval + live-docs RAG, vision/voice/video (all MLX), code-rendered math/arch figures (matplotlib/manim/TikZ).

Requirements

Apple Silicon, 128 GB unified memory (M5-class recommended), macOS 26/27+. MLX ≥ 0.31.
The architecture (glm_moe_dsa: MLA + DSA sparse attention) needs the bundled patch (glm_moe_dsa.py
- install_glm_dsa_patch.py) — current stock mlx_lm can't load it. Native support is landing upstream (ml-explore/mlx-lm PR #1410); once it merges, recent mlx_lm loads this model with no patch — the bundled patch is the interim loader for older versions.
⚠️ Raise the GPU memory ceiling — required. The model needs ~101.6 GB; macOS caps the GPU working set at ~110 GB by default, so it OOM-crashes (Metal command-buffer timeout) on long generations. Fix before serving:
```
sudo sysctl iogpu.wired_limit_mb=122000        # 122 GB; one-shot (resets on reboot)
sudo bash dist/install_gpu_limit.sh            # OR: persist it via a LaunchDaemon
```
Without this the model appears to "randomly crash" — it's just memory-starved.

Use it

python dist/install_glm_dsa_patch.py          # patch mlx_lm (venv AND LM Studio's bundled engine)
GLM_STREAM_EVAL=0 python -m mlx_lm.server --model models/GLM-5.2-q3a4-v4 \
    --adapter-path heal/adapters-v4           # serve (OpenAI-compatible); v2 + heal/adapters also ship
# query it — `enable_thinking` toggles the reasoning trace (GLM-specific; off = faster, on = harder problems):
curl -s localhost:8080/v1/chat/completions -H 'Content-Type: application/json' \
  -d '{"messages":[{"role":"user","content":"Write a typed debounce in TypeScript."}],"chat_template_kwargs":{"enable_thinking":true}}'
# drive the 51-tool agent on your repo:
python scripts/57_tool_agent.py --repo /path/to/your/repo --apply --task "..." --test "cargo test"
# speed: try --dsa-block-size 32/64/128 (free, pick fastest). External draft is Metal-unstable here; MTP self-spec is the real path.

In LM Studio: run the patch, fully quit + reopen, then load the model.

Design — elite, not just competent (full guide + copy-paste system prompt: design/DESIGN.md, with 9 movement-grounded gold seeds): the base prior reverts to the average of its training (hex + arbitrary spacing), so steer + gate it. Prepend src/design_canon.py's CANON (oklch-only · 8px grid · 1.25 type scale · WCAG · bespoke — no Bootstrap/Tailwind/framework cookie-cutter) as the system prompt for elite output today; audit_design() gates eliteness (OKLCH/grid/scale + rejects framework boilerplate) and the constrained decoder bans non-OKLCH tokens; scripts/76_design_flywheel.py (generate→audit→keep-only-elite→SFT) heals the native prior so it designs elite with no prompt at all.

Performance (M5 Max 128 GB, v4)

Metric	Value
Size	99 GB (from 381 GB mxfp4 / ~1.5 TB bf16)
HumanEval pass@1	19/20 (95%), single-shot
Math GSM8K	8/12 (66%) — recovered from v3's 0/5 (code-first balanced calibration kept the math super-experts alive through the prune)
Algebra (SymPy-checked)	3/4 (75%)
Decode speed	11.3 tok/s (no draft) — see the speed note in limitations
Verified-decode checker	TS 0.3 ms · Python ~0 ms · Rust 34 ms

Benchmark honesty: every number is contamination-checked — HumanEval, GSM8K, and miniF2F test problems are not in the training data (0 % / 0 % / 0.4 % near-dup), so they're reasoned, not memorized. Method + full training-data provenance/licenses: TRAINING_DATA.md.

Which version for your runtime (June 2026 — MLX is now everywhere on Apple Silicon)

Runtime	MLX (this repo)	GGUF (with the family)
`mlx_lm` (CLI / server)	✅ native	—
LM Studio	✅ Mac (dual-backend)	✅ Win/Linux
Ollama 0.19+	✅ Mac (MLX engine, since Mar 2026)	✅ 0.30 (llama.cpp)
macMLX	✅ native (SwiftUI + OpenAI API)	—
`llama.cpp`	—	✅
mlx-swift apps	✅ when `glm_moe_dsa` lands in mlx-swift-lm	—

MLX is the native Apple-Silicon path — mlx_lm · LM Studio (Mac) · Ollama 0.19+ · macMLX all run it (MLX beats llama.cpp ~30-40% on M5). GGUF (shipped with the family) covers llama.cpp + Windows/Linux. Every MLX runtime gets this model the moment glm_moe_dsa lands upstream (mlx-lm PR #1410) — or today via install_glm_dsa_patch.py, which scans every mlx_lm install (LM Studio's, Ollama's, your venv's).

Roadmap — the Demolition family (shrink, keep the soul)

Same masters-trained soul (design · dataviz · code · security · math · prose · architecture · research), every Mac — the elite training lives in the facet-inclusive calibration + heal corpus, which are size-agnostic:

  99GB  : ████████ (baseline, this model)
  64GB  : should hold ~baseline   (96 GB Macs)
  48GB  : should hold high        (64 GB Macs)
  28GB  : the squeeze — watch which facets dip   (36-48 GB Macs)
  14GB  : ⚗️ where does the soul start to break?  (24 GB Macs)
   7GB  : ⚗️ the floor             (16 GB laptops)

Each size: facet-calib → prune harder → quantize → heal (the soul corpus) → soul-retention scorecard (% elite per facet). See design/DESIGN.md.

Honest limitations

Specialist: ~70% of experts pruned — strong in the target niche, weaker outside it. Not the full 743B.
Speed ~11 tok/s decode (reading pace; 3 min for long thinking-ON answers). Partly MLX's still-naive DSA attention kernels (mlx #837 / #3402 — improves for free as MLX matures), partly the bandwidth cost of a 743B-class MoE on a laptop. Measured dead-ends (don't bother): 4-bit re-quant is slower for single-token decode (bandwidth-bound, smaller wins); active-experts 8→4 gives no win at batch=1. Real path: --dsa-block-size sweep (free) → upstream MLX → MTP self-speculative (2.6×, a port for this arch). Not a quant change.
Multilingual ability reduced (optional vocab-trim drops ~31% of tokens).
Design is competent but not yet design-soul-elite (correct structure, but missed OKLCH/grid when tested) — the design-canon heal closes this.
Prompt-cache can OOM under heavy concurrent load. The external speculative draft is Metal-unstable on this MoE — MTP self-speculative is the right path; the external draft is not recommended.

Attribution & license

MIT. Base model © Z.ai (zai-org/GLM-5.2, MIT-licensed) — so this derivative is MIT too: free to use, modify, and redistribute with attribution to Z.ai. The demolition / healing / 51-tool agent tooling is this repo's contribution.