Instructions to use jstarks/Qwen3.5-9B-Weight-Space-Programming with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="jstarks/Qwen3.5-9B-Weight-Space-Programming") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] pipe(text=messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("jstarks/Qwen3.5-9B-Weight-Space-Programming") model = AutoModelForMultimodalLM.from_pretrained("jstarks/Qwen3.5-9B-Weight-Space-Programming", device_map="auto") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=256) print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M # Run inference directly in the terminal: llama cli -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M # Run inference directly in the terminal: llama cli -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama cli -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
Use Docker
docker model run hf.co/jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
- Llama new
- LM Studio
- Jan
- vLLM
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "jstarks/Qwen3.5-9B-Weight-Space-Programming" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "jstarks/Qwen3.5-9B-Weight-Space-Programming", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
- SGLang
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "jstarks/Qwen3.5-9B-Weight-Space-Programming" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "jstarks/Qwen3.5-9B-Weight-Space-Programming", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "jstarks/Qwen3.5-9B-Weight-Space-Programming" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "jstarks/Qwen3.5-9B-Weight-Space-Programming", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Ollama
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with Ollama:
ollama run hf.co/jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
- Unsloth Desktop
- Pi
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:9931/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with Docker Model Runner:
docker model run hf.co/jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
- Lemonade
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
Run and chat with the model
lemonade run user.Qwen3.5-9B-Weight-Space-Programming-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
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://127.0.0.1:9931/v1 hermes config set model.default jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use jstarks/Qwen3.5-9B-Weight-Space-Programming with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M
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://127.0.0.1:9931/v1 \ --custom-model-id "jstarks/Qwen3.5-9B-Weight-Space-Programming:Q4_K_M" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
- Qwen3.5-9B-Weight-Space-Programming (WSP v3.0 โ 9-Subspace Dynamic LoRA Architecture)
- 1. Per-Suite 100-Task Breakdown (
Base->WSP v3.0) - 2. All 5 Models Evaluated Uniformly (
100Tasks Each) - 3. Terminal-Bench 2.0 Cross-Model Evaluation (All 10 Models, 20 Stateful Tasks)
- 4. Synthesized Rank-16
f32Subspaces (S1..S9) & Safetensors Patches - 5. Quickstart on Any Machine (
run_wsp.py/ OpenAI Proxy / Terminal Chat)
- 1. Per-Suite 100-Task Breakdown (
Qwen3.5-9B-Weight-Space-Programming (WSP v3.0 โ 9-Subspace Dynamic LoRA Architecture)
Training-free, closed-form 9-Subspace Weight-Space Programming (WSP v3.0) for Qwen/Qwen3.5-9B (32 layers, hidden_size=4096, intermediate_size=12288, swiglu MLP, 1+w RMSNorm, synthesis_mode=bf16_full_32layer_forward_hooks), evaluated end-to-end across 100 unified benchmark tasks (30 Disjoint Holdout across S1..S9 + 70 Established Public Benchmarks) and 20 live stateful Terminal-Bench 2.0 container tasks in WSL2.
Uniform 100-Task + Terminal-Bench 2.0 Result (
qwen35_9b):
- Unified 100-Task Suite: Base
84/100 (84.0%)-> WSP v3.088/100 (88.0%)(+4.0 pp)- Holdout (
S1..S9,30tasks):26/30 -> 28/30- Established Public (
70tasks:20GSM8K,15HumanEval,15BFCL Irrelevance,20NL2Bash):58/70 -> 60/70- Terminal-Bench 2.0 (
20live WSL2 bash tasks): Base7/20 (35.0%)-> WSP7/20 (35.0%)(0.0 pp;68.4 tok/s)- Speed & VRAM (
-c 16384 -ctk q8_0 -ctv q8_0):72.5 tok/s, load delta5552 MiB, peak delta5650 MiB- Cross-Hardware Fit: RTX 2060 (
6 GB): Tight (~5.3 GiB @ 4k before desktop) | RTX 4070 Ti (12 GB): Yes (16k solo or 8k cascade Tier-2)
1. Per-Suite 100-Task Breakdown (Base -> WSP v3.0)
| Suite | Tasks | Base Pass | WSP v3.0 Pass | Avg Completion Tokens (Base โ WSP) |
|---|---|---|---|---|
S1_THINK (Holdout S1: Think Closure) |
3 |
1/3 (33.3%) |
1/3 (33.3%) |
1024.0 โ 1024.0 |
S2_RESTRAINT (Holdout S2: Tool Restraint & Anti-Sycophancy) |
4 |
4/4 (100.0%) |
4/4 (100.0%) |
361.2 โ 366.2 |
S3_ROUTER (Holdout S3: Multi-Tool Router) |
5 |
5/5 (100.0%) |
5/5 (100.0%) |
116.2 โ 122.6 |
S4_CODE (Holdout S4: Mid-Band Code Synthesis) |
2 |
2/2 (100.0%) |
2/2 (100.0%) |
907.0 โ 1003.0 |
S5_TURN2_JSON (Holdout S5: Turn-2 Structured JSON) |
3 |
3/3 (100.0%) |
3/3 (100.0%) |
49.0 โ 69.3 |
S6_HANDOFF (Holdout S6: Escalation Handoff Gate) |
4 |
4/4 (100.0%) |
4/4 (100.0%) |
151.5 โ 151.5 |
S7_INJ_GUARD (Holdout S7: Indirect Injection Guard) |
3 |
3/3 (100.0%) |
3/3 (100.0%) |
97.0 โ 97.0 |
S8_GROUNDING (Holdout S8: Grounded Abstention) |
3 |
1/3 (33.3%) |
3/3 (100.0%) |
86.3 โ 67.0 |
S9_ERR_RECOV (Holdout S9: Error Self-Recovery) |
3 |
3/3 (100.0%) |
3/3 (100.0%) |
61.7 โ 51.0 |
EST_GSM8K (Established: OpenAI GSM8K (20)) |
20 |
14/20 (70.0%) |
14/20 (70.0%) |
997.3 โ 997.3 |
EST_HUMANEVAL (Established: HumanEval pass@1 (15)) |
15 |
15/15 (100.0%) |
15/15 (100.0%) |
576.4 โ 574.6 |
EST_BFCL_IRR (Established: BFCL v3 Irrelevance (15)) |
15 |
13/15 (86.7%) |
13/15 (86.7%) |
332.6 โ 330.7 |
EST_NL2BASH (Established: InterCode-NL2Bash (20)) |
20 |
16/20 (80.0%) |
18/20 (90.0%) |
319.4 โ 308.9 |
2. All 5 Models Evaluated Uniformly (100 Tasks Each)
Every candidate in the wsp-5model-pipeline study was evaluated under identical decoding (temperature=0.0, seed=42, -c 16384 -ctk q8_0 -ctv q8_0) across all 30 disjoint S1..S9 holdout tasks and all 70 cached established tasks:
Model (slug) |
Hugging Face Repo | GGUF (Q4_K_M) |
Base (100 Tasks) |
WSP v3.0 (100 Tasks) |
Net Delta | Holdout (/30) |
Est. (/70) |
Speed (tok/s) |
Peak VRAM Delta (16k) |
RTX 2060 (6 GB) |
RTX 4070 Ti (12 GB) |
|---|---|---|---|---|---|---|---|---|---|---|---|
qwen35_4b |
Qwen3.5-4B-Weight-Space-Programming |
2.553 GiB |
72/100 (72.0%) |
89/100 (89.0%) |
+17.0 pp |
24 โ 29/30 |
48 โ 60/70 |
117.0 tok/s |
3370 MiB |
Yes (~2.7 GiB spare @ 16k) | Yes (~8.0 GiB spare @ 16k) |
nemotron3_4b |
Nemotron-3-Nano-4B-Weight-Space-Programming |
2.642 GiB |
81/100 (81.0%) |
81/100 (81.0%) |
+0.0 pp |
26 โ 28/30 |
55 โ 53/70 |
131.1 tok/s |
3133 MiB |
Yes (~3.0 GiB spare @ 16k) | Yes (fastest 4B @ 16k) |
qwen35_9b |
Qwen3.5-9B-Weight-Space-Programming |
5.290 GiB |
84/100 (84.0%) |
88/100 (88.0%) |
+4.0 pp |
26 โ 28/30 |
58 โ 60/70 |
72.5 tok/s |
5650 MiB |
Tight (~5.3 GiB @ 4k before desktop) | Yes (16k solo or 8k cascade Tier-2) |
qwen35_2b |
Qwen3.5-2B-Weight-Space-Programming |
1.193 GiB |
57/100 (57.0%) |
55/100 (55.0%) |
-2.0 pp |
23 โ 23/30 |
34 โ 32/70 |
226.5 tok/s |
1979 MiB |
Yes (~4.4 GiB spare @ 16k) | Yes (ultra-light leaf / cascade Tier-1) |
gemma4_12b |
Gemma-4-12B-Weight-Space-Programming |
6.633 GiB |
50/100 (50.0%) |
52/100 (52.0%) |
+2.0 pp |
23 โ 24/30 |
27 โ 28/70 |
51.3 tok/s |
7664 MiB |
No (> 6144 MiB VRAM) | Yes (16k solo; ~3.8 GiB free) |
3. Terminal-Bench 2.0 Cross-Model Evaluation (All 10 Models, 20 Stateful Tasks)
Evaluated live inside clean per-task /app sandboxes in WSL2 (5 max turns, [EXIT_CODE=...] feedback, zero grading errors across all 20 runs โ full per-task traces in benchmarks/terminal_bench_results.json):
Model (slug) |
Hugging Face Repo | WSP Variant | Base (20 Tasks) |
WSP (20 Tasks) |
Net Delta | WSP Breakdown (E / M / H) |
Avg Tokens (Base โ WSP) |
Speed (tok/s) |
|---|---|---|---|---|---|---|---|---|
qwen35_0_8b |
Qwen3.5-0.8B-iPhone-Agent-Weight-Space-Programming |
Dynamic 4S Edge-TB BF16 Hooks | 2/20 (10.0%) |
2/20 (10.0%) |
0.0 pp |
2/10 E ยท 0/9 M ยท 0/1 H |
579.2 โ 654.9 |
32.2 |
qwen35_2b |
Qwen3.5-2B-Weight-Space-Programming |
Dynamic 9S LoRA (Proxy) | 3/20 (15.0%) |
3/20 (15.0%) |
0.0 pp |
2/10 E ยท 1/9 M ยท 0/1 H |
715.8 โ 677.7 |
213.1 |
qwen35_4b |
Qwen3.5-4B-Weight-Space-Programming |
Dynamic 11S LoRA (Hybrid 9S + TB S10/S11) | 3/20 (15.0%) |
4/20 (20.0%) |
+5.0 pp |
3/10 E ยท 1/9 M ยท 0/1 H |
604.6 โ 650.0 |
107.7 |
nemotron3_4b |
Nemotron-3-Nano-4B-Weight-Space-Programming |
Dynamic 9S LoRA (BF16 Hooks) | 3/20 (15.0%) |
3/20 (15.0%) |
0.0 pp |
3/10 E ยท 0/9 M ยท 0/1 H |
830.1 โ 649.6 |
118.0 |
qwen35_9b |
Qwen3.5-9B-Weight-Space-Programming |
Dynamic 11S LoRA (BF16 9S + TB S10/S11) | 7/20 (35.0%) |
7/20 (35.0%) |
0.0 pp |
5/10 E ยท 1/9 M ยท 1/1 H |
686.0 โ 700.2 |
68.4 |
gemma4_e2b |
gemma-4-E2B-it-iPhone-Agent-Weight-Space-Programming |
Dynamic 4S Edge-TB BF16 Hooks | 4/20 (20.0%) |
4/20 (20.0%) |
0.0 pp |
3/10 E ยท 0/9 M ยท 1/1 H |
415.2 โ 426.6 |
20.9 |
gemma4_e4b |
gemma-4-E4B-it-weight-space-programming |
Dynamic 4S Edge-TB BF16 Hooks | 4/20 (20.0%) |
4/20 (20.0%) |
0.0 pp |
3/10 E ยท 0/9 M ยท 1/1 H |
723.4 โ 498.6 |
16.4 |
gemma4_12b |
Gemma-4-12B-Weight-Space-Programming |
Dynamic 9S LoRA (BF16 Hooks) | 6/20 (30.0%) |
6/20 (30.0%) |
0.0 pp |
5/10 E ยท 0/9 M ยท 1/1 H |
1010.6 โ 928.0 |
50.7 |
gemma4_26b_a4b |
Gemma-4-26B-A4B-Weight-Space-Programming |
MoE Mid-Attn S4 (scale=2.0) |
5/20 (25.0%) |
6/20 (30.0%) |
+5.0 pp |
4/10 E ยท 1/9 M ยท 1/1 H |
934.4 โ 896.3 |
17.8 |
qwen38_27b |
Qwen3.8-27B-Weight-Space-Programming |
Dynamic 4S Mid+Late LoRA | 6/20 (30.0%) |
8/20 (40.0%) |
+10.0 pp |
6/10 E ยท 1/9 M ยท 1/1 H |
1013.6 โ 1075.3 |
6.5 |
Fair-evaluation parity note:
qwen35_0_8bstops cleanly on[248046 (<|im_end|>), 248044 (<|endoftext|>)](eliminating multi-turn<tool_response>hallucination loops where pre-fix Base emitted2068.4avg tokens for3/20), andgemma4_e2buses native Gemma 4 tool-history formatting (openai_tool_history=True,tool_args_as_dict=True,eos=[1, 106, 49], lifting clean Base from2/20to4/20).
4. Synthesized Rank-16 f32 Subspaces (S1..S9) & Safetensors Patches
- GGUF LoRA Adapters (
llama-server):adapters/wsp_s1_think.gguf .. wsp_s9_error_recovery.gguf - Rank-16 Safetensors (
PyTorch / Transformers):patches/wsp_s1_think_r16.safetensors .. wsp_s9_error_recovery_r16.safetensors - Combined Unit-Sum Dense BF16 Delta Patch:
patches/wsp_qwen35_9b_patch.safetensors(864.00 MiB)
| Slot | GGUF LoRA (llama.cpp) |
Safetensors (PyTorch) |
Target Layers | ||\Delta W||_F |
Cross Leak | Syntax Leak | Role |
|---|---|---|---|---|---|---|---|
id=0 |
adapters/wsp_s1_think.gguf |
patches/wsp_s1_think_r16.safetensors (5.0 MiB) |
[23, 27, 29, 30, 31] |
2.3674 |
0.193% |
0.230% |
Fast </think> / `<channel |
id=1 |
adapters/wsp_s2_restraint.gguf |
patches/wsp_s2_restraint_r16.safetensors (5.0 MiB) |
[23, 27, 29, 30, 31] |
0.6294 |
0.162% |
0.058% |
Tool restraint & anti-sycophancy pushback |
id=2 |
adapters/wsp_s3_router.gguf |
patches/wsp_s3_router_r16.safetensors (5.0 MiB) |
[23, 27, 29, 30, 31] |
0.6575 |
0.067% |
0.041% |
Multi-tool boundary router (exec, web_search, read_file, write_file) |
id=3 |
adapters/wsp_s4_mid_code.gguf |
patches/wsp_s4_mid_code_r16.safetensors (4.0 MiB) |
[7, 11, 15, 19] |
0.5113 |
0.062% |
0.015% |
Mid-network algorithmic code synthesis |
id=4 |
adapters/wsp_s5_json_schema.gguf |
patches/wsp_s5_json_schema_r16.safetensors (5.0 MiB) |
[23, 27, 29, 30, 31] |
0.5181 |
0.190% |
0.041% |
Turn-2 structured JSON synthesis gate with tools attached |
id=5 |
adapters/wsp_s6_handoff.gguf |
patches/wsp_s6_handoff_r16.safetensors (5.0 MiB) |
[23, 27, 29, 30, 31] |
0.7943 |
0.140% |
0.102% |
Bounded subtask vs. unrecoverable escalate_handoff gate |
id=6 |
adapters/wsp_s7_injection_guard.gguf |
patches/wsp_s7_injection_guard_r16.safetensors (5.0 MiB) |
[23, 27, 29, 30, 31] |
0.5290 |
0.131% |
0.051% |
Indirect prompt-injection defense inside tool responses |
id=7 |
adapters/wsp_s8_grounding.gguf |
patches/wsp_s8_grounding_r16.safetensors (5.0 MiB) |
[23, 27, 29, 30, 31] |
0.6088 |
0.180% |
0.050% |
Grounded context fidelity & missing-field abstention |
id=8 |
adapters/wsp_s9_error_recovery.gguf |
patches/wsp_s9_error_recovery_r16.safetensors (5.0 MiB) |
[23, 27, 29, 30, 31] |
0.5621 |
0.230% |
0.045% |
Recoverable local tool-error self-correction |
5. Quickstart on Any Machine (run_wsp.py / OpenAI Proxy / Terminal Chat)
Option A: One-Command Auto-Download + Dynamic 9-Subspace Proxy
Works on any Windows, Linux, or macOS computer with Python + llama.cpp (llama-server):
pip install huggingface_hub
hf download jstarks/Qwen3.5-9B-Weight-Space-Programming run_wsp.py --local-dir ./qwen35_9b-wsp
python ./qwen35_9b-wsp/run_wsp.py --chat
Omit --chat to expose an OpenAI-compatible dynamic routing proxy on http://127.0.0.1:8090/v1 (with /health and /v1/chat/completions), or pass --download-only to download only the GGUF + 9 GGUF LoRAs + metadata (--include-bf16 also fetches safetensors patches / HF files).
Option B: Manual llama-server + scripts/dynamic_router_9s.py
llama-server -m Qwen3.5-9B-Q4_K_M.gguf --host 127.0.0.1 --port 8089 \
-ngl 99 -np 1 -c 16384 -ctk q8_0 -ctv q8_0 -b 512 -ub 256 \
--jinja --reasoning-format deepseek --lora-init-without-apply --lora adapters/wsp_s1_think.gguf --lora adapters/wsp_s2_restraint.gguf --lora adapters/wsp_s3_router.gguf --lora adapters/wsp_s4_mid_code.gguf --lora adapters/wsp_s5_json_schema.gguf --lora adapters/wsp_s6_handoff.gguf --lora adapters/wsp_s7_injection_guard.gguf --lora adapters/wsp_s8_grounding.gguf --lora adapters/wsp_s9_error_recovery.gguf
Option C: PyTorch / Hugging Face In-Memory Patching (scripts/apply_patch_9s.py)
from transformers import AutoModelForCausalLM
from scripts.apply_patch_9s import REGIME_PRESETS, apply_subspace_scales_to_hf_model
model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3.5-9B", torch_dtype="bfloat16")
apply_subspace_scales_to_hf_model(model, "patches", REGIME_PRESETS["general_fast_think"])
- Downloads last month
- 449
4-bit