Instructions to use H3tY/qwen3.5-lumenraw-optimize with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use H3tY/qwen3.5-lumenraw-optimize 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 H3tY/qwen3.5-lumenraw-optimize:Q4_K_M # Run inference directly in the terminal: llama cli -hf H3tY/qwen3.5-lumenraw-optimize:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf H3tY/qwen3.5-lumenraw-optimize:Q4_K_M # Run inference directly in the terminal: llama cli -hf H3tY/qwen3.5-lumenraw-optimize: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 H3tY/qwen3.5-lumenraw-optimize:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama cli -hf H3tY/qwen3.5-lumenraw-optimize:Q4_K_M
Use Docker
docker model run hf.co/H3tY/qwen3.5-lumenraw-optimize:Q4_K_M
- Llama new
- LM Studio
- Jan
- vLLM
How to use H3tY/qwen3.5-lumenraw-optimize with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "H3tY/qwen3.5-lumenraw-optimize" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "H3tY/qwen3.5-lumenraw-optimize", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }'Use Docker
docker model run hf.co/H3tY/qwen3.5-lumenraw-optimize:Q4_K_M
- Ollama
How to use H3tY/qwen3.5-lumenraw-optimize with Ollama:
ollama run hf.co/H3tY/qwen3.5-lumenraw-optimize:Q4_K_M
- Unsloth Desktop
- Pi
How to use H3tY/qwen3.5-lumenraw-optimize with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf H3tY/qwen3.5-lumenraw-optimize: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": "H3tY/qwen3.5-lumenraw-optimize:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use H3tY/qwen3.5-lumenraw-optimize with Docker Model Runner:
docker model run hf.co/H3tY/qwen3.5-lumenraw-optimize:Q4_K_M
- Lemonade
How to use H3tY/qwen3.5-lumenraw-optimize with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull H3tY/qwen3.5-lumenraw-optimize:Q4_K_M
Run and chat with the model
lemonade run user.qwen3.5-lumenraw-optimize-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use H3tY/qwen3.5-lumenraw-optimize with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf H3tY/qwen3.5-lumenraw-optimize: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 H3tY/qwen3.5-lumenraw-optimize:Q4_K_M
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use H3tY/qwen3.5-lumenraw-optimize with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf H3tY/qwen3.5-lumenraw-optimize: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 "H3tY/qwen3.5-lumenraw-optimize: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 LumenRAW Optimize (4B)
A fine-tune of Qwen3.5-4B for LUMEN RAW, a RAW photo editor. It does two jobs, locally and in Chinese or English:
- Natural-language editing: turns a request such as "ๅๆๅคฉ็ฉบ๏ผไฟ็ไบๅฑ" or "make it warmer", with the photo's statistics and optionally a preview image, into LUMEN RAW's edit JSON (global sliders, HSL, colour grading, curves, LUT strength, AI masks for sky / subject / water / vegetation โฆ).
- Operating LUMEN RAW through its 32 MCP tools: import, open, edit, masks, ratings, export, save.
It is not a general chat model. It expects the prompts LUMEN RAW's tools use, which are in prompts/.
It was trained with LoRA on answers from a larger teacher model, DeepSeek-V4.1-Flash-1017 (through DeepSeek's API). The training set included 400 corrections, cases where the teacher took over from this model's own mistakes (section Training).
Files
| File | Size | Use |
|---|---|---|
qwen3.5-lumenraw-optimize-Q4_K_M.gguf |
2.8 GB | The model for llama.cpp / LM Studio (recommended) |
qwen3.5-lumenraw-optimize-Q8_0.gguf |
4.6 GB | Same model, 8-bit: no dangerous action in the tests, 20โ35% slower |
mmproj-qwen3.5-lumenraw-optimize-f16.gguf |
0.67 GB | Vision encoder (needed for requests with a preview image) |
adapter/ |
85 MB | The LoRA adapter for Qwen/Qwen3.5-4B (revision 851bf6e), PEFT format |
prompts/system_prompt.txt |
System prompt for natural-language editing | |
prompts/agent_prompt.txt, prompts/tools.json |
System prompt and tool definitions (OpenAI function format) for tool use |
Usage
Tested on Windows on 2026-11-17 with LM Studio 0.4.14, Ollama 0.40.16 and llama.cpp, each with a picture.
LM Studio: search qwen3.5-lumenraw-optimize in Discover and download Q4_K_M (3.49 GB; the vision
file comes with it). Set the context length to 8192 or more, then load it. The local server is
http://127.0.0.1:1234/v1, model id qwen3.5-lumenraw-optimize.
Ollama: ollama run hf.co/H3tY/qwen3.5-lumenraw-optimize:Q4_K_M downloads the model and the vision
file (3.5 GB in all).
In LUMEN RAW 2.3: open the natural-language panel's settings, choose LM Studio or Ollama, pick this model,
set thinking to off and tick the preview image option. LUMEN RAW 2.3 sends its own built-in prompt rather than
prompts/system_prompt.txt, and with it this model did 88.0% of the held-out set (the original Qwen3.5-4B:
82.9%), against 98.1% with the prompt it was trained on.
llama.cpp:
llama-server -m qwen3.5-lumenraw-optimize-Q4_K_M.gguf --mmproj mmproj-qwen3.5-lumenraw-optimize-f16.gguf \
--jinja -c 32768 -ngl 999
Thinking is off by default. The model was trained and tested with thinking off, so its chat template
leaves thinking out unless a request sets "chat_template_kwargs": {"enable_thinking": true}. The tests
used temperature: 0.3.
Natural-language editing. The system message is prompts/system_prompt.txt. The user message has
the form LUMEN RAW builds:
็
ง็ไฟกๆฏ๏ผ{"ๆ ผๅผ": "NEF", "ๅฐบๅฏธ": "6048 ร 4024", "็ธๆบ": "โฆ", โฆ}
ๅฝๅๆๆ็็ป่ฎกไฟกๆฏ๏ผ{"ไบฎๅบฆๅไฝ๏ผ0โ1๏ผ": {โฆ}, "้ซๅ
ๆบขๅบๆฏไพ": 0.003, โฆ, "่ฒๅฝฉๅพๅ": "ๆดไฝๅๆ"}
ๅฝๅๅๆฐ๏ผๆชๅๅบ็้ฝๆฏ้ป่ฎคๅผ๏ผ๏ผ{"adjustments": {"temperature": 28.0}}
็จๆทๆไปค๏ผๅๆๅคฉ็ฉบ๏ผไฟ็ไบๅฑ
With a preview, the image (JPEG, long edge 768) goes in the same message. The answer is one JSON object,
for example {"explanation": "โฆ", "masks": [{"region": "sky", "adjustments": {"exposure": -0.6, "highlights": -30}}]}.
Tool use. The system message is prompts/agent_prompt.txt and the tools are prompts/tools.json.
The user message is the request followed by the editor's state, e.g.
Export the selected photos to E:/batch as tif\n\nCurrent photo: D:/photos/DSC_0001.NEF; selected: โฆ.
Validate what the model returns before applying it, as LUMEN RAW does. Answers can occasionally be malformed (see Limitations).
Evaluation
The benchmarks were run with LUMEN RAW's own validator and command layer, with no repair of the model's output.
- Seed set (102): hand-written cases.
- Held-out set (158): test phrasings and test photos.
- Hard set (260): compound, vague, multi-turn, settings-preserving and tool cases.
- Photo set (46): vision.
No benchmark phrasing or photo was in the training data: photos are split by their sha256, and phrasings by template.
| Did what was asked | Seed 102 | Held-out 158 | Hard 260 | Hard: tool cases |
|---|---|---|---|---|
| Teacher (DeepSeek-V4.1-Flash-1017) | 97.1% | 98.1% | 98.5% | 99.2% |
| Qwen3.5-4B, first round (NL only), Q4_K_M | 95.1% | 97.5% | 88.1% | 83.5% |
| This model, Q4_K_M | 93.1% | 98.1% | 96.5% | 100% |
| This model, Q8_0 | 97.1% | 98.7% | 95.0% | 100% |
| Strictly valid as given (seed / held-out / hard) | Dangerous actions (all 520) | Median request time | |
|---|---|---|---|
| Teacher | 93.8% / 90.5% / 95.0% | 1 (0.19%) | 1.3โ1.6 s (API) |
| First round, Q4_K_M | 97.5% / 97.3% / 97.1% | 21 (4.04%) | 0.5โ0.8 s |
| This model, Q4_K_M | 95.0% / 97.3% / 100% | 2 (0.38%) | 0.5โ0.9 s |
| This model, Q8_0 | 96.2% / 96.0% / 99.3% | 0 | 0.6โ1.2 s |
The rows for this model are the files published here, re-tested on 2026-11-17. Before the re-export (see Changes) the same weights scored Q4_K_M 92.2% / 98.7% / 96.5% with 2 dangerous actions and Q8_0 96.1% / 98.7% / 95.8% with 1. Splitting the merged tensors changes the rounding of the matrix products, not the weights, so a few answers sampled at temperature 0.3 come out differently.
A dangerous action is an unrequested reset or removal of a LUT or mask, an export over an original or an existing file without being asked, or a file written when nothing should change. Local times are on one AMD RX 9070 XT. Photo set (Q4_K_M): 97.8% with statistics + image, 100% with the image only.
Limitations
- "Export over the original file." When the user explicitly asks to export over the RAW file, the model
still calls
export_photoonto the original withoverwrite: true. This accounts for both remaining dangerous actions. LUMEN RAW's export command refuses it, so no original is lost there. In any other host, keep such a guard: never give the model a tool that can destroy files unchecked. - 2/520 is a point estimate. The 95% upper bound of the dangerous-action rate is about 1.2%.
- Changing one's mind across turns is weak. After "warmer", "็ฎไบ๏ผ่ฟๆฏๅทไธ็นๅง" should end up cooler than the original; the model often only takes some of the warmth back (6/15, teacher 15/15).
- Occasional malformed answers. For example, an
hslobject written inside themaskslist, a mask exposure on a ยฑ100 scale (the range is โ5โฆ5 EV), orgrainunderadjustmentsinstead ofeffects. - Scope. It was tested only on LUMEN RAW's tasks, in Chinese and English. The edits it proposes are AI-generated; tell your users so.
Training
- Method: LoRA (r 16, ฮฑ 32, dropout 0).
- Targets: q/k/v/o of the 8 full-attention layers and gate/up/down of all 32 layers.
- 21.2 M trainable parameters (0.46% of the model). Merged, the 128 target matrices change by about 1% each (0.57โ1.48%).
- Frozen: the vision encoder, the projections of the 24 linear-attention layers, the embeddings and the output layer.
- Data: 1,016 training samples (137 more for validation), all with thinking off.
- 666 answers from the teacher: 598 natural-language, 68 tool use.
- 350 corrections: the first-round model answered training prompts; where it failed or did something dangerous, the teacher continued from the model's own state and only the teacher's turns were trained.
- Photos: openly licensed (CC0) photos, training and validation splits only.
- Schedule: 2 epochs, 342 steps of 8 examples, learning rate 1e-4, up to 16,384 tokens per example. Validation loss 0.330 โ 0.324.
- Hardware: one AMD Radeon RX 9070 XT (16 GB), ROCm PyTorch on Windows, about 14 h.
The teacher's outputs were obtained under the DeepSeek Open Platform Terms of Service (effective 2026-04-29). Section 4.2 assigns the outputs to the user and lists "training other models (such as model distillation)" as a permitted use. This model is not affiliated with or endorsed by DeepSeek, Alibaba Cloud (Qwen) or any camera maker.
Changes
- 2026-11-17: the Q4_K_M and Q8_0 files were re-exported. The weights are unchanged. Each linear-attention
layer now stores its tensors separately:
attn_qkv+attn_gateandssm_beta+ssm_alpha, row slices of the mergedattn_qkv_gateandssm_beta_alpha, with no re-quantisation. With the merged tensors, LM Studio 0.4.14 and Ollama 0.40.16 failed with "tensor 'blk.0.ssm_beta.weight' not found". The chat template was also changed to leave thinking off by default. The first files are still in the repo history.
License
Apache-2.0, like the base model (see LICENSE). Changes from Qwen3.5-4B: LoRA fine-tuning as described
above, merged into the weights and quantised with llama.cpp.
ไธญๆ่ฏดๆ
็จ้๏ผ ่ฟๆฏ LUMEN RAW ไฟฎๅพ่ฝฏไปถ็ไธ็จๆฌๅฐๆจกๅ๏ผๅบไบ Qwen3.5-4B ๅพฎ่ฐ๏ผๅไธคไปถไบ๏ผ
- ๆไธญๆๆ่ฑๆ็ไฟฎๅพ่ฆๆฑ่ฝฌๆขๆ LUMEN RAW ็ๅๆฐ JSON๏ผ
- ้่ฟ 32 ไธช MCP ๅทฅๅ ทๆไฝ LUMEN RAWใ
ๅฎไธๆฏ้็จ่ๅคฉๆจกๅ๏ผ้่ฆ้
ๅ prompts/ ้็ๆ็คบ่ฏไฝฟ็จ๏ผๅนถไธ่ฆๅ
ณ้ญๆ่ใ
่ฎญ็ป๏ผ ๆๅธๆฏ DeepSeek-V4.1-Flash-1017๏ผ่ฎญ็ป้ๅ ๆฌ 400 ๆก็บ ้ๆ ทๆฌ๏ผๅณๆๅธๆฅๆๆนๆญฃ่ฟไธชๆจกๅ่ชๅทฑๅ้็ๅ็ญใ
่ฏๆต๏ผ Q4_K_M ็ๆฌๅจ็งๅญ้ใๆต่ฏ้ใๅฐ้พ้ๅ ฑ 520 ๆก้๏ผๅฑ้ฉๆไฝ 2 ๆฌก๏ผ0.38%๏ผ๏ผๅฐ้พ้ๅๅฐ่ฆๆฑ็ไบ 96.5%๏ผๆๅธๆฏ 98.5%ใ
ๅทฒ็ฅ้ฎ้ข๏ผ
- ็จๆทๆ็กฎ่ฆๆฑโๅฏผๅบ่ฆ็ๅๆไปถโๆถ๏ผๆจกๅไปไผ็ งๅใLUMEN RAW ็ๅฏผๅบๅฝไปคไผๆ็ป่ฟไธช่ฐ็จ๏ผๅจๅซ็่ฝฏไปถ้ไฝฟ็จๆถ๏ผ่ฏทไฟ็่ฟ็ฑปไฟๆคใ
- ๅค่ฝฎๅฏน่ฏ้ๆนไธปๆ็่ฏทๆฑๅค็ๅพไธๅฅฝใ
่ฎธๅฏ๏ผ Apache-2.0ใ
- Downloads last month
- 170
4-bit
8-bit