Instructions to use Lonelyguyse1/halide-vision with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Lonelyguyse1/halide-vision with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="Lonelyguyse1/halide-vision", 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?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("Lonelyguyse1/halide-vision")
model = AutoModelForMultimodalLM.from_pretrained("Lonelyguyse1/halide-vision", 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=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

llama-cpp-python

How to use Lonelyguyse1/halide-vision with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Lonelyguyse1/halide-vision",
	filename="minicpm-v-4.6-merged-v3-q4_k_m.gguf",
)

llm.create_chat_completion(
	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"
					}
				}
			]
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use Lonelyguyse1/halide-vision 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 Lonelyguyse1/halide-vision:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Lonelyguyse1/halide-vision:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Lonelyguyse1/halide-vision:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Lonelyguyse1/halide-vision:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf Lonelyguyse1/halide-vision:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Lonelyguyse1/halide-vision:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf Lonelyguyse1/halide-vision:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Lonelyguyse1/halide-vision:Q4_K_M

Use Docker

docker model run hf.co/Lonelyguyse1/halide-vision:Q4_K_M

LM Studio
Jan

vLLM

How to use Lonelyguyse1/halide-vision with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Lonelyguyse1/halide-vision"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Lonelyguyse1/halide-vision",
		"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/Lonelyguyse1/halide-vision:Q4_K_M

SGLang

How to use Lonelyguyse1/halide-vision 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 "Lonelyguyse1/halide-vision" \
    --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": "Lonelyguyse1/halide-vision",
		"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 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 "Lonelyguyse1/halide-vision" \
        --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": "Lonelyguyse1/halide-vision",
		"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"
						}
					}
				]
			}
		]
	}'

Ollama
How to use Lonelyguyse1/halide-vision with Ollama:
```
ollama run hf.co/Lonelyguyse1/halide-vision:Q4_K_M
```

Unsloth Studio

How to use Lonelyguyse1/halide-vision with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Lonelyguyse1/halide-vision to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Lonelyguyse1/halide-vision to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Lonelyguyse1/halide-vision to start chatting

How to use Lonelyguyse1/halide-vision with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Lonelyguyse1/halide-vision:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "Lonelyguyse1/halide-vision:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Lonelyguyse1/halide-vision with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Lonelyguyse1/halide-vision: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://127.0.0.1:8080/v1
hermes config set model.default Lonelyguyse1/halide-vision:Q4_K_M

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use Lonelyguyse1/halide-vision with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Lonelyguyse1/halide-vision: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://127.0.0.1:8080/v1 \
  --custom-model-id "Lonelyguyse1/halide-vision: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"

Docker Model Runner
How to use Lonelyguyse1/halide-vision with Docker Model Runner:
```
docker model run hf.co/Lonelyguyse1/halide-vision:Q4_K_M
```

Lemonade

How to use Lonelyguyse1/halide-vision with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Lonelyguyse1/halide-vision:Q4_K_M

Run and chat with the model

lemonade run user.halide-vision-Q4_K_M

List all available models

lemonade list

Lonelyguyse1 commited on Jun 14

Commit

627d324

verified ·

1 Parent(s): d76ad9e

Upload Halide MiniCPM-V 4.6 merged checkpoint from minicpm-v-4.6-merged-v7-crack-curriculum-r1-ckpt625

Browse files

Files changed (2) hide show

README.md +31 -33
model-00001-of-00002.safetensors +1 -1

README.md CHANGED Viewed

@@ -1,55 +1,53 @@
----
 license: apache-2.0
 base_model: openbmb/MiniCPM-V-4.6
 library_name: transformers
 pipeline_tag: image-text-to-text
 tags:
   - film
-  - analog-photography
-  - image-to-text
   - defect-detection
   - minicpm-v
 ---
 # Halide Vision
-Fine-tuned MiniCPM-V 4.6 checkpoint for Project Halide, an edge-native diagnostic engine for analog film scans.
-Owner: [Lonelyguyse1](https://huggingface.co/Lonelyguyse1)
-Project Space: <https://huggingface.co/spaces/Lonelyguyse1/project-halide>
-## Intended Use
-This model is used by Project Halide to extract visible defect candidates from analog film scans. It emits JSON bounding boxes for these labels:
-- `dust`
-- `dirt`
-- `scratch`
-- `long_hair`
-- `short_hair`
-The downstream Halide app validates the JSON, drops invalid or low-confidence boxes, merges near duplicates, and sends the cleaned evidence to Nemotron-Mini-4B-Instruct for diagnostic reasoning.
-## Runtime Notes
 - Base model: `openbmb/MiniCPM-V-4.6`
-- Recommended image setting: `downsample_mode="4x"`
-- Runtime target: GPU-backed Gradio Space, Modal GPU job, or another CUDA runtime
-- Local CPU inference is intentionally blocked in the Halide app
-## Artifacts
-This repo contains the merged checkpoint and a llama.cpp-oriented GGUF artifact:
-- `model-00001-of-00002.safetensors`
-- `model-00002-of-00002.safetensors`
-- `minicpm-v-4.6-merged-v3-q4_k_m.gguf`
-## Limitations
-This is a small-data domain adaptation model. It should be used as an inspection aid, not as a final archival judgment. Natural image content can resemble scratches or hair-like artifacts, so Project Halide uses prompt-level non-film rejection, schema validation, confidence filtering when available, and diagnosis text that keeps uncertainty visible.
-## License
-Apache 2.0.

+---
 license: apache-2.0
 base_model: openbmb/MiniCPM-V-4.6
 library_name: transformers
 pipeline_tag: image-text-to-text
 tags:
   - film
+  - computer-vision
   - defect-detection
   - minicpm-v
 ---
 # Halide Vision
+Halide Vision is a MiniCPM-V 4.6 checkpoint fine-tuned for analog film-scan
+defect extraction. It is maintained by
+[Lonelyguyse1](https://huggingface.co/Lonelyguyse1) for
+[Project Halide](https://github.com/Lonelyguyse1/Project-Halide).
+The model emits JSON defect proposals for dust, dirt, scratches, hair-like
+surface contamination, emulsion damage, chemical stains, and light leaks. The
+Project Halide runtime validates the JSON schema, removes low-confidence or
+duplicate boxes, and uses tiled inspection when large scans hide thin crack
+networks at full-frame scale.
+## Training Summary
 - Base model: `openbmb/MiniCPM-V-4.6`
+- Training method: LoRA fine-tuning with LLaMA-Factory, merged for inference
+- Curriculum: FilmDamageSimulator annotations, procedural film-defect positives,
+  hard clean negatives, and a v7 crack curriculum
+- Held-out private negatives: used only for evaluation, not for training
+## Held-Out Smoke Result
+Final v7 checkpoint with 960 px tiled fallback:
+| Sample | Expected surface condition | Result |
+| --- | --- | --- |
+| negative1 | Long scratches across portrait | 8 defects |
+| negative2 | Abraded emulsion and dirt patches | 9 defects |
+| negative3 | Severe emulsion damage and debris | 6 defects |
+| negative4 | Near-clean hard negative | 0 defects |
+| negative5 | Broad lifted crack network | 45 defects |
+## Runtime Notes
+This model is intended to run inside Project Halide with GPU inference. The
+runtime refuses local CPU model inference and does not call cloud inference APIs.
+Use the model as an inspection aid. It can over-box broad damage regions, and
+film metadata should be treated as context unless verified by notes or edge
+marks.

model-00001-of-00002.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5b9a11d760a6f71ce7846cc710ae45bde0668909be1ca2361280fb928d287977
 size 1995700560

 version https://git-lfs.github.com/spec/v1
+oid sha256:eaaf0f204600222d235e7d3e2afbbe47514ae2fdcfb19418f4f9e4206ee65e87
 size 1995700560