Image-Text-to-Text
GGUF
English
Portuguese
ollama
llama.cpp
local-ai
on-device
low-vram
4gb-vram
small-model
tiny-llm
vision
multimodal
thinking
reasoning
chain-of-thought
distillation
tool-calling
qwen
qwen-4b
4b
qwen3.8
qwen3.5
iq3_xxs
imatrix
experimental
conversational
Eval Results (legacy)
Instructions to use sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF 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 sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF 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 sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS # Run inference directly in the terminal: llama cli -hf sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS # Run inference directly in the terminal: llama cli -hf sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
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 sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS # Run inference directly in the terminal: ./build/bin/llama cli -hf sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
Use Docker
docker model run hf.co/sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
- Llama new
- LM Studio
- Jan
- vLLM
How to use sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF", "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/sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
- Ollama
How to use sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF with Ollama:
ollama run hf.co/sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
- Unsloth Desktop
- Pi
How to use sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
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": "sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF with Docker Model Runner:
docker model run hf.co/sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
- Lemonade
How to use sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
Run and chat with the model
lemonade run user.Qwen-4B-Bugio-Vision-Thinking-GGUF-IQ3_XXS
List all available models
lemonade list
- Hermes Agent
How to use sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
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 sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS
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 "sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF:IQ3_XXS" \ --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"
Download examples/transcripts.json from sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF: direct link, hf CLI and curl.
- Browser
- Download file 9.62 kB
-
https://huggingface.co/sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF/resolve/main/examples/transcripts.json
- Command line
-
hf download hf://sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF/examples/transcripts.json
-
curl -L -o transcripts.json https://huggingface.co/sysrobson/Qwen-4B-Bugio-Vision-Thinking-GGUF/resolve/main/examples/transcripts.json
9.62 kB
| { | |
| "model": "bugio-4b", | |
| "results": [ | |
| { | |
| "demo": "vision-fix-ts-error", | |
| "think": false, | |
| "prompt": "This is a screenshot from my editor. Explain the error in one sentence, then give the corrected line 4.", | |
| "thinking": "", | |
| "content": "The error occurs because the `reduce` callback's second parameter `i` is the accumulator object from the previous iteration, which is a `number`, but the code tries to access `i.qty` without a type assertion, causing a type mismatch.\n\n```ts\nreturn items.reduce((sum, i) => sum + i.price * i.qty, 0)\n```\n\nThe fix is to add the type assertion `: number` to the second parameter `i` of the `reduce` callback, so TypeScript correctly infers that `i` is an `Item` object.", | |
| "tool_calls": null, | |
| "tokens": 120, | |
| "tokens_per_second": 15.3, | |
| "seconds": 30.9 | |
| }, | |
| { | |
| "demo": "vision-ui-to-code", | |
| "think": true, | |
| "prompt": "Recreate this screen as a single React + TypeScript component with inline styles. Code only.", | |
| "thinking": "The user wants a recreation of a \"Create account\" form using React and TypeScript.\nI need to:\n1. Set up a React component structure.\n2. Use inline styles (or inline style objects) as requested.\n3. Replicate the layout:\n * Centered card on a light background.\n * Header: \"Create account\" title, \"Start your free trial today\" subtext.\n * Two input fields: Email and Password.\n * A green \"Sign up\" button.\n * A link below the button: \"Already have an account? Log in\".\n\nI will use a simple `useState` to handle basic interactivity (like clicking the button or toggling focus) to make it feel \"live,\" even if no backend logic is present. I'll stick to standard HTML inputs styled with CSS classes or inline styles to keep it clean.\n\nPlan:\n- Component: `CreateAccountForm`.\n- State: `email`, `password`, `isSuccess`, `isError`.\n- Styles: I will use a `const styles` object approach or inline styles for simplicity, adhering to the prompt's \"inline styles\" request. I'll use a `style` object for cleaner code within the component.", | |
| "content": "```tsx\nimport React, { useState } from 'react';\n\ntype FormState = {\n email: string;\n password: string;\n emailError: string | null;\n passwordError: string | null;\n};\n\nconst CreateAccountForm: React.FC = () => {\n const [formData, setFormData] = useState<FormState>({\n email: '',\n password: '',\n emailError: null,\n passwordError: null,\n });\n\n const handleChange = (e: React.ChangeEvent<HTMLInputElement>) => {\n const { name, value } = e.target;\n setFormData((prev) => ({\n ...prev,\n [name]: value,\n }));\n };\n\n const handleSubmit = (e: React.FormEvent) => {\n e.preventDefault();\n // Basic validation\n if (formData.email && formData.email.includes('@')) {\n setFormData((prev) => ({ ...prev, emailError: null }));\n } else {\n setFormData((prev) => ({ ...prev, emailError: 'Please enter a valid email.' }));\n }\n\n if (formData.password && formData.password.length >= 8) {\n setFormData((prev) => ({ ...prev, passwordError: null }));\n } else {\n setFormData((prev) => ({ ...prev, passwordError: 'Password must be at least 8 characters.' }));\n }\n };\n\n return (\n <div style={styles.container}>\n <div style={styles.card}>\n <h1 style={styles.title}>Create account</h1>\n <p style={styles.subtitle}>Start your free trial today</p>\n\n <form onSubmit={handleSubmit} style={styles.form}>\n <div style={styles.inputGroup}>\n <label style={styles.label}>Email</label>\n <input\n style={styles.input}\n type=\"email\"\n value={formData.email}\n onChange={handleChange}\n />\n </div>\n\n <div style={styles.inputGroup}>\n <label style={styles.label}>Password</label>\n <input\n style={styles.input}\n type=\"password\"\n value={formData.password}\n onChange={handleChange}\n />\n </div>\n\n <button type=\"submit\" style={styles.button}>\n Sign up\n </button>\n </form>\n\n <a href=\"#\" style={styles.link}>\n Already have an account? Log in\n </a>\n </div>\n </div>\n );\n};\n\nconst styles = {\n container: {\n display: 'flex',\n justifyContent: 'center',\n alignItems: 'center',\n minHeight: '100vh',\n backgroundColor: '#F2F5F9',\n padding: '20px',\n fontFamily: 'Arial, sans-serif',\n },\n card: {\n backgroundColor: '#ffffff',\n borderRadius: '8px',\n padding: '40px',\n maxWidth: '400px',\n width: '100%',\n boxShadow: '0 0 10px rgba(0, 0, 0, 0.1)',\n display: 'flex',\n flexDirection: 'column',\n gap: '16px',\n textAlign: 'center',\n },\n title: {\n margin: '0',\n fontSize: '24px',\n color: '#111',\n },\n subtitle: {\n margin: '0',\n color: '#666',\n fontSize: '14px',\n },\n form: {\n display: 'flex',\n flexDirection: 'column',\n gap: '12px',\n },\n inputGroup: {\n display: 'flex',\n flexDirection: 'column',\n gap: '4px',\n },\n label: {\n margin: '0',\n fontSize: '14px',\n color: '#333',\n fontWeight: '700',\n },\n input: {\n padding: '12px',\n borderRadius: '8px',\n border: '1px solid #e0e0e0',\n backgroundColor: '#f9f9f9',\n fontSize: '16px',\n color: '#333',\n outline: 'none',\n },\n button: {\n padding: '12px',\n borderRadius: '8px',\n backgroundColor: '#008098',\n color: '#ffffff',\n fontSize: '16px',\n fontWeight: '700',\n border: 'none',\n cursor: 'pointer',\n outline: 'none',\n },\n link: {\n display: 'block',\n margin: '12px',\n color: '#008098',\n textDecoration: 'none',\n fontSize: '14px',\n cursor: 'pointer',\n },\n};\n\nexport default CreateAccountForm;\n```\n\nThis component recreates the design exactly as shown in the image.\n\n**Key features of this implementation:**\n\n- **State Management:** Uses `useState` to handle the form inputs and basic validation logic (checking for valid email format and password length).\n- **Styling:** All styles are defined within a single `styles` object using inline CSS classes.\n- **Layout:** The card is centered on the screen using Flexbox.\n- **Interactivity:** The buttons and inputs are styled to match the green/white theme of the original image.", | |
| "tool_calls": null, | |
| "tokens": 1545, | |
| "tokens_per_second": 15.0, | |
| "seconds": 106.9 | |
| }, | |
| { | |
| "demo": "code-typescript", | |
| "think": false, | |
| "prompt": "Write a TypeScript function groupBy<T, K extends PropertyKey>(items: T[], key: (item: T) => K): Record<K, T[]>. Code only.", | |
| "thinking": "", | |
| "content": "```typescript\nfunction groupBy<T, K extends PropertyKey>(\n items: T[],\n key: (item: T) => K\n): Record<K, T[]> {\n const groups: Record<K, T[]> = {} as Record<K, T[]>;\n\n for (const item of items) {\n const key = key(item);\n if (!groups[key]) {\n groups[key] = [];\n }\n groups[key].push(item);\n }\n\n return groups;\n}\n```\n\n**Usage example:**\n\n```typescript\nconst users = [\n { name: \"Alice\", age: 25 },\n { name: \"Bob\", age: 30 },\n { name: \"Charlie\", age: 25 },\n];\n\nconst groupedByAge = groupBy(users, (user) => user.age);\n// { 25: [ {name:\"Alice\",age:25}, {name:\"Charlie\",age:25} ], 30: [ {name:\"Bob\",age:30} ] }\n```\n\nThis function groups an array of objects based on the result of a provided key function, returning a `Record<K, T[]>` where `K` is the type of the key.", | |
| "tool_calls": null, | |
| "tokens": 274, | |
| "tokens_per_second": 15.7, | |
| "seconds": 20.3 | |
| }, | |
| { | |
| "demo": "thinking-logic", | |
| "think": true, | |
| "prompt": "A train leaves at 14:35 and arrives at 17:10. How long is the trip in minutes?", | |
| "thinking": "The user wants to find the duration of a train trip.\n\n**Given:**\n- Departure time: 14:35 (2:35 PM)\n- Arrival time: 17:10 (5:10 PM)\n\n**Step 1: Calculate the time difference.**\n- From 14:35 to 17:10\n\n**Step 2: Break it down.**\n- From 14:35 to 17:35 would be exactly 3 hours. But the train arrives at 17:10, which is 25 minutes before 17:35.\n- 3 hours = 180 minutes\n- Subtract 25 minutes: 180 - 25 = 155 minutes\n\n**Verification:**\n- 14:35 + 155 minutes = 14:35 + 1 hour = 15:35 + 50 minutes = 16:25 + 30 minutes = 16:55 + 1 hour = 17:35 - 25 minutes = 17:10.\n\nThe answer is 155 minutes.", | |
| "content": "The trip takes **155 minutes**.\n\n**Breakdown:**\n- From 14:35 to 17:10 is **2 hours and 35 minutes**.\n- 2 hours = 120 minutes\n- 120 + 35 = **155 minutes**", | |
| "tool_calls": null, | |
| "tokens": 338, | |
| "tokens_per_second": 15.7, | |
| "seconds": 24.3 | |
| }, | |
| { | |
| "demo": "tool-call", | |
| "think": false, | |
| "prompt": "What's the weather in Lisbon? Use the tool.", | |
| "thinking": "", | |
| "content": "", | |
| "tool_calls": [ | |
| { | |
| "id": "iuMcTZgHGwCYegLk7KMvlT6E3HO6YbQQ", | |
| "function": { | |
| "index": 0, | |
| "name": "get_weather", | |
| "arguments": { | |
| "city": "Lisbon" | |
| } | |
| } | |
| } | |
| ], | |
| "tokens": 28, | |
| "tokens_per_second": 15.9, | |
| "seconds": 6.1 | |
| }, | |
| { | |
| "demo": "tool-result", | |
| "think": false, | |
| "prompt": "{\"temp_c\": 21, \"sky\": \"sunny\"}", | |
| "thinking": "", | |
| "content": "The weather in Lisbon is currently **21°C** and **sunny**.", | |
| "tool_calls": null, | |
| "tokens": 17, | |
| "tokens_per_second": 16.8, | |
| "seconds": 3.9 | |
| } | |
| ] | |
| } |