Instructions to use roshan-soni/qwen-3.5-2b-q4-k-m with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use roshan-soni/qwen-3.5-2b-q4-k-m 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 roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf roshan-soni/qwen-3.5-2b-q4-k-m: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 roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf roshan-soni/qwen-3.5-2b-q4-k-m: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 roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M

Use Docker

docker model run hf.co/roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M

LM Studio
Jan
Ollama
How to use roshan-soni/qwen-3.5-2b-q4-k-m with Ollama:
```
ollama run hf.co/roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M
```
Unsloth Desktop

How to use roshan-soni/qwen-3.5-2b-q4-k-m with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf roshan-soni/qwen-3.5-2b-q4-k-m: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:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use roshan-soni/qwen-3.5-2b-q4-k-m with Docker Model Runner:
```
docker model run hf.co/roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M
```

Lemonade

How to use roshan-soni/qwen-3.5-2b-q4-k-m with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M

Run and chat with the model

lemonade run user.qwen-3.5-2b-q4-k-m-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use roshan-soni/qwen-3.5-2b-q4-k-m with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf roshan-soni/qwen-3.5-2b-q4-k-m: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 roshan-soni/qwen-3.5-2b-q4-k-m:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use roshan-soni/qwen-3.5-2b-q4-k-m with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf roshan-soni/qwen-3.5-2b-q4-k-m: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 "roshan-soni/qwen-3.5-2b-q4-k-m: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"

roshan-soni commited on Jun 24

Commit

e5e1eb5

verified ·

1 Parent(s): 80f4123

Upload README.md with huggingface_hub

Browse files

Files changed (1) hide show

README.md +15 -30

README.md CHANGED Viewed

@@ -4,24 +4,27 @@ license: apache-2.0
 language:
   - en
   - hi
 library_name: llama.cpp
 tags:
   - gguf
   - llama.cpp
-  - qwen
   - quantized
   - q4_k_m
-base_model: Qwen/Qwen2.5-1.5B-Instruct
 ---
-# Qwen2.5-1.5B-Instruct Q4_K_M GGUF
-This repository contains a GGUF quantized version of **Qwen2.5-1.5B-Instruct** optimized for local inference with llama.cpp and compatible applications.
 ## Model Details
-* Base Model: Qwen2.5-1.5B-Instruct
 * Format: GGUF
 * Quantization: Q4_K_M
 * Intended Use: Local inference, experimentation, educational projects, and edge deployment
@@ -30,13 +33,13 @@ This repository contains a GGUF quantized version of **Qwen2.5-1.5B-Instruct** o
 The model was converted from the original Hugging Face checkpoint to GGUF format and quantized using llama.cpp's `Q4_K_M` quantization scheme.
-Q4_K_M provides a strong balance between:
 * Model quality
 * Memory usage
 * Inference speed
-making it suitable for consumer hardware and mobile devices.
 ## Files
@@ -52,35 +55,17 @@ making it suitable for consumer hardware and mobile devices.
   -p "Explain machine learning in simple terms."
 ```
-## Usage with Ollama
-Create a Modelfile:
-```text
-FROM ./qwen-q4_k_m.gguf
-```
-Then:
-```bash
-ollama create qwen-local -f Modelfile
-ollama run qwen-local
-```
 ## Hardware Requirements
-Approximate requirements:
 * Disk Space: ~1–2 GB
-* RAM: 2–4 GB minimum
-* Recommended RAM: 4 GB+
-Actual requirements depend on context length and runtime settings.
 ## Base Model
-This repository contains a quantized derivative of the original Qwen model. Please refer to the original model repository for training details, benchmarks, limitations, and licensing information.
 ## Disclaimer
-This repository only provides a quantized GGUF conversion of the original model. All credit for model training and development belongs to the original Qwen team.

 language:
   - en
   - hi
 library_name: llama.cpp
 tags:
   - gguf
   - llama.cpp
+  - qwen3
   - quantized
   - q4_k_m
+base_model: Qwen/Qwen3.5-2B
 ---
+# Qwen3.5-2B Q4_K_M GGUF
+This repository contains a GGUF quantized version of **Qwen3.5-2B** optimized for local inference with llama.cpp and compatible applications.
 ## Model Details
+* Base Model: Qwen3.5-2B
 * Format: GGUF
 * Quantization: Q4_K_M
 * Intended Use: Local inference, experimentation, educational projects, and edge deployment
 The model was converted from the original Hugging Face checkpoint to GGUF format and quantized using llama.cpp's `Q4_K_M` quantization scheme.
+Q4_K_M provides an excellent balance between:
 * Model quality
 * Memory usage
 * Inference speed
+making it suitable for laptops, desktops, and mobile devices.
 ## Files
   -p "Explain machine learning in simple terms."
 ```
 ## Hardware Requirements
 * Disk Space: ~1–2 GB
+* RAM: 3–6 GB recommended
 ## Base Model
+This repository contains a quantized derivative of the original Qwen3.5-2B model.
+Please refer to the original model repository for training details, benchmarks, limitations, and licensing information.
 ## Disclaimer
+This repository only provides a GGUF quantization of the original model. All credit for model training and development belongs to the Qwen team.