Instructions to use betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10 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 betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10
# Run inference directly in the terminal:
llama cli -hf betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10
# Run inference directly in the terminal:
llama cli -hf betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

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 betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10
# Run inference directly in the terminal:
./llama-cli -hf betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

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 betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10
# Run inference directly in the terminal:
./build/bin/llama-cli -hf betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

Use Docker

docker model run hf.co/betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

LM Studio
Jan
Ollama
How to use betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10 with Ollama:
```
ollama run hf.co/betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10
```
Unsloth Desktop

How to use betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

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": "betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10 with Docker Model Runner:
```
docker model run hf.co/betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10
```

Lemonade

How to use betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

Run and chat with the model

lemonade run user.ralph-v2-qwen3-8b-ternary-af-892aaa10-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

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 betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

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 "betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10" \
  --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"

ralph-v2-qwen3-8b-ternary-af-892aaa10

Compressed derivative of Qwen/Qwen3-8B submitted to the ternary bit-tier of Bittensor subnet 40 (Ralph, model compression). Architecture and parameter count are unchanged from the parent; weights are re-stored at reduced bit-width.

License

Released under the Apache License 2.0 — see LICENSE. This is a derivative of Qwen3-8B, itself Apache-2.0; the parent's copyright and attribution notices, a description of what changed, the data used and the third-party tooling involved are preserved in NOTICE. The exact artifact is model.gguf at revision 55060519c6f179ea1d28ee9e6c333f33bfc03bd3 (SHA-256 e299a0b16fa1240ce73bcc1d7d73be60c7cab0fb5c951398678cc059560210dc); adding this model card and license files does not alter that file.

Model overview


Parent	Qwen/Qwen3-8B (Apache-2.0)
Architecture	unchanged from parent
Format	GGUF (llama.cpp)
Bit tier	ternary (≤2.5 bits/weight)
File size	2,513,979,008 bytes
Measured bits/weight	~2.43
Method	Ternary quantization-aware distillation of the parent: a fixed ternary weight structure (per-block scales, 8-bit embedding/output tensors) with an fp32 master trained by straight-through estimation to minimise KL divergence to Qwen3-8B on continuation prefixes up to 4,096 tokens (about 2,000 steps on 8×H200, plus two short anti-repetition cool-downs). Architecture and parameter count unchanged (8,190,427,136 weight-bearing parameters).

Upstream terms to preserve: Qwen3-8B's Apache-2.0 license and attribution (kept in NOTICE); llama.cpp's MIT license for the GGUF tooling; the dataset licenses listed in NOTICE (Apache-2.0, MIT, CC-BY-4.0 — attribution only).

Downloads last month: 159

GGUF

Model size

8B params

Architecture

qwen3

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for betaprogramming/ralph-v2-qwen3-8b-ternary-af-892aaa10

Base model

Qwen/Qwen3-8B-Base

Finetuned

Qwen/Qwen3-8B

Quantized

(462)

this model