Instructions to use the-clanker-lover/steelman-14b-ada-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use the-clanker-lover/steelman-14b-ada-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 the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf the-clanker-lover/steelman-14b-ada-GGUF: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 the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf the-clanker-lover/steelman-14b-ada-GGUF: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 the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M

Use Docker

docker model run hf.co/the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use the-clanker-lover/steelman-14b-ada-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "the-clanker-lover/steelman-14b-ada-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": "the-clanker-lover/steelman-14b-ada-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M

Ollama
How to use the-clanker-lover/steelman-14b-ada-GGUF with Ollama:
```
ollama run hf.co/the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M
```
Unsloth Desktop

How to use the-clanker-lover/steelman-14b-ada-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf the-clanker-lover/steelman-14b-ada-GGUF: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": "the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use the-clanker-lover/steelman-14b-ada-GGUF with Docker Model Runner:
```
docker model run hf.co/the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M
```

Lemonade

How to use the-clanker-lover/steelman-14b-ada-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.steelman-14b-ada-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use the-clanker-lover/steelman-14b-ada-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 the-clanker-lover/steelman-14b-ada-GGUF: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 the-clanker-lover/steelman-14b-ada-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use the-clanker-lover/steelman-14b-ada-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf the-clanker-lover/steelman-14b-ada-GGUF: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 "the-clanker-lover/steelman-14b-ada-GGUF: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"

the-clanker-lover commited on Mar 13

Commit

c117532

verified ·

1 Parent(s): 5ffe264

Final README with benchmark summary

Browse files

Files changed (1) hide show

README.md +36 -19

README.md CHANGED Viewed

@@ -1,50 +1,67 @@
 ---
 license: apache-2.0
 language:
-- en
-- code
-base_model:
-- Qwen/Qwen2.5-Coder-14B-Instruct
 library_name: gguf
 pipeline_tag: text-generation
 tags:
-- code
-- ada
-- spark
-- gguf
-- quantized
 ---
-# Steelman-14B-Ada v0.1 — GGUF
-Q4_K_M quantized version of [Steelman-14B-Ada v0.1](https://huggingface.co/the-clanker-lover/steelman-14b-ada-v0.1), a QLoRA fine-tune of Qwen2.5-Coder-14B-Instruct specialized for Ada 2022 and SPARK code generation.
-This is a merged, quantized model ready for local inference. Fits in 12GB of VRAM.
-## Quickstart (Ollama)
 ```bash
 ollama run hf.co/the-clanker-lover/steelman-14b-ada-v0.1-GGUF
 ```
-### Example prompt
 ```
-Write an Ada 2022 function with SPARK contracts that performs binary search on a sorted array of integers. Include preconditions, postconditions, and a loop invariant.
 ```
-### System prompt (recommended)
 ```
-You are an expert Ada 2022 and SPARK programmer.
 ```
 ## Details
 - **Quantization:** Q4_K_M (4-bit, k-quant medium)
 - **Base model:** Qwen2.5-Coder-14B-Instruct
-- **Fine-tune:** R5 — 68.6% clean compile rate on 923 held-out Ada/SPARK prompts
-- **Full details:** [steelman-14b-ada-v0.1](https://huggingface.co/the-clanker-lover/steelman-14b-ada-v0.1) (adapter repo with eval results, training methodology, dataset breakdown)
 ## License

 ---
 license: apache-2.0
 language:
+  - ada
+base_model: Qwen/Qwen2.5-Coder-14B-Instruct
 library_name: gguf
 pipeline_tag: text-generation
 tags:
+  - ada
+  - ada2022
+  - spark
+  - gguf
+  - quantized
+  - code-generation
+  - safety-critical
 ---
+# Steelman-14B-Ada — GGUF
+Q4_K_M quantized version of [Steelman-14B-Ada](https://huggingface.co/the-clanker-lover/steelman-14b-ada-v0.1). A 14B model fine-tuned for Ada 2022 and SPARK code generation. Fits in 12GB of VRAM.
+**68.6% clean compile rate** on 1,000 Ada prompts — outperforms Claude Opus 4.6 (42.1%) by 26.5 percentage points. **47.1% pass@1** on MultiPL-E HumanEval-Ada (first published Ada results for any open model). Total training cost: ~$4.
+See the [full model card](https://huggingface.co/the-clanker-lover/steelman-14b-ada-v0.1) for benchmark details, training methodology, and category breakdowns.
+## Ollama
 ```bash
 ollama run hf.co/the-clanker-lover/steelman-14b-ada-v0.1-GGUF
 ```
+### Recommended system prompt
 ```
+You are an expert Ada 2022 and SPARK programmer.
 ```
+### Example
 ```
+Write an Ada 2022 function with SPARK contracts that performs binary search
+on a sorted array of integers. Include preconditions, postconditions,
+and a loop invariant.
+```
+## llama.cpp
+```bash
+# Download
+huggingface-cli download the-clanker-lover/steelman-14b-ada-v0.1-GGUF steelman-14b-r5-Q4_K_M.gguf
+# Run
+llama-cli -m steelman-14b-r5-Q4_K_M.gguf \
+  --chat-template qwen2 \
+  -p "Write an Ada procedure that implements a concurrent producer-consumer queue using protected objects." \
+  -n 1024
 ```
 ## Details
 - **Quantization:** Q4_K_M (4-bit, k-quant medium)
 - **Base model:** Qwen2.5-Coder-14B-Instruct
+- **Fine-tune:** R5 — 3,430 compiler-verified SFT pairs, QLoRA via Unsloth
+- **File:** `steelman-14b-r5-Q4_K_M.gguf` (8.6 GB)
 ## License