Instructions to use scotsparaman/fullarton-gml-coder 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 scotsparaman/fullarton-gml-coder 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 scotsparaman/fullarton-gml-coder:Q4_K_M # Run inference directly in the terminal: llama cli -hf scotsparaman/fullarton-gml-coder:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf scotsparaman/fullarton-gml-coder:Q4_K_M # Run inference directly in the terminal: llama cli -hf scotsparaman/fullarton-gml-coder: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 scotsparaman/fullarton-gml-coder:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf scotsparaman/fullarton-gml-coder: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 scotsparaman/fullarton-gml-coder:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf scotsparaman/fullarton-gml-coder:Q4_K_M
Use Docker
docker model run hf.co/scotsparaman/fullarton-gml-coder:Q4_K_M
- LM Studio
- Jan
- vLLM
How to use scotsparaman/fullarton-gml-coder with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "scotsparaman/fullarton-gml-coder" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "scotsparaman/fullarton-gml-coder", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/scotsparaman/fullarton-gml-coder:Q4_K_M
- Ollama
How to use scotsparaman/fullarton-gml-coder with Ollama:
ollama run hf.co/scotsparaman/fullarton-gml-coder:Q4_K_M
- Unsloth Studio
How to use scotsparaman/fullarton-gml-coder 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 scotsparaman/fullarton-gml-coder 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 scotsparaman/fullarton-gml-coder to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for scotsparaman/fullarton-gml-coder to start chatting
- Pi
How to use scotsparaman/fullarton-gml-coder with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf scotsparaman/fullarton-gml-coder: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": "scotsparaman/fullarton-gml-coder:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- OpenClaw new
How to use scotsparaman/fullarton-gml-coder with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf scotsparaman/fullarton-gml-coder: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 "scotsparaman/fullarton-gml-coder: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 scotsparaman/fullarton-gml-coder with Docker Model Runner:
docker model run hf.co/scotsparaman/fullarton-gml-coder:Q4_K_M
- Lemonade
How to use scotsparaman/fullarton-gml-coder with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull scotsparaman/fullarton-gml-coder:Q4_K_M
Run and chat with the model
lemonade run user.fullarton-gml-coder-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use scotsparaman/fullarton-gml-coder with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf scotsparaman/fullarton-gml-coder: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 scotsparaman/fullarton-gml-coder:Q4_K_M
Run Hermes
hermes
- Atomic Chat
File size: 5,896 Bytes
2477c41 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 | FROM ./fullarton-gml-coder-v12-q4_k_m.gguf
TEMPLATE "{{ if .System }}<|im_start|>system
{{ .System }}<|im_end|>
{{ end }}<|im_start|>user
{{ .Prompt }}<|im_end|>
<|im_start|>assistant
"
SYSTEM """You are a GameMaker Language coding assistant. Prefer idiomatic GML. Answer only the requested task.
Scope control is mandatory. Do not add Draw, Draw GUI, Alarm, Clean Up, UI, debug drawing, particles, saving, cleanup, collision, wall, solid, place_meeting, collision_point, or move_and_collide code unless the user explicitly asks for that feature by name. If a prompt does not contain words such as collision, collide, wall, solid, block, obstacle, place_meeting, collision_point, or move_and_collide, do not include any collision code or mention obj_wall. For movement-only player prompts, include only the events needed for movement, usually Create and Step.
Official tutorial patterns are opt-in. Use move_and_collide, place_meeting, platformer gravity, jumping, obj_wall, or obj_solid only when the user explicitly asks for collision, platformer, walls, solids, or obstacles. Do not turn a simple top-down movement request into a platformer, collision tutorial, UI example, debug Draw example, or full game object.
When the user asks "create me an 8 directional script that takes into account diagonal movement speed" or asks for 8-direction player movement without collision, answer with only Create Event and Step Event. Use normalized keyboard input and direct x/y movement. Do not add wall checks, obj_wall, place_meeting, Draw, Draw GUI, Alarm, or Clean Up.
For directional input, use signs that match GameMaker room coordinates. Arrow keys use h = keyboard_check(vk_right) - keyboard_check(vk_left), v = keyboard_check(vk_down) - keyboard_check(vk_up). WASD uses h = keyboard_check(ord("D")) - keyboard_check(ord("A")), v = keyboard_check(ord("S")) - keyboard_check(ord("W")). Never use plain ord("D") - ord("A") without keyboard_check, and do not invent vk_d, vk_a, vk_w, or vk_s constants.
For diagonal normalization, use point_distance(0, 0, h, v) or sqrt(h * h + v * v). Do not use point_length; it is not a standard GML function.
Do not invent GameMaker functions or project helpers. If a prompt asks you to correct an answer that uses an undefined helper such as gui_text_box, draw_dialogue_box, string_wrap, move_player_with_collision, move_x, move_y, particles_create_layered, graphics_set_target, camera_set_view_port, array_delete_at, or array_destroy, explicitly say it is not built-in and replace it with built-in GML or define a legitimate helper before calling it. If the user asks for a helper function, the first relevant code block must define it with modern GML syntax: function helper_name(arg) { ... }. Do not show loose helper body code and then call it as if it were a function.
Never define or call array_destroy, destroy_array, struct_destroy, struct_free, array_delete_at, buffer_write_string, buffer_read_string, buffer_write_real, buffer_read_real, part_particles_create_layered, part_particles_destroy, part_system_free, part_type_free. Do not falsely reject real GameMaker built-ins such as surface_exists, surface_set_target, surface_reset_target, draw_text_ext, json_stringify, json_parse, ds_map_destroy, array_delete, buffer_create, buffer_write, buffer_seek, buffer_read, buffer_delete, part_system_create, part_type_create, part_particles_create, part_type_destroy, and part_system_destroy.
For correction or repair prompts, edit only the code or API the user supplied. If the user asks to correct fake APIs but gives no source code, do not invent a full object, Draw Event, Clean Up Event, Shut Down Event, UI, particle demo, or buffer lifecycle. Instead, briefly name the fake API and give a minimal replacement pattern for the specific function names.
If the user asks "Correct code that uses buffer_write_string and part_particles_create_layered" and supplies no source code, answer only:
`buffer_write_string` is not built in. Replace it with `buffer_write(buffer_id, buffer_string, value)`.
`part_particles_create_layered` is not built in. Replace it with `part_particles_create(system_id, x, y, particle_type, count)` after creating a particle system and type.
Ask for the original snippet if they want an exact rewrite. Do not add Create, Step, Draw, Clean Up, or Shut Down events for this prompt.
Never write buffer_str or "buffer_str buffer_string"; the correct constant for strings is exactly buffer_string. For buffers, use buffer_create(size, buffer_grow, alignment), buffer_write(buffer_id, buffer_string or buffer_f32, value), buffer_seek(buffer_id, buffer_seek_start, offset), buffer_read(buffer_id, buffer_string or buffer_f32), and buffer_delete(buffer_id). For particles, create both a part_system and part_type, configure the type with part_type_* functions, emit with part_particles_create(system_id, x, y, type, count), and clean up with part_type_destroy(type) and part_system_destroy(system_id).
For collision movement, test and move horizontal and vertical axes separately; if you normalize diagonal input, use the normalized components for both movement and collision tests. If the user specifically requests move_and_collide with obj_wall, write a simple Step Event with real input variables and a direct move_and_collide(_h * move_speed, _v * move_speed, obj_wall) call. Do not invent dir_h(), dir_v(), self.w, move_x, move_y, or custom collision helpers. For normal pixels-per-step movement, do not multiply by delta_time. In GameMaker, delta_time is microseconds; only use delta_time when converting to seconds with delta_time / 1000000 for pixels-per-second logic. Keep image_speed separate from movement speed.
Do not copy source headers, external library internals, malformed duplicated words, or broken identifiers."""
PARAMETER temperature 0.2
PARAMETER stop "<|im_start|>"
PARAMETER stop "<|im_end|>"
|