Instructions to use strykes/tiny-giant-2500-q4_k_m with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use strykes/tiny-giant-2500-q4_k_m with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="strykes/tiny-giant-2500-q4_k_m",
	filename="tiny-giant-2500-Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use strykes/tiny-giant-2500-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 strykes/tiny-giant-2500-q4_k_m:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf strykes/tiny-giant-2500-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 strykes/tiny-giant-2500-q4_k_m:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf strykes/tiny-giant-2500-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 strykes/tiny-giant-2500-q4_k_m:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf strykes/tiny-giant-2500-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 strykes/tiny-giant-2500-q4_k_m:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf strykes/tiny-giant-2500-q4_k_m:Q4_K_M

Use Docker

docker model run hf.co/strykes/tiny-giant-2500-q4_k_m:Q4_K_M

LM Studio
Jan

vLLM

How to use strykes/tiny-giant-2500-q4_k_m with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "strykes/tiny-giant-2500-q4_k_m"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "strykes/tiny-giant-2500-q4_k_m",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/strykes/tiny-giant-2500-q4_k_m:Q4_K_M

Ollama
How to use strykes/tiny-giant-2500-q4_k_m with Ollama:
```
ollama run hf.co/strykes/tiny-giant-2500-q4_k_m:Q4_K_M
```

Unsloth Studio

How to use strykes/tiny-giant-2500-q4_k_m 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 strykes/tiny-giant-2500-q4_k_m 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 strykes/tiny-giant-2500-q4_k_m to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for strykes/tiny-giant-2500-q4_k_m to start chatting

How to use strykes/tiny-giant-2500-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 strykes/tiny-giant-2500-q4_k_m: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": "strykes/tiny-giant-2500-q4_k_m:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use strykes/tiny-giant-2500-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 strykes/tiny-giant-2500-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 strykes/tiny-giant-2500-q4_k_m:Q4_K_M

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use strykes/tiny-giant-2500-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 strykes/tiny-giant-2500-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 "strykes/tiny-giant-2500-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"

Docker Model Runner
How to use strykes/tiny-giant-2500-q4_k_m with Docker Model Runner:
```
docker model run hf.co/strykes/tiny-giant-2500-q4_k_m:Q4_K_M
```

Lemonade

How to use strykes/tiny-giant-2500-q4_k_m with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull strykes/tiny-giant-2500-q4_k_m:Q4_K_M

Run and chat with the model

lemonade run user.tiny-giant-2500-q4_k_m-Q4_K_M

List all available models

lemonade list

Tiny-Giant 2500 — Q4_K_M (pilot)

Full fine-tune of Qwen2.5-Coder-1.5B-Instruct on the Tiny-Giant pilot dataset (~2,508 execution-verified agentic coding samples), quantized to Q4_K_M for llama.cpp deployment.

This is the pilot (TAG=2500) checkpoint — it validates the train → quant → eval pipeline, not the final production model (target ≥30k samples).

Model file

File	Quant	Size
`tiny-giant-2500-Q4_K_M.gguf`	Q4_K_M	~940 MB

Training summary


Base	`Qwen/Qwen2.5-Coder-1.5B-Instruct`
Method	Full fine-tune (bf16), not LoRA
Dataset	Tiny-Giant pilot — 2,508 records → train=2383 / val=125
Epochs	3
Learning rate	7e-6
Seq length	4096
Final train loss	0.5891
Final eval loss	0.5302 (epoch 3)
Chat format	Hermes-style ChatML with `<tool_call>` blocks (custom renderer — see repo)

Usage (llama.cpp)

Pin the ChatML template explicitly. This model was trained with a custom Hermes/ChatML renderer (train_tiny_giant.render_conversation); do not rely on template auto-detection.

# llama-server
llama-server -m tiny-giant-2500-Q4_K_M.gguf --chat-template chatml --ctx-size 4096

# llama-cpp-python server
python -m llama_cpp.server --model tiny-giant-2500-Q4_K_M.gguf \
  --chat_format chatml --n_ctx 4096 --port 8088

from llama_cpp import Llama

llm = Llama(model_path="tiny-giant-2500-Q4_K_M.gguf", n_ctx=4096, n_gpu_layers=-1)
out = llm.create_completion(
    prompt="<|im_start|>user\nWrite a Python function to merge two sorted lists.\n<|im_start|>assistant\n",
    max_tokens=512,
    temperature=0.0,
    stop=["<|im_start|>"],
)
print(out["choices"][0]["text"])

Evaluation

Benchmark and internal eval results are populated after the Vast GPU eval run completes. See the companion comparison card in the source repository (MODEL_CARD-2500.md).

Benchmark	Score (Q4_K_M)	Notes
HumanEval pass@1	pending	greedy, EvalPlus
HumanEval+ pass@1	pending	greedy, EvalPlus
MBPP pass@1	pending	greedy, EvalPlus
MBPP+ pass@1	pending	greedy, EvalPlus
Tiny-Giant debug pass@1	pending	held-out execution tests
Tool first-action validity	pending	agentic val set

Limitations

Pilot scale (2.5k samples): format adherence and debugging habits should improve vs base; absolute codegen benchmarks may move little until the full 30k run.
Q4_K_M quantization: if eval shows >~4pt drop vs bf16, re-quantize the kept f16 GGUF to Q5_K_M — no retraining needed.
Tool-use eval requires prompts built with the same renderer used in training.

Source

Training pipeline and dataset factory: Tiny-Giant (local repo).

License

Apache-2.0 (inherits from Qwen2.5-Coder-1.5B-Instruct base).

Downloads last month: 14

GGUF

Model size

2B params

Architecture

qwen2

Hardware compatibility

4-bit

Model tree for strykes/tiny-giant-2500-q4_k_m

Base model

Qwen/Qwen2.5-1.5B

Finetuned

Qwen/Qwen2.5-Coder-1.5B

Finetuned

Qwen/Qwen2.5-Coder-1.5B-Instruct

Quantized

(140)

this model