Instructions to use KiteFishAI/Minnow-Math-2B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KiteFishAI/Minnow-Math-2B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="KiteFishAI/Minnow-Math-2B")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("KiteFishAI/Minnow-Math-2B")
model = AutoModelForCausalLM.from_pretrained("KiteFishAI/Minnow-Math-2B")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use KiteFishAI/Minnow-Math-2B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KiteFishAI/Minnow-Math-2B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KiteFishAI/Minnow-Math-2B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/KiteFishAI/Minnow-Math-2B

SGLang

How to use KiteFishAI/Minnow-Math-2B with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "KiteFishAI/Minnow-Math-2B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KiteFishAI/Minnow-Math-2B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "KiteFishAI/Minnow-Math-2B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KiteFishAI/Minnow-Math-2B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use KiteFishAI/Minnow-Math-2B with Docker Model Runner:
```
docker model run hf.co/KiteFishAI/Minnow-Math-2B
```

Minnow-Math-2B / README.md

anuj0456

Update README.md

f9909b3 verified 3 months ago

preview code

raw

history blame contribute delete

1.26 kB

	---
	language:
	- en
	license: apache-2.0
	tags:
	- math
	- reasoning
	- mathematics
	- causal-lm
	- text-generation
	library_name: transformers
	pipeline_tag: text-generation
	model_name: Minnow-Math-2B
	---

	# 🐟 Minnow-Math-2B

	Minnow-Math-2B is a 2B-parameter language model by Kitefish, focused on mathematical reasoning, symbolic understanding, and structured problem solving.

	This is an early release and part of our ongoing effort to build strong, efficient models for reasoning-heavy tasks.

	---

	## ✨ What this model is good at

	- Basic to intermediate math problem solving
	- Step-by-step reasoning for equations and word problems
	- Understanding mathematical symbols and structure
	- Educational and experimentation use cases

	---

	## 🚀 Quick start

	```python
	from transformers import AutoTokenizer, AutoModelForCausalLM

	tokenizer = AutoTokenizer.from_pretrained("kitefish/Minnow-Math-2B")
	model = AutoModelForCausalLM.from_pretrained(
	"kitefish/Minnow-Math-2B",
	torch_dtype="auto",
	device_map="auto"
	)

	prompt = "Solve: 2x + 5 = 13"
	inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

	outputs = model.generate(**inputs, max_new_tokens=100)
	print(tokenizer.decode(outputs[0], skip_special_tokens=True))