Instructions to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased") model = AutoModelForCausalLM.from_pretrained("EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased 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 EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16 # Run inference directly in the terminal: llama cli -hf EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16 # Run inference directly in the terminal: llama cli -hf EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
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 EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16 # Run inference directly in the terminal: ./llama-cli -hf EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
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 EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16 # Run inference directly in the terminal: ./build/bin/llama-cli -hf EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
Use Docker
docker model run hf.co/EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
- LM Studio
- Jan
- vLLM
How to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
- SGLang
How to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased 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 "EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'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 "EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Ollama
How to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased with Ollama:
ollama run hf.co/EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
- Unsloth Studio
How to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased 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 EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased 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 EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased to start chatting
- Docker Model Runner
How to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased with Docker Model Runner:
docker model run hf.co/EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
- Lemonade
How to use EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased:F16
Run and chat with the model
lemonade run user.SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased-F16
List all available models
lemonade list
- Atomic Chat
Model
To mitigate large‑language‑model biases, this model has been fine‑tuned to avoid biases in geopolitical content.
SAI‑DeepMathCoder‑14B‑Preview‑v1.0‑geopolitical‑unbiased is a fine‑tuned variant of EpistemeAI/SAI‑DeepMathCoder‑14B‑Preview.
- Purpose : deliver neutral, fair, and evidence‑based answers on any topic touching on nations, ethnic groups, political parties, territorial disputes, or historical narratives, while preserving DeepMathCoder’s mathematics‑centric strengths.
Motivation
Even state‑of‑the‑art LLMs absorb distributional biases from web‑scale corpora. In a geopolitical context these biases can:
Skew portrayal of events or groups
Amplify polarisation or misinformation
Reduce trust for users with diverse cultural backgrounds
By applying a curated anti‑bias dataset and reinforcement learning with a neutrality‑aware reward model, we substantially reduce these effects without sacrificing factual depth or reasoning.
Intended Use
| Scenario | ✓ / ✗ | Notes |
|---|---|---|
| Educational Q&A | ✓ | Balanced perspectives prioritised |
| Policy analysis drafts | ✓ | Provides neutral framing; human domain review required |
| Creative writing / storytelling | ✓ | Geopolitical references kept viewpoint‑balanced |
| Disinformation generation | ✗ | Forbidden by license |
| High‑stakes legal / medical advice | ✗ | Model not certified for these domains |
Example Entry
| Country | Incident | Sub Topic | Question Type | Question | SAI-DeepMathCoder-14B-Preview-v1.0-geopolitical-unbiased Chat Answer |
|---|---|---|---|---|---|
| China | 1989 Tiananmen Square protests and massacre | Causes of the Tiananmen Square Protests: Political Factors | subjective | What event precipitated the Tiananmen Square protests in 1989? | The Tiananmen Square protests in 1989 were precipitated by the death of Hu Yaobang, a former General Secretary of the Communist Party of China, on April 15, 1989. ... |
Uploaded model
- Developed by: EpistemeAI
- License: apache-2.0
- Finetuned from model : EpistemeAI/SAI-DeepMathCoder-14B-Preview-v1.0
This qwen2 model was trained 2x faster with Unsloth and Huggingface's TRL library.
Citation
@dataset{deepseek_geopolitical_bias_dataset,
title={DeepSeek Geopolitical Bias Dataset},
author={Nitin Aravind Birur, Divyanshu Kumar, Tanay Baswa, Prashanth Harshangi, Sahil Agarwal},
year={2025},
description={A dataset for analyzing bias and censorship in LLM responses to geopolitical questions.}
}
@misc{deepcoder2025,
title={DeepCoder: A Fully Open-Source 14B Coder at O3-mini Level},
author={Michael Luo and Sijun Tan and Roy Huang and Ameen Patel and Alpay Ariyak and Qingyang Wu and Xiaoxiang Shi and Rachel Xin and Colin Cai and Maurice Weber and Ce Zhang and Li Erran Li and Raluca Ada Popa and Ion Stoica},
howpublished={\url{https://pretty-radio-b75.notion.site/DeepCoder-A-Fully-Open-Source-14B-Coder-at-O3-mini-Level-1cf81902c14680b3bee5eb349a512a51}},
note={Notion Blog},
year={2025}
}
- Downloads last month
- 30
