Instructions to use shibatch/tinymoe2m with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use shibatch/tinymoe2m with Transformers:

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("shibatch/tinymoe2m", dtype="auto")

llama-cpp-python

How to use shibatch/tinymoe2m with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="shibatch/tinymoe2m",
	filename="tinymoe2m.BF16.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use shibatch/tinymoe2m with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf shibatch/tinymoe2m:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf shibatch/tinymoe2m:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf shibatch/tinymoe2m:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf shibatch/tinymoe2m: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 shibatch/tinymoe2m:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf shibatch/tinymoe2m: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 shibatch/tinymoe2m:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf shibatch/tinymoe2m:Q4_K_M

Use Docker

docker model run hf.co/shibatch/tinymoe2m:Q4_K_M

LM Studio
Jan
Ollama
How to use shibatch/tinymoe2m with Ollama:
```
ollama run hf.co/shibatch/tinymoe2m:Q4_K_M
```

Unsloth Studio

How to use shibatch/tinymoe2m 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 shibatch/tinymoe2m 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 shibatch/tinymoe2m to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for shibatch/tinymoe2m to start chatting

Docker Model Runner
How to use shibatch/tinymoe2m with Docker Model Runner:
```
docker model run hf.co/shibatch/tinymoe2m:Q4_K_M
```

Lemonade

How to use shibatch/tinymoe2m with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull shibatch/tinymoe2m:Q4_K_M

Run and chat with the model

lemonade run user.tinymoe2m-Q4_K_M

List all available models

lemonade list

shibatch commited on about 1 hour ago

Commit

6ec8a17

verified ·

1 Parent(s): ad0e5ec

Upload README.md with huggingface_hub

Browse files

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -17,7 +17,7 @@ This repository provides an ultra-lightweight Mixtral model variant (a Mixture-o
 Following extensive long-context scaling evaluations, this asset has been calibrated to a **4,096 token context window (4k)** with an adjusted **RoPE base frequency (`rope_theta`) of 15,000.0** to prevent numerical saturation under FP32 precision boundaries while maintaining sharp localized attention coordinates.
-It is designed specifically for debugging custom inference engines (such as `vulformer`), and native tensor compilers against MoE-specific runtime features. These include Gating network weight allocation, token distribution/gathering (Scatter/Gather loops), and the weighted addition combining multiple independent expert outputs.
 ---

 Following extensive long-context scaling evaluations, this asset has been calibrated to a **4,096 token context window (4k)** with an adjusted **RoPE base frequency (`rope_theta`) of 15,000.0** to prevent numerical saturation under FP32 precision boundaries while maintaining sharp localized attention coordinates.
+It is designed specifically for debugging custom inference engines, and native tensor compilers against MoE-specific runtime features. These include Gating network weight allocation, token distribution/gathering (Scatter/Gather loops), and the weighted addition combining multiple independent expert outputs.
 ---