Instructions to use MethodWhite/materia-v1-1b-bpe with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use MethodWhite/materia-v1-1b-bpe 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 MethodWhite/materia-v1-1b-bpe:Q4_0
# Run inference directly in the terminal:
llama cli -hf MethodWhite/materia-v1-1b-bpe:Q4_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf MethodWhite/materia-v1-1b-bpe:Q4_0
# Run inference directly in the terminal:
llama cli -hf MethodWhite/materia-v1-1b-bpe:Q4_0

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 MethodWhite/materia-v1-1b-bpe:Q4_0
# Run inference directly in the terminal:
./llama-cli -hf MethodWhite/materia-v1-1b-bpe:Q4_0

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 MethodWhite/materia-v1-1b-bpe:Q4_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf MethodWhite/materia-v1-1b-bpe:Q4_0

Use Docker

docker model run hf.co/MethodWhite/materia-v1-1b-bpe:Q4_0

LM Studio
Jan
Ollama
How to use MethodWhite/materia-v1-1b-bpe with Ollama:
```
ollama run hf.co/MethodWhite/materia-v1-1b-bpe:Q4_0
```

Unsloth Studio

How to use MethodWhite/materia-v1-1b-bpe 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 MethodWhite/materia-v1-1b-bpe 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 MethodWhite/materia-v1-1b-bpe to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for MethodWhite/materia-v1-1b-bpe to start chatting

Atomic Chat new
Docker Model Runner
How to use MethodWhite/materia-v1-1b-bpe with Docker Model Runner:
```
docker model run hf.co/MethodWhite/materia-v1-1b-bpe:Q4_0
```

Lemonade

How to use MethodWhite/materia-v1-1b-bpe with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull MethodWhite/materia-v1-1b-bpe:Q4_0

Run and chat with the model

lemonade run user.materia-v1-1b-bpe-Q4_0

List all available models

lemonade list

M.A.T.E.R.I.A. V1 1B BPE

⚠️ Este repositorio se llamaba anteriormente materia-v4-1b-bpe. La versión "V4" es en realidad la primera versión entrenada (V1) del modelo. Las versiones V1-V3 previas eran prototipos conceptuales no entrenados.

Checkpoint del entrenamiento 1B de M.A.T.E.R.I.A. V1 — arquitectura JEPA-First con HSAQ.

Checkpoints

Archivo original	Descripción
`checkpoint_epoch1.pt`	Epoch 1 — entrenamiento 1B BPE
`checkpoint_epoch2.pt`	Epoch 2 — con HSAQ activo

Modelo base

Archivo	Descripción
`materia-v4.basemateria`	Base model weights (formato .basemateria)
`materia-v4.materia`	Weight module

GGUF (Ollama/Llama.cpp)

Archivo	Quant
`materia-v4-llama-compat.gguf`	FP16 compatible
`materia-v4.Q4_0.gguf`	Q4_0 cuantizado

Arquitectura

JEPA-First con Spectral Coupling Analysis (K = 2.781042)
GQA 8Q:4KV heads + Flash Attention 2
RoPE NTK-aware + SwiGLU
LIF-SNN (16 neuronas) + SSM
HSAQ — máscara de activación dispersa adaptativa

Uso

from huggingface_hub import hf_hub_download
ckpt = hf_hub_download('MethodWhite/materia-v1-1b-bpe', 'checkpoint_epoch2.pt')

Referencias

Downloads last month: 70

GGUF

Model size

1B params

Architecture

llama

Hardware compatibility

4-bit

View +1 variant

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for MethodWhite/materia-v1-1b-bpe

Quantizations

1 model