Instructions to use masahiroid/pyrite-stream with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use masahiroid/pyrite-stream with MLX:
# Make sure mlx-lm is installed # pip install --upgrade mlx-lm # if on a CUDA device, also pip install mlx[cuda] # Generate text with mlx-lm from mlx_lm import load, generate model, tokenizer = load("masahiroid/pyrite-stream") prompt = "Once upon a time in" text = generate(model, tokenizer, prompt=prompt, verbose=True) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- MLX LM
How to use masahiroid/pyrite-stream with MLX LM:
Generate or start a chat session
# Install MLX LM uv tool install mlx-lm # Generate some text mlx_lm.generate --model "masahiroid/pyrite-stream" --prompt "Once upon a time"
- Atomic Chat
Download __init__.py from masahiroid/pyrite-stream: direct link, hf CLI and curl.
- Browser
- Download file 536 Bytes
-
https://huggingface.co/masahiroid/pyrite-stream/resolve/main/__init__.py
- Command line
-
hf download hf://masahiroid/pyrite-stream/__init__.py
-
curl -L -o __init__.py https://huggingface.co/masahiroid/pyrite-stream/resolve/main/__init__.py
536 Bytes
| """Project Pyrite Phase 2: MoE-Stream Engine. | |
| Decoupling Resident Attention and Dynamic NVMe-Streamed Experts | |
| for 100B+ MoEs on Unified Memory Consumer Hardware. | |
| Author: Masahiro NAKATSUGAWA (中津川 昌弘) | |
| Affiliations: n.p.minerva & Co. / AntenneFrance | |
| """ | |
| from .expert_registry import SafetensorsIndex, ModelSafetensorsIndex | |
| from .moe_stream_layer import StreamingMoELayer, ExpertLRUCache | |
| __version__ = "0.2.0" | |
| __all__ = [ | |
| "SafetensorsIndex", | |
| "ModelSafetensorsIndex", | |
| "StreamingMoELayer", | |
| "ExpertLRUCache", | |
| ] | |