Pinned Loading

1. Llama 3.2 inference in mlx - functio... Llama 3.2 inference in mlx - functional style / no nn.Module

   1

   # Uses entropix's tokenizer & downloaded weights. Saving this file into entropix's repo should work

   2

   3

   import math

   4

   from typing import NamedTuple, List

   5

   from pathlib import Path

2. mnist.c mnist.c Public

   MNIST classifier training in simple, raw C/CUDA. Inspired by llm.c

   C

3. cuda-matmuls cuda-matmuls Public

   cuda matmul experimentations

   Cuda

4. minimal objc python bridge minimal objc python bridge

   1

   from typing import Dict, Tuple, Union, Any

   2

   import functools, ctypes, ctypes.util

   3

   import libobjc  # clang2py generated bindings to objc/objc.h and objc/runtime.h

   4

   # note: The Objective-C runtime does not expose enough information to provide completely automatic bindings of all APIs. source: https://pyobjc.readthedocs.io/en/latest/metadata/index.html

   5

   # though it's possible to parse bridgesupport files

5. torch_to_ggml torch_to_ggml Public archive

   convert a saved pytorch model to gguf and generate as much corresponding ggml c code as possible

   Python 13 5

6. bloom-filter bloom-filter Public

   A simple bloom filter implementation in java

   Python