ios_llama8b / meta.yaml
joonhyuk7's picture
Upload folder using huggingface_hub
838f1bb verified
model_info:
name: anemll-Meta-Llama-3-8B-ctx2048
version: 0.3.4
description: |
Demonstarates running Meta-Llama-3-8B on Apple Neural Engine
Context length: 2048
Batch size: 1
Chunks: 4
license: MIT
author: Anemll
framework: Core ML
language: Python
architecture: llama
parameters:
context_length: 2048
batch_size: 1
lut_embeddings: none
lut_ffn: 4
lut_lmhead: 4
num_chunks: 4
model_prefix: llama
embeddings: llama_embeddings.mlmodelc
lm_head: llama_lm_head_lut4.mlmodelc
ffn: llama_FFN_PF_lut4.mlmodelc
split_lm_head: 8