model_info:
  name: anemll-qwen3_0.6b_model_original-ctx1024
  version: 0.3.0
  description: |
    Demonstarates running qwen3_0.6b_model_original on Apple Neural Engine
    Context length: 1024
    Batch size: 64
    Chunks: 1
  license: MIT
  author: Anemll
  framework: Core ML
  language: Python
  parameters:
    context_length: 1024
    batch_size: 64
    lut_embeddings: none
    lut_ffn: none
    lut_lmhead: none
    num_chunks: 1
    model_prefix: qwen0.6_float16
    embeddings: qwen0.6_float16_embeddings.mlmodelc
    lm_head: qwen0.6_float16_lm_head.mlmodelc
    ffn: qwen0.6_float16_FFN_PF.mlmodelc
    split_lm_head: 16