The interactive file manager requires Javascript. Please enable it or use sftp or scp.
You may still browse the files here.

Download Latest Version coremltools-8.3.0.tar.gz (1.6 MB)
Email in envelope

Get an email when there's a new version of Core ML Tools

Home / 8.2
Name Modified Size InfoDownloads / Week
Parent folder
coremltools-8.2.tar.gz 2025-01-22 1.5 MB
0
coremltools-8.2-cp312-none-manylinux1_x86_64.whl 2025-01-22 2.2 MB
0
coremltools-8.2-cp312-none-macosx_11_0_arm64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp312-none-macosx_10_15_x86_64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp311-none-manylinux1_x86_64.whl 2025-01-22 2.2 MB
0
coremltools-8.2-cp311-none-macosx_11_0_arm64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp311-none-macosx_10_15_x86_64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp310-none-manylinux1_x86_64.whl 2025-01-22 2.2 MB
0
coremltools-8.2-cp310-none-macosx_11_0_arm64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp310-none-macosx_10_15_x86_64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp39-none-manylinux1_x86_64.whl 2025-01-22 2.2 MB
0
coremltools-8.2-cp39-none-macosx_11_0_arm64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp39-none-macosx_10_15_x86_64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp38-none-manylinux1_x86_64.whl 2025-01-22 2.3 MB
0
coremltools-8.2-cp38-none-macosx_11_0_arm64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp38-none-macosx_10_15_x86_64.whl 2025-01-22 2.6 MB
0
coremltools-8.2-cp37-none-manylinux1_x86_64.whl 2025-01-22 2.3 MB
0
coremltools-8.2-cp37-none-macosx_10_15_x86_64.whl 2025-01-22 2.6 MB
0
README.md 2025-01-21 833 Bytes
0
Release Notes source code.tar.gz 2025-01-21 27.2 MB
0
Release Notes source code.zip 2025-01-21 30.4 MB
0
Totals: 21 Items   101.3 MB 0
  • Increase conversion support coverage for models produced by torch.export
    • All common ops have been covered, that is 83% parity with our mature torch.jit.trace converter
    • A new model supported: torchaudio wav2vec
  • A new transformers graph pass "common::scaled_dot_product_attention_sliced_q" for improved performance of sdpa calculation with long sequence lengths (e.g. Depth-Anything model, with sequence length 1814, transformed by the graph pass was 34% faster and used 45% less memory on ANE)
  • New torch ops supported
    • native_group_norm
    • bool type triu
  • Bug fixes
    • Incorrect behaviour of torch.linspace
    • Wrong power 2 - sqrt fusion when power exponent is a tensor

Special thanks to our open source community contributors for this release: @RGooBS24 @fukatani @twoertwein @kasper0406

Source: README.md, updated 2025-01-21