Open
Conversation
The MIL scalar types used shorthand syntax (string("x"), int32(1)) that
only works on M4. Changed to the canonical verbose format that CoreML's
own compiler emits (tensor<string, []>("x"), tensor<int32, []>(1)).
Also targets program(1.0) with <ios16> instead of program(1.3)/<ios18>,
and simplifies buildInfo to just coremlc-version.
For conv-based kernels, adds runtime fp16 I/O fallback — M1/M2 ANE
doesn't support the cast op (fp32<->fp16), so on first compile failure
it retries with native fp16 inputs/outputs and does the conversion on
the CPU side. The fallback is persisted across exec() restarts.
Note: matmul and scaled_dot_product_attention ops still fail on M1/M2 —
these are M4+ ANE ops. The attention tests (test_ane_causal_attn,
test_ane_sdpa5, test_full_fused attention part) require M4 hardware.
Conv-based kernels (training, QKV projections, FFN) work on all generations.
Tested on M1 Pro, macOS 26.3 (Tahoe).
|
you need to fix the other files too not just the training dir files to make it work on m1/m2. |
Author
|
From what I see should be complete now although I'd like to fix attn too, would be keen to train a nice UNet like this |
codegen-sh bot
referenced
this pull request
in dermitchell1993/ANE
Mar 2, 2026
Port upstream PR #6 (imperatormk) - fixes MIL scalar type syntax from M4-only shorthand to canonical verbose format that compiles on all Apple Silicon (M1/M2/M3/M4). Changes: - program(1.3) to program(1.0), ios18 to ios16 target - Scalar type shorthand to canonical verbose format - Simplified buildInfo dict (no M4-specific version strings) - fp16 I/O fallback: g_fp16_io flag with auto-retry on compile failure for M1/M2 where cast op is unsupported - Dynamic IOSurface byte calculation (bpe: 2 for fp16, 4 for fp32) Tested on M1 Pro, macOS 26.3 (per upstream PR author).
dev-erik
added a commit
to dev-erik/ANE
that referenced
this pull request
Mar 3, 2026
…program(1.0), ios16 target, tensor types across 18 files
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
MIL scalar types used shorthand syntax that only compiles on M4. Changed to the canonical verbose format that CoreML's compiler emits (
tensor<string, []>("x")instead ofstring("x")), targetsprogram(1.0)/<ios16>, and simplifies buildInfo.For conv kernels, adds runtime fp16 I/O fallback — M1/M2 ANE can't do the
castop, so on first compile failure it retries with fp16 inputs/outputs and converts on the CPU side.Tested on M1 Pro, macOS 26.3.