-
Notifications
You must be signed in to change notification settings - Fork 850
Pull requests: apple/coremltools
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Stop ct.convert from rewriting the traced model's TorchScript graph
#2859
opened Sep 13, 2026 by
Om-singhaI
Contributor
Loading…
Fix FP16 weight detection in NeuralNetworkBuilder.make_updatable (only first layer was checked)
#2858
opened Sep 12, 2026 by
Anai-Guo
Loading…
Clarify supported and preferred devices in quantization docs
#2857
opened Sep 12, 2026 by
cadamcat
Loading…
Fix rank-3 softmax conversion to NeuralNetwork SoftmaxLayer
#2856
opened Sep 12, 2026 by
cpruijsen
Loading…
4 of 5 tasks
Guard against zero scale for all-zero blocks in weight quantization
#2851
opened Sep 8, 2026 by
TusharND12
Loading…
Reject inverted RangeDim bounds when upper_bound is positive
#2849
opened Sep 5, 2026 by
rioyu123
Loading…
Do not count ceil_mode overhang in the avg_pool divisor
#2844
opened Aug 30, 2026 by
LeSingh1
Contributor
Loading…
Add sym_float and trunc support for dynamic float-scale interpolate
#2843
opened Aug 27, 2026 by
SakshamKapoor2911
Contributor
Loading…
Only take the single-bool-mask index path when the mask is the only index
#2842
opened Aug 27, 2026 by
LeSingh1
Contributor
Loading…
Support torch.gather indices smaller than the input
#2841
opened Aug 27, 2026 by
LeSingh1
Contributor
Loading…
Keep the integer dtype in torch.clamp when only one bound is given
#2839
opened Aug 27, 2026 by
LeSingh1
Contributor
Loading…
Fix: shape ops are never folded into quantized weights, leaving them mirrored vs dense
#2831
opened Aug 24, 2026 by
kasper0406
Contributor
Loading…
Use FindPython3 consistently for source builds
#2830
opened Aug 23, 2026 by
cdeil
Contributor
Loading…
Acquire the GIL when releasing NumPy-backed Core ML inputs
#2829
opened Aug 23, 2026 by
cdeil
Contributor
Loading…
Fix weight_id collision when materializing multiple source functions
#2828
opened Aug 23, 2026 by
kasper0406
Contributor
Loading…
Fix SIGSEGV: release prediction input arrays with the GIL held
#2827
opened Aug 21, 2026 by
kasper0406
Contributor
Loading…
Fix rms_norm eps inflation from max-|x| rescale; guard all-zero input
#2822
opened Aug 19, 2026 by
nschlaepfer
Loading…
Add InstanceNorm3d alias-conversion regression coverage
#2820
opened Aug 17, 2026 by
sfeng01-tech
Loading…
Support PyTorch angle and trig ops (deg2rad, rad2deg, hypot)
#2817
opened Aug 15, 2026 by
jaseem-kurikkal
Contributor
Loading…
Decompose Softplus in TF frontend to avoid fp16 overflow on ANE
#2814
opened Aug 10, 2026 by
axiom-of-choice
Loading…
Read torch rand's dtype from the argument that holds it
#2813
opened Aug 9, 2026 by
LeSingh1
Contributor
Loading…
Previous Next
ProTip!
Mix and match filters to narrow down what you’re looking for.