-
Notifications
You must be signed in to change notification settings - Fork 923
Insights: ml-explore/mlx
Overview
-
- 9 Merged pull requests
- 2 Open pull requests
- 3 Closed issues
- 0 New issues
Could not load contribution data
Please try again later
1 Release published by 1 person
-
v0.16.3
published
Aug 12, 2024
9 Pull requests merged by 4 people
-
Fix compiling with space in paths
#1332 merged
Aug 15, 2024 -
Fix optimizer reloading from checkpoint
#1329 merged
Aug 15, 2024 -
Read arrays from files faster
#1330 merged
Aug 15, 2024 -
GPU mx.sign for complex64
#1326 merged
Aug 14, 2024 -
Fix transformer encoder with
norm_first=False
#1327 merged
Aug 13, 2024 -
Add
isfinite
#1318 merged
Aug 13, 2024 -
Fix: Preserve input dtype in Dropout layer output
#1323 merged
Aug 13, 2024 -
Patch
#1320 merged
Aug 12, 2024 -
Add memory_efficient_threshold kwarg to sdpa kernel
#1319 merged
Aug 12, 2024
2 Pull requests opened by 2 people
-
Custom Metal Kernels from Python
#1325 opened
Aug 13, 2024 -
move has no effect on const variables
#1333 opened
Aug 16, 2024
3 Issues closed by 1 person
-
[BUG] Unable to load from a saved checkpoint, `KeyError` for all dropout modules...
#1328 closed
Aug 15, 2024 -
[BUG] Convergence issue in MLX when compared to PyTorch
#1324 closed
Aug 14, 2024 -
[BUG] Dropout not preserving `dtype`
#1321 closed
Aug 13, 2024
5 Unresolved conversations
Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.
-
Support for mlx.float64
#799 commented on
Aug 12, 2024 • 0 new comments -
[Feature] Cholesky decomposition
#1026 commented on
Aug 13, 2024 • 0 new comments -
[Performance] PyTorch (MPS) is faster than MLX in backward of convolution layer
#1313 commented on
Aug 13, 2024 • 0 new comments -
Pinv
#875 commented on
Aug 14, 2024 • 0 new comments -
Fix scatter, reduce
#1300 commented on
Aug 15, 2024 • 0 new comments