issues
search
ml-explore
/
mlx
MLX: An array framework for Apple silicon
https://ml-explore.github.io/mlx/
MIT License
17.37k
stars
1.01k
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
3 and 6 bit quantization
#1613
barronalex
opened
12 hours ago
2
Contiguous op + primitive
#1612
awni
closed
8 hours ago
2
Matrix Attention kernel
#1610
jagrit06
opened
1 day ago
6
fix typo: accross -> across
#1609
cocoa-xu
closed
1 day ago
0
Reduce specializations
#1607
awni
closed
8 hours ago
1
Fix formatting after last PR
#1606
angeloskath
closed
1 day ago
0
[Enhancement Request] SDFA does not support head dimension of size 192 (capped at 128)
#1604
s-tlgh
opened
2 days ago
0
Fix view scalar bug segfault
#1603
awni
closed
2 days ago
0
[BUG] Unable to build recent commits related to osx deployment
#1602
chimezie
closed
2 days ago
1
Fix cmake version
#1601
awni
closed
2 days ago
0
Fix concatenate vmap
#1600
angeloskath
closed
2 days ago
0
[BUG] really bizarre behavior from vmap
#1599
feitreim
closed
2 days ago
3
2-Pass Sdpa Inference Kernel
#1597
angeloskath
closed
3 days ago
0
Allocate raw JSON metadata buffer on the heap, and limit its size
#1596
xnorai
closed
3 days ago
1
Use osx deployment target to pick Metal version
#1595
awni
closed
3 days ago
0
Fix dispatch threads for a few kernels
#1594
awni
closed
3 days ago
0
[BUG] mx.fft.irfftn is inconsistent with np
#1592
stockeh
closed
4 days ago
1
More buffer donation with no-ops
#1591
awni
closed
3 days ago
0
Fix sibling leak
#1590
awni
closed
3 days ago
0
Faster indexing math in a few kernels
#1589
awni
closed
3 days ago
1
[BUG] leak of array descriptors in random split
#1588
davidkoski
closed
3 days ago
0
Fix donation in sdpa
#1587
awni
closed
1 week ago
0
Add orthogonal initialization
#1586
noahfarr
opened
1 week ago
4
[QUESTION] can we integrate direct MPSGraph features in mlx ?
#1585
thegodone
opened
1 week ago
0
Dispatch bf16 at run time when using the JIT
#1584
awni
closed
6 days ago
0
Add utilities to distributions
#1583
noahfarr
opened
1 week ago
2
[Question] Masked SDPA attention kernel?
#1582
EricLBuehler
closed
1 week ago
2
[BUG] implementing entropy in adam return fix loss without converging
#1581
thegodone
closed
3 days ago
2
[FEATURE] implement adam-mini
#1580
thegodone
opened
1 week ago
0
OOB QMV fix
#1579
barronalex
closed
1 week ago
1
Cpu fast quantize
#1578
awni
closed
1 week ago
0
Fix OOB access in qmv
#1577
barronalex
closed
1 week ago
1
Throw for invalid *fft{2,n} cases and remove test
#1575
awni
closed
1 week ago
0
Fix per-example mask + docs in sdpa
#1574
awni
closed
1 week ago
0
Fully wrap the command encoder
#1572
awni
closed
1 week ago
1
[RFC] Multithread stream
#1570
awni
opened
2 weeks ago
1
Add groups in nn.Conv2d
#1569
awni
closed
2 weeks ago
1
Bfs width limit
#1568
awni
closed
1 week ago
6
Use dispatch method everywhere
#1566
awni
closed
2 weeks ago
0
v0.20.0
#1565
barronalex
closed
2 weeks ago
0
Add split_k `qvm` for long context
#1564
barronalex
closed
2 weeks ago
0
Fix gather vmap
#1563
awni
closed
2 weeks ago
1
Rms error for wrong weight size
#1562
awni
closed
2 weeks ago
0
Use fewer barriers
#1561
awni
closed
2 weeks ago
0
Sdpa fix
#1558
angeloskath
closed
2 weeks ago
1
No extra reshape
#1557
awni
closed
2 weeks ago
0
Fix inconsistent example in "Automatic Vectorization" docs
#1556
chrisoffner3d
closed
2 weeks ago
0
Fix typo "it's" -> "its" in docs for "Lazy Evaluation"
#1555
chrisoffner3d
closed
2 weeks ago
1
Add python 3.13 to circle
#1553
awni
closed
2 weeks ago
0
Return an mlx array excluding a set of columns and rows similar to np.delete
#1552
kyrollosyanny
opened
2 weeks ago
0
Next