geohot
cee9fc7540
new_ignore
2025-03-27 13:16:30 +08:00
geohot
444d6279ac
flod better
2025-03-27 12:03:15 +08:00
geohot
5dd59a6096
touchup
2025-03-27 00:23:58 +08:00
geohot
6bec82b918
on 7
2025-03-27 00:05:18 +08:00
geohot
a436d7542f
up 7
2025-03-27 00:00:40 +08:00
George Hotz and GitHub
31ffa1607e
Merge branch 'master' into dsp_search
2025-03-26 21:43:22 +08:00
George Hotz and GitHub
1e6e75e39a
little changes from dsp branch ( #9582 )
...
* little changes from dsp branch
* not that one
* need the where
* Revert "need the where"
This reverts commit 140f89c878 .
2025-03-26 20:01:21 +08:00
Andrey and GitHub
7b865ed03d
use tuple in isinstance for type checking ( #9583 )
2025-03-26 19:36:48 +08:00
geohot
e0d63696d7
cleanups
2025-03-26 17:55:48 +08:00
geohot
905f847d10
fix dq
2025-03-26 17:43:10 +08:00
geohot
9e19cdfbbe
e kernels
2025-03-26 16:49:23 +08:00
geohot
f7b38fa94c
make that 8
2025-03-26 16:46:49 +08:00
geohot
880b4a5e47
put that back
2025-03-26 15:34:57 +08:00
geohot
2e4cae342b
less terrible first
2025-03-26 15:21:38 +08:00
geohot
f1ff18acec
prepad kernel weights
2025-03-26 12:13:46 +08:00
George Hotz and GitHub
60cbfe4222
Merge branch 'master' into dsp_search
2025-03-26 10:50:00 +08:00
nimlgen and GitHub
4cf2b68ca8
am_smi: fix init for newer versions ( #9559 )
2025-03-25 23:48:05 +07:00
geohot
311df3ff21
fixes
2025-03-25 19:14:57 +08:00
geohot
f6e64a5e8e
optional conv
2025-03-25 19:00:48 +08:00
geohot
a9f1227625
4 faster
2025-03-25 18:32:27 +08:00
geohot
74c2587ef4
4 -> 8
2025-03-25 18:31:43 +08:00
geohot
66a90a3c92
ugh, fast 2?
2025-03-25 17:08:35 +08:00
geohot
0d76b0d461
acc2
2025-03-25 15:33:38 +08:00
geohot
29920b74d5
unsafe disable on device
2025-03-25 13:57:57 +08:00
geohot
ccd18a803c
faster?
2025-03-25 13:41:34 +08:00
geohot
943bde47ab
fast k26
2025-03-25 13:32:11 +08:00
geohot
0d10c7ae2f
working on kernel 15
2025-03-25 12:00:16 +08:00
Priyank Patel and GitHub
4f5e03bd60
better fix inplace detach ( #9557 )
2025-03-24 22:50:28 +08:00
geohot
3cab6a3d4a
3x3
2025-03-24 16:36:53 +08:00
geohot
22a56cbaea
something for 8
2025-03-24 16:16:59 +08:00
geohot
afd61730b4
kernel 5
2025-03-24 16:10:08 +08:00
geohot
536556434b
padding ish
2025-03-24 15:57:03 +08:00
geohot
2eb9241329
better conv
2025-03-24 13:07:14 +08:00
George Hotz and GitHub
554a490751
Merge branch 'master' into dsp_search
2025-03-24 12:29:22 +08:00
George Hotz and GitHub
74d98eafb8
add onnx frontend stub [pr] ( #9558 )
2025-03-24 12:24:34 +08:00
geohot
651c678edf
work
2025-03-24 09:49:53 +08:00
geohot
3274bd2d81
output
2025-03-23 15:13:00 +08:00
geohotstan and GitHub
309afa20b7
add Tensor.max_unpool2d ( #9518 )
...
* why does max_unpool2d feel slower than out.gradient ...
* slightly cleaner
* what happened to ruff
* need to think about this some more
* slightly faster now?
* clean up, 1 more failing edge case
* ok good
* working TINY_BACKEND
* nit doc wording
* retry CI
2025-03-22 12:11:33 -04:00
geohot
30f4d64148
rules
2025-03-22 19:17:16 +08:00
geohot
2634975d5a
5 and 8
2025-03-22 19:14:04 +08:00
geohot
fd73ec2b1b
knum
2025-03-22 18:59:54 +08:00
geohot
e1d2bec4a4
opt
2025-03-22 18:52:56 +08:00
geohot
1b4e9f5e91
more opt rules
2025-03-22 18:07:31 +08:00
geohot
25c023bcbe
more
2025-03-22 17:49:34 +08:00
geohot
26b02a037c
fix 33
2025-03-22 17:17:47 +08:00
geohot
dca95428a5
touch
2025-03-22 11:05:36 +08:00
Francis Lata and GitHub
eb95825eea
RetinaNet dataloader ( #9442 )
...
* retinanet dataloader
* remove batch_size from generate_anchors
* refactor kits19 dataset tests
* add tests for dataloader
* fix testing setup and cleanups
* remove unused import
2025-03-21 13:36:41 -04:00
geohot
af94addb3a
ish
2025-03-21 17:46:45 +08:00
geohot
0416b0998d
revert those
2025-03-21 17:15:38 +08:00
George Hotz and GitHub
8e555c586c
switch quantization to unsigned/unsigned + add Ops.REDUCE ( #9527 )
...
* switch quantization to unsigned/unsigned + add Ops.REDUCE
* tests
* nhwc + replay pkl
2025-03-21 17:02:37 +08:00