Default Branch

5adf185f86 · Fix update_modules() when providing a subset (#2308) · Updated 2025-06-21 08:19:46 +08:00

Branches

fd1d0821d2 · Make sure softmax doesn't change the actual max · Updated 2025-06-23 14:34:32 +08:00

0
11

870208eff5 · Start sdpa vector · Updated 2025-06-17 08:38:39 +08:00

6
1

c830b5a9f9 · fix metal kernel linking issue on cuda · Updated 2025-06-11 03:17:29 +08:00

29
2

0b38729f41 · rebase · Updated 2025-06-04 09:03:47 +08:00

28
58
fft

83762691ba · Fix four step fft · Updated 2025-05-09 05:14:59 +08:00    zhangyiss

31
6

7c99acb799 · split logsumexp · Updated 2025-05-07 08:10:14 +08:00    zhangyiss

32
1

998404ada4 · Get trellis to run · Updated 2025-04-26 22:02:20 +08:00    zhangyiss

71
3

11f73d6e89 · Double buffer keys for vector sdpa · Updated 2025-04-22 15:19:11 +08:00    zhangyiss

60
1

4c46e17a5d · Update benchmark output · Updated 2025-04-16 01:50:06 +08:00    zhangyiss

70
1

67ec27d515 · synch before reading memory in test · Updated 2025-04-08 05:37:32 +08:00    zhangyiss

84
4

066336b60e · load q4_k from gguf · Updated 2025-04-04 01:56:12 +08:00    zhangyiss

94
1

688e421184 · only interrupt during an eval · Updated 2025-03-19 22:56:26 +08:00    zhangyiss

141
2

127de8821e · Fix the sig_handler check · Updated 2025-03-08 09:31:06 +08:00    zhangyiss

155
2

c5073fc452 · Ensure we only have one copy of the fence · Updated 2025-03-05 15:37:15 +08:00    zhangyiss

164
3

4c1dfa58b7 · xor op on arrays (#1875) · Updated 2025-02-17 16:24:53 +08:00    zhangyiss

191
0
Included

4515866024 · Change the linux test to ubuntu 24.04 · Updated 2025-01-21 14:58:05 +08:00    zhangyiss

245
9

f14b4d72de · Remove unnecessary copy from winograd · Updated 2025-01-07 06:06:03 +08:00    zhangyiss

264
1

c02e14c264 · Add the 3bit packed qmm_t · Updated 2024-12-18 14:16:30 +08:00    zhangyiss

300
19

0c1155faf5 · binding + tests · Updated 2024-12-10 04:57:36 +08:00    zhangyiss

318
3

82a956c1d9 · fix test · Updated 2024-12-07 02:26:54 +08:00    zhangyiss

337
5