GPUCodeForces/S1
ZZZJ e5ff0163ad fixes DequantizeLinear #88 2025-12-09 20:05:08 +08:00
..
1 fixes #1 2025-10-29 10:49:42 +08:00
2 fixes groupnorm #2 2025-10-30 15:55:47 +08:00
3 fixes softmax #3 2025-10-30 17:54:54 +08:00
4 fixes geglu #4 2025-10-30 18:06:44 +08:00
5 fixes reglu #5 2025-10-30 21:30:18 +08:00
6 finish layernorm #6 2025-10-31 15:24:46 +08:00
7 finish Instancenorm #7 2025-11-03 17:18:15 +08:00
8 fixes localresponsenorm #8 2025-11-04 10:26:53 +08:00
9 fix import missing 2025-10-18 11:10:45 +08:00
10 optimize batchnorm1d operator 2025-10-30 12:32:48 +08:00
11 change dir name 2025-11-06 16:18:38 +08:00
12 redirect to 30 2025-11-12 09:41:29 +08:00
13 finish evonorm #13 2025-11-04 17:14:00 +08:00
14 fixes bcewithlogitsloss #14 2025-11-06 10:39:11 +08:00
15 fixes hingeembeddingloss #15 2025-11-05 16:37:13 +08:00
16 fixes marginrankingloss #16 2025-11-06 10:47:24 +08:00
17 finish crossentropyloss #17 2025-11-05 22:15:59 +08:00
18 finish circleloss #18 2025-11-06 12:01:38 +08:00
19 finish infonceloss #19 2025-11-06 13:16:31 +08:00
20 finish poissonnllloss #20 2025-11-06 14:13:58 +08:00
21 fixes upsamplingnearest2d #21 2025-11-06 17:38:05 +08:00
22 finish ReflectionPad3d #22 2025-11-07 15:06:18 +08:00
23 finish ReflectionPad2d #23 2025-11-07 15:10:54 +08:00
24 finish ReflectionPad1d #24 2025-11-07 15:19:07 +08:00
25 add a matmul operator 2025-11-11 13:38:45 +08:00
26 fixes #26 2025-11-12 09:40:59 +08:00
27 fixes 2025-11-11 18:46:54 +08:00
28 fixes mseloss #28 2025-11-12 09:29:33 +08:00
29 fixes kldivloss #29 2025-11-11 19:15:50 +08:00
30 redirect to 30 2025-11-12 09:41:29 +08:00
31 finish GaussianNLLLoss #31 2025-11-12 11:17:34 +08:00
32 finish NLLLoss #32 2025-11-12 11:22:16 +08:00
33 finish TanimotoCoefficient #33 2025-11-12 11:39:18 +08:00
35 fixed softmarginloss 2025-11-12 14:06:06 +08:00
36 finish multimarginloss 2025-11-12 14:48:12 +08:00
37 fixes pairwisedistance #37 2025-11-12 15:03:58 +08:00
38 fixes pairwisedistance #38 2025-11-12 15:07:00 +08:00
39 finish TripletMarginWithDistanceLoss #39 2025-11-12 15:11:24 +08:00
40 finish SmoothL1Loss #40 2025-11-12 16:15:55 +08:00
41 finish PearsonCorrelation #41 2025-11-12 21:22:03 +08:00
42 finish LogCoshLoss #42 2025-11-12 21:27:42 +08:00
43 finish linear_gelu operator #43 2025-11-13 10:10:33 +08:00
44 finish multilabelmarginloss #44 2025-11-13 10:58:25 +08:00
Example 强调提交规范 2025-11-13 16:19:48 +08:00
Ljy123_#1 finish layernorm #1 2025-11-13 21:31:19 +08:00
Ljy123_#3 finish gelu 2025-11-17 21:41:40 +08:00
Ljy123_#6 optimzied softmax 2025-11-17 22:01:20 +08:00
Ljy123_#7 finish swish 2025-11-18 21:00:38 +08:00
Ljy123_#8 finish rmsnorm 2025-11-18 21:02:51 +08:00
Ljy123_#9 finish affine-relu 2025-12-01 20:34:28 +08:00
Ljy123_#11 finish bias-swish #11 2025-12-01 20:38:09 +08:00
Ljy123_#12 finish glu-fuse #12 2025-12-01 20:39:36 +08:00
Ljy123_#13 optimized affine-relu #13 2025-12-01 20:41:09 +08:00
Ljy123_#14 finish relu-square #14 2025-12-01 20:44:34 +08:00
Ljy123_#15 finish softplus-bias #15 2025-12-01 20:46:04 +08:00
Ljy123_#16 finish sigmoid-clamp-affine #15 2025-12-01 20:48:04 +08:00
Ljy123_#17 finish mish-bias #17 2025-12-01 23:39:33 +08:00
Ljy123_#18 finish softsign-affine #18 2025-12-01 23:40:55 +08:00
Ljy123_#20 finish tanshrink affine #20 2025-12-01 23:42:56 +08:00
Ljy123_#21 optimized sigmoid (power-sigmoid-gate) #21 2025-12-01 23:44:40 +08:00
Ljy123_#22 finish gaussian-bias #22 2025-12-02 15:21:35 +08:00
Ljy123_#23 optimized sigmoid (hardsigmoid-affine) #23 2025-12-02 15:23:15 +08:00
Ljy123_#24 finish spherical-affine #24 2025-12-02 15:24:48 +08:00
Ljy123_#25 finish swish2bias #25 2025-12-02 15:26:08 +08:00
Ljy123_#27 optimized bias (log1p-square-bias) #27 2025-12-02 15:28:42 +08:00
Ljy123_#28 finish sqrt-relu-affine #28 2025-12-03 18:18:49 +08:00
Ljy123_#29 finish expm1 #29 2025-12-03 18:20:01 +08:00
Ljy123_#30 finish softclip-affine #30 2025-12-03 18:21:31 +08:00
Ljy123_#31 finish leaky-swish-bias #31 2025-12-03 18:22:45 +08:00
Ljy123_#33 finish softmax-channel-affine #33 2025-12-03 18:25:39 +08:00
Ljy123_#34 finish contrastive-gate #34 2025-12-03 18:26:57 +08:00
Ljy123_#35 finish channel-rms-gate #35 2025-12-03 23:27:38 +08:00
Ljy123_#36 finish prototype-cosine-gate #36 2025-12-03 23:28:58 +08:00
Ljy123_#37 optimized softmax-affine (use group) #37 2025-12-03 23:30:21 +08:00
Ljy123_#38 finish spatial-diff gate #38 2025-12-03 23:32:42 +08:00
Ljy123_#39 finish range-gate-affine #39 2025-12-03 23:34:47 +08:00
Ljy123_#40 optimized channel-affine(temp-softmax) #40 2025-12-03 23:37:37 +08:00
Ljy123_#41 finish logsigmoid (affine-gate) #41 2025-12-03 23:40:06 +08:00
Ljy123_#42 finish exp-square #42 2025-12-03 23:41:39 +08:00
Lwh20070813_#1 optimized PearsonCorraletion #1 2025-11-23 14:06:09 +08:00
Lwh20070813_#2 optimized LogCoshLoss #2 2025-11-23 14:07:37 +08:00
ZZZJ_#1 fixed cosineloss #1 2025-11-17 18:23:48 +08:00
ZZZJ_#88 fixes DequantizeLinear #88 2025-12-09 20:05:08 +08:00
gsd123_#1 fix reglu #1 2025-11-14 10:50:31 +08:00
gsd123_#2 fix geglu #2 2025-11-14 10:52:35 +08:00
gsd123_#3 fix layernorm #3 2025-11-14 10:54:10 +08:00
gsd123_#4 fix Instancenrm #4 2025-11-14 10:55:28 +08:00
gsd123_#5 fix evonorm #5 2025-11-14 10:56:40 +08:00
gsd123_#6 fix CrossEntropyLoss #6 2025-11-14 10:57:42 +08:00
gsd123_#7 fix poissonnllloss #7 2025-11-14 10:58:43 +08:00
gsd123_#8 fix infonceloss #8 2025-11-14 10:59:44 +08:00
gsd123_#9 fix circleloss#9 2025-11-14 11:01:22 +08:00
gsd123_#10 fix ReflectionPad1d #10 2025-11-14 11:02:36 +08:00
gsd123_#11 fix ReflectionPad2d #11 2025-11-14 11:03:48 +08:00
gsd123_#12 fix ReflectionPad3d #12 2025-11-14 11:04:58 +08:00
gsd123_#13 fix SmoothL1Loss #13 2025-11-14 11:06:01 +08:00
gsd123_#14 fix TripletMarginWithDistanceLoss #14 2025-11-14 11:07:01 +08:00
gsd123_#15 fix CosineEmbeddingLoss #15 2025-11-14 11:08:28 +08:00
gsd123_#16 finish CauchyLoss #16 2025-11-17 14:14:54 +08:00
gsd123_#17 finish DiceLoss #17 2025-11-18 16:51:32 +08:00
gsd123_#18 finish FocalLoss #18 2025-11-18 16:52:46 +08:00
gsd123_#19 finish Generator #19 2025-11-18 16:54:17 +08:00
gsd123_#20 finish WelschLoss #20 2025-11-18 16:55:28 +08:00
gsd123_#21 finish CharbonnierLoss #21 2025-11-25 19:47:41 +08:00
gsd123_#22 finish contrastiveloss #22 2025-11-25 19:49:13 +08:00
gsd123_#23 finish huberloss #23 2025-11-25 19:50:22 +08:00
gsd123_#24 finish kldivloss #24 2025-11-25 19:51:35 +08:00
gsd123_#25 finish softmax2d #24 2025-11-25 19:52:36 +08:00
gsd123_#26 finish CanberraDistance #26 2025-11-25 20:09:17 +08:00
gsd123_#27 finish SorensenDice #27 2025-11-25 20:25:33 +08:00
gsd123_#28 finish clarksDistance #28 2025-11-25 21:02:00 +08:00
gsd123_#29 finish chebyshevDistance #29 2025-11-25 21:04:02 +08:00
gsd123_#30 finish mahalanobisDistance #30 2025-11-25 21:06:02 +08:00
gsd123_#31 finish Mish #31 2025-11-26 10:55:57 +08:00
gsd123_#32 finish Adaptivepiecewiselinear #32 2025-12-03 09:26:08 +08:00
gsd123_#33 finish AHAF #33 2025-12-03 09:27:27 +08:00
gsd123_#34 finish Colu #34 2025-12-03 09:28:48 +08:00
gsd123_#35 finish Elish #35 2025-12-03 09:29:43 +08:00
gsd123_#36 finish hardmish #36 2025-12-03 09:30:52 +08:00
gsd123_#37 finish PDELU #37 2025-12-03 09:31:51 +08:00
gsd123_#38 finish Phish #38 2025-12-03 09:32:56 +08:00
gsd123_#39 finish PiecewiseLinearUnit #39 2025-12-03 09:33:57 +08:00
gsd123_#40 finish Serlu #40 2025-12-03 09:35:04 +08:00
gsd123_#41 finish CLL #41 2025-12-03 11:16:05 +08:00
gsd123_#42 finish CosLU #42 2025-12-03 11:17:00 +08:00
gsd123_#43 finish CosReLU #43 2025-12-03 11:18:05 +08:00
gsd123_#44 finish DELU #44 2025-12-03 11:18:59 +08:00
gsd123_#45 finish FReLU #45 2025-12-03 11:20:21 +08:00
gsd123_#46 finish FunnelActivationforVisualRecognition #46 2025-12-03 11:21:50 +08:00
gsd123_#48 finish InvMultiquadratic #48 2025-12-03 11:23:56 +08:00
gsd123_#49 finish ISRLU #49 2025-12-03 11:24:47 +08:00
gsd123_#50 finish ISRU #50 2025-12-03 11:25:40 +08:00
gsd123_#51 finish marcsing #51 2025-12-03 11:26:31 +08:00
gsd123_#52 finish ModReLU #52 2025-12-03 11:27:40 +08:00
gsd123_#53 finish Flatten-T #53 2025-12-03 13:05:56 +08:00
gsd123_#54 finish Multiquadratic #54 2025-12-03 11:30:01 +08:00
gsd123_#55 finish Flatten-T Swish #53 2025-12-03 13:14:50 +08:00
gsd123_#56 finish ShiftedSoftPlus #56 2025-12-03 13:28:04 +08:00
gsd123_#57 finish SinReLU #57 2025-12-03 11:32:56 +08:00
gsd123_#58 finish SQNL #58 2025-12-03 11:34:01 +08:00
gsd123_#59 finish SQRBF #59 2025-12-03 11:34:52 +08:00
gsd123_#60 finish QReLu #60 2025-12-03 11:37:12 +08:00
gsd123_#61 finish SquaredReLU #61 2025-12-03 11:38:05 +08:00
hli28146_#1 finish tripletmarginloss #1 2025-11-17 16:11:21 +08:00
hli28146_#2 finish multimarginloss 2025-11-17 16:18:25 +08:00
hli28146_#3 finish softmarginloss #3 2025-11-17 16:22:19 +08:00
hli28146_#4 finish huberloss #4 2025-11-17 16:26:20 +08:00
hli28146_#5 finish multilabelmarginloss #5 2025-11-17 16:30:36 +08:00
hli28146_#6 finish lppool1d #6 2025-11-18 10:35:21 +08:00
hli28146_#7 finish cholesky #7 2025-11-14 18:41:56 +08:00
hli28146_#8 finish vecdot #8 2025-11-17 22:28:01 +08:00
hli28146_#9 finish cross #9 2025-11-17 22:30:08 +08:00
hli28146_#10 finish logsumexp #10 2025-11-17 22:36:15 +08:00
hli28146_#11 finish cumsum #11 2025-11-25 16:20:09 +08:00
hli28146_#12 finish gempool #12 2025-11-26 00:03:49 +08:00
hli28146_#13 finish giouloss #13 2025-11-26 00:34:03 +08:00
hli28146_#14 finish wingloss #14 2025-11-26 16:53:48 +08:00
hli28146_#15 finish poly-1 crossentropy loss #15 2025-11-26 18:46:10 +08:00
hli28146_#16 finish poly-1 focalloss #16 2025-11-26 19:50:54 +08:00
hli28146_#17 finish bprloss #17 2025-11-26 19:26:05 +08:00
hli28146_#18 finish logish #18 2025-11-27 00:00:53 +08:00
hli28146_#19 finish LiSHT #19 2025-11-27 00:11:37 +08:00
hli28146_#20 finish tanhexp #20 2025-11-27 00:19:19 +08:00
hli28146_#21 finish cosfaceloss #21 2025-12-02 14:15:27 +08:00
hli28146_#22 finish arcfaceloss #22 2025-12-02 14:22:02 +08:00
hli28146_#23 finish spherefaceloss #23 2025-12-02 14:27:23 +08:00
hli28146_#24 finish Flatten-T swish #24 2025-12-02 15:04:40 +08:00
hli28146_#25 finish ACON-C #25 2025-12-02 17:21:13 +08:00
hli28146_#26 finish meta-ACON #26 2025-12-02 17:42:30 +08:00
hli28146_#27 finish SmeLU #27 2025-12-02 20:12:19 +08:00
hli28146_#28 finish Phish #28 2025-12-02 20:26:34 +08:00
hli28146_#29 finish GCU #29 2025-12-02 20:43:36 +08:00
hli28146_#30 finish ASU #30 2025-12-02 21:03:39 +08:00
uucoco_#1 fix GaussianNLLLoss #1 2025-11-13 20:28:09 +08:00
uucoco_#2 fix NLLLoss #2 2025-11-13 20:38:27 +08:00
uucoco_#3 fix TanimotoCoefficient #3 2025-11-13 20:40:59 +08:00
uucoco_#4 finish HammingDistance #4 2025-11-13 20:45:47 +08:00
uucoco_#5 finish BhattacharyyaDistance #5 2025-11-13 20:48:52 +08:00
uucoco_#6 finish HellingerDistance #6 2025-11-13 20:50:33 +08:00
uucoco_#7 finish MinkowskiDistance #7 2025-11-13 20:52:16 +08:00
uucoco_#8 finish ChannelShuffle #8 2025-11-16 13:23:46 +08:00
uucoco_#9 finish JaccardSimilarity #9 2025-11-16 13:27:30 +08:00
uucoco_#10 finish CharbonnierLoss #10 2025-11-18 21:54:57 +08:00
uucoco_#11 finish IntraClassCorrelation #11 2025-11-18 21:56:49 +08:00
uucoco_#12 finish CoVariance #12 2025-11-18 22:00:30 +08:00
uucoco_#13 finish Variance #13 2025-11-18 22:01:36 +08:00
uucoco_#14 finish softmin #14 2025-11-23 19:50:59 +08:00
uucoco_#15 finish softplus #15 2025-11-23 19:52:17 +08:00
uucoco_#16 finish softshrink #16 2025-11-23 19:53:30 +08:00
uucoco_#17 finish IOULoss#17 2025-11-23 20:44:51 +08:00
uucoco_#18 finish softsign #18 2025-11-23 20:58:11 +08:00
uucoco_#20 finish SoftExponential #20 2025-12-02 20:23:18 +08:00
uucoco_#21 finish SoftClip #21 2025-12-02 20:24:29 +08:00
uucoco_#22 finish LogWeightedSumExp #22 2025-12-02 20:25:50 +08:00
uucoco_#23 finish LogSigmoid #23 2025-12-02 20:26:55 +08:00
uucoco_#24 finish LogMeanExp #24 2025-12-02 20:28:25 +08:00
uucoco_#25 finish HardTanh #25 2025-12-02 20:29:31 +08:00
uucoco_#26 finish HardSwish #26 2025-12-02 20:30:31 +08:00
uucoco_#27 finish HardSigmoid #27 2025-12-02 20:31:45 +08:00
uucoco_#28 finish hardshrink #28 2025-12-02 20:33:01 +08:00
wut0n_#1 feat: Add high-performance LayerNorm operator with vectorized Welford algorithm #1 2025-11-19 21:21:51 +08:00
wut0n_#2 feat:add focal loss optimization 2025-11-20 18:35:27 +08:00
wut0n_#3 feat:Rmsnorm #3 2025-11-20 21:12:55 +08:00
wut0n_#4 feat:Instancenorm #4 2025-11-20 21:15:22 +08:00
wut0n_#5 feat: add high-performance GroupNorm CUDA implementation 2025-11-21 00:50:30 +08:00
wut0n_#6 feat:add high performance diceloss #6 2025-11-21 19:40:36 +08:00
wut0n_#7 feat:add high performance tripletloss #7 2025-11-23 16:19:02 +08:00
wut0n_#8 feat:add high performance Bceloss #8 2025-11-22 16:38:47 +08:00
wut0n_#9 feat:Add high performance L1loss #9 2025-11-21 21:21:58 +08:00
wut0n_#10 feat:add high performance Mseloss #10 2025-11-21 21:51:41 +08:00
wut0n_#11 feat:add high performance contrastiveloss #11 2025-11-23 14:02:40 +08:00
wut0n_#12 feat:add high performance #12 2025-11-23 16:21:47 +08:00
wut0n_#13 feat:add high performance hingeloss #13 2025-11-23 16:23:34 +08:00
wut0n_#14 fixs:repair performance euclidean #14 2025-11-25 15:41:36 +08:00
wut0n_#15 feat:add high performance cosinedistance #15 2025-11-23 16:27:11 +08:00
wut0n_#16 feat:add high performance Chebyshev distance #16 2025-11-23 23:29:18 +08:00
wut0n_#17 feat:add high performance Manhattan distance #17 2025-11-23 23:31:47 +08:00
wut0n_#18 feat:add high performance Minkowski distance #18 2025-11-23 23:33:23 +08:00
wut0n_#19 feat:add high performance cosine+contrasstive #19 2025-12-04 14:56:27 +08:00
wut0n_#20 feat:add good performance Hamming #20 2025-11-25 14:02:11 +08:00
wut0n_#21 fix:add high performance canberra #21 2025-12-03 00:09:03 +08:00
wut0n_#22 feat:add good performance Braycurtis #22 2025-11-25 17:11:19 +08:00
wut0n_#23 feat:add high performance squared_euclidean #23 2025-11-25 17:13:35 +08:00
wut0n_#24 feat:add high performance cosine #24 2025-11-25 17:15:34 +08:00
wut0n_#25 feat:add high performance weighted_sum #25 2025-11-25 17:17:36 +08:00
wut0n_#26 feat:add good variance #26 2025-11-25 17:19:36 +08:00
wut0n_#27 feat:add high performance sigmoid derivative #27 2025-12-03 00:26:44 +08:00
wut0n_#28 feat:add high performance Logsumexp #28 2025-12-03 00:30:17 +08:00
wut0n_#29 feat:add high performance Logbeta #29 2025-12-03 00:32:11 +08:00
wut0n_#31 feat:add high performance focallossfuse #31 2025-12-04 14:59:40 +08:00
wut0n_#32 feat:add high performance rmsnormresidual #32 2025-12-04 15:01:24 +08:00
wwmm_#1 fixes rope #1 2025-11-17 20:58:14 +08:00
wwmm_#2 fixes cosinesimilarity #2 2025-11-17 21:06:28 +08:00
wwmm_#3 fixes embeddingbag #3 2025-11-17 21:12:52 +08:00
zizi05_#1 feat:add mish #1 2025-11-22 15:02:47 +08:00
zizi05_#3 feat:add pairwise_distance #3 2025-11-22 15:44:18 +08:00
zizi05_#4 feat:add a selu_clip #4 2025-11-22 15:50:40 +08:00
zizi05_#6 feat:add a bilinear #6 2025-11-23 16:14:27 +08:00
zizi05_#7 feat:add a swishgelu#7 2025-11-26 10:20:24 +08:00
zizi05_#8 feat:add a hardmish #8 2025-11-25 15:58:07 +08:00
zizi05_#9 feat:add a tanhshrink #9 2025-11-25 16:17:37 +08:00
zizi05_10 finish add a swish_layernorm #10 2025-11-25 16:51:57 +08:00
zizi_05#5 feat:add a localresponsenormalizetion_simple #5 2025-11-23 14:56:12 +08:00
README.md Update README.md 2025-11-13 16:32:02 +08:00

README.md

提交前的注意事项

  • 确保已经阅读了赛题入门代码解读的内容
  • 你的提交包含类似于S1/Example下的四份完整文件
  • 选手提交时的run_code.py应于S1/Example/run_code.py中的保持一致