[KMCompiler][Nvidia] add operator linalg_lu_factor by YangLong114514 · Pull Request #4945 · flagos-ai/FlagGems

YangLong114514 · 2026-07-24T03:58:30Z

PR Category

Operator

Type of Change

Other

Description

Add operator linalg_lu_factor with triton.
Computes the LU decomposition of a matrix with partial pivoting. Given input A (*, m, n), returns the compact LU matrix and pivot indices (*, min(m,n), int32), satisfying P @ A = L @ U. L is implicitly stored as a unit lower-triangular matrix in the lower part of LU; U is stored in the upper part. Supports pivot=False to disable pivoting (CUDA only).

Issue

Progress

Change is properly reviewed (1 reviewer required, 2 recommended).
Change is responded to an issue.
Change is fully covered by a UT.

Performance

test_linalg_lu_factor.py
Operator: linalg_lu_factor  Performance Test (dtype=torch.float32, mode=kernel,level=comprehensive)
Status       Torch Latency (ms)    Gems Latency (ms)         Gems Speedup          Size Detail
-----------------------------------------------------------------------------------------------
SUCCESS               0.064656            0.021792               2.967          ([torch.Size([16, 16])], {'pivot': True})
SUCCESS               0.070112            0.013632               5.143          ([torch.Size([16, 16])], {'pivot': False})
SUCCESS               0.086528            0.050240               1.722          ([torch.Size([32, 32])], {'pivot': True})
SUCCESS               0.071248            0.029728               2.397          ([torch.Size([32, 32])], {'pivot': False})
SUCCESS               0.127808            0.155072               0.824          ([torch.Size([64, 64])], {'pivot': True})
SUCCESS               0.070944            0.086560               0.820          ([torch.Size([64, 64])], {'pivot': False})
SUCCESS               0.214432            0.583872               0.367          ([torch.Size([128, 128])], {'pivot': True})
SUCCESS               0.097408            0.134176               0.726          ([torch.Size([128, 128])], {'pivot': False})
SUCCESS               0.408256            1.523168               0.268          ([torch.Size([256, 256])], {'pivot': True})
SUCCESS               0.140752            1.028960               0.137          ([torch.Size([256, 256])], {'pivot': False})
SUCCESS               7.762320            2.945776               2.635          ([torch.Size([1024, 512])], {'pivot': True})
SUCCESS               1.893728            1.478832               1.281          ([torch.Size([1024, 512])], {'pivot': False})
SUCCESS               0.107296            0.025232               4.252          ([torch.Size([32, 16])], {'pivot': True})
SUCCESS               0.068832            0.017472               3.940          ([torch.Size([32, 16])], {'pivot': False})
SUCCESS               0.113856            0.025536               4.459          ([torch.Size([16, 32])], {'pivot': True})
SUCCESS               0.075360            0.015904               4.738          ([torch.Size([16, 32])], {'pivot': False})
SUCCESS               0.326016            0.171392               1.902          ([torch.Size([128, 64])], {'pivot': True})
SUCCESS               0.123744            0.078080               1.585          ([torch.Size([128, 64])], {'pivot': False})
SUCCESS               0.367392            0.256448               1.433          ([torch.Size([64, 128])], {'pivot': True})
SUCCESS               0.174848            0.106368               1.644          ([torch.Size([64, 128])], {'pivot': False})
SUCCESS               0.061184            0.050368               1.215          ([torch.Size([4, 32, 32])], {'pivot': True})
SUCCESS               0.052672            0.029792               1.768          ([torch.Size([4, 32, 32])], {'pivot': False})
SUCCESS               0.045312            0.023200               1.953          ([torch.Size([128, 16, 16])], {'pivot': True})
SUCCESS               0.045456            0.014784               3.075          ([torch.Size([128, 16, 16])], {'pivot': False})
SUCCESS              35.109056           37.323967               0.941          ([torch.Size([1024, 512, 512])], {'pivot': True})
SUCCESS              18.171968           26.555040               0.684          ([torch.Size([1024, 512, 512])], {'pivot': False})
SUCCESS             128.263489          144.178818               0.890          ([torch.Size([4096, 512, 512])], {'pivot': True})
SUCCESS              66.834274          103.271805               0.647          ([torch.Size([4096, 512, 512])], {'pivot': False})


Operator: linalg_lu_factor  Performance Test (dtype=torch.float64, mode=kernel,level=comprehensive)
Status       Torch Latency (ms)    Gems Latency (ms)         Gems Speedup          Size Detail
-----------------------------------------------------------------------------------------------
SUCCESS               0.071872            0.026912               2.671          ([torch.Size([16, 16])], {'pivot': True})
SUCCESS               0.101792            0.019392               5.249          ([torch.Size([16, 16])], {'pivot': False})
SUCCESS               0.095360            0.057760               1.651          ([torch.Size([32, 32])], {'pivot': True})
SUCCESS               0.103712            0.037824               2.742          ([torch.Size([32, 32])], {'pivot': False})
SUCCESS               0.145952            0.172736               0.845          ([torch.Size([64, 64])], {'pivot': True})
SUCCESS               0.112928            0.128352               0.880          ([torch.Size([64, 64])], {'pivot': False})
SUCCESS               0.261568            0.729184               0.359          ([torch.Size([128, 128])], {'pivot': True})
SUCCESS               0.291440            0.279744               1.042          ([torch.Size([128, 128])], {'pivot': False})
SUCCESS               0.647136            1.605216               0.403          ([torch.Size([256, 256])], {'pivot': True})
SUCCESS               0.703360            1.191024               0.591          ([torch.Size([256, 256])], {'pivot': False})
SUCCESS               9.551360            4.537216               2.105          ([torch.Size([1024, 512])], {'pivot': True})
SUCCESS               3.759536            1.695520               2.217          ([torch.Size([1024, 512])], {'pivot': False})
SUCCESS               0.129872            0.030464               4.263          ([torch.Size([32, 16])], {'pivot': True})
SUCCESS               0.076320            0.020864               3.658          ([torch.Size([32, 16])], {'pivot': False})
SUCCESS               0.129696            0.036640               3.540          ([torch.Size([16, 32])], {'pivot': True})
SUCCESS               0.084368            0.024736               3.411          ([torch.Size([16, 32])], {'pivot': False})
SUCCESS               0.388000            0.237184               1.636          ([torch.Size([128, 64])], {'pivot': True})
SUCCESS               0.166528            0.160096               1.040          ([torch.Size([128, 64])], {'pivot': False})
SUCCESS               0.423264            0.282880               1.496          ([torch.Size([64, 128])], {'pivot': True})
SUCCESS               0.216224            0.177024               1.221          ([torch.Size([64, 128])], {'pivot': False})
SUCCESS               0.078624            0.058016               1.355          ([torch.Size([4, 32, 32])], {'pivot': True})
SUCCESS               0.062784            0.037920               1.656          ([torch.Size([4, 32, 32])], {'pivot': False})
SUCCESS               0.061088            0.032512               1.879          ([torch.Size([128, 16, 16])], {'pivot': True})
SUCCESS               0.058272            0.022528               2.587          ([torch.Size([128, 16, 16])], {'pivot': False})
SUCCESS             191.890244           38.351360               5.003          ([torch.Size([1024, 512, 512])], {'pivot': True})
SUCCESS             171.316925           30.698463               5.581          ([torch.Size([1024, 512, 512])], {'pivot': False})
SUCCESS             751.502991          148.350876               5.066          ([torch.Size([4096, 512, 512])], {'pivot': True})
SUCCESS             669.608276          119.028770               5.626          ([torch.Size([4096, 512, 512])], {'pivot': False})


Operator: linalg_lu_factor_out  Performance Test (dtype=torch.float32, mode=kernel,level=comprehensive)
Status       Torch Latency (ms)    Gems Latency (ms)         Gems Speedup          Size Detail
-----------------------------------------------------------------------------------------------
SUCCESS               0.061440            0.021760               2.824          ([torch.Size([16, 16])], {'pivot': True, 'out': [torch.Size([16, 16]), torch.Size([16])]})
SUCCESS               0.073440            0.013696               5.362          ([torch.Size([16, 16])], {'pivot': False, 'out': [torch.Size([16, 16]), torch.Size([16])]})
SUCCESS               0.083520            0.050208               1.663          ([torch.Size([32, 32])], {'pivot': True, 'out': [torch.Size([32, 32]), torch.Size([32])]})
SUCCESS               0.073120            0.029536               2.476          ([torch.Size([32, 32])], {'pivot': False, 'out': [torch.Size([32, 32]), torch.Size([32])]})
SUCCESS               0.125184            0.154720               0.809          ([torch.Size([64, 64])], {'pivot': True, 'out': [torch.Size([64, 64]), torch.Size([64])]})
SUCCESS               0.072992            0.086368               0.845          ([torch.Size([64, 64])], {'pivot': False, 'out': [torch.Size([64, 64]), torch.Size([64])]})
SUCCESS               0.215104            0.579808               0.371          ([torch.Size([128, 128])], {'pivot': True, 'out': [torch.Size([128, 128]), torch.Size([128])]})
SUCCESS               0.097888            0.134176               0.730          ([torch.Size([128, 128])], {'pivot': False, 'out': [torch.Size([128, 128]), torch.Size([128])]})
SUCCESS               0.405440            1.527488               0.265          ([torch.Size([256, 256])], {'pivot': True, 'out': [torch.Size([256, 256]), torch.Size([256])]})
SUCCESS               0.139648            1.032560               0.135          ([torch.Size([256, 256])], {'pivot': False, 'out': [torch.Size([256, 256]), torch.Size([256])]})
SUCCESS               7.353504            2.930576               2.509          ([torch.Size([1024, 512])], {'pivot': True, 'out': [torch.Size([1024, 512]), torch.Size([512])]})
SUCCESS               1.842512            1.619568               1.138          ([torch.Size([1024, 512])], {'pivot': False, 'out': [torch.Size([1024, 512]), torch.Size([512])]})
SUCCESS               0.109184            0.025536               4.276          ([torch.Size([32, 16])], {'pivot': True, 'out': [torch.Size([32, 16]), torch.Size([16])]})
SUCCESS               0.074496            0.017792               4.187          ([torch.Size([32, 16])], {'pivot': False, 'out': [torch.Size([32, 16]), torch.Size([16])]})
SUCCESS               0.116928            0.025472               4.590          ([torch.Size([16, 32])], {'pivot': True, 'out': [torch.Size([16, 32]), torch.Size([16])]})
SUCCESS               0.078176            0.015968               4.896          ([torch.Size([16, 32])], {'pivot': False, 'out': [torch.Size([16, 32]), torch.Size([16])]})
SUCCESS               0.321904            0.170880               1.884          ([torch.Size([128, 64])], {'pivot': True, 'out': [torch.Size([128, 64]), torch.Size([64])]})
SUCCESS               0.127680            0.109760               1.163          ([torch.Size([128, 64])], {'pivot': False, 'out': [torch.Size([128, 64]), torch.Size([64])]})
SUCCESS               0.361056            0.256352               1.408          ([torch.Size([64, 128])], {'pivot': True, 'out': [torch.Size([64, 128]), torch.Size([64])]})
SUCCESS               0.175536            0.110240               1.592          ([torch.Size([64, 128])], {'pivot': False, 'out': [torch.Size([64, 128]), torch.Size([64])]})
SUCCESS               0.064592            0.050016               1.291          ([torch.Size([4, 32, 32])], {'pivot': True, 'out': [torch.Size([4, 32, 32]), torch.Size([4, 32])]})
SUCCESS               0.054816            0.029632               1.850          ([torch.Size([4, 32, 32])], {'pivot': False, 'out': [torch.Size([4, 32, 32]), torch.Size([4, 32])]})
SUCCESS               0.048768            0.023136               2.108          ([torch.Size([128, 16, 16])], {'pivot': True, 'out': [torch.Size([128, 16, 16]), torch.Size([128, 16])]})
SUCCESS               0.046656            0.014624               3.190          ([torch.Size([128, 16, 16])], {'pivot': False, 'out': [torch.Size([128, 16, 16]), torch.Size([128, 16])]})
SUCCESS              38.076351           37.257504               1.022          ([torch.Size([1024, 512, 512])], {'pivot': True, 'out': [torch.Size([1024, 512, 512]), torch.Size([1024, 512])]})
SUCCESS              20.318992           26.447680               0.768          ([torch.Size([1024, 512, 512])], {'pivot': False, 'out': [torch.Size([1024, 512, 512]), torch.Size([1024, 512])]})
SUCCESS             135.640930          143.988678               0.942          ([torch.Size([4096, 512, 512])], {'pivot': True, 'out': [torch.Size([4096, 512, 512]), torch.Size([4096, 512])]})
SUCCESS              75.481506          103.056580               0.732          ([torch.Size([4096, 512, 512])], {'pivot': False, 'out': [torch.Size([4096, 512, 512]), torch.Size([4096, 512])]})

Operator: linalg_lu_factor_out  Performance Test (dtype=torch.float64, mode=kernel,level=comprehensive)
Status       Torch Latency (ms)    Gems Latency (ms)         Gems Speedup          Size Detail
-----------------------------------------------------------------------------------------------
SUCCESS               0.068032            0.096512               0.705          ([torch.Size([16, 16])], {'pivot': True, 'out': [torch.Size([16, 16]), torch.Size([16])]})
SUCCESS               0.098112            0.088672               1.106          ([torch.Size([16, 16])], {'pivot': False, 'out': [torch.Size([16, 16]), torch.Size([16])]})
SUCCESS               0.092448            0.109280               0.846          ([torch.Size([32, 32])], {'pivot': True, 'out': [torch.Size([32, 32]), torch.Size([32])]})
SUCCESS               0.105888            0.105376               1.005          ([torch.Size([32, 32])], {'pivot': False, 'out': [torch.Size([32, 32]), torch.Size([32])]})
SUCCESS               0.149472            0.172576               0.866          ([torch.Size([64, 64])], {'pivot': True, 'out': [torch.Size([64, 64]), torch.Size([64])]})
SUCCESS               0.114208            0.128576               0.888          ([torch.Size([64, 64])], {'pivot': False, 'out': [torch.Size([64, 64]), torch.Size([64])]})
SUCCESS               0.264032            0.729216               0.362          ([torch.Size([128, 128])], {'pivot': True, 'out': [torch.Size([128, 128]), torch.Size([128])]})
SUCCESS               0.287456            0.280192               1.026          ([torch.Size([128, 128])], {'pivot': False, 'out': [torch.Size([128, 128]), torch.Size([128])]})
SUCCESS               0.649664            1.600832               0.406          ([torch.Size([256, 256])], {'pivot': True, 'out': [torch.Size([256, 256]), torch.Size([256])]})
SUCCESS               0.705216            1.190208               0.593          ([torch.Size([256, 256])], {'pivot': False, 'out': [torch.Size([256, 256]), torch.Size([256])]})
SUCCESS               9.250032            4.519264               2.047          ([torch.Size([1024, 512])], {'pivot': True, 'out': [torch.Size([1024, 512]), torch.Size([512])]})
SUCCESS               3.758384            1.586720               2.369          ([torch.Size([1024, 512])], {'pivot': False, 'out': [torch.Size([1024, 512]), torch.Size([512])]})
SUCCESS               0.121536            0.098560               1.233          ([torch.Size([32, 16])], {'pivot': True, 'out': [torch.Size([32, 16]), torch.Size([16])]})
SUCCESS               0.079360            0.089728               0.884          ([torch.Size([32, 16])], {'pivot': False, 'out': [torch.Size([32, 16]), torch.Size([16])]})
SUCCESS               0.132512            0.105264               1.259          ([torch.Size([16, 32])], {'pivot': True, 'out': [torch.Size([16, 32]), torch.Size([16])]})
SUCCESS               0.088544            0.092096               0.961          ([torch.Size([16, 32])], {'pivot': False, 'out': [torch.Size([16, 32]), torch.Size([16])]})
SUCCESS               0.386208            0.236736               1.631          ([torch.Size([128, 64])], {'pivot': True, 'out': [torch.Size([128, 64]), torch.Size([64])]})
SUCCESS               0.213312            0.159936               1.334          ([torch.Size([128, 64])], {'pivot': False, 'out': [torch.Size([128, 64]), torch.Size([64])]})
SUCCESS               0.420032            0.283184               1.483          ([torch.Size([64, 128])], {'pivot': True, 'out': [torch.Size([64, 128]), torch.Size([64])]})
SUCCESS               0.217696            0.176960               1.230          ([torch.Size([64, 128])], {'pivot': False, 'out': [torch.Size([64, 128]), torch.Size([64])]})
SUCCESS               0.080128            0.116608               0.687          ([torch.Size([4, 32, 32])], {'pivot': True, 'out': [torch.Size([4, 32, 32]), torch.Size([4, 32])]})
SUCCESS               0.064160            0.111680               0.574          ([torch.Size([4, 32, 32])], {'pivot': False, 'out': [torch.Size([4, 32, 32]), torch.Size([4, 32])]})
SUCCESS               0.064768            0.108400               0.597          ([torch.Size([128, 16, 16])], {'pivot': True, 'out': [torch.Size([128, 16, 16]), torch.Size([128, 16])]})
SUCCESS               0.055872            0.096192               0.581          ([torch.Size([128, 16, 16])], {'pivot': False, 'out': [torch.Size([128, 16, 16]), torch.Size([128, 16])]})
SUCCESS             194.085022           38.248032               5.074          ([torch.Size([1024, 512, 512])], {'pivot': True, 'out': [torch.Size([1024, 512, 512]), torch.Size([1024, 512])]})
SUCCESS             171.838211           30.832191               5.573          ([torch.Size([1024, 512, 512])], {'pivot': False, 'out': [torch.Size([1024, 512, 512]), torch.Size([1024, 512])]})
SUCCESS             760.495239          148.134521               5.134          ([torch.Size([4096, 512, 512])], {'pivot': True, 'out': [torch.Size([4096, 512, 512]), torch.Size([4096, 512])]})
SUCCESS             678.771606          118.811234               5.713          ([torch.Size([4096, 512, 512])], {'pivot': False, 'out': [torch.Size([4096, 512, 512]), torch.Size([4096, 512])]})

CLAassistant · 2026-07-24T03:58:36Z

All committers have signed the CLA.

Caeruleann · 2026-08-01T08:28:08Z

/test | linalg_lu_factor:thead

Caeruleann · 2026-08-01T08:28:17Z

/test | linalg_lu_factor:hygon

Caeruleann · 2026-08-01T08:28:29Z

/test | linalg_lu_factor:iluvatar

Caeruleann · 2026-08-01T08:28:46Z

/test | linalg_lu_factor:metax

github-actions · 2026-08-01T08:36:37Z

Test Environment

Env	Setting	Env	Setting
Time	`2026-08-01 16:36:28`	FlagGems	`5.3.2.post1.dev64+g58f1403a2`
Arch	`x86_64`	OS	`ubuntu 22.04`
Python	`3.10.20`	Torch	`2.9.0`
FlagTree	`0.5.1+hcu3.1`	Triton	`3.1.0`
Vendor	`hygon`	Device	`cuda`

Test Result

Accuracy Result

Summary

Status	Duration	Total	Passed	Skipped	Failed
Passed	87.42 sec	10	10	0	0

Performance Result

Summary

Status	Duration	Reason
Passed	319.11 seconds	All tests completed.

Benchmark data

Dtype	Speedup	Base	Gems	Shape
fp32	6.255	-	-	-
	1.997	0.114	0.057	`[[[16,16]],{'pivot':True}]`
	3.139	0.105	0.033	`[[[16,16]],{'pivot':False}]`
	1.338	0.183	0.137	`[[[32,32]],{'pivot':True}]`
	1.519	0.121	0.080	`[[[32,32]],{'pivot':False}]`
	0.833	0.310	0.372	`[[[64,64]],{'pivot':True}]`
	0.941	0.208	0.221	`[[[64,64]],{'pivot':False}]`
	0.659	0.627	0.952	`[[[128,128]],{'pivot':True}]`
	0.768	0.416	0.541	`[[[128,128]],{'pivot':False}]`
	0.632	1.311	2.074	`[[[256,256]],{'pivot':True}]`
	0.802	0.762	0.951	`[[[256,256]],{'pivot':False}]`
	0.681	3.890	5.711	`[[[1024,512]],{'pivot':True}]`
	0.738	2.157	2.921	`[[[1024,512]],{'pivot':False}]`
	1.942	0.123	0.063	`[[[32,16]],{'pivot':True}]`
	2.920	0.105	0.036	`[[[32,16]],{'pivot':False}]`
	1.776	0.111	0.062	`[[[16,32]],{'pivot':True}]`
	2.676	0.095	0.036	`[[[16,32]],{'pivot':False}]`
	0.894	0.362	0.405	`[[[128,64]],{'pivot':True}]`
	1.131	0.245	0.217	`[[[128,64]],{'pivot':False}]`
	1.658	0.873	0.527	`[[[64,128]],{'pivot':True}]`
	1.861	0.553	0.297	`[[[64,128]],{'pivot':False}]`
	0.950	0.132	0.138	`[[[4,32,32]],{'pivot':True}]`
	1.289	0.104	0.080	`[[[4,32,32]],{'pivot':False}]`
	1.954	0.116	0.059	`[[[128,16,16]],{'pivot':True}]`
	3.298	0.110	0.033	`[[[128,16,16]],{'pivot':False}]`
	36.919	2805.690	75.996	`[[[1024,512,512]],{'pivot':True}]`
	29.793	1488.760	49.970	`[[[1024,512,512]],{'pivot':False}]`
	39.685	11218.884	282.699	`[[[4096,512,512]],{'pivot':True}]`
	32.340	5956.508	184.182	`[[[4096,512,512]],{'pivot':False}]`

📎 Download the full log here

github-actions · 2026-08-01T08:36:54Z

Test Environment

Env	Setting	Env	Setting
Time	`2026-08-01 16:36:46`	FlagGems	`5.3.2.post1.dev59+g58f1403a2`
Arch	`x86_64`	OS	`ubuntu 24.04`
Python	`3.12.13`	Torch	`2.9.0`
FlagTree	`None`	Triton	`3.5.0`
Vendor	`thead`	Device	`cuda`

Test Result

Accuracy Result

Summary

Status	Duration	Total	Passed	Skipped	Failed
Passed	125.22 sec	10	10	0	0

Performance Result

Summary

Status	Duration	Reason
Passed	335.95 seconds	All tests completed.

Benchmark data

Dtype	Speedup	Base	Gems	Shape
fp32	7.493	-	-	-
	1.683	0.038	0.023	`[[[16,16]],{'pivot':True}]`
	2.934	0.036	0.012	`[[[16,16]],{'pivot':False}]`
	1.169	0.069	0.059	`[[[32,32]],{'pivot':True}]`
	1.213	0.039	0.032	`[[[32,32]],{'pivot':False}]`
	0.622	0.140	0.225	`[[[64,64]],{'pivot':True}]`
	0.495	0.063	0.128	`[[[64,64]],{'pivot':False}]`
	0.419	0.298	0.712	`[[[128,128]],{'pivot':True}]`
	0.292	0.130	0.444	`[[[128,128]],{'pivot':False}]`
	0.459	0.648	1.413	`[[[256,256]],{'pivot':True}]`
	0.603	0.479	0.795	`[[[256,256]],{'pivot':False}]`
	0.513	2.179	4.245	`[[[1024,512]],{'pivot':True}]`
	2.547	5.916	2.323	`[[[1024,512]],{'pivot':False}]`
	1.482	0.042	0.028	`[[[32,16]],{'pivot':True}]`
	2.387	0.036	0.015	`[[[32,16]],{'pivot':False}]`
	1.619	0.046	0.028	`[[[16,32]],{'pivot':True}]`
	2.368	0.036	0.015	`[[[16,32]],{'pivot':False}]`
	0.630	0.179	0.284	`[[[128,64]],{'pivot':True}]`
	0.793	0.129	0.163	`[[[128,64]],{'pivot':False}]`
	0.433	0.145	0.334	`[[[64,128]],{'pivot':True}]`
	0.356	0.067	0.187	`[[[64,128]],{'pivot':False}]`
	1.259	0.077	0.061	`[[[4,32,32]],{'pivot':True}]`
	1.282	0.043	0.033	`[[[4,32,32]],{'pivot':False}]`
	2.076	0.047	0.023	`[[[128,16,16]],{'pivot':True}]`
	3.303	0.041	0.012	`[[[128,16,16]],{'pivot':False}]`
	31.767	1496.585	47.111	`[[[1024,512,512]],{'pivot':True}]`
	54.226	2013.840	37.138	`[[[1024,512,512]],{'pivot':False}]`
	33.881	5985.705	176.669	`[[[4096,512,512]],{'pivot':True}]`
	58.990	8056.931	136.581	`[[[4096,512,512]],{'pivot':False}]`

📎 Download the full log here

github-actions · 2026-08-01T08:41:48Z

Test Environment

Env	Setting	Env	Setting
Time	`2026-08-01 16:41:34`	FlagGems	`5.3.2.post1.dev59+g58f1403a2`
Arch	`x86_64`	OS	`ubuntu 22.04`
Python	`3.12.13`	Torch	`2.8.0+metax3.7.2.0`
FlagTree	`3.1.0+metax3.7.2.0`	Triton	`3.1.0`
Vendor	`metax`	Device	`cuda`

Test Result

Accuracy Result

Summary

Status	Duration	Total	Passed	Skipped	Failed
Passed	90.02 sec	10	10	0	0

Performance Result

Summary

Status	Duration	Reason
Timeout	601.22 seconds	All tests completed.

Benchmark data

No performance data collected!

📎 Download the full log here

YangLong114514 added 2 commits July 24, 2026 03:37

[KMCompiler] add operator linalg_lu_factor

cfcbd02

[KMCompiler] update tests

18f4c2e

YangLong114514 requested review from 0x45f, 103yiran, Caeruleann, bin913, douxetpur, hellojack163, huangyiqun, tengqm and w1120029931-bit as code owners July 24, 2026 03:58

github-actions Bot added benchmark ops/aten core tests size/XLarge labels Jul 24, 2026

This comment was marked as low quality.

Sign in to view

Caeruleann and others added 6 commits July 24, 2026 14:27

Merge branch 'master' into KMCompiler-linalg_lu_factor

0fa7055

[KMcompiler] add linalg_lu_factor_out

7117c65

Merge branch 'master' into KMCompiler-linalg_lu_factor

94f8f42

[KMCompiler] update test_linalg_lu_factor.py

05ab002

[KMCompiler] update ref cpu

99700c2

Merge branch 'master' into KMCompiler-linalg_lu_factor

e12a813

This comment was marked as outdated.

Sign in to view

YangLong114514 added 4 commits July 30, 2026 03:57

update benchmark

8eb9259

open out

3250689

Merge branch 'master' into KMCompiler-linalg_lu_factor

5f98158

update factor_out

202d5dc

This comment was marked as low quality.

Sign in to view

update benchmark

89d7435

This comment was marked as duplicate.

Sign in to view

This comment was marked as outdated.

Sign in to view

This comment was marked as duplicate.

Sign in to view

This comment was marked as outdated.

Sign in to view

YangLong114514 added 2 commits July 31, 2026 11:13

Merge branch 'master' into KMCompiler-linalg_lu_factor

bc37ebe

iluvatar update

7865723

github-actions Bot added the vendor/Iluvatar label Aug 1, 2026

YangLong114514 added 3 commits August 1, 2026 15:12

Merge branch 'master' into KMCompiler-linalg_lu_factor

e64f247

iluvatar update

db27ae5

delete iluvatar

58f1403

github-actions Bot removed the vendor/Iluvatar label Aug 1, 2026

This comment was marked as low quality.

Sign in to view

Uh oh!

Conversation

YangLong114514 commented Jul 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Category

Type of Change

Description

Issue

Progress

Performance

Uh oh!

CLAassistant commented Jul 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

This comment was marked as low quality.

This comment was marked as low quality.

This comment was marked as outdated.

This comment was marked as outdated.

This comment was marked as outdated.

This comment was marked as outdated.

This comment was marked as outdated.

This comment was marked as low quality.

This comment was marked as low quality.

This comment was marked as duplicate.

This comment was marked as duplicate.

This comment was marked as duplicate.

This comment was marked as duplicate.

This comment was marked as outdated.

This comment was marked as outdated.

This comment was marked as duplicate.

This comment was marked as outdated.

Caeruleann commented Aug 1, 2026

Uh oh!

Caeruleann commented Aug 1, 2026

Uh oh!

Caeruleann commented Aug 1, 2026

Uh oh!

Caeruleann commented Aug 1, 2026

Uh oh!

This comment was marked as low quality.

github-actions Bot commented Aug 1, 2026

Test Environment

Test Result

Accuracy Result

Summary

Performance Result

Summary

Benchmark data

Uh oh!

github-actions Bot commented Aug 1, 2026

Test Environment

Test Result

Accuracy Result

Summary

Performance Result

Summary

Benchmark data

Uh oh!

github-actions Bot commented Aug 1, 2026

Test Environment

Test Result

Accuracy Result

Summary

Performance Result

Summary

Benchmark data

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

YangLong114514 commented Jul 24, 2026 •

edited

Loading

CLAassistant commented Jul 24, 2026 •

edited

Loading