Skip to content

Actions: pytorch/ao

Actions

Run Regression Tests on ROCm

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
regression_test_rocm.yml
1,105 workflow runs
1,105 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

move float8 blockwise kernels out of prototype (#3256) Run Regression Tests on ROCm #1105: Commit 1e473ed pushed by vkuzo
1h 5m 45s main
main
1h 5m 45s
properly skip float8 inference tests without fbgemm (#3255) Run Regression Tests on ROCm #1104: Commit d3bec87 pushed by vkuzo
1h 10m 43s main
main
1h 10m 43s
[mxfp8 moe training] update readme and tests (#3260) Run Regression Tests on ROCm #1103: Commit 0f05b40 pushed by danielvegamyhre
1h 12m 52s main
main
1h 12m 52s
[CPU] Improve INT8 SDPA template (#3230) Run Regression Tests on ROCm #1102: Commit 3577306 pushed by Xia-Weiwen
1h 15m 34s main
main
1h 15m 34s
[mxfp8 moe training] initialize zero tensor differently to avoid d2h … Run Regression Tests on ROCm #1101: Commit d84f5b8 pushed by danielvegamyhre
1h 14m 25s main
main
1h 14m 25s
Add quantize_ nn.Parameter support (#3083) Run Regression Tests on ROCm #1100: Commit 50a555b pushed by jcaip
1h 12m 59s main
main
1h 12m 59s
[mxfp8 moe training] bench and profile mxfp8 a2a fwd and bwd separate… Run Regression Tests on ROCm #1099: Commit dffb3a0 pushed by danielvegamyhre
1h 10m 29s main
main
1h 10m 29s
[mxfp8 moe training] simplify e8m0 -> fp32 calc (#3201) Run Regression Tests on ROCm #1098: Commit d089c6a pushed by danielvegamyhre
1h 11m 47s main
main
1h 11m 47s
[mxfp8 moe training] integrate triton quant/dequant kernels into mxfp… Run Regression Tests on ROCm #1097: Commit 6a62549 pushed by danielvegamyhre
1h 21m 56s main
main
1h 21m 56s
[mxfp8 moe training] add triton kernel for mxfp8 dequantization (#3195) Run Regression Tests on ROCm #1096: Commit 7537d99 pushed by danielvegamyhre
1h 25m 5s main
main
1h 25m 5s
create a separate test for mx and nv serialization (#3251) Run Regression Tests on ROCm #1095: Commit 41a0778 pushed by vkuzo
1h 13m 50s main
main
1h 13m 50s
nvfp4: support inference_mode and rank 3 (#3240) Run Regression Tests on ROCm #1094: Commit 03c2d28 pushed by vkuzo
1h 13m 49s main
main
1h 13m 49s
mx: support inference_mode and rank 3+ (#3238) Run Regression Tests on ROCm #1093: Commit 14eff10 pushed by vkuzo
1h 24m 7s main
main
1h 24m 7s
Only convert to int4 preshuffled tensor in H100 (#3245) Run Regression Tests on ROCm #1092: Commit f303f4c pushed by jerryzh168
1h 8m 22s main
main
1h 8m 22s
[moe training] update readme with links, cleanup (#3239) Run Regression Tests on ROCm #1091: Commit e9c7bea pushed by danielvegamyhre
1h 17m 30s main
main
1h 17m 30s
Fix Wq size check Run Regression Tests on ROCm #1090: Commit 53b5efd pushed by meta-codesync bot
1h 17m 26s main
main
1h 17m 26s
docs: fix qat description in README.md (#3212) Run Regression Tests on ROCm #1089: Commit 204cd48 pushed by andrewor14
1h 15m 6s main
main
1h 15m 6s
Add Unsloth + QAT blog to latest news (#3227) Run Regression Tests on ROCm #1088: Commit f3fc5e7 pushed by andrewor14
58m 13s main
main
58m 13s
Update README.md (#3225) Run Regression Tests on ROCm #1087: Commit 54a7b85 pushed by jerryzh168
52m 42s main
main
52m 42s
[mxfp8 moe training] make compile vs triton for dim0 cast configurabl… Run Regression Tests on ROCm #1086: Commit 7e68d5e pushed by danielvegamyhre
55m 14s main
main
55m 14s
enable custom MKN in inference roofline script (#3224) Run Regression Tests on ROCm #1085: Commit beee153 pushed by vkuzo
53m 51s main
main
53m 51s
mxfp8 inference roofline: add fusion to observed (#3223) Run Regression Tests on ROCm #1084: Commit 6452b4a pushed by vkuzo
1h 8m 57s main
main
1h 8m 57s
fix mxfp8 matmul benchmark (#3221) Run Regression Tests on ROCm #1083: Commit 13595c5 pushed by vkuzo
1h 8m 0s main
main
1h 8m 0s
[moe training] change api _scaled_grouped_mm -> _quantize_then_scaled… Run Regression Tests on ROCm #1082: Commit 82ae011 pushed by danielvegamyhre
1h 1m 22s main
main
1h 1m 22s
fix is metadata func (#3220) Run Regression Tests on ROCm #1081: Commit bd69a73 pushed by jerryzh168
54m 42s main
main
54m 42s