Skip to content

Actions: meta-pytorch/autoparallel

Actions

Test TorchTitan Integration

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
test_torchtitan.yml
1,535 workflow runs
1,535 workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Add doc_packed_attn: AutoParallel-shardable document-packed varlen attention Test TorchTitan Integration #1538: Pull request #508 opened by fmassa
1h 39m 54s fmassa/packed_doc_attention
fmassa/packed_doc_attention
1h 39m 54s
Emit autoparallel solution as a TORCH_TRACE artifact (#507) Test TorchTitan Integration #1537: Commit 14682dc pushed by fmassa
18m 22s main
main
18m 22s
Emit autoparallel solution as a TORCH_TRACE artifact Test TorchTitan Integration #1536: Pull request #507 opened by fmassa
16m 55s fmassa/torch_trace_solution
fmassa/torch_trace_solution
16m 55s
Topo-span-bounded FSDP bucketing for AutoParallel (#481) Test TorchTitan Integration #1535: Commit 7ad1142 pushed by fmassa
28m 4s main
main
28m 4s
Replay first-partitioner save decisions in torch.compile via _ReplayS… Test TorchTitan Integration #1534: Commit c21df74 pushed by fmassa
18m 58s main
main
18m 58s
Replay first-partitioner save decisions in torch.compile via _ReplaySavesPartitioner Test TorchTitan Integration #1533: Pull request #485 synchronize by fmassa
16m 5s fmassa/new_partitioner
fmassa/new_partitioner
16m 5s
Update to use CUDA 13.0 (#506) Test TorchTitan Integration #1532: Commit 39421a2 pushed by fmassa
29m 32s main
main
29m 32s
Replay first-partitioner save decisions in torch.compile via _ReplaySavesPartitioner Test TorchTitan Integration #1531: Pull request #485 synchronize by fmassa
17m 39s fmassa/new_partitioner
fmassa/new_partitioner
17m 39s
Replay first-partitioner save decisions in torch.compile via _ReplaySavesPartitioner Test TorchTitan Integration #1530: Pull request #485 synchronize by fmassa
7m 21s fmassa/new_partitioner
fmassa/new_partitioner
7m 21s
Update to use CUDA 13.0 Test TorchTitan Integration #1529: Pull request #506 opened by fmassa
13m 0s fmassa/update_ci
fmassa/update_ci
13m 0s
Replay first-partitioner save decisions in torch.compile via _ReplaySavesPartitioner Test TorchTitan Integration #1528: Pull request #485 synchronize by fmassa
17m 0s fmassa/new_partitioner
fmassa/new_partitioner
17m 0s
Topo-span-bounded FSDP bucketing for AutoParallel Test TorchTitan Integration #1527: Pull request #481 synchronize by fmassa
17m 45s fmassa/better_bucketing
fmassa/better_bucketing
17m 45s
Use upstream DTensor sharding strategies (#505) Test TorchTitan Integration #1526: Commit 9fa2781 pushed by aditvenk
16m 1s main
main
16m 1s
Use upstream DTensor sharding strategies Test TorchTitan Integration #1525: Pull request #505 synchronize by aditvenk
17m 37s aditvenk/stack/17
aditvenk/stack/17
17m 37s
Make AC recompute test topology-robust (#490) Test TorchTitan Integration #1524: Commit 1aff15c pushed by aditvenk
12m 41s main
main
12m 41s
Use upstream DTensor sharding strategies Test TorchTitan Integration #1523: Pull request #505 synchronize by aditvenk
19m 43s aditvenk/stack/17
aditvenk/stack/17
19m 43s
Use upstream DTensor sharding strategies Test TorchTitan Integration #1522: Pull request #505 opened by aditvenk
11s aditvenk/stack/17
aditvenk/stack/17
11s
Use upstream DTensor index strategy Test TorchTitan Integration #1521: Pull request #493 synchronize by aditvenk
21m 4s aditvenk/stack/5
aditvenk/stack/5
21m 4s
Use upstream DTensor scatter strategy Test TorchTitan Integration #1520: Pull request #502 synchronize by aditvenk
16m 25s aditvenk/stack/14
aditvenk/stack/14
16m 25s
Use upstream DTensor reshape strategy Test TorchTitan Integration #1519: Pull request #503 synchronize by aditvenk
16m 35s aditvenk/stack/15
aditvenk/stack/15
16m 35s
Use upstream DTensor convolution backward strategy Test TorchTitan Integration #1518: Pull request #501 synchronize by aditvenk
16m 17s aditvenk/stack/13
aditvenk/stack/13
16m 17s
Use upstream DTensor select scatter strategy Test TorchTitan Integration #1517: Pull request #497 synchronize by aditvenk
16m 31s aditvenk/stack/9
aditvenk/stack/9
16m 31s
Use upstream DTensor expand strategy Test TorchTitan Integration #1516: Pull request #504 synchronize by aditvenk
13m 35s aditvenk/stack/16
aditvenk/stack/16
13m 35s
Make AC recompute test topology-robust Test TorchTitan Integration #1515: Pull request #490 synchronize by aditvenk
23m 37s aditvenk/stack/2
aditvenk/stack/2
23m 37s
Use upstream DTensor convert strategy Test TorchTitan Integration #1514: Pull request #495 synchronize by aditvenk
20m 12s aditvenk/stack/7
aditvenk/stack/7
20m 12s