Skip to content

Actions: lean-dojo/TorchLean

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
118 workflow runs
118 workflow runs

Filter by Workflow

⌥ + click/return to exclude

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
feat(cuda): TexTable — layered 1-D lookup-table textures with point/hardware filtering CI #9: Pull request #24 opened by NicolasRouquette
2h 3m 48s NicolasRouquette:cuda-texture-table
NicolasRouquette:cuda-texture-table
2h 3m 48s
feat(cuda): bound the device block reuse cache with an env byte cap CI #8: Pull request #10 synchronize by NicolasRouquette
2h 12m 46s NicolasRouquette:cuda-cache-byte-cap
NicolasRouquette:cuda-cache-byte-cap
2h 12m 46s
feat(cuda): fused scaledProdExp buffer op — exp((c·x)·y) in one kernel CI #7: Pull request #23 opened by NicolasRouquette
1h 54m 11s NicolasRouquette:cuda-scaled-prod-exp
NicolasRouquette:cuda-scaled-prod-exp
1h 54m 11s
feat(cuda): scoped device-memory arena (withCudaArena) + use-after-free detector CI #6: Pull request #22 opened by NicolasRouquette
1h 35m 44s NicolasRouquette:cuda-arena-scope
NicolasRouquette:cuda-arena-scope
1h 35m 44s
feat(cuda): bundle + auto-link the glibc≥2.38 isoc23 link shim CI #5: Pull request #18 synchronize by NicolasRouquette
2h 5m 3s NicolasRouquette:cuda-isoc23-shim
NicolasRouquette:cuda-isoc23-shim
2h 5m 3s
feat(cuda): bundle + auto-link the glibc≥2.38 isoc23 link shim Compile blueprint #58: Pull request #18 synchronize by NicolasRouquette
1h 3m 45s NicolasRouquette:cuda-isoc23-shim
NicolasRouquette:cuda-isoc23-shim
1h 3m 45s
Strengthen numerics, runtime validation, and documentation Compile blueprint #57: Commit 14bc499 pushed by Robertboy18
1h 39m 36s main
main
1h 39m 36s
Strengthen numerics, runtime validation, and documentation CI #4: Commit 14bc499 pushed by Robertboy18
2h 1m 7s main
main
2h 1m 7s
Refine architecture, proofs, and documentation CI #3: Commit 4db7a43 pushed by Robertboy18
2h 8m 45s main
main
2h 8m 45s
Refine architecture, proofs, and documentation Compile blueprint #56: Commit 4db7a43 pushed by Robertboy18
1h 41m 32s main
main
1h 41m 32s
Harden numerical semantics and verification Compile blueprint #54: Commit c2be276 pushed by Robertboy18
1h 32m 5s main
main
1h 32m 5s
Harden numerical semantics and verification CI #1: Commit c2be276 pushed by Robertboy18
1h 58m 26s main
main
1h 58m 26s
rewrite docs and delete dead code Compile blueprint #53: Pull request #20 opened by wadkisson
1h 31m 12s wadkisson:doc-fix
wadkisson:doc-fix
1h 31m 12s
feat(cuda): bundle + auto-link the glibc≥2.38 isoc23 link shim Compile blueprint #52: Pull request #18 synchronize by NicolasRouquette
1h 38m 13s NicolasRouquette:cuda-isoc23-shim
NicolasRouquette:cuda-isoc23-shim
1h 38m 13s
Major numerical semantics, backend, API, and documentation upgrade Compile blueprint #51: Commit 8c3a1e6 pushed by Robertboy18
3h 25m 3s main
main
3h 25m 3s
Make CUDA GPT-2 training defaults fast with LibTorch SDPA. Compile blueprint #50: Pull request #19 opened by wadkisson
1h 37m 47s wadkisson:cuda-train-defaults
wadkisson:cuda-train-defaults
1h 37m 47s
feat(cuda): bundle + auto-link the glibc≥2.38 isoc23 link shim Compile blueprint #49: Pull request #18 opened by NicolasRouquette
1h 33m 6s NicolasRouquette:cuda-isoc23-shim
NicolasRouquette:cuda-isoc23-shim
1h 33m 6s
Strengthen floating-point semantics and CUDA runtime Compile blueprint #48: Commit 0a7baaf pushed by Robertboy18
1h 33m 54s main
main
1h 33m 54s
Scale training runtime and harden CUDA ownership Compile blueprint #47: Commit 5f7d3ee pushed by Robertboy18
1h 23m 6s main
main
1h 23m 6s
Major upgrade to Lean 4.32 Compile blueprint #45: Commit 852c9c6 pushed by Robertboy18
1h 30m 23s main
main
1h 30m 23s
Add README citation entry Compile blueprint #44: Commit c1043f1 pushed by Robertboy18
1h 38m 17s main
main
1h 38m 17s
Swap in PyTorch kernels for MHA Compile blueprint #42: Pull request #13 synchronize by wadkisson
1h 36m 20s wadkisson:attention-pr
wadkisson:attention-pr
1h 36m 20s
Fix scientific ML docs source link Compile blueprint #40: Commit a0c5bdf pushed by Robertboy18
1h 35m 18s main
main
1h 35m 18s
Update TorchLean runtime, verification, and website Compile blueprint #39: Commit b0afc9d pushed by Robertboy18
2m 38s main
main
2m 38s
Add token-id LM training, CUDA defaults, and kernel launch hygiene Compile blueprint #38: Pull request #11 opened by wadkisson
1h 27m 25s wadkisson:main
wadkisson:main
1h 27m 25s