Skip to content

Actions: MatN23/AdaptiveTrainingSystem

Actions

Code Test Suite

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
test.yml
236 workflow runs
236 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
Align vocab_size default; reset tokenizer default Code Test Suite #236: Commit dbdbe34 pushed by MatN23
2m 31s main
main
2m 31s
Update trainer.py Code Test Suite #235: Commit 6d7e127 pushed by MatN23
2m 56s main
main
2m 56s
Optimize CUDA kernels, MoE routing, and training utils Code Test Suite #234: Commit c9e58ed pushed by MatN23
2m 41s main
main
2m 41s
Update dataset.py Code Test Suite #233: Commit b7c3107 pushed by MatN23
2m 35s main
main
2m 35s
Add robust loss/grad wrappers and dataset filtering Code Test Suite #232: Commit be236a4 pushed by MatN23
2m 37s main
main
2m 37s
Add KV-cache inference and MoE/tokenizer fixes Code Test Suite #231: Commit bfc7150 pushed by MatN23
2m 30s main
main
2m 30s
Update trainer.py Code Test Suite #230: Commit 2a186df pushed by MatN23
6m 10s main
main
6m 10s
Make CUDA/MPS checks safe and bound histories Code Test Suite #229: Commit cad9f58 pushed by MatN23
2m 24s main
main
2m 24s
Improve OOM recovery and memory handling Code Test Suite #228: Commit d758e6d pushed by MatN23
1m 53s main
main
1m 53s
Refactor fused_grad_clip CUDA & Python API Code Test Suite #227: Commit 61aec1c pushed by MatN23
2m 10s main
main
2m 10s
Enforce CUDA master switch in Main.py Code Test Suite #226: Commit 5321252 pushed by MatN23
2m 26s main
main
2m 26s
Prevent CUDAGraphs overwrite; refactor grad clip Code Test Suite #225: Commit 4cc69d5 pushed by MatN23
2m 33s main
main
2m 33s
Update fused_grad_clip.cu Code Test Suite #224: Commit 2c2d104 pushed by MatN23
2m 38s main
main
2m 38s
Add CLI profiles and config to benchmark script Code Test Suite #223: Commit 8e5a04f pushed by MatN23
2m 33s main
main
2m 33s
Add MoE benchmark and improve CUDA build linking Code Test Suite #222: Commit 7f02c1c pushed by MatN23
2m 38s main
main
2m 38s
Add auto-build, arch handling and MoE/runtime toggles Code Test Suite #221: Commit f12ec34 pushed by MatN23
2m 40s main
main
2m 40s
Add Metal MoE import checks; improve test runner Code Test Suite #220: Commit 216c41c pushed by MatN23
2m 44s main
main
2m 44s
npm(deps-dev): bump electron-builder from 24.13.3 to 26.7.0 Code Test Suite #219: Pull request #38 opened by dependabot bot
2m 37s dependabot/npm_and_yarn/electron-builder-26.7.0
dependabot/npm_and_yarn/electron-builder-26.7.0
2m 37s
npm(deps-dev): bump electron from 28.3.3 to 40.1.0 Code Test Suite #218: Pull request #37 opened by dependabot bot
2m 45s dependabot/npm_and_yarn/electron-40.1.0
dependabot/npm_and_yarn/electron-40.1.0
2m 45s
pip(deps): bump transformers from 4.36.2 to 5.0.0 Code Test Suite #217: Pull request #36 opened by dependabot bot
3m 16s dependabot/pip/transformers-5.0.0
dependabot/pip/transformers-5.0.0
3m 16s
Optimize CUDA SwiGLU and FusedMLPBlock kernels Code Test Suite #216: Commit 4749f59 pushed by MatN23
2m 38s main
main
2m 38s
Integrate cuBLASLt for fused MLP and optimize kernels Code Test Suite #215: Commit 871138d pushed by MatN23
2m 59s main
main
2m 59s
Delete test_triton_ops.py Code Test Suite #214: Commit 85dd588 pushed by MatN23
2m 54s main
main
2m 54s
Refactor fused MLP block to use new CUDA kernel Code Test Suite #213: Commit 120cd3a pushed by MatN23
1m 51s main
main
1m 51s
Remove Triton FP8 support and deprecated kernels Code Test Suite #212: Commit e59f08e pushed by MatN23
1m 53s main
main
1m 53s