Skip to content

Actions: kaelvalen/prism

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
37 workflow runs
37 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
Update docs with ablation results: shared_expert keeps utilization, b… CI #37: Commit 2f1130a pushed by kaelvalen
49s main
main
49s
Add MoM spike status section: gate results, lbal0 collapse finding, C… CI #36: Commit d62531b pushed by kaelvalen
57s main
main
57s
Cap eval scan elements directly (batch × chunk ≤ 512) — robust on 8GB… CI #35: Commit 2aee010 pushed by kaelvalen
1m 4s main
main
1m 4s
Shrink eval chunks to 256 beyond 1k context + empty CUDA cache before… CI #34: Commit f73cf54 pushed by kaelvalen
52s main
main
52s
Merge branch 'main' of github.com:kaelvalen/prism CI #33: Commit bbd2f12 pushed by kaelvalen
44s main
main
44s
Update README.md CI #32: Commit ec14901 pushed by kaelvalen
1m 13s main
main
1m 13s
Update README.md CI #31: Commit a240161 pushed by kaelvalen
58s main
main
58s
Add gradient_checkpointing support CI #30: Commit 19dfc1f pushed by kaelvalen
1m 0s main
main
1m 0s
Update RTX5090 benchmark to use batch size 32 CI #29: Commit 60d5783 pushed by kaelvalen
1m 4s main
main
1m 4s
Add modality support to baselines and benchmarks CI #28: Commit 626e9b9 pushed by kaelvalen
1m 4s main
main
1m 4s
Enable multi-label ECG training with BCE loss CI #27: Commit 1f363c4 pushed by kaelvalen
1m 9s main
main
1m 9s
Refactor Hillis-Steele recurrence for in-place updates CI #26: Commit da8c364 pushed by kaelvalen
1m 7s main
main
1m 7s
Update docs and config examples for per-layer tokens and inference CI #25: Commit 0c34133 pushed by kaelvalen
1m 8s main
main
1m 8s
Require CUDA and non-float32 for FLA delta backend CI #24: Commit ba13862 pushed by kaelvalen
1m 8s main
main
1m 8s
Add gccLib and update LD_LIBRARY_PATH CI #23: Commit ebd7000 pushed by kaelvalen
1m 6s main
main
1m 6s
Enable lazy import of PRISMForClassification CI #22: Commit 8e554f7 pushed by kaelvalen
1m 11s main
main
1m 11s
Export block module symbols CI #21: Commit 8a3ec00 pushed by kaelvalen
1m 7s main
main
1m 7s
Add flake.lock CI #20: Commit 012d85e pushed by kaelvalen
1m 4s main
main
1m 4s
Add RNG utilities for PyTorch NumPy Python CI #19: Commit a450f57 pushed by kaelvalen
1m 3s main
main
1m 3s
Add multi-label PTB-XL support and tests CI #18: Commit f2da258 pushed by kaelvalen
50s main
main
50s
refactor: clean up whitespace in trainer.py for improved readability CI #17: Commit 51137d0 pushed by kaelvalen
52s main
main
52s
refactor: improve configuration validation and enhance data loading l… CI #16: Commit 8c99164 pushed by kaelvalen
43s main
main
43s
refactor: streamline code formatting in delta module and tests CI #15: Commit 18c36bb pushed by kaelvalen
42s main
main
42s
chore: add uv.lock to .gitignore to prevent tracking of lock file CI #14: Commit b12ba62 pushed by kaelvalen
50s main
main
50s
chore: remove uv.lock file to streamline dependency management CI #13: Commit 25021d9 pushed by kaelvalen
46s main
main
46s