Skip to content

Actions: QingGo/qwen35-ple

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
406 workflow runs
406 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
round169: correction -- E0 is the pretrained table, not an aligned ar… CI #406: Commit 224a4a8 pushed by QingGo
2m 9s main
main
2m 9s
round169 phases 9-10: the headline is memory (93.7%); the unseen regi… CI #405: Commit 2ff5106 pushed by QingGo
1m 43s main
main
1m 43s
playbook 2.13(d): a re-run that keeps the old markers, and a guard th… CI #404: Commit f99eb87 pushed by QingGo
1m 58s main
main
1m 58s
round169 phase 10 (retry): the row fill belongs after the constructio… CI #403: Commit e9af811 pushed by QingGo
1m 27s main
main
1m 27s
round169 phase 10: the shuffle result turns on the headline, not just… CI #402: Commit 4fed72c pushed by QingGo
1m 36s main
main
1m 36s
round169 phase 9: is the unseen-regime gain the row's content, or jus… CI #401: Commit 9c911c9 pushed by QingGo
1m 55s main
main
1m 55s
round169 phase 8: a quarter of the value was in a regime nobody had s… CI #400: Commit d7e0cda pushed by QingGo
2m 2s main
main
2m 2s
round169: the offline gate study, with the leak refused in code CI #399: Commit 06b48f0 pushed by QingGo
2m 7s main
main
2m 7s
round169 phase 8: record what a deployable gate needs, and the regime… CI #398: Commit a73f392 pushed by QingGo
1m 55s main
main
1m 55s
round169: the cross-row coupling, measured instead of asserted CI #397: Commit e317b52 pushed by QingGo
2m 11s main
main
2m 11s
round169: the top-1 gate says GAIN, and soft shrinkage beats the hard… CI #396: Commit 961368a pushed by QingGo
1m 58s main
main
1m 58s
round169 phase 7: re-run the gate, and the two guards phase 6 was mis… CI #395: Commit faf6d87 pushed by QingGo
1m 43s main
main
1m 43s
round169 phase 5: row selection holds, the band decomposition does not CI #394: Commit 2789cef pushed by QingGo
2m 4s main
main
2m 4s
round169: phase 6 -- the top-1 gate, then the A/B the derivation pred… CI #393: Commit cc41929 pushed by QingGo
2m 5s main
main
2m 5s
round170: derive the optimal step for a sparse embedding row, and fin… CI #392: Commit d7bf3c3 pushed by QingGo
1m 53s main
main
1m 53s
round169: the corrected band table predicts a positive arm, and phase… CI #391: Commit 7a7731b pushed by QingGo
1m 50s main
main
1m 50s
round169: the per-position record indexed per-trigram arrays with str… CI #390: Commit 4e3a483 pushed by QingGo
2m 9s main
main
2m 9s
round169: the ladder's frozen stopping rule was decided on a corrupte… CI #389: Commit 041fd45 pushed by QingGo
2m 5s main
main
2m 5s
round169: the capacity curve saturates, and two queue-script lessons CI #388: Commit 0e32f0e pushed by QingGo
2m 6s main
main
2m 6s
round169: the lr=1e-3 verdict was a diverged optimiser, and the band … CI #387: Commit 50a2c12 pushed by QingGo
1m 37s main
main
1m 37s
round169: rung 3 carries the trigram, and training the rows makes it … CI #386: Commit 4f2bfe4 pushed by QingGo
2m 7s main
main
2m 7s
round166: 2B/4B replication of the unsaturated regime, and the length… CI #385: Commit eefb3f2 pushed by QingGo
1m 15s main
main
1m 15s
round165B + paper: repairing the read-out does not make table content… CI #384: Commit 6e6b2f4 pushed by QingGo
1m 17s main
main
1m 17s
paper: compress main text 12 -> 8 pages by moving detail to the appendix CI #383: Commit 8fc8df3 pushed by QingGo
1m 11s main
main
1m 11s
paper: major revision responding to review round 1 CI #382: Commit e3774d4 pushed by QingGo
1m 14s main
main
1m 14s