jgrusewski 833cc07d0e fix: training profile must not override GPU profile batch_size ...

dqn-production.toml batch_size=8192 was overriding GPU profile batch_size=64
on RTX 3050 via apply_to(), causing OOM/hang. GPU profile is the sole
authority for batch_size — training profile controls learning params only.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-04-02 23:42:12 +02:00

..

a100.toml

fix: restore batch_size to GPU profiles — profile defaults, not auto-scaling

2026-04-02 21:26:24 +02:00

default.toml

fix: restore batch_size to GPU profiles — profile defaults, not auto-scaling

2026-04-02 21:26:24 +02:00

h100.toml

fix: restore batch_size to GPU profiles — profile defaults, not auto-scaling

2026-04-02 21:26:24 +02:00

rtx3050.toml

fix: training profile must not override GPU profile batch_size

2026-04-02 23:42:12 +02:00