Satyam-53 8d4e05acef Remove CUDA to CPU and vice versa conversion ...

And more network payload logging

2026-09-07 18:49:49 -07:00

..

test_gpu_worker.py

[perf] Disable every offload path on unified memory, unblocking MiniMax H3 generation on one GB10 (#1715)

2026-08-26 15:32:56 -07:00

test_minimax_h3_disaggregated.py

Remove CUDA to CPU and vice versa conversion

2026-09-07 18:49:49 -07:00

test_multiproc_executor.py

[bugfix] Harden OpenAI serving after post-merge review (#1782)

2026-08-28 22:29:24 -07:00

test_ray_distributed_executor.py

[feat]: run FastH3 across two DGX Sparks with Ray sequence parallel (#1803)

2026-09-01 01:37:23 -07:00