Evan Quiney fac6832e5f fix warmup consistency for slow machines (#1748) ...

a fix from pr #1643 which is now stale - should make prefill more
consistent on very slow machines

## testing
qwen-3.5-35b-a3b loads normally
gpt-oss-120b-mxfp4-q8 loads normally

2026-03-24 16:45:55 +00:00

..

auto_bench.sh

fix warmup consistency for slow machines (#1748)

2026-03-24 16:45:55 +00:00

eval_tool_calls.sh

Fix tool calling (#1529)

2026-02-18 20:29:18 +00:00

get_all_models_on_cluster.py

auto bench (#1405)

2026-02-06 15:35:46 +00:00

headless_runner.py

Normalize TextGenerationTaskParams.input to list[InputMessage] (#1360)

2026-02-03 06:01:56 -08:00

run_exo_on.sh

cancel downloads for deleted instances (#1393)

2026-02-05 18:16:43 +00:00

start_distributed_test.py

improve distributed testing (#1300)

2026-02-02 18:25:39 +00:00