75ef5993ce
resolve_attention_implementation calls is_torchelastic_launched() which does not exist in the incomplete torch.distributed shipped with the Windows ROCm wheel, causing a warning on every model config load in the server process. Add it to the stub table alongside the four helpers already patched in _determine_attention_impl_for_gpu_estimate. Also adds two tests: one confirming the new stub and one confirming all five core distributed helpers are covered. |
||
|---|---|---|
| .. | ||
| conftest.py | ||
| smoke_test_llama_prebuilt.py | ||
| smoke_test_parallel_studio_home.py | ||
| test_install_llama_prebuilt_logic.py | ||
| test_llama_pr_force_and_source.py | ||
| test_pr4562_bugfixes.py | ||
| test_rocm_support.py | ||
| test_selection_logic.py | ||