[Frontend][torch.compile] CompilationConfig Overhaul (#20283): name change compilation level to compilation mode, deprecation compilation level (#26355)

Signed-off-by: morrison-turnansky <mturnans@redhat.com>
Signed-off-by: Morrison Turnansky <mturnans@redhat.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
This commit is contained in:
Morrison Turnansky
2025-10-15 02:51:16 +00:00
committed by GitHub
co-authored by Luka Govedič
parent e66d787bce
commit 96b9aa5aa0
42 changed files with 270 additions and 248 deletions
+2 -2
View File
@@ -56,7 +56,7 @@ def make_compiler(compilation_config: CompilationConfig) -> CompilerInterface:
return InductorAdaptor()
else:
assert compilation_config.backend == "eager", (
"Custom backends not supported with CompilationLevel.PIECEWISE"
"Custom backends not supported with CompilationMode.VLLM_COMPILE"
)
logger.debug("Using EagerAdaptor")
@@ -481,7 +481,7 @@ def set_model_tag(tag: str):
class VllmBackend:
"""The compilation backend for `torch.compile` with vLLM.
It is used for compilation level of `CompilationLevel.PIECEWISE`,
It is used for compilation mode of `CompilationMode.VLLM_COMPILE`,
where we customize the compilation.
The major work of this backend is to split the graph into