Enable modelopt gemma3 nvfp4/fp8, make workflow more robust (#22771)
Signed-off-by: Zhiyu Cheng <zhiyuc@nvidia.com> Signed-off-by: Michael Goin <mgoin64@gmail.com> Co-authored-by: Michael Goin <mgoin64@gmail.com>
This commit is contained in:
@@ -31,8 +31,11 @@ logger = init_logger(__name__)
|
||||
|
||||
def make_compiler(compilation_config: CompilationConfig) -> CompilerInterface:
|
||||
if compilation_config.use_inductor:
|
||||
if envs.VLLM_USE_STANDALONE_COMPILE and is_torch_equal_or_newer(
|
||||
"2.8.0.dev"):
|
||||
# Use standalone compile only if requested, version is new enough,
|
||||
# and the symbol actually exists in this PyTorch build.
|
||||
if (envs.VLLM_USE_STANDALONE_COMPILE
|
||||
and is_torch_equal_or_newer("2.8.0.dev")
|
||||
and hasattr(torch._inductor, "standalone_compile")):
|
||||
logger.debug("Using InductorStandaloneAdaptor")
|
||||
return InductorStandaloneAdaptor()
|
||||
else:
|
||||
|
||||
Reference in New Issue
Block a user