[Log] Optimize Startup Log (#26740)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
@@ -368,11 +368,13 @@ class UnquantizedFusedMoEMethod(FusedMoEMethodBase, CustomOp):
|
||||
logger.info_once(
|
||||
"FlashInfer CUTLASS MoE is available for EP"
|
||||
" but not enabled, consider setting"
|
||||
" VLLM_USE_FLASHINFER_MOE_FP16=1 to enable it."
|
||||
" VLLM_USE_FLASHINFER_MOE_FP16=1 to enable it.",
|
||||
scope="local",
|
||||
)
|
||||
elif self.moe.moe_parallel_config.dp_size > 1:
|
||||
logger.info_once(
|
||||
"FlashInfer CUTLASS MoE is currently not available for DP."
|
||||
"FlashInfer CUTLASS MoE is currently not available for DP.",
|
||||
scope="local",
|
||||
)
|
||||
self.flashinfer_cutlass_moe = None # type: ignore
|
||||
|
||||
|
||||
@@ -311,9 +311,10 @@ class DefaultModelLoader(BaseModelLoader):
|
||||
loaded_weights = load_weights_and_online_quantize(self, model, model_config)
|
||||
|
||||
self.counter_after_loading_weights = time.perf_counter()
|
||||
logger.info(
|
||||
logger.info_once(
|
||||
"Loading weights took %.2f seconds",
|
||||
self.counter_after_loading_weights - self.counter_before_loading_weights,
|
||||
scope="local",
|
||||
)
|
||||
# We only enable strict check for non-quantized models
|
||||
# that have loaded weights tracking currently.
|
||||
|
||||
@@ -416,7 +416,7 @@ def download_weights_from_hf(
|
||||
e,
|
||||
)
|
||||
|
||||
logger.info("Using model weights format %s", allow_patterns)
|
||||
logger.debug("Using model weights format %s", allow_patterns)
|
||||
# Use file lock to prevent multiple processes from
|
||||
# downloading the same model weights at the same time.
|
||||
with get_lock(model_name_or_path, cache_dir):
|
||||
|
||||
Reference in New Issue
Block a user