Skip to content

GitLab

Unverified Commit d726a7b0 authored by Nick Hill's avatar Nick Hill Committed by GitHub
Browse files

[BugFix] Unblock use of LoRA with data parallel mode (#30220) 


Signed-off-by: default avatarNick Hill <nhill@redhat.com>
parent 344b50d5
Hide whitespace changes
Inline Side-by-side
Showing with 4 additions and 1 deletion
vllm/v1/metrics/loggers.py
View file @ d726a7b0
...@@ -952,7 +952,10 @@ class PrometheusStatLogger(AggregateStatLoggerBase): ...@@ -952,7 +952,10 @@ class PrometheusStatLogger(AggregateStatLoggerBase):
self.gauge_lora_info: Gauge | None = None self.gauge_lora_info: Gauge | None = None
if vllm_config.lora_config is not None: if vllm_config.lora_config is not None:
if len(self.engine_indexes) > 1: if len(self.engine_indexes) > 1:
raise NotImplementedError("LoRA in DP mode is not supported yet.") logger.warning(
"vllm:lora_requests_info prometheus metrics may be "
"incorrect/misleading with data parallel deployments."
)
self.labelname_max_lora = "max_lora" self.labelname_max_lora = "max_lora"
self.labelname_waiting_lora_adapters = "waiting_lora_adapters" self.labelname_waiting_lora_adapters = "waiting_lora_adapters"
self.labelname_running_lora_adapters = "running_lora_adapters" self.labelname_running_lora_adapters = "running_lora_adapters"
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment