Mirrors
/
hf_text-generation-inference
mirror of https://github.com/huggingface/text-generation-inference.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

misc(deps): update ompi from 4.1.6 to 4.1.7rc1 to avoid strange deadlock

This commit is contained in:
Morgan Funtowicz 2024-10-28 17:24:08 +01:00
parent 512225474a
commit 3bb78a8266
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
Dockerfile_trtllm
View File

@ -1,5 +1,5 @@
ARG CUDA_ARCH_LIST="75-real;80-real;86-real;89-real;90-real"
ARG OMPI_VERSION="4.1.6"
ARG OMPI_VERSION="4.1.7rc1"
# Build dependencies resolver stage
FROM lukemathwalker/cargo-chef:latest AS chef

2
backends/trtllm/lib/backend.cpp
View File

@ -52,7 +52,7 @@ huggingface::tgi::backends::GetParallelConfig(const size_t worldSize, const std:
if (worldSize > 1) {
SPDLOG_INFO("Detected sharded engine deployment, using orchestrator mode");
mode = tle::CommunicationMode::kORCHESTRATOR;
orchestratorConfig = std::make_optional<tle::OrchestratorConfig>(true, workerPath, nullptr, true);
orchestratorConfig = std::make_optional<tle::OrchestratorConfig>(true, workerPath);
} else {
SPDLOG_INFO("Detected single engine deployment, using leader mode");
}