unified management of environment variables for vlm cuda ipc transport #14501
unified management of environment variables for vlm cuda ipc transport #14501yuan-luo merged 1 commit intosgl-project:mainfrom
Conversation
Summary of ChangesHello @yhyang201, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request refactors the management of environment variables pertinent to VLM CUDA IPC transport. It consolidates the definition and default values of these variables into a central Highlights
Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here. You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension. Footnotes
|
There was a problem hiding this comment.
Code Review
This pull request refactors the management of environment variables by introducing a unified Envs class, which is a great improvement for code maintainability and consistency. The changes correctly replace direct environment variable access with the new centralized approach. My review includes a few minor suggestions to further improve code style and consistency.
| SGLANG_RESIZE_RESAMPLE = EnvStr("") | ||
|
|
||
| # VLM Item CUDA IPC Transport | ||
| SGLANG_USE_CUDA_IPC_TRANSPORT=EnvBool(False) |
There was a problem hiding this comment.
For better readability and to adhere to PEP 8 style guidelines, please add spaces around the assignment operator.
| SGLANG_USE_CUDA_IPC_TRANSPORT=EnvBool(False) | |
| SGLANG_USE_CUDA_IPC_TRANSPORT = EnvBool(False) |
| _is_npu = is_npu() | ||
|
|
||
| SGL_USE_CUDA_IPC = get_bool_env_var("SGLANG_USE_CUDA_IPC_TRANSPORT") | ||
| SGL_USE_CUDA_IPC = envs.SGLANG_USE_CUDA_IPC_TRANSPORT.get() |
There was a problem hiding this comment.
For consistency with the project's naming conventions and the ongoing effort to deprecate the SGL_ prefix (as seen in environ.py), consider renaming this variable. A name like SGLANG_USE_CUDA_IPC or USE_CUDA_IPC_TRANSPORT would be more consistent. If you apply this change, please remember to update its usages throughout the file.
| MM_ITEM_MEMORY_POOL_RECYCLE_INTERVAL = ( | ||
| 0.05 | ||
| if not get_float_env_var("SGLANG_MM_ITEM_MEM_POOL_RECYCLE_INTERVAL_SEC") | ||
| else get_float_env_var("SGLANG_MM_ITEM_MEM_POOL_RECYCLE_INTERVAL_SEC") | ||
| envs.SGLANG_MM_ITEM_MEM_POOL_RECYCLE_INTERVAL_SEC.get() | ||
| ) |
There was a problem hiding this comment.
The parentheses around this assignment are redundant and can be removed for cleaner code.
MM_ITEM_MEMORY_POOL_RECYCLE_INTERVAL = envs.SGLANG_MM_ITEM_MEM_POOL_RECYCLE_INTERVAL_SEC.get()|
/tag-and-rerun-ci |
…n3_pp * 'main' of https://github.com/sgl-project/sglang: (74 commits) [bug fix][pp] fix inconsistent latency between tp (sgl-project#15379) Fix warp illegal instruction in kimi k2 thinking PCG (sgl-project#15306) Fix gpt-oss yarn with `truncate` argument (sgl-project#14270) Monkey patch deepseek-ocr's `v_head_dim` (sgl-project#15384) [model-gateway] Replace PolicyRegistry RwLock with DashMap for lock-free policy lookups (sgl-project#15361) [PP] Fix dynamic chunking strategy for PP (sgl-project#15372) Fix issue: ENABLE_BELOW_SM90 cannot be enabled on aarch64 CPU (sgl-project#12967) Split test_piecewise_cuda_graph.py to optimize CI resource usage (sgl-project#15290) unified management of environment variables for vlm cuda ipc transport (sgl-project#14501) Mistral Large 3 NVFP4 TRTLLM MoE support (sgl-project#15049) fix: adjust time for test_epd_disaggregation.py (sgl-project#15354) Add doc for qwen3 next (sgl-project#15337) feat: DeepSeek-V3.2 Streaming tool call output (sgl-project#15278) Feature/trtllm mha workspace size configurable sgl-project#15089 (sgl-project#15131) [VLM] Support cos sin cache for Qwen3-VL & GLM-4.1V (sgl-project#15205) [Deepseek V3.2] Support Overlap Spec + NSA (sgl-project#15307) Add request-level timestamp for when prefill finishes (sgl-project#14860) [CI] Migrate LoRA tests to test/registered/lora/ (sgl-project#15176) Reserve more memory for DeepSeekOCR model and adjust server start timeout for DeepGEMM to reduce flakiness (sgl-project#15277) Fix condition check for require_gathered_buffer (sgl-project#15328) ...
Motivation
Modifications
Accuracy Tests
Benchmarking and Profiling
Checklist