biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[MM] Allow skipping memory profiling for multimodal models. (#22950)

Browse Source 
Signed-off-by: Roger Wang <hey@rogerw.me>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
This commit is contained in:
Roger Wang
2025-08-15 04:41:38 -07:00
committed by GitHub
parent 3e6dd40016
commit 49252cf59e
4 changed files with 120 additions and 89 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
vllm/engine/arg_utils.py
View File

@@ -350,6 +350,7 @@ class EngineArgs:
MultiModalConfig.mm_processor_kwargs
disable_mm_preprocessor_cache: bool = False # DEPRECATED
mm_processor_cache_gb: int = MultiModalConfig.mm_processor_cache_gb
skip_mm_profiling: bool = MultiModalConfig.skip_mm_profiling
# LoRA fields
enable_lora: bool = False
enable_lora_bias: bool = LoRAConfig.bias_enabled
@@ -716,6 +717,8 @@ class EngineArgs:
multimodal_group.add_argument(
"--interleave-mm-strings",
**multimodal_kwargs["interleave_mm_strings"])
multimodal_group.add_argument("--skip-mm-profiling",
**multimodal_kwargs["skip_mm_profiling"])
# LoRA related configs
lora_kwargs = get_kwargs(LoRAConfig)
@@ -918,6 +921,7 @@ class EngineArgs:
limit_mm_per_prompt=self.limit_mm_per_prompt,
interleave_mm_strings=self.interleave_mm_strings,
media_io_kwargs=self.media_io_kwargs,
skip_mm_profiling=self.skip_mm_profiling,
use_async_output_proc=not self.disable_async_output_proc,
config_format=self.config_format,
mm_processor_kwargs=self.mm_processor_kwargs,