A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 08:30 UTC
[Model] Add support for Qwen2.5-Omni-7B-AWQ (Qwen2_5OmniForConditionalGeneration) (#18647)
Yuanhao WU committed
May 24, 2025 at 17:15 UTC
a859320575651cb310be2064f7054d1c91547f77
2 files changed
+3
tests/models/registry.py
+2
@@ -398,6 +398,8 @@ _MULTIMODAL_EXAMPLE_MODELS = {
398
"Qwen2_5_VLForConditionalGeneration": _HfExamplesInfo("Qwen/Qwen2.5-VL-3B-Instruct"), # noqa: E501
399
"Qwen2_5OmniModel": _HfExamplesInfo("Qwen/Qwen2.5-Omni-3B",
400
min_transformers_version="4.52"),
401
+ "Qwen2_5OmniForConditionalGeneration": _HfExamplesInfo("Qwen/Qwen2.5-Omni-7B-AWQ", # noqa: E501
402
+ min_transformers_version="4.52"),
403
"SkyworkR1VChatModel": _HfExamplesInfo("Skywork/Skywork-R1V-38B"),
404
"SmolVLMForConditionalGeneration": _HfExamplesInfo("HuggingFaceTB/SmolVLM2-2.2B-Instruct"), # noqa: E501
405
"UltravoxModel": _HfExamplesInfo("fixie-ai/ultravox-v0_5-llama-3_2-1b", # noqa: E501
vllm/model_executor/models/registry.py
+1
@@ -208,6 +208,7 @@ _MULTIMODAL_MODELS = {
208
"Qwen2_5_VLForConditionalGeneration": ("qwen2_5_vl", "Qwen2_5_VLForConditionalGeneration"), # noqa: E501
209
"Qwen2AudioForConditionalGeneration": ("qwen2_audio", "Qwen2AudioForConditionalGeneration"), # noqa: E501
210
"Qwen2_5OmniModel": ("qwen2_5_omni_thinker", "Qwen2_5OmniThinkerForConditionalGeneration"), # noqa: E501
211
+ "Qwen2_5OmniForConditionalGeneration": ("qwen2_5_omni_thinker", "Qwen2_5OmniThinkerForConditionalGeneration"), # noqa: E501
212
"UltravoxModel": ("ultravox", "UltravoxModel"),
213
"Phi4MMForCausalLM": ("phi4mm", "Phi4MMForCausalLM"),
214
# [Encoder-decoder]