[Model] Support Llama4 in vLLM (#16104)

2025-04-05 21:01:00 -07:00
parent 63375f0cdb
commit c575232395
35 changed files with 2369 additions and 142 deletions
--- a/tests/models/registry.py
+++ b/tests/models/registry.py
@@ -337,6 +337,7 @@ _MULTIMODAL_EXAMPLE_MODELS = {
                                                         tokenizer="facebook/bart-base",
                                                         trust_remote_code=True),  # noqa: E501
    "MllamaForConditionalGeneration": _HfExamplesInfo("meta-llama/Llama-3.2-11B-Vision-Instruct"),  # noqa: E501
+    "Llama4ForConditionalGeneration": _HfExamplesInfo("meta-llama/Llama-4-Scout-17B-16E-Instruct"),  # noqa: E501
    "WhisperForConditionalGeneration": _HfExamplesInfo("openai/whisper-large-v3"),  # noqa: E501
 }