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

[VLM] Support multimodal inputs for Florence-2 models (#13320)

Browse Source
This commit is contained in:
Isotr0py
2025-02-27 18:06:41 +08:00
committed by GitHub
parent 788f284b53
commit edf309ebbe
13 changed files with 1075 additions and 114 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
tests/models/decoder_only/audio_language/test_ultravox.py
View File

@ -15,7 +15,7 @@ from ....conftest import HfRunner, VllmRunner
from ....utils import RemoteOpenAIServer
from ...utils import check_logprobs_close
MODEL_NAME = "fixie-ai/ultravox-v0_5-llama-3_2-1b"
MODEL_NAME = "fixie-ai/ultravox-v0_4"
AudioTuple = Tuple[np.ndarray, int]
@ -187,7 +187,7 @@ def run_multi_audio_test(
@pytest.mark.core_model
@pytest.mark.parametrize("dtype", ["half"])
@pytest.mark.parametrize("dtype", ["bfloat16"])
@pytest.mark.parametrize("max_tokens", [128])
@pytest.mark.parametrize("num_logprobs", [5])
@pytest.mark.parametrize("vllm_kwargs", [