Skip to content

GitLab

Unverified Commit c07a7331 authored by Jiangyun Zhu's avatar Jiangyun Zhu Committed by GitHub
Browse files

[CI] enable idefics3 and fuyu-8b test in multimodal test (#23790) 


Signed-off-by: default avatarzjy0516 <riverclouds.zhu@qq.com>
parent 22feac8e
Show whitespace changes
Inline Side-by-side
Showing with 15 additions and 21 deletions
tests/models/multimodal/generation/test_common.py
View file @ c07a7331
...@@ -189,23 +189,21 @@ VLM_TEST_SETTINGS = { ...@@ -189,23 +189,21 @@ VLM_TEST_SETTINGS = {
}, },
marks=[pytest.mark.core_model], marks=[pytest.mark.core_model],
), ),
# FIXME(Isotr0py): Enable this test after "idefics3-transformers": VLMTestInfo(
# https://github.com/huggingface/transformers/pull/39470 released models=["HuggingFaceTB/SmolVLM-256M-Instruct"],
# "idefics3-transformers": VLMTestInfo( test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE),
# models=["HuggingFaceTB/SmolVLM-256M-Instruct"], prompt_formatter=lambda img_prompt:f"<|begin_of_text|>User:{img_prompt}<end_of_utterance>\nAssistant:", # noqa: E501
# test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE), img_idx_to_prompt=lambda idx: "<image>",
# prompt_formatter=lambda img_prompt:f"<|begin_of_text|>User:{img_prompt}<end_of_utterance>\nAssistant:", # noqa: E501 max_model_len=8192,
# img_idx_to_prompt=lambda idx: "<image>", max_num_seqs=2,
# max_model_len=8192, auto_cls=AutoModelForImageTextToText,
# max_num_seqs=2, hf_output_post_proc=model_utils.idefics3_trunc_hf_output,
# auto_cls=AutoModelForImageTextToText, image_size_factors=[(0.25, 0.5, 1.0)],
# hf_output_post_proc=model_utils.idefics3_trunc_hf_output, vllm_runner_kwargs={
# image_size_factors=[(0.25, 0.5, 1.0)], "model_impl": "transformers",
# vllm_runner_kwargs={ },
# "model_impl": "transformers", marks=[pytest.mark.core_model],
# }, ),
# marks=[pytest.mark.core_model],
# ),
# Pixel values from processor are not 4D or 5D arrays # Pixel values from processor are not 4D or 5D arrays
"qwen2_5_vl-transformers": VLMTestInfo( "qwen2_5_vl-transformers": VLMTestInfo(
models=["Qwen/Qwen2.5-VL-3B-Instruct"], models=["Qwen/Qwen2.5-VL-3B-Instruct"],
...@@ -322,10 +320,6 @@ VLM_TEST_SETTINGS = { ...@@ -322,10 +320,6 @@ VLM_TEST_SETTINGS = {
vllm_output_post_proc=model_utils.fuyu_vllm_to_hf_output, vllm_output_post_proc=model_utils.fuyu_vllm_to_hf_output,
num_logprobs=10, num_logprobs=10,
image_size_factors=[(), (0.25,), (0.25, 0.25, 0.25), (0.25, 0.2, 0.15)], image_size_factors=[(), (0.25,), (0.25, 0.25, 0.25), (0.25, 0.2, 0.15)],
# FIXME(Isotr0py): This model is broken in Transformers v4.54.1, we
# should enable this again after the fix is released:
# https://github.com/huggingface/transformers/pull/39915
marks=[pytest.mark.skip("HF model is broken")],
), ),
"gemma3": VLMTestInfo( "gemma3": VLMTestInfo(
models=["google/gemma-3-4b-it"], models=["google/gemma-3-4b-it"],
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment