Search results

screenspot 分数对齐

Status: Open.

#1391 In open-compass/VLMEvalKit;

· mathCrazyy opened on Dec 30, 2025

Video Holmes llm judge is not used in evaluation?

Status: Open.

#1390 In open-compass/VLMEvalKit;

· gyfastas opened on Dec 30, 2025

Qwen3-VL-4B-Instruct模型评估ocrbench结果和官方公布结果不一致

Status: Open.

#1389 In open-compass/VLMEvalKit;

· shibefore opened on Dec 30, 2025

模型评估结果和官方公布结果不一致

Status: Open.

#1388 In open-compass/VLMEvalKit;

· shorlockhxk opened on Dec 29, 2025

我的环境现在只支持我要使用的模型，但VLMEvalKit 在导入时尝试加载所有模型的依赖

Status: Open.

#1383 In open-compass/VLMEvalKit;

· cdllI opened on Dec 27, 2025

如何使用提前下载好的本地数据集，在哪里能配置路径呢

Status: Open.

#1373 In open-compass/VLMEvalKit;

· junfeiZTE opened on Dec 24, 2025

Update evaluation records of InternVL3

Status: Open.

#1372 In open-compass/VLMEvalKit;

· QYY1028 opened on Dec 24, 2025

Reproducibility of deepseek-vl inference results

Status: Open.

#1367 In open-compass/VLMEvalKit;

· kydxh opened on Dec 19, 2025

Hope to include Qwen3

Status: Open.

#1366 In open-compass/VLMEvalKit;

· Gao-anqi opened on Dec 18, 2025

How can I use VLMEvalKit to test the DeepSeek-OCR model and obtain its accuracy score on OmniDocBench?

Status: Open.

#1361 In open-compass/VLMEvalKit;

· zq1134241486-beep opened on Dec 16, 2025

Hope to open slack for everyone to communicate

Status: Open.

#1337 In open-compass/VLMEvalKit;

· wenba0 opened on Dec 3, 2025

MMVet 数据集问题

Status: Open.

#1336 In open-compass/VLMEvalKit;

· zsgd901 opened on Dec 2, 2025