updates in documentation (#3223)

Author: dtrawins

demos/continuous_batching/accuracy/README.md

@@ -80,6 +80,7 @@ export OPENAI_COMPATIBLE_API_URL=http://localhost:8000/v3
 export OPENAI_COMPATIBLE_API_KEY="unused"
 git clone https://github.com/EvolvingLMMs-Lab/lmms-eval
 cd lmms-eval
+git checkout 4471ad311e620ed6cf3a0419d8ba6f18f8fb1cb3  # https://github.com/EvolvingLMMs-Lab/lmms-eval/issues/625
 pip install -e . --extra-index-url "https://download.pytorch.org/whl/cpu"
 python -m lmms_eval \
     --model openai_compatible \
@@ -101,7 +102,7 @@ openai_compatible (model_version=OpenGVLab/InternVL2_5-8B,max_retries=1), gen_kw
 |--------|-------|------|-----:|--------------------|---|--------:|---|------|
 |mme     |Yaml   |none  |     0|mme_cognition_score |↑  | 600.3571|±  |   N/A|
 |mme     |Yaml   |none  |     0|mme_perception_score|↑  |1618.2984|±  |   N/A|
-|mmmu_val|      0|none  |     0|mmmu_acc            |↑  |   0.5100|±  |   N/A|
+|mmmu_val|      0|none  |     0|mmmu_acc            |↑  |   0.5322|±  |   N/A|
 
 ```
 

demos/continuous_batching/vlm/README.md

@@ -46,6 +46,7 @@ python export_model.py text_generation --source_model OpenGVLab/InternVL2_5-8B -
 > **Note:** Change the `--weight-format` to quantize the model to `int8` or `int4` precision to reduce memory consumption and improve performance.
 
 > **Note:** You can change the model used in the demo out of any topology [tested](https://github.com/openvinotoolkit/openvino.genai/blob/master/SUPPORTED_MODELS.md#visual-language-models) with OpenVINO.
+Be aware that QwenVL models executed on GPU might experience execution errors with very high resolution images. In case of such behavior, it is recommended to reduce the parameter `max_pixels` in `preprocessor_config.json`.
 
 You should have a model folder like below:
 ```