Closed Leon1207 closed 9 months ago
Thanks for your reply. I have another question, how can I get the file format_mme_bench_v1.2.json from MME bench?
Thanks for your reply. I have another question, how can I get the file format_mme_bench_v1.2.json from MME bench?
I have uploaded the file at here. https://github.com/Q-MM/PureMM/blob/main/eval/format_mme_bench_v1.2.json You should modified the mme folder path with your own path to use it.
Thanks for your work. But when I tested a random image with the prompt in MME benchmark, I got the rubbish result. I have made sure that the model weights are loaded correctly, do you have any advice to solve it?
Is the output proper when you test with the mme images?
---Original--- From: @.> Date: Fri, Jan 19, 2024 21:42 PM To: @.>; Cc: @.**@.>; Subject: Re: [Q-MM/PureMM] What is the required conda environment for thiscode? thanks! (Issue #3)
Thanks for your work. But when I tested a random image with the prompt in MME benchmark, I got the rubbish result. I have made sure that the model weights are loaded correctly, do you have any advice to solve it? image.png (view on web)
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>
conda create -n puremm python=3.10 -y conda activate puremm pip install -r requirements.txt