Cui Junbo
Cui Junbo
This issue does not provide a reproducible context and requires more information to help resolve it. If you still need assistance, please provide your code environment and running code to...
Based on the error message, the issue lies in the `bitsandbytes` package not being properly installed or found. You can try the following steps to resolve this issue: 1. **Ensure...
that means you do not install all python env for minicpm-v this will make that error disappear, but can't warranties do not have other error. ``` conda activate MiniCPMV pip...
@wodsoe 目前我们正在集中力量推进llamacpp的集成和进一步提升模型性能。如您有时间，欢迎您成为minicpm-v开源社区开发者之一，欢迎提交fp8版本模型～
https://github.com/open-compass/VLMEvalKit You can try this implement. We processed input as interleaved images and texts, and we use the vlmevalkit calculation method in order to compare with other models. @LongIslandWithoutIceTea
@qyc-98
你好，这是正常的。如果需要开到更高，可以尝试使用gradient_accumulation_steps来调整
Hi, is your V100 32G or 16G of GPU memory, 32G is fine with the current code, you may need to compress the image input to about 500 pixels wide...
for 16GB, finetuning only llm parameters or gradient_accumulation_steps with only 1 bs or gradient checkpointing maybe needed, we do not have the environment to test it. You can have a...
On IOS or ipadOS, the cli way is working correctly with the model using llamacpp, but we don't have the manpower to wrap it into swift, if you want to...