MiniCPM-V
MiniCPM-V copied to clipboard
MiniCPM-V 4.5: A GPT-4o Level MLLM for Single Image, Multi Image and High-FPS Video Understanding on Your Phone
VisionEncoder里面的vit,与训练 权重 用的是idefics2还是 hf4m那个啊?
Very impressive work for the community! I am wondering how to deploy the MiniCPM-V 2.5 with vllm, and I have found that the repo https://github.com/OpenBMB/vllm/commit/94fcfc6582a347de3b26b377169bc78d06184c98 only support MiniCPM-V 2.0. Looking...
I'm trying to fine-tune MiniCPM-V2.5, but I get the following error: f"{','.join(outputs.keys())}. For reference, the inputs it received are { ^ SyntaxError: unterminated string literal (detected at line 38)
想使用vllm部署多卡推理,发现教程中提示的minicpmv_example.py 文件 并不存在vllm库中。
device_map='auto' > File "MiniCPM-Llama3-V-2_5/modeling_minicpmv.py", line 416, in chat res, vision_hidden_states = self.generate( File "MiniCPM-Llama3-V-2_5/modeling_minicpmv.py", line 326, in generate ) = self.get_vllm_embedding(model_inputs) File "MiniCPM-Llama3-V-2_5/modeling_minicpmv.py", line 148, in get_vllm_embedding cur_vllm_emb.scatter_(0, image_indices.view(-1, 1).repeat(1,...
Mac推理报错
PYTORCH_ENABLE_MPS_FALLBACK=1 python macinfer.py /Users/peng/miniconda3/envs/MiniCPMV/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`....
多卡推理报错,提示中间计算部分变量不在同一个 device 上
I've been running a few tests on an M1 pro - with a simple "whats in the image." I've noticed the model gets language confused: продавец I can see a...