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
ipad上支持吗
### 是否已有关于该错误的issue或讨论? | Is there an existing issue / discussion for this? - [ ] 我已经搜索过已有的issues和讨论 | I have searched the existing issues / discussions ### 该问题是否在FAQ中有解答? | Is there...
### 是否已有关于该错误的issue或讨论? | Is there an existing issue / discussion for this? - [X] 我已经搜索过已有的issues和讨论 | I have searched the existing issues / discussions ### 该问题是否在FAQ中有解答? | Is there an...
Original code will load model to CPU memory in float32, and will require 36GB+ CPU memory. My modification will use less CPU memory (< 20GB).
### 起始日期 | Start Date _No response_ ### 实现PR | Implementation PR _No response_ ### 相关Issues | Reference Issues _No response_ ### 摘要 | Summary 模型能上传到https://ollama.com/library吗? ### 基本示例 | Basic...
Does MiniCPM-V support multiple image? Asking because there are image comparison tasks in OpenCampus benchmark. How are those cases handled?
any help is appreciated (: Loading checkpoint shards: 14%|███████ | 1/7 [00:13
### 是否已有关于该错误的issue或讨论? | Is there an existing issue / discussion for this? - [x] 我已经搜索过已有的issues和讨论 | I have searched the existing issues / discussions ### 该问题是否在FAQ中有解答? | Is there an...
finetune/readme.md下面写的加载lora方法的方法只有一部分,请问具体要怎么加载测试呢?
zero3支持
修改了huggingface上我们模型的resampler和minicpmv模型文件,这个pr需要和huggingface的pr一起提交,主要解决了目前zero3需要强制聚拢参数的问题,以及主模型运行时某些变量没有及时被deepspeed发送到对应的显卡上导致minicpmv2不能使用zero3算法微调