Wuyingwen comments

Results 5 comments of


                                            Wuyingwen

训练时报错AttributeError: 'DeepSpeedCPUAdam' object has no attribute 'ds_opt_adam'

maybe you need to update your gcc version, this could solve this problem.

训练时报错AttributeError: 'DeepSpeedCPUAdam' object has no attribute 'ds_opt_adam'

解决了，通过把gcc版本提高

Best Practices for Inference and Fine-Tuning with MiniCPM-V 2.6

您好，我们测试了您提供的 CUDA_VISIBLE_DEVICES=0 swift infer --model_type minicpm-v-v2_6-chat --model_id_or_path openbmb/MiniCPM-V-2_6 以及 video测试代码。发现对视频的测试结果，似乎只依赖于视频第一帧。我们尝试了多次对视频OCR的提取，结果显示都只会输出第一帧的OCR结果。请问能提供具体的测试代码(.py文件)地址么？我们想check一下数据处理的部分，是否只读取了视频第一帧的信息。

Training of Stage 3 , 第三阶段训练，代码中实际的训练参数与论文不符

在训练代码 train.py 开始训练前打印训练参数，其中包含下列项： transformer.mm_projector.image_spatial_proj.0.weight transformer.mm_projector.image_spatial_proj.0.bias transformer.mm_projector.image_spatial_proj.2.weight transformer.mm_projector.image_spatial_proj.2.bias

ImportError: cannot import name '_expand_mask' from 'transformers.models.clip.modeling_clip'

you can copy the following code into corresponding transformer libarary to solve the problem def _expand_mask(mask: torch.Tensor, dtype: torch.dtype, tgt_len: Optional[int] = None): bsz, src_len = mask.size() tgt_len = tgt_len...