Cristhine
commented
10 months ago 运行时环境是A100 安装的是pytorch版本是1.12 cuda版本是11.2 执行代码里面的python demo/demo_lazy.py --config-file configs/LVISCOCOCOCOSTUFF_O365_OID_VGR_SA1B_REFCOCO_GQA_PhraseCut_Flickr30k/ape_deta/ape_deta_vitl_eva02_clip_vlf_lsj1024_cp_16x4_1080k.py --input demo/examples/Pisa.jpg --output save_draw/ --confidence-threshold 0.1 --text-prompt 'person,car,chess piece of horse head' --with-box --with-mask --with-sseg --opts train.init_checkpoint=checkpoint/ape_deta_vitl_eva02_clip_vlf_lsj1024_cp_16x4_1080k_mdl_20230829_162438/model_final.pth model.model_language.cache_dir="" model.model_vision.select_box_nums_for_evaluation=500 model.model_vision.text_feature_bank_reset=True model.model_vision.backbone.net.xattn=False
然后报错 File "/ssd1/vis/duyuting/anaconda3/envs/ape/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, kwargs) File "/ssd1/vis/duyuting/code/APE/ape/layers/multi_scale_deform_attn.py", line 334, in forward output = MultiScaleDeformableAttnFunction.apply( File "/ssd1/vis/duyuting/code/APE/ape/layers/multi_scale_deform_attn.py", line 44, in forward output = torch.ops.ape.ms_deform_attn_forward( File "/ssd1/vis/duyuting/anaconda3/envs/ape/lib/python3.10/site-packages/torch/_ops.py", line 143, in call* return self._op(args, **kwargs or {}) RuntimeError: Not compiled with GPU support
求助安装环境哪里不一致 尝试使用作者给的docker环境，在执行上述推理命令的时候也是报错缺少各种库
shenyunhang
chengyansheng24
wdc233
运行时环境是A100 安装的是pytorch版本是1.12 cuda版本是11.2 执行代码里面的python demo/demo_lazy.py --config-file configs/LVISCOCOCOCOSTUFF_O365_OID_VGR_SA1B_REFCOCO_GQA_PhraseCut_Flickr30k/ape_deta/ape_deta_vitl_eva02_clip_vlf_lsj1024_cp_16x4_1080k.py --input demo/examples/Pisa.jpg --output save_draw/ --confidence-threshold 0.1 --text-prompt 'person,car,chess piece of horse head' --with-box --with-mask --with-sseg --opts train.init_checkpoint=checkpoint/ape_deta_vitl_eva02_clip_vlf_lsj1024_cp_16x4_1080k_mdl_20230829_162438/model_final.pth model.model_language.cache_dir="" model.model_vision.select_box_nums_for_evaluation=500 model.model_vision.text_feature_bank_reset=True model.model_vision.backbone.net.xattn=False
然后报错 File "/ssd1/vis/duyuting/anaconda3/envs/ape/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, *kwargs) File "/ssd1/vis/duyuting/code/APE/ape/layers/multi_scale_deform_attn.py", line 334, in forward output = MultiScaleDeformableAttnFunction.apply( File "/ssd1/vis/duyuting/code/APE/ape/layers/multi_scale_deform_attn.py", line 44, in forward output = torch.ops.ape.ms_deform_attn_forward( File "/ssd1/vis/duyuting/anaconda3/envs/ape/lib/python3.10/site-packages/torch/_ops.py", line 143, in call return self._op(args, **kwargs or {}) RuntimeError: Not compiled with GPU support