search

martinkersner / train-CRF-RNN

Train CRF-RNN for Semantic Image Segmentation
Other
199 stars 92 forks source link

Error "MULTI_STAGE_MEANFIELD" and Failed to parse NetParameter file #49

Open matthiasmace opened 7 years ago

matthiasmace commented 7 years ago

Dear Martin In am willing to make the training run on MacOS. I am able to use CRFasRNN but when I want I run solve.py, I get the following errors : WARNING: Logging before InitGoogleLogging() is written to STDERR I1107 23:19:23.737943 3726271424 solver.cpp:54] Initializing solver from parameters: test_iter: 261 test_interval: 1333 base_lr: 1e-13 display: 50 max_iter: 100000 lr_policy: "fixed" momentum: 0.99 weight_decay: 0.0005 snapshot: 1000 snapshot_prefix: "models/train" solver_mode: GPU net: "TVG_CRFRNN_COCO_VOC_TRAIN_3_CLASSES.prototxt" test_initialization: false I1107 23:19:23.738425 3726271424 solver.cpp:96] Creating training net from net file: TVG_CRFRNN_COCO_VOC_TRAIN_3_CLASSES.prototxt [libprotobuf ERROR google/protobuf/text_format.cc:287] Error parsing text-format caffe.NetParameter: 224:3: Unknown enumeration value of "MULTI_STAGE_MEANFIELD" for field "type". F1107 23:19:23.739708 3726271424 upgrade_proto.cpp:932] Check failed: ReadProtoFromTextFile(param_file, param) Failed to parse NetParameter file: TVG_CRFRNN_COCO_VOC_TRAIN_3_CLASSES.prototxt Check failure stack trace:

Any idea please ?

Sam813 commented 6 years ago

hi, first, check if you install correct caffe version which has the MULTI_STAGE... layer. second, consider checking the prototext. It is written in the older caffe version, for instance the layers must be replaced with layer, also CONVOLUTION with "Convolution". a better way is to cross check it with network prototext (TVG_CRFRNN_new_deploy.prototxt)