model optimizer error

[Vedansh_Surjan]

I got the followinf error while converting a caffe model into IR format..Not sure what the error is for.

Can anyone please let me know how to rectify this

[JAIVIN_J_Intel]

Hi Vedansh,

Could you please provide information about the following?

• Which caffe model are you using? Is it a public model or a custom trained model?
• What Model Optimizer command have you used?
• Also, could you please verify that you are using one of the Supported Topologies and has Supported Caffe Layers?

Regards,

Jaivin

[Vedansh_Surjan]

Hi Jalvin,

I have used the model files from the link https://github.com/liuziwei7/fashion-landmarks

I have used command 

python mo_caffe.py --input_model /home/vedanshsurjan/Projects/Fashion/Landmark_detection/fashion-landmarks/models/FLD_upper/FLD_upper_models/stage3_easy.caffemodel --input_proto /home/vedanshsurjan/Projects/Fashion/Landmark_detection/fashion-landmarks/models/FLD_upper/cascade.prototxt --output_dir /home/vedanshsurjan/Desktop/openvino/models_ir/landmark_detection/upper/stage3 --model_name upper_model_s3_easy

[JAIVIN_J_Intel]

Hi Vedansh,

Thanks for sharing the information.

To use the Model Optimizer for a multi-input model with two inputs, provide a new shape for each input in the order they are passed to the Model Optimizer. Please refer the documentation for using Caffe*-Specific Conversion Parameters.

You may also refer Examples of CLI Commands and use the parameters accordingly.

Regards,

Jaivin

[Vedansh_Surjan]

Hi Jalvin 

I included the input shape command but it did not resolve the issue

Command code :

python mo_caffe.py --input_model /home/vedanshsurjan/Projects/Fashion/Landmark_detection/fashion-landmarks/models/FLD_upper/FLD_upper_models/stage3_easy.caffemodel --input_proto /home/vedanshsurjan/Projects/Fashion/Landmark_detection/fashion-landmarks/models/FLD_upper/cascade.prototxt --output_dir /home/vedanshsurjan/Desktop/openvino/models_ir/landmark_detection/upper/stage3/ --model_name upper_model_s3_easy --input data,last_stage_data --input_shape [1,3,224,224],[1,1,1,12]