- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
When i try to compile my company's caffe model,there is a error"[Error 15] Toolkit Error: Parser Not Supported: None".
Is it wrong with the model?And ow can i modify it?
My company's model has MemoryData layer,but there is no Input layer,like this:
name:"PoyaNet"
layer {
name: "data"
type: "MemoryData"
top: "data"
top: "label"
transform_param {
mirror: false
mean_value: 127.5
mean_value: 127.5
mean_value: 127.5
scale: 0.0078125
}
memory_data_param {
batch_size: 1
channels:3
height:112
width:96
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
kernel_size: 3
stride: 2
pad: 1
weight_filler {
type: "xavier"
}
bias_filler {
type: "constant"
value: 0
}
}
}
*
*
*
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
@InSilence It looks like your prototxt file is expecting input from a labeled data input. This is likely because your prototxt file is set up for training. You must convert your training prototxt file into a deployment prototxt file for use with input that isn't labeled. This will include removing layers that expected labeled input and data layers. More information can be found here: https://github.com/BVLC/caffe/wiki/Using-a-Trained-Network:-Deploy.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
@Tome_at_Intel Okay,thank you for your response! :)
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
@Tome_at_Intel I modify the network prototxt file like the following code:
name:"PoyaNet"
input: "data"
input_shape: {
dim: 1
dim: 3
dim: 112
dim: 96
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
kernel_size: 3
stride: 2
pad: 1
weight_filler {
type: "xavier"
}
bias_filler {
type: "constant"
value: 0
}
}
}
but it also error :(
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
@Tome_at_Intel My prototxt file extensions is "xxx.prot",i modify it to "xxx.prototxt" ,It's done!
![](/skins/images/7FC17B7B85029576C25F1E43CE255B51/responsive_peak/images/icon_anonymous_message.png)
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page