search

keplr-io / quiver

Interactive convnet features visualization for Keras
https://keplr-io.github.io/quiver/
MIT License
1.76k stars 223 forks source link

No model on the GUI #8

Closed kretes closed 7 years ago

kretes commented 7 years ago

After running quiver with a tensorflow model - I see images from my directory but I do not see the model on the left.

some errors in web console.

quiver

server code:

from keras.models import load_model

model = load_model('/project/results/models/final/202imgs-101-101-in768-d16-conv3-drop-fc-sgd-lr-0_00025-from400epochs.h5')
print(model.summary())

from quiver_engine import server
server.launch(model, input_folder='/project/data/class_0')

(this code is loading my model, but the same happens with simple keras model created here)

server logs:

Using TensorFlow backend.
____________________________________________________________________________________________________
Layer (type)                     Output Shape          Param #     Connected to                     
====================================================================================================
convolution2d_1 (Convolution2D)  (None, 768, 768, 16)  448         convolution2d_input_1[0][0]      
____________________________________________________________________________________________________
activation_1 (Activation)        (None, 768, 768, 16)  0           convolution2d_1[0][0]            
____________________________________________________________________________________________________
dropout_1 (Dropout)              (None, 768, 768, 16)  0           activation_1[0][0]               
____________________________________________________________________________________________________
maxpooling2d_1 (MaxPooling2D)    (None, 384, 384, 16)  0           dropout_1[0][0]                  
____________________________________________________________________________________________________
convolution2d_2 (Convolution2D)  (None, 384, 384, 16)  2320        maxpooling2d_1[0][0]             
____________________________________________________________________________________________________
activation_2 (Activation)        (None, 384, 384, 16)  0           convolution2d_2[0][0]            
____________________________________________________________________________________________________
dropout_2 (Dropout)              (None, 384, 384, 16)  0           activation_2[0][0]               
____________________________________________________________________________________________________
maxpooling2d_2 (MaxPooling2D)    (None, 192, 192, 16)  0           dropout_2[0][0]                  
____________________________________________________________________________________________________
convolution2d_3 (Convolution2D)  (None, 192, 192, 16)  2320        maxpooling2d_2[0][0]             
____________________________________________________________________________________________________
activation_3 (Activation)        (None, 192, 192, 16)  0           convolution2d_3[0][0]            
____________________________________________________________________________________________________
dropout_3 (Dropout)              (None, 192, 192, 16)  0           activation_3[0][0]               
____________________________________________________________________________________________________
maxpooling2d_3 (MaxPooling2D)    (None, 96, 96, 16)    0           dropout_3[0][0]                  
____________________________________________________________________________________________________
convolution2d_4 (Convolution2D)  (None, 96, 96, 16)    2320        maxpooling2d_3[0][0]             
____________________________________________________________________________________________________
activation_4 (Activation)        (None, 96, 96, 16)    0           convolution2d_4[0][0]            
____________________________________________________________________________________________________
dropout_4 (Dropout)              (None, 96, 96, 16)    0           activation_4[0][0]               
____________________________________________________________________________________________________
maxpooling2d_4 (MaxPooling2D)    (None, 48, 48, 16)    0           dropout_4[0][0]                  
____________________________________________________________________________________________________
convolution2d_5 (Convolution2D)  (None, 48, 48, 16)    2320        maxpooling2d_4[0][0]             
____________________________________________________________________________________________________
activation_5 (Activation)        (None, 48, 48, 16)    0           convolution2d_5[0][0]            
____________________________________________________________________________________________________
dropout_5 (Dropout)              (None, 48, 48, 16)    0           activation_5[0][0]               
____________________________________________________________________________________________________
maxpooling2d_5 (MaxPooling2D)    (None, 24, 24, 16)    0           dropout_5[0][0]                  
____________________________________________________________________________________________________
convolution2d_6 (Convolution2D)  (None, 24, 24, 16)    2320        maxpooling2d_5[0][0]             
____________________________________________________________________________________________________
activation_6 (Activation)        (None, 24, 24, 16)    0           convolution2d_6[0][0]            
____________________________________________________________________________________________________
dropout_6 (Dropout)              (None, 24, 24, 16)    0           activation_6[0][0]               
____________________________________________________________________________________________________
maxpooling2d_6 (MaxPooling2D)    (None, 12, 12, 16)    0           dropout_6[0][0]                  
____________________________________________________________________________________________________
convolution2d_7 (Convolution2D)  (None, 12, 12, 16)    2320        maxpooling2d_6[0][0]             
____________________________________________________________________________________________________
activation_7 (Activation)        (None, 12, 12, 16)    0           convolution2d_7[0][0]            
____________________________________________________________________________________________________
dropout_7 (Dropout)              (None, 12, 12, 16)    0           activation_7[0][0]               
____________________________________________________________________________________________________
maxpooling2d_7 (MaxPooling2D)    (None, 6, 6, 16)      0           dropout_7[0][0]                  
____________________________________________________________________________________________________
flatten_1 (Flatten)              (None, 576)           0           maxpooling2d_7[0][0]             
____________________________________________________________________________________________________
dropout_8 (Dropout)              (None, 576)           0           flatten_1[0][0]                  
____________________________________________________________________________________________________
dense_1 (Dense)                  (None, 10)            5770        dropout_8[0][0]                  
____________________________________________________________________________________________________
activation_8 (Activation)        (None, 10)            0           dense_1[0][0]                    
____________________________________________________________________________________________________
dense_2 (Dense)                  (None, 1)             11          activation_8[0][0]               
____________________________________________________________________________________________________
activation_9 (Activation)        (None, 1)             0           dense_2[0][0]                    
====================================================================================================
Total params: 20149
____________________________________________________________________________________________________
None
172.17.42.1 - - [2016-11-15 21:23:33] "GET / HTTP/1.1" 200 802 0.003623
172.17.42.1 - - [2016-11-15 21:23:33] "GET /app.c1d3d21cfba8c1aabed6806773c8f0ba.css HTTP/1.1" 200 5131 0.001499
172.17.42.1 - - [2016-11-15 21:23:33] "GET /vendor.e63f1eee61ce7b325e64.js HTTP/1.1" 200 87557 0.001622
172.17.42.1 - - [2016-11-15 21:23:33] "GET /app.a9499a05f30e8bb1c4af.js HTTP/1.1" 200 1179313 0.005261
172.17.42.1 - - [2016-11-15 21:23:34] "GET /model HTTP/1.1" 200 10407 0.002802
172.17.42.1 - - [2016-11-15 21:23:34] "GET /inputs HTTP/1.1" 200 1079 0.001351
172.17.42.1 - - [2016-11-15 21:23:34] "GET /favicon.ico HTTP/1.1" 200 16181 0.001027
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161102_201212.jpg HTTP/1.1" 200 200431 0.002388
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161103_090645.jpg HTTP/1.1" 200 319399 0.003071
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161103_090541.jpg HTTP/1.1" 200 252399 0.003076
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161112121536_52.15833_21.06697_1000006.jpg HTTP/1.1" 200 1138946 0.006618
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161102_182022.jpg HTTP/1.1" 200 319155 0.003340
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161109101939_52.1751_21.06294_1000001.jpg HTTP/1.1" 200 1370579 0.010491
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161103_081455.jpg HTTP/1.1" 200 174565 0.002427
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161113121455_52.18729_21.06157_1000003.jpg HTTP/1.1" 200 1225303 0.007116
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161110084644_52.1601_21.07184_1000004.jpg HTTP/1.1" 200 1072800 0.040332
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161103_194107.jpg HTTP/1.1" 200 168064 0.002330
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161102_201612.jpg HTTP/1.1" 200 255676 0.002810
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161112122059_52.17202_21.05814_1000006.jpg HTTP/1.1" 200 1475996 0.007180
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161113121610_52.1845_21.06018_1000003.jpg HTTP/1.1" 200 1451522 0.016294
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161113104730_52.1602_21.06501_1000001.jpg HTTP/1.1" 200 1092566 0.012721
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161103_192955.jpg HTTP/1.1" 200 195772 0.002486
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161113121520_52.18648_21.06136_1000003.jpg HTTP/1.1" 200 1581348 0.012774
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161109101834_52.16985_21.06818_1000001.jpg HTTP/1.1" 200 1478398 0.007960
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161102_182112.jpg HTTP/1.1" 200 216239 0.002271
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161113121434_52.18714_21.0619_1000003.jpg HTTP/1.1" 200 1096357 0.004649
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161102_090437.jpg HTTP/1.1" 200 394772 0.003571
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/IMG_20161103_194208.jpg HTTP/1.1" 200 153944 0.001994
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161113124924_52.16035_21.07183_1000004.jpg HTTP/1.1" 200 1314513 0.006314
172.17.42.1 - - [2016-11-15 21:23:34] "GET /input-file/20161113104727_52.1602_21.06501_1000001.jpg HTTP/1.1" 200 1004405 0.005401
172.17.42.1 - - [2016-11-15 21:26:01] "GET / HTTP/1.1" 200 802 0.001103
172.17.42.1 - - [2016-11-15 21:26:01] "GET /app.c1d3d21cfba8c1aabed6806773c8f0ba.css HTTP/1.1" 200 5131 0.001246
172.17.42.1 - - [2016-11-15 21:26:01] "GET /app.a9499a05f30e8bb1c4af.js HTTP/1.1" 200 1179313 0.006761
172.17.42.1 - - [2016-11-15 21:26:01] "GET /vendor.e63f1eee61ce7b325e64.js HTTP/1.1" 200 87557 0.001366
172.17.42.1 - - [2016-11-15 21:26:01] "GET /model HTTP/1.1" 200 10407 0.003014
172.17.42.1 - - [2016-11-15 21:26:01] "GET /inputs HTTP/1.1" 200 1079 0.000988

Anything I can do to help debug the issue?

jakebian commented 7 years ago

Looks like I didn't test this thoroughly with the Sequential API (not the functional API), but looks like a client side graph generation issue, should be quick, loooking into it.

jakebian commented 7 years ago

https://github.com/jakebian/quiver/commit/6c7116e858172293d52cb18ebffd2733ed67bf91

Properly supported sequential models and tested on your model in your gist

screenshot from 2016-11-15 14-22-09

WIll close this issue once you verify that it works!

kretes commented 7 years ago

yes after reinstalling from pip release I get this working.

I have two other issues which I will put on a seperate tickets