Mercurial > ift6266

annotate deep/stacked_dae/v_guillaume/nist_sda_retrieve.py @ 436:0ca069550abd

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Added : single class version of SDA
author Guillaume Sicard <guitch21@gmail.com>
date Mon, 03 May 2010 06:14:05 -0400
parents
children
rev   line source
436
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
1 #!/usr/bin/python
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
2 # coding: utf-8
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
3
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
4 import ift6266
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
5 import pylearn
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
6
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
7 import numpy
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
8 import theano
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
9 import time
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
10
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
11 import pylearn.version
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
12 import theano.tensor as T
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
13 from theano.tensor.shared_randomstreams import RandomStreams
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
14
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
15 import copy
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
16 import sys
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
17 import os
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
18 import os.path
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
19
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
20 from jobman import DD
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
21 import jobman, jobman.sql
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
22 from pylearn.io import filetensor
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
23
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
24 from utils import produit_cartesien_jobs
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
25 from copy import copy
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
26
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
27 from sgd_optimization import SdaSgdOptimizer
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
28
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
29 #from ift6266.utils.scalar_series import *
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
30 from ift6266.utils.seriestables import *
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
31 import tables
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
32
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
33 from ift6266 import datasets
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
34 from config2 import *
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
35
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
36 '''
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
37 Function called by jobman upon launching each job
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
38 Its path is the one given when inserting jobs: see EXPERIMENT_PATH
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
39 '''
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
40 def jobman_entrypoint(state, channel):
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
41 # record mercurial versions of each package
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
42 pylearn.version.record_versions(state,[theano,ift6266,pylearn])
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
43 # TODO: remove this, bad for number of simultaneous requests on DB
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
44 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
45
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
46 # For test runs, we don't want to use the whole dataset so
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
47 # reduce it to fewer elements if asked to.
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
48 rtt = None
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
49 if state.has_key('reduce_train_to'):
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
50 rtt = state['reduce_train_to']
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
51 elif REDUCE_TRAIN_TO:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
52 rtt = REDUCE_TRAIN_TO
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
53
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
54 if state.has_key('decrease_lr'):
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
55 decrease_lr = state['decrease_lr']
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
56 else :
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
57 decrease_lr = 0
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
58
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
59 n_ins = 32*32
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
60 n_outs = 62 # 10 digits, 26*2 (lower, capitals)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
61
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
62 examples_per_epoch = NIST_ALL_TRAIN_SIZE
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
63 #To be sure variables will not be only in the if statement
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
64 PATH = ''
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
65 nom_reptrain = ''
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
66 nom_serie = ""
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
67 if state['pretrain_choice'] == 0:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
68 PATH=PATH_NIST
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
69 nom_pretrain='NIST'
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
70 nom_serie="series_NIST.h5"
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
71 elif state['pretrain_choice'] == 1:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
72 PATH=PATH_P07
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
73 nom_pretrain='P07'
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
74 nom_serie="series_P07.h5"
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
75
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
76 series = create_series(state.num_hidden_layers,nom_serie)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
77
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
78 print "Creating optimizer with state, ", state
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
79
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
80 optimizer = SdaSgdOptimizer(dataset=datasets.nist_all(),
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
81 hyperparameters=state, \
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
82 n_ins=n_ins, n_outs=n_outs,\
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
83 examples_per_epoch=examples_per_epoch, \
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
84 series=series,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
85 max_minibatches=rtt)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
86
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
87 parameters=[]
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
88 #Number of files of P07 used for pretraining
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
89 nb_file=0
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
90 ## if state['pretrain_choice'] == 0:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
91 ## print('\n\tpretraining with NIST\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
92 ## optimizer.pretrain(datasets.nist_all())
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
93 ## elif state['pretrain_choice'] == 1:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
94 ## #To know how many file will be used during pretraining
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
95 ## nb_file = state['pretraining_epochs_per_layer']
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
96 ## state['pretraining_epochs_per_layer'] = 1 #Only 1 time over the dataset
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
97 ## if nb_file >=100:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
98 ## sys.exit("The code does not support this much pretraining epoch (99 max with P07).\n"+
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
99 ## "You have to correct the code (and be patient, P07 is huge !!)\n"+
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
100 ## "or reduce the number of pretraining epoch to run the code (better idea).\n")
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
101 ## print('\n\tpretraining with P07')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
102 ## optimizer.pretrain(datasets.nist_P07(min_file=0,max_file=nb_file))
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
103
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
104 print ('Retrieve pre-train done earlier ( '+nom_pretrain+' )')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
105
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
106
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
107
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
108 sys.stdout.flush()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
109 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
110
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
111 #Set some of the parameters used for the finetuning
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
112 if state.has_key('finetune_set'):
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
113 finetune_choice=state['finetune_set']
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
114 else:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
115 finetune_choice=FINETUNE_SET
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
116
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
117 if state.has_key('max_finetuning_epochs'):
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
118 max_finetune_epoch_NIST=state['max_finetuning_epochs']
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
119 else:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
120 max_finetune_epoch_NIST=MAX_FINETUNING_EPOCHS
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
121
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
122 if state.has_key('max_finetuning_epochs_P07'):
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
123 max_finetune_epoch_P07=state['max_finetuning_epochs_P07']
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
124 else:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
125 max_finetune_epoch_P07=max_finetune_epoch_NIST
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
126
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
127 #Decide how the finetune is done
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
128
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
129 if finetune_choice == 0:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
130 print('\n\n\tfinetune with NIST\n\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
131 optimizer.reload_parameters(PATH+'params_pretrain.txt')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
132 optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=1,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
133 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
134 if finetune_choice == 1:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
135 print('\n\n\tfinetune with P07\n\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
136 optimizer.reload_parameters(PATH+'params_pretrain.txt')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
137 optimizer.finetune(datasets.nist_P07(),datasets.nist_all(),max_finetune_epoch_P07,ind_test=0,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
138 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
139 if finetune_choice == 2:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
140 print('\n\n\tfinetune with P07 followed by NIST\n\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
141 optimizer.reload_parameters(PATH+'params_pretrain.txt')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
142 optimizer.finetune(datasets.nist_P07(),datasets.nist_all(),max_finetune_epoch_P07,ind_test=20,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
143 optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=21,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
144 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
145 if finetune_choice == 3:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
146 print('\n\n\tfinetune with NIST only on the logistic regression on top (but validation on P07).\n\
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
147 All hidden units output are input of the logistic regression\n\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
148 optimizer.reload_parameters(PATH+'params_pretrain.txt')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
149 optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=1,special=1,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
150
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
151
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
152 if finetune_choice==-1:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
153 print('\nSERIE OF 4 DIFFERENT FINETUNINGS')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
154 print('\n\n\tfinetune with NIST\n\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
155 sys.stdout.flush()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
156 optimizer.reload_parameters(PATH+'params_pretrain.txt')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
157 optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=1,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
158 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
159 print('\n\n\tfinetune with P07\n\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
160 sys.stdout.flush()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
161 optimizer.reload_parameters(PATH+'params_pretrain.txt')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
162 optimizer.finetune(datasets.nist_P07(),datasets.nist_all(),max_finetune_epoch_P07,ind_test=0,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
163 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
164 print('\n\n\tfinetune with P07 (done earlier) followed by NIST (written here)\n\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
165 sys.stdout.flush()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
166 optimizer.reload_parameters('params_finetune_P07.txt')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
167 optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=21,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
168 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
169 print('\n\n\tfinetune with NIST only on the logistic regression on top.\n\
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
170 All hidden units output are input of the logistic regression\n\n')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
171 sys.stdout.flush()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
172 optimizer.reload_parameters(PATH+'params_pretrain.txt')
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
173 optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=1,special=1,decrease=decrease_lr)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
174 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
175
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
176 channel.save()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
177
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
178 return channel.COMPLETE
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
179
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
180 # These Series objects are used to save various statistics
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
181 # during the training.
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
182 def create_series(num_hidden_layers, nom_serie):
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
183
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
184 # Replace series we don't want to save with DummySeries, e.g.
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
185 # series['training_error'] = DummySeries()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
186
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
187 series = {}
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
188
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
189 basedir = os.getcwd()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
190
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
191 h5f = tables.openFile(os.path.join(basedir, nom_serie), "w")
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
192
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
193 # reconstruction
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
194 reconstruction_base = \
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
195 ErrorSeries(error_name="reconstruction_error",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
196 table_name="reconstruction_error",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
197 hdf5_file=h5f,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
198 index_names=('epoch','minibatch'),
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
199 title="Reconstruction error (mean over "+str(REDUCE_EVERY)+" minibatches)")
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
200 series['reconstruction_error'] = \
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
201 AccumulatorSeriesWrapper(base_series=reconstruction_base,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
202 reduce_every=REDUCE_EVERY)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
203
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
204 # train
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
205 training_base = \
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
206 ErrorSeries(error_name="training_error",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
207 table_name="training_error",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
208 hdf5_file=h5f,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
209 index_names=('epoch','minibatch'),
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
210 title="Training error (mean over "+str(REDUCE_EVERY)+" minibatches)")
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
211 series['training_error'] = \
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
212 AccumulatorSeriesWrapper(base_series=training_base,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
213 reduce_every=REDUCE_EVERY)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
214
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
215 # valid and test are not accumulated/mean, saved directly
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
216 series['validation_error'] = \
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
217 ErrorSeries(error_name="validation_error",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
218 table_name="validation_error",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
219 hdf5_file=h5f,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
220 index_names=('epoch','minibatch'))
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
221
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
222 series['test_error'] = \
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
223 ErrorSeries(error_name="test_error",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
224 table_name="test_error",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
225 hdf5_file=h5f,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
226 index_names=('epoch','minibatch'))
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
227
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
228 param_names = []
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
229 for i in range(num_hidden_layers):
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
230 param_names += ['layer%d_W'%i, 'layer%d_b'%i, 'layer%d_bprime'%i]
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
231 param_names += ['logreg_layer_W', 'logreg_layer_b']
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
232
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
233 # comment out series we don't want to save
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
234 series['params'] = SharedParamsStatisticsWrapper(
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
235 new_group_name="params",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
236 base_group="/",
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
237 arrays_names=param_names,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
238 hdf5_file=h5f,
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
239 index_names=('epoch',))
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
240
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
241 return series
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
242
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
243 # Perform insertion into the Postgre DB based on combination
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
244 # of hyperparameter values above
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
245 # (see comment for produit_cartesien_jobs() to know how it works)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
246 def jobman_insert_nist():
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
247 jobs = produit_cartesien_jobs(JOB_VALS)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
248
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
249 db = jobman.sql.db(JOBDB)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
250 for job in jobs:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
251 job.update({jobman.sql.EXPERIMENT: EXPERIMENT_PATH})
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
252 jobman.sql.insert_dict(job, db)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
253
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
254 print "inserted"
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
255
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
256 if __name__ == '__main__':
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
257
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
258 args = sys.argv[1:]
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
259
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
260 #if len(args) > 0 and args[0] == 'load_nist':
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
261 # test_load_nist()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
262
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
263 if len(args) > 0 and args[0] == 'jobman_insert':
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
264 jobman_insert_nist()
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
265
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
266 elif len(args) > 0 and args[0] == 'test_jobman_entrypoint':
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
267 chanmock = DD({'COMPLETE':0,'save':(lambda:None)})
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
268 jobman_entrypoint(DD(DEFAULT_HP_NIST), chanmock)
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
269
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
270 else:
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
271 print "Bad arguments"
0ca069550abd Added : single class version of SDA
Guillaume Sicard <guitch21@gmail.com>
parents:
diff changeset
272