Esempio n. 1
0
def index():
    if g.user:
        runs = db.query_all_runs(session['user_id'])
        if len(runs) > 0:
            for run in runs:
                if run['descr'] == 'Not started':
                    db.clean_run(run_id=run['id'])
            return render_template('home/index.html',
                                   runs=runs,
                                   logged_in=True)
        else:
            return render_template('home/index.html', logged_in=True)
    else:
        return render_template('home/index.html', logged_in=False)
Esempio n. 2
0
def delete_run():
    runs = db.query_all_runs(user_id=session['user_id'])
    run_id = int(runs[int(request.form['index']) - 1]['id'])

    # Will cancel runs if they are currently in queue
    ids = db.query_get_job_ids(run_id)
    for id in ids:
        if id in current_app.task_queue.jobs:
            cancel_job(job_id=id, connection=current_app.redis)

    # Deletes all files associated with run and sets live = 0 in database (which will cancel run if it is currently in process and checkpoint is reached)
    db.clean_run(run_id=run_id)
    username, title = db.query_username_title(run_id=run_id)
    logger.info('User #{} ({}) deleted Run #{} ({})'.format(
        session['user_id'], username, run_id, title))
    return ''
Esempio n. 3
0
def tabular():
    if request.method == 'POST':
        if 'cancel' in request.form:
            db.clean_run(run_id=session['run_id'])
            return redirect(url_for('index'))

        if 'back' in request.form:
            db.clean_run(run_id=session['run_id'])
            return redirect(url_for('create.create'))

        dep_var = request.form['dep_var']
        cont_inputs = request.form.getlist('cont_inputs')
        int_inputs = request.form.getlist('int_inputs')
        num_epochs = cs.TABULAR_DEFAULT_NUM_EPOCHS if request.form[
            'num_epochs'] == '' else int(request.form['num_epochs'])
        error = cu.validate_tabular_choices(dep_var=dep_var,
                                            cont_inputs=cont_inputs,
                                            int_inputs=int_inputs)
        if error:
            flash(error)
        else:
            db.query_add_depvar(run_id=session['run_id'], depvar=dep_var)
            db.query_add_cont_inputs(run_id=session['run_id'],
                                     cont_inputs=cont_inputs)
            session['dep_var'] = dep_var
            session['cont_inputs'] = cont_inputs
            session['int_inputs'] = int_inputs
            session['num_epochs'] = num_epochs

            if 'advanced_options' in request.form:
                session['advanced_options'] = True
                return redirect(url_for('create.tabular_advanced'))
            elif 'specify_output' in request.form:
                session['advanced_options'] = False
                return redirect(url_for('create.specify_output'))
            else:
                raise Exception('Invalid Request')

    cols = cu.parse_tabular_cols(run_id=session['run_id'])
    return render_template('create/tabular.html',
                           title=session['title'],
                           cols=cols,
                           default_num_epochs='{:,d}'.format(
                               cs.TABULAR_DEFAULT_NUM_EPOCHS),
                           max_num_epochs=cs.TABULAR_MAX_NUM_EPOCHS)
Esempio n. 4
0
def specify_output():
    if session['format'] == 'Tabular':
        dep_choices = cu.parse_tabular_dep(run_id=session['run_id'],
                                           dep_var=session['dep_var'])
    else:  # Image
        dep_choices = session['dep_choices']

    if request.method == 'POST':
        if 'cancel' in request.form:
            db.clean_run(run_id=session['run_id'])
            return redirect(url_for('index'))

        if 'back' in request.form:
            if session['format'] == 'Tabular':
                if session['advanced_options']:
                    return redirect(url_for('create.tabular_advanced'))
                else:
                    return redirect(url_for('create.tabular'))
            else:  # Image
                if session['advanced_options']:
                    return redirect(url_for('create.image_advanced'))
                else:
                    return redirect(url_for('create.image'))

        cu.create_gen_dict(request_form=request.form,
                           directory=cs.RUN_FOLDER,
                           username=g.user['username'],
                           title=session['title'])
        return redirect(url_for('create.success'))

    return render_template('create/specify_output.html',
                           title=session['title'],
                           dep_var=session['dep_var'],
                           dep_choices=dep_choices,
                           max_examples_per_class='{:,d}'.format(
                               cs.MAX_EXAMPLE_PER_CLASS))
Esempio n. 5
0
def success():
    if request.method == 'POST':
        if 'cancel' in request.form:
            db.clean_run(run_id=session['run_id'])
            return redirect(url_for('index'))

        cmd = 'redis-cli ' + (
            '-h redis-server ' if cs.DOCKERIZED else
            '') + 'ping'  # Check to make sure redis server is up
        if os.system(cmd) != 0:
            db.query_set_status(run_id=session['run_id'],
                                status_id=cs.STATUS_DICT['Error'])
            e = 'Redis server is not set up to handle requests.'
            logger.exception('Error: %s', e)
            raise NameError('Error: ' + e)

        db.query_set_status(run_id=session['run_id'],
                            status_id=cs.STATUS_DICT['Training kicked off'])
        if session['format'] == 'Tabular':
            # Load advanced settings (or defaults)
            bs = session['tabular_batch_size'] if session[
                'advanced_options'] else cs.TABULAR_DEFAULT_BATCH_SIZE
            tabular_init_params = session['tabular_init_params'] if session[
                'advanced_options'] else cs.TABULAR_CGAN_INIT_PARAMS
            tabular_eval_freq = session['tabular_eval_freq'] if session[
                'advanced_options'] else cs.TABULAR_DEFAULT_EVAL_FREQ
            tabular_eval_params = session['tabular_eval_params'] if session[
                'advanced_options'] else cs.TABULAR_EVAL_PARAM_GRID
            tabular_eval_folds = session['tabular_eval_folds'] if session[
                'advanced_options'] else cs.TABULAR_EVAL_FOLDS
            tabular_test_size = session['tabular_test_size'] if session[
                'advanced_options'] else cs.TABULAR_DEFAULT_TEST_SIZE

            # Commence tabular run
            make_dataset = current_app.task_queue.enqueue(
                'CSDGAN.pipeline.data.make_tabular_dataset.make_tabular_dataset',
                args=(session['run_id'], g.user['username'], session['title'],
                      session['dep_var'], session['cont_inputs'],
                      session['int_inputs'], tabular_test_size))
            train_model = current_app.task_queue.enqueue(
                'CSDGAN.pipeline.train.train_tabular_model.train_tabular_model',
                args=(session['run_id'], g.user['username'], session['title'],
                      session['num_epochs'], bs, tabular_init_params,
                      tabular_eval_freq, tabular_eval_params,
                      tabular_eval_folds),
                depends_on=make_dataset,
                job_timeout=-1)
            generate_data = current_app.task_queue.enqueue(
                'CSDGAN.pipeline.generate.generate_tabular_data.generate_tabular_data',
                args=(session['run_id'], g.user['username'], session['title']),
                depends_on=train_model)
        else:  # Image
            # Load advanced settings (or defaults)
            image_init_params = session['image_init_params'] if session[
                'advanced_options'] else cs.IMAGE_CGAN_INIT_PARAMS
            image_eval_freq = session['image_eval_freq'] if session[
                'advanced_options'] else cs.IMAGE_DEFAULT_EVAL_FREQ

            # Commence image run
            make_dataset = current_app.task_queue.enqueue(
                'CSDGAN.pipeline.data.make_image_dataset.make_image_dataset',
                args=(session['run_id'], g.user['username'], session['title'],
                      session['folder'], session['bs'], session['x_dim'],
                      session['splits']))
            train_model = current_app.task_queue.enqueue(
                'CSDGAN.pipeline.train.train_image_model.train_image_model',
                args=(session['run_id'], g.user['username'], session['title'],
                      session['num_epochs'], session['bs'], session['nc'],
                      session['num_channels'], image_init_params,
                      image_eval_freq),
                depends_on=make_dataset,
                job_timeout=-1)
            generate_data = current_app.task_queue.enqueue(
                'CSDGAN.pipeline.generate.generate_image_data.generate_image_data',
                args=(session['run_id'], g.user['username'], session['title']),
                depends_on=train_model)
        db.query_add_job_ids(run_id=session['run_id'],
                             data_id=make_dataset.get_id(),
                             train_id=train_model.get_id(),
                             generate_id=generate_data.get_id())
        logger.info('User #{} ({}) kicked off a {} Run #{} ({})'.format(
            g.user['id'], g.user['username'], session['format'],
            session['run_id'], session['title']))
        return redirect(url_for('index'))

    return render_template('create/success.html', title=session['title'])
Esempio n. 6
0
def image_advanced():
    if request.method == 'POST':
        if 'cancel' in request.form:
            db.clean_run(run_id=session['run_id'])
            return redirect(url_for('index'))

        if 'back' in request.form:
            return redirect(url_for('create.image'))

        error = None
        image_init_params = {}

        try:
            image_init_params['netG_lr'] = float(
                request.form['netG_lr']) if request.form[
                    'netG_lr'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['netG_lr']
            image_init_params['netD_lr'] = float(
                request.form['netD_lr']) if request.form[
                    'netD_lr'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['netD_lr']
            image_init_params['netE_lr'] = float(
                request.form['netE_lr']) if request.form[
                    'netE_lr'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['netE_lr']
        except ValueError:
            error = 'Please input a valid number for learning rates.'

        if error:
            flash(error)
        else:
            image_init_params['netG_beta1'] = float(
                request.form['netG_beta1']) if request.form[
                    'netG_beta1'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'netG_beta1']
            image_init_params['netG_beta2'] = float(
                request.form['netG_beta2']) if request.form[
                    'netG_beta2'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'netG_beta2']
            image_init_params['netD_beta1'] = float(
                request.form['netD_beta1']) if request.form[
                    'netD_beta1'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'netD_beta1']
            image_init_params['netD_beta2'] = float(
                request.form['netD_beta2']) if request.form[
                    'netD_beta2'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'netD_beta2']
            image_init_params['netE_beta1'] = float(
                request.form['netE_beta1']) if request.form[
                    'netE_beta1'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'netE_beta1']
            image_init_params['netE_beta2'] = float(
                request.form['netE_beta2']) if request.form[
                    'netE_beta2'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'netE_beta2']
            image_init_params['netG_wd'] = float(
                request.form['netG_wd']) if request.form[
                    'netG_wd'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['netG_wd']
            image_init_params['netD_wd'] = float(
                request.form['netD_wd']) if request.form[
                    'netD_wd'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['netD_wd']
            image_init_params['netE_wd'] = float(
                request.form['netE_wd']) if request.form[
                    'netE_wd'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['netE_wd']
            image_init_params['label_noise'] = float(
                request.form['label_noise']) if request.form[
                    'label_noise'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'label_noise']
            image_init_params['discrim_noise'] = float(
                request.form['discrim_noise']) if request.form[
                    'discrim_noise'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'discrim_noise']
            image_init_params['nz'] = int(
                request.form['nz']
            ) if request.form['nz'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['nz']
            image_init_params['sched_netG'] = int(
                request.form['sched_netG']) if request.form[
                    'sched_netG'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'sched_netG']
            image_init_params['netG_nf'] = int(
                request.form['netG_nf']) if request.form[
                    'netG_nf'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['netG_nf']
            image_init_params['netD_nf'] = int(
                request.form['netD_nf']) if request.form[
                    'netD_nf'] != '' else cs.IMAGE_CGAN_INIT_PARAMS['netD_nf']
            image_init_params['fake_data_set_size'] = int(
                request.form['fake_data_set_size']) if request.form[
                    'fake_data_set_size'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'fake_data_set_size']
            image_init_params['eval_num_epochs'] = int(
                request.form['eval_num_epochs']) if request.form[
                    'eval_num_epochs'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                        'eval_num_epochs']
            image_init_params['early_stopping_patience'] = int(
                request.form['early_stopping_patience']
            ) if request.form[
                'early_stopping_patience'] != '' else cs.IMAGE_CGAN_INIT_PARAMS[
                    'early_stopping_patience']

            if 'label_noise_linear_anneal' not in request.form:
                image_init_params[
                    'label_noise_linear_anneal'] = cs.IMAGE_CGAN_INIT_PARAMS[
                        'label_noise_linear_anneal']
            elif request.form['label_noise_linear_anneal'] == 'True':
                image_init_params['label_noise_linear_anneal'] = True
            else:
                image_init_params['label_noise_linear_anneal'] = False

            if 'discrim_noise_linear_anneal' not in request.form:
                image_init_params[
                    'discrim_noise_linear_anneal'] = cs.IMAGE_CGAN_INIT_PARAMS[
                        'discrim_noise_linear_anneal']
            elif request.form['discrim_noise_linear_anneal'] == 'True':
                image_init_params['discrim_noise_linear_anneal'] = True
            else:
                image_init_params['discrim_noise_linear_anneal'] = False

            image_eval_freq = int(
                request.form['image_eval_freq']) if request.form[
                    'image_eval_freq'] != '' else cs.IMAGE_DEFAULT_EVAL_FREQ

            session['image_init_params'] = image_init_params
            session['image_eval_freq'] = image_eval_freq

            session['advanced_options'] = True
            return redirect(url_for('create.specify_output'))

    return render_template('create/image_advanced.html',
                           title=session['title'],
                           default_params=cs.IMAGE_CGAN_INIT_PARAMS,
                           default_eval_freq=cs.IMAGE_DEFAULT_EVAL_FREQ)
Esempio n. 7
0
def image():
    x_dim, num_channels, summarized_df = cu.parse_image_folder(
        username=g.user['username'],
        title=session['title'],
        file=session['folder'])
    if request.method == 'POST':
        if 'cancel' in request.form:
            db.clean_run(run_id=session['run_id'])
            return redirect(url_for('index'))

        if 'back' in request.form:
            db.clean_run(run_id=session['run_id'])
            return render_template('create/create.html',
                                   available_formats=cs.AVAILABLE_FORMATS)

        dep_choices = list(summarized_df.index)
        nc = len(dep_choices)
        dep_var = cs.IMAGE_DEFAULT_CLASS_NAME if request.form[
            'dep_var'] == '' else request.form['dep_var']
        x_dim = x_dim if request.form['x_dim_width'] == '' or request.form[
            'x_dim_length'] == '' else (int(request.form['x_dim_width']),
                                        int(request.form['x_dim_length']))
        bs = cs.IMAGE_DEFAULT_BATCH_SIZE if request.form['bs'] == '' else int(
            request.form['bs'])

        if all((request.form['splits_0'] == '', request.form['splits_1'] == '',
                request.form['splits_2'] == '')):
            splits = cs.IMAGE_DEFAULT_TRAIN_VAL_TEST_SPLITS
        else:
            splits = request.form['splits_0'], request.form[
                'splits_1'], request.form['splits_2']

        num_epochs = cs.IMAGE_DEFAULT_NUM_EPOCHS if request.form[
            'num_epochs'] == '' else int(request.form['num_epochs'])
        error = cu.validate_image_choices(dep_var=dep_var,
                                          x_dim=x_dim,
                                          bs=bs,
                                          splits=splits,
                                          num_epochs=num_epochs,
                                          num_channels=num_channels)
        if error:
            flash(error)
        else:
            db.query_add_depvar(run_id=session['run_id'], depvar=dep_var)
            session['dep_choices'] = dep_choices
            session['dep_var'] = dep_var
            session['nc'] = nc
            session['x_dim'] = x_dim
            session['bs'] = bs
            session['splits'] = splits
            session['num_epochs'] = num_epochs
            session['num_channels'] = num_channels

            if 'advanced_options' in request.form:
                session['advanced_options'] = True
                return redirect(url_for('create.image_advanced'))
            elif 'specify_output' in request.form:
                session['advanced_options'] = False
                return redirect(url_for('create.specify_output'))
            else:
                raise Exception('Invalid Request')

    return render_template(
        'create/image.html',
        title=session['title'],
        default_x_dim=x_dim,
        max_x_dim=cs.IMAGE_MAX_X_DIM,
        summarized_df=summarized_df,
        default_dep_var=cs.IMAGE_DEFAULT_CLASS_NAME,
        default_bs=cs.IMAGE_DEFAULT_BATCH_SIZE,
        max_bs=cs.IMAGE_MAX_BS,
        default_splits=cs.IMAGE_DEFAULT_TRAIN_VAL_TEST_SPLITS,
        default_num_epochs=cs.IMAGE_DEFAULT_NUM_EPOCHS,
        max_num_epochs=cs.IMAGE_MAX_NUM_EPOCHS)
Esempio n. 8
0
def tabular_advanced():
    if request.method == 'POST':
        if 'cancel' in request.form:
            db.clean_run(run_id=session['run_id'])
            return redirect(url_for('index'))

        if 'back' in request.form:
            return redirect(url_for('create.tabular'))

        error = None
        tabular_init_params = {}
        tabular_eval_params = {}

        try:
            tabular_init_params['netG_lr'] = float(
                request.form['netG_lr']
            ) if request.form[
                'netG_lr'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['netG_lr']
            tabular_init_params['netD_lr'] = float(
                request.form['netD_lr']
            ) if request.form[
                'netD_lr'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['netD_lr']
        except ValueError:
            error = 'Please input a valid number for learning rates.'

        try:
            tabular_eval_params['tol'] = [
                float(request.form['tol'])
            ] if request.form['tol'] != '' else cs.TABULAR_EVAL_PARAM_GRID[
                'tol']
        except ValueError:
            error = 'Please input a valid number for tolerance.'

        if error:
            flash(error)
        else:
            tabular_init_params['nz'] = int(
                request.form['nz']) if request.form[
                    'nz'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['nz']
            tabular_init_params['netG_beta1'] = float(
                request.form['netG_beta1']) if request.form[
                    'netG_beta1'] != '' else cs.TABULAR_CGAN_INIT_PARAMS[
                        'netG_beta1']
            tabular_init_params['netG_beta2'] = float(
                request.form['netG_beta2']) if request.form[
                    'netG_beta2'] != '' else cs.TABULAR_CGAN_INIT_PARAMS[
                        'netG_beta2']
            tabular_init_params['netD_beta1'] = float(
                request.form['netD_beta1']) if request.form[
                    'netD_beta1'] != '' else cs.TABULAR_CGAN_INIT_PARAMS[
                        'netD_beta1']
            tabular_init_params['netD_beta2'] = float(
                request.form['netD_beta2']) if request.form[
                    'netD_beta2'] != '' else cs.TABULAR_CGAN_INIT_PARAMS[
                        'netD_beta2']
            tabular_init_params['netG_wd'] = float(
                request.form['netG_wd']
            ) if request.form[
                'netG_wd'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['netG_wd']
            tabular_init_params['netD_wd'] = float(
                request.form['netD_wd']
            ) if request.form[
                'netD_wd'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['netD_wd']
            tabular_init_params['label_noise'] = float(
                request.form['label_noise']) if request.form[
                    'label_noise'] != '' else cs.TABULAR_CGAN_INIT_PARAMS[
                        'label_noise']
            tabular_init_params['discrim_noise'] = float(
                request.form['discrim_noise']) if request.form[
                    'discrim_noise'] != '' else cs.TABULAR_CGAN_INIT_PARAMS[
                        'discrim_noise']
            tabular_init_params['nz'] = int(
                request.form['nz']) if request.form[
                    'nz'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['nz']
            tabular_init_params['sched_netG'] = int(
                request.form['sched_netG']) if request.form[
                    'sched_netG'] != '' else cs.TABULAR_CGAN_INIT_PARAMS[
                        'sched_netG']
            tabular_init_params['netG_H'] = int(
                request.form['netG_H']) if request.form[
                    'netG_H'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['netG_H']
            tabular_init_params['netD_H'] = int(
                request.form['netD_H']) if request.form[
                    'netD_H'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['netD_H']

            tabular_eval_params['C'] = [
                float(request.form['C'])
            ] if request.form['C'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['C']
            tabular_eval_params['l1_ratio'] = [
                float(request.form['l1_ratio'])
            ] if request.form[
                'l1_ratio'] != '' else cs.TABULAR_CGAN_INIT_PARAMS['l1_ratio']

            if 'label_noise_linear_anneal' not in request.form:
                tabular_init_params[
                    'label_noise_linear_anneal'] = cs.TABULAR_CGAN_INIT_PARAMS[
                        'label_noise_linear_anneal']
            elif request.form['label_noise_linear_anneal'] == 'True':
                tabular_init_params['label_noise_linear_anneal'] = True
            else:
                tabular_init_params['label_noise_linear_anneal'] = False

            if 'discrim_noise_linear_anneal' not in request.form:
                tabular_init_params[
                    'discrim_noise_linear_anneal'] = cs.TABULAR_CGAN_INIT_PARAMS[
                        'discrim_noise_linear_anneal']
            elif request.form['discrim_noise_linear_anneal'] == 'True':
                tabular_init_params['discrim_noise_linear_anneal'] = True
            else:
                tabular_init_params['discrim_noise_linear_anneal'] = False

            tabular_eval_freq = int(
                request.form['tabular_eval_freq']
            ) if request.form[
                'tabular_eval_freq'] != '' else cs.TABULAR_DEFAULT_EVAL_FREQ
            tabular_test_size = float(
                request.form['ts']
            ) if request.form['ts'] != '' else cs.TABULAR_DEFAULT_TEST_SIZE
            tabular_batch_size = int(
                request.form['bs']
            ) if request.form['bs'] != '' else cs.TABULAR_DEFAULT_BATCH_SIZE
            tabular_eval_folds = int(
                request.form['cv']
            ) if request.form['cv'] != '' else cs.TABULAR_EVAL_FOLDS

            session['tabular_init_params'] = tabular_init_params
            session['tabular_eval_params'] = tabular_eval_params
            session['tabular_eval_freq'] = tabular_eval_freq
            session['tabular_test_size'] = tabular_test_size
            session['tabular_batch_size'] = tabular_batch_size
            session['tabular_eval_folds'] = tabular_eval_folds

            session['advanced_options'] = True
            return redirect(url_for('create.specify_output'))

    return render_template('create/tabular_advanced.html',
                           title=session['title'],
                           default_params=cs.TABULAR_CGAN_INIT_PARAMS,
                           default_test_size=cs.TABULAR_DEFAULT_TEST_SIZE,
                           default_batch_size=cs.TABULAR_DEFAULT_BATCH_SIZE,
                           default_eval_param=cs.TABULAR_EVAL_PARAM_GRID,
                           default_eval_folds=cs.TABULAR_EVAL_FOLDS)