| # ! pip install slackclient | |
| # get your token: https://api.slack.com/custom-integrations/legacy-tokens | |
| # composing messages: https://api.slack.com/messaging/composing | |
| import os | |
| from slack import WebClient | |
| sc = WebClient(token='<your_token>', run_async=True) | |
| sc.chat_postMessage(channel='#tasks-notifications', text='', icon_emoji=':smile:', username='') # username is the "sender" name |
Please check the direct links to the feedback form for each talk!
Rate and provide us your feedback to help us running a better conference, and to help the speaker improve.
Ravioli vs Pelmeni software architecture (microservices vs services) by Anastasiia Tymoshchuk: https://python.it/feedback-1515
TemPy! Una alternativa al templating tradizionale usando solo Python by Federico Cerchiari:
| def get_callbacks(self, model_prefix='Model'): | |
| """ | |
| Creates a list of callbacks that can be used during training to create a | |
| snapshot ensemble of the model. | |
| Args: | |
| model_prefix: prefix for the filename of the weights. | |
| Returns: list of 3 callbacks [ModelCheckpoint, LearningRateScheduler, | |
| SnapshotModelCheckpoint] which can be provided to the 'fit' function | |
| """ | |
| if not os.path.exists('weights/'): |
| def clr(self): | |
| cycle = np.floor(1 + self.clr_iterations / (2 * self.step_size)) | |
| x = np.abs(self.clr_iterations / self.step_size - 2 * cycle + 1) | |
| if self.scale_mode == 'cycle': | |
| return self.base_lr + (self.max_lr - self.base_lr) * \ | |
| np.maximum(0, (1 - x)) * self.scale_fn(cycle) | |
| else: | |
| return self.base_lr + (self.max_lr - self.base_lr) * \ | |
| np.maximum(0, (1 - x)) * self.scale_fn(self.clr_iterations) |
| def step_decay(epoch): | |
| initial_lrate = 0.1 | |
| drop = 0.5 | |
| epochs_drop = 10.0 | |
| lrate = initial_lrate * math.pow(drop, math.floor((1+epoch)/epochs_drop)) | |
| return lrate | |
| # ... | |
| lrate = LearningRateScheduler(step_decay) |
| # Compile model | |
| epochs = 50 | |
| learning_rate = 0.1 | |
| decay_rate = learning_rate / epochs | |
| momentum = 0.8 | |
| sgd = SGD(lr=learning_rate, momentum=momentum, decay=decay_rate, nesterov=False) | |
| model.compile(loss='binary_crossentropy', optimizer=sgd, metrics=['accuracy']) |