nbertagnolli

## mnist_dropout.ipynb

      
              1 file
            
          
              0 forks
            
          
              0 comments
            
          
              0 stars
            
          
                nbertagnolli
                / mnist_dropout.ipynb
            
            
              Last active
              January 7, 2022 23:55
            
              
                A notebook describing how to implement dropout in pytorch from scratch. For the associated article https://medium.com/p/67f08a87ccff/edit.
              
          
      View mnist_dropout.ipynb
  
    
      Sorry, something went wrong. Reload?
      Sorry, we cannot display this file.
      Sorry, this file is invalid so it cannot be displayed.
      
          Viewer requires iframe.
      
    
## true_dropout.py
class TrueDropout(torch.nn.Module):

    def __init__(self, p: float=0.5):
        super(TrueDropout, self).__init__()
        self.p = p
        if self.p < 0 or self.p > 1:
            raise ValueError("p must be a probability")

    def forward(self, x):
        if self.training:

## mlp_mnist.py
class MNISTModel(torch.nn.Module):

    def __init__(self):
        super(MNISTModel, self).__init__()
        self.layer_1 = nn.Linear(28 * 28, 512)
        self.layer_2 = nn.Linear(512, 512)
        self.layer_3 = nn.Linear(512, 10)
        self.dropout = Dropout(.5)

    def forward(self, x):

## dropout_basic.py
class Dropout(torch.nn.Module):

    def __init__(self, p: float=0.5):
        super(Dropout, self).__init__()
        self.p = p
        if self.p < 0 or self.p > 1:
            raise ValueError("p must be a probability")

    def forward(self, x):
        if self.training:

## Dockerfile.ui
FROM python:3.8

RUN mkdir /ui
COPY requirements.txt /ui
COPY . /ui
WORKDIR /ui

RUN apt-get update && apt-get install -y build-essential libapr1-dev libssl-dev openssh-client

# Install all necessary libraries into a pyenv environment

## Dockerfile.api
FROM python:3.8

# Setup the api directory for your project code
RUN mkdir /api
COPY . /api
COPY requirements.txt /api
WORKDIR /api


# Install all necessary libraries into a pyenv environment

## ds-app-docker-compose.yml
version: '3.8'
services:
    movie-app-api:
        image: movie-app-api
        ports:
            - 8000:8000
        volumes:
            - ./api:/api
        networks:
            - bridge_network

## streamlit_app.py
import streamlit as st
import requests
import json

url = "http://localhost:8000/review/predict"

# Write out the header for the page
st.write("# Movie Review Analyzer")

# Accept user text input

## fast_api_post_example.py
from fastapi import FastAPI
from pydantic import BaseModel

# Initialize the FastAPI App
app = FastAPI()

# Create a Data Model so FastAPI can read the data from the request.
class Review(BaseModel):
    text: str

## sklearn_lambda_handler.py
import boto3
import json
import os
import pickle


s3 = boto3.resource("s3")
BUCKET_NAME = "nic-sklearn-models"
	class TrueDropout(torch.nn.Module):

	def __init__(self, p: float=0.5):
	super(TrueDropout, self).__init__()
	self.p = p
	if self.p < 0 or self.p > 1:
	raise ValueError("p must be a probability")

	def forward(self, x):
	if self.training:
	class MNISTModel(torch.nn.Module):

	def __init__(self):
	super(MNISTModel, self).__init__()
	self.layer_1 = nn.Linear(28 * 28, 512)
	self.layer_2 = nn.Linear(512, 512)
	self.layer_3 = nn.Linear(512, 10)
	self.dropout = Dropout(.5)

	def forward(self, x):
	class Dropout(torch.nn.Module):

	def __init__(self, p: float=0.5):
	super(Dropout, self).__init__()
	self.p = p
	if self.p < 0 or self.p > 1:
	raise ValueError("p must be a probability")

	def forward(self, x):
	if self.training:
	FROM python:3.8

	RUN mkdir /ui
	COPY requirements.txt /ui
	COPY . /ui
	WORKDIR /ui

	RUN apt-get update && apt-get install -y build-essential libapr1-dev libssl-dev openssh-client

	# Install all necessary libraries into a pyenv environment
	FROM python:3.8

	# Setup the api directory for your project code
	RUN mkdir /api
	COPY . /api
	COPY requirements.txt /api
	WORKDIR /api


	# Install all necessary libraries into a pyenv environment
	version: '3.8'
	services:
	movie-app-api:
	image: movie-app-api
	ports:
	- 8000:8000
	volumes:
	- ./api:/api
	networks:
	- bridge_network
	import streamlit as st
	import requests
	import json

	url = "http://localhost:8000/review/predict"

	# Write out the header for the page
	st.write("# Movie Review Analyzer")

	# Accept user text input
	from fastapi import FastAPI
	from pydantic import BaseModel

	# Initialize the FastAPI App
	app = FastAPI()

	# Create a Data Model so FastAPI can read the data from the request.
	class Review(BaseModel):
	text: str
	import boto3
	import json
	import os
	import pickle


	s3 = boto3.resource("s3")
	BUCKET_NAME = "nic-sklearn-models"