Source code for baselines.mix_of_exps

import copy
import math
from pyexpat import model
import torch
import numpy as np
import torch.nn as nn
import torch.nn.functional as F
import argparse
import os
import random
import shutil
import time
import torch.utils.data as data
import sys
import pickle
import logging
from tqdm import tqdm

sys.path.append("..")
from helpers.utils import *
from helpers.metrics import *
from .basemethod import BaseMethod

eps_cst = 1e-8

# This really doesn't work well on no bencmark,



[docs]class MixtureOfExperts(BaseMethod): """Implementation of Madras et al., 2018""" def __init__(self, model, device, plotting_interval=100): self.plotting_interval = plotting_interval self.model = model self.device = device
[docs] def mixtures_of_experts_loss(self, outputs, human_is_correct, labels): """ Implmentation of Mixtures of Experts loss from Madras et al., 2018 """ batch_size = outputs.size()[0] # batch_size human_loss = torch.cuda.FloatTensor(1 - human_is_correct * 1.0) rejector_probability = torch.sigmoid( outputs[:, -1] + eps_cst ) # probability of rejection outputs_class = F.softmax(outputs[:, :-1], dim=1) classifier_loss = -torch.log2( outputs_class[range(batch_size), labels] + eps_cst ) loss = ( classifier_loss * (1 - rejector_probability) + human_loss * rejector_probability ) return torch.sum(loss) / batch_size
[docs] def fit_epoch(self, dataloader, optimizer, verbose=True, epoch=1): """ Fit the model for one epoch """ batch_time = AverageMeter() losses = AverageMeter() top1 = AverageMeter() end = time.time() self.model.train() for batch, (data_x, data_y, hum_preds) in enumerate(dataloader): data_x = data_x.to(self.device) data_y = data_y.to(self.device) hum_preds = hum_preds.to(self.device) m = (hum_preds == data_y) * 1 m = torch.tensor(m).to(self.device) outputs = self.model(data_x) # apply softmax to outputs loss = self.mixtures_of_experts_loss(outputs, m, data_y) optimizer.zero_grad() loss.backward() optimizer.step() prec1 = accuracy(outputs.data, data_y, topk=(1,))[0] losses.update(loss.data.item(), data_x.size(0)) top1.update(prec1.item(), data_x.size(0)) # measure elapsed time batch_time.update(time.time() - end) end = time.time() if torch.isnan(loss): print("Nan loss") logging.warning(f"NAN LOSS") break if verbose and batch % self.plotting_interval == 0: logging.info( "Epoch: [{0}][{1}/{2}]\t" "Time {batch_time.val:.3f} ({batch_time.avg:.3f})\t" "Loss {loss.val:.4f} ({loss.avg:.4f})\t" "Prec@1 {top1.val:.3f} ({top1.avg:.3f})".format( epoch, batch, len(dataloader), batch_time=batch_time, loss=losses, top1=top1, ) )
[docs] def fit( self, dataloader_train, dataloader_val, dataloader_test, epochs, optimizer, lr, verbose=True, test_interval=5, scheduler=None, ): optimizer = optimizer(self.model.parameters(), lr=lr) if scheduler is not None: scheduler = scheduler(optimizer, len(dataloader_train)*epochs) for epoch in tqdm(range(epochs)): self.fit_epoch(dataloader_train, optimizer, verbose, epoch) if verbose and epoch % test_interval == 0: data_test = self.test(dataloader_val) logging.info(compute_deferral_metrics(data_test)) if scheduler is not None: scheduler.step() final_test = self.test(dataloader_test) return compute_deferral_metrics(final_test)
[docs] def test(self, dataloader): defers_all = [] truths_all = [] hum_preds_all = [] rej_score = [] predictions_all = [] # classifier only self.model.eval() with torch.no_grad(): for batch, (data_x, data_y, hum_preds) in enumerate(dataloader): data_x = data_x.to(self.device) data_y = data_y.to(self.device) hum_preds = hum_preds.to(self.device) outputs = self.model(data_x) outputs_soft = F.softmax(outputs[:, :-1], dim=1) _, predicted_class = torch.max(outputs_soft.data, 1) predictions_all.extend(predicted_class.cpu().numpy()) rejector_outputs = torch.sigmoid(outputs[:, -1]) defers_all.extend((rejector_outputs.cpu().numpy() >= 0.5).astype(int)) truths_all.extend(data_y.cpu().numpy()) hum_preds_all.extend(hum_preds.cpu().numpy()) rej_score.extend(rejector_outputs.cpu().numpy()) # convert to numpy defers_all = np.array(defers_all) truths_all = np.array(truths_all) hum_preds_all = np.array(hum_preds_all) predictions_all = np.array(predictions_all) data = { "defers": defers_all, "labels": truths_all, "hum_preds": hum_preds_all, "preds": predictions_all, "rej_score": rej_score, } return data