Trying to Apply Lesson 1 & 2 - Blood Cell Identification

Raymond-Wu · June 14, 2018, 9:47pm

Hello all, I’m having trouble with applying what was taught in Lesson 1 & 2 to this Kaggle dataset. I’m getting an error “ValueError: too many values to unpack (expected 2)” and don’t know how to proceed.

Along with what I’ve done thus far.

github.com

RaymondDashWu/Test-Scripts/blob/master/Blood Cell Identification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.imports import *\n",

This file has been truncated. show original

A couple concerns/possibilities I’ve had:

This was a very unbalanced dataset with over half of the items being “Neutrophil”
This dataset is meant for Tensorflow (I’m guessing) since it included labeled XMLs. I’m not sure if that makes any difference to the CNN I’m trying to apply it to
The labeling of the images doesn’t match the CSV. The pictures are a seemingly random string whereas the label.csv counts upward.