A Learning Curve is a plot of the prediction accuracy/error versus the training set size. It shows us how the model gets better at predicting as we increase the number of instances used.
Einstein Vision has an API named Get Model Learning Curve that returns the metrics for each epoch in a model. These metrics show you the f1 score, accuracy, confusion matrix, test accuracy, and so on for each training iteration performed to create the model.
The CURL request is as follows:
curl -X GET -H "Authorization: Bearer <TOKEN>" -H "Cache-Control: no-cache" https://api.einstein.ai/v1/vision/models/7JWKAHFKRXTMNLJCEF2DR5CJ46QU/lc
The Response structure consists of detailed metrics data and epoch results. The response format is as follows:
{
"object": "list",
"data": [
{
"epoch": 1,
"metricsData": {
"f1": [
0.7499999999999999,
0.8333333333333333
],
"labels": [
"Roses",
"Lilies"
],
"testAccuracy": 0.8,
"trainingLoss": 0.2573,
"confusionMatrix": [
[
3,
2
],
[
0,
5
]
],
"trainingAccuracy": 0.8809
},
"epochResults": [
{
"exampleName": "601053842.jpg-Rose1",
"expectedLabel": "Roses",
"predictedLabel": "Roses"
},
{
"exampleName": "578339672.jpg-Mountains",
"expectedLabel": "Roses",
"predictedLabel": "Roses"
},
{
"epoch": 2,
"metricsData": {
"f1": [
0.7499999999999999,
0.8333333333333333
],
"labels": [
"Roses",
"Lilies"
],
........
],
"object": "learningcurve"
}
]
}