PHP Search Engine

Dec 1 In-Class Exercise Thread .

Please post your solutions to the Dec 1 In-Class Exercise Thread.

Best,

Chris

Please post your solutions to the Dec 1 In-Class Exercise Thread. Best, Chris

-- Dec 1 In-Class Exercise Thread

The graphs differ slightly because the data is shuffled for each run and the validation data batches used would most likely be different for each epoch/run.

(Edited: 2021-12-01)

((resource:epoch_accuracy (1).png|Resource Description for epoch_accuracy (1).png)) The graphs differ slightly because the data is shuffled for each run and the validation data batches used would most likely be different for each epoch/run. ((resource:run_2021_m_01-16_14_10_train.png|Resource Description for run_2021_m_01-16_14_10_train.png))

-- Dec 1 In-Class Exercise Thread

Epoch Accuracy graph:

How similar are the curves? Why aren't they the same? The curves are similar in a way they are rising towards better accuracy every epoch. The curves are not exactly the same because the weights are initialized differently in every run, also the batches are randomly chosen in every run.

Main Graph:

Epoch Accuracy graph:((resource:Screen Shot 2021-12-01 at 16.24.00.png|Resource Description for Screen Shot 2021-12-01 at 16.24.00.png)) How similar are the curves? Why aren't they the same? The curves are similar in a way they are rising towards better accuracy every epoch. The curves are not exactly the same because the weights are initialized differently in every run, also the batches are randomly chosen in every run. Main Graph: ((resource:Screen Shot 2021-12-01 at 16.24.29.png|Resource Description for Screen Shot 2021-12-01 at 16.24.29.png))

-- Dec 1 In-Class Exercise Thread

The curves are similar in the sense that they go towards higher accuracy in each iteration. However, they are not exactly the same because with each run the data is shuffled (chosen randomly).

(Edited: 2021-12-01)

((resource:Screen Shot 2021-12-01 at 4.24.17 PM.png|Resource Description for Screen Shot 2021-12-01 at 4.24.17 PM.png)) The curves are similar in the sense that they go towards higher accuracy in each iteration. However, they are not exactly the same because with each run the data is shuffled (chosen randomly). ((resource:run_2021_m_02-00_07_32_train.png|Resource Description for run_2021_m_02-00_07_32_train.png))

-- Dec 1 In-Class Exercise Thread

The graphs differ slightly because the data is shuffled for each run and the validation data batches used would most likely be different for each epoch/run.

(Edited: 2021-12-01)

((resource:Screen Shot 2021-12-01 at 5.04.51 PM.png|Resource Description for Screen Shot 2021-12-01 at 5.04.51 PM.png)) The graphs differ slightly because the data is shuffled for each run and the validation data batches used would most likely be different for each epoch/run. ((resource:Screen Shot 2021-12-01 at 4.50.37 PM.png|Resource Description for Screen Shot 2021-12-01 at 4.50.37 PM.png))

-- Dec 1 In-Class Exercise Thread

The validation data batches used are different for each epoch/run. and also data has been shuffled for each run. This is why the graphs differ slightly.

(Edited: 2021-12-01)

((resource:10. tensorboard.JPG|Resource Description for 10. tensorboard.JPG)) ((resource:11. tensorboard.JPG|Resource Description for 11. tensorboard.JPG)) The validation data batches used are different for each epoch/run. and also data has been shuffled for each run. This is why the graphs differ slightly.

-- Dec 1 In-Class Exercise Thread

The graphs are not similar because the data is shuffled for each run and the validation data batches change for every epoch.

(Edited: 2021-12-01)

((resource:Picture2.png|Resource Description for Picture2.png)) ((resource:Picture1.png|Resource Description for Picture1.png)) The graphs are not similar because the data is shuffled for each run and the validation data batches change for every epoch.

-- Dec 1 In-Class Exercise Thread

The curves are similar in that each iteration progresses toward greater precision. They are not identical, though, because the data is scrambled with each iteration

((resource:Screen Shot 2021-12-02 at 6.46.59 PM.png|Resource Description for Screen Shot 2021-12-02 at 6.46.59 PM.png)) ((resource:Screen Shot 2021-12-02 at 6.47.21 PM.png|Resource Description for Screen Shot 2021-12-02 at 6.47.21 PM.png)) The curves are similar in that each iteration progresses toward greater precision. They are not identical, though, because the data is scrambled with each iteration

-- Dec 1 In-Class Exercise Thread

Because the data is shuffled for each run, and the validation data batches utilized are most likely different for each epoch or run, the graphs differ slightly.

((resource:Screen Shot 2021-12-02 at 6.55.38 PM.png|Resource Description for Screen Shot 2021-12-02 at 6.55.38 PM.png)) ((resource:Screen Shot 2021-12-02 at 6.54.55 PM.png|Resource Description for Screen Shot 2021-12-02 at 6.54.55 PM.png)) Because the data is shuffled for each run, and the validation data batches utilized are most likely different for each epoch or run, the graphs differ slightly.

-- Dec 1 In-Class Exercise Thread

Over here Graphs are different because the data is shuffled for each run and the validation data batches change for every epoch.

((resource:Capture.JPG|Resource Description for Capture.JPG)) Over here Graphs are different because the data is shuffled for each run and the validation data batches change for every epoch.