In this project, we generate our own Simpsons TV scripts using RNNs. We are using part of the Simpsons dataset of scripts from 27 seasons. The Neural Network we build will generate a new TV script for a scene at Moe's Tavern.
The original data set is downloaded from the Simpsons dataset of scripts from 27 seasons. It contains:
Roughly the number of unique words: 11492
Number of scenes: 262
Average number of sentences in each scene: 15.248091603053435
Number of lines: 4257
Average number of words in each line: 11.50434578341555
Here is an example that the RNN model generated after training 60 epochs (it achieved training loss 0.165). The prime word is moe_szyslak
moe_szyslak: ah, she got homesick for her old life, diving for tourist pennies in a micronesian swamp.
homer_simpson: oh, cold shoulder, right! andalay!
moe_szyslak:(too big) hanh?!!!
all:(ad lib) oh, sweet mother fat tony.
moe_szyslak:(sighs) i think
barney_gumble: uh-oh. my wife was the last of the family-owned teams.
lenny_leonard: i tried to return my season ticket, but they wouldn't an day, ever will the pigs.
wayne: i'm really more like me, and get to dinner.
walther_hotenhoffer: nein... nein... nein... nein... nein... we got the cash. my clientele-- let it ring; /mr. plow is a ticket. my wife is, moe. i don't know nothin' about how i do anything.
moe_szyslak: i had to get edna out of springfield, and that's it?
moe_szyslak: no, i tells ya, homer.
homer_simpson:(looking at watch)
- Clone the repository and navigate to the downloaded folder.
https://github.com/TetsumichiUmada/tv_script_generation.git
cd tv_script_generation
- Install required packages.
pip install -r requirements.txt
- Open the notebook.
jupyter notebook tv_script_generation.ipynb