GPT2-Rick-n-Morty-with-SpongeBob

This project generates Rick and Morty with SpongeBob script using GPT-2 model.

Fine tuning data 1: Kaggle

Fine tuning data 2: Kaggle

Model download: Hugging face

Fine-tuning notebook: colab

Model information

Base model: e-tony/gpt2-rnm
Epoch: 2
Train runtime: 790.0612 secs
Loss: 2.8569

How to use:

* First, Choose Rick and Morty or SpongeBob character name.
* Second, Fill what the character will say in text. This will be base of script.
* And then, Fill number in length. Text is created as long as "length". I recommend between 100 and 300.
* If length is so big, generate time will be long.

Post parameter

name: The Rick and Morty and SpongeBob character name.
text: The base of script.
length: The size of generated text.

Output foramt

{"0": [[character name, dialog], [character name, dialog], ...]}

Image reference

static/IMG_Reference.md

* With CLI *

Input example

curl -X POST "https://master-gpt2-rick-n-morty-with-sponge-bob-fpem123.endpoint.ainize.ai/gen" -H "accept: application/json" -H "Content-Type: multipart/form-data" -F "name=Rick" -F "text=Morty! SpongeBob!" -F "length=100"

Output example

{
  "0": [
    [
      "Rick",
      " Morty! SpongeBob! You're the guy!"
    ],
    [
      "SpongeBob",
      " Oh, hi, Morty!"
    ],
    [
      "Rick",
      " I need to hear the story for you because here it is."
    ],
    [
      "Morty",
      " [sighs and falls on his bed playing Bubble Bass music, crying] My goodness, I wish I could live in peace for the rest of my life! [laughs] [walks away]"
    ],
    [
      "Rick",
      " Well, that's not all. [SpongeBob and Rick get into bed as they are"
    ]
  ]
}

* With swagger *

API page: Ainize

* With a Demo *

Demo page: End-point

---