File size: 980 Bytes
666c35a
 
bf96279
 
 
 
 
 
 
f8a68c6
bf96279
 
5ed6b47
666c35a
876cc8e
 
 
 
 
 
 
 
 
 
 
 
7485c52
876cc8e
 
7485c52
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
---
license: other
widget:
- text: "There was a ghost"
  example_title: "First Prompt used in video"
- text: "I was playing Terraria but then"
  example_title: "Second prompt used in video"
inference:
  parameters:
    temperature: 0.65
    repetition_penalty: 1.1
    min_length: 128
    max_length: 468
---

This is the model trained for this video:

https://www.youtube.com/watch?v=OEPL5Tm3mmQ

Due to hardware limitations, I trained this model with only a batch size of 2. (I know this isn't ideal).

The quality of the model may be affected.

After training was complete, the best model according to a hold-out set was used.

This model was trained using a filtered version of this dataset:
 
https://www.kaggle.com/datasets/thomaskonstantin/3500-popular-creepypastas

This dataset had a lot of blank entries and missing text.

Please subscribe to my YouTube Channel for bad quality videos and poorly trained models. 

https://www.youtube.com/channel/UCLXxfueCPZRZnyGFWJ07uqA