File size: 3,193 Bytes
97037a8 d23d2f6 3ab8618 97037a8 cfaffad 97037a8 cfaffad 97037a8 cfaffad 20eaaec |
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 |
---
language:
- en
tags:
- text generation
- pytorch
- causal-lm
license: apache-2.0
datasets:
- custom
widget:
- text: "style: Pilsner\nbatch_size: 20\nefficiency: 75\nboil_size:"
example_title: "Pilsener"
- text: "style: IPA\nbatch_size: 20\nefficiency: 75\nboil_size:"
example_title: "IPA"
- text: "style: Scottish Ale\nbatch_size: 20\nefficiency: 75\nboil_size:"
example_title: "Scottish Ale"
inference:
parameters:
do_sample: true
top_k: 10
top_p: 0.99
max_length: 500
---
# GPT-Neo 125M finetuned with beer recipes
## Model Description
GPT-Neo 125M is a transformer model based on EleutherAI's replication of the GPT-3 architecture https://huggingface.co/EleutherAI/gpt-neo-125M.
It generates recipes for brewing beer in a YAML-like format which can be easily used for different purposes.
## Training data
This model was trained on a custom dataset of ~ 76,800 beer recipes from the internet. It includes recipes for the following
styles of beer:
* Strong American Ale
* Pale American Ale
* India Pale Ale (IPA)
* Standard American Beer
* Stout
* English Pale Ale
* IPA
* American Porter and Stout
* Sour Ale
* Irish Beer
* Strong British Ale
* Belgian and French Ale
* German Wheat and Rye Beer
* Czech Lager
* Spice/Herb/Vegetable Beer
* Specialty Beer
* American Ale
* Pilsner
* Belgian Ale
* Strong Belgian Ale
* Bock
* Brown British Beer
* German Wheat Beer
* Fruit Beer
* Amber Malty European Lager
* Pale Malty European Lager
* British Bitter
* Amber and Brown American Beer
* Light Hybrid Beer
* Pale Commonwealth Beer
* American Wild Ale
* European Amber Lager
* Belgian Strong Ale
* International Lager
* Amber Bitter European Lager
* Light Lager
* Scottish and Irish Ale
* European Sour Ale
* Trappist Ale
* Strong European Beer
* Porter
* Historical Beer
* Pale Bitter European Beer
* Amber Hybrid Beer
* Smoke Flavored/Wood-Aged Beer
* Spiced Beer
* Dark European Lager
* Alternative Fermentables Beer
* Mead
* Strong Ale
* Dark British Beer
* Scottish Ale
* Smoked Beer
* English Brown Ale
* Dark Lager
* Cider or Perry
* Wood Beer
### How to use
You can use this model directly with a pipeline for text generation. This example generates a different recipe each time it's run:
```py
>>> from transformers import pipeline
>>> generator = pipeline('text-generation', model='b3ck1/gpt-neo-125M-finetuned-beer-recipes')
>>> generator("style: Pilsner\nbatch_size: 20\nefficiency: 75\nboil_size:", do_sample=True, min_length=50, max_length=500)
>>> print(output[0]['generated_text'])
style: Pilsner
batch_size: 20
efficiency: 70
boil_size: 24
boil_time: 60
fermentables:
- name: Pale Ale
type: Grain
amount: 6.5
hops:
- name: Saaz
alpha: 3.5
use: Boil
time: 60
amount: 0.06
- name: Saaz
alpha: 3.5
use: Boil
time: 30
amount: 0.06
- name: Saaz
alpha: 3.5
use: Boil
time: 10
amount: 0.06
- name: Saaz
alpha: 3.5
use: Boil
time: 0
amount: 0.06
yeasts:
- name: Safale - American Ale Yeast US-05
amount: 0.11
min_temperature: 12
max_temperature: 25
primary_temp: null
mash_steps:
- step_temp: 65
step_time: 60
miscs: []
```
### See this model in action
This model was used to build https://beerai.net. |