Spaces:
Sleeping
Sleeping
File size: 3,006 Bytes
03a8026 9c2c875 03a8026 |
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 |
from transformers import pipeline, set_seed
from transformers import AutoTokenizer
import re
from utils import ext
from utils.ext import pure_comma_separation
from decouple import config
import os
from utils.api import generate_cook_image
from utils.translators.translate_recepie import translate_recepie
from utils.translators.translate_input import translate_input
model_name_or_path = "flax-community/t5-recipe-generation"
task = "text2text-generation"
tokenizer = AutoTokenizer.from_pretrained(
model_name_or_path, cache_dir='./cache')
generator = pipeline(task, model=model_name_or_path,
tokenizer=model_name_or_path)
prefix = "items: "
chef_top = {
"max_length": 512,
"min_length": 64,
"no_repeat_ngram_size": 3,
"do_sample": True,
"top_k": 60,
"top_p": 0.95,
"num_return_sequences": 1,
"return_tensors": True,
"return_text": False
}
chef_beam = {
"max_length": 512,
"min_length": 64,
"no_repeat_ngram_size": 3,
"early_stopping": True,
"num_beams": 5,
"length_penalty": 1.5,
"num_return_sequences": 1
}
generation_kwargs = {
"max_length": 512,
"min_length": 64,
"no_repeat_ngram_size": 3,
"do_sample": True,
"top_k": 60,
"top_p": 0.95
}
def load_api():
api_key = config("API_KEY")
api_id = config("API_ID")
return {"KEY": api_key, "ID": api_id}
def skip_special_tokens_and_prettify(text):
data = {"title": "", "ingredients": [], "directions": []}
text = text + '$'
pattern = r"(\w+:)(.+?(?=\w+:|\$))"
for match in re.findall(pattern, text):
if match[0] == 'title:':
data["title"] = match[1]
elif match[0] == 'ingredients:':
data["ingredients"] = [ing.strip() for ing in match[1].split(',')]
elif match[0] == 'directions:':
data["directions"] = [d.strip() for d in match[1].split('.')]
else:
pass
data["ingredients"] = ext.ingredients(
data["ingredients"])
data["directions"] = ext.directions(data["directions"])
data["title"] = ext.title(data["title"])
return data
def generation_function(texts, lang="en"):
langs = ['ru', 'en']
api_credentials = load_api()
if lang != "en" and lang in langs:
texts = translate_input(texts, lang)
output_ids = generator(
texts,
** chef_top
)[0]["generated_token_ids"]
recepie = tokenizer.decode(output_ids, skip_special_tokens=False)
generated_recipe = skip_special_tokens_and_prettify(recepie)
if lang != "en" and lang in langs:
generated_recipe = translate_recepie(generated_recipe, lang)
cook_image = generate_cook_image(
generated_recipe['title'], app_id=api_credentials['ID'], app_key=api_credentials['KEY'])
generated_recipe["image"] = cook_image
return generated_recipe
items = [
"macaroni, butter, salt, bacon, milk, flour, pepper, cream corn",
"provolone cheese, bacon, bread, ginger"
]
|