Spaces:
Runtime error
Runtime error
Commit
•
7acf2ae
1
Parent(s):
c6c114f
Update eng_to_aslGloss_app.py
Browse files- eng_to_aslGloss_app.py +172 -168
eng_to_aslGloss_app.py
CHANGED
@@ -1,180 +1,150 @@
|
|
1 |
-
from pathlib import Path
|
2 |
-
import gradio as gr
|
3 |
-
#import openai
|
4 |
import os
|
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 |
-
def
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
def
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
60 |
|
61 |
-
def getGlossFromText(query):
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
-
|
74 |
-
def getASLGloss(testQs):
|
75 |
-
|
76 |
-
|
77 |
-
|
78 |
-
|
79 |
-
|
80 |
-
|
81 |
-
|
82 |
-
|
83 |
-
|
84 |
-
|
85 |
-
|
86 |
|
87 |
-
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
|
93 |
|
94 |
|
95 |
|
96 |
-
def main():
|
97 |
-
|
98 |
-
title = "English to ASL Gloss"
|
99 |
-
#description = """Translate English text to ASL Gloss"""
|
100 |
-
description = "This program uses GPT4 alongside prompt engineering to \
|
101 |
-
translate English text to ASL gloss.\n \
|
102 |
-
<b>Type in the English sentence you would like to translate into ASL Gloss.</b> \
|
103 |
-
\n \n This program was last updated on February 27, 2024, and uses GPT4-Turbo (0125 preview version) \
|
104 |
-
\n\n \
|
105 |
-
\n \n This version of EngToASLGloss contains superscript notation which adds \
|
106 |
-
grammatical context to assist in ASL generation. \
|
107 |
-
\n Below are the guidelines we are using to express grammatical concepts \
|
108 |
-
in ASL gloss.\
|
109 |
-
Anything within the angle brackets < > indicates this additional grammatical notation.\
|
110 |
-
If the angle brackets are directly next to a word, the notation inside \
|
111 |
-
the angle brackets is associate with just that word, e.g. WILL < A >. \
|
112 |
-
If the angle brackets are next to a whitespace after a word,\
|
113 |
-
the notation inside the angle bracket is associated with all of the words\
|
114 |
-
before it, up until a comma, another angle bracket, or a double space.\
|
115 |
-
\n \n This sentence is an example of this rule:\
|
116 |
-
\n NEXT-YEAR < Ti >, MY FIANCE < T >, TWO-OF-US MARRY \< A \>.\
|
117 |
-
\n\r \
|
118 |
-
\n The superscript notation options that will appear in results are as follows:\
|
119 |
-
\n Ti marks time\
|
120 |
-
\n T marks topic\
|
121 |
-
\n A marks comment\
|
122 |
-
\n Y/N marks yes-no question\
|
123 |
-
\n WHQ marks wh-question\
|
124 |
-
\n RHQ marks rhetorical question\
|
125 |
-
\n < Cond > marks conditional sentences\
|
126 |
-
\n lower case marks directional verbs\
|
127 |
-
\n ++ marks emphesis ('very' or 'a lot of')\
|
128 |
-
\n \# marks lexical fingerspelling \
|
129 |
-
\n \- marks space between individual letters of fingerspelling\
|
130 |
-
\n \n <b>Note: This is a prototype and is still in development. \
|
131 |
-
Do not use it in a production deployment.</b> \
|
132 |
-
\n For additional details on how the program works, please see \
|
133 |
-
[the README](https://huggingface.co/spaces/rrakov/EngTexToASLGloss/blob/main/README.md)"
|
134 |
-
|
135 |
-
interface = gr.Interface(
|
136 |
-
fn=getGlossFromText,
|
137 |
-
inputs="textbox",
|
138 |
-
outputs="text",
|
139 |
-
title = title,
|
140 |
-
description = description)
|
141 |
-
#examples = [[("Prompt: Every year I buy my dad a gift \n", "Result: EVERY-YEAR<Ti>, MY DAD GIFT<T>, ME BUY<A>")]])
|
142 |
-
# examples=[["Every year I buy my dad a gift"], ["I always look forward to the family vacation"],
|
143 |
-
# ["If I don't travel often, I am sad."]])
|
144 |
-
interface.launch()
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
if __name__ == "__main__":
|
149 |
-
main()
|
150 |
-
|
151 |
|
152 |
-
#
|
153 |
-
#
|
154 |
-
#
|
155 |
-
#
|
156 |
-
#
|
157 |
-
#
|
158 |
-
#
|
159 |
-
#
|
160 |
-
#
|
161 |
-
#
|
162 |
-
|
163 |
-
#
|
164 |
-
#
|
165 |
-
#
|
166 |
-
#
|
167 |
-
#
|
168 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
169 |
|
170 |
# interface = gr.Interface(
|
171 |
-
# fn=
|
172 |
# inputs="textbox",
|
173 |
# outputs="text",
|
174 |
# title = title,
|
175 |
-
# description = description
|
176 |
-
# examples=[["
|
177 |
-
#
|
|
|
178 |
# interface.launch()
|
179 |
|
180 |
|
@@ -182,9 +152,43 @@ if __name__ == "__main__":
|
|
182 |
# if __name__ == "__main__":
|
183 |
# main()
|
184 |
|
185 |
-
|
186 |
-
#
|
187 |
-
#
|
188 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
189 |
|
190 |
|
|
|
|
|
|
|
|
|
1 |
import os
|
2 |
+
secret = os.getenv("VerySecret")
|
3 |
+
print(secret)
|
4 |
+
|
5 |
+
# from pathlib import Path
|
6 |
+
# import gradio as gr
|
7 |
+
# #import openai
|
8 |
+
# import os
|
9 |
+
# import tiktoken
|
10 |
+
# from openai import OpenAI
|
11 |
+
|
12 |
+
|
13 |
+
# # Set secret key
|
14 |
+
# #HF_TOKEN = os.getenv("NextStar")
|
15 |
+
|
16 |
+
# # Set client and secret key
|
17 |
+
# client = OpenAI(api_key=os.getenv("NextStar"))
|
18 |
+
|
19 |
+
# #Set prompt engineering paths (so globally available)
|
20 |
+
# inStructionPath = "intro_instructions_combine.txt"
|
21 |
+
# inRulesPath = "formatting_rules_expanded.txt"
|
22 |
+
# inExamplesPath = "examples_longer1.txt"
|
23 |
+
# inDialoguesPath = "examples_dialogues.txt"
|
24 |
+
|
25 |
+
# #Set to read in prompting files
|
26 |
+
# def openReadFiles(inpath):
|
27 |
+
# infile = Path (inpath)
|
28 |
+
# with open(infile) as f:
|
29 |
+
# data = f.read()
|
30 |
+
# return data
|
31 |
+
|
32 |
+
|
33 |
+
# # Set up prompting data (so globally available)
|
34 |
+
# instruct = openReadFiles(inStructionPath)
|
35 |
+
# rules = openReadFiles(inRulesPath)
|
36 |
+
# examples = openReadFiles(inExamplesPath)
|
37 |
+
# exampleDialogues = openReadFiles(inDialoguesPath)
|
38 |
+
|
39 |
+
|
40 |
+
# def formatQuery(engText):
|
41 |
+
# """Add prompt instructions to English text for GPT4"""
|
42 |
+
# instruct = "Now, translate the following sentences to perfect ASL gloss using the grammatical, syntactic, and notation rules you just learned. \n\n"
|
43 |
+
# query = instruct+engText
|
44 |
+
# return query
|
45 |
+
|
46 |
+
|
47 |
+
# def num_tokens_from_string(string: str, encoding_name: str) -> int:
|
48 |
+
# """Returns the number of tokens in a text string."""
|
49 |
+
# encoding = tiktoken.get_encoding(encoding_name)
|
50 |
+
# num_tokens = len(encoding.encode(string))
|
51 |
+
# return num_tokens
|
52 |
+
|
53 |
+
|
54 |
+
# def checkTokens(tokens):
|
55 |
+
# """Checks tokens to ensrue we can translate to ASL gloss"""
|
56 |
+
# goAhead = None
|
57 |
+
# if tokens >= 553:
|
58 |
+
# print(f"Cannot translate to ASL gloss at this time: too many tokens ({tokens})")
|
59 |
+
# goAhead = False
|
60 |
+
# else:
|
61 |
+
# goAhead = True
|
62 |
+
# print(f"Number of tokens is acceptable: can continue translating")
|
63 |
+
# return goAhead
|
64 |
|
65 |
+
# def getGlossFromText(query):
|
66 |
+
# """Sets all for getting ASL gloss"""
|
67 |
+
# text = formatQuery(query)
|
68 |
+
# tokens = num_tokens_from_string(text, "cl100k_base")
|
69 |
+
# goAhead = checkTokens(tokens)
|
70 |
+
# if goAhead == True:
|
71 |
+
# results = getASLGloss(text)
|
72 |
+
# else:
|
73 |
+
# results = "Too many tokens: cannot translate"
|
74 |
+
# return results
|
75 |
+
|
76 |
+
|
77 |
+
|
78 |
+
# def getASLGloss(testQs):
|
79 |
+
# """Get ASL gloss from OpenAI using our prompt engineering"""
|
80 |
+
# #openai.api_key = HF_TOKENS
|
81 |
+
# completion = client.chat.completions.create(
|
82 |
+
# model = 'gpt-4-0125-preview',
|
83 |
+
# messages = [
|
84 |
+
# {"role": "system", "content": instruct},
|
85 |
+
# {"role": "system", "content": rules},
|
86 |
+
# {"role": "system", "content": examples},
|
87 |
+
# {"role": "system", "content": exampleDialogues},
|
88 |
+
# {"role": "user", "content": testQs},
|
89 |
+
# ],
|
90 |
|
91 |
+
# temperature = 0
|
92 |
+
# )
|
93 |
+
# #results = completion['choices'][0]['message']['content']
|
94 |
+
# results = completion.choices[0].message.content
|
95 |
+
# return results
|
96 |
|
97 |
|
98 |
|
99 |
|
100 |
+
# def main():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
101 |
|
102 |
+
# title = "English to ASL Gloss"
|
103 |
+
# #description = """Translate English text to ASL Gloss"""
|
104 |
+
# description = "This program uses GPT4 alongside prompt engineering to \
|
105 |
+
# translate English text to ASL gloss.\n \
|
106 |
+
# <b>Type in the English sentence you would like to translate into ASL Gloss.</b> \
|
107 |
+
# \n \n This program was last updated on February 27, 2024, and uses GPT4-Turbo (0125 preview version) \
|
108 |
+
# \n\n \
|
109 |
+
# \n \n This version of EngToASLGloss contains superscript notation which adds \
|
110 |
+
# grammatical context to assist in ASL generation. \
|
111 |
+
# \n Below are the guidelines we are using to express grammatical concepts \
|
112 |
+
# in ASL gloss.\
|
113 |
+
# Anything within the angle brackets < > indicates this additional grammatical notation.\
|
114 |
+
# If the angle brackets are directly next to a word, the notation inside \
|
115 |
+
# the angle brackets is associate with just that word, e.g. WILL < A >. \
|
116 |
+
# If the angle brackets are next to a whitespace after a word,\
|
117 |
+
# the notation inside the angle bracket is associated with all of the words\
|
118 |
+
# before it, up until a comma, another angle bracket, or a double space.\
|
119 |
+
# \n \n This sentence is an example of this rule:\
|
120 |
+
# \n NEXT-YEAR < Ti >, MY FIANCE < T >, TWO-OF-US MARRY \< A \>.\
|
121 |
+
# \n\r \
|
122 |
+
# \n The superscript notation options that will appear in results are as follows:\
|
123 |
+
# \n Ti marks time\
|
124 |
+
# \n T marks topic\
|
125 |
+
# \n A marks comment\
|
126 |
+
# \n Y/N marks yes-no question\
|
127 |
+
# \n WHQ marks wh-question\
|
128 |
+
# \n RHQ marks rhetorical question\
|
129 |
+
# \n < Cond > marks conditional sentences\
|
130 |
+
# \n lower case marks directional verbs\
|
131 |
+
# \n ++ marks emphesis ('very' or 'a lot of')\
|
132 |
+
# \n \# marks lexical fingerspelling \
|
133 |
+
# \n \- marks space between individual letters of fingerspelling\
|
134 |
+
# \n \n <b>Note: This is a prototype and is still in development. \
|
135 |
+
# Do not use it in a production deployment.</b> \
|
136 |
+
# \n For additional details on how the program works, please see \
|
137 |
+
# [the README](https://huggingface.co/spaces/rrakov/EngTexToASLGloss/blob/main/README.md)"
|
138 |
|
139 |
# interface = gr.Interface(
|
140 |
+
# fn=getGlossFromText,
|
141 |
# inputs="textbox",
|
142 |
# outputs="text",
|
143 |
# title = title,
|
144 |
+
# description = description)
|
145 |
+
# #examples = [[("Prompt: Every year I buy my dad a gift \n", "Result: EVERY-YEAR<Ti>, MY DAD GIFT<T>, ME BUY<A>")]])
|
146 |
+
# # examples=[["Every year I buy my dad a gift"], ["I always look forward to the family vacation"],
|
147 |
+
# # ["If I don't travel often, I am sad."]])
|
148 |
# interface.launch()
|
149 |
|
150 |
|
|
|
152 |
# if __name__ == "__main__":
|
153 |
# main()
|
154 |
|
155 |
+
|
156 |
+
# # def getAnswer(query, texts = texts, embeddings = embeddings):
|
157 |
+
# # docsearch = FAISS.from_texts(texts, embeddings)
|
158 |
+
# # docs = docsearch.similarity_search(query)
|
159 |
+
# # chain = load_qa_chain(OpenAI(openai_api_key = HF_TOKEN, temperature=0), chain_type="map_reduce", return_map_steps=False)
|
160 |
+
# # response = chain({"input_documents": docs, "question": query}, return_only_outputs=True)
|
161 |
+
# # #interum_q = list(response.keys())
|
162 |
+
# # interum_a = list(response.values())
|
163 |
+
# # q = query
|
164 |
+
# # a = interum_a[0]
|
165 |
+
# # return a
|
166 |
+
|
167 |
+
# # # query = "describe the fisher database"
|
168 |
+
# # # docs = docsearch.similarity_search(query)
|
169 |
+
# # # chain = load_qa_chain(OpenAI(openai_api_key = "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", temperature=0), chain_type="map_reduce", return_map_steps=False)
|
170 |
+
# # # chain({"input_documents": docs, "question": query}, return_only_outputs=True)
|
171 |
+
# # title = "Query the S Drive!"
|
172 |
+
# # description = """This QA system will answer questions based on information in [data descriptions](https://indeocorp-my.sharepoint.com/:x:/g/personal/rrakov_sorenson_com/EWhs_Gpp9nNEukR7iJLd4mQBPREngKdRGYpT545jX8mY4Q?e=9EeEWF)"""
|
173 |
+
|
174 |
+
# # interface = gr.Interface(
|
175 |
+
# # fn=getAnswer,
|
176 |
+
# # inputs="textbox",
|
177 |
+
# # outputs="text",
|
178 |
+
# # title = title,
|
179 |
+
# # description = description,
|
180 |
+
# # examples=[["Where is the Fisher database?"], ["Where is the Defined Crowd audio?"], ["Do we have any Spanish audio data?"],
|
181 |
+
# # ["How many audio files do we have in the CallHome database?"]])
|
182 |
+
# # interface.launch()
|
183 |
+
|
184 |
+
|
185 |
+
|
186 |
+
# # if __name__ == "__main__":
|
187 |
+
# # main()
|
188 |
+
|
189 |
+
# # def main():
|
190 |
+
# # results = setMode()
|
191 |
+
# # print (results)
|
192 |
+
# # main()
|
193 |
|
194 |
|