Omartificial-Intelligence-Space
commited on
Commit
•
bb49240
1
Parent(s):
6cafc91
update app.py
Browse files
app.py
CHANGED
@@ -4,11 +4,9 @@ from wikipediaapi import Wikipedia
|
|
4 |
import textwrap
|
5 |
import numpy as np
|
6 |
from openai import OpenAI
|
7 |
-
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
8 |
-
import os
|
9 |
|
10 |
# Function to process the input and generate the output
|
11 |
-
def process_query(wiki_page, embed_dim, query,
|
12 |
model_mapping = {
|
13 |
"Arabic-mpnet-base-all-nli-triplet": "Omartificial-Intelligence-Space/Arabic-mpnet-base-all-nli-triplet",
|
14 |
"Arabic-all-nli-triplet-Matryoshka": "Omartificial-Intelligence-Space/Arabic-all-nli-triplet-Matryoshka",
|
@@ -17,78 +15,70 @@ def process_query(wiki_page, embed_dim, query, mode):
|
|
17 |
"Marbert-all-nli-triplet-Matryoshka": "Omartificial-Intelligence-Space/Marbert-all-nli-triplet-Matryoshka"
|
18 |
}
|
19 |
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
wiki = Wikipedia('RAGBot/0.0', 'ar')
|
24 |
doc = wiki.page(wiki_page).text
|
25 |
paragraphs = doc.split('\n\n') # chunking
|
|
|
26 |
for i, p in enumerate(paragraphs):
|
27 |
wrapped_text = textwrap.fill(p, width=100)
|
28 |
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
query_embed = model.encode(query, normalize_embeddings=True)
|
35 |
-
similarities = np.dot(docs_embed, query_embed.T)
|
36 |
-
top_3_idx = np.argsort(similarities, axis=0)[-3:][::-1].tolist()
|
37 |
-
most_similar_documents = [paragraphs[idx] for idx in top_3_idx]
|
38 |
-
|
39 |
-
CONTEXT = ""
|
40 |
-
for p in most_similar_documents:
|
41 |
-
wrapped_text = textwrap.fill(p, width=100)
|
42 |
-
CONTEXT += wrapped_text + "\n\n"
|
43 |
-
|
44 |
-
prompt = f"""
|
45 |
-
use the following CONTEXT to answer the QUESTION at the end.
|
46 |
-
If you don't know the answer, just say that you don't know, don't try to make up an answer.
|
47 |
-
CONTEXT: {CONTEXT}
|
48 |
-
QUESTION: {query}
|
49 |
-
"""
|
50 |
-
|
51 |
-
if mode == "OpenAI":
|
52 |
-
client = OpenAI(api_key=openai_api_key)
|
53 |
-
response = client.chat.completions.create(
|
54 |
-
model="gpt-4",
|
55 |
-
messages=[
|
56 |
-
{"role": "user", "content": prompt},
|
57 |
-
]
|
58 |
-
)
|
59 |
-
responses[model_name] = response.choices[0].message.content
|
60 |
-
|
61 |
-
elif mode == "OpenSource":
|
62 |
-
tokenizer = AutoTokenizer.from_pretrained("google/gemini-2b", use_auth_token=hf_token)
|
63 |
-
model = AutoModelForCausalLM.from_pretrained("google/gemini-2b", use_auth_token=hf_token)
|
64 |
-
generator = pipeline('text-generation', model=model, tokenizer=tokenizer)
|
65 |
-
response = generator(prompt, max_length=512, num_return_sequences=1)
|
66 |
-
responses[model_name] = response[0]['generated_text']
|
67 |
-
|
68 |
-
return "\n\n".join([f"Model: {model_name}\nResponse: {response}" for model_name, response in responses.items()])
|
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 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
import textwrap
|
5 |
import numpy as np
|
6 |
from openai import OpenAI
|
|
|
|
|
7 |
|
8 |
# Function to process the input and generate the output
|
9 |
+
def process_query(wiki_page, model_name, embed_dim, query, api_key):
|
10 |
model_mapping = {
|
11 |
"Arabic-mpnet-base-all-nli-triplet": "Omartificial-Intelligence-Space/Arabic-mpnet-base-all-nli-triplet",
|
12 |
"Arabic-all-nli-triplet-Matryoshka": "Omartificial-Intelligence-Space/Arabic-all-nli-triplet-Matryoshka",
|
|
|
15 |
"Marbert-all-nli-triplet-Matryoshka": "Omartificial-Intelligence-Space/Marbert-all-nli-triplet-Matryoshka"
|
16 |
}
|
17 |
|
18 |
+
model_path = model_mapping[model_name]
|
19 |
+
model = SentenceTransformer(model_path, trust_remote_code=True, truncate_dim=embed_dim)
|
|
|
20 |
wiki = Wikipedia('RAGBot/0.0', 'ar')
|
21 |
doc = wiki.page(wiki_page).text
|
22 |
paragraphs = doc.split('\n\n') # chunking
|
23 |
+
|
24 |
for i, p in enumerate(paragraphs):
|
25 |
wrapped_text = textwrap.fill(p, width=100)
|
26 |
|
27 |
+
docs_embed = model.encode(paragraphs, normalize_embeddings=True)
|
28 |
+
query_embed = model.encode(query, normalize_embeddings=True)
|
29 |
+
similarities = np.dot(docs_embed, query_embed.T)
|
30 |
+
top_3_idx = np.argsort(similarities, axis=0)[-3:][::-1].tolist()
|
31 |
+
most_similar_documents = [paragraphs[idx] for idx in top_3_idx]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
|
33 |
+
CONTEXT = ""
|
34 |
+
for i, p in enumerate(most_similar_documents):
|
35 |
+
wrapped_text = textwrap.fill(p, width=100)
|
36 |
+
CONTEXT += wrapped_text + "\n\n"
|
37 |
+
|
38 |
+
prompt = f"""
|
39 |
+
use the following CONTEXT to answer the QUESTION at the end.
|
40 |
+
If you don't know the answer, just say that you don't know, don't try to make up an answer.
|
41 |
+
CONTEXT: {CONTEXT}
|
42 |
+
QUESTION: {query}
|
43 |
+
"""
|
44 |
+
|
45 |
+
client = OpenAI(api_key=api_key)
|
46 |
+
response = client.chat.completions.create(
|
47 |
+
model="gpt-4o",
|
48 |
+
messages=[
|
49 |
+
{"role": "user", "content": prompt},
|
50 |
+
]
|
51 |
)
|
52 |
|
53 |
+
return response.choices[0].message.content
|
54 |
+
|
55 |
+
# Define the interface
|
56 |
+
wiki_page_input = gr.Textbox(label="Wikipedia Page (in Arabic)")
|
57 |
+
query_input = gr.Textbox(label="Query (in Arabic)")
|
58 |
+
api_key_input = gr.Textbox(label="OpenAI API Key", type="password")
|
59 |
+
|
60 |
+
model_choice = gr.Dropdown(
|
61 |
+
choices=[
|
62 |
+
"Arabic-mpnet-base-all-nli-triplet",
|
63 |
+
"Arabic-all-nli-triplet-Matryoshka",
|
64 |
+
"Arabert-all-nli-triplet-Matryoshka",
|
65 |
+
"Arabic-labse-Matryoshka",
|
66 |
+
"Marbert-all-nli-triplet-Matryoshka"
|
67 |
+
],
|
68 |
+
label="Choose Embedding Model"
|
69 |
+
)
|
70 |
+
|
71 |
+
embed_dim_choice = gr.Dropdown(
|
72 |
+
choices=[768, 512, 256, 128, 64],
|
73 |
+
label="Embedding Dimension"
|
74 |
+
)
|
75 |
+
|
76 |
+
output_text = gr.Textbox(label="Output")
|
77 |
+
|
78 |
+
gr.Interface(
|
79 |
+
fn=process_query,
|
80 |
+
inputs=[wiki_page_input, model_choice, embed_dim_choice, query_input, api_key_input],
|
81 |
+
outputs=output_text,
|
82 |
+
title="Arabic Wiki RAG",
|
83 |
+
description="Choose a Wikipedia page, embedding model, and dimension to answer a query in Arabic."
|
84 |
+
).launch()
|