Spaces:
Runtime error
Runtime error
tabs, xlsx analyzer
#2
by
cececerece
- opened
- README.md +1 -1
- app.py +15 -186
- requirements.txt +1 -3
README.md
CHANGED
@@ -5,9 +5,9 @@ colorFrom: purple
|
|
5 |
colorTo: pink
|
6 |
sdk: gradio
|
7 |
sdk_version: 3.27.0
|
8 |
-
python_version: 3.10.9
|
9 |
app_file: app.py
|
10 |
pinned: false
|
11 |
duplicated_from: fffiloni/langchain-chat-with-pdf-openai
|
12 |
---
|
|
|
13 |
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
5 |
colorTo: pink
|
6 |
sdk: gradio
|
7 |
sdk_version: 3.27.0
|
|
|
8 |
app_file: app.py
|
9 |
pinned: false
|
10 |
duplicated_from: fffiloni/langchain-chat-with-pdf-openai
|
11 |
---
|
12 |
+
|
13 |
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
app.py
CHANGED
@@ -9,11 +9,6 @@ from langchain.embeddings import OpenAIEmbeddings
|
|
9 |
from langchain.vectorstores import Chroma
|
10 |
from langchain.chains import ConversationalRetrievalChain
|
11 |
from langchain import PromptTemplate
|
12 |
-
from transformers import Pix2StructForConditionalGeneration, Pix2StructProcessor
|
13 |
-
import requests
|
14 |
-
from PIL import Image
|
15 |
-
import torch
|
16 |
-
|
17 |
|
18 |
|
19 |
# _template = """Given the following conversation and a follow up question, rephrase the follow up question to be a standalone question.
|
@@ -33,32 +28,12 @@ import torch
|
|
33 |
# =========
|
34 |
# Answer in Markdown:"""
|
35 |
|
36 |
-
torch.hub.download_url_to_file('https://raw.githubusercontent.com/vis-nlp/ChartQA/main/ChartQA%20Dataset/val/png/20294671002019.png', 'chart_example.png')
|
37 |
-
torch.hub.download_url_to_file('https://raw.githubusercontent.com/vis-nlp/ChartQA/main/ChartQA%20Dataset/test/png/multi_col_1081.png', 'chart_example_2.png')
|
38 |
-
torch.hub.download_url_to_file('https://raw.githubusercontent.com/vis-nlp/ChartQA/main/ChartQA%20Dataset/test/png/18143564004789.png', 'chart_example_3.png')
|
39 |
-
torch.hub.download_url_to_file('https://sharkcoder.com/files/article/matplotlib-bar-plot.png', 'chart_example_4.png')
|
40 |
-
|
41 |
-
|
42 |
-
model_name = "google/matcha-chartqa"
|
43 |
-
model = Pix2StructForConditionalGeneration.from_pretrained(model_name)
|
44 |
-
processor = Pix2StructProcessor.from_pretrained(model_name)
|
45 |
-
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
46 |
-
model.to(device)
|
47 |
-
|
48 |
-
def filter_output(output):
|
49 |
-
return output.replace("<0x0A>", "")
|
50 |
-
|
51 |
-
def chart_qa(image, question):
|
52 |
-
inputs = processor(images=image, text=question, return_tensors="pt").to(device)
|
53 |
-
predictions = model.generate(**inputs, max_new_tokens=512)
|
54 |
-
return filter_output(processor.decode(predictions[0], skip_special_tokens=True))
|
55 |
-
|
56 |
def loading_pdf():
|
57 |
return "Loading..."
|
58 |
|
59 |
|
60 |
def pdf_changes(pdf_doc, open_ai_key):
|
61 |
-
if
|
62 |
os.environ['OPENAI_API_KEY'] = open_ai_key
|
63 |
loader = OnlinePDFLoader(pdf_doc.name)
|
64 |
documents = loader.load()
|
@@ -108,180 +83,34 @@ css="""
|
|
108 |
"""
|
109 |
|
110 |
title = """
|
111 |
-
<div style="text-align: center;">
|
112 |
<h1>YnP LangChain Test </h1>
|
113 |
<p style="text-align: center;">Please specify OpenAI Key before use</p>
|
114 |
</div>
|
115 |
"""
|
116 |
|
117 |
|
118 |
-
|
119 |
-
# with gr.Column(elem_id="col-container"):
|
120 |
-
# gr.HTML(title)
|
121 |
-
|
122 |
-
# with gr.Column():
|
123 |
-
# openai_key = gr.Textbox(label="You OpenAI API key", type="password")
|
124 |
-
# pdf_doc = gr.File(label="Load a pdf", file_types=['.pdf'], type="file")
|
125 |
-
# with gr.Row():
|
126 |
-
# langchain_status = gr.Textbox(label="Status", placeholder="", interactive=False)
|
127 |
-
# load_pdf = gr.Button("Load pdf to langchain")
|
128 |
-
|
129 |
-
# chatbot = gr.Chatbot([], elem_id="chatbot").style(height=350)
|
130 |
-
# question = gr.Textbox(label="Question", placeholder="Type your question and hit Enter ")
|
131 |
-
# submit_btn = gr.Button("Send Message")
|
132 |
-
|
133 |
-
# load_pdf.click(loading_pdf, None, langchain_status, queue=False)
|
134 |
-
# load_pdf.click(pdf_changes, inputs=[pdf_doc, openai_key], outputs=[langchain_status], queue=False)
|
135 |
-
# question.submit(add_text, [chatbot, question], [chatbot, question]).then(
|
136 |
-
# bot, chatbot, chatbot
|
137 |
-
# )
|
138 |
-
# submit_btn.click(add_text, [chatbot, question], [chatbot, question]).then(
|
139 |
-
# bot, chatbot, chatbot)
|
140 |
-
|
141 |
-
# demo.launch()
|
142 |
-
|
143 |
-
|
144 |
-
"""functions"""
|
145 |
-
|
146 |
-
def load_file():
|
147 |
-
return "Loading..."
|
148 |
-
|
149 |
-
def load_xlsx(name):
|
150 |
-
import pandas as pd
|
151 |
-
|
152 |
-
xls_file = rf'{name}'
|
153 |
-
data = pd.read_excel(xls_file)
|
154 |
-
return data
|
155 |
-
|
156 |
-
def table_loader(table_file, open_ai_key):
|
157 |
-
import os
|
158 |
-
from langchain.llms import OpenAI
|
159 |
-
from langchain.agents import create_pandas_dataframe_agent
|
160 |
-
from pandas import read_csv
|
161 |
-
|
162 |
-
global agent
|
163 |
-
if open_ai_key is not None:
|
164 |
-
os.environ['OPENAI_API_KEY'] = open_ai_key
|
165 |
-
else:
|
166 |
-
return "Enter API"
|
167 |
-
|
168 |
-
if table_file.name.endswith('.xlsx') or table_file.name.endswith('.xls'):
|
169 |
-
data = load_xlsx(table_file.name)
|
170 |
-
agent = create_pandas_dataframe_agent(OpenAI(temperature=0), data)
|
171 |
-
return "Ready!"
|
172 |
-
elif table_file.name.endswith('.csv'):
|
173 |
-
data = read_csv(table_file.name)
|
174 |
-
agent = create_pandas_dataframe_agent(OpenAI(temperature=0), data)
|
175 |
-
return "Ready!"
|
176 |
-
else:
|
177 |
-
return "Wrong file format! Upload excel file or csv!"
|
178 |
-
|
179 |
-
def run(query):
|
180 |
-
from langchain.callbacks import get_openai_callback
|
181 |
-
|
182 |
-
with get_openai_callback() as cb:
|
183 |
-
response = (agent.run(query))
|
184 |
-
costs = (f"Total Cost (USD): ${cb.total_cost}")
|
185 |
-
output = f'{response} \n {costs}'
|
186 |
-
return output
|
187 |
-
|
188 |
-
def respond(message, chat_history):
|
189 |
-
import time
|
190 |
-
|
191 |
-
bot_message = run(message)
|
192 |
-
chat_history.append((message, bot_message))
|
193 |
-
time.sleep(0.5)
|
194 |
-
return "", chat_history
|
195 |
-
|
196 |
-
|
197 |
-
with gr.Blocks() as demo:
|
198 |
with gr.Column(elem_id="col-container"):
|
199 |
gr.HTML(title)
|
200 |
-
key = gr.Textbox(
|
201 |
-
show_label=False,
|
202 |
-
placeholder="Your OpenAI key",
|
203 |
-
type = 'password',
|
204 |
-
).style(container=False)
|
205 |
-
|
206 |
-
# PDF processing tab
|
207 |
-
with gr.Tab("PDFs"):
|
208 |
|
209 |
-
with gr.
|
210 |
-
|
211 |
-
|
|
|
212 |
langchain_status = gr.Textbox(label="Status", placeholder="", interactive=False)
|
213 |
load_pdf = gr.Button("Load pdf to langchain")
|
214 |
-
|
215 |
-
|
216 |
-
|
217 |
-
|
218 |
-
|
219 |
-
with gr.Row():
|
220 |
-
|
221 |
-
with gr.Column(scale=1):
|
222 |
-
chatbot = gr.Chatbot([], elem_id="chatbot").style(height=350)
|
223 |
-
|
224 |
-
with gr.Row():
|
225 |
-
|
226 |
-
with gr.Column(scale=0.85):
|
227 |
-
question = gr.Textbox(
|
228 |
-
show_label=False,
|
229 |
-
placeholder="Enter text and press enter, or upload an image",
|
230 |
-
).style(container=False)
|
231 |
-
|
232 |
-
with gr.Column(scale=0.15, min_width=0):
|
233 |
-
clr_btn = gr.Button("Clear!")
|
234 |
-
|
235 |
load_pdf.click(loading_pdf, None, langchain_status, queue=False)
|
236 |
-
load_pdf.click(pdf_changes, inputs=[pdf_doc,
|
237 |
question.submit(add_text, [chatbot, question], [chatbot, question]).then(
|
238 |
bot, chatbot, chatbot
|
239 |
)
|
240 |
-
|
241 |
-
|
242 |
-
with gr.Tab("Spreadsheets"):
|
243 |
-
with gr.Row():
|
244 |
-
|
245 |
-
with gr.Column(scale=0.5):
|
246 |
-
status_sh = gr.Textbox(label="Status", placeholder="", interactive=False)
|
247 |
-
load_table = gr.Button("Load csv|xlsx to langchain")
|
248 |
-
|
249 |
-
with gr.Column(scale=0.5):
|
250 |
-
raw_table = gr.File(label="Load a table file (xls or csv)", file_types=['.csv, xlsx, xls'], type="file")
|
251 |
-
|
252 |
-
|
253 |
-
with gr.Row():
|
254 |
-
|
255 |
-
with gr.Column(scale=1):
|
256 |
-
chatbot_sh = gr.Chatbot([], elem_id="chatbot").style(height=350)
|
257 |
-
|
258 |
-
|
259 |
-
with gr.Row():
|
260 |
-
|
261 |
-
with gr.Column(scale=0.85):
|
262 |
-
question_sh = gr.Textbox(
|
263 |
-
show_label=False,
|
264 |
-
placeholder="Enter text and press enter, or upload an image",
|
265 |
-
).style(container=False)
|
266 |
-
|
267 |
-
with gr.Column(scale=0.15, min_width=0):
|
268 |
-
clr_btn = gr.Button("Clear!")
|
269 |
-
|
270 |
-
load_table.click(load_file, None, status_sh, queue=False)
|
271 |
-
load_table.click(table_loader, inputs=[raw_table, key], outputs=[status_sh], queue=False)
|
272 |
-
|
273 |
-
question_sh.submit(respond, [question_sh, chatbot_sh], [question_sh, chatbot_sh])
|
274 |
-
clr_btn.click(lambda: None, None, chatbot_sh, queue=False)
|
275 |
-
|
276 |
|
277 |
-
with gr.Tab("Charts"):
|
278 |
-
image = gr.Image(type="pil", label="Chart")
|
279 |
-
question = gr.Textbox(label="Question")
|
280 |
-
load_chart = gr.Button("Load chart and question!")
|
281 |
-
answer = gr.Textbox(label="Model Output")
|
282 |
-
|
283 |
-
load_chart.click(chart_qa, [image, question], answer)
|
284 |
-
|
285 |
-
|
286 |
-
demo.queue(concurrency_count=3)
|
287 |
demo.launch()
|
|
|
9 |
from langchain.vectorstores import Chroma
|
10 |
from langchain.chains import ConversationalRetrievalChain
|
11 |
from langchain import PromptTemplate
|
|
|
|
|
|
|
|
|
|
|
12 |
|
13 |
|
14 |
# _template = """Given the following conversation and a follow up question, rephrase the follow up question to be a standalone question.
|
|
|
28 |
# =========
|
29 |
# Answer in Markdown:"""
|
30 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
31 |
def loading_pdf():
|
32 |
return "Loading..."
|
33 |
|
34 |
|
35 |
def pdf_changes(pdf_doc, open_ai_key):
|
36 |
+
if openai_key is not None:
|
37 |
os.environ['OPENAI_API_KEY'] = open_ai_key
|
38 |
loader = OnlinePDFLoader(pdf_doc.name)
|
39 |
documents = loader.load()
|
|
|
83 |
"""
|
84 |
|
85 |
title = """
|
86 |
+
<div style="text-align: center;max-width: 700px;">
|
87 |
<h1>YnP LangChain Test </h1>
|
88 |
<p style="text-align: center;">Please specify OpenAI Key before use</p>
|
89 |
</div>
|
90 |
"""
|
91 |
|
92 |
|
93 |
+
with gr.Blocks(css=css) as demo:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
94 |
with gr.Column(elem_id="col-container"):
|
95 |
gr.HTML(title)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
96 |
|
97 |
+
with gr.Column():
|
98 |
+
openai_key = gr.Textbox(label="You OpenAI API key", type="password")
|
99 |
+
pdf_doc = gr.File(label="Load a pdf", file_types=['.pdf'], type="file")
|
100 |
+
with gr.Row():
|
101 |
langchain_status = gr.Textbox(label="Status", placeholder="", interactive=False)
|
102 |
load_pdf = gr.Button("Load pdf to langchain")
|
103 |
+
|
104 |
+
chatbot = gr.Chatbot([], elem_id="chatbot").style(height=350)
|
105 |
+
question = gr.Textbox(label="Question", placeholder="Type your question and hit Enter ")
|
106 |
+
submit_btn = gr.Button("Send Message")
|
107 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
108 |
load_pdf.click(loading_pdf, None, langchain_status, queue=False)
|
109 |
+
load_pdf.click(pdf_changes, inputs=[pdf_doc, openai_key], outputs=[langchain_status], queue=False)
|
110 |
question.submit(add_text, [chatbot, question], [chatbot, question]).then(
|
111 |
bot, chatbot, chatbot
|
112 |
)
|
113 |
+
submit_btn.click(add_text, [chatbot, question], [chatbot, question]).then(
|
114 |
+
bot, chatbot, chatbot)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
115 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
116 |
demo.launch()
|
requirements.txt
CHANGED
@@ -3,6 +3,4 @@ tiktoken
|
|
3 |
chromadb
|
4 |
langchain
|
5 |
unstructured
|
6 |
-
unstructured[local-inference]
|
7 |
-
pandas
|
8 |
-
tabulate
|
|
|
3 |
chromadb
|
4 |
langchain
|
5 |
unstructured
|
6 |
+
unstructured[local-inference]
|
|
|
|