Amritpal Singh
commited on
Commit
·
b751517
1
Parent(s):
a001878
re-old one need UI change
Browse files
README.md
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
---
|
| 2 |
-
title:
|
| 3 |
-
emoji: "
|
| 4 |
colorFrom: blue
|
| 5 |
colorTo: blue
|
| 6 |
sdk: streamlit
|
|
@@ -9,4 +9,4 @@ app_file: app.py
|
|
| 9 |
pinned: false
|
| 10 |
---
|
| 11 |
|
| 12 |
-
|
|
|
|
| 1 |
---
|
| 2 |
+
title: DocAnalyzer Pro
|
| 3 |
+
emoji: "📄"
|
| 4 |
colorFrom: blue
|
| 5 |
colorTo: blue
|
| 6 |
sdk: streamlit
|
|
|
|
| 9 |
pinned: false
|
| 10 |
---
|
| 11 |
|
| 12 |
+
DocAnalyzer Pro is a Streamlit app for document question answering and summarization using Hugging Face Transformers.
|
app.py
CHANGED
|
@@ -3,6 +3,10 @@ os.environ["STREAMLIT_SERVER_PORT"] = "8501"
|
|
| 3 |
os.environ["STREAMLIT_SERVER_HEADLESS"] = "true"
|
| 4 |
|
| 5 |
import streamlit as st
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
from transformers import pipeline
|
| 7 |
from PyPDF2 import PdfReader
|
| 8 |
import docx
|
|
@@ -11,34 +15,13 @@ import psutil
|
|
| 11 |
from pathlib import Path
|
| 12 |
import torch
|
| 13 |
|
| 14 |
-
|
| 15 |
-
st.set_page_config(page_title="LexPilot", layout="wide")
|
| 16 |
-
|
| 17 |
-
# Sidebar with project info
|
| 18 |
-
with st.sidebar:
|
| 19 |
-
st.title("LexPilot™")
|
| 20 |
-
st.markdown(
|
| 21 |
-
"""
|
| 22 |
-
LexPilot™ ingests text, PDF, and Word files to instantly analyze contracts.
|
| 23 |
-
It delivers concise summaries and lets you ask targeted questions—
|
| 24 |
-
giving fast, precise insights to speed up legal and procurement reviews.
|
| 25 |
-
"""
|
| 26 |
-
)
|
| 27 |
-
st.markdown("---")
|
| 28 |
-
st.write("### System Status")
|
| 29 |
-
try:
|
| 30 |
-
device_status = 'GPU ✅' if torch.cuda.is_available() else 'CPU ⚠️'
|
| 31 |
-
except:
|
| 32 |
-
device_status = 'CPU (torch not configured)'
|
| 33 |
-
st.text(f"Device: {device_status}")
|
| 34 |
-
st.text(f"Memory: {psutil.virtual_memory().percent}% used")
|
| 35 |
-
st.text(f"CPU: {psutil.cpu_percent()}% used")
|
| 36 |
|
| 37 |
-
# Setup cache directory for models
|
| 38 |
def setup_environment():
|
| 39 |
cache_dir = Path(".cache/models")
|
| 40 |
try:
|
| 41 |
cache_dir.mkdir(exist_ok=True, parents=True)
|
|
|
|
| 42 |
except Exception as e:
|
| 43 |
st.error(f"Failed to create cache directory: {e}")
|
| 44 |
return cache_dir
|
|
@@ -48,19 +31,23 @@ cache_dir = setup_environment()
|
|
| 48 |
@st.cache_resource(ttl=3600)
|
| 49 |
def load_models():
|
| 50 |
try:
|
|
|
|
| 51 |
qa_model = pipeline(
|
| 52 |
"question-answering",
|
| 53 |
model="distilbert-base-cased-distilled-squad",
|
| 54 |
device=-1
|
| 55 |
)
|
|
|
|
|
|
|
| 56 |
summarizer_model = pipeline(
|
| 57 |
"summarization",
|
| 58 |
-
model="sshleifer/distilbart-cnn-6-6",
|
| 59 |
device=-1
|
| 60 |
)
|
|
|
|
| 61 |
return {'qa': qa_model, 'summarizer': summarizer_model}
|
| 62 |
except Exception as e:
|
| 63 |
-
st.error(f"Failed to load models: {e}")
|
| 64 |
st.stop()
|
| 65 |
|
| 66 |
models = load_models()
|
|
@@ -101,8 +88,8 @@ def generate_summary(text, max_length=150):
|
|
| 101 |
st.error(f"Summarization failed: {e}")
|
| 102 |
return ""
|
| 103 |
|
| 104 |
-
|
| 105 |
-
st.
|
| 106 |
|
| 107 |
with st.expander("📤 Upload Document", expanded=True):
|
| 108 |
uploaded_file = st.file_uploader("Choose PDF/DOCX", type=["pdf", "docx"])
|
|
@@ -141,7 +128,15 @@ with tab2:
|
|
| 141 |
st.success(f"Generated in {time.time()-start_time:.1f}s")
|
| 142 |
st.markdown(f"**Summary:**\n\n{summary}")
|
| 143 |
|
| 144 |
-
|
| 145 |
-
|
| 146 |
-
|
| 147 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
os.environ["STREAMLIT_SERVER_HEADLESS"] = "true"
|
| 4 |
|
| 5 |
import streamlit as st
|
| 6 |
+
st.set_page_config(page_title="DocAnalyzer Pro", layout="wide")
|
| 7 |
+
|
| 8 |
+
st.write("Step 1: Imports done")
|
| 9 |
+
|
| 10 |
from transformers import pipeline
|
| 11 |
from PyPDF2 import PdfReader
|
| 12 |
import docx
|
|
|
|
| 15 |
from pathlib import Path
|
| 16 |
import torch
|
| 17 |
|
| 18 |
+
st.write("Step 2: Libraries imported")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
|
|
|
| 20 |
def setup_environment():
|
| 21 |
cache_dir = Path(".cache/models")
|
| 22 |
try:
|
| 23 |
cache_dir.mkdir(exist_ok=True, parents=True)
|
| 24 |
+
st.write(f"Step 3: Cache directory setup at {cache_dir}")
|
| 25 |
except Exception as e:
|
| 26 |
st.error(f"Failed to create cache directory: {e}")
|
| 27 |
return cache_dir
|
|
|
|
| 31 |
@st.cache_resource(ttl=3600)
|
| 32 |
def load_models():
|
| 33 |
try:
|
| 34 |
+
st.info("Loading QA model...")
|
| 35 |
qa_model = pipeline(
|
| 36 |
"question-answering",
|
| 37 |
model="distilbert-base-cased-distilled-squad",
|
| 38 |
device=-1
|
| 39 |
)
|
| 40 |
+
st.info("QA model loaded.")
|
| 41 |
+
st.info("Loading summarizer model...")
|
| 42 |
summarizer_model = pipeline(
|
| 43 |
"summarization",
|
| 44 |
+
model="sshleifer/distilbart-cnn-6-6", # smaller model for faster startup
|
| 45 |
device=-1
|
| 46 |
)
|
| 47 |
+
st.info("Summarizer model loaded.")
|
| 48 |
return {'qa': qa_model, 'summarizer': summarizer_model}
|
| 49 |
except Exception as e:
|
| 50 |
+
st.error(f"❌ Failed to load models: {e}")
|
| 51 |
st.stop()
|
| 52 |
|
| 53 |
models = load_models()
|
|
|
|
| 88 |
st.error(f"Summarization failed: {e}")
|
| 89 |
return ""
|
| 90 |
|
| 91 |
+
st.title("📄 DocAnalyzer Pro")
|
| 92 |
+
st.write("Step 4: UI started")
|
| 93 |
|
| 94 |
with st.expander("📤 Upload Document", expanded=True):
|
| 95 |
uploaded_file = st.file_uploader("Choose PDF/DOCX", type=["pdf", "docx"])
|
|
|
|
| 128 |
st.success(f"Generated in {time.time()-start_time:.1f}s")
|
| 129 |
st.markdown(f"**Summary:**\n\n{summary}")
|
| 130 |
|
| 131 |
+
with st.expander("⚙️ System Status"):
|
| 132 |
+
try:
|
| 133 |
+
device_status = 'GPU ✅' if torch.cuda.is_available() else 'CPU ⚠️'
|
| 134 |
+
except:
|
| 135 |
+
device_status = 'CPU (torch not configured)'
|
| 136 |
+
st.code(f"""
|
| 137 |
+
Models loaded: {', '.join(models.keys())}
|
| 138 |
+
Device: {device_status}
|
| 139 |
+
Memory: {psutil.virtual_memory().percent}% used
|
| 140 |
+
CPU: {psutil.cpu_percent()}% used
|
| 141 |
+
Cache location: {cache_dir}
|
| 142 |
+
""")
|