style change

README.md

@@ -1,46 +0,0 @@
-# Sentiment Analysis and Voice Recorder App
-
-This project is a **Streamlit-based application** that combines voice recording and sentiment analysis. It allows users to record audio, transcribe the speech into text, analyze its sentiment, and visualize the sentiment results as a histogram. Sentiment scores range from -10 (furious) to +10 (peaceful).
-
----
-
-## Features
-- **Voice Recording**: Users can record audio directly in the app.
-- **Speech-to-Text**: Converts recorded audio into text using a transcription service.
-- **Sentiment Analysis**: Analyzes the sentiment of the conversation using a pre-trained model or API (e.g., Gemini).
-- **Histogram Visualization**: Displays the sentiment scores of individual sentences in a histogram.
-- **Export**: Saves the histogram as a PNG file.
-
----
-
-## How to setup
-1. Create environment and install dependencies
-```bash
-conda create -n setiment python=3.12
-conda activate setiment
-pip install -r requirements.txt
-```
-
-2. Create secrets.toml in the .streamlit folder and insert GOOGLE_API_KEY and GOOGLE_APPLICATION_CREDENTIALS
-
-## How to run
-
-```bash
-streamlit run streamlit_app.py
-```
-
-## How it works
-
-**streamlit_app.py** is the ui in streamlit
-
-**translation.py** gets the transcription from the recorded audio using google speech-to-text api
-
-**gemini_call.py** calls gemini on transcription in order to perform the sentiment analysis and makes a plot which is saved in a png
-
-**sentiment_analysis.py** is combining the 2 files above
-
-## Where to find it
-
-Path on **rtx**  : /home/vladp/Projects/sentimentAnalysis/
-
-Env : setiment

streamlit_app.py

@@ -63,7 +63,9 @@ def display_message(role, content, image_path=None):
 
 def main():
     """Main function that loops and can record audio and dispaly the messages from gemini"""
-    st.title("Streamlit Voice Recorder")
+    st.title("Verbal Sentiment Tracker")
+    
+    st.write("This demo lets you record your voice, automatically transcribes the audio, and analyzes the emotional tone of your words over time. By generating a histogram of sentiment evolution, it helps uncover how feelings shift throughout a conversation—useful for scenarios like customer calls, where ending on a positive note can make all the difference.")
 
     logo_path = "zega_logo.svg"
     if os.path.exists(logo_path):

style.css

@@ -29,14 +29,17 @@ h4 {
 
 /* Butoane */
 
-button {
+button,
+.myButton {
 	font-weight: 600 !important;
 }
-button:hover {
+button:hover,
+.myButton:hover {
 	color: #FA6C03 !important;
 	border-color: #FA6C03 !important;
 }
-button:focus {
+button:focus,
+.myButton:focus {
 	color: #fff !important;
 	background-color: #FA6C03 !important;
 }
@@ -57,7 +60,8 @@ button:focus {
 }
 
 /* Select */
-.st-ae {
+.st-ae,
+.st-az {
 	font-family: inherit !important;
 }
 
@@ -67,6 +71,6 @@ button:focus {
 }
 
 /* Other fonts */
-p {
-	font-family: 'Inter Tight', Arial, Helvetica, sans-serif !important;
+[class^="st-emotion-cache-"] {
+  font-family: 'Inter Tight', Arial, Helvetica, sans-serif !important;
 }