Spaces:
Sleeping
Sleeping
samyak152002
commited on
Commit
•
c201eb6
1
Parent(s):
a9fb77e
Update app.py
Browse files
app.py
CHANGED
@@ -1,16 +1,72 @@
|
|
1 |
import gradio as gr
|
2 |
-
|
|
|
|
|
|
|
3 |
|
4 |
-
# Load the
|
5 |
-
|
|
|
6 |
|
7 |
-
#
|
8 |
-
|
9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
return result
|
11 |
|
12 |
-
#
|
13 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
|
15 |
# Launch the app
|
16 |
-
|
|
|
|
1 |
import gradio as gr
|
2 |
+
import torch
|
3 |
+
import torch.nn.functional as F
|
4 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
5 |
+
import pandas as pd
|
6 |
|
7 |
+
# Load the dataset and create label mappings
|
8 |
+
df = pd.read_csv('bert_train.csv') # Update with the correct path
|
9 |
+
df["label"] = df["Label"]
|
10 |
|
11 |
+
# Create int2label and label2int mappings
|
12 |
+
int2label = {i: disease for i, disease in enumerate(df['label'].unique())}
|
13 |
+
label2int = {v: k for k, v in int2label.items()}
|
14 |
+
|
15 |
+
# Load the model and tokenizer
|
16 |
+
model_name = "your-huggingface-model-path" # Replace with your Hugging Face model path
|
17 |
+
model = AutoModelForSequenceClassification.from_pretrained(model_name)
|
18 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
19 |
+
|
20 |
+
# Function to classify text and return the top 3 diseases
|
21 |
+
def classify_text(text):
|
22 |
+
# Set device: GPU if available, else CPU
|
23 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
24 |
+
|
25 |
+
# Move the model to the correct device
|
26 |
+
model.to(device)
|
27 |
+
|
28 |
+
# Tokenize the input and move it to the same device as the model
|
29 |
+
inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True).to(device)
|
30 |
+
|
31 |
+
# Perform inference
|
32 |
+
outputs = model(**inputs)
|
33 |
+
|
34 |
+
# Get the logits (raw scores)
|
35 |
+
logits = outputs.logits
|
36 |
+
|
37 |
+
# Apply softmax to convert logits into probabilities
|
38 |
+
probabilities = F.softmax(logits, dim=1)
|
39 |
+
|
40 |
+
# Convert the probabilities tensor to a list for easy display
|
41 |
+
prob_list = probabilities[0].tolist()
|
42 |
+
|
43 |
+
# Zip together the disease labels with their respective probabilities
|
44 |
+
disease_probs = {int2label[i]: prob for i, prob in enumerate(prob_list)}
|
45 |
+
|
46 |
+
# Sort the diseases by their probabilities in descending order
|
47 |
+
sorted_disease_probs = dict(sorted(disease_probs.items(), key=lambda item: item[1], reverse=True))
|
48 |
+
|
49 |
+
# Get the top 3 diseases
|
50 |
+
top_3_diseases = list(sorted_disease_probs.items())[:3]
|
51 |
+
|
52 |
+
# Format the result for display
|
53 |
+
result = "\n".join([f"{disease}: {prob:.4f}" for disease, prob in top_3_diseases])
|
54 |
+
|
55 |
return result
|
56 |
|
57 |
+
# Gradio interface
|
58 |
+
def predict_disease(text):
|
59 |
+
return classify_text(text)
|
60 |
+
|
61 |
+
# Define the Gradio interface
|
62 |
+
iface = gr.Interface(
|
63 |
+
fn=predict_disease,
|
64 |
+
inputs="text",
|
65 |
+
outputs="text",
|
66 |
+
title="Disease Prediction",
|
67 |
+
description="Enter your symptoms, and the model will predict the top 3 most likely diseases with probabilities."
|
68 |
+
)
|
69 |
|
70 |
# Launch the app
|
71 |
+
if __name__ == "__main__":
|
72 |
+
iface.launch()
|