Update app.py

app.py

@@ -1,28 +1,44 @@
 import streamlit as st
-from transformers import AutoTokenizer, AutoModelForCausalLM
-import torch
+import tensorflow as tf
+import pickle
+import numpy as np
 
+# ----------------------
 # Load model & tokenizer
+# ----------------------
 @st.cache_resource
-def load_model():
-    tokenizer = AutoTokenizer.from_pretrained(".")
-    model = AutoModelForCausalLM.from_pretrained(".")
-    return tokenizer, model
+def load_assets():
+    model = tf.keras.models.load_model("nextword_model.h5")  # your saved model
+    with open("tokenizer.pkl", "rb") as f:                   # your saved tokenizer
+        tokenizer = pickle.load(f)
+    return model, tokenizer
 
-tokenizer, model = load_model()
+model, tokenizer = load_assets()
 
-st.title("📝 Next Word Prediction App")
-st.write("Type a sentence and let the model suggest the next word!")
+# ----------------------
+# Streamlit UI
+# ----------------------
+st.title("📝 LSTM Next Word Prediction")
+st.write("Type a sentence and let the LSTM suggest the next word!")
 
-# User input
-text = st.text_input("Enter your sentence:", "")
+# Input text
+text = st.text_input("Enter your sentence:")
 
-if st.button("Predict Next Word") and text:
-    inputs = tokenizer(text, return_tensors="pt")
-    with torch.no_grad():
-        outputs = model.generate(**inputs, max_new_tokens=1)
-    prediction = tokenizer.decode(outputs[0], skip_special_tokens=True)
+def predict_next_word(model, tokenizer, text, max_len=20):
+    """Predict next word from input text using trained LSTM model."""
+    seq = tokenizer.texts_to_sequences([text])[0]
+    seq = tf.keras.preprocessing.sequence.pad_sequences([seq], maxlen=max_len-1, padding='pre')
+
+    preds = model.predict(seq, verbose=0)[0]
+    next_index = np.argmax(preds)
+    for word, index in tokenizer.word_index.items():
+        if index == next_index:
+            return word
+    return None
 
-    # Extract only the new part
-    predicted_next = prediction[len(text):].strip()
-    st.success(f"**Predicted next word:** {predicted_next}")
+if st.button("Predict Next Word") and text:
+    predicted_word = predict_next_word(model, tokenizer, text)
+    if predicted_word:
+        st.success(f"**Predicted next word:** {predicted_word}")
+    else:
+        st.warning("Could not predict a word. Try another input.")