Update app.py

app.py

@@ -3,6 +3,11 @@ from langchain.chains import LLMChain
 from langchain_community.llms import CTransformers
 from langchain_core.prompts import PromptTemplate
 from sentence_transformers import SentenceTransformer
+from sklearn.cluster import KMeans
+from nltk.tokenize import word_tokenize
+import numpy as np
+import scipy.spatial
+from scipy.spatial.distance import cosine
 
 
 def generate_prompts(user_input):
@@ -12,45 +17,49 @@ def generate_prompts(user_input):
     )
     config = {'max_new_tokens': 2048, 'temperature': 0.7, 'context_length': 4096}
     llm = CTransformers(model="TheBloke/Mistral-7B-Instruct-v0.1-GGUF",
-                        config=config)
+                        config=config,
+                        threads=os.cpu_count())
     hub_chain = LLMChain(prompt = prompt_template, llm = llm)
 
     input_data = {"Question": user_input}
 
     # Here you would integrate your prompt template with your model
     # For demonstration, this is just a placeholder
-    generated_prompts = hub_chain.run(input_data)  
+    generated_prompts = hub_chain.run(input_data)  # Modify this part based on how you run the model
     questions_list = generated_prompts.split('\n') 
     
 
     formatted_questions = "\n".join(f"Question: {question}" for i, question in enumerate(questions_list) if question.strip())
     questions_list = formatted_questions.split("Question:")[1:]
+
     
     return questions_list
 
-def answer_question(prompt):
-    prompt_template = PromptTemplate.from_template(
+def answer_question(prompt, model_name):
+    prompt_template = PromptTemplate(
         input_variables=["Question"],
-        template=f"give one answer for {prompt} and do not consider the number behind it."
+        template=f"Give a short answer to this question '{prompt}' and do not consider the number behind it."
     )
-    config = {'max_new_tokens': 2048, 'temperature': 0.7, 'context_length': 4096}
-    llm = CTransformers(model="TheBloke/Llama-2-7B-Chat-GGML",
+    config = {'max_new_tokens': 512, 'temperature': 0.7, 'context_length': 512}
+    llm = CTransformers(model=model_name, #"TheBloke/Llama-2-7B-Chat-GGML",
                         config=config,
                         threads=os.cpu_count())
     hub_chain = LLMChain(prompt = prompt_template, llm = llm)
+
     input_data = {"Question": prompt}
-    generated_answer = hub_chain.run(input_data)  
-    return generated_answer
 
-    def calculate_similarity(word, other_words, model, threshold=0.5):
-        embeddings_word = model.encode([word])
-        embeddings_other_words = model.encode(other_words)
-        for i, embedding in enumerate(embeddings_other_words):
-            similarity = 1 - scipy.spatial.distance.cosine(embeddings_word[0], embedding)
-            if similarity > threshold and similarity < 0.85:
-                return i, similarity
-        return None, None
 
+    generated_answer = hub_chain.run(input_data)  
+    
+    return generated_answer
+def calculate_similarity(word, other_words, model, threshold=0.5):
+    embeddings_word = model.encode([word])
+    embeddings_other_words = model.encode(other_words)
+    for i, embedding in enumerate(embeddings_other_words):
+        similarity = 1 - scipy.spatial.distance.cosine(embeddings_word[0], embedding)
+        if similarity > threshold and similarity < 0.85:
+            return i, similarity
+    return None, None
 
 def highlight_words_within_cluster(sentences, model, exclude_words):
     # Create a dictionary to map words to color codes
@@ -102,9 +111,6 @@ def highlight_words_within_cluster(sentences, model, exclude_words):
     return highlighted_sentences
 
 # Rest of the code, including the cluster_sentences function, remains the same
-
-exclude_words = {"a", "the", "for", "from", "of", "in","over", "as", "on", "is", "am", "have", "an","has", "had", "and", "by", "it", "its", "those", "these", "was", "were", "their", "them", "I", "you", "also", "your", "me", "after"}
-
 def cluster_sentences(sentences, model, num_clusters=3):
     embeddings = model.encode(sentences)
     kmeans = KMeans(n_clusters=num_clusters)
@@ -215,6 +221,6 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
             prompt_input = gr.Textbox(label="Enter your email address", placeholder="Enter Your Email Address")
         with gr.Row():
             generate_button = gr.Button("Submit", variant="primary")
-
+            
 # Launch the Gradio app
 demo.launch()