chore: autopublish 2024-04-07T17:03:29Z

notebooks/RAG_LLM_experiments.ipynb

@@ -305,13 +305,19 @@
     "        total_cost = self.embedder.generate_embeddings()\n",
     "        self.embedder.create_faiss_index()\n",
     "        return total_cost\n",
-    "        \n",
+    "\n",
     "    def find_similar_documents(self, query_text, num_results):\n",
     "        similar_docs = self.embedder.search_similar_sections(query_text, num_results)\n",
     "        return similar_docs\n",
     "\n",
     "    def determine_expertise_and_prepare_prompt(\n",
-    "        self, user_query, similar_docs, inference_model, max_completion_tokens, temperature):\n",
+    "        self,\n",
+    "        user_query,\n",
+    "        similar_docs,\n",
+    "        inference_model,\n",
+    "        max_completion_tokens,\n",
+    "        temperature,\n",
+    "    ):\n",
     "        self.model_inference_manager.set_model(inference_model)\n",
     "        identified_expertise_area, expertise_area_usage = (\n",
     "            self.model_inference_manager.determine_expertise_area(\n",
@@ -321,7 +327,9 @@
     "        # Calculate the cost for determining expertise area\n",
     "        expertise_area_cost = 0  # Default to 0\n",
     "        if expertise_area_usage:\n",
-    "            expertise_area_cost = self.model_inference_manager.calculate_cost(expertise_area_usage)\n",
+    "            expertise_area_cost = self.model_inference_manager.calculate_cost(\n",
+    "                expertise_area_usage\n",
+    "            )\n",
     "        context_enhanced_prompt = self.model_inference_manager.prepare_prompt_for_llm(\n",
     "            identified_expertise_area, user_query, similar_docs\n",
     "        )\n",
@@ -334,7 +342,7 @@
     "            context_enhanced_prompt, max_completion_tokens, temperature\n",
     "        )\n",
     "        # Calculate the cost for querying the model\n",
-    "        response_cost = 0  # Default to 0 \n",
+    "        response_cost = 0  # Default to 0\n",
     "        if response_usage:\n",
     "            response_cost = self.model_inference_manager.calculate_cost(response_usage)\n",
     "\n",
@@ -492,12 +500,14 @@
     ")\n",
     "\n",
     "# Step 3: Determine expertise area and prepare the prompt\n",
-    "context_enhanced_prompt, expertise_area_cost = query_pipeline.determine_expertise_and_prepare_prompt(\n",
-    "    user_query=query_text,\n",
-    "    similar_docs=similar_docs,\n",
-    "    inference_model=\"gpt-3.5-turbo-0125\",\n",
-    "    max_completion_tokens=150,\n",
-    "    temperature=0.2,\n",
+    "context_enhanced_prompt, expertise_area_cost = (\n",
+    "    query_pipeline.determine_expertise_and_prepare_prompt(\n",
+    "        user_query=query_text,\n",
+    "        similar_docs=similar_docs,\n",
+    "        inference_model=\"gpt-3.5-turbo-0125\",\n",
+    "        max_completion_tokens=150,\n",
+    "        temperature=0.2,\n",
+    "    )\n",
     ")\n",
     "print(f\"Cost for determining expertise area: ${expertise_area_cost}\")\n",
     "\n",
@@ -581,12 +591,14 @@
     ")\n",
     "\n",
     "# Step 3: Determine expertise area and prepare the prompt\n",
-    "context_enhanced_prompt, expertise_area_cost = query_pipeline.determine_expertise_and_prepare_prompt(\n",
-    "    user_query=query_text,\n",
-    "    similar_docs=similar_docs,\n",
-    "    inference_model=\"gpt-3.5-turbo-0125\",\n",
-    "    max_completion_tokens=150,\n",
-    "    temperature=0.2,\n",
+    "context_enhanced_prompt, expertise_area_cost = (\n",
+    "    query_pipeline.determine_expertise_and_prepare_prompt(\n",
+    "        user_query=query_text,\n",
+    "        similar_docs=similar_docs,\n",
+    "        inference_model=\"gpt-3.5-turbo-0125\",\n",
+    "        max_completion_tokens=150,\n",
+    "        temperature=0.2,\n",
+    "    )\n",
     ")\n",
     "print(f\"Cost for determining expertise area: ${expertise_area_cost}\")\n",
     "\n",