From 0b0a0e09c074d56c399e1b623748c0b00bf89932 Mon Sep 17 00:00:00 2001
From: Owen <ocolegro@gmail.com>
Date: Tue, 31 Oct 2023 10:50:07 -0400
Subject: [PATCH] fix chat logic

---
 docs/source/api/main.rst                 | 22 ++++++++--------------
 sciphi/interface/llm/sciphi_interface.py |  2 +-
 2 files changed, 9 insertions(+), 15 deletions(-)

diff --git a/docs/source/api/main.rst b/docs/source/api/main.rst
index 06a523c..fcd755f 100644
--- a/docs/source/api/main.rst
+++ b/docs/source/api/main.rst
@@ -7,6 +7,7 @@ Endpoint Overview
 -----------------
 
 1. **Search**: This endpoint allows you to fetch related documents based on a set of queries. The documents are retrieved by re-ranked similarity search over embeddings produced by the `facebook/contriever <https://huggingface.co/facebook/contriever>`_. As of now, only Wikipedia is embedded, but there are plans to expand this to a more comprehensive corpus using state-of-the-art embedding methods.
+
 2. **OpenAI Formatted LLM Request (v1)**: SciPhi models are served via an API that is compatible with the OpenAI API.
 
 Detailed Endpoint Descriptions
@@ -35,12 +36,6 @@ A list of lists containing Document objects, where each list corresponds to the
         -H "Content-Type: application/json" \
         -d '{"queries": ["What is general relativity?", "Who is Albert Einstein?"], "top_k": 5}'
 
-The expected response is:
-
-.. code-block:: none
-
-    [[{"id":14678539,"title":"General Relativity and Gravitation","text":"General Relativity and Gravitation General Re ...
-
 
 SciPhi v1 Endpoints
 ~~~~~~~~~~~~~~~~~~~
@@ -60,7 +55,7 @@ SciPhi adheres to the API specification of OpenAI's API, allowing compatibility
          "temperature": 0.7
        }'
 
-The expected response is:
+**Response**:
 
 .. code-block:: json
 
@@ -77,15 +72,14 @@ The expected response is:
                 "finish_reason":"length"
             }
         ],
-        "usage":
-            {
-                "prompt_tokens":7,
-                "total_tokens":15,
-                "completion_tokens":8
-            }
+        "usage": {
+            "prompt_tokens":7,
+            "total_tokens":15,
+            "completion_tokens":8
+        }
     }
 
 API Key and Signup
 ------------------
 
-To access the SciPhi API, you need an API key. If you don't possess one, you can sign up `here <https://www.sciphi.ai/signup>`_. Ensure you include the API key in your request headers as shown in the examples.
\ No newline at end of file
+To access the SciPhi API, you need an API key. If you don't possess one, you can sign up `here <https://www.sciphi.ai/signup>`_. Ensure you include the API key in your request headers as shown in the examples.
diff --git a/sciphi/interface/llm/sciphi_interface.py b/sciphi/interface/llm/sciphi_interface.py
index ed8107d..f26ee7a 100644
--- a/sciphi/interface/llm/sciphi_interface.py
+++ b/sciphi/interface/llm/sciphi_interface.py
@@ -99,7 +99,7 @@ def get_chat_completion(
             prompt = f"### System:\n{SciPhiLLMInterface.ALPACA_CHAT_SYSTEM_PROMPT}.\n\n{prompt}"
 
         context = self.rag_interface.get_contexts([last_user_message])[0]
-        prompt += f"{SciPhiFormatter.RETRIEVAL_TOKEN} {SciPhiFormatter.INIT_PARAGRAPH_TOKEN}{context}{SciPhiFormatter.END_PARAGRAPH_TOKEN}"
+        prompt += f"### Response:\n{SciPhiFormatter.RETRIEVAL_TOKEN} {SciPhiFormatter.INIT_PARAGRAPH_TOKEN}{context}{SciPhiFormatter.END_PARAGRAPH_TOKEN}"
         latest_completion = self.model.get_instruct_completion(
             prompt, generation_config
         ).strip()