init

.env

@@ -0,0 +1,7 @@
+RUNPOD_KEY = 80SUW0PDAG3SDFSS2JA0E5QFCIMECU7SCXUTNH3A
+RUNPOD_URL = https://api.runpod.ai/v2/vllm-s4x8uw9y3qv296/openai/v1
+# QDRANT_URL = https://2fee3feb-59b4-43ca-adef-ffaaacd4e0c6.europe-west3-0.gcp.cloud.qdrant.io
+QDRANT_URL = https://7fcf6c04-2f9f-40f3-b26b-c19885487d64.europe-west3-0.gcp.cloud.qdrant.io
+# QDRANT_KEY = X2zkjP-ruS0uvdDEWbj2ZxeMVtmDbG1WXH6lQdGUnOYjT_-ODzqWZQ
+QDRANT_KEY = c-p0ej858dskulydMoLpokCLK2Q2phWXTUHjrKBYYRz3y-g7YSKZ4A
+OPENAI_API_KEY = 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

.gitignore

@@ -0,0 +1 @@
+.env

app.py

@@ -1,62 +1,95 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from openai import OpenAI
+import os
+from dotenv import load_dotenv
+from langchain.prompts import PromptTemplate
+from qdrant_client import QdrantClient
+from qdrant_client.http.models import Distance, VectorParams
+from langchain_qdrant import QdrantVectorStore
+from langchain_openai import OpenAIEmbeddings
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+load_dotenv()
+
+RUNPOD_KEY = os.getenv("RUNPOD_KEY")
+RUNPOD_URL = os.getenv("RUNPOD_URL")
+model = OpenAI(api_key=RUNPOD_KEY, base_url= RUNPOD_URL)
 
+QDRANT_URL = os.getenv("QDRANT_URL")
+QDRANT_KEY = os.getenv("QDRANT_KEY")
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+print(QDRANT_URL)
+print(QDRANT_KEY)
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+print(OPENAI_API_KEY)
 
-    messages.append({"role": "user", "content": message})
+client = QdrantClient(QDRANT_URL, api_key=QDRANT_KEY)
+collection_name = "search_engine"
 
-    response = ""
+embeddings = OpenAIEmbeddings(
+    model="text-embedding-3-small",
+    openai_api_key=OPENAI_API_KEY
+)
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+qdrant = QdrantVectorStore(
+    client=client,
+    collection_name=collection_name,
+    embedding=embeddings
+)
 
-        response += token
-        yield response
 
+
+
+promtp_template = """
+Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
+
+### Instruction:
+{instruction}
+
+### Input:
+{input}
+
+### Response:
 """
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
+
+prompt = PromptTemplate(
+    input_variables=["instruction", "input"],
+    template=promtp_template,
+    )   
+
+def prompt_template(query):
+    
+    results = qdrant.similarity_search( query=query, k=3 )
+
+    _ctx = ''
+
+    for i, result in enumerate(results):
+        _ctx += f'Content {i}: {result.page_content}\n-----\n'
+
+    _prompt = prompt.format(instruction=query, input=_ctx)
+
+    return _prompt
+
+def generate_response(prompt):
+        
+    response = model.chat.completions.create(
+    model="cenrak/llama3.1_fineTuned_model",
+    messages=[
+    {"role": "system", "content": "You are a helpful assistant."},
+    {"role": "user", "content": prompt},
+    ]
+    )
+
+    return response.choices[0].message
+
+def main(query, history):
+    prompt = prompt_template(query)
+    resault = generate_response(prompt)
+
+
+    return resault.content
+
+demo = gr.ChatInterface(fn=main, title = "News GPT")
 
 
 if __name__ == "__main__":

requirements.txt

@@ -1 +1,10 @@
-huggingface_hub==0.22.2
+openai 
+python-dotenv
+spacy
+langchain
+langchain_community
+langchain-openai
+langchain-qdrant 
+qdrant-client
+tqdm 
+spacy