finishing the main feature

.env

@@ -0,0 +1,7 @@
+COLLECTION_NAME=sales_role_play
+COLLECTION_SIZE=1536
+QDRANT_PORT=6333
+QDRANT_HOST=e47c5f9f-b98c-4eda-b28d-062e0cdbebda.eu-central-1-0.aws.cloud.qdrant.io
+QDRANT_API_KEY=VIO8ss9siM8p48NpLYIhjHVKG4S5sufwFvD8g3AiSnXz_hXwQLi7tQ
+OPENAI_API_KEY=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
+OPENAI_ORGANIZATION=org-vzVB2Aj8Ipkxkee7l6w0e2GH

app.py

@@ -1,9 +1,22 @@
 import gradio as gr
+from qdrant import qdrant_manager
+from openai_manager import openai_manager
 
 
-def greet(name):
-    return "Hello " + name + "!!"
+def generate(keywords):
+    try:
+        keywords_list = list(map(lambda x: x.strip(), keywords.split(",")))
+    except:
+        keywords_list = []
+        gr.Warning("Please use ',' to separate Keywords")
 
+    print("kewords", " ".join(keywords_list))
+    embedding = openai_manager.get_embeddings("hello my name is zinou")
 
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
+    points = qdrant_manager.search_point(query_vector=embedding[0])
+
+    return openai_manager.shots(points, " ".join(keywords_list))
+
+
+iface = gr.Interface(fn=generate, inputs="text", outputs="text")
 iface.launch()

openai_manager.py

@@ -0,0 +1,104 @@
+import os
+from dotenv import load_dotenv
+import openai
+
+
+class OpenAiManager:
+    def __init__(self):
+        load_dotenv()
+        openai.api_key = os.getenv("OPENAI_API_KEY")
+        openai.organization = os.getenv("OPENAI_ORGANIZATION")
+
+    def get_completion(
+        self,
+        prompt,
+        model="text-davinci-003",
+        max_tokens=128,
+        temperature=0,
+    ):
+        response = None
+        try:
+            response = openai.Completion.create(
+                prompt=prompt,
+                max_tokens=max_tokens,
+                model=model,
+                temperature=temperature,
+            )["choices"][0]["text"]
+
+        except Exception as err:
+            print(f"Sorry, There was a problem \n\n {err}")
+
+        return response
+
+    def get_chat_completion(self, prompt, model="gpt-3.5-turbo"):
+        response = None
+        try:
+            response = (
+                openai.ChatCompletion.create(
+                    model=model,
+                    messages=[
+                        {
+                            "role": "system",
+                            "content": prompt,
+                        }
+                    ],
+                )
+                .choices[0]
+                .message.content.strip()
+            )
+
+        except Exception as err:
+            print(f"Sorry, There was a problem \n\n {err}")
+
+        return response
+
+    def get_embedding(self, prompt, model="text-embedding-ada-002"):
+        prompt = prompt.replace("\n", " ")
+
+        embedding = None
+        try:
+            embedding = openai.Embedding.create(input=[prompt], model=model)["data"][0][
+                "embedding"
+            ]
+
+        except Exception as err:
+            print(f"Sorry, There was a problem {err}")
+
+        return embedding
+
+    def get_embeddings(self, prompts, model="text-embedding-ada-002"):
+        prompts = [prompt.replace("\n", " ") for prompt in prompts]
+
+        embeddings = None
+        try:
+            embeddings = openai.Embedding.create(input=prompts, model=model)["data"]
+
+        except Exception as err:
+            print(f"Sorry, There was a problem {err}")
+
+        return [embedding["embedding"] for embedding in embeddings]
+
+    def shots(self, examples, keywords):
+        prompt = []
+
+        for example in examples:
+            prompt.append(
+                f"""
+                keywords: {example.payload["keywords"]}
+                script: {example.payload["example"]}
+                """
+            )
+
+        prompt.append(
+            f"""
+            keywords: {keywords}
+            script: 
+            """
+        )
+
+        prompt = "\n\n".join(prompt)
+
+        return self.get_chat_completion(prompt=prompt)
+
+
+openai_manager = OpenAiManager()

qdrant.py

@@ -0,0 +1,101 @@
+import os
+from qdrant_client import QdrantClient
+from qdrant_client.http import models
+
+from dotenv import load_dotenv
+from uuid import uuid4
+
+load_dotenv()
+
+COLLECTION_NAME = os.getenv("COLLECTION_NAME")
+COLLECTION_SIZE = os.getenv("COLLECTION_SIZE")
+QDRANT_PORT = os.getenv("QDRANT_PORT")
+QDRANT_HOST = os.getenv("QDRANT_HOST")
+QDRANT_API_KEY = os.getenv("QDRANT_API_KEY")
+
+
+class QdrantManager:
+    """
+    A class for managing collectionsget_collection_info in the Qdrant database.
+
+    Args:
+        collection_name (str): The name of the collection to manage.
+        collection_size (int): The maximum number of documents in the collection.
+        port (int): The port number for the Qdrant API.
+        host (str): The hostname or IP address for the Qdrant server.
+        api_key (str): The API key for authenticating with the Qdrant server.
+        recreate_collection (bool): Whether to recreate the collection if it already exists.
+
+    Attributes:
+        client (qdrant_client.QdrantClient): The Qdrant client object for interacting with the API.
+    """
+
+    def __init__(
+        self,
+        collection_name=COLLECTION_NAME,
+        collection_size: int = COLLECTION_SIZE,
+        port: int = QDRANT_PORT,
+        host=QDRANT_HOST,
+        api_key=QDRANT_API_KEY,
+        recreate_collection: bool = False,
+    ):
+        self.collection_name = collection_name
+        self.collection_size = collection_size
+        self.host = host
+        self.port = port
+        self.api_key = api_key
+
+        self.client = QdrantClient(host=host, port=port, api_key=api_key)
+        self.setup_collection(collection_size, recreate_collection)
+
+    def setup_collection(self, collection_size: int, recreate_collection: bool):
+        if recreate_collection:
+            self.recreate_collection()
+
+        try:
+            collection_info = self.get_collection_info()
+            current_collection_size = collection_info["vector_size"]
+
+            if current_collection_size != int(collection_size):
+                raise ValueError(
+                    f"""
+                    Existing collection {self.collection_name} has different collection size
+                    To use the new collection configuration, you need to recreate the collection as it already exists with a different configuration.
+                    use recreate_collection = True.
+                    """
+                )
+
+        except Exception as e:
+            print(e)
+
+    def recreate_collection(self):
+        self.client.recreate_collection(
+            collection_name=self.collection_name,
+            vectors_config=models.VectorParams(
+                size=self.collection_size, distance=models.Distance.COSINE
+            ),
+        )
+
+    def get_collection_info(self):
+        collection_info = self.client.get_collection(
+            collection_name=self.collection_name
+        )
+
+        return {
+            "points_count": int(collection_info.points_count),
+            "vectors_count": int(collection_info.vectors_count),
+            "indexed_vectors_count": int(collection_info.indexed_vectors_count),
+            "vector_size": int(collection_info.config.params.vectors.size),
+        }
+
+    def search_point(self, query_vector, limit=5):
+        response = self.client.search(
+            collection_name=self.collection_name,
+            query_vector=query_vector,
+            limit=limit,
+        )
+
+        return response
+
+
+qdrant_manager = QdrantManager()

requirements.txt

@@ -0,0 +1,4 @@
+gradio
+qdrant-client
+openai
+python-dotenv