vanna 0.6.2__tar.gz → 0.6.3__tar.gz

{vanna-0.6.2 → vanna-0.6.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vanna
-Version: 0.6.2
+Version: 0.6.3
 Summary: Generate SQL queries from natural language
 Author-email: Zain Hoda <zain@vanna.ai>
 Requires-Python: >=3.9
@@ -40,7 +40,10 @@ Requires-Dist: opensearch-dsl ; extra == "all"
 Requires-Dist: transformers ; extra == "all"
 Requires-Dist: pinecone-client ; extra == "all"
 Requires-Dist: pymilvus[model] ; extra == "all"
+Requires-Dist: weaviate-client ; extra == "all"
 Requires-Dist: anthropic ; extra == "anthropic"
+Requires-Dist: boto3 ; extra == "bedrock"
+Requires-Dist: botocore ; extra == "bedrock"
 Requires-Dist: google-cloud-bigquery ; extra == "bigquery"
 Requires-Dist: chromadb ; extra == "chromadb"
 Requires-Dist: clickhouse_connect ; extra == "clickhouse"
@@ -67,11 +70,13 @@ Requires-Dist: fastembed ; extra == "qdrant"
 Requires-Dist: snowflake-connector-python ; extra == "snowflake"
 Requires-Dist: tox ; extra == "test"
 Requires-Dist: vllm ; extra == "vllm"
+Requires-Dist: weaviate-client ; extra == "weaviate"
 Requires-Dist: zhipuai ; extra == "zhipuai"
 Project-URL: Bug Tracker, https://github.com/vanna-ai/vanna/issues
 Project-URL: Homepage, https://github.com/vanna-ai/vanna
 Provides-Extra: all
 Provides-Extra: anthropic
+Provides-Extra: bedrock
 Provides-Extra: bigquery
 Provides-Extra: chromadb
 Provides-Extra: clickhouse
@@ -92,6 +97,7 @@ Provides-Extra: qdrant
 Provides-Extra: snowflake
 Provides-Extra: test
 Provides-Extra: vllm
+Provides-Extra: weaviate
 Provides-Extra: zhipuai
 
 

{vanna-0.6.2 → vanna-0.6.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "flit_core.buildapi"
 
 [project]
 name = "vanna"
-version = "0.6.2"
+version = "0.6.3"
 authors = [
   { name="Zain Hoda", email="zain@vanna.ai" },
 ]
@@ -33,7 +33,7 @@ bigquery = ["google-cloud-bigquery"]
 snowflake = ["snowflake-connector-python"]
 duckdb = ["duckdb"]
 google = ["google-generativeai", "google-cloud-aiplatform"]
-all = ["psycopg2-binary", "db-dtypes", "PyMySQL", "google-cloud-bigquery", "snowflake-connector-python", "duckdb", "openai", "mistralai", "chromadb", "anthropic", "zhipuai", "marqo", "google-generativeai", "google-cloud-aiplatform", "qdrant-client", "fastembed", "ollama", "httpx", "opensearch-py", "opensearch-dsl", "transformers", "pinecone-client", "pymilvus[model]"]
+all = ["psycopg2-binary", "db-dtypes", "PyMySQL", "google-cloud-bigquery", "snowflake-connector-python", "duckdb", "openai", "mistralai", "chromadb", "anthropic", "zhipuai", "marqo", "google-generativeai", "google-cloud-aiplatform", "qdrant-client", "fastembed", "ollama", "httpx", "opensearch-py", "opensearch-dsl", "transformers", "pinecone-client", "pymilvus[model]","weaviate-client"]
 test = ["tox"]
 chromadb = ["chromadb"]
 openai = ["openai"]
@@ -49,3 +49,5 @@ pinecone = ["pinecone-client", "fastembed"]
 opensearch = ["opensearch-py", "opensearch-dsl"]
 hf = ["transformers"]
 milvus = ["pymilvus[model]"]
+bedrock = ["boto3", "botocore"]
+weaviate = ["weaviate-client"]

{vanna-0.6.2 → vanna-0.6.3}/src/vanna/base/base.py

@@ -182,7 +182,7 @@ class VannaBase(ABC):
         """
 
         # If the llm_response contains a CTE (with clause), extract the last sql between WITH and ;
-        sqls = re.findall(r"WITH.*?;", llm_response, re.DOTALL)
+        sqls = re.findall(r"\bWITH\b .*?;", llm_response, re.DOTALL)
         if sqls:
             sql = sqls[-1]
             self.log(title="Extracted SQL", message=f"{sql}")

vanna-0.6.3/src/vanna/bedrock/__init__.py

@@ -0,0 +1 @@
+from .bedrock_converse import Bedrock_Converse

vanna-0.6.3/src/vanna/bedrock/bedrock_converse.py

@@ -0,0 +1,85 @@
+from ..base import VannaBase
+
+try:
+    import boto3
+    from botocore.exceptions import ClientError
+except ImportError:
+    raise ImportError("Please install boto3 and botocore to use Amazon Bedrock models")
+
+class Bedrock_Converse(VannaBase):
+    def __init__(self, client=None, config=None):
+        VannaBase.__init__(self, config=config)
+
+        # default parameters
+        self.temperature = 0.0
+        self.max_tokens = 500
+        
+        if client is None:
+            raise ValueError(
+                "A valid Bedrock runtime client must be provided to invoke Bedrock models"
+            )
+        else:
+            self.client = client
+        
+        if config is None:
+            raise ValueError(
+                "Config is required with model_id and inference parameters"
+            )
+        
+        if "modelId" not in config:
+            raise ValueError(
+                "config must contain a modelId to invoke"
+            )
+        else:
+            self.model = config["modelId"]
+        
+        if "temperature" in config:
+            self.temperature = config["temperature"]
+        
+        if "max_tokens" in config:
+            self.max_tokens = config["max_tokens"]
+
+    def system_message(self, message: str) -> dict:
+        return {"role": "system", "content": message}
+
+    def user_message(self, message: str) -> dict:
+        return {"role": "user", "content": message}
+
+    def assistant_message(self, message: str) -> dict:
+        return {"role": "assistant", "content": message}
+
+    def submit_prompt(self, prompt, **kwargs) -> str:
+        inference_config = {
+            "temperature": self.temperature,
+            "maxTokens": self.max_tokens
+        }
+        additional_model_fields = {
+            "top_p": 1,  # setting top_p value for nucleus sampling
+        }
+
+        system_message = None
+        no_system_prompt = []
+        for prompt_message in prompt:
+            role = prompt_message["role"]
+            if role == "system":
+                system_message = prompt_message["content"]
+            else:
+                no_system_prompt.append({"role": role, "content":[{"text": prompt_message["content"]}]})
+
+        converse_api_params = {
+            "modelId": self.model,
+            "messages": no_system_prompt,
+            "inferenceConfig": inference_config,
+            "additionalModelRequestFields": additional_model_fields
+        }
+
+        if system_message:
+            converse_api_params["system"] = [{"text": system_message}]
+
+        try:
+            response = self.client.converse(**converse_api_params)
+            text_content = response["output"]["message"]["content"][0]["text"]
+            return text_content
+        except ClientError as err:
+            message = err.response["Error"]["Message"]
+            raise Exception(f"A Bedrock client error occurred: {message}")

{vanna-0.6.2 → vanna-0.6.3}/src/vanna/google/gemini_chat.py

@@ -7,7 +7,7 @@ class GoogleGeminiChat(VannaBase):
         VannaBase.__init__(self, config=config)
 
         # default temperature - can be overrided using config
-        self.temperature = 0.7      
+        self.temperature = 0.7
 
         if "temperature" in config:
             self.temperature = config["temperature"]
@@ -31,7 +31,7 @@ class GoogleGeminiChat(VannaBase):
         else:
             # Authenticate using VertexAI
             from vertexai.preview.generative_models import GenerativeModel
-            self.chat_model = GenerativeModel("gemini-pro")
+            self.chat_model = GenerativeModel(model_name)
 
     def system_message(self, message: str) -> any:
         return message

{vanna-0.6.2 → vanna-0.6.3}/src/vanna/hf/hf.py

@@ -6,13 +6,15 @@ from ..base import VannaBase
 
 class Hf(VannaBase):
     def __init__(self, config=None):
-        model_name = self.config.get(
-            "model_name", None
-        )  # e.g. meta-llama/Meta-Llama-3-8B-Instruct
-        self.tokenizer = AutoTokenizer.from_pretrained(model_name)
+        model_name_or_path = self.config.get(
+            "model_name_or_path", None
+        )  # e.g. meta-llama/Meta-Llama-3-8B-Instruct or local path to the model checkpoint files
+        # list of quantization methods supported by transformers package: https://huggingface.co/docs/transformers/main/en/quantization/overview
+        quantization_config = self.config.get("quantization_config", None)
+        self.tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
         self.model = AutoModelForCausalLM.from_pretrained(
-            model_name,
-            torch_dtype="auto",
+            model_name_or_path,
+            quantization_config=quantization_config,
             device_map="auto",
         )
 

vanna-0.6.3/src/vanna/weaviate/__init__.py

@@ -0,0 +1 @@
+from .weaviate_vector import WeaviateDatabase

vanna-0.6.3/src/vanna/weaviate/weaviate_vector.py

@@ -0,0 +1,174 @@
+import weaviate
+import weaviate.classes as wvc
+from fastembed import TextEmbedding
+
+from vanna.base import VannaBase
+
+
+class WeaviateDatabase(VannaBase):
+
+    def __init__(self, config=None):
+        """
+        Initialize the VannaEnhanced class with the provided configuration.
+
+        :param config: Dictionary containing configuration parameters.
+
+        params:
+        weaviate_url (str): Weaviate cluster URL while using weaviate cloud,
+        weaviate_api_key (str): Weaviate API key while using weaviate cloud,
+        weaviate_port (num): Weaviate port while using local weaviate,
+        weaviate_grpc (num): Weaviate gRPC port while using local weaviate,
+        fastembed_model (str): Fastembed model name for text embeddings. BAAI/bge-small-en-v1.5 by default.
+
+        """
+        super().__init__(config=config)
+
+        if config is None:
+            raise ValueError("config is required")
+
+        self.fastembed_model = config.get("fastembed_model", "BAAI/bge-small-en-v1.5")
+        self.weaviate_api_key = config.get("weaviate_api_key")
+        self.weaviate_url = config.get("weaviate_url")
+        self.weaviate_port = config.get("weaviate_port")
+        self.weaviate_grpc_port = config.get("weaviate_grpc", 50051)
+
+        if not self.weaviate_api_key and not self.weaviate_port:
+            raise ValueError("Add proper credentials to connect to weaviate")
+
+        self.weaviate_client = self._initialize_weaviate_client()
+        self.embeddings = TextEmbedding(model_name=self.fastembed_model)
+
+        self.training_data_cluster = {
+            "sql": "SQLTrainingDataEntry",
+            "ddl": "DDLEntry",
+            "doc": "DocumentationEntry"
+        }
+
+        self._create_collections_if_not_exist()
+
+    def _create_collections_if_not_exist(self):
+        properties_dict = {
+            self.training_data_cluster['ddl']: [
+                wvc.config.Property(name="description", data_type=wvc.config.DataType.TEXT),
+            ],
+            self.training_data_cluster['doc']: [
+                wvc.config.Property(name="description", data_type=wvc.config.DataType.TEXT),
+            ],
+            self.training_data_cluster['sql']: [
+                wvc.config.Property(name="sql", data_type=wvc.config.DataType.TEXT),
+                wvc.config.Property(name="natural_language_question", data_type=wvc.config.DataType.TEXT),
+            ]
+        }
+
+        for cluster, properties in properties_dict.items():
+            if not self.weaviate_client.collections.exists(cluster):
+                self.weaviate_client.collections.create(
+                    name=cluster,
+                    properties=properties
+                )
+
+    def _initialize_weaviate_client(self):
+        if self.weaviate_api_key:
+            return weaviate.connect_to_wcs(
+                cluster_url=self.weaviate_url,
+                auth_credentials=weaviate.auth.AuthApiKey(self.weaviate_api_key),
+                additional_config=weaviate.config.AdditionalConfig(timeout=(10, 300)),
+                skip_init_checks=True
+            )
+        else:
+            return weaviate.connect_to_local(
+                port=self.weaviate_port,
+                grpc_port=self.weaviate_grpc_port,
+                additional_config=weaviate.config.AdditionalConfig(timeout=(10, 300)),
+                skip_init_checks=True
+            )
+
+    def generate_embedding(self, data: str, **kwargs):
+            embedding_model = TextEmbedding(model_name=self.fastembed_model)
+            embedding = next(embedding_model.embed(data))
+            return embedding.tolist()
+
+
+    def _insert_data(self, cluster_key: str, data_object: dict, vector: list) -> str:
+        self.weaviate_client.connect()
+        response = self.weaviate_client.collections.get(self.training_data_cluster[cluster_key]).data.insert(
+            properties=data_object,
+            vector=vector
+        )
+        self.weaviate_client.close()
+        return response
+
+    def add_ddl(self, ddl: str, **kwargs) -> str:
+        data_object = {
+            "description": ddl,
+        }
+        response = self._insert_data('ddl', data_object, self.generate_embedding(ddl))
+        return f'{response}-ddl'
+
+    def add_documentation(self, doc: str, **kwargs) -> str:
+        data_object = {
+            "description": doc,
+        }
+        response = self._insert_data('doc', data_object, self.generate_embedding(doc))
+        return f'{response}-doc'
+
+    def add_question_sql(self, question: str, sql: str, **kwargs) -> str:
+        data_object = {
+            "sql": sql,
+            "natural_language_question": question,
+        }
+        response = self._insert_data('sql', data_object, self.generate_embedding(question))
+        return f'{response}-sql'
+
+    def _query_collection(self, cluster_key: str, vector_input: list, return_properties: list, limit: int = 3) -> list:
+        self.weaviate_client.connect()
+        collection = self.weaviate_client.collections.get(self.training_data_cluster[cluster_key])
+        response = collection.query.near_vector(
+            near_vector=vector_input,
+            limit=limit,
+            return_properties=return_properties
+        )
+        response_list = [item.properties for item in response.objects]
+        self.weaviate_client.close()
+        return response_list
+
+    def get_related_ddl(self, question: str, **kwargs) -> list:
+        vector_input = self.generate_embedding(question)
+        response_list = self._query_collection('ddl', vector_input, ["description"])
+        return [item["description"] for item in response_list]
+
+    def get_related_documentation(self, question: str, **kwargs) -> list:
+        vector_input = self.generate_embedding(question)
+        response_list = self._query_collection('doc', vector_input, ["description"])
+        return [item["description"] for item in response_list]
+
+    def get_similar_question_sql(self, question: str, **kwargs) -> list:
+        vector_input = self.generate_embedding(question)
+        response_list = self._query_collection('sql', vector_input, ["sql", "natural_language_question"])
+        return [{"question": item["natural_language_question"], "sql": item["sql"]} for item in response_list]
+
+    def get_training_data(self, **kwargs) -> list:
+        self.weaviate_client.connect()
+        combined_response_list = []
+        for collection_name in self.training_data_cluster.values():
+            if self.weaviate_client.collections.exists(collection_name):
+                collection = self.weaviate_client.collections.get(collection_name)
+                response_list = [item.properties for item in collection.iterator()]
+                combined_response_list.extend(response_list)
+        self.weaviate_client.close()
+        return combined_response_list
+
+    def remove_training_data(self, id: str, **kwargs) -> bool:
+        self.weaviate_client.connect()
+        success = False
+        if id.endswith("-sql"):
+            id = id.replace('-sql', '')
+            success = self.weaviate_client.collections.get(self.training_data_cluster['sql']).data.delete_by_id(id)
+        elif id.endswith("-ddl"):
+            id = id.replace('-ddl', '')
+            success = self.weaviate_client.collections.get(self.training_data_cluster['ddl']).data.delete_by_id(id)
+        elif id.endswith("-doc"):
+            id = id.replace('-doc', '')
+            success = self.weaviate_client.collections.get(self.training_data_cluster['doc']).data.delete_by_id(id)
+        self.weaviate_client.close()
+        return success