missing files from rebase

config/settings.py

@@ -0,0 +1,44 @@
+import os
+
+from dotenv import find_dotenv, load_dotenv
+
+_ = load_dotenv(find_dotenv())
+
+LOG_LEVEL = os.environ.get("LOG_LEVEL", "DEBUG")
+
+PROJECT_PATH = os.environ.get("PROJECT_PATH")
+OPENSEARCH_URL = os.environ.get("OPENSEARCH_URL")
+OPENSEARCH_BATCH_SIZE = int(os.environ.get("OPENSEARCH_BATCH_SIZE", 500))
+OPENSEARCH_ENDPOINT_NAME = os.environ.get("OPENSEARCH_ENDPOINT_NAME", "vstore")
+OPENSEARCH_INDEX_NAME = os.environ.get("OPENSEARCH_INDEX_NAME", "")
+EMBEDDING_ENDPOINT_NAME = os.environ.get(
+    "EMBEDDING_ENDPOINT_NAME", "huggingface-sentencesimilarity"
+)
+
+S3_LOADER_BUCKET = os.environ.get("S3_LOADER_BUCKET", "")
+S3_LOADER_FILE_NAME = os.environ.get("S3_LOADER_FILE_NAME", "")
+AWS_REGION = os.environ.get("AWS_REGION", "eu-west-2")
+ENV = os.environ.get("ENV", "prod")
+
+
+LOADER_CONFIG = os.environ.get("LOADER_CONFIG", "s3_loader")  # "file_loader"
+VECTOR_STORE_CONFIG = os.environ.get(
+    "VECTOR_STORE_CONFIG", "opensearch"
+)  # "opensearch"
+LLM_MODEL = os.environ.get("LLM_MODEL", "hosted_llm")  # "local_llm"
+
+AWS_REGION = os.environ.get("AWS_REGION", "eu-west-2")
+AWS_SERVICE_ROLE_ARN = os.environ.get("AWS_SERVICE_ROLE_ARN", "arn:aws:iam::")
+AWS_SAGEMAKER_ENDPOINT = os.environ.get(
+    "AWS_SAGEMAKER_ENDPOINT", "llama-meta-textgeneration"
+)
+
+MODEL = "claude-v3-sonnet"
+SUMMARISE_API = os.environ.get("SUMMARISE_API")
+SUMMARISE_URL = os.environ.get("SUMMARISE_URL")
+FACTCHECK_API = os.environ.get("FACTCHECK_API")
+FACTCHECK_URL = os.environ.get("FACTCHECK_URL")
+GLOSSERY_API = os.environ.get("GLOSSERY_API")
+GLOSSERY_URL = os.environ.get("GLOSSERY_URL")
+CONVERSATION_API = os.environ.get("CONVERSATION_API")
+CONVERSATION_URL = os.environ.get("CONVERSATION_URL")

hackathon/constants/constants.py

@@ -0,0 +1,2 @@
+METADATA_COLUMNS = []
+CONTENT_COLUMNS = [""]

hackathon/llm/chain_config.py

@@ -0,0 +1,64 @@
+from dataclasses import dataclass
+from operator import itemgetter
+from typing import Dict, List
+
+from langchain.schema.output_parser import StrOutputParser
+from langchain_core.output_parsers.transform import BaseTransformOutputParser
+
+from hackathon.llm.prompts.core import PromptTemplate
+
+
+@dataclass
+class ChainType:
+    """
+    A Chain type that a ChainConfig can be set to, this will determine which LLM Chain this config is set up with.
+    """
+
+    identifier: str
+    description: str
+
+
+SINGLE_CHAIN = ChainType(
+    identifier="SINGLE_CHAIN",
+    description="A simple chain which takes var_input for the prompt, a prompt, llm and an output parser",
+)
+
+
+@dataclass
+class ChainConfig:
+    """
+    Configuration for an LLM chain.
+    The class uses a custome initilisation method to set up various cnfiguration parameters.
+
+    Attributes:
+        name (str): The name of the LLM chain.
+        prompt (PromptTemplate): prompt template for the chain to use
+        input_values (List[str]): input values the prompt takes
+        input_format (str) = format of the input for the input values defaults to 'Dict' (Currently not functional)
+        out_parser (BaseTransformOutputParser) = output parser for the LLM Chain to use.
+    """
+
+    name: str
+    prompt: PromptTemplate
+    input_values: List[str]
+    input_format: str = "Dict"
+    # context:Optional[str]  = None
+    out_parser: BaseTransformOutputParser = StrOutputParser
+    chain_type: ChainType = SINGLE_CHAIN
+
+    def __init__(
+        self,
+        name: str,
+        prompt: PromptTemplate,
+        chain_type: ChainType = SINGLE_CHAIN,
+        input_values: List[str] = ["input"],
+        input_format: str = "Dict",
+        # context: Optional[str] = None,
+        out_parser=StrOutputParser,
+    ):
+        self.name = name
+        self.prompt = prompt
+        self.input_format = input_format
+        self.var_input = {key: itemgetter(key) for key in input_values}
+        self.out_parser = out_parser
+        self.chain_type = chain_type

hackathon/llm/prompts/core.py

@@ -0,0 +1,41 @@
+from langchain.prompts.prompt import PromptTemplate
+
+B_INST, E_INST = "[INST]", "[/INST]"
+B_SYS, E_SYS = "<<SYS>>\n", "\n<</SYS>>\n\n"
+
+_advice_system_prompt = """
+
+"""
+_advice_instruction = """
+===  ===                                           
+{}
+===  ===
+{context}
+=======
+Question: , {}.
+=======
+Answer:
+"""
+
+
+_advice_prompt = (
+    B_INST + B_SYS + _advice_system_prompt + E_SYS + _advice_instruction + E_INST
+)
+
+ADVICE_PROMPT = PromptTemplate.from_template(_advice_prompt)
+
+
+SUMMARISE_PROMPT = PromptTemplate.from_template(
+    """
+Summarise the 
+
+: {}
+----------------------
+ANSWER:
+"""
+)
+
+document_retrieval_prompt = PromptTemplate.from_template(
+    """
+"""
+)