john-adeojo

from haystack.nodes import PreProcessor
from haystack.utils import convert_files_to_docs
from haystack.document_stores import FAISSDocumentStore
from sqlalchemy import create_engine
from haystack.nodes import EmbeddingRetriever

# pre-process docs 
def preprocess_docs(doc_dir):
    all_docs = convert_files_to_docs(dir_path=doc_dir)
    preprocessor = PreProcessor(

john-adeojo

from haystack.agents import Agent
from haystack.nodes import PromptNode, PromptTemplate, AnswerParser
from haystack.agents import Tool

def create_agent(document_qa, API_KEY):
    react_prompt = PromptTemplate(
        prompt="""You are a helpful and knowledgeable agent. To achieve your goal of answering complex questions
    correctly, you have access to the following tools:
    
    {tool_names_with_descriptions}

john-adeojo

from haystack.nodes import EmbeddingRetriever, FARMReader
from haystack.pipelines import ExtractiveQAPipeline
from haystack.pipelines import Pipeline

def make_document_qa_pipeline(document_store):
    retriever = EmbeddingRetriever(
        document_store=document_store,
        embedding_model="sentence-transformers/all-mpnet-base-v2"
    )
    document_store.update_embeddings(retriever)

john-adeojo

from haystack.nodes import PreProcessor
from haystack.utils import convert_files_to_docs
from haystack.document_stores import FAISSDocumentStore
from sqlalchemy import create_engine

# pre-process docs 
def preprocess_docs(doc_dir):
    all_docs = convert_files_to_docs(dir_path=doc_dir)
    preprocessor = PreProcessor(
        clean_empty_lines=True,

john-adeojo

import transformers
import evaluate
from sklearn.metrics import f1_score, precision_score, recall_score, accuracy_score

f1_metric = evaluate.load("f1")
recall_metric = evaluate.load("recall")
accuracy_metric = evaluate.load("accuracy")
precision_metric = evaluate.load("precision")

def compute_metrics(eval_pred):

john-adeojo

# Setting up the LoRA model
import os
os.environ["CUDA_VISIBLE_DEVICES"]="0"
from transformers import AutoModelForSequenceClassification
from peft import LoraConfig, get_peft_model, TaskType

MODEL ="xlm-roberta-large"

config = LoraConfig(
    task_type="SEQ_CLS",

john-adeojo

import openai

def return_topic_args(query, API_KEY):

    openai.api_key = API_KEY
    
    content = "You parse the arguments defined in the prompt"
    prompt = f""" From the {query}, Return the news topic required """
    
    response = openai.ChatCompletion.create(

john-adeojo

import openai

def function_calling_multiple_function(API_KEY, query):
    # Step 1: send the conversation and available functions to GPT
    functions = [
        {
            "name": "get_current_news",
            "description": "Get the latest news headline for a given topic",
            "parameters": {
                "type": "object",

john-adeojo

import openai

def generate_response(topic, API_KEY):

    """
    Generates a random news headline for a given ticker
    """
    
    openai.api_key = API_KEY
    

john-adeojo

def function_calling_one_function(API_KEY, query):
    # Step 1: send the conversation and available functions to GPT
    functions = [
        {
            "name": "get_temperature",
            "description": "Get the temperature at a given location and date",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {