flâneur

Amy Bot.ipynb - Colaboratory

colab.research.google.com · 4 words · saved by 1 readers

Amy Bot.ipynb_ File Edit View Insert Runtime Tools Help Comment Share Files .. .config sample_data amy_iMessages.csv Drop files to upload them to session storage Code Text import requests import pandas as pd import numpy as np from sklearn.metrics.pairwise import cosine_similarity df = pd.read_csv("amy_iMessages.csv") query = df['text'].tolist()[:100] # embeddings from iMessages r = requests.post("http://66.228.39.142/embed_batch", json={'query':query, 'model': 'multi-qa-mpnet-base-dot-v1'}) text_embeddings = r.json() len(text_embeddings) 100 # new message input input = "late" r = requests.post("http://66.228.39.142/embed_batch", json={'query':[input], 'model': 'multi-qa-mpnet-base-dot-v1'}) input_embedding = r.json() # find the top 5 similar text messages similarities = cosine_similarity(text_embeddings, input_embedding) similarities = [s[0] for s in similarities] indecies = np.argpartition(similarities, -5)[-5:] convos = set([df['conversation'][i] for i in indecies]) corpus = "" for

Google Colab Sign in

saved by

related reading