veekaybee

Anti-hype LLM reading list

Goals: Add links that are reasonable and good explanations of how stuff works. No hype and no vendor content if possible. Practical first-hand accounts of models in prod eagerly sought.

Foundational Concepts

Pre-Transformer Models

CharlyWargnier

        df = pd.read_csv("reuters-allcats.csv")
        csv = df.to_csv(index=False)       
        b64 = base64.b64encode(csv.encode()).decode()
        href = f'<a href="data:file/csv;base64,{b64}" download="myfilename.csv">Try a sample first</a>'

 
        st.markdown("👆 Upload a .csv file.  " + href, unsafe_allow_html=True)
      
        st.stop()

eliasdabbas

from urllib.parse import urlsplit

import advertools as adv


sites = [
    'https://www.who.int',
    'https://www.nytimes.com',
    'https://www.washingtonpost.com',
]

Sven-Bo

import json

import requests  # pip install requests
import streamlit as st  # pip install streamlit
from streamlit_lottie import st_lottie  # pip install streamlit-lottie

# GitHub: https://github.com/andfanilo/streamlit-lottie
# Lottie Files: https://lottiefiles.com/

def load_lottiefile(filepath: str):

ksv-muralidhar

import pandas as pd
from bs4 import BeautifulSoup
import requests as r
import streamlit as st

st.markdown('<h1 style="background-color: gainsboro; padding-left: 10px; padding-bottom: 20px;">Search Engine Scraper</h1>', unsafe_allow_html=True)
query = st.text_input('', help='Enter the search string and hit Enter/Return')
query = query.replace(" ", "+") #replacing the spaces in query result with +

if query: #Activates the code below on hitting Enter/Return in the search textbox

mtanco

import os
import time
from h2o_wave import main, app, Q, ui, data
import pandas as pd
import numpy as np


@app('/')
async def serve(q: Q):
    print(q.args)

mtanco

# Plot / Dataframe
# Examples of how to format pandas data when plotting
# Use the `tolist()` function on `df.columns` and `df.values` along with Wave's `data` class 
# ---
from h2o_wave import site, data, ui, main
import pandas as pd
import numpy as np

# Page to hold our charts
page = site['/demo']

hamletbatista

import streamlit as st
from ludwig.api import LudwigModel
import pandas as pd

st.cache(show_spinner=False)
def load_model():
  #Update with the path to the Ludwig trained model
  model = LudwigModel.load("results/experiment_run_1/model/")
  return model

andfanilo

{
  "nodes": [
    {
      "id": "0",
      "name": "Myriel",
      "itemStyle": null,
      "symbolSize": 10,
      "x": null,
      "y": null,
      "attributes": {

hamletbatista

from urllib.parse import urlparse
import re

url="https://www.amazon.com/SanDisk-128GB-microSDXC-Memory-Adapter/dp/B073JYC4XM/"

print(set(re.split("[/-]", urlparse(url).path)))

#output
#{'', 'B073JYC4XM', 'dp', '128GB', 'microSDXC', 'Memory', 'SanDisk', 'Adapter'}