bigsnarfdude

(harness) vincent@virus:~/Downloads$ cat bleu_text.py
from nltk.translate.bleu_score import sentence_bleu
reference = [
    'this is a dog'.split(),
    'it is dog'.split(),
    'dog it is'.split(),
    'a dog, it is'.split()
]
candidate = 'it is dog'.split()
print('BLEU score -> {}'.format(sentence_bleu(reference, candidate )))

bigsnarfdude

# requires A100 40GB - 30gb VRAM

from transformers import AutoModelForCausalLM, AutoTokenizer

device = "cuda"
model = AutoModelForCausalLM.from_pretrained("kaist-ai/mistral-orpo-capybara-7k").to(device)
tokenizer = AutoTokenizer.from_pretrained("kaist-ai/mistral-orpo-capybara-7k")
query = [{'role': 'user', 'content': 'Tell me how AI is like the Industrial Revolution'}]
prompt = tokenizer.apply_chat_template(query, tokenize=False, add_generation_prompt=True)
inputs = tokenizer (prompt, return_tensors='pt').to(device)

bigsnarfdude

import os
import time
import datetime

import pandas as pd
import seaborn as sns
import numpy as np
import random

import matplotlib.pyplot as plt

bigsnarfdude

bigsnarfdude

The AI Revolution:

Echoes of the Industrial Age

The rise of Artificial Intelligence (AI) is often compared to the Industrial Revolution, and for good reason. Both represent periods of significant technological advancement that have fundamentally reshaped societies and economies. While separated by centuries, striking parallels emerge in their impact on labor, production, and the overall fabric of human life.

Transformation of Labor:

Industrial Revolution:

bigsnarfdude

#!/bin/zsh

# Check if the video filename is provided as an argument
if [ $# -eq 0 ]; then
    echo "Please provide the video filename as an argument."
    exit 1
fi

video_filename=$1

bigsnarfdude

bigsnarfdude

Instruction-following prompts for ChatGPT, GPT-3.5, GPT-4

This project consists of prompts for ChatGPT and GPT-3.5 models, designed to assist with writing, analysis, and comprehension tasks. There are numerous prompts below that you can use to generate content for your projects, debug your code, find solutions to problems, or simply learn more about what these models can do. By using the appropriate instructional verbs, you can guide the models to solve any language-related tasks.

InstructGPT (more recent versions refered as GPT-3.5) are series language models that has been trained using instructions and human feedback to better understand and align with a user's intent, producing more accurate and appropriate outputs.

ChatGPT also uses instructGPT method but in a dialogue form to understand user instruction along and generate outputs based on user's instruct.

GPT4 More powerful than any GPT-3.5

bigsnarfdude

Wed Apr 24 00:45:56 2024
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.129.03             Driver Version: 535.129.03   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA A100-SXM4-40GB          On  | 00000000:07:00.0 Off |                    0 |
| N/A   56C    P0             367W / 400W |  20643MiB / 40960MiB |    100%      Default |

bigsnarfdude

# pretraining -> supervised instruction-finetuning -> RLHF

import json
import tiktoken

# open file
def extract_text_from_jsonl(file_path):
    prompts = []
    completions = []
    with open(file_path, 'r') as file: