kirill-fedyanin

class,text
Pos, by phil curtolo mel gibson   braveheart   gave a gripping performance as the father of a young kidnapped boy in ron howard s ransom    gibson plays tom mullen   a wealthy business tycoon whose past actions are coming back to haunt him as a deranged psychopath   played by gary sinise   forrest gump     and his band of low life thugs kidnap his only son for $2 million    tom and his wife   kate   played by rene russo   tin cup   were instructed not to inform the police   but they contacted the fbi    minutes later   an elite team of agents led by delroy lindo   broken arrow   are in tom s house and wiring every phone    the plot sounds average   just like most other kidnap movies that you ve already seen   and it was nothing more than that    that is until about half way through the movie    suddenly   tom goes to the fox 5 news room and makes a live broadcast saying     this is your ransom    but this is as close as you will ever get to it    instead   i am offering this money as a reward on yo

kirill-fedyanin

--- South of No North  Stories of the Buried Life.pdf ---
 

Charles Bukowski 
South Of No North. Stories of the buried life 
 
  
Table of contents 
Loneliness..................................................... ............................................................... ........... 1 
Bop bop against that curtain................................... ............................................................... .5  

kirill-fedyanin

"""
Script to quickly check that model answers make sense and not complete garbage
Also useful to check if a checkpoint runs at all
python scripts/arabic_sanity.py --model-name='mistralai/Mistral-7B-v0.1'
python scripts/arabic_sanity.py --model-name='./data/zephyr-7b-sft-lora/checkpoint-12'
"""
from argparse import ArgumentParser
from datasets import load_dataset
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

kirill-fedyanin

# sudo apt install sox -y
import os
from argparse import ArgumentParser
from time import sleep
import boto3


parser = ArgumentParser()
parser.add_argument('--instance-name', default='a100-kirill-2')
args = parser.parse_args()

kirill-fedyanin

================== raw =======================

Hello!Hi!How are you?

================== one_shot =======================

A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions.
### Human: Got any creative ideas for a 10 year old’s birthday?
### Assistant: Of course! Here are some creative ideas for a 10-year-old's birthday party:
1. Treasure Hunt: Organize a treasure hunt in your backyard or nearby park. Create clues and riddles for the kids to solve, leading them to hidden treasures and surprises.

kirill-fedyanin

# Essentially, hf model is just a git repo and you need to push data there.
# Create new model page on huggingface.co and put your credentials and model name in script below
# then you need your username and copy (or create) token from https://huggingface.co/settings/tokens
# Run this from folder with checkpoint
# Edit: lol, you could actually do just `huggingface-cli upload <username>/<model> .`

git init
git remote add origin https://<username>:<token>@huggingface.co/<username>/<model>
git fetch
git checkout main

kirill-fedyanin

BATCH_SIZE = args.batch_size
NUM_WORKERS = 4
MEAN = torch.tensor([0.485, 0.456, 0.406]) * 255
STD = torch.tensor([0.229, 0.224, 0.225]) * 255

train_pipeline = [
    RandomResizedCropRGBImageDecoder((224, 224)),
    RandomHorizontalFlip(),
    ToTensor(),
    ToDevice('cuda:0', non_blocking=True),

kirill-fedyanin

const model = tf.sequential();
model.add(tf.layers.dense({ units: 5, activation: 'sigmoid', inputShape: [1]}));
model.add(tf.layers.dense({ units: 1, activation: 'sigmoid'}));
model.compile({loss: 'meanSquaredError', optimizer: 'adadelta'});
const xs = tf.tensor1d([1, 2, 3, 4, 5, 6, 7, 8, 9]);
const ys = tf.tensor1d([0, 0, 0, 0, 0, 1, 1, 1, 1]);
model.fit(xs, ys, {epochs: 1000}).then(h => {
   console.log("Loss after Epoch " + " : " + h.history.loss[0]);
   model.predict(xs).print();
});

kirill-fedyanin

import keras
import numpy as np

model = keras.Sequential()
model.add(keras.layers.Dense(5, activation='sigmoid', input_shape=(1,)))
model.add(keras.layers.Dense(1, activation='sigmoid'))
model.compile('adadelta', 'mse', metrics=['accuracy'])

xs = np.array(range(1, 10))
ys = np.array([[0] for _ in range(5)] + [[1] for _ in range(4)])

kirill-fedyanin

upstream puma {                                                                 
  server unix:///home/ct_cyrill/apps/sendit/tmp/sockets/puma.sock;              
}                                                                               
                                                                                
server {                                                                        
  listen 80 default_server deferred;                                            
  server_name localhost;                                        
                                                                                
  root /home/ct_cyrill/apps/sendit/public;                                      
  access_log /home/ct_cyrill/apps/sendit/log/nginx.access.log;