minesh1291

Setup

Install TigerVNC -

$ sudo apt install tigervnc-standalone-server

Configure VNC Server by running

minesh1291

#!/bin/bash

# Check if the script is run with root privileges
if [ "$EUID" -ne 0 ]; then
  echo "Please run as root"
  exit
fi

# Determine the log file based on the distribution
if [ -f /var/log/auth.log ]; then

minesh1291

# train_grpo.py
import re
import torch
from datasets import load_dataset, Dataset
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import LoraConfig
from trl import GRPOConfig, GRPOTrainer

# Load and prep dataset

minesh1291

import numpy as np
from sklearn.linear_model import SGDClassifier
from sklearn.datasets import make_classification
from sklearn.metrics import log_loss
import matplotlib.pyplot as plt

np.random.seed(42)
# Create synthetic classification dataset
X, y = make_classification(
    n_samples=500, n_features=50, n_informative=15, random_state=42

minesh1291

import pandas as pd
import numpy as np

# Example DataFrame with random target values
df = pd.DataFrame({
    'label': np.random.normal(size=1000)  # 100 random values between 0 and 1
})

# Step 1: Bin the target values to create a frequency distribution
df['label_bin'] = pd.cut(df['label'], bins=10)

minesh1291

minesh1291

"""
This model integrates the MoE concept within a Transformer architecture. Each token's
representation is processed by a subset of experts, determined by the gating mechanism.
This architecture allows for efficient and specialized handling of different aspects of the
data, aiming for the adaptability and efficiency noted in the Mixtral 8x7B model's design
philosophy. The model activates only a fraction of the available experts for each token,
significantly reducing the computational resources needed compared to activating all experts
for all tokens.
"""

minesh1291

# %%
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import scipy.signal
from gluonts.dataset.repository import get_dataset, dataset_names
from gluonts.dataset.util import to_pandas
from gluonts.dataset.common import ListDataset
from gluonts.torch import SimpleFeedForwardEstimator
from lightning.pytorch.callbacks.early_stopping import EarlyStopping

minesh1291

# Works on gluonts dev branch as of May 30th, 2023
# Assumes "m5-forecasting-accuracy" folder with data next to the script
# Data is obtained from https://www.kaggle.com/c/m5-forecasting-accuracy

import pandas as pd
from pathlib import Path
from gluonts.dataset.pandas import PandasDataset

# Load data from csv files

minesh1291

from os import path
from bokeh.models import Button, Div
from bokeh.layouts import column
from bokeh.document import without_document_lock
from bokeh.io import curdoc
from zmq_subprocess import ZmqSubProcessClient


ok_button = Button(label="ok")
div = Div()