ypeleg

import random
import numpy as np
import pandas as pd
from collections import Counter, defaultdict


def stratified_group_k_fold(X, y, groups, k, seed=None):
    labels_num = np.max(y) + 1
    y_counts_per_group = defaultdict(lambda: np.zeros(labels_num))
    y_distr = Counter()

ypeleg

import tensorflow as tf
from tensorflow.keras.layers import Activation
from tensorflow.keras.utils import get_custom_objects

class Mish(Activation):
    '''
    Mish Activation Function.
    .. math::
        mish(x) = x * tanh(softplus(x)) = x * tanh(ln(1 + e^{x}))
    Shape:

ypeleg

def tf_graph_wrapper(func):
    """Wraps a class method with a tf.Graph context manager"""
    @wraps(func)
    def wrapper(self, *args, **kwargs):
        with self._graph.as_default():
            return func(self, *args, **kwargs)
    return wrapper


def tf_init(func):

ypeleg

# -*- coding: utf-8 -*-

import os
import sys
import keras
import shutil
import numpy as np
import pandas as pd
import keras.backend as K
from keras.models import Model

ypeleg

import os
import math
import pickle
import shutil
import joblib
import warnings
import datetime
import numpy as np

ypeleg

import tensorflow as tf
from copy import deepcopy
import tensorflow.keras.backend as K
from tensorflow.keras.optimizers import Adam
from tensorflow.keras.models import Model, clone_model
from tensorflow.keras.layers import Input, Dropout, Dense, ReLU, BatchNormalization, Activation, Concatenate

class DynamicNet(object):
    def __init__(self, c0 = None, lr = None, concat_input=False, additive_boosting=False, encoder_layers=None):

ypeleg

# Cleaned and minimal perceiver transformer, originally from code  https://github.com/Rishit-dagli/Perceiver
# Original paper: Perceiver: General Perception with Iterative Attention. Jaegle et al.  https://arxiv.org/pdf/2103.03206.pdf.

import math
import tensorflow as tf
from typing import Callable
from einops import rearrange, repeat

ypeleg

import json
import pickle
import struct
import zipfile
import numpy as np
from sentencepiece import SentencePieceProcessor

def rms_norm(x): return (x / np.sqrt(np.square(x).mean(-1, keepdims=True) + 1e-6))
def softmax(x): return (np.exp(x - np.max(x, axis=-1, keepdims=True))) / np.sum((np.exp(x - np.max(x, axis=-1, keepdims=True))), axis=-1, keepdims = True)

ypeleg

import llama
import torch
import pandas as pd
from torch.utils.data import Dataset, random_split
from transformers import TrainingArguments, Trainer

MODEL = 'decapoda-research/llama-7b-hf'
DATA_FILE_PATH = 'elon_musk_tweets.csv'

ypeleg

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('NousResearch/Nous-Hermes-Llama2-13b', device_map = 'auto')
tokenizer = AutoTokenizer.from_pretrained('NousResearch/Nous-Hermes-Llama2-13b')
model.eval()

print(tokenizer('yes')) # [1, 4871]
print(tokenizer.decode(4874)) # yes