hppRC

{"lastUpload":"2020-04-04T07:45:57.757Z","extensionVersion":"v3.4.3"}

hppRC

Python      3 hrs 45 mins  ██████████▌░░░░░░░░░░  50.0%
Markdown    2 hrs 38 mins  ███████▍░░░░░░░░░░░░░  35.1%
TypeScript  37 mins        █▋░░░░░░░░░░░░░░░░░░░   8.2%
JSON        19 mins        ▉░░░░░░░░░░░░░░░░░░░░   4.2%
JavaScript  5 mins         ▎░░░░░░░░░░░░░░░░░░░░   1.3%

hppRC

N_TRIALS=10
STORAGE=sqlite:///example.db

STUDY_NAME=`optuna create-study --storage $STORAGE`
DISTRIBUTIONS=`cat distributions.json`

for _ in `seq 1 $N_TRIALS`; do
    trial=`optuna ask \
        --storage $STORAGE \
        --study-name $STUDY_NAME \

hppRC

from tqdm import tqdm

for i in tqdm(list(range(100)), position=0):
    for batch in tqdm(list(range(10000000)), position=1):
        pass

hppRC

import hoge.a

print(__file__)

hppRC

ORIGIN_DIR=$(pwd)

JUMANPP_DIR="$HOME/.local/share/jumanpp"

mkdir -p $JUMANPP_DIR
cd $JUMANPP_DIR

curl -LO https://github.com/ku-nlp/jumanpp/releases/download/v2.0.0-rc3/jumanpp-2.0.0-rc3.tar.xz
tar -xf jumanpp-2.0.0-rc3.tar.xz
cd jumanpp-2.0.0-rc3

hppRC

def format_param_with_unit(num_params: int) -> str:
    if num_params >= 1000 * 1000 * 1000:
        unit = "B"
        num_params /= 1000 * 1000 * 1000
    elif num_params >= 1000 * 1000:
        unit = "M"
        num_params /= 1000 * 1000
    elif num_params >= 1000:
        unit = "K"
        num_params /= 1000

hppRC

from transformers import PreTrainedTokenizer
from vllm import LLM, SamplingParams
from vllm.outputs import RequestOutput

import datasets as ds


def build_input_text(text: str, tokenizer: PreTrainedTokenizer) -> str:
    text = text.strip()

hppRC

import datasets as ds
from konoha import SentenceTokenizer


def title2text():
    dataset: ds.Dataset = ds.load_dataset("globis-university/aozorabunko-clean", split="train", num_proc=16)

    def process(x: dict[str, list]):
        anc_list, pos_list = [], []

hppRC

import random
import uuid
from pathlib import Path

import click
from vllm import LLM, SamplingParams
from vllm.outputs import RequestOutput

import datasets as ds
from src.data.common import normalize_text