yzhangcs

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

import torch
import triton
import triton.language as tl


@triton.jit
def cumsum_matmul_kernel(
    s,

yzhangcs

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

import torch
import torch.nn as nn
import triton
import triton.language as tl
from transformers.models.llama.modeling_llama import LlamaRMSNorm


@triton.jit

yzhangcs

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

from typing import Optional, Tuple

import torch
import torch.nn as nn
import triton
import triton.language as tl
from transformers.models.llama.modeling_llama import (LlamaAttention,
                                                      apply_rotary_pos_emb,

yzhangcs

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

from __future__ import annotations

import copy
from typing import Optional

import torch
import torch.nn as nn
import torch.nn.functional as F

yzhangcs

# Terminal color definitions

class fg:
    BLACK   = '\033[30m'
    RED     = '\033[31m'
    GREEN   = '\033[32m'
    YELLOW  = '\033[33m'
    BLUE    = '\033[34m'
    MAGENTA = '\033[35m'
    CYAN    = '\033[36m'

yzhangcs

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

import re
from datetime import datetime, timedelta

import arxiv


KEYS = ['adversarial', 'algebraic', 'algebratic', 'amr', 'auto-encoding', 'autoencoder', 'autoencoding', 'autoregressive',
        'backward', 'bayes', 'bayesian', 'bethe', 'bilexical', 'bipartite', 'bregman', 'carlo', 'chomsky', 'circuit', 'clique',

yzhangcs

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

from functools import reduce

import torch
import torch.autograd as autograd
from supar.utils.common import MIN
from torch.autograd import Function

yzhangcs

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

import argparse
from functools import partial
from multiprocessing.dummy import Pool

import nltk
from google_trans_new import google_translator
from tqdm import tqdm

yzhangcs

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

import argparse
from functools import partial
from multiprocessing.dummy import Pool

import nltk
from google_trans_new import google_translator
from tqdm import tqdm

yzhangcs

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

import argparse
from collections import Counter


def factorize(tags):
    spans = []
    for i, tag in enumerate(tags):
        if tag.startswith(('B', 'S')):