1
0
mirror of https://github.com/osmarks/nanogpt-experiments.git synced 2024-09-21 03:39:44 +00:00
nanogpt-experiments/find_unused_tokens.py
2024-07-08 19:36:49 +01:00

10 lines
248 B
Python

import numpy as np
import os
data_dir = "."
data = np.memmap(os.path.join(data_dir, 'train.bin'), dtype=np.uint16, mode='r')
datas = set(data)
vocab = set(range(50257))
unused = vocab - datas
unused = sorted(unused)
print(len(unused))
print(unused)