refactor: 从data 模块分离tokenizer

This commit is contained in:
2026-04-04 16:12:58 +08:00
parent b531232a9b
commit bd9741dc5f
16 changed files with 108 additions and 92 deletions
+1 -1
View File
@@ -5,7 +5,7 @@ import torch.distributed as dist
from torch.optim import AdamW
from torch.optim.lr_scheduler import CosineAnnealingLR
from astrai.data.serialization import Checkpoint
from astrai.serialization import Checkpoint
from astrai.parallel.setup import get_rank, spawn_parallel_fn
+1 -1
View File
@@ -2,7 +2,7 @@ import numpy as np
import torch
from astrai.data.dataset import DatasetFactory
from astrai.data.serialization import save_h5
from astrai.serialization import save_h5
def test_dataset_loader_random_paths(base_test_env):