BoFan

1、dataset_info.json文件核对后发现现在的和原项目的一致，但是我把因为将原文件内容只保留5分之1因此，将 "file_sha1": "647f4ad447bd993e4b6b6223d1be15208bab694a"删了内容如下，应该没有问题吧？ "covid_train": { "file_name": "train.json", "columns": { "prompt": "instruction", "query": "input", "response": "output", "history": "history" } }, 2、训练开始前的部分提示，是否正常？ 06/26/2023 18:29:48 - INFO - utils.common - Loading dataset train.json......

detection 我是用自己数据训练 就是那个官方balloon数据，典型小样本，效果太差

4

optimizer = dict(type='SGD', lr=0.05, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=None) # learning policy lr_config = dict( policy='step', warmup='linear', warmup_iters=500, warmup_ratio=0.001, step=[8, 11]) runner = dict(type='EpochBasedRunner', max_epochs=20)

how to search

8

when I asked weather，Beijing is ok，but other is not running  

why don't use tortoise ORM ？

1

This project used ORM is Peewee。 Peewee ORM is a simple and lightweight ORM, but it does not support asynchronous operations. Can you change to tortoise ORM？