hummingbird
/
fastNLP
Not watched
2
0
Fork 0
Code
Releases 13 Wiki Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain
You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
2196 Commits
3 Branches
25 MB
1 Download
 
 
 
 
 
Tree: d4bccf3f6a
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from 'd4bccf3f6a'
${ noResults }
fastNLP/tests/core/controllers
yh d4bccf3f6a 修复Evaluator的evaluate_use_dist_sampler在Trainer中的bug 3 years ago
..
utils Merge branch 'dev0.8.0' of github.com:fastnlp/fastNLP into dev0.8.0 3 years ago
__init__.py 更新 callbacks, controllers, drivers, samplers 以及 tests 中的部分内容 3 years ago
_test_distributed_launch_torch_1.py 1.修改torch测试例中多卡的driver参数 2.修改测试例中的driver.save driver.load为driver.save_checkpoint driver.load_checkpoint 3.添加lstm 3 years ago
_test_distributed_launch_torch_2.py 1.修改torch测试例中多卡的driver参数 2.修改测试例中的driver.save driver.load为driver.save_checkpoint driver.load_checkpoint 3.添加lstm 3 years ago
_test_trainer_fleet.py 1. 支持在不设置backend的情况下运行单卡的paddle程序 2.当通过launch启动且限制显卡时的paddle多卡逻辑 3 years ago
_test_trainer_fleet_outside.py 1. 继续完善部分文档；2.删除 paddle 多卡下的 data_device 功能 3. 将 paddle_utils 下的 get_device_from_visible 函数更名为 _convert_data_device 并进行修改 3 years ago
_test_trainer_jittor.py 1. 支持在不设置backend的情况下运行单卡的paddle程序 2.当通过launch启动且限制显卡时的paddle多卡逻辑 3 years ago
imdb.py 1. 支持在不设置backend的情况下运行单卡的paddle程序 2.当通过launch启动且限制显卡时的paddle多卡逻辑 3 years ago
test_evaluator_torch.py 修复Evaluator的evaluate_use_dist_sampler在Trainer中的bug 3 years ago
test_trainer.py Trainer.on支持*args等形式的输入 3 years ago
test_trainer_event_trigger.py merge 3 years ago
test_trainer_jittor.py 为 test_trainer_jittor 添加 DummyClass 和 pytest.mark.jittor 3 years ago
test_trainer_other_things.py 修改测试例中的Events为Event 3 years ago
test_trainer_paddle.py 1. 支持在不设置backend的情况下运行单卡的paddle程序 2.当通过launch启动且限制显卡时的paddle多卡逻辑 3 years ago
test_trainer_w_evaluator_torch.py 修改了 trainer 和 evaluator 的 torch_kwargs 3 years ago
test_trainer_wo_evaluator_torch.py paddle 和测试例跟进 set_dist_repro_dataloader 函数；修改test_trainer_wo_evaluator_torch.py的bug 3 years ago
Delete task

Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.

cancel
confirm

一款轻量级的自然语言处理（NLP）工具包，目标是减少用户项目中的工程型代码，例如数据处理循环、训练循环、多卡运行等

Python Jupyter Notebook Text CSV Markdown