mirror of
https://github.com/modelscope/modelscope.git
synced 2025-12-25 12:39:25 +01:00
69104c0f8ae88a242b4669d152377cac04d2c274
1. 将 single_gpu_test 与 multi_gpu_test 中的 model.forward 部分分离为 EpochBasedTrainer 中的 evaluation_step,为部分 evaluation 阶段不调用 forward 的模型提供更好的灵活性
2. 重构代码将文本生成模型 Model 层的输入输出统一为 Tensor,Tensor 到 str 的 decode 过程移动到 pipeline 中完成
3. pipeline 后处理添加对中文和中文标点与英文混杂时空格的处理,使 decode 后中英文混杂输出正确
4. 添加 TextGenerationTrainer 修复了部分模型 evaluation 过程 forward 输出单个 token 计算 metrics 的问题
5. 修复了 rouge 无法接收空字符串的问题
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10473768
* [to #45486649]feat: modelscope model version use model repo tag, unsupport branch or commit it, client user-agent header unified
Introduction
ModelScope library is targeted to support training, evaluation and inference for the state of the art models provided by Mind and further support third-party models provided by users outside alibaba.
Design doc
Please refer to alidoc link
Development doc
Please refer to develop.md
ChangeLog
- 20/05/2022 First release version
Refer to change_log.md for more details
Description
Languages
Python
98.7%
Cuda
0.8%
C++
0.4%