-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
enrich(contrast learning): add SimCLR framework
- Loading branch information
Showing
14 changed files
with
180 additions
and
23 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,18 @@ | ||
--- | ||
layout: post | ||
title: Prefix Tunning | ||
date: 2023-11-19 | ||
tags: [prefix-tunning, peft] | ||
categories: | ||
- nlp | ||
--- | ||
|
||
- [ ] Prefix-tunning 是什么,解决什么问题? | ||
- [ ] 是如何解决这个问题的? | ||
- [ ] 有什么特点? | ||
- [ ] 适用场景? | ||
|
||
|
||
主流的 NLP 任务都是 pretraining + fine-tuning 的范式,即在预训练模型的基础上,针对特定任务进行微调。这种方法的优点是简单,但在当下模型越来越大的情况下,fine-tuning 的成本也越来越高。另外,fine-tuning 也有一些缺点,例如,模型的泛化能力不强,对于一些小数据集,模型的效果很差。针对这些问题,有一些研究者提出了一些方法,例如,[《Prefix-Tuning: Optimizing Continuous Prompts for Generation》](https://arxiv.org/pdf/2101.00190.pdf) 就是一种新的 fine-tuning 方法,它可以在不改变模型参数的情况下,通过修改输入的前缀来优化模型的效果。这种方法的优点是可以在不改变模型参数的情况下,优化模型的效果,而且可以在小数据集上取得很好的效果。 | ||
|
||
是一种轻量的 NLG 任务调参技术,解决 fine-tuning 方法在 LLM 上 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
--- | ||
layout: post | ||
title: Prefix Tunning | ||
date: 2023-11-19 | ||
tags: prefix-tunning peft | ||
categories: nlp | ||
author: berrysleaf | ||
--- | ||
* content | ||
{:toc} | ||
|
||
|
||
- [ ] Prefix-tunning 是什么,解决什么问题? | ||
|
||
|
||
|
||
- [ ] 是如何解决这个问题的? | ||
- [ ] 有什么特点? | ||
- [ ] 适用场景? | ||
|
||
|
||
主流的 NLP 任务都是 pretraining + fine-tuning 的范式,即在预训练模型的基础上,针对特定任务进行微调。这种方法的优点是简单,但在当下模型越来越大的情况下,fine-tuning 的成本也越来越高。另外,fine-tuning 也有一些缺点,例如,模型的泛化能力不强,对于一些小数据集,模型的效果很差。针对这些问题,有一些研究者提出了一些方法,例如,[《Prefix-Tuning: Optimizing Continuous Prompts for Generation》](https://arxiv.org/pdf/2101.00190.pdf) 就是一种新的 fine-tuning 方法,它可以在不改变模型参数的情况下,通过修改输入的前缀来优化模型的效果。这种方法的优点是可以在不改变模型参数的情况下,优化模型的效果,而且可以在小数据集上取得很好的效果。 | ||
|
||
是一种轻量的 NLG 任务调参技术,解决 fine-tuning 方法在 LLM 上 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -486,5 +486,13 @@ | |
[ | ||
"2023-11-18", | ||
738 | ||
], | ||
[ | ||
"2023-11-19", | ||
453 | ||
], | ||
[ | ||
"2023-11-20", | ||
292 | ||
] | ||
] |
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Oops, something went wrong.