rachmanino/SelfExtended
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: rachmanino/SelfExtended
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-08-03 19:17:18 +08:00
commit 4b8df0e085
11 changed files with 93841 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
README.md Normal file
View File

@@ -0,0 +1,29 @@
---
license: mit
language:
- en
---
This model is based on pretrained model meta/Llama-2-7b-chat-bf and has been equipped with SelfExtend trick
(implemented with flash attention) to achieve better performance on long context tasks.
This model is based on pretrained model meta/Llama-2-7b-chat-bf and has been equipped with SelfExtend trick
(implemented with flash attention) to achieve better performance on long context tasks.
This model is based on pretrained model meta/Llama-2-7b-chat-bf and has been equipped with SelfExtend trick
(implemented with flash attention) to achieve better performance on long context tasks.
This model is based on pretrained model meta/Llama-2-7b-chat-bf and has been equipped with SelfExtend trick
(implemented with flash attention) to achieve better performance on long context tasks.
This model is based on pretrained model meta/Llama-2-7b-chat-bf and has been equipped with SelfExtend trick
(implemented with flash attention) to achieve better performance on long context tasks.
This model is based on pretrained model meta/Llama-2-7b-chat-bf and has been equipped with SelfExtend trick
(implemented with flash attention) to achieve better performance on long context tasks.
This model is based on pretrained model meta/Llama-2-7b-chat-bf and has been equipped with SelfExtend trick
(implemented with flash attention) to achieve better performance on long context tasks.
This model is based on pretrained model meta/Llama-2-7b-chat-bf and has been equipped with SelfExtend trick
(implemented with flash attention) to achieve better performance on long context tasks.