 Lengyue
|
89474bbb3b
support s1 model structure
|
10 月之前 |
 Whale and Dolphin
|
7902e408c8
[feature]add dataset classs (#775)
|
1 年之前 |
|
Whale and Dolphin
|
40665e1a39
[fix]fix problems to let version 1.5 support sft (#774)
|
1 年之前 |
|
Lengyue
|
dad516d86d
update checkpoint path
|
1 年之前 |
|
Lengyue
|
1d942c8b97
Update 1.2 sft checkpoints, inference.ipynb and READMEs
|
1 年之前 |
|
Lengyue
|
ab61b60f42
optimize config & improve stability
|
1 年之前 |
|
Lengyue
|
c9209e94d9
support finetuning
|
1 年之前 |
 spicysama
|
70de7876f6
优化模型/配置文件选择时的体验 (#228)
|
1 年之前 |
|
spicysama
|
3863865241
V1.1 package (#177)
|
1 年之前 |
|
Lengyue
|
e787d0182f
Use SFT medium as base model
|
1 年之前 |
|
Lengyue
|
fadbdae513
support smart warmup (float)
|
1 年之前 |
|
Lengyue
|
519b7c0b6e
Fix document checkpoint
|
1 年之前 |
|
Lengyue
|
467b7beb56
Support lora tuning & update document
|
1 年之前 |
|
Lengyue
|
ad7187ec11
Update documents
|
1 年之前 |
|
Lengyue
|
1d4b7256b3
Optimize pretrain & finetune receipe, apply better logger
|
1 年之前 |
|
Lengyue
|
ea101fcd0d
Optimize config system & enable faster loading
|
2 年之前 |
|
Lengyue
|
0a0778191c
Update configs
|
2 年之前 |
|
Lengyue
|
5707699dfd
Handle adaptive number of codebooks
|
2 年之前 |
|
Lengyue
|
5f9e2f2380
Update new config and generation scripts
|
2 年之前 |
|
Lengyue
|
39f6902119
Add dropout options to optimize overfitting
|
2 年之前 |
|
Lengyue
|
a86e61a7cd
Update pretrain config
|
2 年之前 |
|
Lengyue
|
c54eecc63f
Implement windowed repeat penalty
|
2 年之前 |
|
Lengyue
|
f5919b7e71
Update finetune toolchain
|
2 年之前 |
|
Lengyue
|
49f84853c4
remove unused configs
|
2 年之前 |
