BlinkDL
|
2388bca7c3
|
dummy data example
|
3 years ago |
BlinkDL
|
778c0a7f58
|
improvements
|
3 years ago |
BlinkDL
|
f81349f127
|
fix
|
3 years ago |
BlinkDL
|
6ab2e71c25
|
finetuning 1.5B model using 16G VRAM
|
3 years ago |
BlinkDL
|
23b0c74950
|
fix
|
3 years ago |
BlinkDL
|
cdb098c0e0
|
fix
|
3 years ago |
BlinkDL
|
8abea9c08d
|
wip
|
3 years ago |
BlinkDL
|
ba6e9e6264
|
rwkv-v4neo test
|
3 years ago |
BlinkDL
|
50587bd65f
|
fix for jit modules
|
3 years ago |
BlinkDL
|
2b4539cd08
|
faster (use torch 1.12.1+cu116 or newer)
|
3 years ago |
BlinkDL
|
2f33901c10
|
no message
|
3 years ago |
BlinkDL
|
6ff859db80
|
no message
|
3 years ago |
BlinkDL
|
09c76b185a
|
no message
|
3 years ago |
PENG Bo
|
c49fd38ba1
|
Update README.md
|
3 years ago |
PENG Bo
|
f3cf7a5ad1
|
Update README.md
|
3 years ago |
PENG Bo
|
5c29eb779c
|
Add files via upload
|
3 years ago |
BlinkDL
|
8cced0383e
|
no message
|
3 years ago |
PENG Bo
|
f46cf49852
|
Update README.md
|
3 years ago |
BlinkDL
|
2815260d83
|
better
|
3 years ago |
BlinkDL
|
dc7e0802d0
|
faster
|
3 years ago |
BlinkDL
|
c43a17cfb3
|
10% faster training
|
3 years ago |
BlinkDL
|
c84e8fd952
|
bugfix
|
3 years ago |
BlinkDL
|
73b96705d7
|
+ fp32 mode (slow but good for verification)
|
3 years ago |
BlinkDL
|
94f618c52a
|
Merge branch 'main' of https://github.com/BlinkDL/RWKV-LM
|
3 years ago |
BlinkDL
|
a1bf15ac40
|
no message
|
3 years ago |
PENG Bo
|
e05c69452d
|
Update README.md
|
3 years ago |
PENG Bo
|
5f1a473845
|
Update README.md
|
3 years ago |
BlinkDL
|
6299c087a4
|
fixed VRAM consumpition
|
3 years ago |
PENG Bo
|
cb520e0f15
|
Update README.md
|
3 years ago |
PENG Bo
|
a01d915fcc
|
Update README.md
|
3 years ago |
PENG Bo
|
4c2080aadd
|
Add files via upload
|
3 years ago |
BlinkDL
|
c1f7a72724
|
saves some VRAM for 1 GPU training
|
3 years ago |
PENG Bo
|
69e7cfbf39
|
Update README.md
|
3 years ago |
PENG Bo
|
f757518277
|
Update README.md
|
3 years ago |
PENG Bo
|
2bddd576cd
|
Update README.md
|
3 years ago |
PENG Bo
|
1949ed8619
|
Update README.md
|
3 years ago |
BlinkDL
|
68c486ad10
|
supports RWKV-4 pile models
|
3 years ago |
BlinkDL
|
61b7c429df
|
no message
|
3 years ago |
PENG Bo
|
8d4fed7128
|
Update README.md
|
3 years ago |
BlinkDL
|
7cdc8d3164
|
no message
|
3 years ago |
BlinkDL
|
13bb641007
|
no message
|
3 years ago |
BlinkDL
|
f79137b524
|
supports megatron bin+idx format
|
3 years ago |
PENG Bo
|
aa67870849
|
Update README.md
|
3 years ago |
BlinkDL
|
4c8a1a467b
|
no message
|
3 years ago |
BlinkDL
|
083f9504c6
|
+ bf16 mode (more stable)
|
3 years ago |
PENG Bo
|
46eebd98ca
|
Add files via upload
|
3 years ago |
PENG Bo
|
c0c4ffc7b4
|
Update README.md
|
3 years ago |
BlinkDL
|
6667ad18c2
|
more training tips
|
3 years ago |
PENG Bo
|
5f6e9356a2
|
Update README.md
|
3 years ago |
BlinkDL
|
165dfd1b9e
|
RWKV-4 with DeepSpeed & FP16 & Better CUDA Kernel
|
3 years ago |