Bjorno/Illustrious_NoobAI_Archive
Updated
M. S.
Bjorno
AI & ML interests
None yet
Recent Activity
updated a model 1 day ago
Bjorno/Illustrious_NoobAI_Archive updated a model 1 day ago
Bjorno/SDXL_Archive updated a model 1 day ago
Bjorno/SD_15_ArchiveOrganizations
None yet
updated 3
models 1 day ago
updated a
model 2 days ago
published a
model 2 days ago
published a
model about 1 month ago
published a
model about 2 months ago
updated a
model about 2 months ago
published 2
models 3 months ago
reacted to Lewdiculous's post with 🚀 over 2 years ago
Post
72,184
More context for your Pascal GPU or older!
Update: Now available in the official releases of KoboldCpp!
[releases] https://github.com/LostRuins/koboldcpp/releases/latest
These are great news for all the users with GTX 10XX, P40...
Flash Attention implementation for older NVIDIA GPUs without requiring Tensor Cores has come to llama.cpp in the last few days, and should be merged in the next version of KoboldCpp, you can already try it with another fork or by building it.
[Mentioned KCPP fork] https://github.com/Nexesenex/kobold.cpp/releases/latest
[PR] https://github.com/ggerganov/llama.cpp/pull/7188
You should expect less VRAM usage for the same context, allowing you to experience higher contexts with your current GPU.
There have also been reported final tokens/second speed improvements for inference, so that's also grand!
If you have tried it, I'd like to hear your experiences with --flashattention so far, especially for this implementation and for the large number of Pascal (GTX 10XX, P40...) cards.
Discussion linked bellow, with more links to relevant information:
https://huggingface.co/LWDCLS/LLM-Discussions/discussions/11
Cheers!
Update: Now available in the official releases of KoboldCpp!
[releases] https://github.com/LostRuins/koboldcpp/releases/latest
These are great news for all the users with GTX 10XX, P40...
Flash Attention implementation for older NVIDIA GPUs without requiring Tensor Cores has come to llama.cpp in the last few days, and should be merged in the next version of KoboldCpp, you can already try it with another fork or by building it.
[Mentioned KCPP fork] https://github.com/Nexesenex/kobold.cpp/releases/latest
[PR] https://github.com/ggerganov/llama.cpp/pull/7188
You should expect less VRAM usage for the same context, allowing you to experience higher contexts with your current GPU.
There have also been reported final tokens/second speed improvements for inference, so that's also grand!
If you have tried it, I'd like to hear your experiences with --flashattention so far, especially for this implementation and for the large number of Pascal (GTX 10XX, P40...) cards.
Discussion linked bellow, with more links to relevant information:
https://huggingface.co/LWDCLS/LLM-Discussions/discussions/11
Cheers!
- 24 replies
...is it broken?
2
#1 opened over 2 years ago
by 
lisabdunlap
Weirdness with offloading to VRAM
3
#3 opened over 2 years ago
by Bjorno
Very long responses
4
#3 opened over 2 years ago
by Bjorno
Help. Getting Gibberish Responses?
6
#1 opened over 2 years ago
by deleted
What's the difference?
3
#2 opened over 2 years ago
by 
Clevyby