mirror of
https://github.com/LostRuins/koboldcpp.git
synced 2025-09-16 03:49:42 +00:00
GradientAI Auto ROPE Base calculation (#910)
* GradientAI Auto ROPE Base calculation https://gradient.ai/blog/scaling-rotational-embeddings-for-long-context-language-models has a formula that better fits the ideal rope scaling. Tested with Lllama3, checked calculation is correct for llama2. Retains logic for not scaling rope if under trained CTX. * add in solar scaling logic Solar based models require the context values to be multiplied by 8. This is (i'm guessing) because the positions as based on a 32k context, but sliding window of 4k. * Update model_adapter.h adding in tensor count to identify solar models based on tensor count of 435. * Update model_adapter.cpp add in n_tensor count for solar identification * refactor and cleanup GradientAI rope scaling --------- Co-authored-by: Concedo <39025047+LostRuins@users.noreply.github.com>
This commit is contained in:
parent
49e4c3fd7b
commit
1e72b65c38
3 changed files with 39 additions and 22 deletions
|
@ -56,6 +56,7 @@ enum GGUFArch
|
|||
ARCH_FALCON = 1,
|
||||
ARCH_PHI = 2,
|
||||
ARCH_MAMBA = 3,
|
||||
ARCH_SOLAR = 4,
|
||||
};
|
||||
|
||||
struct FileFormatExtraMeta
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue