 Georgi Gerganov
|
c642bc014c
kv-cache : separate recurrent vs non-recurrent impl (#12799)
|
9 月之前 |
 Xuan-Son Nguyen
|
d2b2031e5f
llama : (mrope) allow using normal 1D position for text token (#13138)
|
9 月之前 |
 Juk Armstrong
|
daa422881a
llama : DeepSeek V2/V3 MLA implementation (#12801)
|
9 月之前 |
|
Xuan-Son Nguyen
|
1466621e73
llama : Support llama 4 text-only (#12791)
|
10 月之前 |
|
Georgi Gerganov
|
75422e8bc4
graph : normalize Q, K, V shapes + sync cross attention (#12449)
|
10 月之前 |
|
Georgi Gerganov
|
c522ce4143
graph : simplify attn input build for unified KV cache (#12381)
|
10 月之前 |
|
Georgi Gerganov
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
10 月之前 |
