 ddh0
|
f6dcda3900
server : context checkpointing for hybrid and recurrent models (#16382)
|
3 months ago |
 Johannes Gäßler
|
e789095502
llama: print memory breakdown on exit (#15860)
|
3 months ago |
 Georgi Gerganov
|
b730706a49
kv-cache : support layer reuse (#15504)
|
4 months ago |
|
Georgi Gerganov
|
d32e03f449
server : add SWA checkpoints (#15293)
|
5 months ago |
 l3utterfly
|
7233358d29
memory : handle saving/loading null layers in recurrent memory (#14675)
|
6 months ago |
|
Georgi Gerganov
|
01612b7409
llama : reuse compute graphs (#14482)
|
6 months ago |
 compilade
|
4a5686da22
llama : support Jamba hybrid Transformer-Mamba models (#7531)
|
6 months ago |
|
compilade
|
bb4f7a9e4e
memory : fix broken batch splits for recurrent cache (#14575)
|
6 months ago |
|
Georgi Gerganov
|
67d1ef23c6
batch : add optional for sequential equal split (#14511)
|
6 months ago |
|
Georgi Gerganov
|
c79184d2d1
batch : add n_used count (#14512)
|
6 months ago |
|
Georgi Gerganov
|
745f11fed0
memory : correctly handle failure in apply() (#14438)
|
6 months ago |
|
Georgi Gerganov
|
43678060c1
recurrent : call balloc split_reset() in init_batch() (#14414)
|
6 months ago |
|
Georgi Gerganov
|
692e3cdd0a
memory : rename interface to llama_memory_context_i (#14296)
|
7 months ago |
|
Georgi Gerganov
|
4c9fdfbe15
ubatch : new splitting logic (#14217)
|
7 months ago |
 Gabe Goodhart
|
edc4a29eff
memory : Hybrid recurrent cache (#13979)
|
7 months ago |
