 Jed Fox
|
3924088512
Remove default arguments from sampling functions (#1343)
|
há 2 anos atrás |
 Evan Jones
|
e216aa0463
llama : only copy used KV cache in get / set state (#1272)
|
há 2 anos atrás |
 Georgi Gerganov
|
0e6cbff1b7
llama : fix compile warnings
|
há 2 anos atrás |
 Robert Brisita
|
2bb992f034
llama : allow 0 as a seed number. (#1275)
|
há 2 anos atrás |
|
Georgi Gerganov
|
70269cae37
llama : fix session load / save (#1263)
|
há 2 anos atrás |
 Alex Klinkhamer
|
90b19bd6ee
llama : let context be const when accessing const data (#1261)
|
há 2 anos atrás |
 Ivan Stepanov
|
dd7eff57d8
llama : new sampling algorithms (#1126)
|
há 2 anos atrás |
 Stephan Walter
|
36d19a603b
Remove Q4_3 which is no better than Q5 (#1218)
|
há 2 anos atrás |
|
Evan Jones
|
1481a9cf25
llama : add session file format and saved sessions in main (#1169)
|
há 2 anos atrás |
|
Georgi Gerganov
|
574406dc7e
ggml : add Q5_0 and Q5_1 quantization (#1187)
|
há 2 anos atrás |
 Ásgeir Bjarni Ingvarsson
|
87a6f846d3
Allow setting the rng seed after initialization. (#1184)
|
há 2 anos atrás |
|
Georgi Gerganov
|
7a32fcb3b2
ggml : add Q8_0 quantization format (rename the old one to Q8_1) (ARM NEON) (#1179)
|
há 2 anos atrás |
|
Georgi Gerganov
|
c4fe84fb0d
llama : refactor get / set state + remove redundant kv cache API (#1143)
|
há 2 anos atrás |
 xaedes
|
b6e7f9b09e
llama : add api for getting/setting the complete state: rng, logits, embedding and kv_cache (#1105)
|
há 2 anos atrás |
 Kawrakow
|
38de86a711
llama : multi-threaded quantization (#1075)
|
há 2 anos atrás |
|
Georgi Gerganov
|
e0305ead3a
ggml : add Q4_3 quantization (#1082)
|
há 2 anos atrás |
|
Georgi Gerganov
|
77a73403ca
ggml : add new Q4_2 quantization (ARM only) (#1046)
|
há 2 anos atrás |
 slaren
|
315a95a4d3
Add LoRA support (#820)
|
há 2 anos atrás |
|
Georgi Gerganov
|
9190e8eac8
llama : merge llama_internal.h into llama.h
|
há 2 anos atrás |
|
Stephan Walter
|
e7f6997f89
Don't crash on ftype (formerly f16) == 4 (#917)
|
há 2 anos atrás |
|
Stephan Walter
|
3e6e70d8e8
Add enum llama_ftype, sync ggml_type to model files (#709)
|
há 2 anos atrás |
 comex
|
f963b63afa
Rewrite loading code to try to satisfy everyone:
|
há 2 anos atrás |
 unbounded
|
62cfc54f77
Add quantize-stats command for testing quantization (#728)
|
há 2 anos atrás |
 Christian Falch
|
e986f94829
Added api for getting/setting the kv_cache (#685)
|
há 2 anos atrás |
 Justine Tunney
|
78ca9838ee
Make loading weights 10-100x faster
|
há 2 anos atrás |
 anzz1
|
a5c42c4b13
Fix typo in llama.h (#593)
|
há 2 anos atrás |
|
anzz1
|
7f4c5c6651
llama : fix linkage with mingw (#551)
|
há 2 anos atrás |
|
Stephan Walter
|
436e561931
all : be more strict about converting float to double (#458)
|
há 2 anos atrás |
|
Stephan Walter
|
c1f885067c
ggml : introduce structs for the q4 data blocks (#356)
|
há 2 anos atrás |
|
Georgi Gerganov
|
03f7e33560
Cleanup STL headers + fix embedding examples + minor stuff
|
há 2 anos atrás |
