 Frank Mai
|
84eb2f4fad
docs: introduce gpustack and gguf-parser (#8873)
|
пре 1 година |
 DavidKorczynski
|
1262e7ed13
grammar-parser : fix possible null-deref (#9004)
|
пре 1 година |
|
DavidKorczynski
|
df5478fbea
ggml: fix div-by-zero (#9003)
|
пре 1 година |
 Liu Jia
|
2589292cde
Fix a spelling mistake (#9001)
|
пре 1 година |
 Georgi Gerganov
|
d3ae0ee8d7
py : fix requirements check '==' -> '~=' (#8982)
|
пре 1 година |
|
Georgi Gerganov
|
5ef07e25ac
server : handle models with missing EOS token (#8997)
|
пре 1 година |
 compilade
|
4134999e01
gguf-py : Numpy dequantization for most types (#8939)
|
пре 1 година |
|
Georgi Gerganov
|
8cd1bcfd3f
flake.lock: Update (#8979)
|
пре 1 година |
 Neo Zhang
|
a21c6fd450
update guide (#8909)
|
пре 1 година |
 fairydreaming
|
33309f661a
llama : check all graph nodes when searching for result_embd_pooled (#8956)
|
пре 1 година |
 Markus Tavenrath
|
7c5bfd57f8
Optimize Vulkan backend for better CPU performance and less GPU synchronization overhead. (#8943)
|
пре 1 година |
 slaren
|
6e02327e8b
metal : fix uninitialized abort_callback (#8968)
|
пре 1 година |
 Xuan Son Nguyen
|
7eb23840ed
llama : default n_swa for phi-3 (#8931)
|
пре 1 година |
|
fairydreaming
|
7c3f55c100
Add support for encoder-only T5 models (#8900)
|
пре 1 година |
 Matteo Mortari
|
911b437f22
gguf-py : fix double call to add_architecture() (#8952)
|
пре 1 година |
|
Georgi Gerganov
|
b72942fac9
Merge commit from fork
|
пре 1 година |
|
fairydreaming
|
6afd1a99dc
llama : add support for lora adapters in T5 model (#8938)
|
пре 1 година |
|
Georgi Gerganov
|
272e3bd95e
make : fix llava obj file race (#8946)
|
пре 1 година |
|
Georgi Gerganov
|
45a55b91aa
llama : better replace_all (cont) (#8926)
|
пре 1 година |
 tc-mb
|
3071c0a5f2
llava : support MiniCPM-V-2.5 (#7599)
|
пре 1 година |
|
Georgi Gerganov
|
4305b57c80
sync : ggml
|
пре 1 година |
 Matt Stephenson
|
70c0ea3560
whisper : use vulkan as gpu backend when available (whisper/2302)
|
пре 1 година |
 Daniel Bevenius
|
5b2c04f492
embedding : add --pooling option to README.md [no ci] (#8934)
|
пре 1 година |
|
Daniel Bevenius
|
6f6496bb09
llama : fix typo in llama_tensor_get_type comment [no ci] (#8937)
|
пре 1 година |
 Mathieu Geli
|
daef3ab233
server : add one level list nesting for embeddings (#8936)
|
пре 1 година |
|
compilade
|
345a686d82
llama : reduce useless copies when saving session (#8916)
|
пре 1 година |
|
compilade
|
3a14e00366
gguf-py : simplify support for quant types (#8838)
|
пре 1 година |
|
Georgi Gerganov
|
afd27f01fe
scripts : sync cann files (#0)
|
пре 1 година |
|
Georgi Gerganov
|
366d486c16
scripts : fix sync filenames (#0)
|
пре 1 година |
|
Georgi Gerganov
|
e44a561ab0
sync : ggml
|
пре 1 година |
