 Josh Ramer
|
fed0108491
Scripting & documenting debugging one test without anything else in the loop. (#7096)
|
1 年之前 |
 Xuan Son Nguyen
|
72c177c1f6
fix system prompt handling (#7153)
|
1 年之前 |
 compilade
|
5a419926b0
convert-hf : support bfloat16 conversion (#7158)
|
1 年之前 |
 Georgi Gerganov
|
fae9d234b6
sync : ggml
|
1 年之前 |
 Justina Cho
|
f5ef34e428
feat: implemented sigmoid function (ggml/806)
|
1 年之前 |
 Borislav Stanimirov
|
ef0d5e3ec9
build: fix and ignore msvc warnings (ggml/805)
|
1 年之前 |
 CrispStrobe
|
3292733f95
convert : skip unaccessible HF repos (#7210)
|
1 年之前 |
 Steve Grubb
|
988631335a
server : free llama_batch on exit (#7212)
|
1 年之前 |
 Haoxiang Fei
|
f99e1e456e
llama : lookup word in vocab before doing BPE merges (#7193)
|
1 年之前 |
 Johannes Gäßler
|
5ae3426b0b
server: fix reported top tokens for temperature 0 (#7203)
|
1 年之前 |
 Joan Fontanals
|
b83cc3f5b3
llama : add Jina Embeddings architecture (#6826)
|
1 年之前 |
|
Georgi Gerganov
|
9cb317f77e
ggml : full ALiBi support (#7192)
|
1 年之前 |
 slaren
|
e849648888
llama-bench : add pp+tg test type (#7199)
|
1 年之前 |
|
Georgi Gerganov
|
18e437665c
metal : fix flash attention kernel requirements (#7169)
|
1 年之前 |
|
Georgi Gerganov
|
8c660242d7
convert : print "ignore_merges" field
|
1 年之前 |
|
slaren
|
25c6e82e7a
llama : use n_vocab to differentiate between mistral 7B and llama3 8B (#7200)
|
1 年之前 |
 Justine Tunney
|
4e3880978f
Fix memory bug in grammar parser (#7194)
|
1 年之前 |
 HanishKVC
|
f89fe2732c
Main+: optionally allow special tokens from user in interactive mode (#7097)
|
1 年之前 |
 Andrei
|
d11afd6652
llava : fix moondream support (#7163)
|
1 年之前 |
 Ouadie EL FAROUKI
|
8c570c9496
Minor arithmetic improvement to mmvq wrapper kernel (#7172)
|
1 年之前 |
|
slaren
|
eaf4bd8b39
eval-callback : fix conversion to float (#7184)
|
1 年之前 |
 0cc4m
|
befddd0f15
Vulkan Bugfixes and Improvements (#7084)
|
1 年之前 |
|
Georgi Gerganov
|
d46dbc76f8
readme : add scheduled server workflow status badge
|
1 年之前 |
 l3utterfly
|
0961d86604
readme : add app (#6371)
|
1 年之前 |
 jaime-m-p
|
43248e5594
llama3 custom regex split (#6965)
|
1 年之前 |
|
Johannes Gäßler
|
a743d76a01
CUDA: generalize FP16 fattn vec kernel (#7061)
|
1 年之前 |
 Galunid
|
f31ec120bc
Add warning if token is invalid (#7173)
|
1 年之前 |
 Daniel Bevenius
|
fd9f92b154
llama : update llama_timings.n_p_eval setting (#7160)
|
1 年之前 |
 Sigbjørn Skjæret
|
22842164bc
gguf-py : add special token modification capability (#7166)
|
1 年之前 |
 Albert Jin
|
4734524882
opencl : alignment size converted from bits to bytes (#7090)
|
1 年之前 |
