 Johannes Gäßler
|
af0a5b6163
server: fix incorrectly reported token probabilities (#7125)
|
пре 1 година |
 maor-ps
|
03fb8a002d
If first token generated from the server is the stop word the server will crash (#7038)
|
пре 1 година |
 Georgi Gerganov
|
9c67c2773d
ggml : add Flash Attention (#5021)
|
пре 1 година |
 Olivier Chafik
|
8843a98c2b
Improve usability of --model-url & related flags (#6930)
|
пре 1 година |
 Pierrick Hymbert
|
b7368332e2
ci: server: tests python env on github container ubuntu latest / fix n_predict (#6935)
|
пре 1 година |
|
Pierrick Hymbert
|
0c4d489e29
quantize: add imatrix and dataset metadata in GGUF (#6658)
|
пре 1 година |
|
Pierrick Hymbert
|
7f5ff558ee
server: stop generation at `n_ctx_train` if `n_predict` is not set (#6638)
|
пре 1 година |
 Kyle Mistele
|
37246b1031
common : revert showing control tokens by default for server (#6860)
|
пре 1 година |
|
Johannes Gäßler
|
28103f4832
Server: fix seed for multiple slots (#6835)
|
пре 1 година |
 Pedro Cuenca
|
b97bc3966e
llama : support Llama 3 HF conversion (#6745)
|
пре 1 година |
|
Olivier Chafik
|
ab9a3240a9
JSON schema conversion: ⚡️ faster repetitions, min/maxLength for strings, cap number length (#6555)
|
пре 1 година |
|
Pierrick Hymbert
|
24ee66ed0d
server : coherent log output for KV cache full (#6637)
|
пре 1 година |
 Jared Van Bortel
|
1b67731e18
BERT tokenizer fixes (#6498)
|
пре 1 година |
 Jan Boon
|
beea6e1b16
llama : save and restore kv cache for single seq id (#6341)
|
пре 1 година |
|
Georgi Gerganov
|
4399f13fb9
server : remove obsolete --memory-f32 option
|
пре 1 година |
 Xiao-Yong Jin
|
1a43c7254e
server : add option to disable KV offload (#6468)
|
пре 1 година |
 Eric Zhang
|
6902cb7f2e
server : stop gracefully on SIGTERM (#6348)
|
пре 1 година |
 compilade
|
557410b8f0
llama : greatly reduce output buffer memory usage (#6122)
|
пре 1 година |
|
Jan Boon
|
3d032ece8e
server : add `n_discard` parameter (#6300)
|
пре 1 година |
 slaren
|
280345968d
cuda : rename build flag to LLAMA_CUDA (#6299)
|
пре 1 година |
 Xuan Son Nguyen
|
ad3a0505e3
Server: clean up OAI params parsing function (#6284)
|
пре 1 година |
|
Pierrick Hymbert
|
f482bb2e49
common: llama_load_model_from_url split support (#6192)
|
пре 1 година |
|
Olivier Chafik
|
72114edf06
json-schema-to-grammar : fix order of props + non-str const/enum (#6232)
|
пре 1 година |
 Jan Boon
|
6b8bb3a31d
server : fix n_keep always showing as 0 in response (#6211)
|
пре 1 година |
|
Georgi Gerganov
|
68e210b354
server : enable continuous batching by default (#6231)
|
пре 1 година |
|
Olivier Chafik
|
5b7b0ac8df
json-schema-to-grammar improvements (+ added to server) (#5978)
|
пре 1 година |
|
Pierrick Hymbert
|
d01b3c4c32
common: llama_load_model_from_url using --model-url (#6098)
|
пре 1 година |
|
slaren
|
f30ea47a87
llama : add pipeline parallelism support (#6017)
|
пре 1 година |
|
Xuan Son Nguyen
|
99b71c068f
Server: Use multi-task for embeddings endpoint (#6001)
|
пре 1 година |
|
Xuan Son Nguyen
|
caa106d4e0
Server: format error to json (#5961)
|
пре 1 година |
