 Sachin Desai
|
3db4da56a5
chat : support Granite model reasoning and tool call (#14864)
|
hai 5 meses |
 Georgi Gerganov
|
fd1234cb46
llama : add gpt-oss (#15091)
|
hai 5 meses |
 compilade
|
19f68fa5a4
imatrix : warn when GGUF imatrix is saved without .gguf suffix (#15076)
|
hai 5 meses |
|
compilade
|
d31192b4ee
imatrix : use GGUF by default (#14842)
|
hai 5 meses |
 Aman Gupta
|
784524053d
Fix params bug in diffusion example (#14993)
|
hai 5 meses |
 Diego Devesa
|
d6818d06a6
llama : allow other bufts when overriding to CPU, add --no-repack option (#14990)
|
hai 5 meses |
 g2mt
|
94933c8c2e
server : implement universal assisted decoding (#12635)
|
hai 5 meses |
|
Aman Gupta
|
8a4a856277
Add LLaDA 8b Diffusion model (#14771)
|
hai 5 meses |
 Ed Addario
|
d1aa0cc5d1
imatrix: add option to display importance score statistics for a given imatrix file (#12718)
|
hai 6 meses |
|
compilade
|
90083283ec
imatrix : use GGUF to store importance matrices (#9400)
|
hai 6 meses |
|
Georgi Gerganov
|
225e7a1438
llama : add high-throughput mode (#14363)
|
hai 6 meses |
|
Aman Gupta
|
ab14019821
Support diffusion models: Add Dream 7B (#14644)
|
hai 6 meses |
|
Georgi Gerganov
|
6ffd4e9c44
server : pre-calculate EOG logit biases (#14721)
|
hai 6 meses |
 Alawode Oluwandabira
|
17a1f0d2d4
server: Add ability to mount server at prefix (#14544)
|
hai 6 meses |
 matteo
|
caf5681fcb
server : support jinja extra template kwargs (Qwen3 enable_thinking feature), from command line and from client (#13196)
|
hai 6 meses |
 Sigbjørn Skjæret
|
88fc854b4b
llama : improve sep token handling (#14272)
|
hai 7 meses |
 aa956
|
d67341dc18
server : add server parameters for draft model cache type (#13782)
|
hai 7 meses |
|
Georgi Gerganov
|
d3e64b9f49
llama : rework embeddings logic (#14208)
|
hai 7 meses |
 Olivier Chafik
|
c9bbc77931
`server`: update deepseek reasoning format (pass reasoning_content as diffs) (#13933)
|
hai 7 meses |
|
Olivier Chafik
|
cdf94a1802
server: --offline mode (#13804)
|
hai 8 meses |
|
Olivier Chafik
|
e121edc432
`server`: add `--reasoning-budget 0` to disable thinking (incl. qwen3 w/ enable_thinking:false) (#13771)
|
hai 8 meses |
|
Olivier Chafik
|
f5cd27b71d
`server`: streaming of tool calls and thoughts when `--jinja` is on (#12379)
|
hai 8 meses |
 Xuan-Son Nguyen
|
797990c4bc
mtmd : add ultravox audio input (#13623)
|
hai 8 meses |
|
Georgi Gerganov
|
a4090d1174
llama : remove llama_kv_cache_view API + remove deprecated (#13653)
|
hai 8 meses |
|
Georgi Gerganov
|
e298d2fbd0
kv-cache : add SWA support (#13194)
|
hai 8 meses |
 psocolovsky
|
1dfbf2cf3a
common : add load_progress_callback (#13617)
|
hai 8 meses |
 Isaac McFadyen
|
6a2bc8bfb7
server : added --no-prefill-assistant flag (#13608)
|
hai 8 meses |
 Olivier Chafik
|
3198405e98
`common`: add partial regex support (#12808)
|
hai 8 meses |
 Johannes Gäßler
|
10d2af0eaa
llama/ggml: add LLM training support (#10544)
|
hai 8 meses |
 David Huang
|
7f323a589f
Add `--no-op-offload` to improve `-ot` pp perf in MoE models like llama4 400B (#13386)
|
hai 8 meses |
