Default Branch

4ebdf2c74a · ci : use t4-medium for cuda jobs (#29842) · Updated 2026-10-02 10:31:19 -05:00

Branches

56fe6a8187 · models : fix dflash wo_a reshape on load · Updated 2026-08-04 08:41:42 -05:00

1101
1

201e50cc20 · clean up comments · Updated 2026-08-04 08:18:19 -05:00

1117
56

eb75c420f0 · Shared default of 64 for history-based samplers, remove context_size · Updated 2026-08-04 01:02:53 -05:00

1111
2

3dfb7c7fad · server: simplify get_info result handling · Updated 2026-08-03 11:22:56 -05:00

1122
4

7f5903206c · correct to 9931 · Updated 2026-08-03 05:18:50 -05:00

1130
3

ddd9cdb158 · reuse str_to_lower from server-models · Updated 2026-08-02 13:41:31 -05:00

1146
2

1e95e764b1 · fix convert bug · Updated 2026-08-02 09:35:46 -05:00

1141
3

a850ca8c62 · test: fix some CI errors · Updated 2026-08-01 16:15:10 -05:00

1149
1

37e22b750e · improve Security review a bit · Updated 2026-08-01 10:48:31 -05:00

1151
2

55467b6b66 · fix build · Updated 2026-07-31 03:08:09 -05:00

1180
23

10329a2abc · mtmd: add n_embd_head · Updated 2026-07-30 15:14:00 -05:00

1169
1

f531c878c2 · mtmd: add lanczos resize method · Updated 2026-07-30 14:23:15 -05:00

1172
1

76bb9cb0ad · fix ~server_batch() · Updated 2026-07-30 11:31:20 -05:00

1180
2

a0f1a44b8d · rm has_logit_bias · Updated 2026-07-29 09:41:54 -05:00

1188
3

96ece714c4 · metal : per-device tuned (Q, NE) for flash-attn vec (#25750) · Updated 2026-07-29 08:46:01 -05:00

1199
8

72d0d8de4e · common_http_client · Updated 2026-07-28 16:51:35 -05:00

1196
1

f45ca59f0f · server: abstract llama_memory calls to common_memory · Updated 2026-07-28 04:36:16 -05:00

1208
1

be4c137590 · nits 2 · Updated 2026-07-27 14:35:03 -05:00

1220
5

7ecdaafc35 · add example · Updated 2026-07-27 04:45:54 -05:00

1223
2

987e91cae2 · fix lint · Updated 2026-07-26 18:06:31 -05:00

1229
2