Tags

Tags give the ability to mark specific points in history as being important

b1617

05cd6e50 · server : recognize cache_prompt parameter in OAI API (#4347) · Dec 06, 2023
b1616

caa92492 · common : fix compile warning · Dec 06, 2023
b1615

da5eaef1 · speculative : support `--color` (#4343) · Dec 06, 2023
b1614

5f6e0c0d · grammar : pre-computed pieces + reserve mem + less string copies (#4330) · Dec 05, 2023
b1613

5aa365d8 · llama : allow overriding GGUF metadata when loading model (#4092) · Dec 05, 2023
b1612

52c8bc3c · sampling : custom samplers order (#4285) · Dec 05, 2023
b1611

e4b76bbe · swift : revert compiler checks for swift package (#4332) · Dec 05, 2023
b1610

23b5e12e · simple : update error message for KV cache check (#4324) · Dec 04, 2023
b1609

d208995c · swift : fix concatenation method to avoid invalid UTF8 stringfication (#4325) · Dec 04, 2023
b1608

5c9f90cb · swift : fix prompt tokenization logic (#4321) · Dec 04, 2023
b1607

4fa44e84 · grammar-parser : fix typo (#4318) · Dec 04, 2023
b1606

fbbc4282 · ggml : reuse ggml_get_n_tasks() in ggml_graph_plan() (#4308) · Dec 03, 2023
b1605

adf3de4f · ggml : fix soft max out-of-bounds access (#4307) · Dec 03, 2023
b1604

33e171d1 · server : fix OpenAI API `stop` field to be optional (#4299) · Dec 03, 2023
b1602

d7b800b8 · llama : pad KV cache size (#4280) · Dec 03, 2023
b1601

5a7d3125 · llama : avoid using "optional" keyword (#4283) · Dec 01, 2023
b1600

d5a1cbde · llama : support optional tensors (#4283) · Dec 01, 2023
b1599

b220222a · swift : fix token_to_piece implementation (#4278) · Dec 01, 2023
b1598

511f52c3 · build : enable libstdc++ assertions for debug builds (#4275) · Dec 01, 2023
b1597

03562f3a · llama : support attention bias on LLaMA architecture (#4283) · Dec 01, 2023

Prev
1
…
42
43
44
45
46
47
48
49
50
…
99
Next