Default Branch

d230ddd763 · llama: fix whole source code rebuilt on each new commit (#28278) · Updated 2026-09-03 23:53:04 +02:00

Branches

f0cbb6ddf6 · iq1_s: turn off SIMD implementation for QK_K = 64 (it does not work) · Updated 2024-02-28 07:28:10 +01:00    Superminaren

8512
6

14d757066b · llama : add llama_kv_cache_compress (EXPERIMENTAL) · Updated 2024-02-27 15:24:40 +01:00    Superminaren

8513
1

608f449880 · swift : fix build · Updated 2024-02-23 18:02:09 +01:00    Superminaren

8544
4

56c047156a · py : minor fixes · Updated 2024-02-22 18:22:56 +01:00    Superminaren

8553
1

5271c75666 · llama : fix K-shift with quantized K (wip) · Updated 2024-02-22 00:28:42 +01:00    Superminaren

8561
1

f249c997a8 · llama : adapt to F16 KQ_pos · Updated 2024-02-19 12:31:02 +01:00    Superminaren

8599
62

412735ec70 · Merge branch 'master' into gg/metal-batched · Updated 2024-02-19 10:25:24 +01:00    Superminaren

8599
6

47c662b0de · fix some spaces added by IDE in math op · Updated 2024-02-18 21:40:35 +01:00    Superminaren

8609
4

974e3cadff · ggml : try another fix · Updated 2024-02-17 17:14:35 +01:00    Superminaren

8628
2

e856bfed3b · hf : add support for --repo and --file · Updated 2024-02-15 14:05:15 +01:00    Superminaren

8642
3

ccd757a174 · convert : fix mistakes from refactoring · Updated 2024-02-13 18:01:30 +01:00    Superminaren

8650
4

5c977221d2 · iq1_s: slightly faster dot product · Updated 2024-02-13 14:18:27 +01:00    Superminaren

8656
15

4246b71ad7 · Fix compiler warnings (shadow variable) · Updated 2024-02-13 07:44:56 +01:00    Superminaren

8659
1

7286b83d3f · BERT WIP · Updated 2024-02-06 23:10:11 +01:00    Superminaren

8712
1

adcf16fd68 · py : fix empty bytes arg · Updated 2024-02-05 18:53:07 +01:00    Superminaren

8722
2

91c453fb11 · One cannot possibly be defining static_assert in a C++ compilation · Updated 2024-02-05 12:22:14 +01:00    Superminaren

8727
2

49a483e0f2 · wip · Updated 2024-02-04 11:34:36 +01:00    Superminaren

8753
60

a647257b47 · cuda : express strides with helper constants · Updated 2024-02-04 10:45:26 +01:00    Superminaren

8753
60

b957b8f5f6 · cuda : add flash_attn kernel (wip) · Updated 2024-02-01 18:49:57 +01:00    Superminaren

8757
39

ac26f27028 · cuda : increase C to 128 for better performance · Updated 2024-02-01 16:08:29 +01:00    Superminaren

8757
61