Tags

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

b2717

c8297c6a · llama : add phi3 support (#6852) · Apr 24, 2024
b2715

192090ba · llamafile : improve sgemm.cpp (#6796) · Apr 22, 2024
b2714

e931888d · ggml : fix calloc argument ordering. (#6820) · Apr 22, 2024
b2712

c0956b09 · ci: fix job are cancelling each other (#6781) · Apr 22, 2024
b2710

5cf5e7d4 · `build`: generate hex dump of server assets during build (#6661) · Apr 21, 2024
b2709

40f74e4d · llama : add option to render special/control tokens (#6807) · Apr 21, 2024
b2708

b9cc76d8 · ggml : fix ggml_backend_cpu_supports_op() for CPY (#0) · Apr 21, 2024
b2707

7dbdba56 · llama : add llama-3 chat template (#6751) · Apr 21, 2024
b2702

b97bc396 · llama : support Llama 3 HF conversion (#6745) · Apr 21, 2024
b2700

aed82f68 · common : try to fix Android CI (#6780) · Apr 20, 2024
b2699

0e4802b2 · ci: add ubuntu latest release and fix missing build number (mac & ubuntu) (#6748) · Apr 19, 2024
b2698

637e9a86 · server: static: upstream upgrade (#6765) · Apr 19, 2024
b2697

9958c81b · Implement the OLMo architecture (#6741) · Apr 19, 2024
b2696

8b1b1f49 · train : add general name (#6752) · Apr 19, 2024
b2694

0d56246f · ggml : group all experts in a single ggml_mul_mat_id (#6505) · Apr 18, 2024
b2692

e11b2e6e · Qwen2 : assume tied weights if lm_head/output weights is missing (#6738) · Apr 18, 2024
b2691

c71bfd73 · llama : fix compatibility with old 2 expert models (#6735) · Apr 18, 2024
b2690

3b8f1ec4 · llamafile : tmp disable + build sgemm.o when needed (#6716) · Apr 17, 2024
b2687

532c1737 · llama : make general.name optional (#6709) · Apr 16, 2024
b2686

666867b7 · ggml : fix llamafile sgemm wdata offsets (#6710) · Apr 16, 2024

Prev
1
…
9
10
11
12
13
14
15
16
17
…
99
Next