Tags

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

b2970

197ff914 · build : remove zig (#7471) · May 22, 2024
b2969

6ff13987 · common : normalize naming style (#7462) · May 22, 2024
b2968

38c03478 · CUDA: fix FA out-of-bounds writes (#7465) · May 22, 2024
b2967

b18532a4 · phi3 : duplicate rope factors in each layer (#7447) · May 22, 2024
b2966

fcda1128 · vulkan: add workaround for iterator boundary check to fix clang-cl debug build (#7426) · May 22, 2024
b2965

03d8900e · llama : add missing model type names (#7445) · May 22, 2024
b2964

9b3d8331 · cuda : fix compile warning (#7454) · May 22, 2024
b2963

95fb0aef · CUDA: remove incorrect precision check (#7454) · May 22, 2024
b2962

3e5faa85 · cuda : fix rope + add tests (#7452) · May 22, 2024
b2961

201cc11a · llama : add phi3 128K model support (#7225) · May 21, 2024
b2958

fcf6538b · CUDA: fix unused warning in mmq.cu (#7442) · May 21, 2024
b2956

11474e75 · examples: cache hf model when --model not provided (#7353) · May 21, 2024
b2955

d8ee9022 · CUDA: deduplicate mmq code (#7397) · May 21, 2024
b2953

917dc8cf · Tokenizer SPM fixes for phi-3 and llama-spm (#7375) · May 20, 2024
b2952

fabf30b4 · llama : remove Persimmon (#7408) · May 21, 2024
b2950

db10f013 · rpc : track allocated buffers (#7411) · May 20, 2024
b2949

3bc10cb4 · server : fix temperature + disable some tests (#7409) · May 20, 2024
b2948

6bf9b66f · [SYCL] Update SYCL upscale operation (#7321) · May 20, 2024
b2946

213e90ed · ggml-opencl, llama: using reserve() if count already known (#7272) · May 20, 2024
b2945

65c58207 · ggml : add loongarch lsx and lasx support (#6454) · May 20, 2024

Prev
1
2
3
4
5
6
7
8
9
…
99
Next