Tags

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

b1056

2e5f70a2 · Added `enum` to `llama_token_get_type` return type (#2774) · Aug 24, 2023
b1054

0d3094f0 · gguf : add rope_freq_base parameter for CodeLlama (#2769) · Aug 24, 2023
b1052

38b16dfc · metal : bug-fix when enable ggml-alloc (#2757) · Aug 24, 2023
b1050

7694adda · Fix for main example getting stuck when -n -2 and --interactive (#2767) · Aug 24, 2023
b1049

fea95c68 · fix convert.py for codellama, add llama 34B to the list of recognized models (#2768) · Aug 24, 2023
b1048

ef955fbd · Tag release with build number (#2732) · Aug 24, 2023
b1047

c3e53b42 · llama : escape all U+2581 in a string (#2750) · Aug 24, 2023
b1046

6e91a1b0 · llama : fix grammar sometimes generating null char (#2756) · Aug 24, 2023
master-cf658ad

cf658adc · llm : add Falcon support (#2717) · Aug 23, 2023
master-a192860

a192860c · minor : fix trailing whitespace · Aug 23, 2023
master-9538524

95385241 · examples : restore the functionality to import llama2.c models (#2685) · Aug 23, 2023
master-5290c38

5290c38e · main : insert bos if no tokens (#2727) · Aug 23, 2023
master-8207214

8207214b · Fix values shown in the quantize tool help (#2735) · Aug 23, 2023
master-62959e7

62959e74 · Strided perplexity (#2714) · Aug 23, 2023
master-b8ad1b6

b8ad1b66 · server : allow json array in prompt or content for direct token input (#2306) · Aug 23, 2023
master-777f42b

777f42ba · Improve handling of special tokens in GGML to GGUF converter (#2725) · Aug 22, 2023
master-46ef5b5

46ef5b5f · llama : fix whitespace escaping in tokenizer (#2724) · Aug 23, 2023
master-c63bb1d

c63bb1d1 · CUDA: use mul_mat_q kernels by default (#2683) · Aug 22, 2023
master-800c963

800c9635 · Fix CUDA softmax by subtracting max value before exp (#2665) · Aug 22, 2023
master-deb7dfc

deb7dfca · gguf : add ftype meta info to the model (#2710) · Aug 22, 2023

Prev
1
…
62
63
64
65
66
67
68
69
70
…
99
Next