Tags

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

b2684

dbceec87 · llama : add StableLM2 12B (#6635) · Apr 16, 2024
b2683

f4dea7da · llama : add qwen2moe (#6074) · Apr 16, 2024
b2681

58227ffd · perplexity : require positive --ctx-size arg (#6695) · Apr 16, 2024
b2680

4fbd8098 · gguf : add special tokens metadata for FIM/Infill (#6689) · Apr 16, 2024
b2679

7593639c · `main`: add --json-schema / -j flag (#6659) · Apr 15, 2024
b2678

132f5579 · llama : fix restoring the number of outputs from state files (#6687) · Apr 15, 2024
b2676

7fc16a2c · swift : linux support (#6590) · Apr 15, 2024
b2675

17e98d4c · fix mul_mat_id() for new input, make the ut pass (#6682) · Apr 15, 2024
b2674

1958f7e0 · llama : add missing kv clear in llama_beam_search (#6664) · Apr 14, 2024
b2673

04fbc5f2 · Add Command R chat template (#6650) · Apr 14, 2024
b2671

422c2aff · Added support for GGML_OP_CLAMP in Metal (#6662) · Apr 14, 2024
b2670

8800226d · Fix --split-max-size (#6655) · Apr 14, 2024
b2669

e689fc4e · [bug fix] convert github repository_owner to lowercase (#6673) · Apr 14, 2024
b2667

de17e3f7 · fix memcpy() crash, add missed cmd in guide, fix softmax (#6622) · Apr 14, 2024
b2666

b5e7285b · CUDA: fix matrix multiplication logic for tests (#6667) · Apr 14, 2024
b2665

4bd0f93e · model: support arch `DbrxForCausalLM` (#6515) · Apr 13, 2024
b2664

ab9a3240 · JSON schema conversion: ⚡ faster repetitions, min/maxLength for strings, cap number length (#6555) · Apr 12, 2024
b2663

fbbc030b · metal : unify mul_mv_id kernels (#6556) · Apr 12, 2024
b2661

24ee66ed · server : coherent log output for KV cache full (#6637) · Apr 12, 2024
b2660

91c73601 · llama : add gguf_remove_key + remove split meta during quantize (#6591) · Apr 12, 2024

Prev
1
…
10
11
12
13
14
15
16
17
18
…
99
Next