Skip to content

Actions: ggml-org/llama.cpp

Actions

Release

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
release.yml
2,500+ workflow runs
2,500+ workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
hexagon: MUL_MAT, MUL_MAT_ID, FLASH_ATTN and GDN cleanup and optimiza… Release #3187: Commit 8f7f3bf pushed by max-krasnyansky
Pending master
master
Pending
hexagon: add gelu_quick (#24007) Release #3186: Commit d178a11 pushed by max-krasnyansky
In progress master
master
In progress
server: real-time reasoning interruption via control endpoint (#23971) Release #3185: Commit 354ebac pushed by allozaur
31m 2s master
master
31m 2s
clean up unused variables warnings (#23975) Release #3184: Commit 1fd5f48 pushed by am17an
56m 0s master
master
56m 0s
opencl: fix compiler warnings for non-adreno path (#23922) Release #3183: Commit 210a657 pushed by lhez
31m 30s master
master
31m 30s
speculative : fix n_outputs_max and remove draft-simple auto-enable (… Release #3182: Commit 5dcb711 pushed by ggerganov
34m 38s master
master
34m 38s
opencl: add basic support for q5_0 and q5_1 (#23548) Release #3181: Commit 27d9ed8 pushed by lhez
2h 21m 49s master
master
2h 21m 49s
vendor : update cpp-httplib to 0.46.1 (#23980) Release #3180: Commit 335abed pushed by ggerganov
2h 48m 33s master
master
2h 48m 33s
llama: limit max outputs of llama_context (#23861) Release #3179: Commit de6f727 pushed by ggerganov
4h 26m 43s master
master
4h 26m 43s
metal: template GLU kernels to support f16/f32 (#23882) Release #3178: Commit 95b8b8e pushed by ggerganov
6h 22m 16s master
master
6h 22m 16s
vulkan: don't hold the device mutex while compiling pipelines (#23641) Release #3177: Commit 55ac090 pushed by 0cc4m
6h 30m 36s master
master
6h 30m 36s
vulkan: reduce host memory lock contention (#23376) Release #3176: Commit bef69f1 pushed by 0cc4m
6h 7m 44s master
master
6h 7m 44s
vocab: add normalizer.lowercase support to WPM (#23899) Release #3175: Commit 5aba536 pushed by ggerganov
6h 16m 35s master
master
6h 16m 35s
TP: quantized KV cache support (#23792) Release #3174: Commit 8e6fff8 pushed by JohannesGaessler
6h 54m 59s master
master
6h 54m 59s
model: Add EXAONE 4.5 implementations (#21733) Release #3173: Commit 48b88c3 pushed by ngxson
5h 11m 10s master
master
5h 11m 10s
vulkan: Block-load Q3_K/Q6_K block data and subtract on 32b ints (#23… Release #3172: Commit 1962000 pushed by 0cc4m
4h 39m 50s master
master
4h 39m 50s
vulkan: Removed unused functions (#23175) Release #3171: Commit f8c0a19 pushed by 0cc4m
4h 2m 2s master
master
4h 2m 2s
common : support manually triggering the reasoning budget end sequenc… Release #3170: Commit 5254a79 pushed by pwilkin
1h 58m 15s master
master
1h 58m 15s
[SYCL] Support Q4_1, Q5_0, Q5_1 in Flash-attention (#23812) Release #3169: Commit a511424 pushed by ggerganov
4h 8m 5s master
master
4h 8m 5s
[SYCL] Add more types in GET_ROWS OP (#23710) Release #3168: Commit 4162522 pushed by ggerganov
3h 16m 29s master
master
3h 16m 29s
sycl : Optimize Q3_K mul_mat by reorder (#23725) Release #3167: Commit 44e211c pushed by ggerganov
2h 43m 25s master
master
2h 43m 25s
server : handle If-None-Match weak ETags (#23916) Release #3166: Commit 6f165c1 pushed by aldehir
34m 5s master
master
34m 5s
vocab : add tokenizer support for jina-embeddings-v2-base-zh (#18756) Release #3165: Commit d4c8e2c pushed by CISC
33m 20s master
master
33m 20s
ui: fix ETag truncation with MSVC compiler (#23917) Release #3164: Commit 3292da0 pushed by allozaur
33m 51s master
master
33m 51s
llama: only use one iGPU device by default (#23897) Release #3163: Commit 22cadc1 pushed by 0cc4m
43m 23s master
master
43m 23s