| 2023-06-10 |
Aisuko | doc : fix wrong address of BLIS.md (#1772) |
commit | commitdiff | tree |
| 2023-06-10 |
Georgi Gerganov | ggml : force no_alloc == false when creating opt tensor... |
commit | commitdiff | tree |
| 2023-06-10 |
Kawrakow | metal : add Q4_1 implementation (#1785) |
commit | commitdiff | tree |
| 2023-06-10 |
Kerfuffle | llama : support requantizing models instead of only... |
commit | commitdiff | tree |
| 2023-06-10 |
Xingchen Song... | ggml : workaround for missing _mm256_setr_m128i in... |
commit | commitdiff | tree |
| 2023-06-10 |
rankaiyx | make : add SSSE3 compilation use case (#1659) |
commit | commitdiff | tree |
| 2023-06-09 |
Robert Sung... | OpenCL: Add release memory (#1741) |
commit | commitdiff | tree |
| 2023-06-09 |
Johannes Gäßler | Windows nvcc workaround (#1753) |
commit | commitdiff | tree |
| 2023-06-09 |
Georgi Gerganov | metal : fix build "tanhf" -> "tanh" |
commit | commitdiff | tree |
| 2023-06-09 |
AT | metal : add GELU implementation (#1770) |
commit | commitdiff | tree |
| 2023-06-09 |
Kawrakow | metal : faster q4_0 (#1775) |
commit | commitdiff | tree |
| 2023-06-08 |
Kawrakow | metal : add Q2_K implementation (#1762) |
commit | commitdiff | tree |
| 2023-06-08 |
Georgi Gerganov | Revert "ggml : load data into int8x16x4_t using vld4q_s... |
commit | commitdiff | tree |
| 2023-06-08 |
le.chang | ggml : load data into int8x16x4_t using vld4q_s8 on... |
commit | commitdiff | tree |
| 2023-06-08 |
Kawrakow | metal : Q6_K implementation (#1752) |
commit | commitdiff | tree |
| 2023-06-08 |
qingfengfenga | Add llama.cpp docker support for non-latin languages... |
commit | commitdiff | tree |
| 2023-06-08 |
Steven Roussey | ggml : fix fprintf warnings (#1720) |
commit | commitdiff | tree |
| 2023-06-08 |
Georgi Gerganov | clang-tidy : restore dot file from accidental deletion |
commit | commitdiff | tree |
| 2023-06-08 |
Kawrakow | metal : add Q4_K implementation (#1733) |
commit | commitdiff | tree |
| 2023-06-08 |
johnson442 | k-quants : add missing compile definition to CMakeLists... |
commit | commitdiff | tree |
| 2023-06-07 |
Georgi Gerganov | k-quants : allow to optionally disable at compile time... |
commit | commitdiff | tree |
| 2023-06-07 |
jacobi petrucciani | flake : update to support metal on m1/m2 (#1724) |
commit | commitdiff | tree |
| 2023-06-07 |
Georgi Gerganov | readme : add June roadmap |
commit | commitdiff | tree |
| 2023-06-07 |
Willy Tarreau | main: add the possibility to open the prompt cache... |
commit | commitdiff | tree |
| 2023-06-06 |
Georgi Gerganov | llama : fix vram_scratch var |
commit | commitdiff | tree |
| 2023-06-06 |
Georgi Gerganov | llama : fix compile warnings |
commit | commitdiff | tree |
| 2023-06-06 |
Johannes Gäßler | Multi GPU support, CUDA refactor, CUDA scratch buffer... |
commit | commitdiff | tree |
| 2023-06-06 |
Georgi Gerganov | metal : add f16 support |
commit | commitdiff | tree |
| 2023-06-06 |
LostRuins | Clblast fixes + enhancements to save VRAM and offload... |
commit | commitdiff | tree |
| 2023-06-06 |
Georgi Gerganov | ggml : fix builds, add ggml-quants-k.o (close #1712... |
commit | commitdiff | tree |
| 2023-06-06 |
Georgi Gerganov | gitignore : add .clang-tidy |
commit | commitdiff | tree |
| 2023-06-06 |
Georgi Gerganov | llama : temporary disable Q6_K output quantization... |
commit | commitdiff | tree |
| 2023-06-06 |
Spencer Sutton | metal : add checks for buffer size (#1706) |
commit | commitdiff | tree |
| 2023-06-05 |
Yuval Peled | docs : add performance troubleshoot + example benchmark... |
commit | commitdiff | tree |
| 2023-06-05 |
Foul-Tarnished | readme : fix typo (#1700) |
commit | commitdiff | tree |
| 2023-06-05 |
mgroeber9110 | llama : consistently catch and throw only exceptions... |
commit | commitdiff | tree |
| 2023-06-05 |
kiltyj | metal : use shared buffers between CPU and GPU (#1696) |
commit | commitdiff | tree |
| 2023-06-05 |
grahameth | ggml : fix internal overflow in ggml_time_us on Windows... |
commit | commitdiff | tree |
| 2023-06-05 |
Georgi Gerganov | ci : disable auto tidy (#1705) |
commit | commitdiff | tree |
| 2023-06-05 |
Kawrakow | ggml : add SOTA 2,3,4,5,6 bit k-quantizations (#1684) |
commit | commitdiff | tree |
| 2023-06-05 |
Henri Vasserman | Increase 3B scratch buffers. (#1698) |
commit | commitdiff | tree |
| 2023-06-05 |
Georgi Gerganov | llama : fix Metal KV cache sync (close #1695) |
commit | commitdiff | tree |
| 2023-06-04 |
Georgi Gerganov | readme : update hot topics |
commit | commitdiff | tree |
| 2023-06-04 |
Georgi Gerganov | llama : Metal inference (#1642) |
commit | commitdiff | tree |
| 2023-06-04 |
0cc4m | OpenCL: Fix duplication of layers in VRAM and RAM,... |
commit | commitdiff | tree |
| 2023-06-03 |
Henri Vasserman | Add info about CUDA_VISIBLE_DEVICES (#1682) |
commit | commitdiff | tree |
| 2023-06-03 |
Jiří Podivín | Docker: change to calling convert.py (#1641) |
commit | commitdiff | tree |
| 2023-06-03 |
Evan Jones | Fix prompt cache saving and chat-persistent rollover... |
commit | commitdiff | tree |
| 2023-05-30 |
Henri Vasserman | OpenLLaMA 3B support (#1588) |
commit | commitdiff | tree |
| 2023-05-29 |
Georgi Gerganov | ggml : sync cgraph import / export API |
commit | commitdiff | tree |
| 2023-05-29 |
Georgi Gerganov | ggml : fix bug in ggml_alibi |
commit | commitdiff | tree |
| 2023-05-29 |
DannyDaemonic | Work around for recalculating logits in cached prompts... |
commit | commitdiff | tree |
| 2023-05-29 |
Jiří Podivín | Adding git in container package dependencies (#1621) |
commit | commitdiff | tree |
| 2023-05-28 |
Johannes Gäßler | LLAMA_DEBUG adds debug symbols (#1617) |
commit | commitdiff | tree |
| 2023-05-28 |
Kerfuffle | Only show -ngl option when relevant + other doc/arg... |
commit | commitdiff | tree |
| 2023-05-28 |
Vladimir Zorin | examples : add --alias option to gpt_params to set... |
commit | commitdiff | tree |
| 2023-05-28 |
Howard Su | opencl : no need to allocate cl_mem on heap (#1612) |
commit | commitdiff | tree |
| 2023-05-28 |
Howard Su | opencl : use strstr to check if fp16 supported (#1611) |
commit | commitdiff | tree |
| 2023-05-27 |
apcameron | ggml : add support for the RISCV architecture (#1616) |
commit | commitdiff | tree |
| 2023-05-27 |
Kerfuffle | Include server in releases + other build system cleanup... |
commit | commitdiff | tree |
| 2023-05-27 |
Henri Vasserman | Add documentation about CLBlast (#1604) |
commit | commitdiff | tree |
| 2023-05-27 |
Henri Vasserman | [CI] Fix openblas (#1613) |
commit | commitdiff | tree |
| 2023-05-27 |
Georgi Gerganov | ggml : add ggml_tensor_overhead() |
commit | commitdiff | tree |
| 2023-05-27 |
Henri Vasserman | [CI] CLBlast: Fix directory name (#1606) |
commit | commitdiff | tree |
| 2023-05-27 |
Georgi Gerganov | ggml : sync ggml core (minor additions, e.g. ggml_get_t... |
commit | commitdiff | tree |
| 2023-05-26 |
Kerfuffle | Some improvements to loading the session with --prompt... |
commit | commitdiff | tree |
| 2023-05-25 |
Johannes Gäßler | cuda : performance optimizations (#1530) |
commit | commitdiff | tree |
| 2023-05-24 |
Henri Vasserman | Update CLBlast to 1.6.0 (#1580) |
commit | commitdiff | tree |
| 2023-05-24 |
Evan Jones | readme : add docs for chat-persistent.sh (#1568) |
commit | commitdiff | tree |
| 2023-05-24 |
Senemu | chat-persistent.sh : use bracket expressions in grep... |
commit | commitdiff | tree |
| 2023-05-23 |
Maarten ter... | Fix handling of "invalid property" when creating OpenCL... |
commit | commitdiff | tree |
| 2023-05-22 |
0cc4m | OpenCL Token Generation Acceleration (#1459) |
commit | commitdiff | tree |
| 2023-05-21 |
Steward Garcia | examples : add server example with REST API (#1443) |
commit | commitdiff | tree |
| 2023-05-21 |
Stefan Sydow | make : .PHONY clean (#1553) |
commit | commitdiff | tree |
| 2023-05-21 |
Georgi Gerganov | ggml : output 3d sizes in ggml_graph_dump_dot() |
commit | commitdiff | tree |
| 2023-05-20 |
Georgi Gerganov | ggml : update WASM SIMD |
commit | commitdiff | tree |
| 2023-05-20 |
Zenix | feature : support blis and other blas implementation... |
commit | commitdiff | tree |
| 2023-05-20 |
Henri Vasserman | OpenCL: Fixes for older devices. (#1435) |
commit | commitdiff | tree |
| 2023-05-20 |
Juuso Alasuutari | llama : define magic numbers as integer constants ... |
commit | commitdiff | tree |
| 2023-05-20 |
Georgi Gerganov | ggml : add ggml_clamp() (#1539) |
commit | commitdiff | tree |
| 2023-05-20 |
Johannes Gäßler | cuda : loading models directly into VRAM, norm calculat... |
commit | commitdiff | tree |
| 2023-05-20 |
Georgi Gerganov | Revert "feature : add blis and other BLAS implementatio... |
commit | commitdiff | tree |
| 2023-05-20 |
Zenix | feature : add blis and other BLAS implementation suppor... |
commit | commitdiff | tree |
| 2023-05-20 |
Georgi Gerganov | llama : add llama_init_backend() API (close #1527) |
commit | commitdiff | tree |
| 2023-05-20 |
DannyDaemonic | Fix for mingw (#1462) |
commit | commitdiff | tree |
| 2023-05-20 |
Maxime | llama : fix name shadowing and C4146 (#1526) |
commit | commitdiff | tree |
| 2023-05-20 |
Georgi Gerganov | llama : fix compile warnings in llama_set_state_data() |
commit | commitdiff | tree |
| 2023-05-20 |
Georgi Gerganov | ggml : fix scalar implementation of Q4_1 dot |
commit | commitdiff | tree |
| 2023-05-19 |
Georgi Gerganov | ggml : use F16 instead of F32 in Q4_0, Q4_1, Q8_0 ... |
commit | commitdiff | tree |
| 2023-05-19 |
Georgi Gerganov | tests : add missing header |
commit | commitdiff | tree |
| 2023-05-19 |
Evan Jones | examples : add persistent chat (#1495) |
commit | commitdiff | tree |
| 2023-05-19 |
Jason McCartney | main : make reverse prompt option act as a stop token... |
commit | commitdiff | tree |
| 2023-05-19 |
David Kennedy | readme : adds WizardLM to the list of supported models... |
commit | commitdiff | tree |
| 2023-05-19 |
Georgi Gerganov | minor : fix compile warnings |
commit | commitdiff | tree |
| 2023-05-18 |
Erik Scholz | make kv_f16 the default for api users (#1517) |
commit | commitdiff | tree |
| 2023-05-18 |
DannyDaemonic | Fixes #1511 lambda issue for w64devkit (mingw) (#1513) |
commit | commitdiff | tree |
| 2023-05-17 |
Stephan Walter | Remove unused n_parts parameter (#1509) |
commit | commitdiff | tree |
| 2023-05-17 |
rankaiyx | benchmark-matmul: Print the average of the test results... |
commit | commitdiff | tree |
| 2023-05-16 |
Tom Jobbins | convert.py: Support models which are stored in a single... |
commit | commitdiff | tree |
| 2023-05-16 |
Ilya Kurdyukov | ~7% faster Q5_1 AVX2 code (#1477) |
commit | commitdiff | tree |
| next |
overview / pkg / ggml / sources / llama.cpp / shortlog