]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2023-05-08 Pavol Rusnakllama : fix hparams shadow (#1367)
2023-05-08 Georgi Gerganovllama : require first token to be BOS (#1303)
2023-05-08 ubik2convert: add ability to convert safetensors files ...
2023-05-08 Johannes GäßlerDocumented CUDA reproducibility, added warning (#1346)
2023-05-07 Henri VassermanCI: add Windows CLBlast and OpenBLAS builds (#1277)
2023-05-07 swittkggml : Allow usage of CLBlast alongside Accelerate...
2023-05-06 Jed FoxRemove default arguments from sampling functions (...
2023-05-05 DaniAndTheWebmakefile: automatic Arch Linux detection (#1332)
2023-05-05 Erik Scholzci : add cublas to windows release (#1271)
2023-05-05 Pavol Rusnakreadme: add missing info (#1324)
2023-05-05 Ionoclast LaboratoriesFix for OpenCL / clbast builds on macOS. (#1329)
2023-05-05 Benjamin LecaillonConvert.py @staticmethod (#1327)
2023-05-04 slarenquantize: make output filename optional, default to...
2023-05-04 Ivan StepanovWrap exceptions in std::exception to verbose output...
2023-05-04 Ivan Stepanovconvert: support DT_BF16 tensors (#1309)
2023-05-04 44670readme : add OpenBuddy link (#1321)
2023-05-04 44670main : add --in-suffix option (#1318)
2023-05-04 Ron Jailallggml : change immintrin.h to intrin.h for compatibility...
2023-05-04 DannyDaemonicOnly escape prompts when used with `-e` (#1311)
2023-05-04 DannyDaemonicUpdate main's README.md with new features (#1296)
2023-05-04 Tomasfix #1224 reverse prompt and multi line (#1297)
2023-05-03 Georgi Gerganovggml : vectorize Q8_0 quantization
2023-05-03 khimarosexamples : read chat prompts from a template file ...
2023-05-03 Georgi Gerganovminor : fix whitespaces (#1302)
2023-05-03 Georgi Gerganovminor : fix trailing whitespaces
2023-05-03 KASRscripts : platform independent script to verify sha256...
2023-05-03 CRD716examples : various prompt and example fixes (#1298)
2023-05-03 Evan Jonesllama : only copy used KV cache in get / set state...
2023-05-03 DannyDaemonicProcess escape sequences given in prompts (#1173)
2023-05-03 DannyDaemonicHandle signals properly on Windows (#1123)
2023-05-03 DannyDaemonicCall sh on build-info.sh (#1294)
2023-05-03 kuvausfix build-info.h for git submodules (#1289)
2023-05-02 slarenfix missing parameters in `llama_init_from_gpt_params...
2023-05-02 Ron Evansexamples : add llama_init_from_gpt_params() common...
2023-05-02 Georgi Gerganovllama : fix compile warnings
2023-05-02 Georgi Gerganovggml : fix 32-bit ARM
2023-05-02 Ron Evansexamples : improve vertical alignment of a few variable...
2023-05-02 Marvin Gießingggml : fix ppc64le build error and make cmake detect...
2023-05-02 Robert Brisitallama : allow 0 as a seed number. (#1275)
2023-05-02 Ron Evansmain : switch input_noecho to input_echo to remove...
2023-05-02 slarenggml: add names to tensors (#1268)
2023-05-01 DannyDaemonicAdd git-based build information for better issue tracki...
2023-05-01 slarencuBLAS: refactor and optimize f16 mat mul performance...
2023-05-01 xloemllama : update stubs for systems without mmap and mlock...
2023-05-01 Kerfuffleggml : fix ggml_used_mem() (#1264)
2023-05-01 Georgi Gerganovllama : fix session load / save (#1263)
2023-05-01 slarencuBLAS: fall back to pageable memory if pinned alloc...
2023-05-01 Alex Klinkhamerllama : let context be const when accessing const data...
2023-04-30 Georgi Gerganovggml : fix UB (int << 31)
2023-04-30 Pavol Rusnakbuild: add armv{6,7,8} support to cmake (#1251)
2023-04-30 jon-chuangcommon : better default number of threads (#934)
2023-04-30 0cc4mggml : add CLBlast q5_0, q5_1, q8_0 dequant kernels...
2023-04-30 Georgi Gerganovggml : add Q5 WASM SIMD + GGML_FTYPE
2023-04-30 Stephan WalterVarious fixes to mat_mul benchmark (#1253)
2023-04-30 Georgi Gerganovggml : fix labels for GGML_OP_ALIBI
2023-04-29 Georgi Gerganovggml : fix 32-bit ARM NEON
2023-04-29 Georgi Gerganovggml : use vzip instead of vuzp for consistency
2023-04-29 Georgi Gerganovggml : fix visibility and unused warnings
2023-04-29 Georgi Gerganovggml : fix #if for f32_f32 mul_mat (CLBlast) (#1229)
2023-04-29 Georgi Gerganovggml : adjust mul_mat_f16 work memory (#1226)
2023-04-29 Georgi Gerganovbuild : fix reference to old llama_util.h
2023-04-29 Georgi Gerganovexamples : fix save-load-state + rename llama-util.h
2023-04-29 Georgi Gerganovcommon : change default parameters to pre-#1126 (#1223)
2023-04-29 Ivan Stepanovllama : new sampling algorithms (#1126)
2023-04-29 slarencuBLAS: use host pinned memory and dequantize while...
2023-04-28 Henri VassermancuBLAS: non-contiguous tensor support (#1215)
2023-04-28 Stephan WalterRemove Q4_3 which is no better than Q5 (#1218)
2023-04-28 Georgi Gerganovreadme : update hot topics
2023-04-28 Georgi Gerganovggml : sync ggml (ggml_alibi)
2023-04-28 CRD716examples : add Jeopardy example (#1168)
2023-04-28 Evan Jonesllama : add session file format and saved sessions...
2023-04-28 Georgi Gerganovggml : add helper debug printf in soft_max
2023-04-28 0cc4mggml : add CLBlast support (#1164)
2023-04-28 Folko-VenCorrecting link to w64devkit (#1214)
2023-04-28 Johannes GäßlerAdd Manjaro CUDA include and lib dirs to Makefile ...
2023-04-28 Yann Folletadd avx2 for dot_q8_0_q8_0, 2x faster than scalar ...
2023-04-26 Stephan Walterggml : slightly faster AVX2 implementation for Q5 ...
2023-04-26 Georgi Gerganovreadme : add quantization info
2023-04-26 Georgi Gerganovggml : add Q5_0 and Q5_1 quantization (#1187)
2023-04-26 Ásgeir Bjarni... Allow setting the rng seed after initialization. (...
2023-04-26 DaniAndTheWebUpdating build instructions to include BLAS support...
2023-04-26 Pavol Rusnakquantize : use `map` to assign quantization type from...
2023-04-25 Stephan WalterUpdate SHA256SUMS after quantization change (#1181)
2023-04-25 ostix360py : cast lora_alpha to int in convert-lora-to-ggml...
2023-04-25 Pavol Rusnaknix: use convert.py instead of legacy wrapper convert...
2023-04-25 Georgi Gerganovggml : add Q8_0 quantization format (rename the old...
2023-04-25 unboundedggml : use full range for Q4_0 and Q4_2 quantization...
2023-04-24 xaedesggml : fix bug in ggml_compute_forward_sum_f32 (#1162)
2023-04-24 Georgi Gerganovggml : export symbols (#1155)
2023-04-24 xaedesexamples : add save_load_state example (#1150)
2023-04-24 Georgi Gerganovllama : increase scratch buffer size for 65B (ref ...
2023-04-24 mgroeber9110examples/main README improvements and some light refact...
2023-04-24 Stephan WalterFix build for gcc 8 and test in CI (#1154)
2023-04-24 slarenFix cuda compilation (#1128)
2023-04-24 Georgi Gerganovllama : refactor get / set state + remove redundant...
2023-04-23 slarenFix LoRA acronym (#1145)
2023-04-23 Georgi Gerganovscripts : add helper scripts to synch ggml repo
2023-04-23 DannyDaemonicAdded README.md for main with examples and explanations...
2023-04-23 Georgi Gerganovggml : do not print perf ops that have not been used...
2023-04-23 Georgi Gerganovggml : better PERF prints + support "LLAMA_PERF=1 make"
next