]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2024-10-10 Diego Devesarpc : add backend registry / device interfaces (#9812)
2024-10-10 R0CKSTARmusa: add docker image support (#9685)
2024-10-10 Diego Devesaexamples : do not use common library in simple example...
2024-10-09 Diego Devesacmake : do not build common library by default when...
2024-10-09 Georgi Gerganovperplexity : fix integer overflow (#9783)
2024-10-09 Georgi Gerganovexamples : remove llama.vim
2024-10-08 Diego Devesaggml : fix BLAS with unsupported types (#9775)
2024-10-08 Xuan Son Nguyenserver : better security control for public deployments...
2024-10-08 standby24x7scripts : fix spelling typo in messages and comments...
2024-10-07 Diego Devesaggml : add backend registry / device interfaces to...
2024-10-07 Andrew Minh... Update building for Android (#9672)
2024-10-07 Georgi Gerganovflake.lock: Update (#9753)
2024-10-07 Georgi Gerganovggml : add metal backend registry / device (#9713)
2024-10-07 Paul Tsochantarismetal : single allocation of encode_async block (#9747)
2024-10-06 Georgi Gerganovcontrib : simplify + minor edits [no ci]
2024-10-06 Georgi Gerganovreadme : fix typo [no ci]
2024-10-06 Georgi Gerganovsync : llama.cpp
2024-10-06 SRHMorrisvulkan : retry allocation with fallback flags (whisper...
2024-10-05 Georgi Gerganovrerank : use [SEP] token instead of [BOS] (#9737)
2024-10-05 Georgi Gerganovsync : ggml
2024-10-05 Georgi Gerganovmetal : zero-init buffer contexts (whisper/0)
2024-10-04 Viet-Anh NGUYEN... Add Llama Assistant (#9744)
2024-10-04 Georgi Gerganovsync : ggml
2024-10-04 Daniel Beveniusggml : fix typo in example usage ggml_gallocr_new ...
2024-10-04 Diego Devesaggml : fixes after sync (ggml/983)
2024-10-04 Xuan Son Nguyenci : fine-grant permission (#9710)
2024-10-04 Daniel KleineFixed RNG seed docs (#9723)
2024-10-03 Georgi Gerganovmetal : remove abort (skip) (ggml/0)
2024-10-03 Georgi Gerganovsync : ggml
2024-10-03 Johannes Gäßlerggml/ex: calculate accuracy in graph, adapt MNIST ...
2024-10-03 Johannes Gäßlerggml: refactor cross entropy loss CPU impl. (ggml/976)
2024-10-03 Jack Mousseaumetal : fix compute pass descriptor autorelease crash...
2024-10-03 Diego Devesaggml-backend : add device description to CPU backend...
2024-10-03 bandotiggml: unify backend logging mechanism (#9709)
2024-10-03 compiladeconvert : handle tokenizer merges format from transform...
2024-10-03 Radoslav Gerganovrpc : enable vulkan (#9714)
2024-10-03 Ouadie EL FAROUKIFixed dequant precision issues in Q4_1 and Q5_1 (#9711)
2024-10-02 Diego Devesaggml-backend : add device and backend reg interfaces...
2024-10-02 Xuan Son Nguyenllama : reduce compile time and binary size (#9712)
2024-10-02 Alberto Cabrera... [SYCL] Initial cmake support of SYCL for AMD GPUs ...
2024-10-02 Radoslav Gerganovvulkan : do not use tensor->extra (#9407)
2024-10-02 Zhenwei Jingguf-split : improve --split and --merge logic (#9619)
2024-10-02 Georgi Gerganovexamples : remove benchmark (#9704)
2024-10-01 Paweł WodnickiUpdate README.md (#9591)
2024-10-01 Georgi Gerganovsync : ggml
2024-10-01 Johannes Gäßlertest: fix OPT_STEP_ADAMW for test-backend-ops (ggml...
2024-10-01 Salvatore Mesoracavulkan : mul_mat: fix UB with small warps (ggml/952)
2024-10-01 Borislav Stanimirovggml : fix ggml_cast (ggml/973)
2024-10-01 Johannes Gäßlerggml: fix gradient allocation logic (ggml/966)
2024-10-01 Georgi Gerganovmetal : reduce command encoding overhead (#9698)
2024-10-01 Georgi Gerganovllama : print correct model type for Llama 3.2 1B and 3B
2024-10-01 compiladeconvert : refactor rope_freqs generation (#9396)
2024-09-30 serhii-nakonFix Docker ROCM builds, use AMDGPU_TARGETS instead...
2024-09-30 compiladeci : reduce severity of unused Pyright ignore comments...
2024-09-30 vbpy : update transfomers version (#9694)
2024-09-30 Georgi Gerganovflake.lock: Update (#9680)
2024-09-30 Ruchira Hasarangaconsole : utf-8 fix for windows stdin (#9690)
2024-09-29 Georgi Gerganovggml : define missing HWCAP flags (#9684)
2024-09-29 Georgi Gerganovsync : ggml
2024-09-29 Johannes GäßlerCUDA: remove bad assert (ggml/972)
2024-09-29 Jeff Bolzvulkan : multithread pipeline creation (ggml/963)
2024-09-29 Jeff Bolzvulkan : fix build for GGML_VULKAN_RUN_TESTS, add TFLOP...
2024-09-29 Salvatore Mesoracavulkan : argsort barriers must be under uniform control...
2024-09-29 Georgi Gerganovggml : fix GGML_MAX_N_THREADS + improve formatting...
2024-09-29 matiaslincommon : ensure llama_batch size does not exceed max...
2024-09-29 nopperlpy : add model class for Chameleon conversion (#9683)
2024-09-29 Georgi Gerganovcontrib : add Resources section (#9675)
2024-09-28 Georgi Gerganovllama : add reranking support (#9510)
2024-09-28 slarentest-backend-ops : use flops for some performance tests...
2024-09-28 Georgi Gerganovllama : add comment about thread-safety [no ci] (#9449)
2024-09-28 Zhenwei Jinvocab : refactor tokenizer to reduce init overhead...
2024-09-28 nopperlllama : add support for Chameleon (#8543)
2024-09-28 Aarni Koskelareadme : add tool (#9655)
2024-09-28 Dan Johanssonggml : add run-time detection of neon, i8mm and sve...
2024-09-28 Markus TavenrathEnable use to the rebar feature to upload buffers to...
2024-09-27 Georgi Gerganovreadme : update hot topics
2024-09-27 Borislav Stanimirovcmake : add option for common library (#9661)
2024-09-26 Neo Zhang Jianyu[SYCL] add missed dll file in package (#9577)
2024-09-26 R0CKSTARmtgpu: enable VMM (#9597)
2024-09-25 Xuan Son Nguyenci : fix docker build number and tag name (#9638)
2024-09-25 Charles Xuggml : remove assert for AArch64 GEMV and GEMM Q4 kerne...
2024-09-25 Xuan Son Nguyenserver : add more env vars, improve gen-docs (#9635)
2024-09-25 Gabe Goodhartllama : add IBM Granite MoE architecture (#9438)
2024-09-25 Dou Xinpengcann: fix crash when llama-bench is running on multiple...
2024-09-24 Eric Zhangggml : add AVX512DQ requirement for AVX512 builds ...
2024-09-24 Georgi Gerganovsync : ggml
2024-09-24 Georgi Gerganovexamples : adapt to ggml.h changes (ggml/0)
2024-09-24 Georgi Gerganovllama : keep track of all EOG tokens in the vocab ...
2024-09-24 Georgi Gerganovlog : add CONT level for continuing previous log entry...
2024-09-24 StrangeBytesDevserver : add newline after chat example (#9616)
2024-09-24 Georgi Gerganovsampling : avoid expensive softmax during greedy sampli...
2024-09-24 Max Krasnyanskythreads: fix msvc build without openmp (#9615)
2024-09-24 Ivancuda: add q8_0->f32 cpy operation (#9571)
2024-09-23 Xuan Son Nguyenserver : add --no-context-shift option (#9607)
2024-09-23 Max Krasnyanskythreads: improve ggml_barrier scaling with large number...
2024-09-23 Riceball LEEreadme : add programmable prompt engine language CLI...
2024-09-23 Georgi Gerganovflake.lock: Update (#9586)
2024-09-23 Srihari-mcwggml : AVX512 gemm for Q4_0_8_8 (#9532)
2024-09-23 Georgi Gerganovperplexity : remove extra new lines after chunks (...
2024-09-23 Georgi Gerganovmetal : use F32 prec for K*Q in vec FA (#9595)
next