command : always-prompt mode (#383)

go : adding features to the go-whisper example, go ci, etc (#384)

ggml : fix bug in new soft max computation

cmake : add -Wno-unused-function + update whisper.js

ggml : when using BLAS start only 1 CPU thread

ggml : fix running tasks with variable number of threads

ggml : unroll ggml_vec_dot_f16 in ggml_compute_forward_flash_attn_f16

whisper : revert accidental MB change

ggml : speed-up soft max via Accelerate + unroll

ggml : use vDSP_sve and vDSP_maxv from Accelerate

ggml : make gcc happy (minor)

talk.wasm : bump memory usage + update whisper.js

cmake : support AVX2 in Windows better (#381)

Revert "tmp"

go : bindings updated so they can be used in third party packages. (#379)

tmp

cmake : change min required version to 3.0 (#351)

command : add "guided-mode" video demo in the README.md

ggml : add SSE3 and fp16 conversion lookup table (#368)

whisper : document POWER VSX support

ggml : reorganize POWER9 ppc64le SIMD code

ggml : change f16 load and store macro arguments

whisper : do not use F16 tensors when in F32 mode (#369)

ci : add emscripten build

main : make whisper_print_segment_callback() more readable (close #371)

cmake : add CMAKE_RUNTIME_OUTPUT_DIRECTORY

ggml : add void to argument-less functions

minor : close #370 + Makefile build info print change

cmake : update to 3.19 (#351)

whisper : use ranged-based for loops for readability

ggml : define MIN / MAX only if not defined (minor)

make : print build information

stream : fix data race on bool + avoid division-by-zero

models : fix HF model URL (close #356)

command : refactor to split command list & general transcription modes (#331)

ggml : improve f16 acceleration for POWER9 ppc64le

whisper : add whisper_n_audio_ctx and check for invalid audio_ctx

models : fix typo in convert-h5-to-ggml.py

cmake : add runtime destination install (#345)

whisper : avoid some memory allocations

main : add -ocsv, aka --output-csv to output a CSV file

make : add i686 arch (close #329)

ggml : barrier refactor + static functions

ggml : simplify the SIMD code (#324)

examples : fix memory leak on failure to load gpt2 model (#323)

examples : small code cleanups (#322)

ggml : use vaddvq_f32 for slightly more efficient reduce

ggml : add f16 acceleration for POWER9 ppc64le

models : fix support for spaces in path (close #315)

whisper : use emplace_back in place of push_back (#319)