Commit Graph

407 Commits (1512545149e9463c0b478cd0203638c501b0ac29)
 

Author SHA1 Message Date
Syahmi Azhar 1512545149
whisper : add loader class to allow loading from buffer and others (#353)
1 year ago
Georgi Gerganov 52a3e0c92a
ggml : improve vec_dot_f16 unrolling in flash_attn_f16
1 year ago
Georgi Gerganov d1ea1220ff
command : clean-up / refactoring / formatting (#383)
1 year ago
David 9c4a1522f6
command : always-prompt mode (#383)
1 year ago
David Thorpe f078a6f20e
go : adding features to the go-whisper example, go ci, etc (#384)
1 year ago
Georgi Gerganov f30b5d322c
ggml : fix bug in new soft max computation
1 year ago
Georgi Gerganov 44efbf7ff1
cmake : add -Wno-unused-function + update whisper.js
1 year ago
Georgi Gerganov d347a59a5f
ggml : when using BLAS start only 1 CPU thread
1 year ago
Georgi Gerganov 6394c906af
ggml : fix running tasks with variable number of threads
1 year ago
Georgi Gerganov 74ffa14e1d
ggml : unroll ggml_vec_dot_f16 in ggml_compute_forward_flash_attn_f16
1 year ago
Georgi Gerganov 65fdcbbbbb
whisper : revert accidental MB change
1 year ago
Georgi Gerganov d61d55cd4b
ggml : speed-up soft max via Accelerate + unroll
1 year ago
Georgi Gerganov d51fc3ee0a
ggml : use vDSP_sve and vDSP_maxv from Accelerate
1 year ago
Georgi Gerganov f82a7dd019
ggml : make gcc happy (minor)
1 year ago
Georgi Gerganov 87dd4a3081
talk.wasm : bump memory usage + update whisper.js
1 year ago
m.bell 41e05c6b1b
cmake : support AVX2 in Windows better (#381)
1 year ago
Georgi Gerganov fa379cb22a
Revert "tmp"
1 year ago
David Thorpe 322f4e6c4e
go : bindings updated so they can be used in third party packages. (#379)
1 year ago
Georgi Gerganov 1652965529
tmp
1 year ago
Georgi Gerganov 6042c7a3be
cmake : change min required version to 3.0 (#351)
1 year ago
Georgi Gerganov 6b351bb669
command : add "guided-mode" video demo in the README.md
1 year ago
Abitofevrything a62170c656
ggml : add SSE3 and fp16 conversion lookup table (#368)
1 year ago
Thomas Fitzsimmons 1944e7c33e whisper : document POWER VSX support
1 year ago
Thomas Fitzsimmons 49a8dd6732 ggml : reorganize POWER9 ppc64le SIMD code
1 year ago
Thomas Fitzsimmons 8c7f642286 ggml : change f16 load and store macro arguments
1 year ago
Georgi Gerganov ad2a4ffa03
whisper : do not use F16 tensors when in F32 mode (#369)
1 year ago
Georgi Gerganov b3c865083e
ci : add emscripten build
1 year ago
Georgi Gerganov a0d4f8e65c
main : make whisper_print_segment_callback() more readable (close #371)
1 year ago
Georgi Gerganov 4a214d2f07
cmake : add CMAKE_RUNTIME_OUTPUT_DIRECTORY
1 year ago
Georgi Gerganov 0a0cfa7985
ggml : add void to argument-less functions
1 year ago
Georgi Gerganov 196d738974
minor : close #370 + Makefile build info print change
1 year ago
Andy Maloney 84c6b42e65
cmake : update to 3.19 (#351)
1 year ago
Andy Maloney dd6d582977 whisper : use ranged-based for loops for readability
1 year ago
Georgi Gerganov d51c5eb906
ggml : define MIN / MAX only if not defined (minor)
1 year ago
Georgi Gerganov 0be6a1afd9
make : print build information
1 year ago
Georgi Gerganov a466c3404d
stream : fix data race on bool + avoid division-by-zero
1 year ago
Georgi Gerganov d629c034a4
models : fix HF model URL (close #356)
1 year ago
Andy Maloney f00509d57c
command : refactor to split command list & general transcription modes (#331)
1 year ago
Thomas Fitzsimmons 424c410c42 ggml : improve f16 acceleration for POWER9 ppc64le
1 year ago
Georgi Gerganov d97e6005e9
whisper : add whisper_n_audio_ctx and check for invalid audio_ctx
1 year ago
Ikko Ashimine 3467230a77 models : fix typo in convert-h5-to-ggml.py
1 year ago
Avik Sengupta a091581eb3
cmake : add runtime destination install (#345)
1 year ago
Georgi Gerganov 68daf6e487
whisper : avoid some memory allocations
1 year ago
Niels Mayer a593b932e4
main : add -ocsv, aka --output-csv to output a CSV file
1 year ago
Georgi Gerganov 9a8ad3db69
make : add i686 arch (close #329)
1 year ago
Georgi Gerganov 4e0b2069e7
ggml : barrier refactor + static functions
1 year ago
Georgi Gerganov ac521a566e
ggml : simplify the SIMD code (#324)
1 year ago
Andy Maloney 331c0bbddc
examples : fix memory leak on failure to load gpt2 model (#323)
1 year ago
Andy Maloney dc90efd504
examples : small code cleanups (#322)
1 year ago
Georgi Gerganov 7282e2109e
ggml : use vaddvq_f32 for slightly more efficient reduce
1 year ago