whispercpp 1.3.6 → 1.3.7
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/.document +3 -0
- data/.rdoc_options +2 -0
- data/README.md +38 -5
- data/Rakefile +18 -3
- data/ext/dependencies.rb +10 -4
- data/ext/dependencies_for_windows.rb +17 -0
- data/ext/extconf.rb +20 -8
- data/ext/options.rb +54 -14
- data/ext/options_for_windows.rb +51 -0
- data/ext/ruby_whisper.c +36 -42
- data/ext/ruby_whisper.h +135 -0
- data/ext/ruby_whisper_context.c +107 -28
- data/ext/ruby_whisper_log_queue.c +180 -0
- data/ext/ruby_whisper_log_settable.h +47 -0
- data/ext/ruby_whisper_parakeet.c +49 -0
- data/ext/ruby_whisper_parakeet_context.c +304 -0
- data/ext/ruby_whisper_parakeet_context_params.c +117 -0
- data/ext/ruby_whisper_parakeet_model.c +84 -0
- data/ext/ruby_whisper_parakeet_params.c +548 -0
- data/ext/ruby_whisper_parakeet_segment.c +157 -0
- data/ext/ruby_whisper_parakeet_token.c +188 -0
- data/ext/ruby_whisper_parakeet_transcribe.cpp +58 -0
- data/ext/ruby_whisper_params.c +256 -65
- data/ext/ruby_whisper_segment.c +6 -6
- data/ext/ruby_whisper_transcribe.cpp +42 -15
- data/ext/sources/CMakeLists.txt +41 -3
- data/ext/sources/CMakePresets.json +95 -0
- data/ext/sources/cmake/parakeet-config.cmake.in +30 -0
- data/ext/sources/cmake/parakeet.pc.in +10 -0
- data/ext/sources/cmake/whisper.pc.in +1 -1
- data/ext/sources/examples/CMakeLists.txt +4 -2
- data/ext/sources/examples/bench/bench.cpp +1 -1
- data/ext/sources/examples/cli/cli.cpp +43 -9
- data/ext/sources/examples/common-ggml.cpp +2 -0
- data/ext/sources/examples/common-whisper.cpp +139 -67
- data/ext/sources/examples/common-whisper.h +11 -0
- data/ext/sources/examples/ffmpeg-transcode.cpp +211 -341
- data/ext/sources/examples/parakeet-cli/CMakeLists.txt +8 -0
- data/ext/sources/examples/parakeet-cli/parakeet-cli.cpp +243 -0
- data/ext/sources/examples/parakeet-quantize/CMakeLists.txt +7 -0
- data/ext/sources/examples/parakeet-quantize/parakeet-quantize.cpp +230 -0
- data/ext/sources/examples/server/server.cpp +199 -163
- data/ext/sources/ggml/CMakeLists.txt +21 -13
- data/ext/sources/ggml/cmake/FindNCCL.cmake +36 -0
- data/ext/sources/ggml/cmake/ggml-config.cmake.in +12 -2
- data/ext/sources/ggml/include/ggml-alloc.h +1 -0
- data/ext/sources/ggml/include/ggml-backend.h +72 -10
- data/ext/sources/ggml/include/ggml-cuda.h +3 -0
- data/ext/sources/ggml/include/ggml-rpc.h +3 -3
- data/ext/sources/ggml/include/ggml.h +101 -9
- data/ext/sources/ggml/include/gguf.h +10 -2
- data/ext/sources/ggml/src/CMakeLists.txt +22 -5
- data/ext/sources/ggml/src/ggml-alloc.c +5 -1
- data/ext/sources/ggml/src/ggml-backend-impl.h +22 -2
- data/ext/sources/ggml/src/ggml-backend-meta.cpp +2263 -0
- data/ext/sources/ggml/src/ggml-backend-reg.cpp +12 -0
- data/ext/sources/ggml/src/ggml-backend.cpp +110 -9
- data/ext/sources/ggml/src/ggml-blas/ggml-blas.cpp +4 -0
- data/ext/sources/ggml/src/ggml-cann/aclnn_ops.cpp +672 -257
- data/ext/sources/ggml/src/ggml-cann/aclnn_ops.h +71 -0
- data/ext/sources/ggml/src/ggml-cann/common.h +20 -10
- data/ext/sources/ggml/src/ggml-cann/ggml-cann.cpp +211 -30
- data/ext/sources/ggml/src/ggml-common.h +11 -0
- data/ext/sources/ggml/src/ggml-cpu/CMakeLists.txt +58 -29
- data/ext/sources/ggml/src/ggml-cpu/amx/amx.cpp +2 -0
- data/ext/sources/ggml/src/ggml-cpu/amx/mmq.cpp +16 -16
- data/ext/sources/ggml/src/ggml-cpu/arch/arm/quants.c +116 -7
- data/ext/sources/ggml/src/ggml-cpu/arch/arm/repack.cpp +65 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/loongarch/quants.c +151 -1
- data/ext/sources/ggml/src/ggml-cpu/arch/powerpc/quants.c +0 -1
- data/ext/sources/ggml/src/ggml-cpu/arch/riscv/quants.c +4279 -1292
- data/ext/sources/ggml/src/ggml-cpu/arch/riscv/repack.cpp +5 -35
- data/ext/sources/ggml/src/ggml-cpu/arch/s390/quants.c +0 -1
- data/ext/sources/ggml/src/ggml-cpu/arch/wasm/quants.c +72 -1
- data/ext/sources/ggml/src/ggml-cpu/arch/x86/quants.c +177 -27
- data/ext/sources/ggml/src/ggml-cpu/arch/x86/repack.cpp +1 -1
- data/ext/sources/ggml/src/ggml-cpu/arch-fallback.h +5 -0
- data/ext/sources/ggml/src/ggml-cpu/cmake/FindSMTIME.cmake +32 -0
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu-impl.h +10 -0
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu.c +95 -5
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu.cpp +2 -0
- data/ext/sources/ggml/src/ggml-cpu/kleidiai/kleidiai.cpp +146 -134
- data/ext/sources/ggml/src/ggml-cpu/llamafile/sgemm.cpp +88 -70
- data/ext/sources/ggml/src/ggml-cpu/ops.cpp +372 -73
- data/ext/sources/ggml/src/ggml-cpu/ops.h +3 -0
- data/ext/sources/ggml/src/ggml-cpu/quants.c +55 -0
- data/ext/sources/ggml/src/ggml-cpu/quants.h +3 -0
- data/ext/sources/ggml/src/ggml-cpu/repack.cpp +3 -0
- data/ext/sources/ggml/src/ggml-cpu/simd-gemm.h +90 -0
- data/ext/sources/ggml/src/ggml-cpu/simd-mappings.h +3 -16
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime.cpp +1402 -687
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime.h +8 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime1_kernels.cpp +597 -2766
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime2_kernels.cpp +5768 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime_env.cpp +320 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime_env.h +55 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime_kernels.h +182 -19
- data/ext/sources/ggml/src/ggml-cpu/spacemit/repack.cpp +1795 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/repack.h +14 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/rvv_kernels.cpp +3178 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/rvv_kernels.h +95 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/spine_barrier.h +34 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/spine_mem_pool.cpp +760 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/spine_mem_pool.h +32 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/spine_tcm.h +409 -0
- data/ext/sources/ggml/src/ggml-cpu/vec.cpp +37 -53
- data/ext/sources/ggml/src/ggml-cpu/vec.h +225 -240
- data/ext/sources/ggml/src/ggml-cuda/CMakeLists.txt +17 -7
- data/ext/sources/ggml/src/ggml-cuda/allreduce.cu +971 -0
- data/ext/sources/ggml/src/ggml-cuda/allreduce.cuh +29 -0
- data/ext/sources/ggml/src/ggml-cuda/argsort.cu +62 -26
- data/ext/sources/ggml/src/ggml-cuda/binbcast.cu +44 -18
- data/ext/sources/ggml/src/ggml-cuda/binbcast.cuh +1 -0
- data/ext/sources/ggml/src/ggml-cuda/common.cuh +242 -28
- data/ext/sources/ggml/src/ggml-cuda/concat.cu +120 -114
- data/ext/sources/ggml/src/ggml-cuda/conv2d-transpose.cu +45 -21
- data/ext/sources/ggml/src/ggml-cuda/conv2d-transpose.cuh +1 -0
- data/ext/sources/ggml/src/ggml-cuda/convert.cu +53 -0
- data/ext/sources/ggml/src/ggml-cuda/convert.cuh +10 -0
- data/ext/sources/ggml/src/ggml-cuda/cpy.cu +14 -6
- data/ext/sources/ggml/src/ggml-cuda/dequantize.cuh +22 -0
- data/ext/sources/ggml/src/ggml-cuda/fattn-common.cuh +278 -44
- data/ext/sources/ggml/src/ggml-cuda/fattn-mma-f16.cuh +331 -130
- data/ext/sources/ggml/src/ggml-cuda/fattn-tile.cu +12 -0
- data/ext/sources/ggml/src/ggml-cuda/fattn-tile.cuh +126 -27
- data/ext/sources/ggml/src/ggml-cuda/fattn-vec.cuh +40 -15
- data/ext/sources/ggml/src/ggml-cuda/fattn-wmma-f16.cu +18 -9
- data/ext/sources/ggml/src/ggml-cuda/fattn.cu +152 -49
- data/ext/sources/ggml/src/ggml-cuda/fattn.cuh +2 -0
- data/ext/sources/ggml/src/ggml-cuda/fwht.cu +101 -0
- data/ext/sources/ggml/src/ggml-cuda/fwht.cuh +4 -0
- data/ext/sources/ggml/src/ggml-cuda/gated_delta_net.cu +84 -35
- data/ext/sources/ggml/src/ggml-cuda/getrows.cu +34 -12
- data/ext/sources/ggml/src/ggml-cuda/ggml-cuda.cu +1069 -609
- data/ext/sources/ggml/src/ggml-cuda/im2col.cu +32 -29
- data/ext/sources/ggml/src/ggml-cuda/mean.cu +4 -2
- data/ext/sources/ggml/src/ggml-cuda/mma.cuh +242 -195
- data/ext/sources/ggml/src/ggml-cuda/mmf.cuh +3 -3
- data/ext/sources/ggml/src/ggml-cuda/mmq.cu +18 -12
- data/ext/sources/ggml/src/ggml-cuda/mmq.cuh +502 -423
- data/ext/sources/ggml/src/ggml-cuda/mmvf.cu +19 -12
- data/ext/sources/ggml/src/ggml-cuda/mmvq.cu +485 -57
- data/ext/sources/ggml/src/ggml-cuda/mmvq.cuh +6 -1
- data/ext/sources/ggml/src/ggml-cuda/norm.cu +36 -10
- data/ext/sources/ggml/src/ggml-cuda/out-prod.cu +23 -7
- data/ext/sources/ggml/src/ggml-cuda/quantize.cu +133 -26
- data/ext/sources/ggml/src/ggml-cuda/quantize.cuh +1 -1
- data/ext/sources/ggml/src/ggml-cuda/reduce_rows.cuh +5 -1
- data/ext/sources/ggml/src/ggml-cuda/rope.cu +11 -4
- data/ext/sources/ggml/src/ggml-cuda/scale.cu +4 -1
- data/ext/sources/ggml/src/ggml-cuda/set-rows.cu +14 -6
- data/ext/sources/ggml/src/ggml-cuda/snake.cu +72 -0
- data/ext/sources/ggml/src/ggml-cuda/snake.cuh +8 -0
- data/ext/sources/ggml/src/ggml-cuda/softcap.cu +4 -1
- data/ext/sources/ggml/src/ggml-cuda/ssm-conv.cu +45 -13
- data/ext/sources/ggml/src/ggml-cuda/ssm-conv.cuh +1 -1
- data/ext/sources/ggml/src/ggml-cuda/ssm-scan.cu +40 -18
- data/ext/sources/ggml/src/ggml-cuda/sumrows.cu +8 -4
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_1-ncols2_16.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_1-ncols2_32.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_1-ncols2_8.cu +2 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_16-ncols2_4.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_2-ncols2_16.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_2-ncols2_32.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_2-ncols2_4.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_2-ncols2_8.cu +2 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_4-ncols2_16.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_4-ncols2_4.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_4-ncols2_8.cu +2 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_8-ncols2_4.cu +1 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_8-ncols2_8.cu +2 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq192-dv128.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq320-dv256.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq512-dv512.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-bf16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-f16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q4_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q4_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q5_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q5_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q8_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-bf16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-bf16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-bf16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-bf16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-bf16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-bf16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-nvfp4.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q1_0.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/top-k.cu +5 -4
- data/ext/sources/ggml/src/ggml-cuda/topk-moe.cu +26 -23
- data/ext/sources/ggml/src/ggml-cuda/unary.cu +31 -2
- data/ext/sources/ggml/src/ggml-cuda/unary.cuh +2 -0
- data/ext/sources/ggml/src/ggml-cuda/vecdotq.cuh +80 -0
- data/ext/sources/ggml/src/ggml-cuda/vendors/cuda.h +7 -2
- data/ext/sources/ggml/src/ggml-cuda/vendors/hip.h +22 -4
- data/ext/sources/ggml/src/ggml-cuda/vendors/musa.h +3 -0
- data/ext/sources/ggml/src/ggml-hexagon/CMakeLists.txt +2 -1
- data/ext/sources/ggml/src/ggml-hexagon/ggml-hexagon.cpp +1428 -743
- data/ext/sources/ggml/src/ggml-hexagon/htp/CMakeLists.txt +45 -7
- data/ext/sources/ggml/src/ggml-hexagon/htp/act-ops.c +53 -84
- data/ext/sources/ggml/src/ggml-hexagon/htp/argsort-ops.c +25 -12
- data/ext/sources/ggml/src/ggml-hexagon/htp/binary-ops.c +165 -184
- data/ext/sources/ggml/src/ggml-hexagon/htp/cmake-toolchain.cmake +5 -5
- data/ext/sources/ggml/src/ggml-hexagon/htp/concat-ops.c +277 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/cpy-ops.c +170 -127
- data/ext/sources/ggml/src/ggml-hexagon/htp/cumsum-ops.c +270 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/diag-ops.c +216 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/fill-ops.c +123 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/flash-attn-ops.c +125 -97
- data/ext/sources/ggml/src/ggml-hexagon/htp/gated-delta-net-ops.c +1148 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/get-rows-ops.c +148 -42
- data/ext/sources/ggml/src/ggml-hexagon/htp/hex-dma.c +2 -2
- data/ext/sources/ggml/src/ggml-hexagon/htp/hex-dma.h +252 -62
- data/ext/sources/ggml/src/ggml-hexagon/htp/hex-dump.h +9 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hex-utils.h +87 -1
- data/ext/sources/ggml/src/ggml-hexagon/htp/hmx-flash-attn-ops.c +1878 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hmx-matmul-ops.c +2066 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hmx-ops.c +6 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hmx-ops.h +88 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hmx-profile.h +34 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hmx-queue.c +158 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hmx-queue.h +134 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hmx-utils.h +200 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/htp-ctx.h +96 -13
- data/ext/sources/ggml/src/ggml-hexagon/htp/htp-ops.h +182 -57
- data/ext/sources/ggml/src/ggml-hexagon/htp/htp_iface.idl +9 -3
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-base.h +71 -3
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-copy.h +27 -10
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-div.h +63 -23
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-exp.h +9 -8
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-flash-attn.h +47 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-log.h +65 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-pow.h +42 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-repl.h +74 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-sigmoid.h +1 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-sin-cos.h +90 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/hvx-utils.h +5 -8
- data/ext/sources/ggml/src/ggml-hexagon/htp/main.c +529 -815
- data/ext/sources/ggml/src/ggml-hexagon/htp/matmul-ops.c +2522 -234
- data/ext/sources/ggml/src/ggml-hexagon/htp/pad-ops.c +547 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/repeat-ops.c +148 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/rope-ops.c +291 -95
- data/ext/sources/ggml/src/ggml-hexagon/htp/set-rows-ops.c +59 -37
- data/ext/sources/ggml/src/ggml-hexagon/htp/softmax-ops.c +121 -133
- data/ext/sources/ggml/src/ggml-hexagon/htp/solve-tri-ops.c +267 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp/ssm-conv.c +244 -151
- data/ext/sources/ggml/src/ggml-hexagon/htp/sum-rows-ops.c +6 -6
- data/ext/sources/ggml/src/ggml-hexagon/htp/unary-ops.c +719 -45
- data/ext/sources/ggml/src/ggml-hexagon/htp/vtcm-utils.h +16 -0
- data/ext/sources/ggml/src/ggml-hexagon/htp-opnode.h +272 -0
- data/ext/sources/ggml/src/ggml-hexagon/libggml-htp.inf +3 -1
- data/ext/sources/ggml/src/ggml-hip/CMakeLists.txt +22 -9
- data/ext/sources/ggml/src/ggml-impl.h +6 -1
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-device.cpp +138 -13
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-device.h +32 -1
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-device.m +164 -28
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-impl.h +80 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-ops.cpp +190 -19
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-ops.h +2 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal.cpp +39 -26
- data/ext/sources/ggml/src/ggml-metal/ggml-metal.metal +823 -322
- data/ext/sources/ggml/src/ggml-musa/CMakeLists.txt +5 -6
- data/ext/sources/ggml/src/ggml-opencl/CMakeLists.txt +54 -5
- data/ext/sources/ggml/src/ggml-opencl/ggml-opencl.cpp +12248 -5907
- data/ext/sources/ggml/src/ggml-opencl/kernels/concat.cl +67 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/cpy.cl +59 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/cvt.cl +1819 -112
- data/ext/sources/ggml/src/ggml-opencl/kernels/gated_delta_net.cl +249 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_mxfp4_f32_ns.cl +306 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q4_0_f32_ns.cl +256 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q4_1_f32_ns.cl +258 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q4_k_f32_ns.cl +283 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q5_0_f32_ns.cl +260 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q5_1_f32_ns.cl +262 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q5_k_f32_ns.cl +288 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q6_k_f32_ns.cl +267 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_iq4_nl_f32.cl +150 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/{mul_mat_Ab_Bi_8x4.cl → gemm_noshuffle_q4_0_f32.cl} +1 -1
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q4_k_f32.cl +172 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q5_0_f32.cl +131 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q5_1_f32.cl +134 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q5_k_f32.cl +176 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q6_k_f32.cl +140 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/{mul_mm_q8_0_f32_8x4.cl → gemm_noshuffle_q8_0_f32.cl} +1 -1
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemm_xmem_f16_f32_os8.cl +233 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_mxfp4_f32_ns.cl +165 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q4_0_f32_ns.cl +120 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q4_1_f32_ns.cl +123 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q4_k_f32_ns.cl +155 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q5_0_f32_ns.cl +123 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q5_1_f32_ns.cl +125 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q5_k_f32_ns.cl +160 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q6_k_f32_ns.cl +141 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_iq4_nl_f32.cl +302 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/{gemv_noshuffle_general.cl → gemv_noshuffle_q4_0_f32.cl} +5 -5
- data/ext/sources/ggml/src/ggml-opencl/kernels/{gemv_noshuffle.cl → gemv_noshuffle_q4_0_f32_spec.cl} +5 -5
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q4_k_f32.cl +318 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q5_0_f32.cl +291 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q5_1_f32.cl +294 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q5_k_f32.cl +326 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q6_k_f32.cl +293 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/get_rows.cl +15 -9
- data/ext/sources/ggml/src/ggml-opencl/kernels/moe_reorder_b.cl +30 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/moe_sort_by_expert.cl +82 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_iq4_nl_f32_l4_lm.cl +171 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q4_k_f32_l4_lm.cl +179 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q5_0_f32_l4_lm.cl +173 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q5_1_f32_l4_lm.cl +175 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q5_k_f32_l4_lm.cl +192 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_iq4_nl_f32.cl +164 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_iq4_nl_f32_flat.cl +202 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_k_f32_flat.cl +196 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_0_f32.cl +241 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_0_f32_flat.cl +243 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_1_f32.cl +243 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_1_f32_flat.cl +247 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_k_f32.cl +187 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_k_f32_flat.cl +203 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q6_k_f32_flat.cl +48 -64
- data/ext/sources/ggml/src/ggml-openvino/ggml-decoder.cpp +15 -5
- data/ext/sources/ggml/src/ggml-openvino/ggml-openvino-extra.cpp +18 -11
- data/ext/sources/ggml/src/ggml-openvino/ggml-openvino.cpp +35 -13
- data/ext/sources/ggml/src/ggml-openvino/ggml-quants.cpp +264 -192
- data/ext/sources/ggml/src/ggml-openvino/openvino/op/rope.cpp +33 -7
- data/ext/sources/ggml/src/ggml-openvino/openvino/op/unary_gelu.cpp +25 -0
- data/ext/sources/ggml/src/ggml-openvino/openvino/op_table.cpp +1 -0
- data/ext/sources/ggml/src/ggml-openvino/openvino/op_table.h +1 -0
- data/ext/sources/ggml/src/ggml-openvino/openvino/rt_info/weightless_caching_attributes.hpp +41 -0
- data/ext/sources/ggml/src/ggml-openvino/openvino/translate_session.cpp +27 -3
- data/ext/sources/ggml/src/ggml-openvino/openvino/utils.cpp +67 -36
- data/ext/sources/ggml/src/ggml-openvino/openvino/utils.h +1 -0
- data/ext/sources/ggml/src/ggml-openvino/utils.cpp +101 -44
- data/ext/sources/ggml/src/ggml-openvino/utils.h +23 -3
- data/ext/sources/ggml/src/ggml-opt.cpp +1 -0
- data/ext/sources/ggml/src/ggml-quants.c +289 -114
- data/ext/sources/ggml/src/ggml-quants.h +3 -0
- data/ext/sources/ggml/src/ggml-rpc/CMakeLists.txt +24 -0
- data/ext/sources/ggml/src/ggml-rpc/ggml-rpc.cpp +167 -311
- data/ext/sources/ggml/src/ggml-rpc/transport.cpp +683 -0
- data/ext/sources/ggml/src/ggml-rpc/transport.h +34 -0
- data/ext/sources/ggml/src/ggml-sycl/CMakeLists.txt +50 -4
- data/ext/sources/ggml/src/ggml-sycl/add-id.cpp +1 -1
- data/ext/sources/ggml/src/ggml-sycl/backend.hpp +3 -1
- data/ext/sources/ggml/src/ggml-sycl/common.cpp +74 -2
- data/ext/sources/ggml/src/ggml-sycl/common.hpp +41 -1
- data/ext/sources/ggml/src/ggml-sycl/convert.cpp +115 -13
- data/ext/sources/ggml/src/ggml-sycl/convert.hpp +9 -0
- data/ext/sources/ggml/src/ggml-sycl/cumsum.cpp +148 -0
- data/ext/sources/ggml/src/ggml-sycl/cumsum.hpp +5 -0
- data/ext/sources/ggml/src/ggml-sycl/dequantize.hpp +663 -0
- data/ext/sources/ggml/src/ggml-sycl/diag.cpp +67 -0
- data/ext/sources/ggml/src/ggml-sycl/diag.hpp +5 -0
- data/ext/sources/ggml/src/ggml-sycl/dmmv.cpp +586 -6
- data/ext/sources/ggml/src/ggml-sycl/element_wise.cpp +1 -90
- data/ext/sources/ggml/src/ggml-sycl/element_wise.hpp +0 -2
- data/ext/sources/ggml/src/ggml-sycl/fattn-buffers.cpp +56 -0
- data/ext/sources/ggml/src/ggml-sycl/fattn-buffers.hpp +63 -0
- data/ext/sources/ggml/src/ggml-sycl/fattn-common.hpp +7 -5
- data/ext/sources/ggml/src/ggml-sycl/fattn-tile.cpp +4 -0
- data/ext/sources/ggml/src/ggml-sycl/fattn-tile.hpp +76 -168
- data/ext/sources/ggml/src/ggml-sycl/fattn-vec.hpp +7 -0
- data/ext/sources/ggml/src/ggml-sycl/fattn.cpp +3 -1
- data/ext/sources/ggml/src/ggml-sycl/fill.cpp +55 -0
- data/ext/sources/ggml/src/ggml-sycl/fill.hpp +5 -0
- data/ext/sources/ggml/src/ggml-sycl/gated_delta_net.cpp +69 -31
- data/ext/sources/ggml/src/ggml-sycl/gated_delta_net.hpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/gemm.hpp +3 -0
- data/ext/sources/ggml/src/ggml-sycl/getrows.cpp +79 -3
- data/ext/sources/ggml/src/ggml-sycl/ggml-sycl.cpp +823 -190
- data/ext/sources/ggml/src/ggml-sycl/im2col.cpp +353 -89
- data/ext/sources/ggml/src/ggml-sycl/im2col.hpp +5 -3
- data/ext/sources/ggml/src/ggml-sycl/mmvq.cpp +1344 -26
- data/ext/sources/ggml/src/ggml-sycl/mmvq.hpp +16 -0
- data/ext/sources/ggml/src/ggml-sycl/pad.cpp +27 -27
- data/ext/sources/ggml/src/ggml-sycl/quants.hpp +71 -0
- data/ext/sources/ggml/src/ggml-sycl/set_rows.cpp +7 -1
- data/ext/sources/ggml/src/ggml-sycl/solve_tri.cpp +172 -0
- data/ext/sources/ggml/src/ggml-sycl/solve_tri.hpp +8 -0
- data/ext/sources/ggml/src/ggml-sycl/ssm_conv.cpp +6 -1
- data/ext/sources/ggml/src/ggml-sycl/ssm_scan.cpp +156 -0
- data/ext/sources/ggml/src/ggml-sycl/ssm_scan.hpp +5 -0
- data/ext/sources/ggml/src/ggml-sycl/sycl_hw.cpp +62 -10
- data/ext/sources/ggml/src/ggml-sycl/sycl_hw.hpp +18 -6
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq512-dv512.cpp +6 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-f16-f16.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-f16-q4_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-f16-q4_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-f16-q5_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-f16-q5_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-f16-q8_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_0-f16.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_0-q4_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_0-q4_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_0-q5_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_0-q5_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_0-q8_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_1-f16.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_1-q4_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_1-q4_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_1-q5_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_1-q5_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q4_1-q8_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_0-f16.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_0-q4_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_0-q4_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_0-q5_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_0-q5_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_0-q8_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_1-f16.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_1-q4_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_1-q4_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_1-q5_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_1-q5_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q5_1-q8_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q8_0-f16.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q8_0-q4_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q8_0-q4_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q8_0-q5_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q8_0-q5_1.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q8_0-q8_0.cpp +1 -0
- data/ext/sources/ggml/src/ggml-sycl/type.hpp +112 -0
- data/ext/sources/ggml/src/ggml-sycl/upscale.cpp +410 -0
- data/ext/sources/ggml/src/ggml-sycl/upscale.hpp +9 -0
- data/ext/sources/ggml/src/ggml-sycl/vecdotq.hpp +215 -53
- data/ext/sources/ggml/src/ggml-virtgpu/ggml-backend-buffer.cpp +4 -0
- data/ext/sources/ggml/src/ggml-virtgpu/ggml-backend-device.cpp +2 -0
- data/ext/sources/ggml/src/ggml-virtgpu/ggml-backend.cpp +2 -0
- data/ext/sources/ggml/src/ggml-virtgpu/virtgpu-shm.cpp +1 -0
- data/ext/sources/ggml/src/ggml-virtgpu/virtgpu.cpp +1 -0
- data/ext/sources/ggml/src/ggml-virtgpu/virtgpu.h +0 -2
- data/ext/sources/ggml/src/ggml-vulkan/CMakeLists.txt +11 -0
- data/ext/sources/ggml/src/ggml-vulkan/ggml-vulkan.cpp +2060 -535
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/CMakeLists.txt +4 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/contig_copy.comp +6 -2
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/conv2d_mm.comp +146 -13
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/copy.comp +3 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/copy_from_quant.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/copy_to_quant.comp +25 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_funcs.glsl +88 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_funcs_cm2.glsl +643 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_nvfp4.comp +32 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q1_0.comp +29 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/diag.comp +0 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dot_product_funcs.glsl +27 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/exp.comp +0 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/feature-tests/coopmat2_decode_vector.comp +7 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn.comp +197 -48
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_base.glsl +60 -59
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp +115 -113
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm2.comp +122 -31
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_dequant.glsl +131 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_mmq_funcs.glsl +203 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/fwht.comp +115 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/gated_delta_net.comp +125 -64
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/generic_binary_head.glsl +0 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/glu_head.glsl +10 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/glu_main.glsl +16 -6
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/im2col.comp +76 -54
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/im2col_3d.comp +0 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/log.comp +0 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec.comp +122 -27
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_iface.glsl +6 -6
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q2_k.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q4_k.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q5_k.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vecq.comp +1 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vecq_funcs.glsl +88 -55
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm.comp +11 -17
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_cm2.comp +43 -10
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_funcs.glsl +159 -125
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq_funcs.glsl +8 -8
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq_shmem_types.glsl +24 -9
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/multi_add.comp +0 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_funcs.glsl +5 -2
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_head.glsl +0 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_params.glsl +3 -2
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/snake.comp +49 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/ssm_conv.comp +11 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/tri.comp +0 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/types.glsl +79 -2
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp +171 -147
- data/ext/sources/ggml/src/ggml-webgpu/CMakeLists.txt +5 -2
- data/ext/sources/ggml/src/ggml-webgpu/ggml-webgpu-shader-lib.hpp +2202 -283
- data/ext/sources/ggml/src/ggml-webgpu/ggml-webgpu.cpp +2610 -1403
- data/ext/sources/ggml/src/ggml-webgpu/pre_wgsl.hpp +37 -7
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/add_id.wgsl +64 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/binary.wgsl +8 -7
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/common_decls.tmpl +76 -95
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/concat.wgsl +19 -1
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/conv2d.wgsl +165 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/{cpy.tmpl.wgsl → cpy.wgsl} +25 -50
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn.wgsl +107 -184
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_quant_staging.tmpl +124 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_tile.wgsl +397 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_vec_blk.wgsl +101 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_vec_reduce.wgsl +84 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_vec_split.wgsl +619 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/gated_delta_net.wgsl +149 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/get_rows.wgsl +183 -78
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/glu.wgsl +155 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/im2col.wgsl +101 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_decls.tmpl +655 -495
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_id.wgsl +195 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_id_gather.wgsl +52 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_id_vec.wgsl +154 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_reg_tile.wgsl +8 -6
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_subgroup_matrix.wgsl +5 -1
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_vec.wgsl +80 -409
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_vec_acc.tmpl +1432 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_vec_q_acc.tmpl +303 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/quant_inner_loops.tmpl +21 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/quantize_q8.wgsl +173 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rms_norm_mul.wgsl +152 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/{rope.tmpl.wgsl → rope.wgsl} +71 -142
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/row_norm.wgsl +153 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/scale.wgsl +6 -4
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/set.wgsl +109 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/set_rows.wgsl +2 -3
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/set_rows_quant.wgsl +224 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/{soft_max.tmpl.wgsl → soft_max.wgsl} +106 -206
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/solve_tri.wgsl +121 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/ssm_conv.wgsl +65 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/ssm_scan.wgsl +193 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/unary.wgsl +68 -48
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/upscale.wgsl +240 -0
- data/ext/sources/ggml/src/ggml-zdnn/ggml-zdnn.cpp +18 -14
- data/ext/sources/ggml/src/ggml-zendnn/CMakeLists.txt +1 -1
- data/ext/sources/ggml/src/ggml-zendnn/ggml-zendnn.cpp +244 -10
- data/ext/sources/ggml/src/ggml.c +110 -28
- data/ext/sources/ggml/src/gguf.cpp +173 -28
- data/ext/sources/include/parakeet.h +342 -0
- data/ext/sources/include/whisper.h +10 -0
- data/ext/sources/media/matmul.png +0 -0
- data/ext/sources/src/CMakeLists.txt +23 -0
- data/ext/sources/src/parakeet-arch.h +188 -0
- data/ext/sources/src/parakeet.cpp +3838 -0
- data/ext/sources/src/whisper.cpp +56 -12
- data/extsources.rb +26 -10
- data/lib/whisper/log_settable.rb +36 -0
- data/lib/whisper/model/uri.rb +13 -1
- data/lib/whisper/output.rb +74 -0
- data/sig/whisper.rbs +411 -62
- data/test/helper.rb +2 -0
- data/test/jfk_reader/jfk_reader.c +50 -7
- data/test/test_callback.rb +1 -0
- data/test/test_package.rb +6 -5
- data/test/test_parakeet.rb +28 -0
- data/test/test_parakeet_callback.rb +107 -0
- data/test/test_parakeet_context.rb +116 -0
- data/test/test_parakeet_context_params.rb +24 -0
- data/test/test_parakeet_model.rb +21 -0
- data/test/test_parakeet_params.rb +78 -0
- data/test/test_parakeet_segment.rb +42 -0
- data/test/test_parakeet_token.rb +73 -0
- data/test/test_params.rb +2 -0
- data/test/test_vad_segment.rb +1 -1
- data/test/test_whisper.rb +24 -6
- data/whispercpp.gemspec +2 -2
- metadata +215 -281
- data/ext/sources/bindings/javascript/CMakeLists.txt +0 -41
- data/ext/sources/bindings/javascript/emscripten.cpp +0 -93
- data/ext/sources/bindings/javascript/libwhisper.worker.js +0 -1
- data/ext/sources/bindings/javascript/package.json +0 -26
- data/ext/sources/bindings/javascript/whisper.js +0 -19
- data/ext/sources/examples/addon.node/CMakeLists.txt +0 -31
- data/ext/sources/examples/addon.node/__test__/whisper.spec.js +0 -133
- data/ext/sources/examples/addon.node/addon.cpp +0 -557
- data/ext/sources/examples/addon.node/index.js +0 -59
- data/ext/sources/examples/addon.node/package.json +0 -16
- data/ext/sources/examples/addon.node/vad-example.js +0 -132
- data/ext/sources/examples/bench.wasm/CMakeLists.txt +0 -49
- data/ext/sources/examples/bench.wasm/emscripten.cpp +0 -87
- data/ext/sources/examples/bench.wasm/index-tmpl.html +0 -285
- data/ext/sources/examples/coi-serviceworker.js +0 -146
- data/ext/sources/examples/command/CMakeLists.txt +0 -10
- data/ext/sources/examples/command/command.cpp +0 -802
- data/ext/sources/examples/command/commands.txt +0 -9
- data/ext/sources/examples/command.wasm/CMakeLists.txt +0 -50
- data/ext/sources/examples/command.wasm/emscripten.cpp +0 -327
- data/ext/sources/examples/command.wasm/index-tmpl.html +0 -415
- data/ext/sources/examples/generate-karaoke.sh +0 -57
- data/ext/sources/examples/helpers.js +0 -191
- data/ext/sources/examples/livestream.sh +0 -112
- data/ext/sources/examples/lsp/CMakeLists.txt +0 -10
- data/ext/sources/examples/lsp/lsp.cpp +0 -471
- data/ext/sources/examples/lsp/whisper.vim +0 -362
- data/ext/sources/examples/python/test_whisper_processor.py +0 -7
- data/ext/sources/examples/python/whisper_processor.py +0 -54
- data/ext/sources/examples/server/bench.js +0 -29
- data/ext/sources/examples/server.py +0 -120
- data/ext/sources/examples/stream/CMakeLists.txt +0 -10
- data/ext/sources/examples/stream/stream.cpp +0 -437
- data/ext/sources/examples/stream.wasm/CMakeLists.txt +0 -49
- data/ext/sources/examples/stream.wasm/emscripten.cpp +0 -216
- data/ext/sources/examples/stream.wasm/index-tmpl.html +0 -491
- data/ext/sources/examples/sycl/CMakeLists.txt +0 -9
- data/ext/sources/examples/sycl/build.sh +0 -22
- data/ext/sources/examples/sycl/ls-sycl-device.cpp +0 -11
- data/ext/sources/examples/sycl/run-whisper.sh +0 -17
- data/ext/sources/examples/talk-llama/CMakeLists.txt +0 -48
- data/ext/sources/examples/talk-llama/eleven-labs.py +0 -80
- data/ext/sources/examples/talk-llama/llama-adapter.cpp +0 -488
- data/ext/sources/examples/talk-llama/llama-adapter.h +0 -89
- data/ext/sources/examples/talk-llama/llama-arch.cpp +0 -2877
- data/ext/sources/examples/talk-llama/llama-arch.h +0 -628
- data/ext/sources/examples/talk-llama/llama-batch.cpp +0 -919
- data/ext/sources/examples/talk-llama/llama-batch.h +0 -173
- data/ext/sources/examples/talk-llama/llama-chat.cpp +0 -896
- data/ext/sources/examples/talk-llama/llama-chat.h +0 -71
- data/ext/sources/examples/talk-llama/llama-context.cpp +0 -3633
- data/ext/sources/examples/talk-llama/llama-context.h +0 -359
- data/ext/sources/examples/talk-llama/llama-cparams.cpp +0 -5
- data/ext/sources/examples/talk-llama/llama-cparams.h +0 -47
- data/ext/sources/examples/talk-llama/llama-ext.h +0 -12
- data/ext/sources/examples/talk-llama/llama-grammar.cpp +0 -1464
- data/ext/sources/examples/talk-llama/llama-grammar.h +0 -194
- data/ext/sources/examples/talk-llama/llama-graph.cpp +0 -2735
- data/ext/sources/examples/talk-llama/llama-graph.h +0 -1031
- data/ext/sources/examples/talk-llama/llama-hparams.cpp +0 -258
- data/ext/sources/examples/talk-llama/llama-hparams.h +0 -353
- data/ext/sources/examples/talk-llama/llama-impl.cpp +0 -171
- data/ext/sources/examples/talk-llama/llama-impl.h +0 -75
- data/ext/sources/examples/talk-llama/llama-io.cpp +0 -15
- data/ext/sources/examples/talk-llama/llama-io.h +0 -35
- data/ext/sources/examples/talk-llama/llama-kv-cache-iswa.cpp +0 -330
- data/ext/sources/examples/talk-llama/llama-kv-cache-iswa.h +0 -137
- data/ext/sources/examples/talk-llama/llama-kv-cache.cpp +0 -2285
- data/ext/sources/examples/talk-llama/llama-kv-cache.h +0 -389
- data/ext/sources/examples/talk-llama/llama-kv-cells.h +0 -533
- data/ext/sources/examples/talk-llama/llama-memory-hybrid-iswa.cpp +0 -275
- data/ext/sources/examples/talk-llama/llama-memory-hybrid-iswa.h +0 -140
- data/ext/sources/examples/talk-llama/llama-memory-hybrid.cpp +0 -268
- data/ext/sources/examples/talk-llama/llama-memory-hybrid.h +0 -139
- data/ext/sources/examples/talk-llama/llama-memory-recurrent.cpp +0 -1165
- data/ext/sources/examples/talk-llama/llama-memory-recurrent.h +0 -182
- data/ext/sources/examples/talk-llama/llama-memory.cpp +0 -59
- data/ext/sources/examples/talk-llama/llama-memory.h +0 -122
- data/ext/sources/examples/talk-llama/llama-mmap.cpp +0 -752
- data/ext/sources/examples/talk-llama/llama-mmap.h +0 -73
- data/ext/sources/examples/talk-llama/llama-model-loader.cpp +0 -1655
- data/ext/sources/examples/talk-llama/llama-model-loader.h +0 -206
- data/ext/sources/examples/talk-llama/llama-model-saver.cpp +0 -299
- data/ext/sources/examples/talk-llama/llama-model-saver.h +0 -40
- data/ext/sources/examples/talk-llama/llama-model.cpp +0 -9056
- data/ext/sources/examples/talk-llama/llama-model.h +0 -597
- data/ext/sources/examples/talk-llama/llama-quant.cpp +0 -1304
- data/ext/sources/examples/talk-llama/llama-quant.h +0 -1
- data/ext/sources/examples/talk-llama/llama-sampler.cpp +0 -3885
- data/ext/sources/examples/talk-llama/llama-sampler.h +0 -42
- data/ext/sources/examples/talk-llama/llama-vocab.cpp +0 -3970
- data/ext/sources/examples/talk-llama/llama-vocab.h +0 -187
- data/ext/sources/examples/talk-llama/llama.cpp +0 -1194
- data/ext/sources/examples/talk-llama/llama.h +0 -1573
- data/ext/sources/examples/talk-llama/models/afmoe.cpp +0 -190
- data/ext/sources/examples/talk-llama/models/apertus.cpp +0 -125
- data/ext/sources/examples/talk-llama/models/arcee.cpp +0 -135
- data/ext/sources/examples/talk-llama/models/arctic.cpp +0 -137
- data/ext/sources/examples/talk-llama/models/arwkv7.cpp +0 -86
- data/ext/sources/examples/talk-llama/models/baichuan.cpp +0 -123
- data/ext/sources/examples/talk-llama/models/bailingmoe.cpp +0 -143
- data/ext/sources/examples/talk-llama/models/bailingmoe2.cpp +0 -133
- data/ext/sources/examples/talk-llama/models/bert.cpp +0 -184
- data/ext/sources/examples/talk-llama/models/bitnet.cpp +0 -145
- data/ext/sources/examples/talk-llama/models/bloom.cpp +0 -101
- data/ext/sources/examples/talk-llama/models/chameleon.cpp +0 -178
- data/ext/sources/examples/talk-llama/models/chatglm.cpp +0 -132
- data/ext/sources/examples/talk-llama/models/codeshell.cpp +0 -111
- data/ext/sources/examples/talk-llama/models/cogvlm.cpp +0 -102
- data/ext/sources/examples/talk-llama/models/cohere2-iswa.cpp +0 -134
- data/ext/sources/examples/talk-llama/models/command-r.cpp +0 -122
- data/ext/sources/examples/talk-llama/models/dbrx.cpp +0 -122
- data/ext/sources/examples/talk-llama/models/deci.cpp +0 -135
- data/ext/sources/examples/talk-llama/models/deepseek.cpp +0 -142
- data/ext/sources/examples/talk-llama/models/deepseek2.cpp +0 -262
- data/ext/sources/examples/talk-llama/models/delta-net-base.cpp +0 -445
- data/ext/sources/examples/talk-llama/models/dots1.cpp +0 -132
- data/ext/sources/examples/talk-llama/models/dream.cpp +0 -105
- data/ext/sources/examples/talk-llama/models/ernie4-5-moe.cpp +0 -148
- data/ext/sources/examples/talk-llama/models/ernie4-5.cpp +0 -110
- data/ext/sources/examples/talk-llama/models/eurobert.cpp +0 -97
- data/ext/sources/examples/talk-llama/models/exaone-moe.cpp +0 -145
- data/ext/sources/examples/talk-llama/models/exaone.cpp +0 -114
- data/ext/sources/examples/talk-llama/models/exaone4.cpp +0 -123
- data/ext/sources/examples/talk-llama/models/falcon-h1.cpp +0 -111
- data/ext/sources/examples/talk-llama/models/falcon.cpp +0 -120
- data/ext/sources/examples/talk-llama/models/gemma-embedding.cpp +0 -116
- data/ext/sources/examples/talk-llama/models/gemma.cpp +0 -112
- data/ext/sources/examples/talk-llama/models/gemma2-iswa.cpp +0 -128
- data/ext/sources/examples/talk-llama/models/gemma3.cpp +0 -155
- data/ext/sources/examples/talk-llama/models/gemma3n-iswa.cpp +0 -384
- data/ext/sources/examples/talk-llama/models/glm4-moe.cpp +0 -170
- data/ext/sources/examples/talk-llama/models/glm4.cpp +0 -157
- data/ext/sources/examples/talk-llama/models/gpt2.cpp +0 -105
- data/ext/sources/examples/talk-llama/models/gptneox.cpp +0 -144
- data/ext/sources/examples/talk-llama/models/granite-hybrid.cpp +0 -195
- data/ext/sources/examples/talk-llama/models/granite.cpp +0 -210
- data/ext/sources/examples/talk-llama/models/grok.cpp +0 -159
- data/ext/sources/examples/talk-llama/models/grovemoe.cpp +0 -139
- data/ext/sources/examples/talk-llama/models/hunyuan-dense.cpp +0 -132
- data/ext/sources/examples/talk-llama/models/hunyuan-moe.cpp +0 -153
- data/ext/sources/examples/talk-llama/models/internlm2.cpp +0 -120
- data/ext/sources/examples/talk-llama/models/jais.cpp +0 -86
- data/ext/sources/examples/talk-llama/models/jais2.cpp +0 -123
- data/ext/sources/examples/talk-llama/models/jamba.cpp +0 -106
- data/ext/sources/examples/talk-llama/models/kimi-linear.cpp +0 -381
- data/ext/sources/examples/talk-llama/models/lfm2.cpp +0 -196
- data/ext/sources/examples/talk-llama/models/llada-moe.cpp +0 -122
- data/ext/sources/examples/talk-llama/models/llada.cpp +0 -99
- data/ext/sources/examples/talk-llama/models/llama-iswa.cpp +0 -178
- data/ext/sources/examples/talk-llama/models/llama.cpp +0 -175
- data/ext/sources/examples/talk-llama/models/maincoder.cpp +0 -117
- data/ext/sources/examples/talk-llama/models/mamba-base.cpp +0 -289
- data/ext/sources/examples/talk-llama/models/mamba.cpp +0 -54
- data/ext/sources/examples/talk-llama/models/mimo2-iswa.cpp +0 -129
- data/ext/sources/examples/talk-llama/models/minicpm3.cpp +0 -200
- data/ext/sources/examples/talk-llama/models/minimax-m2.cpp +0 -123
- data/ext/sources/examples/talk-llama/models/mistral3.cpp +0 -160
- data/ext/sources/examples/talk-llama/models/models.h +0 -704
- data/ext/sources/examples/talk-llama/models/modern-bert.cpp +0 -109
- data/ext/sources/examples/talk-llama/models/mpt.cpp +0 -126
- data/ext/sources/examples/talk-llama/models/nemotron-h.cpp +0 -162
- data/ext/sources/examples/talk-llama/models/nemotron.cpp +0 -122
- data/ext/sources/examples/talk-llama/models/neo-bert.cpp +0 -104
- data/ext/sources/examples/talk-llama/models/olmo.cpp +0 -121
- data/ext/sources/examples/talk-llama/models/olmo2.cpp +0 -150
- data/ext/sources/examples/talk-llama/models/olmoe.cpp +0 -124
- data/ext/sources/examples/talk-llama/models/openai-moe-iswa.cpp +0 -127
- data/ext/sources/examples/talk-llama/models/openelm.cpp +0 -124
- data/ext/sources/examples/talk-llama/models/orion.cpp +0 -123
- data/ext/sources/examples/talk-llama/models/paddleocr.cpp +0 -122
- data/ext/sources/examples/talk-llama/models/pangu-embedded.cpp +0 -121
- data/ext/sources/examples/talk-llama/models/phi2.cpp +0 -121
- data/ext/sources/examples/talk-llama/models/phi3.cpp +0 -152
- data/ext/sources/examples/talk-llama/models/plamo.cpp +0 -110
- data/ext/sources/examples/talk-llama/models/plamo2.cpp +0 -320
- data/ext/sources/examples/talk-llama/models/plamo3.cpp +0 -128
- data/ext/sources/examples/talk-llama/models/plm.cpp +0 -169
- data/ext/sources/examples/talk-llama/models/qwen.cpp +0 -108
- data/ext/sources/examples/talk-llama/models/qwen2.cpp +0 -126
- data/ext/sources/examples/talk-llama/models/qwen2moe.cpp +0 -151
- data/ext/sources/examples/talk-llama/models/qwen2vl.cpp +0 -117
- data/ext/sources/examples/talk-llama/models/qwen3.cpp +0 -120
- data/ext/sources/examples/talk-llama/models/qwen35.cpp +0 -381
- data/ext/sources/examples/talk-llama/models/qwen35moe.cpp +0 -422
- data/ext/sources/examples/talk-llama/models/qwen3moe.cpp +0 -131
- data/ext/sources/examples/talk-llama/models/qwen3next.cpp +0 -525
- data/ext/sources/examples/talk-llama/models/qwen3vl-moe.cpp +0 -140
- data/ext/sources/examples/talk-llama/models/qwen3vl.cpp +0 -132
- data/ext/sources/examples/talk-llama/models/refact.cpp +0 -94
- data/ext/sources/examples/talk-llama/models/rnd1.cpp +0 -126
- data/ext/sources/examples/talk-llama/models/rwkv6-base.cpp +0 -164
- data/ext/sources/examples/talk-llama/models/rwkv6.cpp +0 -94
- data/ext/sources/examples/talk-llama/models/rwkv6qwen2.cpp +0 -86
- data/ext/sources/examples/talk-llama/models/rwkv7-base.cpp +0 -137
- data/ext/sources/examples/talk-llama/models/rwkv7.cpp +0 -90
- data/ext/sources/examples/talk-llama/models/seed-oss.cpp +0 -124
- data/ext/sources/examples/talk-llama/models/smallthinker.cpp +0 -126
- data/ext/sources/examples/talk-llama/models/smollm3.cpp +0 -128
- data/ext/sources/examples/talk-llama/models/stablelm.cpp +0 -146
- data/ext/sources/examples/talk-llama/models/starcoder.cpp +0 -100
- data/ext/sources/examples/talk-llama/models/starcoder2.cpp +0 -121
- data/ext/sources/examples/talk-llama/models/step35-iswa.cpp +0 -165
- data/ext/sources/examples/talk-llama/models/t5-dec.cpp +0 -166
- data/ext/sources/examples/talk-llama/models/t5-enc.cpp +0 -96
- data/ext/sources/examples/talk-llama/models/wavtokenizer-dec.cpp +0 -149
- data/ext/sources/examples/talk-llama/models/xverse.cpp +0 -108
- data/ext/sources/examples/talk-llama/prompts/talk-alpaca.txt +0 -23
- data/ext/sources/examples/talk-llama/speak +0 -40
- data/ext/sources/examples/talk-llama/speak.bat +0 -1
- data/ext/sources/examples/talk-llama/speak.ps1 +0 -14
- data/ext/sources/examples/talk-llama/talk-llama.cpp +0 -813
- data/ext/sources/examples/talk-llama/unicode-data.cpp +0 -7034
- data/ext/sources/examples/talk-llama/unicode-data.h +0 -20
- data/ext/sources/examples/talk-llama/unicode.cpp +0 -1103
- data/ext/sources/examples/talk-llama/unicode.h +0 -111
- data/ext/sources/examples/wchess/CMakeLists.txt +0 -10
- data/ext/sources/examples/wchess/libwchess/CMakeLists.txt +0 -19
- data/ext/sources/examples/wchess/libwchess/Chessboard.cpp +0 -803
- data/ext/sources/examples/wchess/libwchess/Chessboard.h +0 -33
- data/ext/sources/examples/wchess/libwchess/WChess.cpp +0 -193
- data/ext/sources/examples/wchess/libwchess/WChess.h +0 -63
- data/ext/sources/examples/wchess/libwchess/test-chessboard.cpp +0 -117
- data/ext/sources/examples/wchess/wchess.cmd/CMakeLists.txt +0 -8
- data/ext/sources/examples/wchess/wchess.cmd/wchess.cmd.cpp +0 -253
- data/ext/sources/examples/whisper.wasm/CMakeLists.txt +0 -50
- data/ext/sources/examples/whisper.wasm/emscripten.cpp +0 -118
- data/ext/sources/examples/whisper.wasm/index-tmpl.html +0 -659
- data/ext/sources/ggml/src/ggml-cuda/template-instances/generate_cu_files.py +0 -99
- data/ext/sources/ggml/src/ggml-hexagon/htp/htp-msg.h +0 -155
- data/ext/sources/ggml/src/ggml-hexagon/op-desc.h +0 -153
- data/ext/sources/ggml/src/ggml-opencl/kernels/embed_kernel.py +0 -26
- data/ext/sources/ggml/src/ggml-openvino/openvino/pass/eliminate_zp.cpp +0 -123
- data/ext/sources/ggml/src/ggml-openvino/openvino/pass/eliminate_zp.h +0 -17
- data/ext/sources/ggml/src/ggml-virtgpu/regenerate_remoting.py +0 -333
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rte.glsl +0 -5
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/embed_wgsl.py +0 -182
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/glu.tmpl.wgsl +0 -323
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat.wgsl +0 -718
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rms_norm.wgsl +0 -123
- data/ext/sources/tests/CMakeLists.txt +0 -112
- data/ext/sources/tests/earnings21/eval.mk +0 -58
- data/ext/sources/tests/earnings21/eval.py +0 -68
- data/ext/sources/tests/earnings21/normalizers/__init__.py +0 -2
- data/ext/sources/tests/earnings21/normalizers/basic.py +0 -80
- data/ext/sources/tests/earnings21/normalizers/english.json +0 -1741
- data/ext/sources/tests/earnings21/normalizers/english.py +0 -550
- data/ext/sources/tests/earnings21/requirements.txt +0 -6
- data/ext/sources/tests/en-0-ref.txt +0 -1
- data/ext/sources/tests/en-1-ref.txt +0 -1
- data/ext/sources/tests/en-2-ref.txt +0 -1
- data/ext/sources/tests/es-0-ref.txt +0 -1
- data/ext/sources/tests/librispeech/eval.mk +0 -39
- data/ext/sources/tests/librispeech/eval.py +0 -47
- data/ext/sources/tests/librispeech/normalizers/__init__.py +0 -2
- data/ext/sources/tests/librispeech/normalizers/basic.py +0 -80
- data/ext/sources/tests/librispeech/normalizers/english.json +0 -1741
- data/ext/sources/tests/librispeech/normalizers/english.py +0 -550
- data/ext/sources/tests/librispeech/requirements.txt +0 -6
- data/ext/sources/tests/run-tests.sh +0 -130
- data/ext/sources/tests/test-c.c +0 -3
- data/ext/sources/tests/test-vad-full.cpp +0 -56
- data/ext/sources/tests/test-vad.cpp +0 -83
- data/ext/sources/tests/test-whisper.js +0 -58
- data/lib/whisper/context.rb +0 -15
- data/lib/whisper/segment.rb +0 -58
- /data/ext/sources/ggml/src/ggml-opencl/kernels/{gemv_noshuffle_general_q8_0_f32.cl → gemv_noshuffle_q8_0_f32.cl} +0 -0
metadata
CHANGED
|
@@ -1,15 +1,14 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: whispercpp
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 1.3.
|
|
4
|
+
version: 1.3.7
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Georgi Gerganov
|
|
8
8
|
- Todd A. Fisher
|
|
9
|
-
autorequire:
|
|
10
9
|
bindir: bin
|
|
11
10
|
cert_chain: []
|
|
12
|
-
date:
|
|
11
|
+
date: 1980-01-02 00:00:00.000000000 Z
|
|
13
12
|
dependencies: []
|
|
14
13
|
description: High-performance inference of OpenAI's Whisper automatic speech recognition
|
|
15
14
|
(ASR) model via Ruby
|
|
@@ -21,20 +20,34 @@ extra_rdoc_files:
|
|
|
21
20
|
- LICENSE
|
|
22
21
|
- README.md
|
|
23
22
|
files:
|
|
23
|
+
- ".document"
|
|
24
24
|
- ".gitignore"
|
|
25
|
+
- ".rdoc_options"
|
|
25
26
|
- LICENSE
|
|
26
27
|
- README.md
|
|
27
28
|
- Rakefile
|
|
28
29
|
- ext/.gitignore
|
|
29
30
|
- ext/dependencies.rb
|
|
31
|
+
- ext/dependencies_for_windows.rb
|
|
30
32
|
- ext/extconf.rb
|
|
31
33
|
- ext/options.rb
|
|
34
|
+
- ext/options_for_windows.rb
|
|
32
35
|
- ext/ruby_whisper.c
|
|
33
36
|
- ext/ruby_whisper.h
|
|
34
37
|
- ext/ruby_whisper_context.c
|
|
35
38
|
- ext/ruby_whisper_context_params.c
|
|
36
39
|
- ext/ruby_whisper_error.c
|
|
40
|
+
- ext/ruby_whisper_log_queue.c
|
|
41
|
+
- ext/ruby_whisper_log_settable.h
|
|
37
42
|
- ext/ruby_whisper_model.c
|
|
43
|
+
- ext/ruby_whisper_parakeet.c
|
|
44
|
+
- ext/ruby_whisper_parakeet_context.c
|
|
45
|
+
- ext/ruby_whisper_parakeet_context_params.c
|
|
46
|
+
- ext/ruby_whisper_parakeet_model.c
|
|
47
|
+
- ext/ruby_whisper_parakeet_params.c
|
|
48
|
+
- ext/ruby_whisper_parakeet_segment.c
|
|
49
|
+
- ext/ruby_whisper_parakeet_token.c
|
|
50
|
+
- ext/ruby_whisper_parakeet_transcribe.cpp
|
|
38
51
|
- ext/ruby_whisper_params.c
|
|
39
52
|
- ext/ruby_whisper_segment.c
|
|
40
53
|
- ext/ruby_whisper_token.c
|
|
@@ -46,43 +59,25 @@ files:
|
|
|
46
59
|
- ext/ruby_whisper_vad_segments.c
|
|
47
60
|
- ext/sources/CMakeGraphVizOptions.cmake
|
|
48
61
|
- ext/sources/CMakeLists.txt
|
|
49
|
-
- ext/sources/
|
|
50
|
-
- ext/sources/bindings/javascript/emscripten.cpp
|
|
51
|
-
- ext/sources/bindings/javascript/libwhisper.worker.js
|
|
62
|
+
- ext/sources/CMakePresets.json
|
|
52
63
|
- ext/sources/bindings/javascript/package-tmpl.json
|
|
53
|
-
- ext/sources/bindings/javascript/package.json
|
|
54
|
-
- ext/sources/bindings/javascript/whisper.js
|
|
55
64
|
- ext/sources/cmake/DefaultTargetOptions.cmake
|
|
56
65
|
- ext/sources/cmake/FindFFmpeg.cmake
|
|
57
66
|
- ext/sources/cmake/arm64-apple-clang.cmake
|
|
58
67
|
- ext/sources/cmake/arm64-windows-llvm.cmake
|
|
59
68
|
- ext/sources/cmake/build-info.cmake
|
|
60
69
|
- ext/sources/cmake/git-vars.cmake
|
|
70
|
+
- ext/sources/cmake/parakeet-config.cmake.in
|
|
71
|
+
- ext/sources/cmake/parakeet.pc.in
|
|
61
72
|
- ext/sources/cmake/riscv64-spacemit-linux-gnu-gcc.cmake
|
|
62
73
|
- ext/sources/cmake/whisper-config.cmake.in
|
|
63
74
|
- ext/sources/cmake/whisper.pc.in
|
|
64
75
|
- ext/sources/cmake/x64-windows-llvm.cmake
|
|
65
76
|
- ext/sources/examples/CMakeLists.txt
|
|
66
|
-
- ext/sources/examples/addon.node/CMakeLists.txt
|
|
67
|
-
- ext/sources/examples/addon.node/__test__/whisper.spec.js
|
|
68
|
-
- ext/sources/examples/addon.node/addon.cpp
|
|
69
|
-
- ext/sources/examples/addon.node/index.js
|
|
70
|
-
- ext/sources/examples/addon.node/package.json
|
|
71
|
-
- ext/sources/examples/addon.node/vad-example.js
|
|
72
|
-
- ext/sources/examples/bench.wasm/CMakeLists.txt
|
|
73
|
-
- ext/sources/examples/bench.wasm/emscripten.cpp
|
|
74
|
-
- ext/sources/examples/bench.wasm/index-tmpl.html
|
|
75
77
|
- ext/sources/examples/bench/CMakeLists.txt
|
|
76
78
|
- ext/sources/examples/bench/bench.cpp
|
|
77
79
|
- ext/sources/examples/cli/CMakeLists.txt
|
|
78
80
|
- ext/sources/examples/cli/cli.cpp
|
|
79
|
-
- ext/sources/examples/coi-serviceworker.js
|
|
80
|
-
- ext/sources/examples/command.wasm/CMakeLists.txt
|
|
81
|
-
- ext/sources/examples/command.wasm/emscripten.cpp
|
|
82
|
-
- ext/sources/examples/command.wasm/index-tmpl.html
|
|
83
|
-
- ext/sources/examples/command/CMakeLists.txt
|
|
84
|
-
- ext/sources/examples/command/command.cpp
|
|
85
|
-
- ext/sources/examples/command/commands.txt
|
|
86
81
|
- ext/sources/examples/common-ggml.cpp
|
|
87
82
|
- ext/sources/examples/common-ggml.h
|
|
88
83
|
- ext/sources/examples/common-sdl.cpp
|
|
@@ -94,226 +89,24 @@ files:
|
|
|
94
89
|
- ext/sources/examples/deprecation-warning/CMakeLists.txt
|
|
95
90
|
- ext/sources/examples/deprecation-warning/deprecation-warning.cpp
|
|
96
91
|
- ext/sources/examples/ffmpeg-transcode.cpp
|
|
97
|
-
- ext/sources/examples/generate-karaoke.sh
|
|
98
92
|
- ext/sources/examples/grammar-parser.cpp
|
|
99
93
|
- ext/sources/examples/grammar-parser.h
|
|
100
|
-
- ext/sources/examples/helpers.js
|
|
101
94
|
- ext/sources/examples/json.hpp
|
|
102
|
-
- ext/sources/examples/livestream.sh
|
|
103
|
-
- ext/sources/examples/lsp/CMakeLists.txt
|
|
104
|
-
- ext/sources/examples/lsp/lsp.cpp
|
|
105
|
-
- ext/sources/examples/lsp/whisper.vim
|
|
106
95
|
- ext/sources/examples/miniaudio.h
|
|
107
|
-
- ext/sources/examples/
|
|
108
|
-
- ext/sources/examples/
|
|
96
|
+
- ext/sources/examples/parakeet-cli/CMakeLists.txt
|
|
97
|
+
- ext/sources/examples/parakeet-cli/parakeet-cli.cpp
|
|
98
|
+
- ext/sources/examples/parakeet-quantize/CMakeLists.txt
|
|
99
|
+
- ext/sources/examples/parakeet-quantize/parakeet-quantize.cpp
|
|
109
100
|
- ext/sources/examples/quantize/CMakeLists.txt
|
|
110
101
|
- ext/sources/examples/quantize/quantize.cpp
|
|
111
|
-
- ext/sources/examples/server.py
|
|
112
102
|
- ext/sources/examples/server/CMakeLists.txt
|
|
113
|
-
- ext/sources/examples/server/bench.js
|
|
114
103
|
- ext/sources/examples/server/httplib.h
|
|
115
104
|
- ext/sources/examples/server/server.cpp
|
|
116
105
|
- ext/sources/examples/stb_vorbis.c
|
|
117
|
-
- ext/sources/examples/stream.wasm/CMakeLists.txt
|
|
118
|
-
- ext/sources/examples/stream.wasm/emscripten.cpp
|
|
119
|
-
- ext/sources/examples/stream.wasm/index-tmpl.html
|
|
120
|
-
- ext/sources/examples/stream/CMakeLists.txt
|
|
121
|
-
- ext/sources/examples/stream/stream.cpp
|
|
122
|
-
- ext/sources/examples/sycl/CMakeLists.txt
|
|
123
|
-
- ext/sources/examples/sycl/build.sh
|
|
124
|
-
- ext/sources/examples/sycl/ls-sycl-device.cpp
|
|
125
|
-
- ext/sources/examples/sycl/run-whisper.sh
|
|
126
|
-
- ext/sources/examples/talk-llama/CMakeLists.txt
|
|
127
|
-
- ext/sources/examples/talk-llama/eleven-labs.py
|
|
128
|
-
- ext/sources/examples/talk-llama/llama-adapter.cpp
|
|
129
|
-
- ext/sources/examples/talk-llama/llama-adapter.h
|
|
130
|
-
- ext/sources/examples/talk-llama/llama-arch.cpp
|
|
131
|
-
- ext/sources/examples/talk-llama/llama-arch.h
|
|
132
|
-
- ext/sources/examples/talk-llama/llama-batch.cpp
|
|
133
|
-
- ext/sources/examples/talk-llama/llama-batch.h
|
|
134
|
-
- ext/sources/examples/talk-llama/llama-chat.cpp
|
|
135
|
-
- ext/sources/examples/talk-llama/llama-chat.h
|
|
136
|
-
- ext/sources/examples/talk-llama/llama-context.cpp
|
|
137
|
-
- ext/sources/examples/talk-llama/llama-context.h
|
|
138
|
-
- ext/sources/examples/talk-llama/llama-cparams.cpp
|
|
139
|
-
- ext/sources/examples/talk-llama/llama-cparams.h
|
|
140
|
-
- ext/sources/examples/talk-llama/llama-ext.h
|
|
141
|
-
- ext/sources/examples/talk-llama/llama-grammar.cpp
|
|
142
|
-
- ext/sources/examples/talk-llama/llama-grammar.h
|
|
143
|
-
- ext/sources/examples/talk-llama/llama-graph.cpp
|
|
144
|
-
- ext/sources/examples/talk-llama/llama-graph.h
|
|
145
|
-
- ext/sources/examples/talk-llama/llama-hparams.cpp
|
|
146
|
-
- ext/sources/examples/talk-llama/llama-hparams.h
|
|
147
|
-
- ext/sources/examples/talk-llama/llama-impl.cpp
|
|
148
|
-
- ext/sources/examples/talk-llama/llama-impl.h
|
|
149
|
-
- ext/sources/examples/talk-llama/llama-io.cpp
|
|
150
|
-
- ext/sources/examples/talk-llama/llama-io.h
|
|
151
|
-
- ext/sources/examples/talk-llama/llama-kv-cache-iswa.cpp
|
|
152
|
-
- ext/sources/examples/talk-llama/llama-kv-cache-iswa.h
|
|
153
|
-
- ext/sources/examples/talk-llama/llama-kv-cache.cpp
|
|
154
|
-
- ext/sources/examples/talk-llama/llama-kv-cache.h
|
|
155
|
-
- ext/sources/examples/talk-llama/llama-kv-cells.h
|
|
156
|
-
- ext/sources/examples/talk-llama/llama-memory-hybrid-iswa.cpp
|
|
157
|
-
- ext/sources/examples/talk-llama/llama-memory-hybrid-iswa.h
|
|
158
|
-
- ext/sources/examples/talk-llama/llama-memory-hybrid.cpp
|
|
159
|
-
- ext/sources/examples/talk-llama/llama-memory-hybrid.h
|
|
160
|
-
- ext/sources/examples/talk-llama/llama-memory-recurrent.cpp
|
|
161
|
-
- ext/sources/examples/talk-llama/llama-memory-recurrent.h
|
|
162
|
-
- ext/sources/examples/talk-llama/llama-memory.cpp
|
|
163
|
-
- ext/sources/examples/talk-llama/llama-memory.h
|
|
164
|
-
- ext/sources/examples/talk-llama/llama-mmap.cpp
|
|
165
|
-
- ext/sources/examples/talk-llama/llama-mmap.h
|
|
166
|
-
- ext/sources/examples/talk-llama/llama-model-loader.cpp
|
|
167
|
-
- ext/sources/examples/talk-llama/llama-model-loader.h
|
|
168
|
-
- ext/sources/examples/talk-llama/llama-model-saver.cpp
|
|
169
|
-
- ext/sources/examples/talk-llama/llama-model-saver.h
|
|
170
|
-
- ext/sources/examples/talk-llama/llama-model.cpp
|
|
171
|
-
- ext/sources/examples/talk-llama/llama-model.h
|
|
172
|
-
- ext/sources/examples/talk-llama/llama-quant.cpp
|
|
173
|
-
- ext/sources/examples/talk-llama/llama-quant.h
|
|
174
|
-
- ext/sources/examples/talk-llama/llama-sampler.cpp
|
|
175
|
-
- ext/sources/examples/talk-llama/llama-sampler.h
|
|
176
|
-
- ext/sources/examples/talk-llama/llama-vocab.cpp
|
|
177
|
-
- ext/sources/examples/talk-llama/llama-vocab.h
|
|
178
|
-
- ext/sources/examples/talk-llama/llama.cpp
|
|
179
|
-
- ext/sources/examples/talk-llama/llama.h
|
|
180
|
-
- ext/sources/examples/talk-llama/models/afmoe.cpp
|
|
181
|
-
- ext/sources/examples/talk-llama/models/apertus.cpp
|
|
182
|
-
- ext/sources/examples/talk-llama/models/arcee.cpp
|
|
183
|
-
- ext/sources/examples/talk-llama/models/arctic.cpp
|
|
184
|
-
- ext/sources/examples/talk-llama/models/arwkv7.cpp
|
|
185
|
-
- ext/sources/examples/talk-llama/models/baichuan.cpp
|
|
186
|
-
- ext/sources/examples/talk-llama/models/bailingmoe.cpp
|
|
187
|
-
- ext/sources/examples/talk-llama/models/bailingmoe2.cpp
|
|
188
|
-
- ext/sources/examples/talk-llama/models/bert.cpp
|
|
189
|
-
- ext/sources/examples/talk-llama/models/bitnet.cpp
|
|
190
|
-
- ext/sources/examples/talk-llama/models/bloom.cpp
|
|
191
|
-
- ext/sources/examples/talk-llama/models/chameleon.cpp
|
|
192
|
-
- ext/sources/examples/talk-llama/models/chatglm.cpp
|
|
193
|
-
- ext/sources/examples/talk-llama/models/codeshell.cpp
|
|
194
|
-
- ext/sources/examples/talk-llama/models/cogvlm.cpp
|
|
195
|
-
- ext/sources/examples/talk-llama/models/cohere2-iswa.cpp
|
|
196
|
-
- ext/sources/examples/talk-llama/models/command-r.cpp
|
|
197
|
-
- ext/sources/examples/talk-llama/models/dbrx.cpp
|
|
198
|
-
- ext/sources/examples/talk-llama/models/deci.cpp
|
|
199
|
-
- ext/sources/examples/talk-llama/models/deepseek.cpp
|
|
200
|
-
- ext/sources/examples/talk-llama/models/deepseek2.cpp
|
|
201
|
-
- ext/sources/examples/talk-llama/models/delta-net-base.cpp
|
|
202
|
-
- ext/sources/examples/talk-llama/models/dots1.cpp
|
|
203
|
-
- ext/sources/examples/talk-llama/models/dream.cpp
|
|
204
|
-
- ext/sources/examples/talk-llama/models/ernie4-5-moe.cpp
|
|
205
|
-
- ext/sources/examples/talk-llama/models/ernie4-5.cpp
|
|
206
|
-
- ext/sources/examples/talk-llama/models/eurobert.cpp
|
|
207
|
-
- ext/sources/examples/talk-llama/models/exaone-moe.cpp
|
|
208
|
-
- ext/sources/examples/talk-llama/models/exaone.cpp
|
|
209
|
-
- ext/sources/examples/talk-llama/models/exaone4.cpp
|
|
210
|
-
- ext/sources/examples/talk-llama/models/falcon-h1.cpp
|
|
211
|
-
- ext/sources/examples/talk-llama/models/falcon.cpp
|
|
212
|
-
- ext/sources/examples/talk-llama/models/gemma-embedding.cpp
|
|
213
|
-
- ext/sources/examples/talk-llama/models/gemma.cpp
|
|
214
|
-
- ext/sources/examples/talk-llama/models/gemma2-iswa.cpp
|
|
215
|
-
- ext/sources/examples/talk-llama/models/gemma3.cpp
|
|
216
|
-
- ext/sources/examples/talk-llama/models/gemma3n-iswa.cpp
|
|
217
|
-
- ext/sources/examples/talk-llama/models/glm4-moe.cpp
|
|
218
|
-
- ext/sources/examples/talk-llama/models/glm4.cpp
|
|
219
|
-
- ext/sources/examples/talk-llama/models/gpt2.cpp
|
|
220
|
-
- ext/sources/examples/talk-llama/models/gptneox.cpp
|
|
221
|
-
- ext/sources/examples/talk-llama/models/granite-hybrid.cpp
|
|
222
|
-
- ext/sources/examples/talk-llama/models/granite.cpp
|
|
223
|
-
- ext/sources/examples/talk-llama/models/grok.cpp
|
|
224
|
-
- ext/sources/examples/talk-llama/models/grovemoe.cpp
|
|
225
|
-
- ext/sources/examples/talk-llama/models/hunyuan-dense.cpp
|
|
226
|
-
- ext/sources/examples/talk-llama/models/hunyuan-moe.cpp
|
|
227
|
-
- ext/sources/examples/talk-llama/models/internlm2.cpp
|
|
228
|
-
- ext/sources/examples/talk-llama/models/jais.cpp
|
|
229
|
-
- ext/sources/examples/talk-llama/models/jais2.cpp
|
|
230
|
-
- ext/sources/examples/talk-llama/models/jamba.cpp
|
|
231
|
-
- ext/sources/examples/talk-llama/models/kimi-linear.cpp
|
|
232
|
-
- ext/sources/examples/talk-llama/models/lfm2.cpp
|
|
233
|
-
- ext/sources/examples/talk-llama/models/llada-moe.cpp
|
|
234
|
-
- ext/sources/examples/talk-llama/models/llada.cpp
|
|
235
|
-
- ext/sources/examples/talk-llama/models/llama-iswa.cpp
|
|
236
|
-
- ext/sources/examples/talk-llama/models/llama.cpp
|
|
237
|
-
- ext/sources/examples/talk-llama/models/maincoder.cpp
|
|
238
|
-
- ext/sources/examples/talk-llama/models/mamba-base.cpp
|
|
239
|
-
- ext/sources/examples/talk-llama/models/mamba.cpp
|
|
240
|
-
- ext/sources/examples/talk-llama/models/mimo2-iswa.cpp
|
|
241
|
-
- ext/sources/examples/talk-llama/models/minicpm3.cpp
|
|
242
|
-
- ext/sources/examples/talk-llama/models/minimax-m2.cpp
|
|
243
|
-
- ext/sources/examples/talk-llama/models/mistral3.cpp
|
|
244
|
-
- ext/sources/examples/talk-llama/models/models.h
|
|
245
|
-
- ext/sources/examples/talk-llama/models/modern-bert.cpp
|
|
246
|
-
- ext/sources/examples/talk-llama/models/mpt.cpp
|
|
247
|
-
- ext/sources/examples/talk-llama/models/nemotron-h.cpp
|
|
248
|
-
- ext/sources/examples/talk-llama/models/nemotron.cpp
|
|
249
|
-
- ext/sources/examples/talk-llama/models/neo-bert.cpp
|
|
250
|
-
- ext/sources/examples/talk-llama/models/olmo.cpp
|
|
251
|
-
- ext/sources/examples/talk-llama/models/olmo2.cpp
|
|
252
|
-
- ext/sources/examples/talk-llama/models/olmoe.cpp
|
|
253
|
-
- ext/sources/examples/talk-llama/models/openai-moe-iswa.cpp
|
|
254
|
-
- ext/sources/examples/talk-llama/models/openelm.cpp
|
|
255
|
-
- ext/sources/examples/talk-llama/models/orion.cpp
|
|
256
|
-
- ext/sources/examples/talk-llama/models/paddleocr.cpp
|
|
257
|
-
- ext/sources/examples/talk-llama/models/pangu-embedded.cpp
|
|
258
|
-
- ext/sources/examples/talk-llama/models/phi2.cpp
|
|
259
|
-
- ext/sources/examples/talk-llama/models/phi3.cpp
|
|
260
|
-
- ext/sources/examples/talk-llama/models/plamo.cpp
|
|
261
|
-
- ext/sources/examples/talk-llama/models/plamo2.cpp
|
|
262
|
-
- ext/sources/examples/talk-llama/models/plamo3.cpp
|
|
263
|
-
- ext/sources/examples/talk-llama/models/plm.cpp
|
|
264
|
-
- ext/sources/examples/talk-llama/models/qwen.cpp
|
|
265
|
-
- ext/sources/examples/talk-llama/models/qwen2.cpp
|
|
266
|
-
- ext/sources/examples/talk-llama/models/qwen2moe.cpp
|
|
267
|
-
- ext/sources/examples/talk-llama/models/qwen2vl.cpp
|
|
268
|
-
- ext/sources/examples/talk-llama/models/qwen3.cpp
|
|
269
|
-
- ext/sources/examples/talk-llama/models/qwen35.cpp
|
|
270
|
-
- ext/sources/examples/talk-llama/models/qwen35moe.cpp
|
|
271
|
-
- ext/sources/examples/talk-llama/models/qwen3moe.cpp
|
|
272
|
-
- ext/sources/examples/talk-llama/models/qwen3next.cpp
|
|
273
|
-
- ext/sources/examples/talk-llama/models/qwen3vl-moe.cpp
|
|
274
|
-
- ext/sources/examples/talk-llama/models/qwen3vl.cpp
|
|
275
|
-
- ext/sources/examples/talk-llama/models/refact.cpp
|
|
276
|
-
- ext/sources/examples/talk-llama/models/rnd1.cpp
|
|
277
|
-
- ext/sources/examples/talk-llama/models/rwkv6-base.cpp
|
|
278
|
-
- ext/sources/examples/talk-llama/models/rwkv6.cpp
|
|
279
|
-
- ext/sources/examples/talk-llama/models/rwkv6qwen2.cpp
|
|
280
|
-
- ext/sources/examples/talk-llama/models/rwkv7-base.cpp
|
|
281
|
-
- ext/sources/examples/talk-llama/models/rwkv7.cpp
|
|
282
|
-
- ext/sources/examples/talk-llama/models/seed-oss.cpp
|
|
283
|
-
- ext/sources/examples/talk-llama/models/smallthinker.cpp
|
|
284
|
-
- ext/sources/examples/talk-llama/models/smollm3.cpp
|
|
285
|
-
- ext/sources/examples/talk-llama/models/stablelm.cpp
|
|
286
|
-
- ext/sources/examples/talk-llama/models/starcoder.cpp
|
|
287
|
-
- ext/sources/examples/talk-llama/models/starcoder2.cpp
|
|
288
|
-
- ext/sources/examples/talk-llama/models/step35-iswa.cpp
|
|
289
|
-
- ext/sources/examples/talk-llama/models/t5-dec.cpp
|
|
290
|
-
- ext/sources/examples/talk-llama/models/t5-enc.cpp
|
|
291
|
-
- ext/sources/examples/talk-llama/models/wavtokenizer-dec.cpp
|
|
292
|
-
- ext/sources/examples/talk-llama/models/xverse.cpp
|
|
293
|
-
- ext/sources/examples/talk-llama/prompts/talk-alpaca.txt
|
|
294
|
-
- ext/sources/examples/talk-llama/speak
|
|
295
|
-
- ext/sources/examples/talk-llama/speak.bat
|
|
296
|
-
- ext/sources/examples/talk-llama/speak.ps1
|
|
297
|
-
- ext/sources/examples/talk-llama/talk-llama.cpp
|
|
298
|
-
- ext/sources/examples/talk-llama/unicode-data.cpp
|
|
299
|
-
- ext/sources/examples/talk-llama/unicode-data.h
|
|
300
|
-
- ext/sources/examples/talk-llama/unicode.cpp
|
|
301
|
-
- ext/sources/examples/talk-llama/unicode.h
|
|
302
106
|
- ext/sources/examples/vad-speech-segments/CMakeLists.txt
|
|
303
107
|
- ext/sources/examples/vad-speech-segments/speech.cpp
|
|
304
|
-
- ext/sources/examples/wchess/CMakeLists.txt
|
|
305
|
-
- ext/sources/examples/wchess/libwchess/CMakeLists.txt
|
|
306
|
-
- ext/sources/examples/wchess/libwchess/Chessboard.cpp
|
|
307
|
-
- ext/sources/examples/wchess/libwchess/Chessboard.h
|
|
308
|
-
- ext/sources/examples/wchess/libwchess/WChess.cpp
|
|
309
|
-
- ext/sources/examples/wchess/libwchess/WChess.h
|
|
310
|
-
- ext/sources/examples/wchess/libwchess/test-chessboard.cpp
|
|
311
|
-
- ext/sources/examples/wchess/wchess.cmd/CMakeLists.txt
|
|
312
|
-
- ext/sources/examples/wchess/wchess.cmd/wchess.cmd.cpp
|
|
313
|
-
- ext/sources/examples/whisper.wasm/CMakeLists.txt
|
|
314
|
-
- ext/sources/examples/whisper.wasm/emscripten.cpp
|
|
315
|
-
- ext/sources/examples/whisper.wasm/index-tmpl.html
|
|
316
108
|
- ext/sources/ggml/CMakeLists.txt
|
|
109
|
+
- ext/sources/ggml/cmake/FindNCCL.cmake
|
|
317
110
|
- ext/sources/ggml/cmake/GitVars.cmake
|
|
318
111
|
- ext/sources/ggml/cmake/common.cmake
|
|
319
112
|
- ext/sources/ggml/cmake/ggml-config.cmake.in
|
|
@@ -343,6 +136,7 @@ files:
|
|
|
343
136
|
- ext/sources/ggml/src/ggml-backend-dl.cpp
|
|
344
137
|
- ext/sources/ggml/src/ggml-backend-dl.h
|
|
345
138
|
- ext/sources/ggml/src/ggml-backend-impl.h
|
|
139
|
+
- ext/sources/ggml/src/ggml-backend-meta.cpp
|
|
346
140
|
- ext/sources/ggml/src/ggml-backend-reg.cpp
|
|
347
141
|
- ext/sources/ggml/src/ggml-backend.cpp
|
|
348
142
|
- ext/sources/ggml/src/ggml-blas/CMakeLists.txt
|
|
@@ -380,6 +174,7 @@ files:
|
|
|
380
174
|
- ext/sources/ggml/src/ggml-cpu/binary-ops.cpp
|
|
381
175
|
- ext/sources/ggml/src/ggml-cpu/binary-ops.h
|
|
382
176
|
- ext/sources/ggml/src/ggml-cpu/cmake/FindSIMD.cmake
|
|
177
|
+
- ext/sources/ggml/src/ggml-cpu/cmake/FindSMTIME.cmake
|
|
383
178
|
- ext/sources/ggml/src/ggml-cpu/common.h
|
|
384
179
|
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-impl.h
|
|
385
180
|
- ext/sources/ggml/src/ggml-cpu/ggml-cpu.c
|
|
@@ -403,7 +198,18 @@ files:
|
|
|
403
198
|
- ext/sources/ggml/src/ggml-cpu/spacemit/ime.cpp
|
|
404
199
|
- ext/sources/ggml/src/ggml-cpu/spacemit/ime.h
|
|
405
200
|
- ext/sources/ggml/src/ggml-cpu/spacemit/ime1_kernels.cpp
|
|
201
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/ime2_kernels.cpp
|
|
202
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/ime_env.cpp
|
|
203
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/ime_env.h
|
|
406
204
|
- ext/sources/ggml/src/ggml-cpu/spacemit/ime_kernels.h
|
|
205
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/repack.cpp
|
|
206
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/repack.h
|
|
207
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/rvv_kernels.cpp
|
|
208
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/rvv_kernels.h
|
|
209
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/spine_barrier.h
|
|
210
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/spine_mem_pool.cpp
|
|
211
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/spine_mem_pool.h
|
|
212
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/spine_tcm.h
|
|
407
213
|
- ext/sources/ggml/src/ggml-cpu/traits.cpp
|
|
408
214
|
- ext/sources/ggml/src/ggml-cpu/traits.h
|
|
409
215
|
- ext/sources/ggml/src/ggml-cpu/unary-ops.cpp
|
|
@@ -415,6 +221,8 @@ files:
|
|
|
415
221
|
- ext/sources/ggml/src/ggml-cuda/acc.cuh
|
|
416
222
|
- ext/sources/ggml/src/ggml-cuda/add-id.cu
|
|
417
223
|
- ext/sources/ggml/src/ggml-cuda/add-id.cuh
|
|
224
|
+
- ext/sources/ggml/src/ggml-cuda/allreduce.cu
|
|
225
|
+
- ext/sources/ggml/src/ggml-cuda/allreduce.cuh
|
|
418
226
|
- ext/sources/ggml/src/ggml-cuda/arange.cu
|
|
419
227
|
- ext/sources/ggml/src/ggml-cuda/arange.cuh
|
|
420
228
|
- ext/sources/ggml/src/ggml-cuda/argmax.cu
|
|
@@ -464,6 +272,8 @@ files:
|
|
|
464
272
|
- ext/sources/ggml/src/ggml-cuda/fattn.cuh
|
|
465
273
|
- ext/sources/ggml/src/ggml-cuda/fill.cu
|
|
466
274
|
- ext/sources/ggml/src/ggml-cuda/fill.cuh
|
|
275
|
+
- ext/sources/ggml/src/ggml-cuda/fwht.cu
|
|
276
|
+
- ext/sources/ggml/src/ggml-cuda/fwht.cuh
|
|
467
277
|
- ext/sources/ggml/src/ggml-cuda/gated_delta_net.cu
|
|
468
278
|
- ext/sources/ggml/src/ggml-cuda/gated_delta_net.cuh
|
|
469
279
|
- ext/sources/ggml/src/ggml-cuda/getrows.cu
|
|
@@ -513,6 +323,8 @@ files:
|
|
|
513
323
|
- ext/sources/ggml/src/ggml-cuda/set-rows.cuh
|
|
514
324
|
- ext/sources/ggml/src/ggml-cuda/set.cu
|
|
515
325
|
- ext/sources/ggml/src/ggml-cuda/set.cuh
|
|
326
|
+
- ext/sources/ggml/src/ggml-cuda/snake.cu
|
|
327
|
+
- ext/sources/ggml/src/ggml-cuda/snake.cuh
|
|
516
328
|
- ext/sources/ggml/src/ggml-cuda/softcap.cu
|
|
517
329
|
- ext/sources/ggml/src/ggml-cuda/softcap.cuh
|
|
518
330
|
- ext/sources/ggml/src/ggml-cuda/softmax.cu
|
|
@@ -550,50 +362,65 @@ files:
|
|
|
550
362
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_8-ncols2_8.cu
|
|
551
363
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq112-dv112.cu
|
|
552
364
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq128-dv128.cu
|
|
365
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq192-dv128.cu
|
|
553
366
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq256-dv256.cu
|
|
367
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq320-dv256.cu
|
|
554
368
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq40-dv40.cu
|
|
369
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq512-dv512.cu
|
|
555
370
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq576-dv512.cu
|
|
556
371
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq64-dv64.cu
|
|
557
372
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq72-dv72.cu
|
|
558
373
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq80-dv80.cu
|
|
559
374
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-tile-instance-dkq96-dv96.cu
|
|
375
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-bf16.cu
|
|
376
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-f16.cu
|
|
377
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q4_0.cu
|
|
378
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q4_1.cu
|
|
379
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q5_0.cu
|
|
380
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q5_1.cu
|
|
381
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-bf16-q8_0.cu
|
|
382
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-bf16.cu
|
|
560
383
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-f16.cu
|
|
561
384
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q4_0.cu
|
|
562
385
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q4_1.cu
|
|
563
386
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q5_0.cu
|
|
564
387
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q5_1.cu
|
|
565
388
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q8_0.cu
|
|
389
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-bf16.cu
|
|
566
390
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-f16.cu
|
|
567
391
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q4_0.cu
|
|
568
392
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q4_1.cu
|
|
569
393
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q5_0.cu
|
|
570
394
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q5_1.cu
|
|
571
395
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q8_0.cu
|
|
396
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-bf16.cu
|
|
572
397
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-f16.cu
|
|
573
398
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q4_0.cu
|
|
574
399
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q4_1.cu
|
|
575
400
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q5_0.cu
|
|
576
401
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q5_1.cu
|
|
577
402
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q8_0.cu
|
|
403
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-bf16.cu
|
|
578
404
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-f16.cu
|
|
579
405
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q4_0.cu
|
|
580
406
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q4_1.cu
|
|
581
407
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q5_0.cu
|
|
582
408
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q5_1.cu
|
|
583
409
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q8_0.cu
|
|
410
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-bf16.cu
|
|
584
411
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-f16.cu
|
|
585
412
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q4_0.cu
|
|
586
413
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q4_1.cu
|
|
587
414
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q5_0.cu
|
|
588
415
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q5_1.cu
|
|
589
416
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q8_0.cu
|
|
417
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-bf16.cu
|
|
590
418
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-f16.cu
|
|
591
419
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q4_0.cu
|
|
592
420
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q4_1.cu
|
|
593
421
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q5_0.cu
|
|
594
422
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q5_1.cu
|
|
595
423
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q8_0.cu
|
|
596
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/generate_cu_files.py
|
|
597
424
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_1.cu
|
|
598
425
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_10.cu
|
|
599
426
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_11.cu
|
|
@@ -619,6 +446,8 @@ files:
|
|
|
619
446
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-iq4_nl.cu
|
|
620
447
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-iq4_xs.cu
|
|
621
448
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-mxfp4.cu
|
|
449
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-nvfp4.cu
|
|
450
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q1_0.cu
|
|
622
451
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu
|
|
623
452
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q3_k.cu
|
|
624
453
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q4_0.cu
|
|
@@ -651,21 +480,34 @@ files:
|
|
|
651
480
|
- ext/sources/ggml/src/ggml-hexagon/ggml-hexagon.cpp
|
|
652
481
|
- ext/sources/ggml/src/ggml-hexagon/htp-drv.cpp
|
|
653
482
|
- ext/sources/ggml/src/ggml-hexagon/htp-drv.h
|
|
483
|
+
- ext/sources/ggml/src/ggml-hexagon/htp-opnode.h
|
|
654
484
|
- ext/sources/ggml/src/ggml-hexagon/htp/CMakeLists.txt
|
|
655
485
|
- ext/sources/ggml/src/ggml-hexagon/htp/act-ops.c
|
|
656
486
|
- ext/sources/ggml/src/ggml-hexagon/htp/argsort-ops.c
|
|
657
487
|
- ext/sources/ggml/src/ggml-hexagon/htp/binary-ops.c
|
|
658
488
|
- ext/sources/ggml/src/ggml-hexagon/htp/cmake-toolchain.cmake
|
|
489
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/concat-ops.c
|
|
659
490
|
- ext/sources/ggml/src/ggml-hexagon/htp/cpy-ops.c
|
|
491
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/cumsum-ops.c
|
|
492
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/diag-ops.c
|
|
493
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/fill-ops.c
|
|
660
494
|
- ext/sources/ggml/src/ggml-hexagon/htp/flash-attn-ops.c
|
|
495
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/gated-delta-net-ops.c
|
|
661
496
|
- ext/sources/ggml/src/ggml-hexagon/htp/get-rows-ops.c
|
|
662
497
|
- ext/sources/ggml/src/ggml-hexagon/htp/hex-dma.c
|
|
663
498
|
- ext/sources/ggml/src/ggml-hexagon/htp/hex-dma.h
|
|
664
499
|
- ext/sources/ggml/src/ggml-hexagon/htp/hex-dump.h
|
|
665
500
|
- ext/sources/ggml/src/ggml-hexagon/htp/hex-fastdiv.h
|
|
666
501
|
- ext/sources/ggml/src/ggml-hexagon/htp/hex-utils.h
|
|
502
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hmx-flash-attn-ops.c
|
|
503
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hmx-matmul-ops.c
|
|
504
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hmx-ops.c
|
|
505
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hmx-ops.h
|
|
506
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hmx-profile.h
|
|
507
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hmx-queue.c
|
|
508
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hmx-queue.h
|
|
509
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hmx-utils.h
|
|
667
510
|
- ext/sources/ggml/src/ggml-hexagon/htp/htp-ctx.h
|
|
668
|
-
- ext/sources/ggml/src/ggml-hexagon/htp/htp-msg.h
|
|
669
511
|
- ext/sources/ggml/src/ggml-hexagon/htp/htp-ops.h
|
|
670
512
|
- ext/sources/ggml/src/ggml-hexagon/htp/htp_iface.idl
|
|
671
513
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-arith.h
|
|
@@ -674,27 +516,35 @@ files:
|
|
|
674
516
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-div.h
|
|
675
517
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-dump.h
|
|
676
518
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-exp.h
|
|
519
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-flash-attn.h
|
|
677
520
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-floor.h
|
|
678
521
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-inverse.h
|
|
522
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-log.h
|
|
523
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-pow.h
|
|
679
524
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-reduce.h
|
|
525
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-repl.h
|
|
680
526
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-scale.h
|
|
681
527
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-sigmoid.h
|
|
528
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-sin-cos.h
|
|
682
529
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-sqrt.h
|
|
683
530
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-types.h
|
|
684
531
|
- ext/sources/ggml/src/ggml-hexagon/htp/hvx-utils.h
|
|
685
532
|
- ext/sources/ggml/src/ggml-hexagon/htp/main.c
|
|
686
533
|
- ext/sources/ggml/src/ggml-hexagon/htp/matmul-ops.c
|
|
534
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/pad-ops.c
|
|
535
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/repeat-ops.c
|
|
687
536
|
- ext/sources/ggml/src/ggml-hexagon/htp/rope-ops.c
|
|
688
537
|
- ext/sources/ggml/src/ggml-hexagon/htp/set-rows-ops.c
|
|
689
538
|
- ext/sources/ggml/src/ggml-hexagon/htp/softmax-ops.c
|
|
539
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/solve-tri-ops.c
|
|
690
540
|
- ext/sources/ggml/src/ggml-hexagon/htp/ssm-conv.c
|
|
691
541
|
- ext/sources/ggml/src/ggml-hexagon/htp/sum-rows-ops.c
|
|
692
542
|
- ext/sources/ggml/src/ggml-hexagon/htp/unary-ops.c
|
|
543
|
+
- ext/sources/ggml/src/ggml-hexagon/htp/vtcm-utils.h
|
|
693
544
|
- ext/sources/ggml/src/ggml-hexagon/htp/worker-pool.c
|
|
694
545
|
- ext/sources/ggml/src/ggml-hexagon/htp/worker-pool.h
|
|
695
546
|
- ext/sources/ggml/src/ggml-hexagon/libdl.h
|
|
696
547
|
- ext/sources/ggml/src/ggml-hexagon/libggml-htp.inf
|
|
697
|
-
- ext/sources/ggml/src/ggml-hexagon/op-desc.h
|
|
698
548
|
- ext/sources/ggml/src/ggml-hip/CMakeLists.txt
|
|
699
549
|
- ext/sources/ggml/src/ggml-impl.h
|
|
700
550
|
- ext/sources/ggml/src/ggml-metal/CMakeLists.txt
|
|
@@ -728,21 +578,52 @@ files:
|
|
|
728
578
|
- ext/sources/ggml/src/ggml-opencl/kernels/diag.cl
|
|
729
579
|
- ext/sources/ggml/src/ggml-opencl/kernels/diag_mask_inf.cl
|
|
730
580
|
- ext/sources/ggml/src/ggml-opencl/kernels/div.cl
|
|
731
|
-
- ext/sources/ggml/src/ggml-opencl/kernels/embed_kernel.py
|
|
732
581
|
- ext/sources/ggml/src/ggml-opencl/kernels/exp.cl
|
|
733
582
|
- ext/sources/ggml/src/ggml-opencl/kernels/expm1.cl
|
|
734
583
|
- ext/sources/ggml/src/ggml-opencl/kernels/fill.cl
|
|
735
584
|
- ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f16.cl
|
|
736
585
|
- ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f32.cl
|
|
737
586
|
- ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f32_f16.cl
|
|
587
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gated_delta_net.cl
|
|
738
588
|
- ext/sources/ggml/src/ggml-opencl/kernels/gelu.cl
|
|
739
589
|
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_mxfp4_f32.cl
|
|
590
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_mxfp4_f32_ns.cl
|
|
591
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q4_0_f32_ns.cl
|
|
592
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q4_1_f32_ns.cl
|
|
593
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q4_k_f32_ns.cl
|
|
594
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q5_0_f32_ns.cl
|
|
595
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q5_1_f32_ns.cl
|
|
596
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q5_k_f32_ns.cl
|
|
597
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_moe_q6_k_f32_ns.cl
|
|
598
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_iq4_nl_f32.cl
|
|
599
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q4_0_f32.cl
|
|
740
600
|
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q4_1_f32.cl
|
|
601
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q4_k_f32.cl
|
|
602
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q5_0_f32.cl
|
|
603
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q5_1_f32.cl
|
|
604
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q5_k_f32.cl
|
|
605
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q6_k_f32.cl
|
|
606
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_noshuffle_q8_0_f32.cl
|
|
607
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemm_xmem_f16_f32_os8.cl
|
|
741
608
|
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_mxfp4_f32.cl
|
|
742
|
-
- ext/sources/ggml/src/ggml-opencl/kernels/
|
|
743
|
-
- ext/sources/ggml/src/ggml-opencl/kernels/
|
|
744
|
-
- ext/sources/ggml/src/ggml-opencl/kernels/
|
|
609
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_mxfp4_f32_ns.cl
|
|
610
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q4_0_f32_ns.cl
|
|
611
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q4_1_f32_ns.cl
|
|
612
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q4_k_f32_ns.cl
|
|
613
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q5_0_f32_ns.cl
|
|
614
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q5_1_f32_ns.cl
|
|
615
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q5_k_f32_ns.cl
|
|
616
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_moe_q6_k_f32_ns.cl
|
|
617
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_iq4_nl_f32.cl
|
|
618
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q4_0_f32.cl
|
|
619
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q4_0_f32_spec.cl
|
|
745
620
|
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q4_1_f32.cl
|
|
621
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q4_k_f32.cl
|
|
622
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q5_0_f32.cl
|
|
623
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q5_1_f32.cl
|
|
624
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q5_k_f32.cl
|
|
625
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q6_k_f32.cl
|
|
626
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_q8_0_f32.cl
|
|
746
627
|
- ext/sources/ggml/src/ggml-opencl/kernels/get_rows.cl
|
|
747
628
|
- ext/sources/ggml/src/ggml-opencl/kernels/glu.cl
|
|
748
629
|
- ext/sources/ggml/src/ggml-opencl/kernels/group_norm.cl
|
|
@@ -750,16 +631,21 @@ files:
|
|
|
750
631
|
- ext/sources/ggml/src/ggml-opencl/kernels/im2col_f32.cl
|
|
751
632
|
- ext/sources/ggml/src/ggml-opencl/kernels/l2_norm.cl
|
|
752
633
|
- ext/sources/ggml/src/ggml-opencl/kernels/mean.cl
|
|
634
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/moe_reorder_b.cl
|
|
635
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/moe_sort_by_expert.cl
|
|
753
636
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul.cl
|
|
754
|
-
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mat_Ab_Bi_8x4.cl
|
|
755
637
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mat_f16_f32.cl
|
|
756
638
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_f16_f32_kq_kqv.cl
|
|
757
639
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_f16_f32_l4_lm.cl
|
|
758
640
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_f32_f32_l4_lm.cl
|
|
641
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_iq4_nl_f32_l4_lm.cl
|
|
759
642
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q4_0_f32_l4_lm.cl
|
|
760
643
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q4_1_f32_l4_lm.cl
|
|
644
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q4_k_f32_l4_lm.cl
|
|
645
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q5_0_f32_l4_lm.cl
|
|
646
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q5_1_f32_l4_lm.cl
|
|
647
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q5_k_f32_l4_lm.cl
|
|
761
648
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q6_k_f32_l4_lm.cl
|
|
762
|
-
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q8_0_f32_8x4.cl
|
|
763
649
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_q8_0_f32_l4_lm.cl
|
|
764
650
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_f16_f16.cl
|
|
765
651
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_f16_f32.cl
|
|
@@ -771,6 +657,8 @@ files:
|
|
|
771
657
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q4_0_f32_8x_flat.cl
|
|
772
658
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q8_0_f32.cl
|
|
773
659
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q8_0_f32_flat.cl
|
|
660
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_iq4_nl_f32.cl
|
|
661
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_iq4_nl_f32_flat.cl
|
|
774
662
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_mxfp4_f32.cl
|
|
775
663
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_mxfp4_f32_flat.cl
|
|
776
664
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_0_f32.cl
|
|
@@ -781,6 +669,13 @@ files:
|
|
|
781
669
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_1_f32.cl
|
|
782
670
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_1_f32_flat.cl
|
|
783
671
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_k_f32.cl
|
|
672
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_k_f32_flat.cl
|
|
673
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_0_f32.cl
|
|
674
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_0_f32_flat.cl
|
|
675
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_1_f32.cl
|
|
676
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_1_f32_flat.cl
|
|
677
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_k_f32.cl
|
|
678
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q5_k_f32_flat.cl
|
|
784
679
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q6_k_f32.cl
|
|
785
680
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q6_k_f32_flat.cl
|
|
786
681
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q8_0_f32.cl
|
|
@@ -842,17 +737,17 @@ files:
|
|
|
842
737
|
- ext/sources/ggml/src/ggml-openvino/openvino/op/set_rows.cpp
|
|
843
738
|
- ext/sources/ggml/src/ggml-openvino/openvino/op/softmax.cpp
|
|
844
739
|
- ext/sources/ggml/src/ggml-openvino/openvino/op/transpose.cpp
|
|
740
|
+
- ext/sources/ggml/src/ggml-openvino/openvino/op/unary_gelu.cpp
|
|
845
741
|
- ext/sources/ggml/src/ggml-openvino/openvino/op/unary_silu.cpp
|
|
846
742
|
- ext/sources/ggml/src/ggml-openvino/openvino/op/view.cpp
|
|
847
743
|
- ext/sources/ggml/src/ggml-openvino/openvino/op_table.cpp
|
|
848
744
|
- ext/sources/ggml/src/ggml-openvino/openvino/op_table.h
|
|
849
|
-
- ext/sources/ggml/src/ggml-openvino/openvino/pass/eliminate_zp.cpp
|
|
850
|
-
- ext/sources/ggml/src/ggml-openvino/openvino/pass/eliminate_zp.h
|
|
851
745
|
- ext/sources/ggml/src/ggml-openvino/openvino/pass/fuse_to_sdpa.cpp
|
|
852
746
|
- ext/sources/ggml/src/ggml-openvino/openvino/pass/fuse_to_sdpa.h
|
|
853
747
|
- ext/sources/ggml/src/ggml-openvino/openvino/pass/mark_decompression_convert_constant_folding.h
|
|
854
748
|
- ext/sources/ggml/src/ggml-openvino/openvino/pass/squeeze_matmul.cpp
|
|
855
749
|
- ext/sources/ggml/src/ggml-openvino/openvino/pass/squeeze_matmul.h
|
|
750
|
+
- ext/sources/ggml/src/ggml-openvino/openvino/rt_info/weightless_caching_attributes.hpp
|
|
856
751
|
- ext/sources/ggml/src/ggml-openvino/openvino/translate_session.cpp
|
|
857
752
|
- ext/sources/ggml/src/ggml-openvino/openvino/translate_session.h
|
|
858
753
|
- ext/sources/ggml/src/ggml-openvino/openvino/utils.cpp
|
|
@@ -864,6 +759,8 @@ files:
|
|
|
864
759
|
- ext/sources/ggml/src/ggml-quants.h
|
|
865
760
|
- ext/sources/ggml/src/ggml-rpc/CMakeLists.txt
|
|
866
761
|
- ext/sources/ggml/src/ggml-rpc/ggml-rpc.cpp
|
|
762
|
+
- ext/sources/ggml/src/ggml-rpc/transport.cpp
|
|
763
|
+
- ext/sources/ggml/src/ggml-rpc/transport.h
|
|
867
764
|
- ext/sources/ggml/src/ggml-sycl/CMakeLists.txt
|
|
868
765
|
- ext/sources/ggml/src/ggml-sycl/add-id.cpp
|
|
869
766
|
- ext/sources/ggml/src/ggml-sycl/add-id.hpp
|
|
@@ -882,18 +779,26 @@ files:
|
|
|
882
779
|
- ext/sources/ggml/src/ggml-sycl/count-equal.hpp
|
|
883
780
|
- ext/sources/ggml/src/ggml-sycl/cpy.cpp
|
|
884
781
|
- ext/sources/ggml/src/ggml-sycl/cpy.hpp
|
|
782
|
+
- ext/sources/ggml/src/ggml-sycl/cumsum.cpp
|
|
783
|
+
- ext/sources/ggml/src/ggml-sycl/cumsum.hpp
|
|
885
784
|
- ext/sources/ggml/src/ggml-sycl/dequantize.hpp
|
|
785
|
+
- ext/sources/ggml/src/ggml-sycl/diag.cpp
|
|
786
|
+
- ext/sources/ggml/src/ggml-sycl/diag.hpp
|
|
886
787
|
- ext/sources/ggml/src/ggml-sycl/dmmv.cpp
|
|
887
788
|
- ext/sources/ggml/src/ggml-sycl/dmmv.hpp
|
|
888
789
|
- ext/sources/ggml/src/ggml-sycl/dpct/helper.hpp
|
|
889
790
|
- ext/sources/ggml/src/ggml-sycl/element_wise.cpp
|
|
890
791
|
- ext/sources/ggml/src/ggml-sycl/element_wise.hpp
|
|
792
|
+
- ext/sources/ggml/src/ggml-sycl/fattn-buffers.cpp
|
|
793
|
+
- ext/sources/ggml/src/ggml-sycl/fattn-buffers.hpp
|
|
891
794
|
- ext/sources/ggml/src/ggml-sycl/fattn-common.hpp
|
|
892
795
|
- ext/sources/ggml/src/ggml-sycl/fattn-tile.cpp
|
|
893
796
|
- ext/sources/ggml/src/ggml-sycl/fattn-tile.hpp
|
|
894
797
|
- ext/sources/ggml/src/ggml-sycl/fattn-vec.hpp
|
|
895
798
|
- ext/sources/ggml/src/ggml-sycl/fattn.cpp
|
|
896
799
|
- ext/sources/ggml/src/ggml-sycl/fattn.hpp
|
|
800
|
+
- ext/sources/ggml/src/ggml-sycl/fill.cpp
|
|
801
|
+
- ext/sources/ggml/src/ggml-sycl/fill.hpp
|
|
897
802
|
- ext/sources/ggml/src/ggml-sycl/gated_delta_net.cpp
|
|
898
803
|
- ext/sources/ggml/src/ggml-sycl/gated_delta_net.hpp
|
|
899
804
|
- ext/sources/ggml/src/ggml-sycl/gemm.hpp
|
|
@@ -931,14 +836,19 @@ files:
|
|
|
931
836
|
- ext/sources/ggml/src/ggml-sycl/set_rows.hpp
|
|
932
837
|
- ext/sources/ggml/src/ggml-sycl/softmax.cpp
|
|
933
838
|
- ext/sources/ggml/src/ggml-sycl/softmax.hpp
|
|
839
|
+
- ext/sources/ggml/src/ggml-sycl/solve_tri.cpp
|
|
840
|
+
- ext/sources/ggml/src/ggml-sycl/solve_tri.hpp
|
|
934
841
|
- ext/sources/ggml/src/ggml-sycl/ssm_conv.cpp
|
|
935
842
|
- ext/sources/ggml/src/ggml-sycl/ssm_conv.hpp
|
|
843
|
+
- ext/sources/ggml/src/ggml-sycl/ssm_scan.cpp
|
|
844
|
+
- ext/sources/ggml/src/ggml-sycl/ssm_scan.hpp
|
|
936
845
|
- ext/sources/ggml/src/ggml-sycl/sycl_hw.cpp
|
|
937
846
|
- ext/sources/ggml/src/ggml-sycl/sycl_hw.hpp
|
|
938
847
|
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq112-dv112.cpp
|
|
939
848
|
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq128-dv128.cpp
|
|
940
849
|
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq256-dv256.cpp
|
|
941
850
|
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq40-dv40.cpp
|
|
851
|
+
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq512-dv512.cpp
|
|
942
852
|
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq576-dv512.cpp
|
|
943
853
|
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq64-dv64.cpp
|
|
944
854
|
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-tile-instance-dkq72-dv72.cpp
|
|
@@ -982,6 +892,9 @@ files:
|
|
|
982
892
|
- ext/sources/ggml/src/ggml-sycl/template-instances/fattn-vec-instance-q8_0-q8_0.cpp
|
|
983
893
|
- ext/sources/ggml/src/ggml-sycl/tsembd.cpp
|
|
984
894
|
- ext/sources/ggml/src/ggml-sycl/tsembd.hpp
|
|
895
|
+
- ext/sources/ggml/src/ggml-sycl/type.hpp
|
|
896
|
+
- ext/sources/ggml/src/ggml-sycl/upscale.cpp
|
|
897
|
+
- ext/sources/ggml/src/ggml-sycl/upscale.hpp
|
|
985
898
|
- ext/sources/ggml/src/ggml-sycl/vecdotq.hpp
|
|
986
899
|
- ext/sources/ggml/src/ggml-sycl/wkv.cpp
|
|
987
900
|
- ext/sources/ggml/src/ggml-sycl/wkv.hpp
|
|
@@ -1015,7 +928,6 @@ files:
|
|
|
1015
928
|
- ext/sources/ggml/src/ggml-virtgpu/ggml-remoting.h
|
|
1016
929
|
- ext/sources/ggml/src/ggml-virtgpu/ggmlremoting_functions.yaml
|
|
1017
930
|
- ext/sources/ggml/src/ggml-virtgpu/include/apir_hw.h
|
|
1018
|
-
- ext/sources/ggml/src/ggml-virtgpu/regenerate_remoting.py
|
|
1019
931
|
- ext/sources/ggml/src/ggml-virtgpu/virtgpu-apir.h
|
|
1020
932
|
- ext/sources/ggml/src/ggml-virtgpu/virtgpu-forward-backend.cpp
|
|
1021
933
|
- ext/sources/ggml/src/ggml-virtgpu/virtgpu-forward-buffer-type.cpp
|
|
@@ -1073,6 +985,8 @@ files:
|
|
|
1073
985
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq4_nl.comp
|
|
1074
986
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq4_xs.comp
|
|
1075
987
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_mxfp4.comp
|
|
988
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_nvfp4.comp
|
|
989
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q1_0.comp
|
|
1076
990
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q2_k.comp
|
|
1077
991
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q3_k.comp
|
|
1078
992
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q4_0.comp
|
|
@@ -1086,20 +1000,25 @@ files:
|
|
|
1086
1000
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/diag.comp
|
|
1087
1001
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/diag_mask_inf.comp
|
|
1088
1002
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/div.comp
|
|
1003
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dot_product_funcs.glsl
|
|
1089
1004
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/elu.comp
|
|
1090
1005
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/exp.comp
|
|
1091
1006
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/feature-tests/bfloat16.comp
|
|
1092
1007
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/feature-tests/coopmat.comp
|
|
1093
1008
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/feature-tests/coopmat2.comp
|
|
1009
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/feature-tests/coopmat2_decode_vector.comp
|
|
1094
1010
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/feature-tests/integer_dot.comp
|
|
1095
1011
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/fill.comp
|
|
1096
1012
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn.comp
|
|
1097
1013
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_base.glsl
|
|
1098
1014
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp
|
|
1099
1015
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm2.comp
|
|
1016
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_dequant.glsl
|
|
1100
1017
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_mask_opt.comp
|
|
1018
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_mmq_funcs.glsl
|
|
1101
1019
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_split_k_reduce.comp
|
|
1102
1020
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/floor.comp
|
|
1021
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/fwht.comp
|
|
1103
1022
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/gated_delta_net.comp
|
|
1104
1023
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/geglu.comp
|
|
1105
1024
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/geglu_erf.comp
|
|
@@ -1174,13 +1093,13 @@ files:
|
|
|
1174
1093
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_params.glsl
|
|
1175
1094
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_vision.comp
|
|
1176
1095
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/round.comp
|
|
1177
|
-
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rte.glsl
|
|
1178
1096
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/scale.comp
|
|
1179
1097
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sgn.comp
|
|
1180
1098
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sigmoid.comp
|
|
1181
1099
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/silu.comp
|
|
1182
1100
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/silu_back.comp
|
|
1183
1101
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sin.comp
|
|
1102
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/snake.comp
|
|
1184
1103
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/soft_max.comp
|
|
1185
1104
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/soft_max_back.comp
|
|
1186
1105
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/soft_max_large1.comp
|
|
@@ -1217,33 +1136,54 @@ files:
|
|
|
1217
1136
|
- ext/sources/ggml/src/ggml-webgpu/ggml-webgpu-shader-lib.hpp
|
|
1218
1137
|
- ext/sources/ggml/src/ggml-webgpu/ggml-webgpu.cpp
|
|
1219
1138
|
- ext/sources/ggml/src/ggml-webgpu/pre_wgsl.hpp
|
|
1139
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/add_id.wgsl
|
|
1220
1140
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/argmax.wgsl
|
|
1221
1141
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/argsort.wgsl
|
|
1222
1142
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/argsort_merge.wgsl
|
|
1223
1143
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/binary.wgsl
|
|
1224
1144
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/common_decls.tmpl
|
|
1225
1145
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/concat.wgsl
|
|
1226
|
-
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/
|
|
1146
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/conv2d.wgsl
|
|
1147
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/cpy.wgsl
|
|
1227
1148
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/cumsum.wgsl
|
|
1228
|
-
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/embed_wgsl.py
|
|
1229
1149
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn.wgsl
|
|
1150
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_quant_staging.tmpl
|
|
1151
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_tile.wgsl
|
|
1152
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_vec_blk.wgsl
|
|
1153
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_vec_reduce.wgsl
|
|
1154
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_vec_split.wgsl
|
|
1155
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/gated_delta_net.wgsl
|
|
1230
1156
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/get_rows.wgsl
|
|
1231
|
-
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/glu.
|
|
1157
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/glu.wgsl
|
|
1158
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/im2col.wgsl
|
|
1232
1159
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/memset.wgsl
|
|
1233
|
-
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat.wgsl
|
|
1234
1160
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_decls.tmpl
|
|
1161
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_id.wgsl
|
|
1162
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_id_gather.wgsl
|
|
1163
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_id_vec.wgsl
|
|
1235
1164
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_reg_tile.wgsl
|
|
1236
1165
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_subgroup_matrix.wgsl
|
|
1237
1166
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_vec.wgsl
|
|
1167
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_vec_acc.tmpl
|
|
1168
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_vec_q_acc.tmpl
|
|
1238
1169
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/pad.wgsl
|
|
1170
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/quant_inner_loops.tmpl
|
|
1171
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/quantize_q8.wgsl
|
|
1239
1172
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/repeat.wgsl
|
|
1240
|
-
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/
|
|
1241
|
-
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rope.
|
|
1173
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rms_norm_mul.wgsl
|
|
1174
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rope.wgsl
|
|
1175
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/row_norm.wgsl
|
|
1242
1176
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/scale.wgsl
|
|
1177
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/set.wgsl
|
|
1243
1178
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/set_rows.wgsl
|
|
1244
|
-
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/
|
|
1179
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/set_rows_quant.wgsl
|
|
1180
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/soft_max.wgsl
|
|
1181
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/solve_tri.wgsl
|
|
1182
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/ssm_conv.wgsl
|
|
1183
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/ssm_scan.wgsl
|
|
1245
1184
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/sum_rows.wgsl
|
|
1246
1185
|
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/unary.wgsl
|
|
1186
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/upscale.wgsl
|
|
1247
1187
|
- ext/sources/ggml/src/ggml-zdnn/CMakeLists.txt
|
|
1248
1188
|
- ext/sources/ggml/src/ggml-zdnn/common.hpp
|
|
1249
1189
|
- ext/sources/ggml/src/ggml-zdnn/ggml-zdnn.cpp
|
|
@@ -1256,7 +1196,9 @@ files:
|
|
|
1256
1196
|
- ext/sources/ggml/src/ggml.c
|
|
1257
1197
|
- ext/sources/ggml/src/ggml.cpp
|
|
1258
1198
|
- ext/sources/ggml/src/gguf.cpp
|
|
1199
|
+
- ext/sources/include/parakeet.h
|
|
1259
1200
|
- ext/sources/include/whisper.h
|
|
1201
|
+
- ext/sources/media/matmul.png
|
|
1260
1202
|
- ext/sources/src/CMakeLists.txt
|
|
1261
1203
|
- ext/sources/src/coreml/whisper-compat.h
|
|
1262
1204
|
- ext/sources/src/coreml/whisper-compat.m
|
|
@@ -1268,36 +1210,14 @@ files:
|
|
|
1268
1210
|
- ext/sources/src/coreml/whisper-encoder.mm
|
|
1269
1211
|
- ext/sources/src/openvino/whisper-openvino-encoder.cpp
|
|
1270
1212
|
- ext/sources/src/openvino/whisper-openvino-encoder.h
|
|
1213
|
+
- ext/sources/src/parakeet-arch.h
|
|
1214
|
+
- ext/sources/src/parakeet.cpp
|
|
1271
1215
|
- ext/sources/src/whisper-arch.h
|
|
1272
1216
|
- ext/sources/src/whisper.cpp
|
|
1273
|
-
- ext/sources/tests/CMakeLists.txt
|
|
1274
|
-
- ext/sources/tests/earnings21/eval.mk
|
|
1275
|
-
- ext/sources/tests/earnings21/eval.py
|
|
1276
|
-
- ext/sources/tests/earnings21/normalizers/__init__.py
|
|
1277
|
-
- ext/sources/tests/earnings21/normalizers/basic.py
|
|
1278
|
-
- ext/sources/tests/earnings21/normalizers/english.json
|
|
1279
|
-
- ext/sources/tests/earnings21/normalizers/english.py
|
|
1280
|
-
- ext/sources/tests/earnings21/requirements.txt
|
|
1281
|
-
- ext/sources/tests/en-0-ref.txt
|
|
1282
|
-
- ext/sources/tests/en-1-ref.txt
|
|
1283
|
-
- ext/sources/tests/en-2-ref.txt
|
|
1284
|
-
- ext/sources/tests/es-0-ref.txt
|
|
1285
|
-
- ext/sources/tests/librispeech/eval.mk
|
|
1286
|
-
- ext/sources/tests/librispeech/eval.py
|
|
1287
|
-
- ext/sources/tests/librispeech/normalizers/__init__.py
|
|
1288
|
-
- ext/sources/tests/librispeech/normalizers/basic.py
|
|
1289
|
-
- ext/sources/tests/librispeech/normalizers/english.json
|
|
1290
|
-
- ext/sources/tests/librispeech/normalizers/english.py
|
|
1291
|
-
- ext/sources/tests/librispeech/requirements.txt
|
|
1292
|
-
- ext/sources/tests/run-tests.sh
|
|
1293
|
-
- ext/sources/tests/test-c.c
|
|
1294
|
-
- ext/sources/tests/test-vad-full.cpp
|
|
1295
|
-
- ext/sources/tests/test-vad.cpp
|
|
1296
|
-
- ext/sources/tests/test-whisper.js
|
|
1297
1217
|
- extsources.rb
|
|
1298
|
-
- lib/whisper/
|
|
1218
|
+
- lib/whisper/log_settable.rb
|
|
1299
1219
|
- lib/whisper/model/uri.rb
|
|
1300
|
-
- lib/whisper/
|
|
1220
|
+
- lib/whisper/output.rb
|
|
1301
1221
|
- sig/whisper.rbs
|
|
1302
1222
|
- test/helper.rb
|
|
1303
1223
|
- test/jfk_reader/.gitignore
|
|
@@ -1308,6 +1228,14 @@ files:
|
|
|
1308
1228
|
- test/test_error.rb
|
|
1309
1229
|
- test/test_model.rb
|
|
1310
1230
|
- test/test_package.rb
|
|
1231
|
+
- test/test_parakeet.rb
|
|
1232
|
+
- test/test_parakeet_callback.rb
|
|
1233
|
+
- test/test_parakeet_context.rb
|
|
1234
|
+
- test/test_parakeet_context_params.rb
|
|
1235
|
+
- test/test_parakeet_model.rb
|
|
1236
|
+
- test/test_parakeet_params.rb
|
|
1237
|
+
- test/test_parakeet_segment.rb
|
|
1238
|
+
- test/test_parakeet_token.rb
|
|
1311
1239
|
- test/test_params.rb
|
|
1312
1240
|
- test/test_segment.rb
|
|
1313
1241
|
- test/test_token.rb
|
|
@@ -1322,7 +1250,6 @@ homepage: https://github.com/ggml-org/whisper.cpp
|
|
|
1322
1250
|
licenses:
|
|
1323
1251
|
- MIT
|
|
1324
1252
|
metadata: {}
|
|
1325
|
-
post_install_message:
|
|
1326
1253
|
rdoc_options:
|
|
1327
1254
|
- "--main"
|
|
1328
1255
|
- README.md
|
|
@@ -1332,15 +1259,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
|
|
|
1332
1259
|
requirements:
|
|
1333
1260
|
- - ">="
|
|
1334
1261
|
- !ruby/object:Gem::Version
|
|
1335
|
-
version: 3.
|
|
1262
|
+
version: 3.3.0
|
|
1336
1263
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
1337
1264
|
requirements:
|
|
1338
1265
|
- - ">="
|
|
1339
1266
|
- !ruby/object:Gem::Version
|
|
1340
1267
|
version: '0'
|
|
1341
1268
|
requirements: []
|
|
1342
|
-
rubygems_version: 3.
|
|
1343
|
-
signing_key:
|
|
1269
|
+
rubygems_version: 3.6.9
|
|
1344
1270
|
specification_version: 4
|
|
1345
1271
|
summary: Ruby whisper.cpp bindings
|
|
1346
1272
|
test_files:
|
|
@@ -1353,6 +1279,14 @@ test_files:
|
|
|
1353
1279
|
- test/test_error.rb
|
|
1354
1280
|
- test/test_model.rb
|
|
1355
1281
|
- test/test_package.rb
|
|
1282
|
+
- test/test_parakeet.rb
|
|
1283
|
+
- test/test_parakeet_callback.rb
|
|
1284
|
+
- test/test_parakeet_context.rb
|
|
1285
|
+
- test/test_parakeet_context_params.rb
|
|
1286
|
+
- test/test_parakeet_model.rb
|
|
1287
|
+
- test/test_parakeet_params.rb
|
|
1288
|
+
- test/test_parakeet_segment.rb
|
|
1289
|
+
- test/test_parakeet_token.rb
|
|
1356
1290
|
- test/test_params.rb
|
|
1357
1291
|
- test/test_segment.rb
|
|
1358
1292
|
- test/test_token.rb
|