whispercpp 1.3.2 → 1.3.4
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/.gitignore +6 -3
- data/README.md +71 -14
- data/Rakefile +20 -7
- data/ext/.gitignore +4 -6
- data/ext/dependencies.rb +36 -24
- data/ext/extconf.rb +1 -1
- data/ext/options.rb +48 -184
- data/ext/ruby_whisper.c +18 -0
- data/ext/ruby_whisper_context.c +43 -12
- data/ext/ruby_whisper_model.c +1 -1
- data/ext/ruby_whisper_params.c +59 -27
- data/ext/ruby_whisper_segment.c +81 -4
- data/ext/ruby_whisper_transcribe.cpp +13 -7
- data/ext/ruby_whisper_vad_params.c +1 -1
- data/ext/sources/CMakeLists.txt +5 -1
- data/ext/sources/bindings/javascript/package.json +1 -1
- data/ext/sources/build-xcframework.sh +24 -0
- data/ext/sources/examples/CMakeLists.txt +1 -0
- data/ext/sources/examples/addon.node/__test__/whisper.spec.js +120 -24
- data/ext/sources/examples/addon.node/addon.cpp +154 -35
- data/ext/sources/examples/addon.node/index.js +10 -5
- data/ext/sources/examples/addon.node/vad-example.js +132 -0
- data/ext/sources/examples/bench/bench.cpp +29 -18
- data/ext/sources/examples/bench.wasm/index-tmpl.html +10 -9
- data/ext/sources/examples/cli/cli.cpp +7 -4
- data/ext/sources/examples/command/command.cpp +58 -32
- data/ext/sources/examples/command.wasm/index-tmpl.html +5 -4
- data/ext/sources/examples/common-ggml.cpp +2 -0
- data/ext/sources/examples/common-whisper.cpp +14 -7
- data/ext/sources/examples/lsp/lsp.cpp +21 -17
- data/ext/sources/examples/quantize/quantize.cpp +3 -0
- data/ext/sources/examples/server/CMakeLists.txt +3 -0
- data/ext/sources/examples/server/server.cpp +193 -35
- data/ext/sources/examples/server.py +6 -1
- data/ext/sources/examples/stream/stream.cpp +10 -2
- data/ext/sources/examples/stream.wasm/emscripten.cpp +6 -6
- data/ext/sources/examples/stream.wasm/index-tmpl.html +82 -5
- data/ext/sources/examples/talk-llama/CMakeLists.txt +3 -0
- data/ext/sources/examples/talk-llama/llama-adapter.cpp +101 -4
- data/ext/sources/examples/talk-llama/llama-adapter.h +6 -0
- data/ext/sources/examples/talk-llama/llama-arch.cpp +756 -15
- data/ext/sources/examples/talk-llama/llama-arch.h +85 -1
- data/ext/sources/examples/talk-llama/llama-batch.cpp +773 -272
- data/ext/sources/examples/talk-llama/llama-batch.h +126 -55
- data/ext/sources/examples/talk-llama/llama-chat.cpp +150 -13
- data/ext/sources/examples/talk-llama/llama-chat.h +8 -0
- data/ext/sources/examples/talk-llama/llama-context.cpp +814 -542
- data/ext/sources/examples/talk-llama/llama-context.h +68 -32
- data/ext/sources/examples/talk-llama/llama-cparams.cpp +1 -1
- data/ext/sources/examples/talk-llama/llama-cparams.h +4 -4
- data/ext/sources/examples/talk-llama/llama-graph.cpp +787 -440
- data/ext/sources/examples/talk-llama/llama-graph.h +333 -153
- data/ext/sources/examples/talk-llama/llama-hparams.cpp +128 -6
- data/ext/sources/examples/talk-llama/llama-hparams.h +80 -17
- data/ext/sources/examples/talk-llama/llama-impl.h +2 -0
- data/ext/sources/examples/talk-llama/llama-kv-cache-iswa.cpp +326 -0
- data/ext/sources/examples/talk-llama/llama-kv-cache-iswa.h +137 -0
- data/ext/sources/examples/talk-llama/llama-kv-cache.cpp +1248 -1967
- data/ext/sources/examples/talk-llama/llama-kv-cache.h +218 -345
- data/ext/sources/examples/talk-llama/llama-kv-cells.h +164 -52
- data/ext/sources/examples/talk-llama/llama-memory-hybrid.cpp +266 -0
- data/ext/sources/examples/talk-llama/llama-memory-hybrid.h +139 -0
- data/ext/sources/examples/talk-llama/llama-memory-recurrent.cpp +1154 -0
- data/ext/sources/examples/talk-llama/llama-memory-recurrent.h +182 -0
- data/ext/sources/examples/talk-llama/llama-memory.cpp +58 -0
- data/ext/sources/examples/talk-llama/llama-memory.h +94 -4
- data/ext/sources/examples/talk-llama/llama-mmap.cpp +1 -1
- data/ext/sources/examples/talk-llama/llama-model-loader.cpp +44 -17
- data/ext/sources/examples/talk-llama/llama-model-loader.h +3 -2
- data/ext/sources/examples/talk-llama/llama-model-saver.cpp +1 -0
- data/ext/sources/examples/talk-llama/llama-model.cpp +11377 -5248
- data/ext/sources/examples/talk-llama/llama-model.h +87 -9
- data/ext/sources/examples/talk-llama/llama-quant.cpp +137 -16
- data/ext/sources/examples/talk-llama/llama-sampling.cpp +226 -126
- data/ext/sources/examples/talk-llama/llama-vocab.cpp +502 -38
- data/ext/sources/examples/talk-llama/llama-vocab.h +46 -0
- data/ext/sources/examples/talk-llama/llama.cpp +76 -17
- data/ext/sources/examples/talk-llama/llama.h +176 -151
- data/ext/sources/examples/talk-llama/talk-llama.cpp +11 -6
- data/ext/sources/examples/talk-llama/unicode.cpp +212 -0
- data/ext/sources/examples/talk-llama/unicode.h +45 -0
- data/ext/sources/examples/vad-speech-segments/speech.cpp +6 -0
- data/ext/sources/examples/wchess/wchess.cmd/wchess.cmd.cpp +6 -2
- data/ext/sources/examples/whisper.wasm/index-tmpl.html +17 -16
- data/ext/sources/ggml/CMakeLists.txt +106 -33
- data/ext/sources/ggml/cmake/common.cmake +24 -0
- data/ext/sources/ggml/cmake/ggml-config.cmake.in +132 -93
- data/ext/sources/ggml/include/ggml-backend.h +18 -2
- data/ext/sources/ggml/include/ggml-cpu.h +2 -0
- data/ext/sources/ggml/include/ggml-metal.h +1 -6
- data/ext/sources/ggml/include/ggml-opt.h +25 -6
- data/ext/sources/ggml/include/ggml-webgpu.h +19 -0
- data/ext/sources/ggml/include/ggml-zdnn.h +17 -0
- data/ext/sources/ggml/include/ggml.h +365 -21
- data/ext/sources/ggml/src/CMakeLists.txt +98 -25
- data/ext/sources/ggml/src/ggml-alloc.c +265 -141
- data/ext/sources/ggml/src/ggml-backend-impl.h +4 -1
- data/ext/sources/ggml/src/ggml-backend-reg.cpp +35 -13
- data/ext/sources/ggml/src/ggml-backend.cpp +266 -60
- data/ext/sources/ggml/src/ggml-blas/CMakeLists.txt +4 -4
- data/ext/sources/ggml/src/ggml-blas/ggml-blas.cpp +5 -4
- data/ext/sources/ggml/src/ggml-cann/CMakeLists.txt +15 -0
- data/ext/sources/ggml/src/ggml-cann/acl_tensor.cpp +3 -1
- data/ext/sources/ggml/src/ggml-cann/aclnn_ops.cpp +903 -717
- data/ext/sources/ggml/src/ggml-cann/aclnn_ops.h +143 -25
- data/ext/sources/ggml/src/ggml-cann/common.h +149 -2
- data/ext/sources/ggml/src/ggml-cann/ggml-cann.cpp +521 -78
- data/ext/sources/ggml/src/ggml-common.h +21 -0
- data/ext/sources/ggml/src/ggml-cpu/CMakeLists.txt +165 -50
- data/ext/sources/ggml/src/ggml-cpu/amx/amx.cpp +5 -3
- data/ext/sources/ggml/src/ggml-cpu/amx/mmq.cpp +11 -10
- data/ext/sources/ggml/src/ggml-cpu/arch/arm/cpu-feats.cpp +94 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/arm/quants.c +3650 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/arm/repack.cpp +1891 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/loongarch/quants.c +2160 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/powerpc/cpu-feats.cpp +82 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/powerpc/quants.c +2305 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/riscv/quants.c +1897 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/riscv/repack.cpp +342 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/s390/quants.c +1468 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/wasm/quants.c +1221 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/x86/quants.c +3820 -0
- data/ext/sources/ggml/src/ggml-cpu/arch/x86/repack.cpp +6307 -0
- data/ext/sources/ggml/src/ggml-cpu/arch-fallback.h +214 -0
- data/ext/sources/ggml/src/ggml-cpu/common.h +18 -3
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu-impl.h +23 -7
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu.c +179 -110
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu.cpp +44 -33
- data/ext/sources/ggml/src/ggml-cpu/{ggml-cpu-hbm.cpp → hbm.cpp} +1 -1
- data/ext/sources/ggml/src/ggml-cpu/kleidiai/kernels.cpp +152 -18
- data/ext/sources/ggml/src/ggml-cpu/kleidiai/kernels.h +7 -1
- data/ext/sources/ggml/src/ggml-cpu/kleidiai/kleidiai.cpp +228 -98
- data/ext/sources/ggml/src/ggml-cpu/llamafile/sgemm.cpp +532 -1124
- data/ext/sources/ggml/src/ggml-cpu/llamafile/sgemm.h +5 -0
- data/ext/sources/ggml/src/ggml-cpu/ops.cpp +3374 -2081
- data/ext/sources/ggml/src/ggml-cpu/ops.h +13 -8
- data/ext/sources/ggml/src/ggml-cpu/quants.c +1193 -0
- data/ext/sources/ggml/src/ggml-cpu/{ggml-cpu-quants.h → quants.h} +34 -0
- data/ext/sources/ggml/src/ggml-cpu/repack.cpp +1982 -0
- data/ext/sources/ggml/src/ggml-cpu/repack.h +120 -0
- data/ext/sources/ggml/src/ggml-cpu/simd-mappings.h +367 -46
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime.cpp +1024 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime.h +13 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime1_kernels.cpp +3196 -0
- data/ext/sources/ggml/src/ggml-cpu/spacemit/ime_kernels.h +26 -0
- data/ext/sources/ggml/src/ggml-cpu/{ggml-cpu-traits.cpp → traits.cpp} +3 -3
- data/ext/sources/ggml/src/ggml-cpu/{ggml-cpu-traits.h → traits.h} +1 -1
- data/ext/sources/ggml/src/ggml-cpu/vec.cpp +272 -35
- data/ext/sources/ggml/src/ggml-cpu/vec.h +794 -142
- data/ext/sources/ggml/src/ggml-cuda/CMakeLists.txt +20 -16
- data/ext/sources/ggml/src/ggml-cuda/add-id.cu +58 -0
- data/ext/sources/ggml/src/ggml-cuda/add-id.cuh +3 -0
- data/ext/sources/ggml/src/ggml-cuda/binbcast.cu +330 -191
- data/ext/sources/ggml/src/ggml-cuda/binbcast.cuh +2 -0
- data/ext/sources/ggml/src/ggml-cuda/common.cuh +291 -81
- data/ext/sources/ggml/src/ggml-cuda/conv-transpose-1d.cu +1 -4
- data/ext/sources/ggml/src/ggml-cuda/conv2d-dw.cu +161 -0
- data/ext/sources/ggml/src/ggml-cuda/conv2d-dw.cuh +5 -0
- data/ext/sources/ggml/src/ggml-cuda/conv2d-transpose.cu +91 -0
- data/ext/sources/ggml/src/ggml-cuda/conv2d-transpose.cuh +4 -0
- data/ext/sources/ggml/src/ggml-cuda/conv2d.cu +166 -0
- data/ext/sources/ggml/src/ggml-cuda/conv2d.cuh +5 -0
- data/ext/sources/ggml/src/ggml-cuda/convert.cu +117 -22
- data/ext/sources/ggml/src/ggml-cuda/convert.cuh +20 -0
- data/ext/sources/ggml/src/ggml-cuda/cpy-utils.cuh +217 -0
- data/ext/sources/ggml/src/ggml-cuda/cpy.cu +64 -307
- data/ext/sources/ggml/src/ggml-cuda/cross-entropy-loss.cu +2 -14
- data/ext/sources/ggml/src/ggml-cuda/dequantize.cuh +14 -40
- data/ext/sources/ggml/src/ggml-cuda/fattn-common.cuh +499 -368
- data/ext/sources/ggml/src/ggml-cuda/fattn-mma-f16.cuh +142 -93
- data/ext/sources/ggml/src/ggml-cuda/fattn-tile.cu +755 -0
- data/ext/sources/ggml/src/ggml-cuda/fattn-tile.cuh +3 -0
- data/ext/sources/ggml/src/ggml-cuda/fattn-vec.cuh +593 -0
- data/ext/sources/ggml/src/ggml-cuda/fattn-wmma-f16.cu +90 -50
- data/ext/sources/ggml/src/ggml-cuda/fattn.cu +185 -198
- data/ext/sources/ggml/src/ggml-cuda/fattn.cuh +2 -0
- data/ext/sources/ggml/src/ggml-cuda/getrows.cu +50 -39
- data/ext/sources/ggml/src/ggml-cuda/ggml-cuda.cu +636 -222
- data/ext/sources/ggml/src/ggml-cuda/im2col.cu +196 -35
- data/ext/sources/ggml/src/ggml-cuda/im2col.cuh +1 -0
- data/ext/sources/ggml/src/ggml-cuda/mean.cu +73 -0
- data/ext/sources/ggml/src/ggml-cuda/mean.cuh +3 -0
- data/ext/sources/ggml/src/ggml-cuda/mma.cuh +198 -45
- data/ext/sources/ggml/src/ggml-cuda/mmf.cu +123 -0
- data/ext/sources/ggml/src/ggml-cuda/mmf.cuh +496 -0
- data/ext/sources/ggml/src/ggml-cuda/mmq.cu +206 -57
- data/ext/sources/ggml/src/ggml-cuda/mmq.cuh +1262 -721
- data/ext/sources/ggml/src/ggml-cuda/mmvf.cu +506 -0
- data/ext/sources/ggml/src/ggml-cuda/{mmv.cuh → mmvf.cuh} +4 -5
- data/ext/sources/ggml/src/ggml-cuda/mmvq.cu +64 -73
- data/ext/sources/ggml/src/ggml-cuda/norm.cu +284 -12
- data/ext/sources/ggml/src/ggml-cuda/norm.cuh +7 -0
- data/ext/sources/ggml/src/ggml-cuda/opt-step-sgd.cu +49 -0
- data/ext/sources/ggml/src/ggml-cuda/opt-step-sgd.cuh +5 -0
- data/ext/sources/ggml/src/ggml-cuda/pad.cu +46 -23
- data/ext/sources/ggml/src/ggml-cuda/pad_reflect_1d.cu +91 -0
- data/ext/sources/ggml/src/ggml-cuda/pad_reflect_1d.cuh +5 -0
- data/ext/sources/ggml/src/ggml-cuda/quantize.cu +12 -10
- data/ext/sources/ggml/src/ggml-cuda/reduce_rows.cuh +53 -0
- data/ext/sources/ggml/src/ggml-cuda/roll.cu +67 -0
- data/ext/sources/ggml/src/ggml-cuda/roll.cuh +5 -0
- data/ext/sources/ggml/src/ggml-cuda/rope.cu +21 -27
- data/ext/sources/ggml/src/ggml-cuda/scale.cu +14 -11
- data/ext/sources/ggml/src/ggml-cuda/set-rows.cu +276 -0
- data/ext/sources/ggml/src/ggml-cuda/set-rows.cuh +7 -0
- data/ext/sources/ggml/src/ggml-cuda/softcap.cu +34 -0
- data/ext/sources/ggml/src/ggml-cuda/softcap.cuh +5 -0
- data/ext/sources/ggml/src/ggml-cuda/softmax.cu +126 -59
- data/ext/sources/ggml/src/ggml-cuda/ssm-conv.cu +10 -2
- data/ext/sources/ggml/src/ggml-cuda/ssm-scan.cu +322 -98
- data/ext/sources/ggml/src/ggml-cuda/sum.cu +6 -10
- data/ext/sources/ggml/src/ggml-cuda/sumrows.cu +23 -19
- data/ext/sources/ggml/src/ggml-cuda/sumrows.cuh +0 -1
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-f16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q4_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q4_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q5_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q5_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q8_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-f16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q4_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q4_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q5_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q5_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q8_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-f16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q4_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q4_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q5_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q5_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q8_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-f16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q4_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q4_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q5_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q5_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q8_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-f16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q4_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q4_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q5_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q5_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q8_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-f16.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q4_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q4_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q5_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q5_1.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q8_0.cu +7 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/generate_cu_files.py +21 -18
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_1.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_10.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_11.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_12.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_13.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_14.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_15.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_16.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_2.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_3.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_4.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_5.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_6.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_7.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_8.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_9.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-mxfp4.cu +5 -0
- data/ext/sources/ggml/src/ggml-cuda/topk-moe.cu +259 -0
- data/ext/sources/ggml/src/ggml-cuda/topk-moe.cuh +14 -0
- data/ext/sources/ggml/src/ggml-cuda/tsembd.cu +3 -3
- data/ext/sources/ggml/src/ggml-cuda/unary.cu +179 -0
- data/ext/sources/ggml/src/ggml-cuda/unary.cuh +15 -0
- data/ext/sources/ggml/src/ggml-cuda/upscale.cu +92 -6
- data/ext/sources/ggml/src/ggml-cuda/vecdotq.cuh +110 -22
- data/ext/sources/ggml/src/ggml-cuda/vendors/cuda.h +4 -0
- data/ext/sources/ggml/src/ggml-cuda/vendors/hip.h +58 -36
- data/ext/sources/ggml/src/ggml-cuda/vendors/musa.h +4 -3
- data/ext/sources/ggml/src/ggml-hip/CMakeLists.txt +14 -2
- data/ext/sources/ggml/src/ggml-impl.h +229 -175
- data/ext/sources/ggml/src/ggml-metal/CMakeLists.txt +21 -17
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-common.cpp +446 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-common.h +52 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-context.h +33 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-context.m +600 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-device.cpp +1376 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-device.h +226 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-device.m +1308 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-impl.h +163 -63
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-ops.cpp +3158 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal-ops.h +82 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal.cpp +718 -0
- data/ext/sources/ggml/src/ggml-metal/ggml-metal.metal +3208 -1575
- data/ext/sources/ggml/src/ggml-musa/CMakeLists.txt +18 -8
- data/ext/sources/ggml/src/ggml-musa/mudnn.cuh +2 -2
- data/ext/sources/ggml/src/ggml-opencl/CMakeLists.txt +32 -0
- data/ext/sources/ggml/src/ggml-opencl/ggml-opencl.cpp +4430 -792
- data/ext/sources/ggml/src/ggml-opencl/kernels/add.cl +107 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/add_id.cl +42 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/argsort.cl +86 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/concat.cl +109 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/conv2d.cl +185 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/conv2d_f16_f32.cl +176 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/cvt.cl +84 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/div.cl +138 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f16.cl +370 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f32.cl +370 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f32_f16.cl +373 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/gelu.cl +27 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/glu.cl +378 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/group_norm.cl +121 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/im2col_f16.cl +1 -1
- data/ext/sources/ggml/src/ggml-opencl/kernels/im2col_f32.cl +1 -1
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul.cl +73 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mat_f16_f32.cl +130 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_f16_f32_l4_lm.cl +132 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_f32_f32_l4_lm.cl +133 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_mxfp4_f32.cl +189 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_mxfp4_f32_flat.cl +176 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q4_0_f32_8x_flat.cl +283 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q8_0_f32.cl +140 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q8_0_f32_flat.cl +222 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_mxfp4_f32.cl +144 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_mxfp4_f32_flat.cl +167 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q8_0_f32.cl +125 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q8_0_f32_flat.cl +202 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/norm.cl +80 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/pad.cl +30 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/repeat.cl +39 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/rms_norm.cl +79 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/scale.cl +3 -2
- data/ext/sources/ggml/src/ggml-opencl/kernels/set_rows.cl +189 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/sigmoid.cl +29 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/softmax_4_f16.cl +34 -13
- data/ext/sources/ggml/src/ggml-opencl/kernels/softmax_4_f32.cl +34 -13
- data/ext/sources/ggml/src/ggml-opencl/kernels/softmax_f16.cl +34 -13
- data/ext/sources/ggml/src/ggml-opencl/kernels/softmax_f32.cl +34 -13
- data/ext/sources/ggml/src/ggml-opencl/kernels/sub.cl +138 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/sum_rows.cl +39 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/tanh.cl +63 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/transpose.cl +20 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/tsembd.cl +48 -0
- data/ext/sources/ggml/src/ggml-opencl/kernels/upscale.cl +120 -0
- data/ext/sources/ggml/src/ggml-opt.cpp +97 -41
- data/ext/sources/ggml/src/ggml-quants.c +117 -24
- data/ext/sources/ggml/src/ggml-quants.h +6 -0
- data/ext/sources/ggml/src/ggml-rpc/ggml-rpc.cpp +85 -62
- data/ext/sources/ggml/src/ggml-sycl/CMakeLists.txt +3 -3
- data/ext/sources/ggml/src/ggml-sycl/backend.hpp +2 -0
- data/ext/sources/ggml/src/ggml-sycl/binbcast.cpp +9 -0
- data/ext/sources/ggml/src/ggml-sycl/binbcast.hpp +6 -0
- data/ext/sources/ggml/src/ggml-sycl/common.hpp +20 -48
- data/ext/sources/ggml/src/ggml-sycl/concat.cpp +13 -17
- data/ext/sources/ggml/src/ggml-sycl/convert.cpp +21 -2
- data/ext/sources/ggml/src/ggml-sycl/cpy.cpp +116 -211
- data/ext/sources/ggml/src/ggml-sycl/cpy.hpp +213 -1
- data/ext/sources/ggml/src/ggml-sycl/dequantize.hpp +32 -0
- data/ext/sources/ggml/src/ggml-sycl/element_wise.cpp +700 -1041
- data/ext/sources/ggml/src/ggml-sycl/element_wise.hpp +20 -9
- data/ext/sources/ggml/src/ggml-sycl/gemm.hpp +17 -26
- data/ext/sources/ggml/src/ggml-sycl/getrows.cpp +2 -96
- data/ext/sources/ggml/src/ggml-sycl/ggml-sycl.cpp +393 -250
- data/ext/sources/ggml/src/ggml-sycl/im2col.cpp +1 -1
- data/ext/sources/ggml/src/ggml-sycl/mmvq.cpp +32 -8
- data/ext/sources/ggml/src/ggml-sycl/quantize.hpp +133 -0
- data/ext/sources/ggml/src/ggml-sycl/quants.hpp +38 -11
- data/ext/sources/ggml/src/ggml-sycl/rope.cpp +125 -21
- data/ext/sources/ggml/src/ggml-sycl/set_rows.cpp +234 -0
- data/ext/sources/ggml/src/ggml-sycl/set_rows.hpp +8 -0
- data/ext/sources/ggml/src/ggml-sycl/sycl_hw.cpp +3 -1
- data/ext/sources/ggml/src/ggml-sycl/sycl_hw.hpp +3 -0
- data/ext/sources/ggml/src/ggml-sycl/tsembd.cpp +4 -3
- data/ext/sources/ggml/src/ggml-sycl/vecdotq.hpp +105 -17
- data/ext/sources/ggml/src/ggml-vulkan/CMakeLists.txt +36 -32
- data/ext/sources/ggml/src/ggml-vulkan/ggml-vulkan.cpp +4198 -1145
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/CMakeLists.txt +4 -12
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/add.comp +41 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/add_id.comp +42 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/argmax.comp +13 -4
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/argsort.comp +39 -29
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/conv2d_mm.comp +349 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/conv_transpose_1d.comp +98 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/copy_from_quant.comp +2 -2
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/copy_to_quant.comp +66 -12
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_funcs.comp +154 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_funcs_cm2.comp +21 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq2_s.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq2_xxs.comp +2 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq3_s.comp +6 -5
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq3_xxs.comp +4 -2
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_mxfp4.comp +32 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q2_k.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q3_k.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q4_k.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q5_k.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q6_k.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/exp.comp +21 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn.comp +69 -24
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_base.comp +60 -20
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp +98 -42
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm2.comp +64 -27
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_split_k_reduce.comp +74 -13
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/geglu.comp +13 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/geglu_erf.comp +27 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/geglu_quick.comp +11 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/gelu_erf.comp +39 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/generic_binary_head.comp +4 -17
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/get_rows.comp +19 -10
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/get_rows_quant.comp +25 -15
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/glu_head.comp +19 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/glu_main.comp +29 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/hardsigmoid.comp +22 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/hardswish.comp +22 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/im2col.comp +18 -14
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/im2col_3d.comp +126 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_base.comp +65 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_nc.comp +11 -7
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vecq.comp +140 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm.comp +144 -531
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_cm2.comp +206 -38
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_funcs.comp +556 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq.comp +12 -5
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq_funcs.comp +15 -9
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/multi_add.comp +111 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/opt_step_sgd.comp +22 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/pad.comp +24 -3
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/quantize_q8_1.comp +53 -3
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/reglu.comp +9 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rms_norm.comp +64 -11
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rms_norm_partials.comp +65 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/roll.comp +46 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_head.comp +1 -4
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_multi.comp +7 -9
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_neox.comp +7 -9
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_norm.comp +7 -9
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rte.comp +5 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/scale.comp +1 -1
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/soft_max.comp +29 -7
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/soft_max_back.comp +4 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sqrt.comp +17 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sum_rows.comp +38 -5
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/swiglu.comp +9 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/swiglu_oai.comp +14 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/timestep_embedding.comp +4 -3
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/types.comp +101 -9
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/upscale.comp +69 -5
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/utils.comp +25 -0
- data/ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp +338 -71
- data/ext/sources/ggml/src/ggml-webgpu/CMakeLists.txt +54 -0
- data/ext/sources/ggml/src/ggml-webgpu/ggml-webgpu.cpp +1558 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/add.tmpl.wgsl +44 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/add_in_place.tmpl.wgsl +41 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/binary_head.tmpl +45 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/common_decls.tmpl +930 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/cpy.wgsl +60 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/embed_wgsl.py +124 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/get_rows.tmpl.wgsl +874 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/memset.wgsl +40 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul.tmpl.wgsl +44 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_in_place.tmpl.wgsl +41 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat.tmpl.wgsl +907 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rms_norm.wgsl +57 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rms_norm_in_place.wgsl +48 -0
- data/ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/set_rows.wgsl +81 -0
- data/ext/sources/ggml/src/ggml-zdnn/CMakeLists.txt +36 -0
- data/ext/sources/ggml/src/ggml-zdnn/common.hpp +59 -0
- data/ext/sources/ggml/src/ggml-zdnn/ggml-zdnn.cpp +628 -0
- data/ext/sources/ggml/src/ggml-zdnn/mmf.cpp +80 -0
- data/ext/sources/ggml/src/ggml-zdnn/mmf.hpp +12 -0
- data/ext/sources/ggml/src/ggml-zdnn/utils.cpp +79 -0
- data/ext/sources/ggml/src/ggml-zdnn/utils.hpp +19 -0
- data/ext/sources/ggml/src/ggml.c +802 -142
- data/ext/sources/ggml/src/ggml.cpp +26 -0
- data/ext/sources/ggml/src/gguf.cpp +32 -4
- data/ext/sources/include/whisper.h +2 -0
- data/ext/sources/src/CMakeLists.txt +2 -0
- data/ext/sources/src/coreml/whisper-compat.h +10 -0
- data/ext/sources/src/coreml/whisper-compat.m +35 -0
- data/ext/sources/src/coreml/whisper-decoder-impl.m +1 -0
- data/ext/sources/src/coreml/whisper-encoder-impl.m +1 -0
- data/ext/sources/src/whisper.cpp +241 -215
- data/ext/sources/tests/CMakeLists.txt +8 -1
- data/ext/sources/tests/test-vad-full.cpp +3 -3
- data/ext/sources/tests/test-vad.cpp +2 -2
- data/extsources.rb +15 -9
- data/lib/whisper/context.rb +15 -0
- data/lib/whisper/model/uri.rb +57 -2
- data/lib/whisper/segment.rb +58 -0
- data/sig/whisper.rbs +75 -38
- data/{tests → test}/helper.rb +1 -12
- data/{tests → test}/test_model.rb +9 -0
- data/test/test_package.rb +51 -0
- data/{tests → test}/test_params.rb +8 -0
- data/test/test_segment.rb +146 -0
- data/{tests → test}/test_whisper.rb +70 -0
- data/whispercpp.gemspec +2 -3
- metadata +246 -191
- data/ext/sources/.dockerignore +0 -3
- data/ext/sources/.github/workflows/bindings-ruby.yml +0 -21
- data/ext/sources/ci/run.sh +0 -336
- data/ext/sources/close-issue.yml +0 -28
- data/ext/sources/ggml/include/ggml-kompute.h +0 -50
- data/ext/sources/ggml/src/ggml-amx/CMakeLists.txt +0 -107
- data/ext/sources/ggml/src/ggml-amx/common.h +0 -94
- data/ext/sources/ggml/src/ggml-amx/ggml-amx.cpp +0 -446
- data/ext/sources/ggml/src/ggml-amx/mmq.cpp +0 -2510
- data/ext/sources/ggml/src/ggml-amx/mmq.h +0 -17
- data/ext/sources/ggml/src/ggml-cann/kernels/CMakeLists.txt +0 -30
- data/ext/sources/ggml/src/ggml-cann/kernels/ascendc_kernels.h +0 -19
- data/ext/sources/ggml/src/ggml-cann/kernels/dup.cpp +0 -234
- data/ext/sources/ggml/src/ggml-cann/kernels/get_row_f16.cpp +0 -197
- data/ext/sources/ggml/src/ggml-cann/kernels/get_row_f32.cpp +0 -190
- data/ext/sources/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp +0 -204
- data/ext/sources/ggml/src/ggml-cann/kernels/get_row_q8_0.cpp +0 -191
- data/ext/sources/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp +0 -218
- data/ext/sources/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp +0 -216
- data/ext/sources/ggml/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp +0 -295
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp +0 -6431
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu-aarch64.h +0 -8
- data/ext/sources/ggml/src/ggml-cpu/ggml-cpu-quants.c +0 -13747
- data/ext/sources/ggml/src/ggml-cuda/fattn-tile-f16.cu +0 -357
- data/ext/sources/ggml/src/ggml-cuda/fattn-tile-f16.cuh +0 -3
- data/ext/sources/ggml/src/ggml-cuda/fattn-tile-f32.cu +0 -365
- data/ext/sources/ggml/src/ggml-cuda/fattn-tile-f32.cuh +0 -3
- data/ext/sources/ggml/src/ggml-cuda/fattn-vec-f16.cuh +0 -482
- data/ext/sources/ggml/src/ggml-cuda/fattn-vec-f32.cuh +0 -472
- data/ext/sources/ggml/src/ggml-cuda/mmv.cu +0 -336
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs256-f16-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs256-f16-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-f16.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_1.cu +0 -5
- data/ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q8_0.cu +0 -5
- data/ext/sources/ggml/src/ggml-kompute/CMakeLists.txt +0 -166
- data/ext/sources/ggml/src/ggml-kompute/ggml-kompute.cpp +0 -2251
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/common.comp +0 -112
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_add.comp +0 -58
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_addrow.comp +0 -25
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_cpy_f16_f16.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_cpy_f16_f32.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_cpy_f32_f16.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_cpy_f32_f32.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_diagmask.comp +0 -30
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_gelu.comp +0 -22
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows.comp +0 -17
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_f16.comp +0 -31
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_f32.comp +0 -31
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_q4_0.comp +0 -38
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_q4_1.comp +0 -39
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_q6_k.comp +0 -44
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_f16.comp +0 -69
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_mat_f32.comp +0 -51
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q4_0.comp +0 -33
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q4_1.comp +0 -35
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q4_k.comp +0 -140
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q6_k.comp +0 -106
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q8_0.comp +0 -73
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mv_q_n.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mv_q_n_pre.comp +0 -28
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_norm.comp +0 -84
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_relu.comp +0 -21
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rmsnorm.comp +0 -53
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rope_neox_f16.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rope_neox_f32.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rope_norm_f16.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rope_norm_f32.comp +0 -52
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_scale.comp +0 -19
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_scale_8.comp +0 -23
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_silu.comp +0 -22
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_softmax.comp +0 -72
- data/ext/sources/ggml/src/ggml-kompute/kompute-shaders/rope_common.comp +0 -71
- data/ext/sources/ggml/src/ggml-metal/ggml-metal.m +0 -5998
- data/tests/test_package.rb +0 -46
- data/tests/test_segment.rb +0 -74
- /data/ext/sources/ggml/src/ggml-cpu/{cpu-feats-x86.cpp → arch/x86/cpu-feats.cpp} +0 -0
- /data/ext/sources/ggml/src/ggml-cpu/{ggml-cpu-hbm.h → hbm.h} +0 -0
- /data/{tests → test}/jfk_reader/.gitignore +0 -0
- /data/{tests → test}/jfk_reader/extconf.rb +0 -0
- /data/{tests → test}/jfk_reader/jfk_reader.c +0 -0
- /data/{tests → test}/test_callback.rb +0 -0
- /data/{tests → test}/test_error.rb +0 -0
- /data/{tests → test}/test_vad.rb +0 -0
- /data/{tests → test}/test_vad_params.rb +0 -0
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: whispercpp
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.3.
|
4
|
+
version: 1.3.4
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Georgi Gerganov
|
@@ -9,7 +9,7 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2025-
|
12
|
+
date: 2025-10-08 00:00:00.000000000 Z
|
13
13
|
dependencies: []
|
14
14
|
description: High-performance inference of OpenAI's Whisper automatic speech recognition
|
15
15
|
(ASR) model via Ruby
|
@@ -38,8 +38,6 @@ files:
|
|
38
38
|
- ext/ruby_whisper_segment.c
|
39
39
|
- ext/ruby_whisper_transcribe.cpp
|
40
40
|
- ext/ruby_whisper_vad_params.c
|
41
|
-
- ext/sources/.dockerignore
|
42
|
-
- ext/sources/.github/workflows/bindings-ruby.yml
|
43
41
|
- ext/sources/CMakeGraphVizOptions.cmake
|
44
42
|
- ext/sources/CMakeLists.txt
|
45
43
|
- ext/sources/bindings/javascript/CMakeLists.txt
|
@@ -49,8 +47,6 @@ files:
|
|
49
47
|
- ext/sources/bindings/javascript/package.json
|
50
48
|
- ext/sources/bindings/javascript/whisper.js
|
51
49
|
- ext/sources/build-xcframework.sh
|
52
|
-
- ext/sources/ci/run.sh
|
53
|
-
- ext/sources/close-issue.yml
|
54
50
|
- ext/sources/cmake/DefaultTargetOptions.cmake
|
55
51
|
- ext/sources/cmake/FindFFmpeg.cmake
|
56
52
|
- ext/sources/cmake/build-info.cmake
|
@@ -63,6 +59,7 @@ files:
|
|
63
59
|
- ext/sources/examples/addon.node/addon.cpp
|
64
60
|
- ext/sources/examples/addon.node/index.js
|
65
61
|
- ext/sources/examples/addon.node/package.json
|
62
|
+
- ext/sources/examples/addon.node/vad-example.js
|
66
63
|
- ext/sources/examples/bench.wasm/CMakeLists.txt
|
67
64
|
- ext/sources/examples/bench.wasm/emscripten.cpp
|
68
65
|
- ext/sources/examples/bench.wasm/index-tmpl.html
|
@@ -141,9 +138,15 @@ files:
|
|
141
138
|
- ext/sources/examples/talk-llama/llama-impl.h
|
142
139
|
- ext/sources/examples/talk-llama/llama-io.cpp
|
143
140
|
- ext/sources/examples/talk-llama/llama-io.h
|
141
|
+
- ext/sources/examples/talk-llama/llama-kv-cache-iswa.cpp
|
142
|
+
- ext/sources/examples/talk-llama/llama-kv-cache-iswa.h
|
144
143
|
- ext/sources/examples/talk-llama/llama-kv-cache.cpp
|
145
144
|
- ext/sources/examples/talk-llama/llama-kv-cache.h
|
146
145
|
- ext/sources/examples/talk-llama/llama-kv-cells.h
|
146
|
+
- ext/sources/examples/talk-llama/llama-memory-hybrid.cpp
|
147
|
+
- ext/sources/examples/talk-llama/llama-memory-hybrid.h
|
148
|
+
- ext/sources/examples/talk-llama/llama-memory-recurrent.cpp
|
149
|
+
- ext/sources/examples/talk-llama/llama-memory-recurrent.h
|
147
150
|
- ext/sources/examples/talk-llama/llama-memory.cpp
|
148
151
|
- ext/sources/examples/talk-llama/llama-memory.h
|
149
152
|
- ext/sources/examples/talk-llama/llama-mmap.cpp
|
@@ -197,22 +200,18 @@ files:
|
|
197
200
|
- ext/sources/ggml/include/ggml-cpp.h
|
198
201
|
- ext/sources/ggml/include/ggml-cpu.h
|
199
202
|
- ext/sources/ggml/include/ggml-cuda.h
|
200
|
-
- ext/sources/ggml/include/ggml-kompute.h
|
201
203
|
- ext/sources/ggml/include/ggml-metal.h
|
202
204
|
- ext/sources/ggml/include/ggml-opencl.h
|
203
205
|
- ext/sources/ggml/include/ggml-opt.h
|
204
206
|
- ext/sources/ggml/include/ggml-rpc.h
|
205
207
|
- ext/sources/ggml/include/ggml-sycl.h
|
206
208
|
- ext/sources/ggml/include/ggml-vulkan.h
|
209
|
+
- ext/sources/ggml/include/ggml-webgpu.h
|
210
|
+
- ext/sources/ggml/include/ggml-zdnn.h
|
207
211
|
- ext/sources/ggml/include/ggml.h
|
208
212
|
- ext/sources/ggml/include/gguf.h
|
209
213
|
- ext/sources/ggml/src/CMakeLists.txt
|
210
214
|
- ext/sources/ggml/src/ggml-alloc.c
|
211
|
-
- ext/sources/ggml/src/ggml-amx/CMakeLists.txt
|
212
|
-
- ext/sources/ggml/src/ggml-amx/common.h
|
213
|
-
- ext/sources/ggml/src/ggml-amx/ggml-amx.cpp
|
214
|
-
- ext/sources/ggml/src/ggml-amx/mmq.cpp
|
215
|
-
- ext/sources/ggml/src/ggml-amx/mmq.h
|
216
215
|
- ext/sources/ggml/src/ggml-backend-impl.h
|
217
216
|
- ext/sources/ggml/src/ggml-backend-reg.cpp
|
218
217
|
- ext/sources/ggml/src/ggml-backend.cpp
|
@@ -226,16 +225,6 @@ files:
|
|
226
225
|
- ext/sources/ggml/src/ggml-cann/aclnn_ops.h
|
227
226
|
- ext/sources/ggml/src/ggml-cann/common.h
|
228
227
|
- ext/sources/ggml/src/ggml-cann/ggml-cann.cpp
|
229
|
-
- ext/sources/ggml/src/ggml-cann/kernels/CMakeLists.txt
|
230
|
-
- ext/sources/ggml/src/ggml-cann/kernels/ascendc_kernels.h
|
231
|
-
- ext/sources/ggml/src/ggml-cann/kernels/dup.cpp
|
232
|
-
- ext/sources/ggml/src/ggml-cann/kernels/get_row_f16.cpp
|
233
|
-
- ext/sources/ggml/src/ggml-cann/kernels/get_row_f32.cpp
|
234
|
-
- ext/sources/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp
|
235
|
-
- ext/sources/ggml/src/ggml-cann/kernels/get_row_q8_0.cpp
|
236
|
-
- ext/sources/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp
|
237
|
-
- ext/sources/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp
|
238
|
-
- ext/sources/ggml/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp
|
239
228
|
- ext/sources/ggml/src/ggml-common.h
|
240
229
|
- ext/sources/ggml/src/ggml-cpu/CMakeLists.txt
|
241
230
|
- ext/sources/ggml/src/ggml-cpu/amx/amx.cpp
|
@@ -243,22 +232,29 @@ files:
|
|
243
232
|
- ext/sources/ggml/src/ggml-cpu/amx/common.h
|
244
233
|
- ext/sources/ggml/src/ggml-cpu/amx/mmq.cpp
|
245
234
|
- ext/sources/ggml/src/ggml-cpu/amx/mmq.h
|
235
|
+
- ext/sources/ggml/src/ggml-cpu/arch-fallback.h
|
236
|
+
- ext/sources/ggml/src/ggml-cpu/arch/arm/cpu-feats.cpp
|
237
|
+
- ext/sources/ggml/src/ggml-cpu/arch/arm/quants.c
|
238
|
+
- ext/sources/ggml/src/ggml-cpu/arch/arm/repack.cpp
|
239
|
+
- ext/sources/ggml/src/ggml-cpu/arch/loongarch/quants.c
|
240
|
+
- ext/sources/ggml/src/ggml-cpu/arch/powerpc/cpu-feats.cpp
|
241
|
+
- ext/sources/ggml/src/ggml-cpu/arch/powerpc/quants.c
|
242
|
+
- ext/sources/ggml/src/ggml-cpu/arch/riscv/quants.c
|
243
|
+
- ext/sources/ggml/src/ggml-cpu/arch/riscv/repack.cpp
|
244
|
+
- ext/sources/ggml/src/ggml-cpu/arch/s390/quants.c
|
245
|
+
- ext/sources/ggml/src/ggml-cpu/arch/wasm/quants.c
|
246
|
+
- ext/sources/ggml/src/ggml-cpu/arch/x86/cpu-feats.cpp
|
247
|
+
- ext/sources/ggml/src/ggml-cpu/arch/x86/quants.c
|
248
|
+
- ext/sources/ggml/src/ggml-cpu/arch/x86/repack.cpp
|
246
249
|
- ext/sources/ggml/src/ggml-cpu/binary-ops.cpp
|
247
250
|
- ext/sources/ggml/src/ggml-cpu/binary-ops.h
|
248
251
|
- ext/sources/ggml/src/ggml-cpu/cmake/FindSIMD.cmake
|
249
252
|
- ext/sources/ggml/src/ggml-cpu/common.h
|
250
|
-
- ext/sources/ggml/src/ggml-cpu/cpu-feats-x86.cpp
|
251
|
-
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp
|
252
|
-
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-aarch64.h
|
253
|
-
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-hbm.cpp
|
254
|
-
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-hbm.h
|
255
253
|
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-impl.h
|
256
|
-
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-quants.c
|
257
|
-
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-quants.h
|
258
|
-
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-traits.cpp
|
259
|
-
- ext/sources/ggml/src/ggml-cpu/ggml-cpu-traits.h
|
260
254
|
- ext/sources/ggml/src/ggml-cpu/ggml-cpu.c
|
261
255
|
- ext/sources/ggml/src/ggml-cpu/ggml-cpu.cpp
|
256
|
+
- ext/sources/ggml/src/ggml-cpu/hbm.cpp
|
257
|
+
- ext/sources/ggml/src/ggml-cpu/hbm.h
|
262
258
|
- ext/sources/ggml/src/ggml-cpu/kleidiai/kernels.cpp
|
263
259
|
- ext/sources/ggml/src/ggml-cpu/kleidiai/kernels.h
|
264
260
|
- ext/sources/ggml/src/ggml-cpu/kleidiai/kleidiai.cpp
|
@@ -267,7 +263,17 @@ files:
|
|
267
263
|
- ext/sources/ggml/src/ggml-cpu/llamafile/sgemm.h
|
268
264
|
- ext/sources/ggml/src/ggml-cpu/ops.cpp
|
269
265
|
- ext/sources/ggml/src/ggml-cpu/ops.h
|
266
|
+
- ext/sources/ggml/src/ggml-cpu/quants.c
|
267
|
+
- ext/sources/ggml/src/ggml-cpu/quants.h
|
268
|
+
- ext/sources/ggml/src/ggml-cpu/repack.cpp
|
269
|
+
- ext/sources/ggml/src/ggml-cpu/repack.h
|
270
270
|
- ext/sources/ggml/src/ggml-cpu/simd-mappings.h
|
271
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/ime.cpp
|
272
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/ime.h
|
273
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/ime1_kernels.cpp
|
274
|
+
- ext/sources/ggml/src/ggml-cpu/spacemit/ime_kernels.h
|
275
|
+
- ext/sources/ggml/src/ggml-cpu/traits.cpp
|
276
|
+
- ext/sources/ggml/src/ggml-cpu/traits.h
|
271
277
|
- ext/sources/ggml/src/ggml-cpu/unary-ops.cpp
|
272
278
|
- ext/sources/ggml/src/ggml-cpu/unary-ops.h
|
273
279
|
- ext/sources/ggml/src/ggml-cpu/vec.cpp
|
@@ -275,6 +281,8 @@ files:
|
|
275
281
|
- ext/sources/ggml/src/ggml-cuda/CMakeLists.txt
|
276
282
|
- ext/sources/ggml/src/ggml-cuda/acc.cu
|
277
283
|
- ext/sources/ggml/src/ggml-cuda/acc.cuh
|
284
|
+
- ext/sources/ggml/src/ggml-cuda/add-id.cu
|
285
|
+
- ext/sources/ggml/src/ggml-cuda/add-id.cuh
|
278
286
|
- ext/sources/ggml/src/ggml-cuda/arange.cu
|
279
287
|
- ext/sources/ggml/src/ggml-cuda/arange.cuh
|
280
288
|
- ext/sources/ggml/src/ggml-cuda/argmax.cu
|
@@ -290,11 +298,18 @@ files:
|
|
290
298
|
- ext/sources/ggml/src/ggml-cuda/concat.cuh
|
291
299
|
- ext/sources/ggml/src/ggml-cuda/conv-transpose-1d.cu
|
292
300
|
- ext/sources/ggml/src/ggml-cuda/conv-transpose-1d.cuh
|
301
|
+
- ext/sources/ggml/src/ggml-cuda/conv2d-dw.cu
|
302
|
+
- ext/sources/ggml/src/ggml-cuda/conv2d-dw.cuh
|
303
|
+
- ext/sources/ggml/src/ggml-cuda/conv2d-transpose.cu
|
304
|
+
- ext/sources/ggml/src/ggml-cuda/conv2d-transpose.cuh
|
305
|
+
- ext/sources/ggml/src/ggml-cuda/conv2d.cu
|
306
|
+
- ext/sources/ggml/src/ggml-cuda/conv2d.cuh
|
293
307
|
- ext/sources/ggml/src/ggml-cuda/convert.cu
|
294
308
|
- ext/sources/ggml/src/ggml-cuda/convert.cuh
|
295
309
|
- ext/sources/ggml/src/ggml-cuda/count-equal.cu
|
296
310
|
- ext/sources/ggml/src/ggml-cuda/count-equal.cuh
|
297
311
|
- ext/sources/ggml/src/ggml-cuda/cp-async.cuh
|
312
|
+
- ext/sources/ggml/src/ggml-cuda/cpy-utils.cuh
|
298
313
|
- ext/sources/ggml/src/ggml-cuda/cpy.cu
|
299
314
|
- ext/sources/ggml/src/ggml-cuda/cpy.cuh
|
300
315
|
- ext/sources/ggml/src/ggml-cuda/cross-entropy-loss.cu
|
@@ -304,12 +319,9 @@ files:
|
|
304
319
|
- ext/sources/ggml/src/ggml-cuda/diagmask.cuh
|
305
320
|
- ext/sources/ggml/src/ggml-cuda/fattn-common.cuh
|
306
321
|
- ext/sources/ggml/src/ggml-cuda/fattn-mma-f16.cuh
|
307
|
-
- ext/sources/ggml/src/ggml-cuda/fattn-tile
|
308
|
-
- ext/sources/ggml/src/ggml-cuda/fattn-tile
|
309
|
-
- ext/sources/ggml/src/ggml-cuda/fattn-
|
310
|
-
- ext/sources/ggml/src/ggml-cuda/fattn-tile-f32.cuh
|
311
|
-
- ext/sources/ggml/src/ggml-cuda/fattn-vec-f16.cuh
|
312
|
-
- ext/sources/ggml/src/ggml-cuda/fattn-vec-f32.cuh
|
322
|
+
- ext/sources/ggml/src/ggml-cuda/fattn-tile.cu
|
323
|
+
- ext/sources/ggml/src/ggml-cuda/fattn-tile.cuh
|
324
|
+
- ext/sources/ggml/src/ggml-cuda/fattn-vec.cuh
|
313
325
|
- ext/sources/ggml/src/ggml-cuda/fattn-wmma-f16.cu
|
314
326
|
- ext/sources/ggml/src/ggml-cuda/fattn-wmma-f16.cuh
|
315
327
|
- ext/sources/ggml/src/ggml-cuda/fattn.cu
|
@@ -321,29 +333,44 @@ files:
|
|
321
333
|
- ext/sources/ggml/src/ggml-cuda/gla.cuh
|
322
334
|
- ext/sources/ggml/src/ggml-cuda/im2col.cu
|
323
335
|
- ext/sources/ggml/src/ggml-cuda/im2col.cuh
|
336
|
+
- ext/sources/ggml/src/ggml-cuda/mean.cu
|
337
|
+
- ext/sources/ggml/src/ggml-cuda/mean.cuh
|
324
338
|
- ext/sources/ggml/src/ggml-cuda/mma.cuh
|
339
|
+
- ext/sources/ggml/src/ggml-cuda/mmf.cu
|
340
|
+
- ext/sources/ggml/src/ggml-cuda/mmf.cuh
|
325
341
|
- ext/sources/ggml/src/ggml-cuda/mmq.cu
|
326
342
|
- ext/sources/ggml/src/ggml-cuda/mmq.cuh
|
327
|
-
- ext/sources/ggml/src/ggml-cuda/
|
328
|
-
- ext/sources/ggml/src/ggml-cuda/
|
343
|
+
- ext/sources/ggml/src/ggml-cuda/mmvf.cu
|
344
|
+
- ext/sources/ggml/src/ggml-cuda/mmvf.cuh
|
329
345
|
- ext/sources/ggml/src/ggml-cuda/mmvq.cu
|
330
346
|
- ext/sources/ggml/src/ggml-cuda/mmvq.cuh
|
331
347
|
- ext/sources/ggml/src/ggml-cuda/norm.cu
|
332
348
|
- ext/sources/ggml/src/ggml-cuda/norm.cuh
|
333
349
|
- ext/sources/ggml/src/ggml-cuda/opt-step-adamw.cu
|
334
350
|
- ext/sources/ggml/src/ggml-cuda/opt-step-adamw.cuh
|
351
|
+
- ext/sources/ggml/src/ggml-cuda/opt-step-sgd.cu
|
352
|
+
- ext/sources/ggml/src/ggml-cuda/opt-step-sgd.cuh
|
335
353
|
- ext/sources/ggml/src/ggml-cuda/out-prod.cu
|
336
354
|
- ext/sources/ggml/src/ggml-cuda/out-prod.cuh
|
337
355
|
- ext/sources/ggml/src/ggml-cuda/pad.cu
|
338
356
|
- ext/sources/ggml/src/ggml-cuda/pad.cuh
|
357
|
+
- ext/sources/ggml/src/ggml-cuda/pad_reflect_1d.cu
|
358
|
+
- ext/sources/ggml/src/ggml-cuda/pad_reflect_1d.cuh
|
339
359
|
- ext/sources/ggml/src/ggml-cuda/pool2d.cu
|
340
360
|
- ext/sources/ggml/src/ggml-cuda/pool2d.cuh
|
341
361
|
- ext/sources/ggml/src/ggml-cuda/quantize.cu
|
342
362
|
- ext/sources/ggml/src/ggml-cuda/quantize.cuh
|
363
|
+
- ext/sources/ggml/src/ggml-cuda/reduce_rows.cuh
|
364
|
+
- ext/sources/ggml/src/ggml-cuda/roll.cu
|
365
|
+
- ext/sources/ggml/src/ggml-cuda/roll.cuh
|
343
366
|
- ext/sources/ggml/src/ggml-cuda/rope.cu
|
344
367
|
- ext/sources/ggml/src/ggml-cuda/rope.cuh
|
345
368
|
- ext/sources/ggml/src/ggml-cuda/scale.cu
|
346
369
|
- ext/sources/ggml/src/ggml-cuda/scale.cuh
|
370
|
+
- ext/sources/ggml/src/ggml-cuda/set-rows.cu
|
371
|
+
- ext/sources/ggml/src/ggml-cuda/set-rows.cuh
|
372
|
+
- ext/sources/ggml/src/ggml-cuda/softcap.cu
|
373
|
+
- ext/sources/ggml/src/ggml-cuda/softcap.cuh
|
347
374
|
- ext/sources/ggml/src/ggml-cuda/softmax.cu
|
348
375
|
- ext/sources/ggml/src/ggml-cuda/softmax.cuh
|
349
376
|
- ext/sources/ggml/src/ggml-cuda/ssm-conv.cu
|
@@ -373,93 +400,59 @@ files:
|
|
373
400
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_8-ncols2_2.cu
|
374
401
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_8-ncols2_4.cu
|
375
402
|
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-mma-f16-instance-ncols1_8-ncols2_8.cu
|
376
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
377
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
378
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
379
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
380
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
381
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
382
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
383
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
384
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
385
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
386
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
387
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
388
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
389
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
390
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
391
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
392
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
393
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
394
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
395
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
396
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
397
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
398
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
399
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
400
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
401
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
402
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
403
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
404
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
405
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
406
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
407
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
408
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
409
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
410
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
411
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-
|
412
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs256-f16-f16.cu
|
413
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-f16.cu
|
414
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_0.cu
|
415
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_1.cu
|
416
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_0.cu
|
417
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_1.cu
|
418
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q8_0.cu
|
419
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-f16.cu
|
420
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_0.cu
|
421
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_1.cu
|
422
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_0.cu
|
423
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_1.cu
|
424
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q8_0.cu
|
425
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-f16.cu
|
426
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_0.cu
|
427
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_1.cu
|
428
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_0.cu
|
429
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_1.cu
|
430
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q8_0.cu
|
431
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-f16.cu
|
432
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_0.cu
|
433
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_1.cu
|
434
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_0.cu
|
435
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_1.cu
|
436
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q8_0.cu
|
437
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-f16.cu
|
438
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_0.cu
|
439
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_1.cu
|
440
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_0.cu
|
441
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_1.cu
|
442
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q8_0.cu
|
443
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-f16.cu
|
444
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_0.cu
|
445
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_1.cu
|
446
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_0.cu
|
447
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_1.cu
|
448
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q8_0.cu
|
449
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-f16.cu
|
450
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_0.cu
|
451
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_1.cu
|
452
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_0.cu
|
453
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_1.cu
|
454
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q8_0.cu
|
455
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs256-f16-f16.cu
|
456
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-f16.cu
|
457
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_0.cu
|
458
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_1.cu
|
459
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_0.cu
|
460
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_1.cu
|
461
|
-
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q8_0.cu
|
403
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-f16.cu
|
404
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q4_0.cu
|
405
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q4_1.cu
|
406
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q5_0.cu
|
407
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q5_1.cu
|
408
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-f16-q8_0.cu
|
409
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-f16.cu
|
410
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q4_0.cu
|
411
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q4_1.cu
|
412
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q5_0.cu
|
413
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q5_1.cu
|
414
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_0-q8_0.cu
|
415
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-f16.cu
|
416
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q4_0.cu
|
417
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q4_1.cu
|
418
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q5_0.cu
|
419
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q5_1.cu
|
420
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q4_1-q8_0.cu
|
421
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-f16.cu
|
422
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q4_0.cu
|
423
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q4_1.cu
|
424
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q5_0.cu
|
425
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q5_1.cu
|
426
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_0-q8_0.cu
|
427
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-f16.cu
|
428
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q4_0.cu
|
429
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q4_1.cu
|
430
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q5_0.cu
|
431
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q5_1.cu
|
432
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q5_1-q8_0.cu
|
433
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-f16.cu
|
434
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q4_0.cu
|
435
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q4_1.cu
|
436
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q5_0.cu
|
437
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q5_1.cu
|
438
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/fattn-vec-instance-q8_0-q8_0.cu
|
462
439
|
- ext/sources/ggml/src/ggml-cuda/template-instances/generate_cu_files.py
|
440
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_1.cu
|
441
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_10.cu
|
442
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_11.cu
|
443
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_12.cu
|
444
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_13.cu
|
445
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_14.cu
|
446
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_15.cu
|
447
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_16.cu
|
448
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_2.cu
|
449
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_3.cu
|
450
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_4.cu
|
451
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_5.cu
|
452
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_6.cu
|
453
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_7.cu
|
454
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_8.cu
|
455
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmf-instance-ncols_9.cu
|
463
456
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-iq1_s.cu
|
464
457
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_s.cu
|
465
458
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_xs.cu
|
@@ -468,6 +461,7 @@ files:
|
|
468
461
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-iq3_xxs.cu
|
469
462
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-iq4_nl.cu
|
470
463
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-iq4_xs.cu
|
464
|
+
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-mxfp4.cu
|
471
465
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu
|
472
466
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q3_k.cu
|
473
467
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q4_0.cu
|
@@ -478,6 +472,8 @@ files:
|
|
478
472
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q5_k.cu
|
479
473
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q6_k.cu
|
480
474
|
- ext/sources/ggml/src/ggml-cuda/template-instances/mmq-instance-q8_0.cu
|
475
|
+
- ext/sources/ggml/src/ggml-cuda/topk-moe.cu
|
476
|
+
- ext/sources/ggml/src/ggml-cuda/topk-moe.cuh
|
481
477
|
- ext/sources/ggml/src/ggml-cuda/tsembd.cu
|
482
478
|
- ext/sources/ggml/src/ggml-cuda/tsembd.cuh
|
483
479
|
- ext/sources/ggml/src/ggml-cuda/unary.cu
|
@@ -492,48 +488,18 @@ files:
|
|
492
488
|
- ext/sources/ggml/src/ggml-cuda/wkv.cuh
|
493
489
|
- ext/sources/ggml/src/ggml-hip/CMakeLists.txt
|
494
490
|
- ext/sources/ggml/src/ggml-impl.h
|
495
|
-
- ext/sources/ggml/src/ggml-kompute/CMakeLists.txt
|
496
|
-
- ext/sources/ggml/src/ggml-kompute/ggml-kompute.cpp
|
497
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/common.comp
|
498
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_add.comp
|
499
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_addrow.comp
|
500
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_cpy_f16_f16.comp
|
501
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_cpy_f16_f32.comp
|
502
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_cpy_f32_f16.comp
|
503
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_cpy_f32_f32.comp
|
504
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_diagmask.comp
|
505
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_gelu.comp
|
506
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows.comp
|
507
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_f16.comp
|
508
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_f32.comp
|
509
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_q4_0.comp
|
510
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_q4_1.comp
|
511
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_getrows_q6_k.comp
|
512
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul.comp
|
513
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_f16.comp
|
514
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_mat_f32.comp
|
515
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q4_0.comp
|
516
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q4_1.comp
|
517
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q4_k.comp
|
518
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q6_k.comp
|
519
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mat_q8_0.comp
|
520
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mv_q_n.comp
|
521
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_mul_mv_q_n_pre.comp
|
522
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_norm.comp
|
523
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_relu.comp
|
524
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rmsnorm.comp
|
525
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rope_neox_f16.comp
|
526
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rope_neox_f32.comp
|
527
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rope_norm_f16.comp
|
528
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_rope_norm_f32.comp
|
529
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_scale.comp
|
530
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_scale_8.comp
|
531
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_silu.comp
|
532
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/op_softmax.comp
|
533
|
-
- ext/sources/ggml/src/ggml-kompute/kompute-shaders/rope_common.comp
|
534
491
|
- ext/sources/ggml/src/ggml-metal/CMakeLists.txt
|
492
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-common.cpp
|
493
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-common.h
|
494
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-context.h
|
495
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-context.m
|
496
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-device.cpp
|
497
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-device.h
|
498
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-device.m
|
535
499
|
- ext/sources/ggml/src/ggml-metal/ggml-metal-impl.h
|
536
|
-
- ext/sources/ggml/src/ggml-metal/ggml-metal.
|
500
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-ops.cpp
|
501
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal-ops.h
|
502
|
+
- ext/sources/ggml/src/ggml-metal/ggml-metal.cpp
|
537
503
|
- ext/sources/ggml/src/ggml-metal/ggml-metal.metal
|
538
504
|
- ext/sources/ggml/src/ggml-musa/CMakeLists.txt
|
539
505
|
- ext/sources/ggml/src/ggml-musa/mudnn.cu
|
@@ -541,41 +507,73 @@ files:
|
|
541
507
|
- ext/sources/ggml/src/ggml-opencl/CMakeLists.txt
|
542
508
|
- ext/sources/ggml/src/ggml-opencl/ggml-opencl.cpp
|
543
509
|
- ext/sources/ggml/src/ggml-opencl/kernels/add.cl
|
510
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/add_id.cl
|
511
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/argsort.cl
|
544
512
|
- ext/sources/ggml/src/ggml-opencl/kernels/clamp.cl
|
513
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/concat.cl
|
514
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/conv2d.cl
|
515
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/conv2d_f16_f32.cl
|
545
516
|
- ext/sources/ggml/src/ggml-opencl/kernels/cpy.cl
|
546
517
|
- ext/sources/ggml/src/ggml-opencl/kernels/cvt.cl
|
547
518
|
- ext/sources/ggml/src/ggml-opencl/kernels/diag_mask_inf.cl
|
519
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/div.cl
|
548
520
|
- ext/sources/ggml/src/ggml-opencl/kernels/embed_kernel.py
|
521
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f16.cl
|
522
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f32.cl
|
523
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/flash_attn_f32_f16.cl
|
549
524
|
- ext/sources/ggml/src/ggml-opencl/kernels/gelu.cl
|
550
525
|
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle.cl
|
551
526
|
- ext/sources/ggml/src/ggml-opencl/kernels/gemv_noshuffle_general.cl
|
552
527
|
- ext/sources/ggml/src/ggml-opencl/kernels/get_rows.cl
|
528
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/glu.cl
|
529
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/group_norm.cl
|
553
530
|
- ext/sources/ggml/src/ggml-opencl/kernels/im2col_f16.cl
|
554
531
|
- ext/sources/ggml/src/ggml-opencl/kernels/im2col_f32.cl
|
555
532
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul.cl
|
556
533
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mat_Ab_Bi_8x4.cl
|
534
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mat_f16_f32.cl
|
535
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_f16_f32_l4_lm.cl
|
536
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mm_f32_f32_l4_lm.cl
|
557
537
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_f16_f16.cl
|
558
538
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_f16_f32.cl
|
559
539
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_f16_f32_1row.cl
|
560
540
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_f16_f32_l4.cl
|
561
541
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_f32_f32.cl
|
542
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_mxfp4_f32.cl
|
543
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_mxfp4_f32_flat.cl
|
544
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q4_0_f32_8x_flat.cl
|
545
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q8_0_f32.cl
|
546
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_id_q8_0_f32_flat.cl
|
547
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_mxfp4_f32.cl
|
548
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_mxfp4_f32_flat.cl
|
562
549
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_0_f32.cl
|
563
550
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_0_f32_1d_16x_flat.cl
|
564
551
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_0_f32_1d_8x_flat.cl
|
565
552
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_0_f32_8x_flat.cl
|
566
553
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q4_0_f32_v.cl
|
567
554
|
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q6_k.cl
|
555
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q8_0_f32.cl
|
556
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/mul_mv_q8_0_f32_flat.cl
|
568
557
|
- ext/sources/ggml/src/ggml-opencl/kernels/norm.cl
|
558
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/pad.cl
|
569
559
|
- ext/sources/ggml/src/ggml-opencl/kernels/relu.cl
|
560
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/repeat.cl
|
570
561
|
- ext/sources/ggml/src/ggml-opencl/kernels/rms_norm.cl
|
571
562
|
- ext/sources/ggml/src/ggml-opencl/kernels/rope.cl
|
572
563
|
- ext/sources/ggml/src/ggml-opencl/kernels/scale.cl
|
564
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/set_rows.cl
|
565
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/sigmoid.cl
|
573
566
|
- ext/sources/ggml/src/ggml-opencl/kernels/silu.cl
|
574
567
|
- ext/sources/ggml/src/ggml-opencl/kernels/softmax_4_f16.cl
|
575
568
|
- ext/sources/ggml/src/ggml-opencl/kernels/softmax_4_f32.cl
|
576
569
|
- ext/sources/ggml/src/ggml-opencl/kernels/softmax_f16.cl
|
577
570
|
- ext/sources/ggml/src/ggml-opencl/kernels/softmax_f32.cl
|
571
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/sub.cl
|
572
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/sum_rows.cl
|
573
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/tanh.cl
|
578
574
|
- ext/sources/ggml/src/ggml-opencl/kernels/transpose.cl
|
575
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/tsembd.cl
|
576
|
+
- ext/sources/ggml/src/ggml-opencl/kernels/upscale.cl
|
579
577
|
- ext/sources/ggml/src/ggml-opt.cpp
|
580
578
|
- ext/sources/ggml/src/ggml-quants.c
|
581
579
|
- ext/sources/ggml/src/ggml-quants.h
|
@@ -618,9 +616,12 @@ files:
|
|
618
616
|
- ext/sources/ggml/src/ggml-sycl/outprod.cpp
|
619
617
|
- ext/sources/ggml/src/ggml-sycl/outprod.hpp
|
620
618
|
- ext/sources/ggml/src/ggml-sycl/presets.hpp
|
619
|
+
- ext/sources/ggml/src/ggml-sycl/quantize.hpp
|
621
620
|
- ext/sources/ggml/src/ggml-sycl/quants.hpp
|
622
621
|
- ext/sources/ggml/src/ggml-sycl/rope.cpp
|
623
622
|
- ext/sources/ggml/src/ggml-sycl/rope.hpp
|
623
|
+
- ext/sources/ggml/src/ggml-sycl/set_rows.cpp
|
624
|
+
- ext/sources/ggml/src/ggml-sycl/set_rows.hpp
|
624
625
|
- ext/sources/ggml/src/ggml-sycl/softmax.cpp
|
625
626
|
- ext/sources/ggml/src/ggml-sycl/softmax.hpp
|
626
627
|
- ext/sources/ggml/src/ggml-sycl/sycl_hw.cpp
|
@@ -638,12 +639,15 @@ files:
|
|
638
639
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/CMakeLists.txt
|
639
640
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/acc.comp
|
640
641
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/add.comp
|
642
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/add_id.comp
|
641
643
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/argmax.comp
|
642
644
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/argsort.comp
|
643
645
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/clamp.comp
|
644
646
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/concat.comp
|
645
647
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/contig_copy.comp
|
646
648
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/conv2d_dw.comp
|
649
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/conv2d_mm.comp
|
650
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/conv_transpose_1d.comp
|
647
651
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/copy.comp
|
648
652
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/copy_from_quant.comp
|
649
653
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/copy_to_quant.comp
|
@@ -662,6 +666,7 @@ files:
|
|
662
666
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq3_xxs.comp
|
663
667
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq4_nl.comp
|
664
668
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_iq4_xs.comp
|
669
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_mxfp4.comp
|
665
670
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q2_k.comp
|
666
671
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q3_k.comp
|
667
672
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q4_0.comp
|
@@ -674,20 +679,30 @@ files:
|
|
674
679
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/dequant_q8_0.comp
|
675
680
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/diag_mask_inf.comp
|
676
681
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/div.comp
|
682
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/exp.comp
|
677
683
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn.comp
|
678
684
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_base.comp
|
679
685
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp
|
680
686
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm2.comp
|
681
687
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_split_k_reduce.comp
|
688
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/geglu.comp
|
689
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/geglu_erf.comp
|
690
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/geglu_quick.comp
|
682
691
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/gelu.comp
|
692
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/gelu_erf.comp
|
683
693
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/gelu_quick.comp
|
684
694
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/generic_binary_head.comp
|
685
695
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/generic_head.comp
|
686
696
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/generic_unary_head.comp
|
687
697
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/get_rows.comp
|
688
698
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/get_rows_quant.comp
|
699
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/glu_head.comp
|
700
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/glu_main.comp
|
689
701
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/group_norm.comp
|
702
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/hardsigmoid.comp
|
703
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/hardswish.comp
|
690
704
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/im2col.comp
|
705
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/im2col_3d.comp
|
691
706
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/l2_norm.comp
|
692
707
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/leaky_relu.comp
|
693
708
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul.comp
|
@@ -708,25 +723,33 @@ files:
|
|
708
723
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q4_k.comp
|
709
724
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q5_k.comp
|
710
725
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q6_k.comp
|
726
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vecq.comp
|
711
727
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm.comp
|
712
728
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_cm2.comp
|
729
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_funcs.comp
|
713
730
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq.comp
|
714
731
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq_funcs.comp
|
732
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/multi_add.comp
|
715
733
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/norm.comp
|
716
734
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/opt_step_adamw.comp
|
735
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/opt_step_sgd.comp
|
717
736
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/pad.comp
|
718
737
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/pool2d.comp
|
719
738
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/quantize_q8_1.comp
|
739
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/reglu.comp
|
720
740
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/relu.comp
|
721
741
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/repeat.comp
|
722
742
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/repeat_back.comp
|
723
743
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rms_norm.comp
|
724
744
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rms_norm_back.comp
|
745
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rms_norm_partials.comp
|
746
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/roll.comp
|
725
747
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_head.comp
|
726
748
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_multi.comp
|
727
749
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_neox.comp
|
728
750
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_norm.comp
|
729
751
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rope_vision.comp
|
752
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/rte.comp
|
730
753
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/scale.comp
|
731
754
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sigmoid.comp
|
732
755
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/silu.comp
|
@@ -734,9 +757,12 @@ files:
|
|
734
757
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sin.comp
|
735
758
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/soft_max.comp
|
736
759
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/soft_max_back.comp
|
760
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sqrt.comp
|
737
761
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/square.comp
|
738
762
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sub.comp
|
739
763
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/sum_rows.comp
|
764
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/swiglu.comp
|
765
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/swiglu_oai.comp
|
740
766
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/tanh.comp
|
741
767
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/test_bfloat16_support.comp
|
742
768
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/test_coopmat2_support.comp
|
@@ -745,13 +771,40 @@ files:
|
|
745
771
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/timestep_embedding.comp
|
746
772
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/types.comp
|
747
773
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/upscale.comp
|
774
|
+
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/utils.comp
|
748
775
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp
|
749
776
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/wkv6.comp
|
750
777
|
- ext/sources/ggml/src/ggml-vulkan/vulkan-shaders/wkv7.comp
|
778
|
+
- ext/sources/ggml/src/ggml-webgpu/CMakeLists.txt
|
779
|
+
- ext/sources/ggml/src/ggml-webgpu/ggml-webgpu.cpp
|
780
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/add.tmpl.wgsl
|
781
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/add_in_place.tmpl.wgsl
|
782
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/binary_head.tmpl
|
783
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/common_decls.tmpl
|
784
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/cpy.wgsl
|
785
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/embed_wgsl.py
|
786
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/get_rows.tmpl.wgsl
|
787
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/memset.wgsl
|
788
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul.tmpl.wgsl
|
789
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_in_place.tmpl.wgsl
|
790
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat.tmpl.wgsl
|
791
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rms_norm.wgsl
|
792
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/rms_norm_in_place.wgsl
|
793
|
+
- ext/sources/ggml/src/ggml-webgpu/wgsl-shaders/set_rows.wgsl
|
794
|
+
- ext/sources/ggml/src/ggml-zdnn/CMakeLists.txt
|
795
|
+
- ext/sources/ggml/src/ggml-zdnn/common.hpp
|
796
|
+
- ext/sources/ggml/src/ggml-zdnn/ggml-zdnn.cpp
|
797
|
+
- ext/sources/ggml/src/ggml-zdnn/mmf.cpp
|
798
|
+
- ext/sources/ggml/src/ggml-zdnn/mmf.hpp
|
799
|
+
- ext/sources/ggml/src/ggml-zdnn/utils.cpp
|
800
|
+
- ext/sources/ggml/src/ggml-zdnn/utils.hpp
|
751
801
|
- ext/sources/ggml/src/ggml.c
|
802
|
+
- ext/sources/ggml/src/ggml.cpp
|
752
803
|
- ext/sources/ggml/src/gguf.cpp
|
753
804
|
- ext/sources/include/whisper.h
|
754
805
|
- ext/sources/src/CMakeLists.txt
|
806
|
+
- ext/sources/src/coreml/whisper-compat.h
|
807
|
+
- ext/sources/src/coreml/whisper-compat.m
|
755
808
|
- ext/sources/src/coreml/whisper-decoder-impl.h
|
756
809
|
- ext/sources/src/coreml/whisper-decoder-impl.m
|
757
810
|
- ext/sources/src/coreml/whisper-encoder-impl.h
|
@@ -787,21 +840,23 @@ files:
|
|
787
840
|
- ext/sources/tests/test-vad.cpp
|
788
841
|
- ext/sources/tests/test-whisper.js
|
789
842
|
- extsources.rb
|
843
|
+
- lib/whisper/context.rb
|
790
844
|
- lib/whisper/model/uri.rb
|
845
|
+
- lib/whisper/segment.rb
|
791
846
|
- sig/whisper.rbs
|
792
|
-
-
|
793
|
-
-
|
794
|
-
-
|
795
|
-
-
|
796
|
-
-
|
797
|
-
-
|
798
|
-
-
|
799
|
-
-
|
800
|
-
-
|
801
|
-
-
|
802
|
-
-
|
803
|
-
-
|
804
|
-
-
|
847
|
+
- test/helper.rb
|
848
|
+
- test/jfk_reader/.gitignore
|
849
|
+
- test/jfk_reader/extconf.rb
|
850
|
+
- test/jfk_reader/jfk_reader.c
|
851
|
+
- test/test_callback.rb
|
852
|
+
- test/test_error.rb
|
853
|
+
- test/test_model.rb
|
854
|
+
- test/test_package.rb
|
855
|
+
- test/test_params.rb
|
856
|
+
- test/test_segment.rb
|
857
|
+
- test/test_vad.rb
|
858
|
+
- test/test_vad_params.rb
|
859
|
+
- test/test_whisper.rb
|
805
860
|
- whispercpp.gemspec
|
806
861
|
homepage: https://github.com/ggml-org/whisper.cpp
|
807
862
|
licenses:
|
@@ -829,16 +884,16 @@ signing_key:
|
|
829
884
|
specification_version: 4
|
830
885
|
summary: Ruby whisper.cpp bindings
|
831
886
|
test_files:
|
832
|
-
-
|
833
|
-
-
|
834
|
-
-
|
835
|
-
-
|
836
|
-
-
|
837
|
-
-
|
838
|
-
-
|
839
|
-
-
|
840
|
-
-
|
841
|
-
-
|
842
|
-
-
|
843
|
-
-
|
844
|
-
-
|
887
|
+
- test/helper.rb
|
888
|
+
- test/jfk_reader/.gitignore
|
889
|
+
- test/jfk_reader/extconf.rb
|
890
|
+
- test/jfk_reader/jfk_reader.c
|
891
|
+
- test/test_callback.rb
|
892
|
+
- test/test_error.rb
|
893
|
+
- test/test_model.rb
|
894
|
+
- test/test_package.rb
|
895
|
+
- test/test_params.rb
|
896
|
+
- test/test_segment.rb
|
897
|
+
- test/test_vad.rb
|
898
|
+
- test/test_vad_params.rb
|
899
|
+
- test/test_whisper.rb
|