@novastera-oss/llamarn 0.3.1 → 0.4.1
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.
- package/README.md +86 -3
- package/RNLlamaCpp.podspec +1 -1
- package/android/CMakeLists.txt +11 -3
- package/android/generated/jni/react/renderer/components/RNLlamaCppSpec/RNLlamaCppSpecJSI.h +49 -4
- package/android/src/main/cpp/include/llama.h +53 -114
- package/android/src/main/jniLibs/arm64-v8a/libggml-base.so +0 -0
- package/android/src/main/jniLibs/arm64-v8a/libggml-cpu.so +0 -0
- package/android/src/main/jniLibs/arm64-v8a/libggml.so +0 -0
- package/android/src/main/jniLibs/arm64-v8a/libllama.so +0 -0
- package/android/src/main/jniLibs/armeabi-v7a/libggml-base.so +0 -0
- package/android/src/main/jniLibs/armeabi-v7a/libggml-cpu.so +0 -0
- package/android/src/main/jniLibs/armeabi-v7a/libggml.so +0 -0
- package/android/src/main/jniLibs/armeabi-v7a/libllama.so +0 -0
- package/android/src/main/jniLibs/x86/libggml-base.so +0 -0
- package/android/src/main/jniLibs/x86/libggml-cpu.so +0 -0
- package/android/src/main/jniLibs/x86/libggml.so +0 -0
- package/android/src/main/jniLibs/x86/libllama.so +0 -0
- package/android/src/main/jniLibs/x86_64/libggml-base.so +0 -0
- package/android/src/main/jniLibs/x86_64/libggml-cpu.so +0 -0
- package/android/src/main/jniLibs/x86_64/libggml.so +0 -0
- package/android/src/main/jniLibs/x86_64/libllama.so +0 -0
- package/cpp/LlamaCppModel.cpp +2 -10
- package/cpp/PureCppImpl.cpp +71 -4
- package/cpp/SystemUtils.cpp +3 -7
- package/cpp/build-info.cpp +2 -2
- package/cpp/llama.cpp/CMakeLists.txt +2 -0
- package/cpp/llama.cpp/CODEOWNERS +1 -1
- package/cpp/llama.cpp/Makefile +6 -1605
- package/cpp/llama.cpp/README.md +5 -1
- package/cpp/llama.cpp/common/arg.cpp +230 -51
- package/cpp/llama.cpp/common/chat-parser.cpp +9 -1
- package/cpp/llama.cpp/common/chat.cpp +539 -8
- package/cpp/llama.cpp/common/chat.h +8 -1
- package/cpp/llama.cpp/common/common.cpp +60 -15
- package/cpp/llama.cpp/common/common.h +64 -15
- package/cpp/llama.cpp/common/speculative.cpp +135 -54
- package/cpp/llama.cpp/common/speculative.h +8 -1
- package/cpp/llama.cpp/convert_hf_to_gguf.py +1216 -109
- package/cpp/llama.cpp/convert_hf_to_gguf_update.py +19 -6
- package/cpp/llama.cpp/convert_lora_to_gguf.py +1 -1
- package/cpp/llama.cpp/flake.nix +0 -5
- package/cpp/llama.cpp/ggml/CMakeLists.txt +6 -3
- package/cpp/llama.cpp/ggml/cmake/ggml-config.cmake.in +71 -70
- package/cpp/llama.cpp/ggml/include/ggml-opt.h +25 -6
- package/cpp/llama.cpp/ggml/include/ggml-zdnn.h +16 -0
- package/cpp/llama.cpp/ggml/include/ggml.h +90 -3
- package/cpp/llama.cpp/ggml/src/CMakeLists.txt +13 -1
- package/cpp/llama.cpp/ggml/src/ggml-alloc.c +1 -0
- package/cpp/llama.cpp/ggml/src/ggml-backend-reg.cpp +10 -0
- package/cpp/llama.cpp/ggml/src/ggml-backend.cpp +113 -17
- package/cpp/llama.cpp/ggml/src/ggml-blas/ggml-blas.cpp +4 -4
- package/cpp/llama.cpp/ggml/src/ggml-cann/CMakeLists.txt +14 -0
- package/cpp/llama.cpp/ggml/src/ggml-cann/aclnn_ops.cpp +701 -585
- package/cpp/llama.cpp/ggml/src/ggml-cann/aclnn_ops.h +13 -3
- package/cpp/llama.cpp/ggml/src/ggml-cann/common.h +52 -0
- package/cpp/llama.cpp/ggml/src/ggml-cann/ggml-cann.cpp +274 -91
- package/cpp/llama.cpp/ggml/src/ggml-common.h +17 -0
- package/cpp/llama.cpp/ggml/src/ggml-cpu/CMakeLists.txt +2 -2
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/arm/quants.c +132 -596
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/arm/repack.cpp +14 -286
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/loongarch/quants.c +90 -569
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/powerpc/quants.c +162 -589
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/riscv/quants.c +55 -341
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/riscv/repack.cpp +3 -58
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/s390/quants.c +371 -298
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/wasm/quants.c +54 -314
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/x86/quants.c +184 -675
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch/x86/repack.cpp +4679 -1657
- package/cpp/llama.cpp/ggml/src/ggml-cpu/arch-fallback.h +33 -2
- package/cpp/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-impl.h +8 -0
- package/cpp/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c +26 -1
- package/cpp/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.cpp +21 -24
- package/cpp/llama.cpp/ggml/src/ggml-cpu/kleidiai/kleidiai.cpp +16 -7
- package/cpp/llama.cpp/ggml/src/ggml-cpu/llamafile/sgemm.cpp +232 -123
- package/cpp/llama.cpp/ggml/src/ggml-cpu/ops.cpp +428 -23
- package/cpp/llama.cpp/ggml/src/ggml-cpu/ops.h +4 -8
- package/cpp/llama.cpp/ggml/src/ggml-cpu/quants.c +35 -0
- package/cpp/llama.cpp/ggml/src/ggml-cpu/quants.h +8 -0
- package/cpp/llama.cpp/ggml/src/ggml-cpu/repack.cpp +458 -46
- package/cpp/llama.cpp/ggml/src/ggml-cpu/repack.h +22 -0
- package/cpp/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h +39 -14
- package/cpp/llama.cpp/ggml/src/ggml-cpu/traits.cpp +2 -2
- package/cpp/llama.cpp/ggml/src/ggml-cpu/traits.h +1 -1
- package/cpp/llama.cpp/ggml/src/ggml-cpu/vec.cpp +20 -1
- package/cpp/llama.cpp/ggml/src/ggml-cpu/vec.h +122 -5
- package/cpp/llama.cpp/ggml/src/ggml-cuda/CMakeLists.txt +9 -11
- package/cpp/llama.cpp/ggml/src/ggml-cuda/add-id.cu +58 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/add-id.cuh +3 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/binbcast.cu +275 -170
- package/cpp/llama.cpp/ggml/src/ggml-cuda/binbcast.cuh +2 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/common.cuh +103 -65
- package/cpp/llama.cpp/ggml/src/ggml-cuda/conv-transpose-1d.cu +1 -4
- package/cpp/llama.cpp/ggml/src/ggml-cuda/conv2d.cu +171 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/conv2d.cuh +5 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/convert.cu +33 -7
- package/cpp/llama.cpp/ggml/src/ggml-cuda/convert.cuh +13 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/cpy-utils.cuh +2 -10
- package/cpp/llama.cpp/ggml/src/ggml-cuda/cpy.cu +3 -4
- package/cpp/llama.cpp/ggml/src/ggml-cuda/dequantize.cuh +14 -40
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn-common.cuh +83 -27
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn-mma-f16.cuh +116 -57
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn-tile-f16.cu +45 -18
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn-tile-f32.cu +56 -29
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn-vec-f16.cuh +61 -39
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn-vec-f32.cuh +70 -49
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn-wmma-f16.cu +70 -21
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn.cu +162 -50
- package/cpp/llama.cpp/ggml/src/ggml-cuda/fattn.cuh +2 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/getrows.cu +5 -4
- package/cpp/llama.cpp/ggml/src/ggml-cuda/ggml-cuda.cu +208 -97
- package/cpp/llama.cpp/ggml/src/ggml-cuda/im2col.cu +46 -35
- package/cpp/llama.cpp/ggml/src/ggml-cuda/mean.cu +56 -2
- package/cpp/llama.cpp/ggml/src/ggml-cuda/mma.cuh +95 -51
- package/cpp/llama.cpp/ggml/src/ggml-cuda/mmf.cu +427 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/mmf.cuh +5 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/mmq.cu +204 -57
- package/cpp/llama.cpp/ggml/src/ggml-cuda/mmq.cuh +252 -168
- package/cpp/llama.cpp/ggml/src/ggml-cuda/{mmv.cu → mmvf.cu} +53 -53
- package/cpp/llama.cpp/ggml/src/ggml-cuda/{mmv.cuh → mmvf.cuh} +3 -3
- package/cpp/llama.cpp/ggml/src/ggml-cuda/mmvq.cu +10 -5
- package/cpp/llama.cpp/ggml/src/ggml-cuda/norm.cu +192 -19
- package/cpp/llama.cpp/ggml/src/ggml-cuda/norm.cuh +5 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/opt-step-sgd.cu +49 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/opt-step-sgd.cuh +5 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/pad_reflect_1d.cu +82 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/pad_reflect_1d.cuh +5 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/reduce_rows.cuh +53 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/roll.cu +67 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/roll.cuh +5 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/set-rows.cu +1 -8
- package/cpp/llama.cpp/ggml/src/ggml-cuda/softcap.cu +34 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/softcap.cuh +5 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/softmax.cu +16 -10
- package/cpp/llama.cpp/ggml/src/ggml-cuda/ssm-scan.cu +153 -71
- package/cpp/llama.cpp/ggml/src/ggml-cuda/sum.cu +6 -10
- package/cpp/llama.cpp/ggml/src/ggml-cuda/sumrows.cu +21 -4
- package/cpp/llama.cpp/ggml/src/ggml-cuda/template-instances/mmq-instance-mxfp4.cu +5 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/unary.cu +75 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/unary.cuh +2 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/vecdotq.cuh +110 -22
- package/cpp/llama.cpp/ggml/src/ggml-cuda/vendors/cuda.h +4 -0
- package/cpp/llama.cpp/ggml/src/ggml-cuda/vendors/hip.h +14 -25
- package/cpp/llama.cpp/ggml/src/ggml-cuda/vendors/musa.h +2 -1
- package/cpp/llama.cpp/ggml/src/ggml-hip/CMakeLists.txt +10 -2
- package/cpp/llama.cpp/ggml/src/ggml-impl.h +61 -0
- package/cpp/llama.cpp/ggml/src/ggml-metal/ggml-metal-impl.h +31 -20
- package/cpp/llama.cpp/ggml/src/ggml-metal/ggml-metal.m +342 -131
- package/cpp/llama.cpp/ggml/src/ggml-metal/ggml-metal.metal +464 -134
- package/cpp/llama.cpp/ggml/src/ggml-musa/CMakeLists.txt +0 -4
- package/cpp/llama.cpp/ggml/src/ggml-opencl/CMakeLists.txt +8 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/ggml-opencl.cpp +1108 -176
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/add.cl +107 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/add_id.cl +42 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/div.cl +66 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/flash_attn_f16.cl +343 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/flash_attn_f32.cl +343 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/flash_attn_f32_f16.cl +346 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/glu.cl +41 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/group_norm.cl +49 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/mul.cl +73 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/mul_mm_f16_f32_l4_lm.cl +132 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/mul_mm_f32_f32_l4_lm.cl +133 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/mul_mv_id_mxfp4_f32.cl +189 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/mul_mv_mxfp4_f32.cl +144 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/norm.cl +80 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/softmax_4_f16.cl +10 -2
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/softmax_4_f32.cl +10 -2
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/softmax_f16.cl +10 -2
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/softmax_f32.cl +10 -2
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/sub.cl +66 -0
- package/cpp/llama.cpp/ggml/src/ggml-opencl/kernels/transpose.cl +20 -0
- package/cpp/llama.cpp/ggml/src/ggml-opt.cpp +97 -41
- package/cpp/llama.cpp/ggml/src/ggml-quants.c +110 -16
- package/cpp/llama.cpp/ggml/src/ggml-quants.h +6 -0
- package/cpp/llama.cpp/ggml/src/ggml-rpc/ggml-rpc.cpp +22 -9
- package/cpp/llama.cpp/ggml/src/ggml-sycl/backend.hpp +1 -0
- package/cpp/llama.cpp/ggml/src/ggml-sycl/cpy.cpp +0 -212
- package/cpp/llama.cpp/ggml/src/ggml-sycl/cpy.hpp +213 -1
- package/cpp/llama.cpp/ggml/src/ggml-sycl/ggml-sycl.cpp +117 -238
- package/cpp/llama.cpp/ggml/src/ggml-sycl/quantize.hpp +133 -0
- package/cpp/llama.cpp/ggml/src/ggml-sycl/set_rows.cpp +94 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/ggml-vulkan.cpp +1666 -633
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/add.comp +41 -1
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/add_id.comp +42 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/argmax.comp +13 -4
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/argsort.comp +39 -29
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/conv2d_mm.comp +107 -43
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/copy_from_quant.comp +2 -2
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/dequant_funcs.comp +18 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/dequant_funcs_cm2.comp +21 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/dequant_mxfp4.comp +32 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/exp.comp +20 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn.comp +21 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_base.comp +16 -1
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp +44 -8
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm2.comp +44 -16
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_split_k_reduce.comp +26 -1
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/generic_binary_head.comp +2 -17
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/glu_head.comp +2 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_base.comp +37 -1
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_nc.comp +11 -7
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm.comp +109 -55
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_cm2.comp +71 -41
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq_funcs.comp +6 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/multi_add.comp +111 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/opt_step_sgd.comp +22 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/rms_norm.comp +49 -11
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/rms_norm_partials.comp +65 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/soft_max.comp +9 -3
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/sqrt.comp +17 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/sum_rows.comp +38 -5
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/swiglu_oai.comp +14 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/types.comp +55 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/utils.comp +25 -0
- package/cpp/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp +75 -20
- package/cpp/llama.cpp/ggml/src/ggml-webgpu/CMakeLists.txt +2 -2
- package/cpp/llama.cpp/ggml/src/ggml-webgpu/ggml-webgpu.cpp +807 -412
- package/cpp/llama.cpp/ggml/src/ggml-webgpu/wgsl-shaders/embed_wgsl.py +72 -22
- package/cpp/llama.cpp/ggml/src/ggml-webgpu/wgsl-shaders/memset.wgsl +8 -8
- package/cpp/llama.cpp/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat.tmpl.wgsl +1794 -0
- package/cpp/llama.cpp/ggml/src/ggml-webgpu/wgsl-shaders/set_rows.wgsl +82 -0
- package/cpp/llama.cpp/ggml/src/ggml-zdnn/CMakeLists.txt +36 -0
- package/cpp/llama.cpp/ggml/src/ggml-zdnn/ggml-zdnn-impl.h +97 -0
- package/cpp/llama.cpp/ggml/src/ggml-zdnn/ggml-zdnn.cpp +846 -0
- package/cpp/llama.cpp/ggml/src/ggml.c +204 -50
- package/cpp/llama.cpp/gguf-py/gguf/constants.py +187 -2
- package/cpp/llama.cpp/gguf-py/gguf/gguf_writer.py +11 -2
- package/cpp/llama.cpp/gguf-py/gguf/quants.py +53 -4
- package/cpp/llama.cpp/gguf-py/gguf/scripts/gguf_convert_endian.py +67 -63
- package/cpp/llama.cpp/gguf-py/gguf/scripts/gguf_new_metadata.py +7 -1
- package/cpp/llama.cpp/gguf-py/gguf/tensor_mapping.py +120 -16
- package/cpp/llama.cpp/gguf-py/gguf/utility.py +5 -1
- package/cpp/llama.cpp/gguf-py/gguf/vocab.py +284 -1
- package/cpp/llama.cpp/gguf-py/tests/test_quants.py +14 -5
- package/cpp/llama.cpp/include/llama.h +53 -114
- package/cpp/llama.cpp/models/templates/ByteDance-Seed-OSS.jinja +171 -0
- package/cpp/llama.cpp/models/templates/README.md +2 -1
- package/cpp/llama.cpp/models/templates/ibm-granite-granite-3.3-2B-Instruct.jinja +59 -0
- package/cpp/llama.cpp/models/templates/openai-gpt-oss-120b.jinja +331 -0
- package/cpp/llama.cpp/models/templates/unsloth-mistral-Devstral-Small-2507.jinja +105 -0
- package/cpp/llama.cpp/requirements/requirements-convert_hf_to_gguf.txt +3 -1
- package/cpp/llama.cpp/requirements/requirements-convert_hf_to_gguf_update.txt +0 -6
- package/cpp/llama.cpp/requirements/requirements-pydantic.txt +1 -1
- package/cpp/llama.cpp/src/CMakeLists.txt +2 -2
- package/cpp/llama.cpp/src/llama-adapter.cpp +68 -4
- package/cpp/llama.cpp/src/llama-adapter.h +3 -0
- package/cpp/llama.cpp/src/llama-arch.cpp +192 -2
- package/cpp/llama.cpp/src/llama-arch.h +18 -0
- package/cpp/llama.cpp/src/llama-batch.cpp +2 -2
- package/cpp/llama.cpp/src/llama-chat.cpp +47 -6
- package/cpp/llama.cpp/src/llama-chat.h +3 -0
- package/cpp/llama.cpp/src/llama-context.cpp +61 -252
- package/cpp/llama.cpp/src/llama-context.h +10 -15
- package/cpp/llama.cpp/src/llama-cparams.h +0 -1
- package/cpp/llama.cpp/src/llama-graph.cpp +180 -85
- package/cpp/llama.cpp/src/llama-graph.h +90 -51
- package/cpp/llama.cpp/src/llama-hparams.cpp +34 -3
- package/cpp/llama.cpp/src/llama-hparams.h +21 -6
- package/cpp/llama.cpp/src/{llama-kv-cache-unified-iswa.cpp → llama-kv-cache-iswa.cpp} +79 -56
- package/cpp/llama.cpp/src/{llama-kv-cache-unified-iswa.h → llama-kv-cache-iswa.h} +30 -28
- package/cpp/llama.cpp/src/{llama-kv-cache-unified.cpp → llama-kv-cache.cpp} +240 -632
- package/cpp/llama.cpp/src/{llama-kv-cache-unified.h → llama-kv-cache.h} +39 -74
- package/cpp/llama.cpp/src/llama-kv-cells.h +21 -21
- package/cpp/llama.cpp/src/llama-memory-hybrid.cpp +41 -35
- package/cpp/llama.cpp/src/llama-memory-hybrid.h +26 -29
- package/cpp/llama.cpp/src/llama-memory-recurrent.cpp +13 -9
- package/cpp/llama.cpp/src/llama-memory-recurrent.h +10 -14
- package/cpp/llama.cpp/src/llama-memory.h +13 -10
- package/cpp/llama.cpp/src/llama-model-loader.cpp +2 -0
- package/cpp/llama.cpp/src/llama-model-loader.h +3 -2
- package/cpp/llama.cpp/src/llama-model.cpp +1959 -419
- package/cpp/llama.cpp/src/llama-model.h +28 -4
- package/cpp/llama.cpp/src/llama-quant.cpp +40 -4
- package/cpp/llama.cpp/src/llama-vocab.cpp +51 -2
- package/cpp/llama.cpp/src/llama-vocab.h +1 -0
- package/cpp/llama.cpp/vendor/minja/chat-template.hpp +16 -7
- package/cpp/llama.cpp/vendor/minja/minja.hpp +47 -12
- package/cpp/rn-completion.cpp +3 -27
- package/ios/generated/RNLlamaCppSpec/RNLlamaCppSpec.h +30 -0
- package/ios/generated/RNLlamaCppSpecJSI.h +49 -4
- package/ios/include/chat.h +8 -1
- package/ios/include/common/minja/chat-template.hpp +16 -7
- package/ios/include/common/minja/minja.hpp +47 -12
- package/ios/include/common.h +64 -15
- package/ios/include/llama.h +53 -114
- package/ios/include/speculative.h +8 -1
- package/ios/libs/llama.xcframework/Info.plist +18 -18
- package/ios/libs/llama.xcframework/ios-arm64/dSYMs/llama.dSYM/Contents/Resources/DWARF/llama +0 -0
- package/ios/libs/llama.xcframework/ios-arm64/dSYMs/llama.dSYM/Contents/Resources/Relocations/aarch64/llama.yml +5557 -5267
- package/ios/libs/llama.xcframework/ios-arm64/llama.framework/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/ios-arm64/llama.framework/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/ios-arm64/llama.framework/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/ios-arm64/llama.framework/llama +0 -0
- package/ios/libs/llama.xcframework/ios-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/DWARF/llama +0 -0
- package/ios/libs/llama.xcframework/ios-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/Relocations/aarch64/llama.yml +5520 -5238
- package/ios/libs/llama.xcframework/ios-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/Relocations/x86_64/llama.yml +4241 -4014
- package/ios/libs/llama.xcframework/ios-arm64_x86_64-simulator/llama.framework/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/ios-arm64_x86_64-simulator/llama.framework/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/ios-arm64_x86_64-simulator/llama.framework/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/ios-arm64_x86_64-simulator/llama.framework/llama +0 -0
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/dSYMs/llama.dSYM/Contents/Resources/DWARF/llama +0 -0
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/dSYMs/llama.dSYM/Contents/Resources/Relocations/aarch64/llama.yml +5519 -5238
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/dSYMs/llama.dSYM/Contents/Resources/Relocations/x86_64/llama.yml +4242 -4016
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/A/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/A/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/A/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/A/llama +0 -0
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/Current/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/Current/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/Current/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/Current/llama +0 -0
- package/ios/libs/llama.xcframework/macos-arm64_x86_64/llama.framework/llama +0 -0
- package/ios/libs/llama.xcframework/tvos-arm64/dSYMs/llama.dSYM/Contents/Resources/DWARF/llama +0 -0
- package/ios/libs/llama.xcframework/tvos-arm64/dSYMs/llama.dSYM/Contents/Resources/Relocations/aarch64/llama.yml +5556 -5267
- package/ios/libs/llama.xcframework/tvos-arm64/llama.framework/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/tvos-arm64/llama.framework/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/tvos-arm64/llama.framework/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/tvos-arm64/llama.framework/llama +0 -0
- package/ios/libs/llama.xcframework/tvos-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/DWARF/llama +0 -0
- package/ios/libs/llama.xcframework/tvos-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/Relocations/aarch64/llama.yml +5519 -5238
- package/ios/libs/llama.xcframework/tvos-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/Relocations/x86_64/llama.yml +4241 -4014
- package/ios/libs/llama.xcframework/tvos-arm64_x86_64-simulator/llama.framework/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/tvos-arm64_x86_64-simulator/llama.framework/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/tvos-arm64_x86_64-simulator/llama.framework/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/tvos-arm64_x86_64-simulator/llama.framework/llama +0 -0
- package/ios/libs/llama.xcframework/xros-arm64/dSYMs/llama.dSYM/Contents/Resources/DWARF/llama +0 -0
- package/ios/libs/llama.xcframework/xros-arm64/dSYMs/llama.dSYM/Contents/Resources/Relocations/aarch64/llama.yml +5553 -5303
- package/ios/libs/llama.xcframework/xros-arm64/llama.framework/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/xros-arm64/llama.framework/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/xros-arm64/llama.framework/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/xros-arm64/llama.framework/llama +0 -0
- package/ios/libs/llama.xcframework/xros-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/DWARF/llama +0 -0
- package/ios/libs/llama.xcframework/xros-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/Relocations/aarch64/llama.yml +5515 -5274
- package/ios/libs/llama.xcframework/xros-arm64_x86_64-simulator/dSYMs/llama.dSYM/Contents/Resources/Relocations/x86_64/llama.yml +4238 -4044
- package/ios/libs/llama.xcframework/xros-arm64_x86_64-simulator/llama.framework/Headers/ggml-opt.h +25 -6
- package/ios/libs/llama.xcframework/xros-arm64_x86_64-simulator/llama.framework/Headers/ggml.h +90 -3
- package/ios/libs/llama.xcframework/xros-arm64_x86_64-simulator/llama.framework/Headers/llama.h +53 -114
- package/ios/libs/llama.xcframework/xros-arm64_x86_64-simulator/llama.framework/llama +0 -0
- package/lib/module/NativeRNLlamaCpp.js.map +1 -1
- package/lib/typescript/src/NativeRNLlamaCpp.d.ts +5 -0
- package/lib/typescript/src/NativeRNLlamaCpp.d.ts.map +1 -1
- package/package.json +1 -2
- package/src/NativeRNLlamaCpp.ts +7 -0
- package/cpp/llama.cpp/ggml/src/ggml-webgpu/wgsl-shaders/mul_mat.wgsl +0 -56
|
@@ -29,15 +29,18 @@
|
|
|
29
29
|
<key>DebugSymbolsPath</key>
|
|
30
30
|
<string>dSYMs</string>
|
|
31
31
|
<key>LibraryIdentifier</key>
|
|
32
|
-
<string>
|
|
32
|
+
<string>tvos-arm64_x86_64-simulator</string>
|
|
33
33
|
<key>LibraryPath</key>
|
|
34
34
|
<string>llama.framework</string>
|
|
35
35
|
<key>SupportedArchitectures</key>
|
|
36
36
|
<array>
|
|
37
37
|
<string>arm64</string>
|
|
38
|
+
<string>x86_64</string>
|
|
38
39
|
</array>
|
|
39
40
|
<key>SupportedPlatform</key>
|
|
40
|
-
<string>
|
|
41
|
+
<string>tvos</string>
|
|
42
|
+
<key>SupportedPlatformVariant</key>
|
|
43
|
+
<string>simulator</string>
|
|
41
44
|
</dict>
|
|
42
45
|
<dict>
|
|
43
46
|
<key>BinaryPath</key>
|
|
@@ -45,7 +48,7 @@
|
|
|
45
48
|
<key>DebugSymbolsPath</key>
|
|
46
49
|
<string>dSYMs</string>
|
|
47
50
|
<key>LibraryIdentifier</key>
|
|
48
|
-
<string>
|
|
51
|
+
<string>tvos-arm64</string>
|
|
49
52
|
<key>LibraryPath</key>
|
|
50
53
|
<string>llama.framework</string>
|
|
51
54
|
<key>SupportedArchitectures</key>
|
|
@@ -53,23 +56,24 @@
|
|
|
53
56
|
<string>arm64</string>
|
|
54
57
|
</array>
|
|
55
58
|
<key>SupportedPlatform</key>
|
|
56
|
-
<string>
|
|
59
|
+
<string>tvos</string>
|
|
57
60
|
</dict>
|
|
58
61
|
<dict>
|
|
59
62
|
<key>BinaryPath</key>
|
|
60
|
-
<string>llama.framework/llama</string>
|
|
63
|
+
<string>llama.framework/Versions/A/llama</string>
|
|
61
64
|
<key>DebugSymbolsPath</key>
|
|
62
65
|
<string>dSYMs</string>
|
|
63
66
|
<key>LibraryIdentifier</key>
|
|
64
|
-
<string>
|
|
67
|
+
<string>macos-arm64_x86_64</string>
|
|
65
68
|
<key>LibraryPath</key>
|
|
66
69
|
<string>llama.framework</string>
|
|
67
70
|
<key>SupportedArchitectures</key>
|
|
68
71
|
<array>
|
|
69
72
|
<string>arm64</string>
|
|
73
|
+
<string>x86_64</string>
|
|
70
74
|
</array>
|
|
71
75
|
<key>SupportedPlatform</key>
|
|
72
|
-
<string>
|
|
76
|
+
<string>macos</string>
|
|
73
77
|
</dict>
|
|
74
78
|
<dict>
|
|
75
79
|
<key>BinaryPath</key>
|
|
@@ -77,35 +81,31 @@
|
|
|
77
81
|
<key>DebugSymbolsPath</key>
|
|
78
82
|
<string>dSYMs</string>
|
|
79
83
|
<key>LibraryIdentifier</key>
|
|
80
|
-
<string>
|
|
84
|
+
<string>ios-arm64</string>
|
|
81
85
|
<key>LibraryPath</key>
|
|
82
86
|
<string>llama.framework</string>
|
|
83
87
|
<key>SupportedArchitectures</key>
|
|
84
88
|
<array>
|
|
85
89
|
<string>arm64</string>
|
|
86
|
-
<string>x86_64</string>
|
|
87
90
|
</array>
|
|
88
91
|
<key>SupportedPlatform</key>
|
|
89
|
-
<string>
|
|
90
|
-
<key>SupportedPlatformVariant</key>
|
|
91
|
-
<string>simulator</string>
|
|
92
|
+
<string>ios</string>
|
|
92
93
|
</dict>
|
|
93
94
|
<dict>
|
|
94
95
|
<key>BinaryPath</key>
|
|
95
|
-
<string>llama.framework/
|
|
96
|
+
<string>llama.framework/llama</string>
|
|
96
97
|
<key>DebugSymbolsPath</key>
|
|
97
98
|
<string>dSYMs</string>
|
|
98
99
|
<key>LibraryIdentifier</key>
|
|
99
|
-
<string>
|
|
100
|
+
<string>xros-arm64</string>
|
|
100
101
|
<key>LibraryPath</key>
|
|
101
102
|
<string>llama.framework</string>
|
|
102
103
|
<key>SupportedArchitectures</key>
|
|
103
104
|
<array>
|
|
104
105
|
<string>arm64</string>
|
|
105
|
-
<string>x86_64</string>
|
|
106
106
|
</array>
|
|
107
107
|
<key>SupportedPlatform</key>
|
|
108
|
-
<string>
|
|
108
|
+
<string>xros</string>
|
|
109
109
|
</dict>
|
|
110
110
|
<dict>
|
|
111
111
|
<key>BinaryPath</key>
|
|
@@ -113,7 +113,7 @@
|
|
|
113
113
|
<key>DebugSymbolsPath</key>
|
|
114
114
|
<string>dSYMs</string>
|
|
115
115
|
<key>LibraryIdentifier</key>
|
|
116
|
-
<string>
|
|
116
|
+
<string>xros-arm64_x86_64-simulator</string>
|
|
117
117
|
<key>LibraryPath</key>
|
|
118
118
|
<string>llama.framework</string>
|
|
119
119
|
<key>SupportedArchitectures</key>
|
|
@@ -122,7 +122,7 @@
|
|
|
122
122
|
<string>x86_64</string>
|
|
123
123
|
</array>
|
|
124
124
|
<key>SupportedPlatform</key>
|
|
125
|
-
<string>
|
|
125
|
+
<string>xros</string>
|
|
126
126
|
<key>SupportedPlatformVariant</key>
|
|
127
127
|
<string>simulator</string>
|
|
128
128
|
</dict>
|