bigdl-core-npu 2.6.0b20250122__cp310-cp310-win_amd64.whl → 2.6.0b20250126__cp310-cp310-win_amd64.whl

bigdl-core-npu/include/llamacpp/log.h

@@ -0,0 +1,92 @@
+#pragma once
+
+#include "ggml.h" // for ggml_log_level
+
+#ifndef __GNUC__
+#    define LOG_ATTRIBUTE_FORMAT(...)
+#elif defined(__MINGW32__)
+#    define LOG_ATTRIBUTE_FORMAT(...) __attribute__((format(gnu_printf, __VA_ARGS__)))
+#else
+#    define LOG_ATTRIBUTE_FORMAT(...) __attribute__((format(printf, __VA_ARGS__)))
+#endif
+
+#define LOG_DEFAULT_DEBUG 1
+#define LOG_DEFAULT_LLAMA 0
+
+// needed by the LOG_TMPL macro to avoid computing log arguments if the verbosity lower
+// set via gpt_log_set_verbosity()
+extern int gpt_log_verbosity_thold;
+
+void gpt_log_set_verbosity_thold(int verbosity); // not thread-safe
+
+// the gpt_log uses an internal worker thread to print/write log messages
+// when the worker thread is paused, incoming log messages are discarded
+struct gpt_log;
+
+struct gpt_log * gpt_log_init();
+struct gpt_log * gpt_log_main(); // singleton, automatically destroys itself on exit
+void             gpt_log_pause (struct gpt_log * log); // pause  the worker thread, not thread-safe
+void             gpt_log_resume(struct gpt_log * log); // resume the worker thread, not thread-safe
+void             gpt_log_free  (struct gpt_log * log);
+
+LOG_ATTRIBUTE_FORMAT(3, 4)
+void gpt_log_add(struct gpt_log * log, enum ggml_log_level level, const char * fmt, ...);
+
+// defaults: file = NULL, colors = false, prefix = false, timestamps = false
+//
+// regular log output:
+//
+//   ggml_backend_metal_log_allocated_size: allocated buffer, size =  6695.84 MiB, ( 6695.91 / 21845.34)
+//   llm_load_tensors: ggml ctx size =    0.27 MiB
+//   llm_load_tensors: offloading 32 repeating layers to GPU
+//   llm_load_tensors: offloading non-repeating layers to GPU
+//
+// with prefix = true, timestamps = true, the log output will look like this:
+//
+//   0.00.035.060 D ggml_backend_metal_log_allocated_size: allocated buffer, size =  6695.84 MiB, ( 6695.91 / 21845.34)
+//   0.00.035.064 I llm_load_tensors: ggml ctx size =    0.27 MiB
+//   0.00.090.578 I llm_load_tensors: offloading 32 repeating layers to GPU
+//   0.00.090.579 I llm_load_tensors: offloading non-repeating layers to GPU
+//
+// I - info    (stdout, V = 0)
+// W - warning (stderr, V = 0)
+// E - error   (stderr, V = 0)
+// D - debug   (stderr, V = LOG_DEFAULT_DEBUG)
+//
+
+void gpt_log_set_file      (struct gpt_log * log, const char * file);       // not thread-safe
+void gpt_log_set_colors    (struct gpt_log * log,       bool   colors);     // not thread-safe
+void gpt_log_set_prefix    (struct gpt_log * log,       bool   prefix);     // whether to output prefix to each log
+void gpt_log_set_timestamps(struct gpt_log * log,       bool   timestamps); // whether to output timestamps in the prefix
+
+// helper macros for logging
+// use these to avoid computing log arguments if the verbosity of the log is higher than the threshold
+//
+// for example:
+//
+//   LOG_DBG("this is a debug message: %d\n", expensive_function());
+//
+// this will avoid calling expensive_function() if LOG_DEFAULT_DEBUG > gpt_log_verbosity_thold
+//
+
+#define LOG_TMPL(level, verbosity, ...) \
+    do { \
+        if ((verbosity) <= gpt_log_verbosity_thold) { \
+            gpt_log_add(gpt_log_main(), (level), __VA_ARGS__); \
+        } \
+    } while (0)
+
+#define LOG(...)             LOG_TMPL(GGML_LOG_LEVEL_NONE, 0,         __VA_ARGS__)
+#define LOGV(verbosity, ...) LOG_TMPL(GGML_LOG_LEVEL_NONE, verbosity, __VA_ARGS__)
+
+#define LOG_INF(...) LOG_TMPL(GGML_LOG_LEVEL_INFO,  0,                 __VA_ARGS__)
+#define LOG_WRN(...) LOG_TMPL(GGML_LOG_LEVEL_WARN,  0,                 __VA_ARGS__)
+#define LOG_ERR(...) LOG_TMPL(GGML_LOG_LEVEL_ERROR, 0,                 __VA_ARGS__)
+#define LOG_DBG(...) LOG_TMPL(GGML_LOG_LEVEL_DEBUG, LOG_DEFAULT_DEBUG, __VA_ARGS__)
+#define LOG_CNT(...) LOG_TMPL(GGML_LOG_LEVEL_CONT,  0,                 __VA_ARGS__)
+
+#define LOG_INFV(verbosity, ...) LOG_TMPL(GGML_LOG_LEVEL_INFO,  verbosity, __VA_ARGS__)
+#define LOG_WRNV(verbosity, ...) LOG_TMPL(GGML_LOG_LEVEL_WARN,  verbosity, __VA_ARGS__)
+#define LOG_ERRV(verbosity, ...) LOG_TMPL(GGML_LOG_LEVEL_ERROR, verbosity, __VA_ARGS__)
+#define LOG_DBGV(verbosity, ...) LOG_TMPL(GGML_LOG_LEVEL_DEBUG, verbosity, __VA_ARGS__)
+#define LOG_CNTV(verbosity, ...) LOG_TMPL(GGML_LOG_LEVEL_CONT,  verbosity, __VA_ARGS__)

{bigdl_core_npu-2.6.0b20250122.dist-info → bigdl_core_npu-2.6.0b20250126.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: bigdl-core-npu
-Version: 2.6.0b20250122
+Version: 2.6.0b20250126
 Summary: Intel® NPU Acceleration Library
 Home-page: https://github.com/intel/intel-npu-acceleration-library
 Author: Alessandro Palla

{bigdl_core_npu-2.6.0b20250122.dist-info → bigdl_core_npu-2.6.0b20250126.dist-info}/RECORD

@@ -1,11 +1,23 @@
 bigdl-core-npu/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-bigdl-core-npu/llama-cli-npu.exe,sha256=yBds0mQgC7L8_5lafxq1o0FRVE1yWQr-LyXBQs7j1mo,41984
-bigdl-core-npu/npu_llm.dll,sha256=tajIrrN0y-VJyVvgRk-Tf7nYug5kAcEM8_rErOt9L64,3925504
+bigdl-core-npu/common.lib,sha256=dszuzlqhqxqlyZzDh3HBvx7Zfr1XX5AOfmpEXWmMphc,6647656
+bigdl-core-npu/ggml.dll,sha256=qETfE0Uwxl_GjqqFFwhKEwC-xnzcIUDz2G3Bea0Gh0M,704000
+bigdl-core-npu/ggml.lib,sha256=gnGRJzDCJ4O25sbrAp-tydMm0j50lhiigyXkpqrgW8M,139692
+bigdl-core-npu/llama-cli-npu.exe,sha256=k0sX6AT1OTr5JUS0tP6_gPWf2VqkIujvaYTouaL4lIg,41984
+bigdl-core-npu/llama.dll,sha256=DfKUiIdCV7VBX-3dtopFmDQeO1tY-yOAILS8w0pCJQk,1573376
+bigdl-core-npu/llama.lib,sha256=Z2dQy5y6QJ1P2DubIjfo4cB-T6a2ZfAMEC7MDShrX8Q,1494020
+bigdl-core-npu/npu_llm.dll,sha256=cIzJ763TQigSFQrcNrCCR4YqWB8gwmJAmxMgGoCCZf4,3924992
 bigdl-core-npu/npu_llm.lib,sha256=nNYF-btjrT9Pzcd31kQV5BsLBSNdA_DSeyK5RnmOc9s,44524
 bigdl-core-npu/include/common.h,sha256=P_WP5vhSSYZPmFAZxLMvhEobKRfksfvQt-O8wmGzFSQ,3059
 bigdl-core-npu/include/npu_llm.h,sha256=siEFqkKnZlzR9cvq7qXiAy5Z65-X14QyWEw__wsLLpA,2605
+bigdl-core-npu/include/llamacpp/arg.h,sha256=XOLwSDqJsCJJgGCkOGSWcA9yGZ0eiiY7GffHdAxFYYY,3110
+bigdl-core-npu/include/llamacpp/common.h,sha256=SaGJ1cAmJWd5D-T_03udb6ADTcJ8VX-bMy0GYGDMpI4,25841
+bigdl-core-npu/include/llamacpp/ggml-alloc.h,sha256=kiWITcUF9Q7kvGSBeCfouV59YyGDNy56VE_0fXXnWHE,3088
+bigdl-core-npu/include/llamacpp/ggml-backend.h,sha256=cUJR-AC7hUbIxukKixR9vUSV38gmp-9eoViaBG8lyio,14347
+bigdl-core-npu/include/llamacpp/ggml.h,sha256=_K-zhVjT6uHa5lXMN3k3AwSrwb9W2q1A7v5HiGQc8iQ,105629
+bigdl-core-npu/include/llamacpp/llama.h,sha256=gMTNuRrEY7z0nkCewKLMZtf5urcSzkQm_HxEs3DJXHI,60967
+bigdl-core-npu/include/llamacpp/log.h,sha256=D6UiundA26yZOh3ci828u5U4ZaBxWb9CwSNu_3dt6CA,4295
 intel_npu_acceleration_library/__init__.py,sha256=ZKTIhGMDjF7P6pF-yX8KWcSXbeHWRk24AO_orsa18f8,536
-intel_npu_acceleration_library/_version.py,sha256=gdf2LKw4K1gTtbTJXqrF4XwHSN85CCMvdYONDeKjdlk,112
+intel_npu_acceleration_library/_version.py,sha256=pfTWD8BzC-mWRwDdCEa_lX70cHhl1MEalYyS5mzHoko,112
 intel_npu_acceleration_library/compiler.py,sha256=3IdgqjamSC8MLexDBJypIeZRiWIcTFnvQSU1LPXUr7Y,6225
 intel_npu_acceleration_library/device.py,sha256=9bn8eVXJa5cXIqgfLsQAdkMVtVUQABb8z0-mQik5jRg,7424
 intel_npu_acceleration_library/dtypes.py,sha256=gdd06Wsc9zIZFHlauUEx4xcK9WGTn1Mu6GkuYDJeA-E,4683
@@ -222,7 +234,7 @@ intel_npu_acceleration_library/external/openvino/torch/__init__.py,sha256=RXLzsf
 intel_npu_acceleration_library/functional/__init__.py,sha256=WWKwKOh6Sgovv7mKctA872TbLP98Pg5m5-MREvUmlAA,204
 intel_npu_acceleration_library/functional/scaled_dot_product_attention.py,sha256=yGUcg4tDQOLuUnP1g74cl-ec8TRr2SuAMcNLlN6qLvE,1620
 intel_npu_acceleration_library/lib/Release/cache.json,sha256=CyrSqZUWo0Ec4_7ydOiuKIC0Gm8AybrGdozUqUuHxBw,8840377
-intel_npu_acceleration_library/lib/Release/intel_npu_acceleration_library.dll,sha256=kI_RiyX3vxhDzophlumvHoaTznew38O1Fq_JMe8YZV0,341504
+intel_npu_acceleration_library/lib/Release/intel_npu_acceleration_library.dll,sha256=BpYRmTopXmzg3fg-MbIpKPqOGwffLMVjSLJLttnpUWI,341504
 intel_npu_acceleration_library/lib/Release/openvino.dll,sha256=m7M119p3JBq2YYJJ2zzCaBDz6XivKK3nNykb8L1cvDU,13244768
 intel_npu_acceleration_library/lib/Release/openvino_auto_batch_plugin.dll,sha256=2v_I9P3Qo0St1bQZMEZscnFOUVvgZQQ0HvQlG3HtTd0,203104
 intel_npu_acceleration_library/lib/Release/openvino_auto_plugin.dll,sha256=e3Aj9CDRHN30dBEdPSk7OCWe52tWfhI4xeXgyFjuDHg,475488
@@ -252,7 +264,7 @@ intel_npu_acceleration_library/nn/functional.py,sha256=UfAKBc0u6RtyaMo14ldH2GpEn
 intel_npu_acceleration_library/nn/linear.py,sha256=Q06SoGQeLaI86nA_ky2GnFC6H2Fw1zyMDILKnpYC2eo,5739
 intel_npu_acceleration_library/nn/llm.py,sha256=P6dz36Yf6BHtzWcftaghC6QaMI_WeRfQwrCbO7fD6hk,15002
 intel_npu_acceleration_library/nn/module.py,sha256=EYxoTq6I_YgBDgTF76GPDxHrT8SupOTDGMzQaomBeq8,12667
-bigdl_core_npu-2.6.0b20250122.dist-info/METADATA,sha256=kIpS9pWRKGbpa7tYLD9eOS-5XQD4S3KSggr70N_cqoA,1762
-bigdl_core_npu-2.6.0b20250122.dist-info/WHEEL,sha256=rzGfZgUcGeKSgIHGYMuqg4xE4VPHxnaldXH6BG0zjVk,101
-bigdl_core_npu-2.6.0b20250122.dist-info/top_level.txt,sha256=iMQZlTsFPJjlD-Y0MqZEP_9ifI0LlbNCJIOTaMoGMjk,46
-bigdl_core_npu-2.6.0b20250122.dist-info/RECORD,,
+bigdl_core_npu-2.6.0b20250126.dist-info/METADATA,sha256=Phpf0TJvqJypcAsoOQmwsksNqAIbGDJOOSsEVh_XXNI,1762
+bigdl_core_npu-2.6.0b20250126.dist-info/WHEEL,sha256=rzGfZgUcGeKSgIHGYMuqg4xE4VPHxnaldXH6BG0zjVk,101
+bigdl_core_npu-2.6.0b20250126.dist-info/top_level.txt,sha256=iMQZlTsFPJjlD-Y0MqZEP_9ifI0LlbNCJIOTaMoGMjk,46
+bigdl_core_npu-2.6.0b20250126.dist-info/RECORD,,

intel_npu_acceleration_library/_version.py

@@ -3,4 +3,4 @@
 # SPDX-License-Identifier: Apache 2.0
 #
 
-__version__ = "2.6.0b20250122"
+__version__ = "2.6.0b20250126"