bigdl-core-npu 2.6.0b20241127__cp311-cp311-win_amd64.whl → 2.6.0b20241129__cp311-cp311-win_amd64.whl

bigdl-core-npu/include/common.h

@@ -29,7 +29,7 @@
 
 struct common_params {
     int32_t n_predict             =    -1; // new tokens to predict
-    std::string model                = ""; // model path                                                    // NOLINT
+    char* model                = ""; // model path                                                    // NOLINT
     std::string prompt               = "";                                                                  // NOLINT
     std::string prompt_file          = ""; // store the external prompt file name                           // NOLINT
 

bigdl-core-npu/include/npu_llm.h

@@ -47,15 +47,19 @@ extern "C" {
 
     EXPORT_API std::string llm_decode(vector<int32_t> tokens);
 
-    EXPORT_API extern NPUModel* load_model_from_file(npu_model_params &model_params, std::string model_str);
+    EXPORT_API void* load_model_from_file(const char* model_path);
+
+    EXPORT_API void load_config_from_file(npu_model_params &model_params, const char* model_path);
 
     EXPORT_API std::string add_chat_template(npu_model_params model_params, std::string input_prompt);
 
-    EXPORT_API float* run_prefill(NPUModel* model, std::vector<int32_t> embd_inp);
+    EXPORT_API float* run_prefill(void* void_model, int32_t* embd_inp_ptr, int32_t embd_inp_size);
+
+    EXPORT_API float* run_decode(void* void_model, int32_t input_token);
 
-    EXPORT_API float* run_decode(NPUModel* model, int32_t input_token);
+    EXPORT_API int32_t llm_sample_token(float* logits, bool greedy_search, int32_t vocab_size);
 
-    EXPORT_API int32_t llm_sample_token(float* logits, bool greedy_search, npu_model_params model_params);
+    EXPORT_API void reset(void* void_model);
 
 #ifdef __cplusplus
 }

{bigdl_core_npu-2.6.0b20241127.dist-info → bigdl_core_npu-2.6.0b20241129.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: bigdl-core-npu
-Version: 2.6.0b20241127
+Version: 2.6.0b20241129
 Summary: Intel® NPU Acceleration Library
 Home-page: https://github.com/intel/intel-npu-acceleration-library
 Author: Alessandro Palla

{bigdl_core_npu-2.6.0b20241127.dist-info → bigdl_core_npu-2.6.0b20241129.dist-info}/RECORD

@@ -1,10 +1,10 @@
 bigdl-core-npu/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-bigdl-core-npu/npu_llm.dll,sha256=QTe8asDJOiXwF7KmE1j2FMQMDV0gugwp365S496WDts,3899392
-bigdl-core-npu/npu_llm.lib,sha256=0ZJqhYtedXgS99Zu9K3WNyfIB_2c1x8yIi533iuqBvU,42990
-bigdl-core-npu/include/common.h,sha256=dNguwq6Q_4vbv9kZeVqrtMYn3yfRh17ykQVPapUMl6k,2638
-bigdl-core-npu/include/npu_llm.h,sha256=SF3oVFtieuyV9UA4KmL0oa5LiJsmnawk_5Tai-7X208,1789
+bigdl-core-npu/npu_llm.dll,sha256=A1Tbjpdw_VuGX4-jWcjFPU3TxTOXl3bnCmrdNPK8uQo,3902464
+bigdl-core-npu/npu_llm.lib,sha256=VuvMfKbbAM7BTiDyFBCK8PGqmfdEOVJVtNQOtdTJI_k,43370
+bigdl-core-npu/include/common.h,sha256=2qMhyPMYet0BzIQ5m_wRXye_A8ileRpnsxp55TvcThE,2632
+bigdl-core-npu/include/npu_llm.h,sha256=aYJxEHcrNsHZdgjtom7pILZtB0eBONB9H-azgQDmo38,1903
 intel_npu_acceleration_library/__init__.py,sha256=ZKTIhGMDjF7P6pF-yX8KWcSXbeHWRk24AO_orsa18f8,536
-intel_npu_acceleration_library/_version.py,sha256=WD_iLLcnaylPzNBSHyrUaRihZyy9wh7pMVq-iv75ncI,112
+intel_npu_acceleration_library/_version.py,sha256=0JBHmQg23E48IvzuouLicFI6jfbU4iioNTqm_rvIxx4,112
 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
@@ -198,7 +198,7 @@ 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=CJ5RqPWJJQwTfeGz-ujOboCKdltYFh6acmIQuLBhyds,333312
+intel_npu_acceleration_library/lib/Release/intel_npu_acceleration_library.dll,sha256=lDqJmGdG1KHgK65VNfSdFQJev3ETwOgNnTpYLRLY-jM,333312
 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
@@ -228,7 +228,7 @@ 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-2.6.0b20241127.dist-info/METADATA,sha256=JlG2fGeBn3KiGYB6KGIWaeo2ymtH86jCYNkofrNySTw,1541
-bigdl_core_npu-2.6.0b20241127.dist-info/WHEEL,sha256=nkBcd8Ko0v5sEcSagm2-x_RVrb8gBSkTa8VFFZ0Mr1o,101
-bigdl_core_npu-2.6.0b20241127.dist-info/top_level.txt,sha256=iMQZlTsFPJjlD-Y0MqZEP_9ifI0LlbNCJIOTaMoGMjk,46
-bigdl_core_npu-2.6.0b20241127.dist-info/RECORD,,
+bigdl_core_npu-2.6.0b20241129.dist-info/METADATA,sha256=aZJgC5QEtfWfCP6WxG_MFUTSPkOs60A6zF6ayPGdThs,1541
+bigdl_core_npu-2.6.0b20241129.dist-info/WHEEL,sha256=nkBcd8Ko0v5sEcSagm2-x_RVrb8gBSkTa8VFFZ0Mr1o,101
+bigdl_core_npu-2.6.0b20241129.dist-info/top_level.txt,sha256=iMQZlTsFPJjlD-Y0MqZEP_9ifI0LlbNCJIOTaMoGMjk,46
+bigdl_core_npu-2.6.0b20241129.dist-info/RECORD,,

intel_npu_acceleration_library/_version.py

@@ -3,4 +3,4 @@
 # SPDX-License-Identifier: Apache 2.0
 #
 
-__version__ = "2.6.0b20241127"
+__version__ = "2.6.0b20241129"