sima_utils.transformer.model.base
=================================

.. py:module:: sima_utils.transformer.model.base


Classes
-------

.. autoapisummary::

   sima_utils.transformer.model.base.FileGenMode
   sima_utils.transformer.model.base.FileGenPrecision
   sima_utils.transformer.model.base.EvalMode
   sima_utils.transformer.model.base.BaseModel


Module Contents
---------------

.. py:class:: FileGenMode



   File generation mode.


   .. py:attribute:: ONNX


   .. py:attribute:: MODEL_SDK_QUANTIZE


   .. py:attribute:: MODEL_SDK_COMPILE


   .. py:attribute:: DEVKIT


   .. py:attribute:: ALL


.. py:class:: FileGenPrecision



   Precision used when generating files.


   .. py:attribute:: BF16
      :value: 'bf16'



   .. py:attribute:: A_BF16_W_INT8
      :value: 'a_bf16_w_int8'



   .. py:attribute:: A_BF16_W_INT4
      :value: 'a_bf16_w_int4'



.. py:class:: EvalMode



   Model evaluation mode.


   .. py:attribute:: HF
      :value: 'hf'



   .. py:attribute:: ONNX
      :value: 'onnx'



   .. py:attribute:: SDK
      :value: 'sdk'



.. py:class:: BaseModel



   Base implementation for visual-language model file generation.

   .. attribute:: cfg

      Configuration of the model.

   .. attribute:: model_name

      Name of the model. This will be used to determine the generated files' names.

   .. attribute:: onnx_path

      Path to store the ONNX files.

   .. attribute:: sima_path

      Path to store the SiMa-specific files.

   .. attribute:: hf_model

      LocalHuggingFaceModel object for obtaining the parameters to generate ONNX files.

   .. attribute:: onnx_file_name

      File name of the generated ONNX file.

   .. attribute:: weight_prefix

      The prefix of weight tensor names in the source model.


   .. py:attribute:: cfg
      :type:  sima_utils.transformer.vlm_config.BaseConfig


   .. py:attribute:: model_name
      :type:  str


   .. py:attribute:: onnx_path
      :type:  pathlib.Path
      :value: 'onnx_files'



   .. py:attribute:: sima_path
      :type:  pathlib.Path
      :value: 'sima_files'



   .. py:attribute:: hf_model
      :type:  sima_utils.transformer.hf_transformer.LocalHuggingFaceModel | None
      :value: None



   .. py:attribute:: vlm_helper
      :type:  sima_utils.transformer.vlm_config.VlmHelper | None
      :value: None



   .. py:method:: gen_files(gen_mode: FileGenMode, *, precision: FileGenPrecision = FileGenPrecision.BF16, log_level: int = logging.NOTSET, resume: bool = False)

      Generates files based on the provided file generation mode.

      :param gen_mode: File generation mode.
      :param precision: The precision to be used for Model SDK quantization mode.
      :param log_level: Logging level.
      :param resume: Set to generate only when the file cannot be found.



   .. py:method:: run_model(eval_mode: EvalMode, ifms: list[numpy.ndarray]) -> list[numpy.ndarray]

      Runs the model based on the evaluation mode.



   .. py:property:: vision_model_name
      :type: str



   .. py:property:: language_model_name
      :type: str



   .. py:property:: onnx_file_name
      :type: pathlib.Path



   .. py:property:: sima_model_sdk_path
      :type: pathlib.Path


      Path to the generated quantized Model SDK files.


   .. py:property:: sima_mpk_path
      :type: pathlib.Path


      Path to the generated MPK files.


   .. py:property:: sdk_file_name
      :type: pathlib.Path


      Path to the generated quantized Model SDK file.


   .. py:property:: mpk_file_name
      :type: pathlib.Path


      Path to the generated quantized Model SDK file.


   .. py:property:: sima_devkit_path
      :type: pathlib.Path


      Path to the generated files for DEVKIT.


   .. py:method:: get_gen_file_name(gen_mode: FileGenMode) -> pathlib.Path


   .. py:method:: gen_onnx_files()

      Generates ONNX files.



   .. py:method:: gen_model_sdk_files(precision: FileGenPrecision, log_level: int)

      Generates quantized Model SDK files.

      :param precision: Precision used for quantization.
      :param log_level: Logging level.



   .. py:method:: gen_mpk_files(log_level: int) -> afe.apis.model.Model

      Generates MPK files.

      :param log_level: Logging level.



   .. py:method:: gen_devkit_files(resume: bool = False)

      Generates files for devkit.



   .. py:method:: check_hf_param(name: str) -> bool

      Checks if a parameter tensor exists in the LocalHuggingFaceModel object.

      :param name: Full name of the parameter.

      :returns: True if the parameter tensor exists.



   .. py:method:: get_hf_param(name: str) -> numpy.ndarray

      Gets the parameter tensor from the LocalHuggingFaceModel object.

      :param name: Full name of the parameter.

      :returns: The parameter tensor in numpy array.



   .. py:method:: create_onnx_builder()

      Creates onnx builder.



   .. py:method:: gen_files_from_model_list(model_list: list[tuple[BaseModel, FileGenPrecision]], gen_mode: FileGenMode, num_processes: int, log_level: int, resume: bool)