^Njv6UddlmZmZmZmZddlmZmZmZddl m Z ddl m Z m Z ddlmZmZddlmZddlmZmZedd d d d ed dddedd dd deddddedddd deddeddd d d ed!d"dded#ddd d$ed%ded&d'd(d d)ed*d+dded,dd-d.d/ed,d0ed1dd-d.d)ed1d0ed2dd-d.d ed2d0ed3d d-d.d4ed5d0ed6d d7d.d8ed6d0ed9dd-d.d:ed9d0ed;d dg Zeeed?<Gd@dAeeeZGdBdCeeZyD)E)AnyIterableSequenceType) NumpyArray OnnxProviderDevice)OnnxOutputContext)define_cache_dir normalize) OnnxTextModelTextEmbeddingWorker)TextEmbeddingBase)DenseModelDescription ModelSourcezBAAI/bge-base-eniz}Text embeddings, Unimodal (text), English, 512 input tokens truncation, Prefixes for queries/documents: necessary, 2023 year.mitgzG?zQdrant/fast-bge-base-enzGhttps://storage.googleapis.com/qdrant-fastembed/fast-bge-base-en.tar.gzT)hfurl_deprecated_tar_structzmodel_optimized.onnx)modeldim descriptionlicense size_in_GBsources model_filezBAAI/bge-base-en-v1.5zText embeddings, Unimodal (text), English, 512 input tokens truncation, Prefixes for queries/documents: not so necessary, 2023 year.gzG?zqdrant/bge-base-en-v1.5-onnx-qzLhttps://storage.googleapis.com/qdrant-fastembed/fast-bge-base-en-v1.5.tar.gzzBAAI/bge-large-en-v1.5g333333?zqdrant/bge-large-en-v1.5-onnx)rz model.onnxzBAAI/bge-small-enigp= ף?zQdrant/bge-small-enzHhttps://storage.googleapis.com/qdrant-fastembed/BAAI-bge-small-en.tar.gzBAAI/bge-small-en-v1.5gx&?zqdrant/bge-small-en-v1.5-onnx-qzBAAI/bge-small-zh-v1.5izText embeddings, Unimodal (text), Chinese, 512 input tokens truncation, Prefixes for queries/documents: not so necessary, 2023 year.g ףp= ?zQdrant/bge-small-zh-v1.5zMhttps://storage.googleapis.com/qdrant-fastembed/fast-bge-small-zh-v1.5.tar.gzz"mixedbread-ai/mxbai-embed-large-v1z}Text embeddings, Unimodal (text), English, 512 input tokens truncation, Prefixes for queries/documents: necessary, 2024 year.z apache-2.0g{Gz?zonnx/model.onnxz#snowflake/snowflake-arctic-embed-xsz"snowflake/snowflake-arctic-embed-sz"snowflake/snowflake-arctic-embed-mgQ?z"Snowflake/snowflake-arctic-embed-mz'snowflake/snowflake-arctic-embed-m-longz~Text embeddings, Unimodal (text), English, 2048 input tokens truncation, Prefixes for queries/documents: necessary, 2024 year.gHzG?z"snowflake/snowflake-arctic-embed-lgRQ?zjinaai/jina-clip-v1zkText embeddings, Multimodal (text&image), English, Prefixes for queries/documents: not necessary, 2024 yearg?zonnx/text_model.onnxsupported_onnx_modelsceZdZdZedeefdZddddejddddf de de dzd e dzd e e dzd eezd ee dzd ede dzde dzdeffd Z dde ee zde de dzdedeef dZededfdZdee efdedee effdZdededeefdZd dZ d!de ee zde dede fdZxZS)"OnnxTextEmbeddingz+Implementation of the Flag Embedding model.returnctS)z Lists the supported models. Returns: list[DenseModelDescription]: A list of DenseModelDescription objects containing the model information. )rclss n/Users/ahmed/devFolder/Ultron/claude-voice/.venv/lib/python3.12/site-packages/fastembed/text/onnx_embedding.py_list_supported_modelsz(OnnxTextEmbedding._list_supported_modelss %$rNF model_name cache_dirthreads providerscuda device_ids lazy_load device_idspecific_model_pathkwargsc  t ||||fi| ||_||_|j | |_||_||_d|_|||_n |j |j d|_|j||_ tt||_ | |_|j|j|j|j |j|_|js|j%yy)a Args: model_name (str): The name of the model to use. cache_dir (str, optional): The path to the cache directory. Can be set using the `FASTEMBED_CACHE_PATH` env variable. Defaults to `fastembed_cache` in the system's temp directory. threads (int, optional): The number of threads single onnxruntime session can use. Defaults to None. providers (Optional[Sequence[OnnxProvider]], optional): The list of onnxruntime providers to use. Mutually exclusive with the `cuda` and `device_ids` arguments. Defaults to None. cuda (Union[bool, Device], optional): Whether to use cuda for inference. Mutually exclusive with `providers` Defaults to Device.AUTO. device_ids (Optional[list[int]], optional): The list of device ids to use for data parallel processing in workers. Should be used with `cuda` equals to `True`, `Device.AUTO` or `Device.CUDA`, mutually exclusive with `providers`. Defaults to None. lazy_load (bool, optional): Whether to load the model during class initialization or on demand. Should be set to True when using multiple-gpu and parallel encoding. Defaults to False. device_id (Optional[int], optional): The device id to use for loading the model in the worker process. specific_model_path (Optional[str], optional): The specific path to the onnx model dir if it should be imported from somewhere else Raises: ValueError: If the model_name is not in the format / e.g. BAAI/bge-base-en. Nr)local_files_onlyr1)super__init__r,r/_select_exposed_session_options_extra_session_optionsr.r-r0_get_model_descriptionmodel_descriptionstrr r*_specific_model_pathdownload_model_local_files_only _model_dirload_onnx_model) selfr)r*r+r,r-r.r/r0r1r2 __class__s r&r6zOnnxTextEmbedding.__init__sF YB6B""&*&J&J6&R#$ &*  &DN __ (!__Q/DN!%! 1, data-parallel encoding will be used, recommended for offline encoding of large datasets. If 0, use all available cores. If None, don't use data-parallel processing, use default onnxruntime threading instead. Returns: List of embeddings, one per document ) r)r*rCrDrEr,r-r.r4r1extra_session_optionsN) _embed_documentsr)r;r*r,r-r.r>r<r8)rArCrDrEr2s r&embedzOnnxTextEmbedding.embeds},)4((  $..)!nn!33 $ 9 9"&"="=    sA;B=B>BzTextEmbeddingWorker[NumpyArray]ctS)N)OnnxTextEmbeddingWorkerr$s r&_get_worker_classz#OnnxTextEmbedding._get_worker_class+s&&r( onnx_inputc |S)z, Preprocess the onnx input. rH)rArNr2s r&_preprocess_onnx_inputz(OnnxTextEmbedding._preprocess_onnx_input/s r(outputc |j}|jdk(r|dddf}t |S|jdk(r |}t |Std|j)NrzUnsupported embedding shape: ) model_outputndim ValueErrorshaper )rArQr2 embeddingsprocessed_embeddingss r&_post_process_onnx_outputz+OnnxTextEmbedding._post_process_onnx_output7sv(( ??a #-ad#3 -.. __ !#- -.. c:o  /' /36 / *  /  =AI8C=(I69ILOI Ir(r!c$eZdZdedededefdZy)rLr)r*r2r"c td||dd|S)N)r)r*r+rH)r!)rAr)r*r2s r&init_embeddingz&OnnxTextEmbeddingWorker.init_embeddingVs) ! !    r(N)rdrerfr;rr!rrrHr(r&rLrLUs-          r(rLN)typingrrrrfastembed.common.typesrrr fastembed.common.onnx_modelr fastembed.common.utilsr r fastembed.text.onnx_text_modelr r"fastembed.text.text_embedding_baser"fastembed.common.model_descriptionrrrri__annotations__r!rLrHr(r&r{s00CC9>M@Q  D(Y#' * %  K/^#' * &  K>? !  D$Z#' * &  K@A) &  K)_#' * 2  DCD$ 3  DDE$ 2  DCD$ 2  DCD$ 7  DHI$ 2  DCD$ #  '45) Cm6t12m`XI)=+DXIv  1*=  r(