BThZddlZddlmZddlmZddlZddlZddlmZm Z m Z m Z m Z m Z mZddlmZddlmZddlmZddlmZdd lmZdd lmZdd lm Z dd l!m"Z"d dl#m$Z$m%Z%m&Z&m'Z'm(Z(m)Z)m*Z*m+Z+d dl,m-Z-m.Z.d dl/m0Z0ddl1m2Z2e Z3e e e%e e3fZ4e de+Z5dZ6gdZ7GddZ8y)N)BytesIO)urljoin)IterableListOptionalUnionDictTypeVarBinaryIO)ManageAssistantsApi)SearchCompletions)Chat) MessageModel)ContextOptionsModel) Assistant) ApiClient) FileModel)ContextOptionsMessageStreamChatResponseCitationStreamChatResponseContentDeltaStreamChatResponseMessageEndStreamChatResponseMessageStart ChatResponseBaseStreamChatResponseChunk)StreamingChatCompletionChunkChatCompletionResponse)ContextResponse)ContextRequestS)bound assistant)gpt-4ozgpt-4.1zo4-minizclaude-3-5-sonnetzclaude-3-7-sonnetzgemini-2.5-proceZdZdefdZdZdZdZ d(dede e ee fd e e d e fd Z d(d ed ede e ee fd e e d e f dZ d(dede e ee fd e e d e fdZd)dede ed e fdZd*de e ee fd ee fdZd*ded e e fdZedeeeeefd eefdZ d+deeeeefde e ee fd edeedfde ed eeee ff dZ! d,deedede ee fde ed ef dZ" d,deedede e ee fde ed ee f dZ# d-deeeeefde e ee fd edeedfde ededed e ee$e ee ffd ee%ee&ffd!Z' d.deedede ede ee fdeded e e$d e%fd"Z( d/deedede ede e ee fded e e$d ee&fd#Z) d0d$e edeeeeefde e ee fd%e e d&e e f d'Z*y)1AssistantModelr$c||_|j|_t|jt|_|r|ni|_|t t d|j|_|jj|_|jj|_ |jj|_ |jj|_ |jj|_ g|_ y)Nz2025-01)host)r$r)r HOST_SUFFIXconfigr DataApiClient_assistant_data_apiname created_at updated_atmetadatastatusctxs)selfr$client_builderr+s /var/www/catia.catastroantioquia-mas.com/valormas/lib/python3.12/site-packages/pinecone_plugins/assistant/models/assistant_model.py__init__zAssistantModel.__init__-s"NN DII{3 &fB #1 }idii$  NN'' ..33..33// nn++  c,t|jSN)strr$r4s r6__str__zAssistantModel.__str__>s4>>""r8c,t|jSr:)reprr$r<s r6__repr__zAssistantModel.__repr__AsDNN##r8c.t|j|Sr:)getattrr$)r4attrs r6 __getattr__zAssistantModel.__getattr__Dst~~t,,r8N file_pathr1timeoutreturnc t|d5}|j|||cdddS#1swYyxYw#t$rtd|dt$rtd|dwxYw)a Uploads a file from the specified path to this assistant for internal processing. :param file_path: The path to the file that needs to be uploaded. :type file_path: str, required :param metadata: Optional metadata dictionary to be attached to the file. :type metadata: Optional[dict[str, any]], optional :param timeout: Specify the number of seconds to wait until file processing is done. If None, wait indefinitely; if >=0, time out after this many seconds; if -1, return immediately and do not wait. Default: None :type timeout: int, optional :return: FileModel object with the following properties: - id: The UUID of the uploaded file. - name: The name of the uploaded file. - created_on: The timestamp of when the file was created. - updated_on: The timestamp of the last update to the file. - metadata: Metadata associated with the file. - status: The status of the file. Example: >>> assistant = (...).assistant.Assistant("assistant_name") >>> file_model = assistant.upload_file(file_path="/path/to/file.txt") # use the default timeout >>> print(file_model) {'created_on': '2024-06-02T19:48:00Z', 'id': '070513b3-022f-4966-b583-a9b12e0920ff', 'metadata': None, 'name': 'tiny_file.txt', 'status': 'Available', 'updated_on': '2024-06-02T19:48:00Z'} rbNzError: The file at z was not found.z"Error: Could not read the file at .)open_upload_file_streamFileNotFoundError ExceptionIOError)r4rEr1rFfiles r6 upload_filezAssistantModel.upload_fileGsN Oi& I$//hH I I I  N1)OLM M O@ 1MN N Os 7+ 74770A'stream file_namec6||_|j|||S)a Uploads a file-like stream to the assistant for internal processing. Note: for text files, the stream must be encoded in UTF-8. :param stream: BytesIO stream containing the file bytes to be uploaded. :type stream: BytesIO, required :param file_name: The file name to associate with the stream. :type file_name: str, required :param metadata: Optional metadata dictionary to be attached to the file. :type metadata: Optional[dict[str, any]], optional :param timeout: Specify the number of seconds to wait until file processing is done. If None, wait indefinitely; if >=0, time out after this many seconds; if -1, return immediately and do not wait. Default: None :type timeout: int, optional :return: FileModel object representing the uploaded file. Example: >>> assistant = (...).assistant.Assistant("assistant_name") Example: >>> from io import BytesIO >>> # Create a BytesIO stream with some binary data. >>> text = "Hello, world!" >>> stream = BytesIO(text.encode("utf-8")) >>> # Instantiate your assistant object (assuming proper initialization). >>> assistant = Assistant("assistant_name") >>> # Upload the stream with a specified file name. >>> file_model = assistant.upload_bytes_stream(stream, "myfile.txt") >>> print(file_model) {'created_on': '2024-06-02T19:48:00Z', 'id': '070513b3-022f-4966-b583-a9b12e0920ff', 'metadata': None, 'name': 'myfile.txt', 'status': 'Available', 'updated_on': '2024-06-02T19:48:00Z'} )r.rL)r4rRrSr1rFs r6upload_bytes_streamz"AssistantModel.upload_bytes_streamvs!\  '''BBr8 file_streamc|rE|jj|jj|t j |n0|jj|jj|}|dk(rt j|S|w|jdk(rtjd|j|j}|jdk(rtd|j|jdk(rgn|jdk(rv|dk\rqtjd|dz}|j|j}|jdk(rtd|j|jdk(r|dk\rq|r|dkrtd |jd t j|S) N)assistant_namerPr1)rXrP ProcessingProcessingFailedzFile processing failed. Error: rz$Please call `pc.assistant.Assistant(z)` to confirm assistant status.)r-rQr$r.jsondumpsr from_openapir2timesleep describe_fileidrN error_message TimeoutError)r4rVr1rF upload_resps r6rLz"AssistantModel._upload_file_streams ((44#'>>#6#6$!ZZ15 --99#'>>#6#6[: "} --k::!((L8JJqM"&"4"4[^^"DK"))-??'*I+JcJcId(eff "((L8 "((L8W\JJqMqLG"&"4"4[^^"DK"))-??'*I+JcJcId(eff "((L8W\7Q; >tyykIhi ))+6 6r8file_id include_urlc|r@|jj|j|t|j }n'|jj|j|}t j |S)a' Describes a file with the specified file_id from this assistant. Includes information on its status and metadata. :param : The file id of the file to be described :type file_id: str, required :param include_url: If True, the signed URL of the file is included in the response. :type include_url: bool, optional :return: FileModel object with the following properties: - id: The UUID of the requested file. - name: The name of the requested file. - created_on: The timestamp of when the file was created. - updated_on: The timestamp of the last update to the file. - metadata: Metadata associated with the file. - status: The status of the file. Example: >>> assistant = (...).assistant.Assistant("model_name") >>> file_model = assistant.upload_file(file_path="/path/to/file.txt") # use the default timeout >>> print(file_model) {'created_on': '2024-06-02T19:48:00Z', 'id': '070513b3-022f-4966-b583-a9b12e0290ff', 'metadata': None, 'name': 'tiny_file.txt', 'status': 'Available', 'updated_on': '2024-06-02T19:48:00Z'} >>> assistant.describe_file(file_id='070513b3-022f-4966-b583-a9b12e0290ff') {'created_on': '2024-06-02T19:48:00Z', 'id': '070513b3-022f-4966-b583-a9b12e0290ff', 'metadata': None, 'name': 'tiny_file.txt', 'status': 'Available', 'updated_on': '2024-06-02T19:48:00Z'} )rXassistant_file_idrhrXrj)r-rbr.r;lowerrr_)r4rgrhrPs r6rbzAssistantModel.describe_filesxJ ++99#yy") ,224:D ++99#yy"):D%%d++r8filterc(|r:|jj|jtj|n$|jj|j}|j Dcgc]}t j|c}Scc}w)aZ Lists all uploaded files in this assistant. :return: List of FileModel objects with the following properties: - id: The UUID of the requested file. - name: The name of the requested file. - created_on: The timestamp of when the file was created. - updated_on: The timestamp of the last update to the file. - metadata: Metadata associated with the file. - status: The status of the file. Example: >>> assistant = (...).assistant.Assistant("model_name") >>> assistant.list_files() [{'created_on': '2024-06-02T19:48:00Z', 'id': '070513b3-022f-4966-b583-a9b12e0290ff', 'metadata': None, 'name': 'tiny_file.txt', 'status': 'Available', 'updated_on': '2024-06-02T19:48:00Z'}, ...] )rm)r- list_filesr.r]r^filesrr_)r4rm files_resprPs r6rozAssistantModel.list_filess{2  $ $ / / $**V"4 0 6))44TYY?  :D9I9IJ &&t,JJJs0Bc|jj|j||dk(ry|@|j|}|rzt j d |j|}|r+nN|j|}|r:|dk\r5t j d|dz} |j|}|r|dk\r5|r |dkrtdjdyy#t $rd}YwxYw#t $rd}YIwxYw) a Deletes a file with the specified file_id from this assistant. :param file_path: The path to the file that needs to be uploaded. :type file_path: str, required :type timeout: int, optional :param timeout: Specify the number of seconds to wait until file processing is done. If None, wait indefinitely; if >=0, time out after this many seconds; if -1, return immediately and do not wait. Default: None Example: >>> assistant = (...).assistant.Assistant("model_name") >>> assistant.delete_file(file_id='070513b3-022f-4966-b583-a9b12e0290ff') # use the default timeout >>> assistant.list_files() [] rkrYN)rgr[rz@Please call the describe_model API ({}) to confirm model status.zDhttps://www.pinecone.io/docs/api/operation/assistant/describe_model/) r- delete_filer.rbr`rarNreformat)r4rgrFrPs r6rszAssistantModel.delete_file$s%"   ,,99 -  b=  ?%%g%6D 1  --g->D%%g%6D7a< 1 1  --g->D 7a< w{V]]^ #7! D ! D s$C!$C2! C/.C/2 D?Dmessagescv|Dcgc])}t|trtj|n|+c}Scc}wr:) isinstancedictr from_dict)clsrumessages r6_parse_messageszAssistantModel._parse_messagesWs0dlmY`j$.G!!'*WTmmms.6model temperaturec|d}n1|tvr)tddjdtDd|j|}|r|j ||||S|j ||||S)a Performs a chat completion request to the following assistant. Use this method if you want the response output to be OpenAI's chat completion format. :param messages: The current context for the chat request. The final element in the list represents the user query to be made from this context. :type messages: List[Message] where Message requires the following: Message: - role: str, the role of the context ('user' or 'agent') - content: str, the content of the context Alternatively, you can pass a list of dictionaries with the following keys: - role: str, the role of the context ('user' or 'agent') - content: str, the content of the context :param model: The large language model to use for answer generation. If this flag is set to 'None', then the model used is OpenAI's GPT-4o. :type model: str | None (default 'gpt-4o' in case `None` is passed) :param temperature: Controls the randomness of the model's output: lower values make responses more deterministic, while higher values increase creativity and variability. If the model does not support a temperature parameter, the parameter will be ignored. :type temperature: float | None (default 0.0 in case `None` is passed) :param filter: Optional dictionary to filter which documents can be used in this query. Use this to narrow down the context for the assistant's response. :type filter: Optional[dict[str, any]] (default None) Example filter: { "genre": {"$ne": "documentary"} } This filter would exclude documents with the genre "documentary" from being used in the query. :param stream: If this flag is turned on, then the return type is an Iterable[StreamingChatCompletionChunk] whether data is returned as a generator/stream :type stream: bool (default false) :return: The default result is a ChatCompletionResponse with the following format: { "choices": [ { "finish_reason": "stop", "index": 0, "message": { "content": "The 2020 World Series was played in Texas at Globe Life Field in Arlington.", "role": "assistant" }, "logprobs": null } ], "id": "chatcmpl-7QyqpwdfhqwajicIEznoc6Q47XAyW", "model": "gpt-3.5-turbo-0613", } However, when stream is set to true, the response is an iterable of StreamingChatCompletionChunks. See examples below: { "choices": [ { "finish_reason": null, "index": 0, "delta": { "content": "The", "role": "" }, "logprobs": null } ], "id": "chatcmpl-7QyqpwdfhqwajicIEznoc6Q47XAyW", "model": "gpt-3.5-turbo-0613", } Example: >>> from pinecone_plugins.assistant.models import Message >>> assistant = (...).assistant.Assistant("planets-km") >>> messages = [Message(role='user', content='How old is the earth')] >>> resp = assistant.chat_completions(messages=messages) >>> print(resp) {'choices': [{'finish_reason': 'stop', 'index': 0, 'message': {'content': 'The age of the Earth is estimated to be ' 'about 4.54 billion years, based on ' 'evidence from radiometric age dating of ' 'meteorite material and Earth rocks, as ' 'well as lunar samples. This estimate has ' 'a margin of error of about 1%.', 'role': 'assistant'}}], 'id': 'chatcmpl-9VmkSD9s7rfP28uScLlheookaSwcB', 'model': 'planets-km'} Streaming example: >>> resp = assistant.chat_completions(messages=messages, stream=True) >>> for chunk in resp: if chunk: print(chunk) [{'choices': [{'finish_reason': 'stop', 'index': 0, 'delta': {'content': 'The age of the Earth is estimated to be ' 'about 4.54 billion years, based on ' 'evidence from radiometric age dating of ' 'meteorite material and Earth rocks, as ' 'well as lunar samples. This estimate has ' 'a margin of error of about 1%.', 'role': 'assistant'}}], 'id': 'chatcmpl-9VmkSD9s7rfP28uScLlheookaSwcB', 'model': 'gpt-4o'}, ... ] r%!Invalid model. Valid options are , c3(K|] }d|d yw`N.0ms r6 z2AssistantModel.chat_completions..=W1!Ah=WrJ)rur}rmr~)MODELS ValueErrorjoinr|_chat_completions_streaming_chat_completions_single)r4rurmrRr}r~s r6chat_completionszAssistantModel.chat_completions[s` =E & 3DII=WPV=W4W3XXYZ ''1 33!v;4 00!v;1 r8c|Dcgc]#}t|j|j%}}||d}|r||d<|||d<tdi|}|jj |j |}tj|Scc}w)Nrolecontent)rur}rmr~)rXsearch_completionsr) rrrChatCompletionsRequestr-chat_completion_assistantr.rr_) r4rur}rmr~ctxkwargs chat_requestresults r6rz'AssistantModel._chat_completions_singlesIQ ADLchh <  '7 %F8   "$/F= !-77 ))CC99D &226:: s(B c#JK|jj}|jd|jd}|dd}|Dcgc] }t |}}|d|d} |r|| d<||| d< t j ||| d d } | j| jD]Z} | s| jd } | jd r| d d} tj| } tj| }|\ycc}w#t$r}t!d|d}~wwxYww)N/chat/z/chat/completionsapplication/jsonzapi-keyz Content-TypeT)rurRr}rmr~<headersr]rFrRutf-8data:r[%Error in chat completions streaming: )r+api_keyr)r.varsrequestspostraise_for_status iter_linesdecode startswithr]loadsrryrNr)r4rur}rmr~rbase_urlrr{rresponselinedata json_datareses r6rz*AssistantModel._chat_completions_streamings8++%%ii[tyyk1BC%7IJ19:gDM::'4%H  &GH   "%0GM " J}}'DH  % % ' ++- ;;w/Dw/#ABx $ 4 0I6@@KCI ;. JDQCHI I Js;D# D  DD  D# json_responseinclude_highlightscontext_optionsc ^|d}n1|tvr)tddjdtDd|r |r td|j|}t |t rt j|n|}|r|j||||||S|j|||||||S) aT Performs a chat request to the following assistant. :param messages: The current context for the chat request. The final element in the list represents the user query to be made from this context. :type messages: List[Message] where Message requires the following: Message: - role: str, the role of the context ('user' or 'agent') - content: str, the content of the context Alternatively, you can pass a list of dictionaries with the following keys: - role: str, the role of the context ('user' or 'agent') - content: str, the content of the context :param model: The large language model to use for answer generation. If this flag is set to 'None', then the model used is OpenAI's GPT-4o. :type model: str | None (default 'gpt-4o' in case `None` is passed) :param temperature: Controls the randomness of the model's output: lower values make responses more deterministic, while higher values increase creativity and variability. If the model does not support a temperature parameter, the parameter will be ignored. :type temperature: float | None (default 0.0 in case `None` is passed) :param json_response: If true, the assistant will be instructed to return a JSON response. Cannot be used with streaming. :type json_response: bool (default False) :param include_highlights: If true, the assistant will be instructed to mark highlights for citations. Highlights are small snippets of text that are relevant to the citation. :type include_highlights: bool (default False) :param filter: Optional dictionary to filter which documents can be used in this query. Use this to narrow down the context for the assistant's response. :type filter: Optional[dict[str, any]] (default None) Example filter: { "genre": {"$ne": "documentary"} } This filter would exclude documents with the genre "documentary" from being used in the query. :param stream: If this flag is turned on, then the return type is an Iterable[StreamChatResponseChunk] whether data is returned as a generator/stream :type stream: bool (default false) :param context_options: Option to control the context snippets sent to the LLM. :type context_options: Optional[ContextOptions] (default None) ContextOptions: - top_k: int, the number of context snippets to use. Default is 16. Maximum is 64. - snippet_size: int, the maximum context snippet size. Default is 2048 tokens. Minimum is 512 tokens. Maximum is 8192 tokens. Alternatively, you can pass a dictionary with the following keys: - top_k: int, the number of context snippets to use. - snippet_size: int, the maximum context snippet size. :return: The default result is a ChatModel with the following format: { "finish_reason": "stop", "index": 0, "message": { "content": "The 2020 World Series was played in Texas at Globe Life Field in Arlington.", "role": "assistant" }, "id": "chatcmpl-7QyqpwdfhqwajicIEznoc6Q47XAyW", "model": "gpt-3.5-turbo-0613", "citations": [ { "position": 3, "references": [ { "file": { 'created_on': '2024-06-02T19:48:00Z', 'id': '070513b3-022f-4966-b583-a9b12e0290ff', 'metadata': None, 'name': 'tiny_file.txt', 'status': 'Available', 'updated_on': '2024-06-02T19:48:00Z' }, "pages": [1, 2, 3], "highlight": Null, } ], } ] } However, when stream is set to true, the response is an stream of StreamChatResponseChunks. This can be one of the following types: - StreamChatResponseMessageStart: {'type': 'message_start', 'id': '0000000000000000468323be9d266e55', 'model': 'gpt-4o-2024-11-20', 'role': 'assistant'} - StreamChatResponseContentDelta {'type': 'content_chunk', 'id': '0000000000000000468323be9d266e55', 'model': 'gpt-4o-2024-11-20', 'delta': {'content': 'The'}} - StreamChatResponseCitation {'type': 'citation', 'id': '0000000000000000116990b44044d21e', 'model': 'gpt-4o-2024-11-20', 'citation': {'position': 247, 'references': [{'id': 's0', 'file': {'status': 'Available', 'id': '985edb6c-f649-4334-8f14-9a16b7039ab6', 'name': 'PEPSICO_2022_10K.pdf', 'size': 2993516, 'metadata': None, 'updated_on': '2024-08-08T15:41:58.839846634Z', 'created_on': '2024-08-08T15:41:07.427879083Z', 'percent_done': 0.0, 'signed_url': 'example.com'}, 'pages': [32]}]}} - StreamChatResponseMessageEnd {'type': 'message_end', 'id': '0000000000000000116990b44044d21e', 'model': 'gpt-4o-2024-11-20', 'finish_reason': 'stop', 'usage': {'prompt_tokens': 1, 'completion_tokens': 1, 'total_tokens': 2}} Example: >>> from pinecone_plugins.assistant.models import Message >>> assistant = (...).assistant.Assistant("planets-km") >>> messages = [Message(role='user', content='How old is the earth')] >>> resp = assistant.chat(messages=messages, context_options=ContextOptions(top_k=10, snippet_size=4096)) >>> print(resp) { "finish_reason": "stop", "index": 0, "message": { 'content': 'The age of the Earth is estimated to be ' 'about 4.54 billion years, based on ' 'evidence from radiometric age dating of ' 'meteorite material and Earth rocks, as ' 'well as lunar samples. This estimate has ' 'a margin of error of about 1%.', 'role': 'assistant' }, 'id': 'chatcmpl-7QyqpwdfhqwajicIEznoc6Q47XAyW', "model": "gpt-3.5-turbo-0613", "citations": [ { "position": 3, "references": [ { "file": { 'created_on': '2024-06-02T19:48:00Z', 'id': '070513b3-022f-4966-b583-a9b12e0290ff', 'metadata': None, 'name': 'tiny_file.txt', 'status': 'Available', 'updated_on': '2024-06-02T19:48:00Z' }, "pages": [1, 2, 3], "highlight": Null, } ], } ] } Streaming example: >>> resp = assistant.chat(messages=messages, stream=True, context_options={"top_k":10, "snippet_size":4096}) >>> for chunk in resp: if chunk: print(chunk) [{'type': 'message_start', 'id': '0000000000000000468323be9d266e55', 'model': 'gpt-4o-2024-11-20', 'role': 'assistant'}, {'type': 'content_chunk', 'id': '0000000000000000468323be9d266e55', 'model': 'gpt-4o-2024-11-20', 'delta': {'content': 'The'}}, ... {'type': 'message_end', 'id': '0000000000000000116990b44044d21e', 'model': 'gpt-4o-2024-11-20', 'finish_reason': 'stop', 'usage': {'prompt_tokens': 1, 'completion_tokens': 1, 'total_tokens': 2}}] r%rrc3(K|] }d|d ywrrrs r6rz&AssistantModel.chat..rrrJz'Cannot use json_response with streaming)rur}rmrrr~)rur}rmrrrr~) rrrr|rwrxrry_chat_streaming _chat_single) r4rurmrRr}r~rrrs r6chatzAssistantModel.chatsz =E & 3DII=WPV=W4W3XXYZ  VFG G''1GQRacgGh.22?Cn} ''vj|O^lw'x x$$heFboEWixFQ$R Rr8c|Dcgc]#}t|j|j%}}||||d} |r|| d<||| d<|Hi} |j|j| d<|j|j| d<| rt d i| | d<t d i| } |jj|j| } tj| Scc}w) Nr)rur}rrrmr~top_k snippet_sizer)rXrr) rrrrrr ChatRequestr-chat_assistantr.rr_) r4rur}r~rmrrrrroptionsr chat_results r6rzAssistantModel._chat_singlesIQ ADLchh <  !*"4   %F8   "$/F= !  &G$$0#2#8#8 ++7*9*F*F',?,J',J()",V, ..==99<> ((557 s(Cc#K|jj}|jd|j}|dd} |D cgc] } t | }} |d||d} |r|| d<||| d<|?i} |j |j | d<|j |j | d <| r| | d < tj|| | d d } | j| jD]}|s|jd }|jdr|dd}tj|}d}|jddk(rt!j"|}n}|jddk(rt%j"|}nS|jddk(rt'j"|}n)|jddk(rt)j"|}|ycc} w#t*$r}t-d|d}~wwxYww)NrrrT)rurRr}rrmr~rrrrrrrr[type message_start content_chunkcitation message_endr)r+rr)r.rrrrrrrrrr]rgetrryrrrrNr)r4rur}r~rmrrrrrr{rrrrrrrrs r6rzAssistantModel._chat_streamings ++%%ii[tyyk2%7IJ19:gDM::'4%_qr  &GH   "%0GM "  &G$$0#2#8#8 ++7*9*F*F'-4)* J}}'DH  % % ' ++- ;;w/Dw/#ABx $ 4 0IC }}V,?<FF%'"v./A<FF%'"v.*<8BB9M"v.-?:DDYOI) -;V JDQCHI I Js<;G8GAG8(?G(C+GG8 G5"G00G55G8queryrrc|s|s|s|st||tdSi}|rE|j|}|Dcgc]#}t|j|j %}}||d<n||d<|r||d<|||d<|||d<t d i|}|jj|j|} tj| Scc}w) a Performs a context request to the following assistant. :param query: The query to be used in the context request. Either one of query or messages may be used to generate the context response. :type query: Optional[str] :param messages: The messages to be used in the context request. Either one of query or messages may be used to generate the context response. :type messages: Optional[List[Message]] :param filter: Optional dictionary to filter which documents can be used in this query. Use this to narrow down the context for the assistant's response. :type filter: Optional[dict[str, any]] (default None) Example filter: { "genre": {"$ne": "documentary"} } This filter would exclude documents with the genre "documentary" from being used in the query. :param top_k: Optional integer to specify the maximum number of context snippets to return. :type top_k: Optional[int] (default None) :param snippet_size: Optional integer to specify the maximum context snippet size. Default is 2048 tokens. Minimum is 512 tokens. Maximum is 8192 tokens. :type snippet_size: Optional[int] (default None) :return: The default result is a ContextResponse with the following format: { "snippets": [ { "type": "text", "content": "The quick brown fox jumps over the lazy dog.", "score": 0.9946, "reference": { "type": "pdf", "file": { "id": "96e6e2de-82b2-494d-8988-7dc88ce2ac01", "metadata": null, "name": "sample.pdf", "percent_done": 1.0, "status": "Available", "created_on": "2024-11-13T14:59:53.369365582Z", "updated_on": "2024-11-13T14:59:55.369365582Z", "signed_url": "https://storage.googleapis.com/..." }, "pages": [1] } } ], "usage": { "completion_tokens": 0, "prompt_tokens": 506, "total_tokens": 506 } } Example: >>> assistant = (...).assistant.Assistant("planets-km") >>> resp = assistant.context(query="What is the age of the earth?") >>> print(resp) zBInvalid Inputs: Exactly one of query or messages must be inputted.rrurrmrr)rXcontext_requestr) printrr|rrrr!r-context_assistantr.rr_) r4rrurmrrrrr raw_responses r6contextzAssistantModel.context'sTxe % "bc c ++H5HT\]S #((CKKH]H]!)F: #F7O %F8   #F7O  #%1F> "(  //AA99+B ++L99'^s(C )NN)Fr:)NFNN)r%NN)NFNNFFN)r%NNFFN)r%NNFN)NNNNN)+__name__ __module__ __qualname__OpenAIAssistantModelr7r=r@rDr;rrxanyintrrQrrUr rLboolrbrrors classmethodrr RawMessager|floatrrrrrrrrr"rrrrrr8r6r'r',s7"6"#$- .2!% -O-O4S>*-O# -O  -Of26%) /C/C/CtCH~. /C c] /C  /Cj.2!% *7*74S>**7# *7  *7X0,S0,x~0,R[0,dK$sCx.!9KT)_K<131#1fnuT']Drs   KKKWXgBBR.C  DM4 #334  C23 e_ :_ :r8