o
    پi9                     @   s0  d Z ddlmZ ddlmZ ddlmZ ddlmZ	 ddl
mZ eejjddd	d
d e	 Zd	dlmZ e dZe Zeee eede ejsde_ded _ded _ded _ded _ded _ded _ded _ded _ded _ded _ded  _ded  _d!ed" _d#ed" _d$ed% _d&ed% _d'ed( _d)ed( _d*ed+ _d,ed+ _d-ed. _d/ed. _d0ed1 _d2ed1 _d3ed4 _d5ed4 _d6ed7 _d8ed7 _d9ed: _d;ed: _d<ed= _d>ed= _d?ed@ _dAed@ _dBedC _dDedC _dEedF _dGedF _dHedI _dJedI _dKedL _dMedL _dNedO _dPedO _dQedR _dSedR _dTedU _dVedU _dWedX _dYedX _dZed[ _d\ed[ _d]ed^ _d_ed^ _d`eda _dbeda _dcedd _deedd _dfedg _dhedg _diedj _dkedj _dledm _dnedm _dS dS )ozGenerated protocol buffer code.    )
descriptor)descriptor_pool)runtime_version)symbol_database)builder          ztrtllm_service.proto)
common_pb2s/  
trtllm_service.prototrtllmcommon.proto"
GenerateRequest

request_id (	)
	tokenized (2.trtllm.TokenizedInput/
sampling_config (2.trtllm.SamplingConfig+
output_config (2.trtllm.OutputConfig

max_tokens (
	streaming (:
guided_decoding (2.trtllm.GuidedDecodingParamsH 
embedding_bias (,
lora_config (2.trtllm.LoraConfigH=
prompt_tuning_config (2.trtllm.PromptTuningConfigH6
multimodal_input (2.trtllm.MultimodalInputH?
kv_cache_retention (2.trtllm.KvCacheRetentionConfigH>
disaggregated_params (2.trtllm.DisaggregatedParamsH6
lookahead_config (2.trtllm.LookaheadConfigH
cache_salt_id (H
arrival_time (H
stop (	
stop_token_ids (

ignore_eos (
bad (	
bad_token_ids (B
_guided_decodingB
_lora_configB
_prompt_tuning_configB
_multimodal_inputB
_kv_cache_retentionB
_disaggregated_paramsB
_lookahead_configB
_cache_salt_idB
_arrival_timeJJ	J	
J
"Y
TokenizedInput
original_text (	
input_token_ids (
query_token_ids ("
SamplingConfig

beam_width (
num_return_sequences (
top_k (H 
top_p (H
	top_p_min (H
top_p_reset_ids (H
top_p_decay (H
seed (H
temperature	 (H

min_tokens
 (H'
beam_search_diversity_rate (H
repetition_penalty (H	
presence_penalty (H

frequency_penalty (H!
prompt_ignore_length (H
length_penalty (H
early_stopping (H!
no_repeat_ngram_size (H
min_p (H
beam_width_array (B
_top_kB
_top_pB

_top_p_minB
_top_p_reset_idsB
_top_p_decayB
_seedB
_temperatureB
_min_tokensB
_beam_search_diversity_rateB
_repetition_penaltyB
_presence_penaltyB
_frequency_penaltyB
_prompt_ignore_lengthB
_length_penaltyB
_early_stoppingB
_no_repeat_ngram_sizeB
_min_p"
OutputConfig
logprobs (H 
prompt_logprobs (H
return_context_logits ( 
return_generation_logits (!
exclude_input_from_output (
return_encoder_output (
return_perf_metrics (B
	_logprobsB
_prompt_logprobs"
GuidedDecodingParams:

guide_type (2&.trtllm.GuidedDecodingParams.GuideType
guide (	"
	GuideType
GUIDE_TYPE_UNSPECIFIED 
GUIDE_TYPE_JSON
GUIDE_TYPE_JSON_SCHEMA
GUIDE_TYPE_REGEX
GUIDE_TYPE_EBNF_GRAMMAR
GUIDE_TYPE_STRUCTURAL_TAG"i

LoraConfig
task_id (
weights (H 
config_json (	HB

_weightsB
_config_json"-
PromptTuningConfig
embedding_table ("%
MultimodalInput

image_data ("=
KvCacheRetentionConfig
policy (	
config_json (	"
DisaggregatedParams=
request_type (2'.trtllm.DisaggregatedParams.RequestType
ctx_request_id (	=
context_phase_params (2.trtllm.ContextPhaseParamsH "w
RequestType'
#REQUEST_TYPE_CONTEXT_AND_GENERATION 
REQUEST_TYPE_CONTEXT_ONLY 
REQUEST_TYPE_GENERATION_ONLYB
_context_phase_params"I
ContextPhaseParams
first_gen_token_id (
kv_cache_blocks ("e
LookaheadConfig
max_window_size (
max_ngram_size (!
max_verification_set_size ("
GenerateResponse

request_id (	,
chunk (2.trtllm.GenerateStreamChunkH ,
complete (2.trtllm.GenerateCompleteH &
error (2.trtllm.GenerateErrorH B

response"
GenerateStreamChunk
	token_ids (
sequence_index (
prompt_tokens (
completion_tokens (
cached_tokens (&
logprobs (2.trtllm.TokenLogprob"
GenerateComplete
output_token_ids (
sequence_index (
finish_reason (	
matched_stop_str (	H 
matched_token_id (H 
prompt_tokens (
completion_tokens (
cached_tokens (&
logprobs (2.trtllm.TokenLogprob-
prompt_logprobs	 (2.trtllm.TokenLogprob.
perf_metrics
 (2.trtllm.PerfMetricsH
context_logits (H
generation_logits (HB
matched_stopB
_perf_metricsB
_context_logitsB
_generation_logits"[
TokenLogprob
token_id (
logprob ((
top_logprobs (2.trtllm.TopLogprob"/

TopLogprob
token_id (
logprob ("
PerfMetrics
arrival_time (
first_scheduled_time (
first_token_time (
last_token_time (
kv_cache_transfer_start (
kv_cache_transfer_end (
kv_cache_size ("<
GenerateError
message (	
type (	
code ("M
EmbedRequest

request_id (	)
	tokenized (2.trtllm.TokenizedInput"M
EmbedResponse

request_id (	
	embedding (
prompt_tokens ("
HealthCheckRequest"%
HealthCheckResponse
status (	""
AbortRequest

request_id (	"1
AbortResponse
success (
message (	"
GetModelInfoRequest"
GetModelInfoResponse
model_id (	
max_input_len (
max_seq_len (
max_batch_size (

vocab_size (
hidden_size (

num_layers (
	num_heads (
supported_features	 (	"
GetServerInfoRequest"
GetServerInfoResponse
version (	
backend (	
tensor_parallel_size (
pipeline_parallel_size (
context_parallel_size (

world_size (2
TrtllmService?
Generate.trtllm.GenerateRequest.trtllm.GenerateResponse04
Embed.trtllm.EmbedRequest.trtllm.EmbedResponseF
HealthCheck.trtllm.HealthCheckRequest.trtllm.HealthCheckResponse4
Abort.trtllm.AbortRequest.trtllm.AbortResponseI
GetModelInfo.trtllm.GetModelInfoRequest.trtllm.GetModelInfoResponseL
GetServerInfo.trtllm.GetServerInfoRequest.trtllm.GetServerInfoResponseZ
GetTokenizer$.smg.grpc.common.GetTokenizerRequest".smg.grpc.common.GetTokenizerChunk0_
SubscribeKvEvents).smg.grpc.common.SubscribeKvEventsRequest.smg.grpc.common.KvEventBatch0bproto3trtllm_service_pb2N/   _GENERATEREQUESTi  i  _TOKENIZEDINPUTix  i{  _SAMPLINGCONFIGi  i  _OUTPUTCONFIGi  i  _GUIDEDDECODINGPARAMSi	  iQ	  _GUIDEDDECODINGPARAMS_GUIDETYPEi	  _LORACONFIGif
  ih
  _PROMPTTUNINGCONFIGi
  i
  _MULTIMODALINPUTi
  i
  _KVCACHERETENTIONCONFIGi
  i
  _DISAGGREGATEDPARAMSi;  i   _DISAGGREGATEDPARAMS_REQUESTTYPEi"  i=  _CONTEXTPHASEPARAMSi  i  _LOOKAHEADCONFIGi  i  _GENERATERESPONSEi  i  _GENERATESTREAMCHUNKiZ  i]  _GENERATECOMPLETEiG  iI  _TOKENLOGPROBi  i  _TOPLOGPROBi  i  _PERFMETRICSi  i  _GENERATEERRORi  i  _EMBEDREQUESTi0  i2  _EMBEDRESPONSEi  i  _HEALTHCHECKREQUESTi  i  _HEALTHCHECKRESPONSEi  i  _ABORTREQUESTi  i  _ABORTRESPONSEi  i  _GETMODELINFOREQUESTi*  i-  _GETMODELINFORESPONSEi  i  _GETSERVERINFOREQUESTi  i   _GETSERVERINFORESPONSEi  i  _TRTLLMSERVICEi'  )__doc__google.protobufr   _descriptorr   _descriptor_poolr   _runtime_versionr   _symbol_databasegoogle.protobuf.internalr   _builderValidateProtobufRuntimeVersionDomainPUBLICDefault_sym_dbr
   r   common__pb2AddSerializedFile
DESCRIPTORglobals_globalsBuildMessageAndEnumDescriptorsBuildTopDescriptorsAndMessages_USE_C_DESCRIPTORS_loaded_options_serialized_start_serialized_end rF   rF   _/home/ubuntu/.local/lib/python3.10/site-packages/smg_grpc_proto/generated/trtllm_service_pb2.py<module>   s   































































