Mirror/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2024-06-07 19:40:48 +00:00
Code Issues Packages Projects Releases Wiki Activity
048b81373d
LocalAI/extra/grpc/autogptq/backend_pb2.py

62 lines
8.9 KiB
Python
Raw Normal View History

feat: add initial AutoGPTQ backend implementation
2023-08-07 20:39:10 +00:00
# -*- coding: utf-8 -*-
# Generated by the protocol buffer compiler. DO NOT EDIT!
# source: backend.proto
"""Generated protocol buffer code."""
from google.protobuf import descriptor as _descriptor
from google.protobuf import descriptor_pool as _descriptor_pool
from google.protobuf import symbol_database as _symbol_database
from google.protobuf.internal import builder as _builder
# @@protoc_insertion_point(imports)
_sym_db = _symbol_database.Default()
feat(speculative-sampling): allow to specify a draft model in the model config (#1052) **Description** This PR fixes #1013. It adds `draft_model` and `n_draft` to the model YAML config in order to load models with speculative sampling. This should be compatible as well with grammars. example: ```yaml backend: llama context_size: 1024 name: my-model-name parameters: model: foo-bar n_draft: 16 draft_model: model-name ``` --------- Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
2023-09-14 15:44:16 +00:00
DESCRIPTOR = _descriptor_pool.Default().AddSerializedFile(b'\n\rbackend.proto\x12\x07\x62\x61\x63kend\"\x0f\n\rHealthMessage\"\x96\x06\n\x0ePredictOptions\x12\x0e\n\x06Prompt\x18\x01 \x01(\t\x12\x0c\n\x04Seed\x18\x02 \x01(\x05\x12\x0f\n\x07Threads\x18\x03 \x01(\x05\x12\x0e\n\x06Tokens\x18\x04 \x01(\x05\x12\x0c\n\x04TopK\x18\x05 \x01(\x05\x12\x0e\n\x06Repeat\x18\x06 \x01(\x05\x12\r\n\x05\x42\x61tch\x18\x07 \x01(\x05\x12\r\n\x05NKeep\x18\x08 \x01(\x05\x12\x13\n\x0bTemperature\x18\t \x01(\x02\x12\x0f\n\x07Penalty\x18\n \x01(\x02\x12\r\n\x05\x46\x31\x36KV\x18\x0b \x01(\x08\x12\x11\n\tDebugMode\x18\x0c \x01(\x08\x12\x13\n\x0bStopPrompts\x18\r \x03(\t\x12\x11\n\tIgnoreEOS\x18\x0e \x01(\x08\x12\x19\n\x11TailFreeSamplingZ\x18\x0f \x01(\x02\x12\x10\n\x08TypicalP\x18\x10 \x01(\x02\x12\x18\n\x10\x46requencyPenalty\x18\x11 \x01(\x02\x12\x17\n\x0fPresencePenalty\x18\x12 \x01(\x02\x12\x10\n\x08Mirostat\x18\x13 \x01(\x05\x12\x13\n\x0bMirostatETA\x18\x14 \x01(\x02\x12\x13\n\x0bMirostatTAU\x18\x15 \x01(\x02\x12\x12\n\nPenalizeNL\x18\x16 \x01(\x08\x12\x11\n\tLogitBias\x18\x17 \x01(\t\x12\r\n\x05MLock\x18\x19 \x01(\x08\x12\x0c\n\x04MMap\x18\x1a \x01(\x08\x12\x16\n\x0ePromptCacheAll\x18\x1b \x01(\x08\x12\x15\n\rPromptCacheRO\x18\x1c \x01(\x08\x12\x0f\n\x07Grammar\x18\x1d \x01(\t\x12\x0f\n\x07MainGPU\x18\x1e \x01(\t\x12\x13\n\x0bTensorSplit\x18\x1f \x01(\t\x12\x0c\n\x04TopP\x18 \x01(\x02\x12\x17\n\x0fPromptCachePath\x18! \x01(\t\x12\r\n\x05\x44\x65\x62ug\x18\" \x01(\x08\x12\x17\n\x0f\x45mbeddingTokens\x18# \x03(\x05\x12\x12\n\nEmbeddings\x18$ \x01(\t\x12\x14\n\x0cRopeFreqBase\x18% \x01(\x02\x12\x15\n\rRopeFreqScale\x18& \x01(\x02\x12\x1b\n\x13NegativePromptScale\x18\' \x01(\x02\x12\x16\n\x0eNegativePrompt\x18( \x01(\t\x12\x0e\n\x06NDraft\x18) \x01(\x05\"\x18\n\x05Reply\x12\x0f\n\x07message\x18\x01 \x01(\x0c\"\xf0\x05\n\x0cModelOptions\x12\r\n\x05Model\x18\x01 \x01(\t\x12\x13\n\x0b\x43ontextSize\x18\x02 \x01(\x05\x12\x0c\n\x04Seed\x18\x03 \x01(\x05\x12\x0e\n\x06NBatch\x18\x04 \x01(\x05\x12\x11\n\tF16Memory\x18\x05 \x01(\x08\x12\r\n\x05MLock\x18\x06 \x01(\x08\x12\x0c\n\x04MMap\x18\x07 \x01(\x08\x12\x11\n\tVocabOnly\x18\x08 \x01(\x08\x12\x0f\n\x07LowVRAM\x18\t \x01(\x08\x12\x12\n\nEmbeddings\x18\n \x01(\x08\x12\x0c\n\x04NUMA\x18\x0b \x01(\x08\x12\x12\n\nNGPULayers\x18\x0c \x01(\x05\x12\x0f\n\x07MainGPU\x18\r \x01(\t\x12\x13\n\x0bTensorSplit\x18\x0e \x01(\t\x12\x0f\n\x07Threads\x18\x0f \x01(\x05\x12\x19\n\x11LibrarySearchPath\x18\x10 \x01(\t\x12\x14\n\x0cRopeFreqBase\x18\x11 \x01(\x02\x12\x15\n\rRopeFreqScale\x18\x12 \x01(\x02\x12\x12\n\nRMSNormEps\x18\x13 \x01(\x02\x12\x0c\n\x04NGQA\x18\x14 \x01(\x05\x12\x11\n\tModelFile\x18\x15 \x01(\t\x12\x0e\n\x06\x44\x65vice\x18\x16 \x01(\t\x12\x11\n\tUseTriton\x18\x17 \x01(\x08\x12\x15\n\rModelBaseName\x18\x18 \x01(\t\x12\x18\n\x10UseFastTokenizer\x18\x19 \x01(\x08\x12\x14\n\x0cPipelineType\x18\x1a \x01(\t\x12\x15\n\rSchedulerType\x18\x1b \x01(\t\x12\x0c\n\x04\x43UDA\x18\x1c \x01(\x08\x12\x10\n\x08\x43\x46GScale\x18\x1d \x01(\x02\x12\x0f\n\x07IMG2IMG\x18\x1e \x01(\x08\x12\x11\n\tCLIPModel\x18\x1f \x01(\t\x12\x15\n\rCLIPSubfolder\x18 \x01(\t\x12\x10\n\x08\x43LIPSkip\x18! \x01(\x05\x12\x11\n\tTokenizer\x18\" \x01(\t\x12\x10\n\x08LoraBase\x18# \x01(\t\x12\x13\n\x0bLoraAdapter\x18$ \x01(\t\x12\x11\n\tNoMulMatQ\x18% \x01(\x08\x12\x12\n\nDraftModel\x18\' \x01(\t\x12\x11\n\tAudioPath\x18& \x01(\t\"*\n\x06Result\x12\x0f\n\x07message\x18\x01 \x01(\t\x12\x0f\n\x07success\x18\x02 \x01(\x08\"%\n\x0f\x45mbeddingResult\x12\x12\n\nembeddings\x18\x01 \x03(\x02\"C\n\x11TranscriptRequest\x12\x0b\n\x03\x64st\x18\x02 \x01(\t\x12\x10\n\x08language\x18\x03 \x01(\t\x12\x0f\n\x07threads\x18\x04 \x01(\r\"N\n\x10TranscriptResult\x12,\n\x08segments\x18\x01 \x03(\x0b\x32\x1a.backend.TranscriptSegment\x12\x0c\n\x04text\x18\x02 \x01(\t\"Y\n\x11TranscriptSegment\x12\n\n\x02id\x18\x01 \x01(\x05\x12\r\n\x05start\x18\x02 \x01(\x03\x12\x0b\n\x03\x65nd\x18\x03 \x01(\x03\x12\x0c\n\x04text\x18\x04 \x01(\t\x12\x0e\n\x06tokens\x18\x05 \x03(\x05\"\xd7\x01\n\x14GenerateImageRequest\x12\x0e\n\x06height\x18\x01 \x01(\x05\x12\r\n\x05width\x18\x02
feat: add initial AutoGPTQ backend implementation
2023-08-07 20:39:10 +00:00
_globals = globals()
_builder.BuildMessageAndEnumDescriptors(DESCRIPTOR, _globals)
_builder.BuildTopDescriptorsAndMessages(DESCRIPTOR, 'backend_pb2', _globals)
if _descriptor._USE_C_DESCRIPTORS == False:
DESCRIPTOR._options = None
DESCRIPTOR._serialized_options = b'\n\031io.skynet.localai.backendB\016LocalAIBackendP\001Z+github.com/go-skynet/LocalAI/pkg/grpc/proto'
feat(diffusers): be consistent with pipelines, support also depthimg2img (#926) Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
2023-08-18 20:06:24 +00:00
_MEMORYUSAGEDATA_BREAKDOWNENTRY._options = None
_MEMORYUSAGEDATA_BREAKDOWNENTRY._serialized_options = b'8\001'
feat: add initial AutoGPTQ backend implementation
2023-08-07 20:39:10 +00:00
_globals['_HEALTHMESSAGE']._serialized_start=26
_globals['_HEALTHMESSAGE']._serialized_end=41
_globals['_PREDICTOPTIONS']._serialized_start=44
feat(speculative-sampling): allow to specify a draft model in the model config (#1052) **Description** This PR fixes #1013. It adds `draft_model` and `n_draft` to the model YAML config in order to load models with speculative sampling. This should be compatible as well with grammars. example: ```yaml backend: llama context_size: 1024 name: my-model-name parameters: model: foo-bar n_draft: 16 draft_model: model-name ``` --------- Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
2023-09-14 15:44:16 +00:00
_globals['_PREDICTOPTIONS']._serialized_end=834
_globals['_REPLY']._serialized_start=836
_globals['_REPLY']._serialized_end=860
_globals['_MODELOPTIONS']._serialized_start=863
_globals['_MODELOPTIONS']._serialized_end=1615
_globals['_RESULT']._serialized_start=1617
_globals['_RESULT']._serialized_end=1659
_globals['_EMBEDDINGRESULT']._serialized_start=1661
_globals['_EMBEDDINGRESULT']._serialized_end=1698
_globals['_TRANSCRIPTREQUEST']._serialized_start=1700
_globals['_TRANSCRIPTREQUEST']._serialized_end=1767
_globals['_TRANSCRIPTRESULT']._serialized_start=1769
_globals['_TRANSCRIPTRESULT']._serialized_end=1847
_globals['_TRANSCRIPTSEGMENT']._serialized_start=1849
_globals['_TRANSCRIPTSEGMENT']._serialized_end=1938
_globals['_GENERATEIMAGEREQUEST']._serialized_start=1941
_globals['_GENERATEIMAGEREQUEST']._serialized_end=2156
_globals['_TTSREQUEST']._serialized_start=2158
_globals['_TTSREQUEST']._serialized_end=2212
_globals['_TOKENIZATIONRESPONSE']._serialized_start=2214
_globals['_TOKENIZATIONRESPONSE']._serialized_end=2268
_globals['_MEMORYUSAGEDATA']._serialized_start=2271
_globals['_MEMORYUSAGEDATA']._serialized_end=2413
_globals['_MEMORYUSAGEDATA_BREAKDOWNENTRY']._serialized_start=2365
_globals['_MEMORYUSAGEDATA_BREAKDOWNENTRY']._serialized_end=2413
_globals['_STATUSRESPONSE']._serialized_start=2416
_globals['_STATUSRESPONSE']._serialized_end=2589
_globals['_STATUSRESPONSE_STATE']._serialized_start=2522
_globals['_STATUSRESPONSE_STATE']._serialized_end=2589
_globals['_BACKEND']._serialized_start=2592
_globals['_BACKEND']._serialized_end=3220
feat: add initial AutoGPTQ backend implementation
2023-08-07 20:39:10 +00:00
# @@protoc_insertion_point(module_scope)
Reference in New Issue Copy Permalink