youngkingdom/dify
1
0
Fork 0
mirror of https://github.com/langgenius/dify.git synced 2026-05-04 17:38:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge main into feat/plugin

Browse Source
This commit is contained in:
Yeuoly
2024-09-26 12:59:06 +08:00
parent ea497f828f a0b0809b1c
commit dc255cc154
69 changed files with 946 additions and 295 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
api/core/app/apps/base_app_runner.py
View File

@ -309,7 +309,7 @@ class AppRunner:
if not prompt_messages:
prompt_messages = result.prompt_messages
if not usage and result.delta.usage:
if result.delta.usage:
usage = result.delta.usage
if not usage:

9
api/core/embedding/cached_embedding.py
View File

@ -5,6 +5,7 @@ from typing import Optional, cast
import numpy as np
from sqlalchemy.exc import IntegrityError
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_manager import ModelInstance
from core.model_runtime.entities.model_entities import ModelPropertyKey
from core.model_runtime.model_providers.__base.text_embedding_model import TextEmbeddingModel
@ -56,7 +57,9 @@ class CacheEmbedding(Embeddings):
for i in range(0, len(embedding_queue_texts), max_chunks):
batch_texts = embedding_queue_texts[i : i + max_chunks]
embedding_result = self._model_instance.invoke_text_embedding(texts=batch_texts, user=self._user)
embedding_result = self._model_instance.invoke_text_embedding(
texts=batch_texts, user=self._user, input_type=EmbeddingInputType.DOCUMENT
)
for vector in embedding_result.embeddings:
try:
@ -100,7 +103,9 @@ class CacheEmbedding(Embeddings):
redis_client.expire(embedding_cache_key, 600)
return list(np.frombuffer(base64.b64decode(embedding), dtype="float"))
try:
embedding_result = self._model_instance.invoke_text_embedding(texts=[text], user=self._user)
embedding_result = self._model_instance.invoke_text_embedding(
texts=[text], user=self._user, input_type=EmbeddingInputType.QUERY
)
embedding_results = embedding_result.embeddings[0]
embedding_results = (embedding_results / np.linalg.norm(embedding_results)).tolist()

10
api/core/embedding/embedding_constant.py Normal file
View File

@ -0,0 +1,10 @@
from enum import Enum
class EmbeddingInputType(Enum):
"""
Enum for embedding input type.
"""
DOCUMENT = "document"
QUERY = "query"

7
api/core/model_manager.py
View File

@ -3,6 +3,7 @@ import os
from collections.abc import Callable, Generator, Sequence
from typing import IO, Literal, Optional, Union, cast, overload
from core.embedding.embedding_constant import EmbeddingInputType
from core.entities.provider_configuration import ProviderConfiguration, ProviderModelBundle
from core.entities.provider_entities import ModelLoadBalancingConfiguration
from core.errors.error import ProviderTokenNotInitError
@ -194,12 +195,15 @@ class ModelInstance:
tools=tools,
)
def invoke_text_embedding(self, texts: list[str], user: Optional[str] = None) -> TextEmbeddingResult:
def invoke_text_embedding(
self, texts: list[str], user: Optional[str] = None, input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT
) -> TextEmbeddingResult:
"""
Invoke large language model
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
if not isinstance(self.model_type_instance, TextEmbeddingModel):
@ -212,6 +216,7 @@ class ModelInstance:
credentials=self.credentials,
texts=texts,
user=user,
input_type=input_type,
)
def get_text_embedding_num_tokens(self, texts: list[str]) -> int:

23
api/core/model_runtime/model_providers/__base/text_embedding_model.py
View File

@ -4,6 +4,7 @@ from typing import Optional
from pydantic import ConfigDict
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import ModelPropertyKey, ModelType
from core.model_runtime.entities.text_embedding_entities import TextEmbeddingResult
from core.model_runtime.model_providers.__base.ai_model import AIModel
@ -20,35 +21,47 @@ class TextEmbeddingModel(AIModel):
model_config = ConfigDict(protected_namespaces=())
def invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke large language model
Invoke text embedding model
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
self.started_at = time.perf_counter()
try:
return self._invoke(model, credentials, texts, user)
return self._invoke(model, credentials, texts, user, input_type)
except Exception as e:
raise self._transform_invoke_error(e)
@abstractmethod
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke large language model
Invoke text embedding model
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
raise NotImplementedError

18
api/core/model_runtime/model_providers/azure_openai/text_embedding/text_embedding.py
View File

@ -7,6 +7,7 @@ import numpy as np
import tiktoken
from openai import AzureOpenAI
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import AIModelEntity, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.validate import CredentialsValidateFailedError
@ -17,8 +18,23 @@ from core.model_runtime.model_providers.azure_openai._constant import EMBEDDING_
class AzureOpenAITextEmbeddingModel(_CommonAzureOpenAI, TextEmbeddingModel):
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
base_model_name = credentials["base_model_name"]
credentials_kwargs = self._to_credential_kwargs(credentials)
client = AzureOpenAI(**credentials_kwargs)

9
api/core/model_runtime/model_providers/baichuan/text_embedding/text_embedding.py
View File

@ -4,6 +4,7 @@ from typing import Optional
from requests import post
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import (
@ -35,7 +36,12 @@ class BaichuanTextEmbeddingModel(TextEmbeddingModel):
api_base: str = "http://api.baichuan-ai.com/v1/embeddings"
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -44,6 +50,7 @@ class BaichuanTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
api_key = credentials["api_key"]

9
api/core/model_runtime/model_providers/bedrock/text_embedding/text_embedding.py
View File

@ -13,6 +13,7 @@ from botocore.exceptions import (
UnknownServiceError,
)
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import (
@ -30,7 +31,12 @@ logger = logging.getLogger(__name__)
class BedrockTextEmbeddingModel(TextEmbeddingModel):
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -39,6 +45,7 @@ class BedrockTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
client_config = Config(region_name=credentials["aws_region"])

9
api/core/model_runtime/model_providers/cohere/text_embedding/text_embedding.py
View File

@ -5,6 +5,7 @@ import cohere
import numpy as np
from cohere.core import RequestOptions
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import (
@ -25,7 +26,12 @@ class CohereTextEmbeddingModel(TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -34,6 +40,7 @@ class CohereTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
# get model properties

1
api/core/model_runtime/model_providers/fireworks/fireworks.yaml
View File

@ -15,6 +15,7 @@ help:
en_US: https://fireworks.ai/account/api-keys
supported_model_types:
- llm
- text-embedding
configurate_methods:
- predefined-model
provider_credential_schema:

12
api/core/model_runtime/model_providers/fireworks/text_embedding/UAE-Large-V1.yaml Normal file
View File

@ -0,0 +1,12 @@
model: WhereIsAI/UAE-Large-V1
label:
zh_Hans: UAE-Large-V1
en_US: UAE-Large-V1
model_type: text-embedding
model_properties:
context_size: 512
max_chunks: 1
pricing:
input: '0.008'
unit: '0.000001'
currency: 'USD'

0
api/core/model_runtime/model_providers/fireworks/text_embedding/__init__.py Normal file
View File

12
api/core/model_runtime/model_providers/fireworks/text_embedding/gte-base.yaml Normal file
View File

@ -0,0 +1,12 @@
model: thenlper/gte-base
label:
zh_Hans: GTE-base
en_US: GTE-base
model_type: text-embedding
model_properties:
context_size: 512
max_chunks: 1
pricing:
input: '0.008'
unit: '0.000001'
currency: 'USD'

12
api/core/model_runtime/model_providers/fireworks/text_embedding/gte-large.yaml Normal file
View File

@ -0,0 +1,12 @@
model: thenlper/gte-large
label:
zh_Hans: GTE-large
en_US: GTE-large
model_type: text-embedding
model_properties:
context_size: 512
max_chunks: 1
pricing:
input: '0.008'
unit: '0.000001'
currency: 'USD'

12
api/core/model_runtime/model_providers/fireworks/text_embedding/nomic-embed-text-v1.5.yaml Normal file
View File

@ -0,0 +1,12 @@
model: nomic-ai/nomic-embed-text-v1.5
label:
zh_Hans: nomic-embed-text-v1.5
en_US: nomic-embed-text-v1.5
model_type: text-embedding
model_properties:
context_size: 8192
max_chunks: 16
pricing:
input: '0.008'
unit: '0.000001'
currency: 'USD'

12
api/core/model_runtime/model_providers/fireworks/text_embedding/nomic-embed-text-v1.yaml Normal file
View File

@ -0,0 +1,12 @@
model: nomic-ai/nomic-embed-text-v1
label:
zh_Hans: nomic-embed-text-v1
en_US: nomic-embed-text-v1
model_type: text-embedding
model_properties:
context_size: 8192
max_chunks: 16
pricing:
input: '0.008'
unit: '0.000001'
currency: 'USD'

151
api/core/model_runtime/model_providers/fireworks/text_embedding/text_embedding.py Normal file
View File

@ -0,0 +1,151 @@
import time
from collections.abc import Mapping
from typing import Optional, Union
import numpy as np
from openai import OpenAI
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.validate import CredentialsValidateFailedError
from core.model_runtime.model_providers.__base.text_embedding_model import TextEmbeddingModel
from core.model_runtime.model_providers.fireworks._common import _CommonFireworks
class FireworksTextEmbeddingModel(_CommonFireworks, TextEmbeddingModel):
"""
Model class for Fireworks text embedding model.
"""
def _invoke(
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
credentials_kwargs = self._to_credential_kwargs(credentials)
client = OpenAI(**credentials_kwargs)
extra_model_kwargs = {}
if user:
extra_model_kwargs["user"] = user
extra_model_kwargs["encoding_format"] = "float"
context_size = self._get_context_size(model, credentials)
max_chunks = self._get_max_chunks(model, credentials)
inputs = []
indices = []
used_tokens = 0
for i, text in enumerate(texts):
# Here token count is only an approximation based on the GPT2 tokenizer
# TODO: Optimize for better token estimation and chunking
num_tokens = self._get_num_tokens_by_gpt2(text)
if num_tokens >= context_size:
cutoff = int(np.floor(len(text) * (context_size / num_tokens)))
# if num tokens is larger than context length, only use the start
inputs.append(text[0:cutoff])
else:
inputs.append(text)
indices += [i]
batched_embeddings = []
_iter = range(0, len(inputs), max_chunks)
for i in _iter:
embeddings_batch, embedding_used_tokens = self._embedding_invoke(
model=model,
client=client,
texts=inputs[i : i + max_chunks],
extra_model_kwargs=extra_model_kwargs,
)
used_tokens += embedding_used_tokens
batched_embeddings += embeddings_batch
usage = self._calc_response_usage(model=model, credentials=credentials, tokens=used_tokens)
return TextEmbeddingResult(embeddings=batched_embeddings, usage=usage, model=model)
def get_num_tokens(self, model: str, credentials: dict, texts: list[str]) -> int:
"""
Get number of tokens for given prompt messages
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:return:
"""
return sum(self._get_num_tokens_by_gpt2(text) for text in texts)
def validate_credentials(self, model: str, credentials: Mapping) -> None:
"""
Validate model credentials
:param model: model name
:param credentials: model credentials
:return:
"""
try:
# transform credentials to kwargs for model instance
credentials_kwargs = self._to_credential_kwargs(credentials)
client = OpenAI(**credentials_kwargs)
# call embedding model
self._embedding_invoke(model=model, client=client, texts=["ping"], extra_model_kwargs={})
except Exception as ex:
raise CredentialsValidateFailedError(str(ex))
def _embedding_invoke(
self, model: str, client: OpenAI, texts: Union[list[str], str], extra_model_kwargs: dict
) -> tuple[list[list[float]], int]:
"""
Invoke embedding model
:param model: model name
:param client: model client
:param texts: texts to embed
:param extra_model_kwargs: extra model kwargs
:return: embeddings and used tokens
"""
response = client.embeddings.create(model=model, input=texts, **extra_model_kwargs)
return [data.embedding for data in response.data], response.usage.total_tokens
def _calc_response_usage(self, model: str, credentials: dict, tokens: int) -> EmbeddingUsage:
"""
Calculate response usage
:param model: model name
:param credentials: model credentials
:param tokens: input tokens
:return: usage
"""
input_price_info = self.get_price(
model=model, credentials=credentials, tokens=tokens, price_type=PriceType.INPUT
)
usage = EmbeddingUsage(
tokens=tokens,
total_tokens=tokens,
unit_price=input_price_info.unit_price,
price_unit=input_price_info.unit,
total_price=input_price_info.total_amount,
currency=input_price_info.currency,
latency=time.perf_counter() - self.started_at,
)
return usage

18
api/core/model_runtime/model_providers/huggingface_hub/text_embedding/text_embedding.py
View File

@ -6,6 +6,7 @@ import numpy as np
import requests
from huggingface_hub import HfApi, InferenceClient
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import AIModelEntity, FetchFrom, ModelType, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
@ -18,8 +19,23 @@ HUGGINGFACE_ENDPOINT_API = "https://api.endpoints.huggingface.cloud/v2/endpoint/
class HuggingfaceHubTextEmbeddingModel(_CommonHuggingfaceHub, TextEmbeddingModel):
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
client = InferenceClient(token=credentials["huggingfacehub_api_token"])
execute_model = model

9
api/core/model_runtime/model_providers/huggingface_tei/text_embedding/text_embedding.py
View File

@ -1,6 +1,7 @@
import time
from typing import Optional
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import AIModelEntity, FetchFrom, ModelPropertyKey, ModelType, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
@ -23,7 +24,12 @@ class HuggingfaceTeiTextEmbeddingModel(TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -38,6 +44,7 @@ class HuggingfaceTeiTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
server_url = credentials["server_url"]

9
api/core/model_runtime/model_providers/hunyuan/text_embedding/text_embedding.py
View File

@ -9,6 +9,7 @@ from tencentcloud.common.profile.client_profile import ClientProfile
from tencentcloud.common.profile.http_profile import HttpProfile
from tencentcloud.hunyuan.v20230901 import hunyuan_client, models
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import (
@ -26,7 +27,12 @@ class HunyuanTextEmbeddingModel(TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -35,6 +41,7 @@ class HunyuanTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""

43
api/core/model_runtime/model_providers/jina/jina.yaml
View File

@ -67,46 +67,3 @@ model_credential_schema:
required: false
type: text-input
default: '8192'
- variable: task
label:
zh_Hans: 下游任务
en_US: Downstream task
placeholder:
zh_Hans: 选择将使用向量模型的下游任务。模型将返回针对该任务优化的向量。
en_US: Select the downstream task for which the embeddings will be used. The model will return the optimized embeddings for that task.
required: false
type: select
options:
- value: retrieval.query
label:
en_US: retrieval.query
- value: retrieval.passage
label:
en_US: retrieval.passage
- value: separation
label:
en_US: separation
- value: classification
label:
en_US: classification
- value: text-matching
label:
en_US: text-matching
- variable: dimensions
label:
zh_Hans: 输出维度
en_US: Output dimensions
placeholder:
zh_Hans: 输入您的输出维度
en_US: Enter output dimensions
required: false
type: text-input
- variable: late_chunking
label:
zh_Hans: 后期分块
en_US: Late chunking
placeholder:
zh_Hans: 应用后期分块技术来利用模型的长上下文功能来生成上下文块向量化。
en_US: Apply the late chunking technique to leverage the model's long-context capabilities for generating contextual chunk embeddings.
required: false
type: switch

29
api/core/model_runtime/model_providers/jina/text_embedding/text_embedding.py
View File

@ -4,6 +4,7 @@ from typing import Optional
from requests import post
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import AIModelEntity, FetchFrom, ModelPropertyKey, ModelType, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
@ -27,7 +28,7 @@ class JinaTextEmbeddingModel(TextEmbeddingModel):
api_base: str = "https://api.jina.ai/v1"
def _to_payload(self, model: str, texts: list[str], credentials: dict) -> dict:
def _to_payload(self, model: str, texts: list[str], credentials: dict, input_type: EmbeddingInputType) -> dict:
"""
Parse model credentials
@ -44,23 +45,20 @@ class JinaTextEmbeddingModel(TextEmbeddingModel):
data = {"model": model, "input": [transform_jina_input_text(model, text) for text in texts]}
task = credentials.get("task")
dimensions = credentials.get("dimensions")
late_chunking = credentials.get("late_chunking")
if task is not None:
data["task"] = task
if dimensions is not None:
data["dimensions"] = int(dimensions)
if late_chunking is not None:
data["late_chunking"] = late_chunking
# model specific parameters
if model == "jina-embeddings-v3":
# set `task` type according to input type for the best performance
data["task"] = "retrieval.query" if input_type == EmbeddingInputType.QUERY else "retrieval.passage"
return data
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -69,6 +67,7 @@ class JinaTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
api_key = credentials["api_key"]
@ -81,7 +80,7 @@ class JinaTextEmbeddingModel(TextEmbeddingModel):
url = base_url + "/embeddings"
headers = {"Authorization": "Bearer " + api_key, "Content-Type": "application/json"}
data = self._to_payload(model=model, texts=texts, credentials=credentials)
data = self._to_payload(model=model, texts=texts, credentials=credentials, input_type=input_type)
try:
response = post(url, headers=headers, data=dumps(data))

11
api/core/model_runtime/model_providers/localai/text_embedding/text_embedding.py
View File

@ -5,6 +5,7 @@ from typing import Optional
from requests import post
from yarl import URL
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import AIModelEntity, FetchFrom, ModelPropertyKey, ModelType, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
@ -22,11 +23,16 @@ from core.model_runtime.model_providers.__base.text_embedding_model import TextE
class LocalAITextEmbeddingModel(TextEmbeddingModel):
"""
Model class for Jina text embedding model.
Model class for LocalAI text embedding model.
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -35,6 +41,7 @@ class LocalAITextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
if len(texts) != 1:

9
api/core/model_runtime/model_providers/minimax/text_embedding/text_embedding.py
View File

@ -4,6 +4,7 @@ from typing import Optional
from requests import post
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import (
@ -34,7 +35,12 @@ class MinimaxTextEmbeddingModel(TextEmbeddingModel):
api_base: str = "https://api.minimax.chat/v1/embeddings"
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -43,6 +49,7 @@ class MinimaxTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
api_key = credentials["minimax_api_key"]

9
api/core/model_runtime/model_providers/mixedbread/text_embedding/text_embedding.py
View File

@ -4,6 +4,7 @@ from typing import Optional
import requests
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import AIModelEntity, FetchFrom, ModelPropertyKey, ModelType, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
@ -27,7 +28,12 @@ class MixedBreadTextEmbeddingModel(TextEmbeddingModel):
api_base: str = "https://api.mixedbread.ai/v1"
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -36,6 +42,7 @@ class MixedBreadTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
api_key = credentials["api_key"]

3
api/core/model_runtime/model_providers/nomic/text_embedding/text_embedding.py
View File

@ -5,6 +5,7 @@ from typing import Optional
from nomic import embed
from nomic import login as nomic_login
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import (
EmbeddingUsage,
@ -46,6 +47,7 @@ class NomicTextEmbeddingModel(_CommonNomic, TextEmbeddingModel):
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -54,6 +56,7 @@ class NomicTextEmbeddingModel(_CommonNomic, TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
embeddings, prompt_tokens, total_tokens = self.embed_text(

9
api/core/model_runtime/model_providers/nvidia/text_embedding/text_embedding.py
View File

@ -4,6 +4,7 @@ from typing import Optional
from requests import post
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import (
@ -27,7 +28,12 @@ class NvidiaTextEmbeddingModel(TextEmbeddingModel):
models: list[str] = ["NV-Embed-QA"]
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -36,6 +42,7 @@ class NvidiaTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
api_key = credentials["api_key"]

9
api/core/model_runtime/model_providers/oci/text_embedding/text_embedding.py
View File

@ -6,6 +6,7 @@ from typing import Optional
import numpy as np
import oci
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import (
@ -41,7 +42,12 @@ class OCITextEmbeddingModel(TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -50,6 +56,7 @@ class OCITextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
# get model properties

9
api/core/model_runtime/model_providers/ollama/text_embedding/text_embedding.py
View File

@ -8,6 +8,7 @@ from urllib.parse import urljoin
import numpy as np
import requests
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import (
AIModelEntity,
@ -38,7 +39,12 @@ class OllamaEmbeddingModel(TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -47,6 +53,7 @@ class OllamaEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""

9
api/core/model_runtime/model_providers/openai/text_embedding/text_embedding.py
View File

@ -6,6 +6,7 @@ import numpy as np
import tiktoken
from openai import OpenAI
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.validate import CredentialsValidateFailedError
@ -19,7 +20,12 @@ class OpenAITextEmbeddingModel(_CommonOpenAI, TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -28,6 +34,7 @@ class OpenAITextEmbeddingModel(_CommonOpenAI, TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
# transform credentials to kwargs for model instance

9
api/core/model_runtime/model_providers/openai_api_compatible/text_embedding/text_embedding.py
View File

@ -7,6 +7,7 @@ from urllib.parse import urljoin
import numpy as np
import requests
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import (
AIModelEntity,
@ -28,7 +29,12 @@ class OAICompatEmbeddingModel(_CommonOaiApiCompat, TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -37,6 +43,7 @@ class OAICompatEmbeddingModel(_CommonOaiApiCompat, TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""

9
api/core/model_runtime/model_providers/openllm/text_embedding/text_embedding.py
View File

@ -5,6 +5,7 @@ from typing import Optional
from requests import post
from requests.exceptions import ConnectionError, InvalidSchema, MissingSchema
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import (
@ -25,7 +26,12 @@ class OpenLLMTextEmbeddingModel(TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -34,6 +40,7 @@ class OpenLLMTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
server_url = credentials["server_url"]

9
api/core/model_runtime/model_providers/perfxcloud/text_embedding/text_embedding.py
View File

@ -7,6 +7,7 @@ from urllib.parse import urljoin
import numpy as np
import requests
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import (
AIModelEntity,
@ -28,7 +29,12 @@ class OAICompatEmbeddingModel(_CommonOaiApiCompat, TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -37,6 +43,7 @@ class OAICompatEmbeddingModel(_CommonOaiApiCompat, TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""

18
api/core/model_runtime/model_providers/replicate/text_embedding/text_embedding.py
View File

@ -4,6 +4,7 @@ from typing import Optional
from replicate import Client as ReplicateClient
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import AIModelEntity, FetchFrom, ModelType, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
@ -14,8 +15,23 @@ from core.model_runtime.model_providers.replicate._common import _CommonReplicat
class ReplicateEmbeddingModel(_CommonReplicate, TextEmbeddingModel):
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
client = ReplicateClient(api_token=credentials["replicate_api_token"], timeout=30)
if "model_version" in credentials:

16
api/core/model_runtime/model_providers/sagemaker/llm/llm.py
View File

@ -84,8 +84,9 @@ class SageMakerLargeLanguageModel(LargeLanguageModel):
Model class for Cohere large language model.
"""
sagemaker_client: Any = None
sagemaker_session: Any = None
predictor: Any = None
sagemaker_endpoint: str = None
def _handle_chat_generate_response(
self,
@ -211,7 +212,7 @@ class SageMakerLargeLanguageModel(LargeLanguageModel):
:param user: unique user id
:return: full response or stream response chunk generator result
"""
if not self.sagemaker_client:
if not self.sagemaker_session:
access_key = credentials.get("aws_access_key_id")
secret_key = credentials.get("aws_secret_access_key")
aws_region = credentials.get("aws_region")
@ -226,11 +227,14 @@ class SageMakerLargeLanguageModel(LargeLanguageModel):
else:
boto_session = boto3.Session()
self.sagemaker_client = boto_session.client("sagemaker")
sagemaker_session = Session(boto_session=boto_session, sagemaker_client=self.sagemaker_client)
sagemaker_client = boto_session.client("sagemaker")
self.sagemaker_session = Session(boto_session=boto_session, sagemaker_client=sagemaker_client)
if self.sagemaker_endpoint != credentials.get("sagemaker_endpoint"):
self.sagemaker_endpoint = credentials.get("sagemaker_endpoint")
self.predictor = Predictor(
endpoint_name=credentials.get("sagemaker_endpoint"),
sagemaker_session=sagemaker_session,
endpoint_name=self.sagemaker_endpoint,
sagemaker_session=self.sagemaker_session,
serializer=serializers.JSONSerializer(),
)

9
api/core/model_runtime/model_providers/sagemaker/text_embedding/text_embedding.py
View File

@ -6,6 +6,7 @@ from typing import Any, Optional
import boto3
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import AIModelEntity, FetchFrom, ModelPropertyKey, ModelType, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
@ -53,7 +54,12 @@ class SageMakerEmbeddingModel(TextEmbeddingModel):
return embeddings
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -62,6 +68,7 @@ class SageMakerEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
# get model properties

39
api/core/model_runtime/model_providers/siliconflow/llm/_position.yaml
View File

@ -1,25 +1,38 @@
- Qwen/Qwen2.5-7B-Instruct
- Qwen/Qwen2.5-14B-Instruct
- Qwen/Qwen2.5-32B-Instruct
- Qwen/Qwen2.5-72B-Instruct
- Qwen/Qwen2.5-Math-72B-Instruct
- Qwen/Qwen2.5-32B-Instruct
- Qwen/Qwen2.5-14B-Instruct
- Qwen/Qwen2.5-7B-Instruct
- Qwen/Qwen2.5-Coder-7B-Instruct
- deepseek-ai/DeepSeek-V2.5
- Qwen/Qwen2-72B-Instruct
- Qwen/Qwen2-57B-A14B-Instruct
- Qwen/Qwen2-7B-Instruct
- Qwen/Qwen2-1.5B-Instruct
- 01-ai/Yi-1.5-34B-Chat
- 01-ai/Yi-1.5-9B-Chat-16K
- 01-ai/Yi-1.5-6B-Chat
- THUDM/glm-4-9b-chat
- deepseek-ai/DeepSeek-V2.5
- deepseek-ai/DeepSeek-V2-Chat
- deepseek-ai/DeepSeek-Coder-V2-Instruct
- THUDM/glm-4-9b-chat
- THUDM/chatglm3-6b
- 01-ai/Yi-1.5-34B-Chat-16K
- 01-ai/Yi-1.5-9B-Chat-16K
- 01-ai/Yi-1.5-6B-Chat
- internlm/internlm2_5-20b-chat
- internlm/internlm2_5-7b-chat
- google/gemma-2-27b-it
- google/gemma-2-9b-it
- meta-llama/Meta-Llama-3-70B-Instruct
- meta-llama/Meta-Llama-3-8B-Instruct
- meta-llama/Meta-Llama-3.1-405B-Instruct
- meta-llama/Meta-Llama-3.1-70B-Instruct
- meta-llama/Meta-Llama-3.1-8B-Instruct
- mistralai/Mixtral-8x7B-Instruct-v0.1
- meta-llama/Meta-Llama-3-70B-Instruct
- meta-llama/Meta-Llama-3-8B-Instruct
- google/gemma-2-27b-it
- google/gemma-2-9b-it
- mistralai/Mistral-7B-Instruct-v0.2
- Pro/Qwen/Qwen2-7B-Instruct
- Pro/Qwen/Qwen2-1.5B-Instruct
- Pro/THUDM/glm-4-9b-chat
- Pro/THUDM/chatglm3-6b
- Pro/01-ai/Yi-1.5-9B-Chat-16K
- Pro/01-ai/Yi-1.5-6B-Chat
- Pro/internlm/internlm2_5-7b-chat
- Pro/meta-llama/Meta-Llama-3.1-8B-Instruct
- Pro/meta-llama/Meta-Llama-3-8B-Instruct
- Pro/google/gemma-2-9b-it

18
api/core/model_runtime/model_providers/siliconflow/text_embedding/text_embedding.py
View File

@ -1,5 +1,6 @@
from typing import Optional
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.text_embedding_entities import TextEmbeddingResult
from core.model_runtime.model_providers.openai_api_compatible.text_embedding.text_embedding import (
OAICompatEmbeddingModel,
@ -16,8 +17,23 @@ class SiliconflowTextEmbeddingModel(OAICompatEmbeddingModel):
super().validate_credentials(model, credentials)
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
self._add_custom_parameters(credentials)
return super()._invoke(model, credentials, texts, user)

9
api/core/model_runtime/model_providers/spark/llm/llm.py
View File

@ -213,18 +213,21 @@ class SparkLargeLanguageModel(LargeLanguageModel):
:param prompt_messages: prompt messages
:return: llm response chunk generator result
"""
completion = ""
for index, content in enumerate(client.subscribe()):
if isinstance(content, dict):
delta = content["data"]
else:
delta = content
completion += delta
assistant_prompt_message = AssistantPromptMessage(
content=delta or "",
)
temp_assistant_prompt_message = AssistantPromptMessage(
content=completion,
)
prompt_tokens = self.get_num_tokens(model, credentials, prompt_messages)
completion_tokens = self.get_num_tokens(model, credentials, [assistant_prompt_message])
completion_tokens = self.get_num_tokens(model, credentials, [temp_assistant_prompt_message])
# transform usage
usage = self._calc_response_usage(model, credentials, prompt_tokens, completion_tokens)

3
api/core/model_runtime/model_providers/tongyi/text_embedding/text_embedding.py
View File

@ -4,6 +4,7 @@ from typing import Optional
import dashscope
import numpy as np
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import (
EmbeddingUsage,
@ -27,6 +28,7 @@ class TongyiTextEmbeddingModel(_CommonTongyi, TextEmbeddingModel):
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -35,6 +37,7 @@ class TongyiTextEmbeddingModel(_CommonTongyi, TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
credentials_kwargs = self._to_credential_kwargs(credentials)

11
api/core/model_runtime/model_providers/upstage/text_embedding/text_embedding.py
View File

@ -7,6 +7,7 @@ import numpy as np
from openai import OpenAI
from tokenizers import Tokenizer
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.validate import CredentialsValidateFailedError
@ -22,7 +23,14 @@ class UpstageTextEmbeddingModel(_CommonUpstage, TextEmbeddingModel):
def _get_tokenizer(self) -> Tokenizer:
return Tokenizer.from_pretrained("upstage/solar-1-mini-tokenizer")
def _invoke(self, model: str, credentials: dict, texts: list[str], user: str | None = None) -> TextEmbeddingResult:
def _invoke(
self,
model: str,
credentials: dict,
texts: list[str],
user: str | None = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -30,6 +38,7 @@ class UpstageTextEmbeddingModel(_CommonUpstage, TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""

2
api/core/model_runtime/model_providers/vertex_ai/llm/gemini-1.5-flash.yaml → api/core/model_runtime/model_providers/vertex_ai/llm/gemini-1.5-flash-001.yaml
View File

@ -1,6 +1,6 @@
model: gemini-1.5-flash-001
label:
en_US: Gemini 1.5 Flash
en_US: Gemini 1.5 Flash 001
model_type: llm
features:
- agent-thought

37
api/core/model_runtime/model_providers/vertex_ai/llm/gemini-1.5-flash-002.yaml Normal file
View File

@ -0,0 +1,37 @@
model: gemini-1.5-flash-002
label:
en_US: Gemini 1.5 Flash 002
model_type: llm
features:
- agent-thought
- vision
model_properties:
mode: chat
context_size: 1048576
parameter_rules:
- name: temperature
use_template: temperature
- name: top_p
use_template: top_p
- name: top_k
label:
en_US: Top k
type: int
help:
en_US: Only sample from the top K options for each subsequent token.
required: false
- name: presence_penalty
use_template: presence_penalty
- name: frequency_penalty
use_template: frequency_penalty
- name: max_output_tokens
use_template: max_tokens
required: true
default: 8192
min: 1
max: 8192
pricing:
input: '0.00'
output: '0.00'
unit: '0.000001'
currency: USD

2
api/core/model_runtime/model_providers/vertex_ai/llm/gemini-1.5-pro.yaml → api/core/model_runtime/model_providers/vertex_ai/llm/gemini-1.5-pro-001.yaml
View File

@ -1,6 +1,6 @@
model: gemini-1.5-pro-001
label:
en_US: Gemini 1.5 Pro
en_US: Gemini 1.5 Pro 001
model_type: llm
features:
- agent-thought

37
api/core/model_runtime/model_providers/vertex_ai/llm/gemini-1.5-pro-002.yaml Normal file
View File

@ -0,0 +1,37 @@
model: gemini-1.5-pro-002
label:
en_US: Gemini 1.5 Pro 002
model_type: llm
features:
- agent-thought
- vision
model_properties:
mode: chat
context_size: 1048576
parameter_rules:
- name: temperature
use_template: temperature
- name: top_p
use_template: top_p
- name: top_k
label:
en_US: Top k
type: int
help:
en_US: Only sample from the top K options for each subsequent token.
required: false
- name: presence_penalty
use_template: presence_penalty
- name: frequency_penalty
use_template: frequency_penalty
- name: max_output_tokens
use_template: max_tokens
required: true
default: 8192
min: 1
max: 8192
pricing:
input: '0.00'
output: '0.00'
unit: '0.000001'
currency: USD

37
api/core/model_runtime/model_providers/vertex_ai/llm/gemini-flash-experimental.yaml Normal file
View File

@ -0,0 +1,37 @@
model: gemini-flash-experimental
label:
en_US: Gemini Flash Experimental
model_type: llm
features:
- agent-thought
- vision
model_properties:
mode: chat
context_size: 1048576
parameter_rules:
- name: temperature
use_template: temperature
- name: top_p
use_template: top_p
- name: top_k
label:
en_US: Top k
type: int
help:
en_US: Only sample from the top K options for each subsequent token.
required: false
- name: presence_penalty
use_template: presence_penalty
- name: frequency_penalty
use_template: frequency_penalty
- name: max_output_tokens
use_template: max_tokens
required: true
default: 8192
min: 1
max: 8192
pricing:
input: '0.00'
output: '0.00'
unit: '0.000001'
currency: USD

37
api/core/model_runtime/model_providers/vertex_ai/llm/gemini-pro-experimental.yaml Normal file
View File

@ -0,0 +1,37 @@
model: gemini-pro-experimental
label:
en_US: Gemini Pro Experimental
model_type: llm
features:
- agent-thought
- vision
model_properties:
mode: chat
context_size: 1048576
parameter_rules:
- name: temperature
use_template: temperature
- name: top_p
use_template: top_p
- name: top_k
label:
en_US: Top k
type: int
help:
en_US: Only sample from the top K options for each subsequent token.
required: false
- name: presence_penalty
use_template: presence_penalty
- name: frequency_penalty
use_template: frequency_penalty
- name: max_output_tokens
use_template: max_tokens
required: true
default: 8192
min: 1
max: 8192
pricing:
input: '0.00'
output: '0.00'
unit: '0.000001'
currency: USD

11
api/core/model_runtime/model_providers/vertex_ai/llm/llm.py
View File

@ -2,6 +2,7 @@ import base64
import io
import json
import logging
import time
from collections.abc import Generator
from typing import Optional, Union, cast
@ -20,7 +21,6 @@ from google.api_core import exceptions
from google.cloud import aiplatform
from google.oauth2 import service_account
from PIL import Image
from vertexai.generative_models import HarmBlockThreshold, HarmCategory
from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunk, LLMResultChunkDelta, LLMUsage
from core.model_runtime.entities.message_entities import (
@ -34,6 +34,7 @@ from core.model_runtime.entities.message_entities import (
ToolPromptMessage,
UserPromptMessage,
)
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.errors.invoke import (
InvokeAuthorizationError,
InvokeBadRequestError,
@ -503,20 +504,12 @@ class VertexAiLargeLanguageModel(LargeLanguageModel):
else:
history.append(content)
safety_settings = {
HarmCategory.HARM_CATEGORY_HARASSMENT: HarmBlockThreshold.BLOCK_NONE,
HarmCategory.HARM_CATEGORY_HATE_SPEECH: HarmBlockThreshold.BLOCK_NONE,
HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT: HarmBlockThreshold.BLOCK_NONE,
HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT: HarmBlockThreshold.BLOCK_NONE,
}
google_model = glm.GenerativeModel(model_name=model, system_instruction=system_instruction)
response = google_model.generate_content(
contents=history,
generation_config=glm.GenerationConfig(**config_kwargs),
stream=stream,
safety_settings=safety_settings,
tools=self._convert_tools_to_glm_tool(tools) if tools else None,
)

10
api/core/model_runtime/model_providers/vertex_ai/text_embedding/text_embedding.py
View File

@ -9,6 +9,7 @@ from google.cloud import aiplatform
from google.oauth2 import service_account
from vertexai.language_models import TextEmbeddingModel as VertexTextEmbeddingModel
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import (
AIModelEntity,
@ -30,7 +31,12 @@ class VertexAiTextEmbeddingModel(_CommonVertexAi, TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -38,6 +44,8 @@ class VertexAiTextEmbeddingModel(_CommonVertexAi, TextEmbeddingModel):
:param model: model name
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
service_account_info = json.loads(base64.b64decode(credentials["vertex_service_account_key"]))

9
api/core/model_runtime/model_providers/volcengine_maas/text_embedding/text_embedding.py
View File

@ -2,6 +2,7 @@ import time
from decimal import Decimal
from typing import Optional
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import (
AIModelEntity,
@ -41,7 +42,12 @@ class VolcengineMaaSTextEmbeddingModel(TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -50,6 +56,7 @@ class VolcengineMaaSTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
if ArkClientV3.is_legacy(credentials):

9
api/core/model_runtime/model_providers/wenxin/text_embedding/text_embedding.py
View File

@ -7,6 +7,7 @@ from typing import Any, Optional
import numpy as np
from requests import Response, post
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.invoke import InvokeError
@ -70,7 +71,12 @@ class WenxinTextEmbeddingModel(TextEmbeddingModel):
return WenxinTextEmbedding(api_key, secret_key)
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -79,6 +85,7 @@ class WenxinTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""

9
api/core/model_runtime/model_providers/xinference/text_embedding/text_embedding.py
View File

@ -3,6 +3,7 @@ from typing import Optional
from xinference_client.client.restful.restful_client import Client, RESTfulEmbeddingModelHandle
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.model_entities import AIModelEntity, FetchFrom, ModelPropertyKey, ModelType, PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
@ -25,7 +26,12 @@ class XinferenceTextEmbeddingModel(TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -40,6 +46,7 @@ class XinferenceTextEmbeddingModel(TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
server_url = credentials["server_url"]

9
api/core/model_runtime/model_providers/zhipuai/text_embedding/text_embedding.py
View File

@ -1,6 +1,7 @@
import time
from typing import Optional
from core.embedding.embedding_constant import EmbeddingInputType
from core.model_runtime.entities.model_entities import PriceType
from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
from core.model_runtime.errors.validate import CredentialsValidateFailedError
@ -15,7 +16,12 @@ class ZhipuAITextEmbeddingModel(_CommonZhipuaiAI, TextEmbeddingModel):
"""
def _invoke(
self, model: str, credentials: dict, texts: list[str], user: Optional[str] = None
self,
model: str,
credentials: dict,
texts: list[str],
user: Optional[str] = None,
input_type: EmbeddingInputType = EmbeddingInputType.DOCUMENT,
) -> TextEmbeddingResult:
"""
Invoke text embedding model
@ -24,6 +30,7 @@ class ZhipuAITextEmbeddingModel(_CommonZhipuaiAI, TextEmbeddingModel):
:param credentials: model credentials
:param texts: texts to embed
:param user: unique user id
:param input_type: input type
:return: embeddings result
"""
credentials_kwargs = self._to_credential_kwargs(credentials)

4
api/core/rag/extractor/extract_processor.py
View File

@ -124,7 +124,7 @@ class ExtractProcessor:
extractor = UnstructuredPPTXExtractor(file_path, unstructured_api_url)
elif file_extension == ".xml":
extractor = UnstructuredXmlExtractor(file_path, unstructured_api_url)
elif file_extension == "epub":
elif file_extension == ".epub":
extractor = UnstructuredEpubExtractor(file_path, unstructured_api_url)
else:
# txt
@ -146,7 +146,7 @@ class ExtractProcessor:
extractor = WordExtractor(file_path, upload_file.tenant_id, upload_file.created_by)
elif file_extension == ".csv":
extractor = CSVExtractor(file_path, autodetect_encoding=True)
elif file_extension == "epub":
elif file_extension == ".epub":
extractor = UnstructuredEpubExtractor(file_path)
else:
# txt

3
api/core/tools/__base/tool_provider.py
View File

@ -64,6 +64,9 @@ class ToolProviderController(ABC):
# check type
credential_schema = credentials_need_to_validate[credential_name]
if not credential_schema.required and credentials[credential_name] is None:
continue
if credential_schema.type in {ProviderConfig.Type.SECRET_INPUT, ProviderConfig.Type.TEXT_INPUT}:
if not isinstance(credentials[credential_name], str):
raise ToolProviderCredentialValidationError(f"credential {credential_name} should be string")