AI/dify
1
0
Fork 0
mirror of https://git.mirrors.martin98.com/https://github.com/langgenius/dify.git synced 2025-08-13 22:25:57 +08:00
Code Issues Packages Projects Releases Wiki Activity

add together ai model setting (#3895)

Browse Source
This commit is contained in:
Jyong 2024-04-26 20:43:17 +08:00 committed by GitHub
parent 045827043d
commit 0ec8b57825
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 108 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

112
api/core/model_runtime/model_providers/togetherai/llm/llm.py
View File

@ -1,9 +1,23 @@
from collections.abc import Generator
from decimal import Decimal
from typing import Optional, Union
from core.model_runtime.entities.llm_entities import LLMResult
from core.model_runtime.entities.message_entities import PromptMessage, PromptMessageTool
from core.model_runtime.entities.model_entities import AIModelEntity
from core.model_runtime.entities.common_entities import I18nObject
from core.model_runtime.entities.llm_entities import LLMMode, LLMResult
from core.model_runtime.entities.message_entities import (
PromptMessage,
PromptMessageTool,
)
from core.model_runtime.entities.model_entities import (
AIModelEntity,
DefaultParameterName,
FetchFrom,
ModelPropertyKey,
ModelType,
ParameterRule,
ParameterType,
PriceConfig,
)
from core.model_runtime.model_providers.openai_api_compatible.llm.llm import OAIAPICompatLargeLanguageModel
@ -36,8 +50,98 @@ class TogetherAILargeLanguageModel(OAIAPICompatLargeLanguageModel):
def get_customizable_model_schema(self, model: str, credentials: dict) -> AIModelEntity:
cred_with_endpoint = self._update_endpoint_url(credentials=credentials)
REPETITION_PENALTY = "repetition_penalty"
TOP_K = "top_k"
features = []
return super().get_customizable_model_schema(model, cred_with_endpoint)
entity = AIModelEntity(
model=model,
label=I18nObject(en_US=model),
model_type=ModelType.LLM,
fetch_from=FetchFrom.CUSTOMIZABLE_MODEL,
features=features,
model_properties={
ModelPropertyKey.CONTEXT_SIZE: int(cred_with_endpoint.get('context_size', "4096")),
ModelPropertyKey.MODE: cred_with_endpoint.get('mode'),
},
parameter_rules=[
ParameterRule(
name=DefaultParameterName.TEMPERATURE.value,
label=I18nObject(en_US="Temperature"),
type=ParameterType.FLOAT,
default=float(cred_with_endpoint.get('temperature', 0.7)),
min=0,
max=2,
precision=2
),
ParameterRule(
name=DefaultParameterName.TOP_P.value,
label=I18nObject(en_US="Top P"),
type=ParameterType.FLOAT,
default=float(cred_with_endpoint.get('top_p', 1)),
min=0,
max=1,
precision=2
),
ParameterRule(
name=TOP_K,
label=I18nObject(en_US="Top K"),
type=ParameterType.INT,
default=int(cred_with_endpoint.get('top_k', 50)),
min=-2147483647,
max=2147483647,
precision=0
),
ParameterRule(
name=REPETITION_PENALTY,
label=I18nObject(en_US="Repetition Penalty"),
type=ParameterType.FLOAT,
default=float(cred_with_endpoint.get('repetition_penalty', 1)),
min=-3.4,
max=3.4,
precision=1
),
ParameterRule(
name=DefaultParameterName.MAX_TOKENS.value,
label=I18nObject(en_US="Max Tokens"),
type=ParameterType.INT,
default=512,
min=1,
max=int(cred_with_endpoint.get('max_tokens_to_sample', 4096)),
),
ParameterRule(
name=DefaultParameterName.FREQUENCY_PENALTY.value,
label=I18nObject(en_US="Frequency Penalty"),
type=ParameterType.FLOAT,
default=float(credentials.get('frequency_penalty', 0)),
min=-2,
max=2
),
ParameterRule(
name=DefaultParameterName.PRESENCE_PENALTY.value,
label=I18nObject(en_US="Presence Penalty"),
type=ParameterType.FLOAT,
default=float(credentials.get('presence_penalty', 0)),
min=-2,
max=2
)
],
pricing=PriceConfig(
input=Decimal(cred_with_endpoint.get('input_price', 0)),
output=Decimal(cred_with_endpoint.get('output_price', 0)),
unit=Decimal(cred_with_endpoint.get('unit', 0)),
currency=cred_with_endpoint.get('currency', "USD")
),
)
if cred_with_endpoint['mode'] == 'chat':
entity.model_properties[ModelPropertyKey.MODE] = LLMMode.CHAT.value
elif cred_with_endpoint['mode'] == 'completion':
entity.model_properties[ModelPropertyKey.MODE] = LLMMode.COMPLETION.value
else:
raise ValueError(f"Unknown completion type {cred_with_endpoint['completion_type']}")
return entity
def get_num_tokens(self, model: str, credentials: dict, prompt_messages: list[PromptMessage],
tools: Optional[list[PromptMessageTool]] = None) -> int: