forked from langgenius/dify
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathhosting_configuration.py
266 lines (214 loc) · 10.1 KB
/
hosting_configuration.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
from typing import Optional
from flask import Flask
from pydantic import BaseModel
from configs import dify_config
from core.entities import DEFAULT_PLUGIN_ID
from core.entities.provider_entities import ProviderQuotaType, QuotaUnit, RestrictModel
from core.model_runtime.entities.model_entities import ModelType
class HostingQuota(BaseModel):
quota_type: ProviderQuotaType
restrict_models: list[RestrictModel] = []
class TrialHostingQuota(HostingQuota):
quota_type: ProviderQuotaType = ProviderQuotaType.TRIAL
quota_limit: int = 0
"""Quota limit for the hosting provider models. -1 means unlimited."""
class PaidHostingQuota(HostingQuota):
quota_type: ProviderQuotaType = ProviderQuotaType.PAID
class FreeHostingQuota(HostingQuota):
quota_type: ProviderQuotaType = ProviderQuotaType.FREE
class HostingProvider(BaseModel):
enabled: bool = False
credentials: Optional[dict] = None
quota_unit: Optional[QuotaUnit] = None
quotas: list[HostingQuota] = []
class HostedModerationConfig(BaseModel):
enabled: bool = False
providers: list[str] = []
class HostingConfiguration:
provider_map: dict[str, HostingProvider]
moderation_config: Optional[HostedModerationConfig] = None
def __init__(self) -> None:
self.provider_map = {}
self.moderation_config = None
def init_app(self, app: Flask) -> None:
if dify_config.EDITION != "CLOUD":
return
self.provider_map[f"{DEFAULT_PLUGIN_ID}/azure_openai/azure_openai"] = self.init_azure_openai()
self.provider_map[f"{DEFAULT_PLUGIN_ID}/openai/openai"] = self.init_openai()
self.provider_map[f"{DEFAULT_PLUGIN_ID}/anthropic/anthropic"] = self.init_anthropic()
self.provider_map[f"{DEFAULT_PLUGIN_ID}/minimax/minimax"] = self.init_minimax()
self.provider_map[f"{DEFAULT_PLUGIN_ID}/spark/spark"] = self.init_spark()
self.provider_map[f"{DEFAULT_PLUGIN_ID}/zhipuai/zhipuai"] = self.init_zhipuai()
self.moderation_config = self.init_moderation_config()
@staticmethod
def init_azure_openai() -> HostingProvider:
quota_unit = QuotaUnit.TIMES
if dify_config.HOSTED_AZURE_OPENAI_ENABLED:
credentials = {
"openai_api_key": dify_config.HOSTED_AZURE_OPENAI_API_KEY,
"openai_api_base": dify_config.HOSTED_AZURE_OPENAI_API_BASE,
"base_model_name": "gpt-35-turbo",
}
quotas: list[HostingQuota] = []
hosted_quota_limit = dify_config.HOSTED_AZURE_OPENAI_QUOTA_LIMIT
trial_quota = TrialHostingQuota(
quota_limit=hosted_quota_limit,
restrict_models=[
RestrictModel(model="gpt-4", base_model_name="gpt-4", model_type=ModelType.LLM),
RestrictModel(model="gpt-4o", base_model_name="gpt-4o", model_type=ModelType.LLM),
RestrictModel(model="gpt-4o-mini", base_model_name="gpt-4o-mini", model_type=ModelType.LLM),
RestrictModel(model="gpt-4-32k", base_model_name="gpt-4-32k", model_type=ModelType.LLM),
RestrictModel(
model="gpt-4-1106-preview", base_model_name="gpt-4-1106-preview", model_type=ModelType.LLM
),
RestrictModel(
model="gpt-4-vision-preview", base_model_name="gpt-4-vision-preview", model_type=ModelType.LLM
),
RestrictModel(model="gpt-35-turbo", base_model_name="gpt-35-turbo", model_type=ModelType.LLM),
RestrictModel(
model="gpt-35-turbo-1106", base_model_name="gpt-35-turbo-1106", model_type=ModelType.LLM
),
RestrictModel(
model="gpt-35-turbo-instruct", base_model_name="gpt-35-turbo-instruct", model_type=ModelType.LLM
),
RestrictModel(
model="gpt-35-turbo-16k", base_model_name="gpt-35-turbo-16k", model_type=ModelType.LLM
),
RestrictModel(
model="text-davinci-003", base_model_name="text-davinci-003", model_type=ModelType.LLM
),
RestrictModel(
model="text-embedding-ada-002",
base_model_name="text-embedding-ada-002",
model_type=ModelType.TEXT_EMBEDDING,
),
RestrictModel(
model="text-embedding-3-small",
base_model_name="text-embedding-3-small",
model_type=ModelType.TEXT_EMBEDDING,
),
RestrictModel(
model="text-embedding-3-large",
base_model_name="text-embedding-3-large",
model_type=ModelType.TEXT_EMBEDDING,
),
],
)
quotas.append(trial_quota)
return HostingProvider(enabled=True, credentials=credentials, quota_unit=quota_unit, quotas=quotas)
return HostingProvider(
enabled=False,
quota_unit=quota_unit,
)
def init_openai(self) -> HostingProvider:
quota_unit = QuotaUnit.CREDITS
quotas: list[HostingQuota] = []
if dify_config.HOSTED_OPENAI_TRIAL_ENABLED:
hosted_quota_limit = dify_config.HOSTED_OPENAI_QUOTA_LIMIT
trial_models = self.parse_restrict_models_from_env("HOSTED_OPENAI_TRIAL_MODELS")
trial_quota = TrialHostingQuota(quota_limit=hosted_quota_limit, restrict_models=trial_models)
quotas.append(trial_quota)
if dify_config.HOSTED_OPENAI_PAID_ENABLED:
paid_models = self.parse_restrict_models_from_env("HOSTED_OPENAI_PAID_MODELS")
paid_quota = PaidHostingQuota(restrict_models=paid_models)
quotas.append(paid_quota)
if len(quotas) > 0:
credentials = {
"openai_api_key": dify_config.HOSTED_OPENAI_API_KEY,
}
if dify_config.HOSTED_OPENAI_API_BASE:
credentials["openai_api_base"] = dify_config.HOSTED_OPENAI_API_BASE
if dify_config.HOSTED_OPENAI_API_ORGANIZATION:
credentials["openai_organization"] = dify_config.HOSTED_OPENAI_API_ORGANIZATION
return HostingProvider(enabled=True, credentials=credentials, quota_unit=quota_unit, quotas=quotas)
return HostingProvider(
enabled=False,
quota_unit=quota_unit,
)
@staticmethod
def init_anthropic() -> HostingProvider:
quota_unit = QuotaUnit.TOKENS
quotas: list[HostingQuota] = []
if dify_config.HOSTED_ANTHROPIC_TRIAL_ENABLED:
hosted_quota_limit = dify_config.HOSTED_ANTHROPIC_QUOTA_LIMIT
trial_quota = TrialHostingQuota(quota_limit=hosted_quota_limit)
quotas.append(trial_quota)
if dify_config.HOSTED_ANTHROPIC_PAID_ENABLED:
paid_quota = PaidHostingQuota()
quotas.append(paid_quota)
if len(quotas) > 0:
credentials = {
"anthropic_api_key": dify_config.HOSTED_ANTHROPIC_API_KEY,
}
if dify_config.HOSTED_ANTHROPIC_API_BASE:
credentials["anthropic_api_url"] = dify_config.HOSTED_ANTHROPIC_API_BASE
return HostingProvider(enabled=True, credentials=credentials, quota_unit=quota_unit, quotas=quotas)
return HostingProvider(
enabled=False,
quota_unit=quota_unit,
)
@staticmethod
def init_minimax() -> HostingProvider:
quota_unit = QuotaUnit.TOKENS
if dify_config.HOSTED_MINIMAX_ENABLED:
quotas: list[HostingQuota] = [FreeHostingQuota()]
return HostingProvider(
enabled=True,
credentials=None, # use credentials from the provider
quota_unit=quota_unit,
quotas=quotas,
)
return HostingProvider(
enabled=False,
quota_unit=quota_unit,
)
@staticmethod
def init_spark() -> HostingProvider:
quota_unit = QuotaUnit.TOKENS
if dify_config.HOSTED_SPARK_ENABLED:
quotas: list[HostingQuota] = [FreeHostingQuota()]
return HostingProvider(
enabled=True,
credentials=None, # use credentials from the provider
quota_unit=quota_unit,
quotas=quotas,
)
return HostingProvider(
enabled=False,
quota_unit=quota_unit,
)
@staticmethod
def init_zhipuai() -> HostingProvider:
quota_unit = QuotaUnit.TOKENS
if dify_config.HOSTED_ZHIPUAI_ENABLED:
quotas: list[HostingQuota] = [FreeHostingQuota()]
return HostingProvider(
enabled=True,
credentials=None, # use credentials from the provider
quota_unit=quota_unit,
quotas=quotas,
)
return HostingProvider(
enabled=False,
quota_unit=quota_unit,
)
@staticmethod
def init_moderation_config() -> HostedModerationConfig:
if dify_config.HOSTED_MODERATION_ENABLED and dify_config.HOSTED_MODERATION_PROVIDERS:
providers = dify_config.HOSTED_MODERATION_PROVIDERS.split(",")
hosted_providers = []
for provider in providers:
if "/" not in provider:
provider = f"{DEFAULT_PLUGIN_ID}/{provider}/{provider}"
hosted_providers.append(provider)
return HostedModerationConfig(enabled=True, providers=hosted_providers)
return HostedModerationConfig(enabled=False)
@staticmethod
def parse_restrict_models_from_env(env_var: str) -> list[RestrictModel]:
models_str = dify_config.model_dump().get(env_var)
models_list = models_str.split(",") if models_str else []
return [
RestrictModel(model=model_name.strip(), model_type=ModelType.LLM)
for model_name in models_list
if model_name.strip()
]