File size: 6,906 Bytes
20f348c |
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 |
import os
from collections.abc import Generator
from time import sleep
import pytest
from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunk, LLMResultChunkDelta
from core.model_runtime.entities.message_entities import AssistantPromptMessage, SystemPromptMessage, UserPromptMessage
from core.model_runtime.entities.model_entities import AIModelEntity
from core.model_runtime.errors.validate import CredentialsValidateFailedError
from core.model_runtime.model_providers.wenxin.llm.llm import ErnieBotLargeLanguageModel
def test_predefined_models():
model = ErnieBotLargeLanguageModel()
model_schemas = model.predefined_models()
assert len(model_schemas) >= 1
assert isinstance(model_schemas[0], AIModelEntity)
def test_validate_credentials_for_chat_model():
sleep(3)
model = ErnieBotLargeLanguageModel()
with pytest.raises(CredentialsValidateFailedError):
model.validate_credentials(
model="ernie-bot", credentials={"api_key": "invalid_key", "secret_key": "invalid_key"}
)
model.validate_credentials(
model="ernie-bot",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
)
def test_invoke_model_ernie_bot():
sleep(3)
model = ErnieBotLargeLanguageModel()
response = model.invoke(
model="ernie-bot",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
prompt_messages=[UserPromptMessage(content="Hello World!")],
model_parameters={
"temperature": 0.7,
"top_p": 1.0,
},
stop=["you"],
user="abc-123",
stream=False,
)
assert isinstance(response, LLMResult)
assert len(response.message.content) > 0
assert response.usage.total_tokens > 0
def test_invoke_model_ernie_bot_turbo():
sleep(3)
model = ErnieBotLargeLanguageModel()
response = model.invoke(
model="ernie-bot-turbo",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
prompt_messages=[UserPromptMessage(content="Hello World!")],
model_parameters={
"temperature": 0.7,
"top_p": 1.0,
},
stop=["you"],
user="abc-123",
stream=False,
)
assert isinstance(response, LLMResult)
assert len(response.message.content) > 0
assert response.usage.total_tokens > 0
def test_invoke_model_ernie_8k():
sleep(3)
model = ErnieBotLargeLanguageModel()
response = model.invoke(
model="ernie-bot-8k",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
prompt_messages=[UserPromptMessage(content="Hello World!")],
model_parameters={
"temperature": 0.7,
"top_p": 1.0,
},
stop=["you"],
user="abc-123",
stream=False,
)
assert isinstance(response, LLMResult)
assert len(response.message.content) > 0
assert response.usage.total_tokens > 0
def test_invoke_model_ernie_bot_4():
sleep(3)
model = ErnieBotLargeLanguageModel()
response = model.invoke(
model="ernie-bot-4",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
prompt_messages=[UserPromptMessage(content="Hello World!")],
model_parameters={
"temperature": 0.7,
"top_p": 1.0,
},
stop=["you"],
user="abc-123",
stream=False,
)
assert isinstance(response, LLMResult)
assert len(response.message.content) > 0
assert response.usage.total_tokens > 0
def test_invoke_stream_model():
sleep(3)
model = ErnieBotLargeLanguageModel()
response = model.invoke(
model="ernie-3.5-8k",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
prompt_messages=[UserPromptMessage(content="Hello World!")],
model_parameters={
"temperature": 0.7,
"top_p": 1.0,
},
stop=["you"],
stream=True,
user="abc-123",
)
assert isinstance(response, Generator)
for chunk in response:
assert isinstance(chunk, LLMResultChunk)
assert isinstance(chunk.delta, LLMResultChunkDelta)
assert isinstance(chunk.delta.message, AssistantPromptMessage)
assert len(chunk.delta.message.content) > 0 if chunk.delta.finish_reason is None else True
def test_invoke_model_with_system():
sleep(3)
model = ErnieBotLargeLanguageModel()
response = model.invoke(
model="ernie-bot",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
prompt_messages=[SystemPromptMessage(content="你是Kasumi"), UserPromptMessage(content="你是谁？")],
model_parameters={
"temperature": 0.7,
"top_p": 1.0,
},
stop=["you"],
stream=False,
user="abc-123",
)
assert isinstance(response, LLMResult)
assert "kasumi" in response.message.content.lower()
def test_invoke_with_search():
sleep(3)
model = ErnieBotLargeLanguageModel()
response = model.invoke(
model="ernie-bot",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
prompt_messages=[UserPromptMessage(content="北京今天的天气怎么样")],
model_parameters={
"temperature": 0.7,
"top_p": 1.0,
"disable_search": True,
},
stop=[],
stream=True,
user="abc-123",
)
assert isinstance(response, Generator)
total_message = ""
for chunk in response:
assert isinstance(chunk, LLMResultChunk)
assert isinstance(chunk.delta, LLMResultChunkDelta)
assert isinstance(chunk.delta.message, AssistantPromptMessage)
total_message += chunk.delta.message.content
print(chunk.delta.message.content)
assert len(chunk.delta.message.content) > 0 if not chunk.delta.finish_reason else True
# there should be 对不起、我不能、不支持……
assert "不" in total_message or "抱歉" in total_message or "无法" in total_message
def test_get_num_tokens():
sleep(3)
model = ErnieBotLargeLanguageModel()
response = model.get_num_tokens(
model="ernie-bot",
credentials={"api_key": os.environ.get("WENXIN_API_KEY"), "secret_key": os.environ.get("WENXIN_SECRET_KEY")},
prompt_messages=[UserPromptMessage(content="Hello World!")],
tools=[],
)
assert isinstance(response, int)
assert response == 10
|