feat: add baichuan llm support (#1294)

Co-authored-by: zxhlyh <jasonapring2015@outlook.com>
This commit is contained in:
takatost
2023-10-10 12:09:26 +08:00
committed by GitHub
parent 677aacc8e3
commit 1d4f019de4
9 changed files with 745 additions and 1 deletions

View File

@@ -35,6 +35,10 @@ WENXIN_SECRET_KEY=
# ZhipuAI Credentials
ZHIPUAI_API_KEY=
# Baichuan Credentials
BAICHUAN_API_KEY=
BAICHUAN_SECRET_KEY=
# ChatGLM Credentials
CHATGLM_API_BASE=

View File

@@ -0,0 +1,81 @@
import json
import os
from unittest.mock import patch
from core.model_providers.models.entity.message import PromptMessage, MessageType
from core.model_providers.models.entity.model_params import ModelKwargs
from core.model_providers.models.llm.baichuan_model import BaichuanModel
from core.model_providers.providers.baichuan_provider import BaichuanProvider
from models.provider import Provider, ProviderType
def get_mock_provider(valid_api_key, valid_secret_key):
return Provider(
id='provider_id',
tenant_id='tenant_id',
provider_name='baichuan',
provider_type=ProviderType.CUSTOM.value,
encrypted_config=json.dumps({
'api_key': valid_api_key,
'secret_key': valid_secret_key,
}),
is_valid=True,
)
def get_mock_model(model_name: str, streaming: bool = False):
model_kwargs = ModelKwargs(
temperature=0.01,
)
valid_api_key = os.environ['BAICHUAN_API_KEY']
valid_secret_key = os.environ['BAICHUAN_SECRET_KEY']
model_provider = BaichuanProvider(provider=get_mock_provider(valid_api_key, valid_secret_key))
return BaichuanModel(
model_provider=model_provider,
name=model_name,
model_kwargs=model_kwargs,
streaming=streaming
)
def decrypt_side_effect(tenant_id, encrypted_api_key):
return encrypted_api_key
@patch('core.helper.encrypter.decrypt_token', side_effect=decrypt_side_effect)
def test_chat_get_num_tokens(mock_decrypt):
model = get_mock_model('baichuan2-53b')
rst = model.get_num_tokens([
PromptMessage(type=MessageType.SYSTEM, content='you are a kindness Assistant.'),
PromptMessage(type=MessageType.HUMAN, content='Who is your manufacturer?')
])
assert rst > 0
@patch('core.helper.encrypter.decrypt_token', side_effect=decrypt_side_effect)
def test_chat_run(mock_decrypt, mocker):
mocker.patch('core.model_providers.providers.base.BaseModelProvider.update_last_used', return_value=None)
model = get_mock_model('baichuan2-53b')
messages = [
PromptMessage(type=MessageType.HUMAN, content='Are you Human? you MUST only answer `y` or `n`?')
]
rst = model.run(
messages,
)
assert len(rst.content) > 0
@patch('core.helper.encrypter.decrypt_token', side_effect=decrypt_side_effect)
def test_chat_stream_run(mock_decrypt, mocker):
mocker.patch('core.model_providers.providers.base.BaseModelProvider.update_last_used', return_value=None)
model = get_mock_model('baichuan2-53b', streaming=True)
messages = [
PromptMessage(type=MessageType.HUMAN, content='Are you Human? you MUST only answer `y` or `n`?')
]
rst = model.run(
messages
)
assert len(rst.content) > 0

View File

@@ -0,0 +1,97 @@
import pytest
from unittest.mock import patch
import json
from langchain.schema import ChatResult, ChatGeneration, AIMessage
from core.model_providers.providers.baichuan_provider import BaichuanProvider
from core.model_providers.providers.base import CredentialsValidateFailedError
from models.provider import ProviderType, Provider
PROVIDER_NAME = 'baichuan'
MODEL_PROVIDER_CLASS = BaichuanProvider
VALIDATE_CREDENTIAL = {
'api_key': 'valid_key',
'secret_key': 'valid_key',
}
def encrypt_side_effect(tenant_id, encrypt_key):
return f'encrypted_{encrypt_key}'
def decrypt_side_effect(tenant_id, encrypted_key):
return encrypted_key.replace('encrypted_', '')
def test_is_provider_credentials_valid_or_raise_valid(mocker):
mocker.patch('core.third_party.langchain.llms.baichuan_llm.BaichuanChatLLM._generate',
return_value=ChatResult(generations=[ChatGeneration(message=AIMessage(content='abc'))]))
MODEL_PROVIDER_CLASS.is_provider_credentials_valid_or_raise(VALIDATE_CREDENTIAL)
def test_is_provider_credentials_valid_or_raise_invalid():
# raise CredentialsValidateFailedError if api_key is not in credentials
with pytest.raises(CredentialsValidateFailedError):
MODEL_PROVIDER_CLASS.is_provider_credentials_valid_or_raise({})
credential = VALIDATE_CREDENTIAL.copy()
credential['api_key'] = 'invalid_key'
credential['secret_key'] = 'invalid_key'
# raise CredentialsValidateFailedError if api_key is invalid
with pytest.raises(CredentialsValidateFailedError):
MODEL_PROVIDER_CLASS.is_provider_credentials_valid_or_raise(credential)
@patch('core.helper.encrypter.encrypt_token', side_effect=encrypt_side_effect)
def test_encrypt_credentials(mock_encrypt):
result = MODEL_PROVIDER_CLASS.encrypt_provider_credentials('tenant_id', VALIDATE_CREDENTIAL.copy())
assert result['api_key'] == f'encrypted_{VALIDATE_CREDENTIAL["api_key"]}'
assert result['secret_key'] == f'encrypted_{VALIDATE_CREDENTIAL["secret_key"]}'
@patch('core.helper.encrypter.decrypt_token', side_effect=decrypt_side_effect)
def test_get_credentials_custom(mock_decrypt):
encrypted_credential = VALIDATE_CREDENTIAL.copy()
encrypted_credential['api_key'] = 'encrypted_' + encrypted_credential['api_key']
encrypted_credential['secret_key'] = 'encrypted_' + encrypted_credential['secret_key']
provider = Provider(
id='provider_id',
tenant_id='tenant_id',
provider_name=PROVIDER_NAME,
provider_type=ProviderType.CUSTOM.value,
encrypted_config=json.dumps(encrypted_credential),
is_valid=True,
)
model_provider = MODEL_PROVIDER_CLASS(provider=provider)
result = model_provider.get_provider_credentials()
assert result['api_key'] == 'valid_key'
assert result['secret_key'] == 'valid_key'
@patch('core.helper.encrypter.decrypt_token', side_effect=decrypt_side_effect)
def test_get_credentials_obfuscated(mock_decrypt):
encrypted_credential = VALIDATE_CREDENTIAL.copy()
encrypted_credential['api_key'] = 'encrypted_' + encrypted_credential['api_key']
encrypted_credential['secret_key'] = 'encrypted_' + encrypted_credential['secret_key']
provider = Provider(
id='provider_id',
tenant_id='tenant_id',
provider_name=PROVIDER_NAME,
provider_type=ProviderType.CUSTOM.value,
encrypted_config=json.dumps(encrypted_credential),
is_valid=True,
)
model_provider = MODEL_PROVIDER_CLASS(provider=provider)
result = model_provider.get_provider_credentials(obfuscated=True)
middle_token = result['api_key'][6:-2]
secret_key_middle_token = result['secret_key'][6:-2]
assert len(middle_token) == max(len(VALIDATE_CREDENTIAL['api_key']) - 8, 0)
assert len(secret_key_middle_token) == max(len(VALIDATE_CREDENTIAL['secret_key']) - 8, 0)
assert all(char == '*' for char in middle_token)
assert all(char == '*' for char in secret_key_middle_token)