dify/api/tests/integration_tests/model_runtime/huggingface_hub/test_llm.py

import os
from collections.abc import Generator

import pytest

from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunk, LLMResultChunkDelta
from core.model_runtime.entities.message_entities import AssistantPromptMessage, UserPromptMessage
from core.model_runtime.errors.validate import CredentialsValidateFailedError
from core.model_runtime.model_providers.huggingface_hub.llm.llm import HuggingfaceHubLargeLanguageModel
from tests.integration_tests.model_runtime.__mock.huggingface import setup_huggingface_mock


@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_hosted_inference_api_validate_credentials(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    with pytest.raises(CredentialsValidateFailedError):
        model.validate_credentials(
            model='HuggingFaceH4/zephyr-7b-beta',
            credentials={
                'huggingfacehub_api_type': 'hosted_inference_api',
                'huggingfacehub_api_token': 'invalid_key'
            }
        )

    with pytest.raises(CredentialsValidateFailedError):
        model.validate_credentials(
            model='fake-model',
            credentials={
                'huggingfacehub_api_type': 'hosted_inference_api',
                'huggingfacehub_api_token': 'invalid_key'
            }
        )

    model.validate_credentials(
        model='HuggingFaceH4/zephyr-7b-beta',
        credentials={
            'huggingfacehub_api_type': 'hosted_inference_api',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY')
        }
    )

@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_hosted_inference_api_invoke_model(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    response = model.invoke(
        model='HuggingFaceH4/zephyr-7b-beta',
        credentials={
            'huggingfacehub_api_type': 'hosted_inference_api',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY')
        },
        prompt_messages=[
            UserPromptMessage(
                content='Who are you?'
            )
        ],
        model_parameters={
            'temperature': 1.0,
            'top_k': 2,
            'top_p': 0.5,
        },
        stop=['How'],
        stream=False,
        user="abc-123"
    )

    assert isinstance(response, LLMResult)
    assert len(response.message.content) > 0

@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_hosted_inference_api_invoke_stream_model(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    response = model.invoke(
        model='HuggingFaceH4/zephyr-7b-beta',
        credentials={
            'huggingfacehub_api_type': 'hosted_inference_api',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY')
        },
        prompt_messages=[
            UserPromptMessage(
                content='Who are you?'
            )
        ],
        model_parameters={
            'temperature': 1.0,
            'top_k': 2,
            'top_p': 0.5,
        },
        stop=['How'],
        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

@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_inference_endpoints_text_generation_validate_credentials(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    with pytest.raises(CredentialsValidateFailedError):
        model.validate_credentials(
            model='openchat/openchat_3.5',
            credentials={
                'huggingfacehub_api_type': 'inference_endpoints',
                'huggingfacehub_api_token': 'invalid_key',
                'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT_GEN_ENDPOINT_URL'),
                'task_type': 'text-generation'
            }
        )

    model.validate_credentials(
        model='openchat/openchat_3.5',
        credentials={
            'huggingfacehub_api_type': 'inference_endpoints',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY'),
            'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT_GEN_ENDPOINT_URL'),
            'task_type': 'text-generation'
        }
    )

@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_inference_endpoints_text_generation_invoke_model(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    response = model.invoke(
        model='openchat/openchat_3.5',
        credentials={
            'huggingfacehub_api_type': 'inference_endpoints',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY'),
            'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT_GEN_ENDPOINT_URL'),
            'task_type': 'text-generation'
        },
        prompt_messages=[
            UserPromptMessage(
                content='Who are you?'
            )
        ],
        model_parameters={
            'temperature': 1.0,
            'top_k': 2,
            'top_p': 0.5,
        },
        stop=['How'],
        stream=False,
        user="abc-123"
    )

    assert isinstance(response, LLMResult)
    assert len(response.message.content) > 0

@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_inference_endpoints_text_generation_invoke_stream_model(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    response = model.invoke(
        model='openchat/openchat_3.5',
        credentials={
            'huggingfacehub_api_type': 'inference_endpoints',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY'),
            'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT_GEN_ENDPOINT_URL'),
            'task_type': 'text-generation'
        },
        prompt_messages=[
            UserPromptMessage(
                content='Who are you?'
            )
        ],
        model_parameters={
            'temperature': 1.0,
            'top_k': 2,
            'top_p': 0.5,
        },
        stop=['How'],
        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

@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_inference_endpoints_text2text_generation_validate_credentials(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    with pytest.raises(CredentialsValidateFailedError):
        model.validate_credentials(
            model='google/mt5-base',
            credentials={
                'huggingfacehub_api_type': 'inference_endpoints',
                'huggingfacehub_api_token': 'invalid_key',
                'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT2TEXT_GEN_ENDPOINT_URL'),
                'task_type': 'text2text-generation'
            }
        )

    model.validate_credentials(
        model='google/mt5-base',
        credentials={
            'huggingfacehub_api_type': 'inference_endpoints',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY'),
            'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT2TEXT_GEN_ENDPOINT_URL'),
            'task_type': 'text2text-generation'
        }
    )

@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_inference_endpoints_text2text_generation_invoke_model(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    response = model.invoke(
        model='google/mt5-base',
        credentials={
            'huggingfacehub_api_type': 'inference_endpoints',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY'),
            'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT2TEXT_GEN_ENDPOINT_URL'),
            'task_type': 'text2text-generation'
        },
        prompt_messages=[
            UserPromptMessage(
                content='Who are you?'
            )
        ],
        model_parameters={
            'temperature': 1.0,
            'top_k': 2,
            'top_p': 0.5,
        },
        stop=['How'],
        stream=False,
        user="abc-123"
    )

    assert isinstance(response, LLMResult)
    assert len(response.message.content) > 0

@pytest.mark.parametrize('setup_huggingface_mock', [['none']], indirect=True)
def test_inference_endpoints_text2text_generation_invoke_stream_model(setup_huggingface_mock):
    model = HuggingfaceHubLargeLanguageModel()

    response = model.invoke(
        model='google/mt5-base',
        credentials={
            'huggingfacehub_api_type': 'inference_endpoints',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY'),
            'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT2TEXT_GEN_ENDPOINT_URL'),
            'task_type': 'text2text-generation'
        },
        prompt_messages=[
            UserPromptMessage(
                content='Who are you?'
            )
        ],
        model_parameters={
            'temperature': 1.0,
            'top_k': 2,
            'top_p': 0.5,
        },
        stop=['How'],
        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_get_num_tokens():
    model = HuggingfaceHubLargeLanguageModel()

    num_tokens = model.get_num_tokens(
        model='google/mt5-base',
        credentials={
            'huggingfacehub_api_type': 'inference_endpoints',
            'huggingfacehub_api_token': os.environ.get('HUGGINGFACE_API_KEY'),
            'huggingfacehub_endpoint_url': os.environ.get('HUGGINGFACE_TEXT2TEXT_GEN_ENDPOINT_URL'),
            'task_type': 'text2text-generation'
        },
        prompt_messages=[
            UserPromptMessage(
                content='Hello World!'
            )
        ]
    )

    assert num_tokens == 7