Model Runtime (#1858)

Co-authored-by: StyleZhang <jasonapring2015@outlook.com>
Co-authored-by: Garfield Dai <dai.hai@foxmail.com>
Co-authored-by: chenhe <guchenhe@gmail.com>
Co-authored-by: jyong <jyong@dify.ai>
Co-authored-by: Joel <iamjoel007@gmail.com>
Co-authored-by: Yeuoly <admin@srmxy.cn>

api/tests/integration_tests/model_runtime/openllm/test_llm.py

@@ -0,0 +1,104 @@
+import os
+import pytest
+
+from typing import Generator
+
+from core.model_runtime.entities.message_entities import AssistantPromptMessage, UserPromptMessage
+from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunkDelta, \
+    LLMResultChunk
+from core.model_runtime.errors.validate import CredentialsValidateFailedError
+from core.model_runtime.model_providers.openllm.llm.llm import OpenLLMLargeLanguageModel
+
+def test_validate_credentials_for_chat_model():
+    model = OpenLLMLargeLanguageModel()
+
+    with pytest.raises(CredentialsValidateFailedError):
+        model.validate_credentials(
+            model='NOT IMPORTANT',
+            credentials={
+                'server_url': 'invalid_key',
+            }
+        )
+
+    model.validate_credentials(
+        model='NOT IMPORTANT',
+        credentials={
+            'server_url': os.environ.get('OPENLLM_SERVER_URL'),
+        }
+    )
+
+def test_invoke_model():
+    model = OpenLLMLargeLanguageModel()
+
+    response = model.invoke(
+        model='NOT IMPORTANT',
+        credentials={
+            'server_url': os.environ.get('OPENLLM_SERVER_URL'),
+        },
+        prompt_messages=[
+            UserPromptMessage(
+                content='Hello World!'
+            )
+        ],
+        model_parameters={
+            'temperature': 0.7,
+            'top_p': 1.0,
+            'top_k': 1,
+        },
+        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():
+    model = OpenLLMLargeLanguageModel()
+
+    response = model.invoke(
+        model='NOT IMPORTANT',
+        credentials={
+            'server_url': os.environ.get('OPENLLM_SERVER_URL'),
+        },
+        prompt_messages=[
+            UserPromptMessage(
+                content='Hello World!'
+            )
+        ],
+        model_parameters={
+            'temperature': 0.7,
+            'top_p': 1.0,
+            'top_k': 1,
+        },
+        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_get_num_tokens():
+    model = OpenLLMLargeLanguageModel()
+
+    response = model.get_num_tokens(
+        model='NOT IMPORTANT',
+        credentials={
+            'server_url': os.environ.get('OPENLLM_SERVER_URL'),
+        },
+        prompt_messages=[
+            UserPromptMessage(
+                content='Hello World!'
+            )
+        ],
+        tools=[]
+    )
+
+    assert isinstance(response, int)
+    assert response == 3