Merge pull request #157 from SylphAI-Inc/main

[release 0.1.0.b6]

docs/source/tutorials/component.rst

@@ -1,3 +1,16 @@
+.. raw:: html
+
+   <div style="display: flex; justify-content: flex-start; align-items: center; margin-bottom: 20px;">
+      <a href="https://colab.research.google.com/drive/1aD0C8-iMB8quIn8FKhrtFAGcrboRNg2C?usp=sharing" target="_blank" style="margin-right: 10px;">
+         <img alt="Try Quickstart in Colab" src="https://colab.research.google.com/assets/colab-badge.svg" style="vertical-align: middle;">
+      </a>
+      <a href="https://github.com/SylphAI-Inc/LightRAG/blob/main/lightrag/lightrag/core/component.py" target="_blank" style="display: flex; align-items: center;">
+         <img src="https://github.githubassets.com/images/modules/logos_page/GitHub-Mark.png" alt="GitHub" style="height: 20px; width: 20px; margin-right: 5px;">
+         <span style="vertical-align: middle;"> Open Source Code</span>
+      </a>
+   </div>
+
+
 Component
 ============
 

docs/source/tutorials/generator.rst

@@ -1,5 +1,18 @@
 .. _generator:
 
+.. raw:: html
+
+   <div style="display: flex; justify-content: flex-start; align-items: center; margin-bottom: 20px;">
+      <a href="https://colab.research.google.com/drive/1gmxeX1UuUxZDouWhkLGQYrD4hAdt9IVX?usp=sharing" target="_blank" style="margin-right: 10px;">
+         <img alt="Try Quickstart in Colab" src="https://colab.research.google.com/assets/colab-badge.svg" style="vertical-align: middle;">
+      </a>
+      <a href="https://github.com/SylphAI-Inc/AdalFlow/blob/main/tutorials/generator_note.py" target="_blank" style="display: flex; align-items: center;">
+         <img src="https://github.githubassets.com/images/modules/logos_page/GitHub-Mark.png" alt="GitHub" style="height: 20px; width: 20px; margin-right: 5px;">
+         <span style="vertical-align: middle;"> Open Source Code</span>
+      </a>
+   </div>
+
+
 Generator
 =========
 
@@ -276,7 +289,7 @@ We will use the `JsonParser` to parse the output back to a `dict` object.
 
     output_format_str = r"""Your output should be formatted as a standard JSON object with two keys:
     {
-        "explaination": "A brief explaination of the concept in one sentence.",
+        "explanation": "A brief explanation of the concept in one sentence.",
         "example": "An example of the concept in a sentence."
     }
     """
@@ -304,7 +317,7 @@ The final prompt is:
     <OUTPUT_FORMAT>
     Your output should be formatted as a standard JSON object with two keys:
         {
-            "explaination": "A brief explaination of the concept in one sentence.",
+            "explanation": "A brief explanation of the concept in one sentence.",
             "example": "An example of the concept in a sentence."
         }
 
@@ -320,7 +333,7 @@ The above printout is:
 .. code-block::
 
     <class 'dict'>
-    {'explaination': 'LLM stands for Large Language Model, which are deep learning models trained on enormous amounts of text data.', 'example': 'An example of a LLM is GPT-3, which can generate human-like text based on the input provided.'}
+    {'explanation': 'LLM stands for Large Language Model, which are deep learning models trained on enormous amounts of text data.', 'example': 'An example of a LLM is GPT-3, which can generate human-like text based on the input provided.'}
 
 Please refer to :doc:`output_parsers` for a more comprehensive guide on the `Parser` components.
 

lightrag/lightrag/components/model_client/ollama_client.py

@@ -1,7 +1,16 @@
 """Ollama ModelClient integration."""
 
 import os
-from typing import Dict, Optional, Any, TypeVar, List, Type, Generator, Union
+from typing import (
+    Dict,
+    Optional,
+    Any,
+    TypeVar,
+    List,
+    Type,
+    Generator as GeneratorType,
+    Union,
+)
 import backoff
 import logging
 import warnings
@@ -21,6 +30,27 @@
 T = TypeVar("T")
 
 
+def parse_stream_response(completion: GeneratorType) -> Any:
+    """Parse the completion to a str. We use the generate with prompt instead of chat with messages."""
+    for chunk in completion:
+        log.debug(f"Raw chunk: {chunk}")
+        yield chunk["response"] if "response" in chunk else None
+
+
+def parse_generate_response(completion: GenerateResponse) -> Any:
+    """Parse the completion to a str. We use the generate with prompt instead of chat with messages."""
+    if "response" in completion:
+        log.debug(f"response: {completion}")
+        return completion["response"]
+    else:
+        log.error(
+            f"Error parsing the completion: {completion}, type: {type(completion)}"
+        )
+        raise ValueError(
+            f"Error parsing the completion: {completion}, type: {type(completion)}"
+        )
+
+
 class OllamaClient(ModelClient):
     __doc__ = r"""A component wrapper for the Ollama SDK client.
 
@@ -146,25 +176,16 @@ def init_async_client(self):
 
         self.async_client = ollama.AsyncClient(host=self._host)
 
+    # NOTE: do not put yield and return in the same function, thus we separate the functions
     def parse_chat_completion(
-        self, completion: Union[GenerateResponse, Generator]
+        self, completion: Union[GenerateResponse, GeneratorType]
     ) -> Any:
         """Parse the completion to a str. We use the generate with prompt instead of chat with messages."""
-        log.debug(f"completion: {completion}, {isinstance(completion, Generator)}")
-        if isinstance(completion, Generator):  # streaming
-            for chunk in completion:
-                log.debug(f"Raw chunk: {chunk}")
-                yield chunk["response"] if "response" in chunk else None
+        log.debug(f"completion: {completion}, {isinstance(completion, GeneratorType)}")
+        if isinstance(completion, GeneratorType):  # streaming
+            return parse_stream_response(completion)
         else:
-            if "response" in completion:
-                return completion["response"]
-            else:
-                log.error(
-                    f"Error parsing the completion: {completion}, type: {type(completion)}"
-                )
-                raise ValueError(
-                    f"Error parsing the completion: {completion}, type: {type(completion)}"
-                )
+            return parse_generate_response(completion)
 
     def parse_embedding_response(
         self, response: Dict[str, List[float]]
@@ -264,18 +285,34 @@ def to_dict(self, exclude: Optional[List[str]] = None) -> Dict[str, Any]:
         return output
 
 
+# TODO: add tests to stream and non stream case
 # if __name__ == "__main__":
 #     from lightrag.core.generator import Generator
-#     from lightrag.components.model_client import OllamaClient
+#     from lightrag.components.model_client import OllamaClient, OpenAIClient
 #     from lightrag.utils import setup_env, get_logger
 
-#     # log = get_logger(level="DEBUG")
+#     log = get_logger(level="DEBUG")
 
 #     setup_env()
 
-#     model_client = OllamaClient()
-#     model_kwargs = {"model": "phi3", "stream": True}
-#     generator = Generator(model_client=model_client, model_kwargs=model_kwargs)
-#     output = generator({"input_str": "What is the capital of France?"})
-#     for chunk in output.data:
-#         print(chunk)
+#     ollama_ai = {
+#         "model_client": OllamaClient(),
+#         "model_kwargs": {
+#             "model": "qwen2:0.5b",
+#             "stream": True,
+#         },
+#     }
+#     open_ai = {
+#         "model_client": OpenAIClient(),
+#         "model_kwargs": {
+#             "model": "gpt-3.5-turbo",
+#             "stream": False,
+#         },
+#     }
+#     # generator = Generator(**open_ai)
+#     # output = generator({"input_str": "What is the capital of France?"})
+#     # print(output)
+
+#     # generator = Generator(**ollama_ai)
+#     # output = generator({"input_str": "What is the capital of France?"})
+#     # print(output)

lightrag/lightrag/core/generator.py

@@ -168,7 +168,6 @@ def _post_call(self, completion: Any) -> GeneratorOutputType:
             response = self.model_client.parse_chat_completion(completion)
         except Exception as e:
             log.error(f"Error parsing the completion {completion}: {e}")
-            # response = str(completion)
             return GeneratorOutput(raw_response=str(completion), error=str(e))
 
         # the output processors operate on the str, the raw_response field.
@@ -227,20 +226,22 @@ def call(
         log.debug(f"api_kwargs: {api_kwargs}")
         output: GeneratorOutputType = None
         # call the model client
+        completion = None
         try:
             completion = self.model_client.call(
                 api_kwargs=api_kwargs, model_type=self.model_type
             )
-
         except Exception as e:
             log.error(f"Error calling the model: {e}")
             output = GeneratorOutput(error=str(e))
+        # process the completion
+        if completion:
+            try:
+                output = self._post_call(completion)
 
-        try:
-            output = self._post_call(completion)
-        except Exception as e:
-            log.error(f"Error processing the output: {e}")
-            output = GeneratorOutput(raw_response=str(completion), error=str(e))
+            except Exception as e:
+                log.error(f"Error processing the output: {e}")
+                output = GeneratorOutput(raw_response=str(completion), error=str(e))
 
         log.info(f"output: {output}")
         return output

lightrag/lightrag/core/types.py

@@ -68,17 +68,17 @@ class ModelClientType:
 
     .. code-block:: python
 
-    from lightrag.components.model_client import CohereAPIClient, TransformersClient, AnthropicAPIClient, GroqAPIClient, OpenAIClient
+        from lightrag.components.model_client import CohereAPIClient, TransformersClient, AnthropicAPIClient, GroqAPIClient, OpenAIClient
 
-    model_client = OpenAIClient()
+        model_client = OpenAIClient()
 
     To this:
 
     .. code-block:: python
 
-    from lightrag.core.types import ModelClientType
+        from lightrag.core.types import ModelClientType
 
-    model_client = ModelClientType.OPENAI()
+        model_client = ModelClientType.OPENAI()
     """
     COHERE = CohereAPIClient
     TRANSFORMERS = TransformersClient

tutorials/generator_note.py

@@ -34,8 +34,8 @@ async def acall(self, query):
 
 @dataclass
 class QAOutput(DataClass):
-    explaination: str = field(
-        metadata={"desc": "A brief explaination of the concept in one sentence."}
+    explanation: str = field(
+        metadata={"desc": "A brief explanation of the concept in one sentence."}
     )
     example: str = field(metadata={"desc": "An example of the concept in a sentence."})