doc clearn up

adalflow/adalflow/eval/g_eval.py

@@ -258,5 +258,5 @@ def compute(
         return final_output, output
 
     def __str__(self) -> str:
-        s = f"llm_judge={self.llm_evaluator}, prompt_kwargs={self.llm_judge.prompt_kwargs}"
+        s = f"llm_judge={self.llm_judge}, prompt_kwargs={self.llm_judge.prompt_kwargs}"
         return s

adalflow/adalflow/optim/text_grad/tgd_optimizer.py

@@ -25,58 +25,6 @@
 
 log = logging.getLogger(__name__)
 
-GLOSSARY_TEXT = r"""
-### Glossary of tags that will be sent to you:
-{# # - <LM_SYSTEM_PROMPT>: The system prompt for the language model.
-# - <LM_INPUT>: The input to the language model.
-# - <LM_OUTPUT>: The output of the language model. #}
-# - <FEEDBACK>: The feedback to the variable.
-# - <CONVERSATION>: The conversation history.
-# - <FOCUS>: The focus of the optimization.
-# - <ROLE>: The role description of the variable."""
-
-# customize the system prompt
-# prompts, solutions to problems, code, or any other text-based variable. -> to the variable type.
-# The optimizer will have an understanding of different variable types.
-
-# Tips:
-# - DO NOT address concerns on the peer variables. The peer variables will be optimized separately.
-# - The instruction needs to be on point, clear, and accurate.
-# You can take the following actions:
-# - You can delete words or phrases that you think are not necessary or you find misleading.
-# - You can add new words or phrases that you think can address the feedback.
-# - You can add sections like "Tips" or "Remember" to fix the issue.
-# - You can be creative and write the variable in a different way.#}
-
-# Think step by step. But keep the reasoning concise!!!
-# Start response with "because" and "since" to provide reasoning.
-# The last line of your response will be the new variable value put in between {{new_variable_start_tag}} and {{new_variable_end_tag}} tags.
-
-# {#You are part of an optimization system that improves the exsiting variable value according to feedback.
-
-# You will be asked to creatively and critically at proposing new variable value.
-# The feedback may be noisy, focus on what is important.
-# {# output format #}
-# You MUST ONLY output the new variable value in the response between {{new_variable_start_tag}} and {{new_variable_end_tag}} tags.
-# {% if instruction_to_optimizer %}
-# USER INSTRUCTION: {{instruction_to_optimizer}}
-# {% endif %}
-# Remember:
-# - Pay attention to role description, and most importantly, the feedback.
-# - Must provide a different value that addresses the feedback.
-# - If <HISTORY_PERFORMANCE> is provided, observe the patterns of the past iterations, pay close attention to high-performing value and aim to improve the variable value.
-# - The new value better not overlap with the scope of peer variables.
-# - Make your best judgment and be creative.
-# TIPS:
-# - Delete words or phrases that you think are not necessary or you find misleading.
-# - Add new words or phrases that you think can address the feedback.
-# - Add sections like "Tips" or "Remember" to fix the issue.
-# - Be eative and write the variable in a different way.#}
-
-# Your task: Propose a new variable value in response to the feedback.
-# 1. Focus on what is essential, even if feedback is noisy.
-# 2. Observe past performance patterns when provided and aim to improve upon high-performing values.
-# 3. Avoid overlap with peer variables' scope.
 
 # Tips:
 # 1. Eliminate unnecessary words or phrases.

adalflow/adalflow/optim/trainer/adal.py

@@ -26,11 +26,12 @@
 
 
 class AdalComponent(Component):
-    """Define a train, eval, and test step for a task pipeline.
+    __doc__ = """Define a train, eval, and test step for a task pipeline.
 
     This serves the following purposes:
-    1. Organize all parts for training a task pipeline organized in one place.
+    1. Organize all parts for training a task pipeline in one place.
     2. Help with debugging and testing before the actual training.
+    3. Adds multi-threading support for training and evaluation.
     """
 
     task: Component

adalflow/adalflow/optim/trainer/trainer.py

@@ -35,7 +35,7 @@
 
 
 class Trainer(Component):
-    r"""We make trainer a component to as a trainer itself is an LLM task pipeline too.
+    __doc__ = r"""Ready to use trainer for LLM task pipeline to optimize all types of parameters.
 
 
     Training set: can be used for passing initial proposed prompt or for few-shot sampling.
@@ -339,7 +339,6 @@ def fit(
                 )
             self.adaltask.configure_teacher_generator()
             print("Configured demo optimizers")
-            # return
         else:
             print("No trainable demo params to optimize")
             self.demo_optimizers = []