o1 support

examples/hello_world.py

@@ -10,9 +10,9 @@ def get_random_adjective():
 def get_random_punctuation():
     return random.choice(["!", "!!", "!!!"])
 
-@ell.simple(model="gpt-3.5-turbo")
+@ell.simple(model="o1-preview")
 def hello(name: str):
-    """You are a helpful and expressive assistant."""
+    # """You are a helpful and expressive assistant."""
     adjective = get_random_adjective()
     punctuation = get_random_punctuation()
     return f"Say a {adjective} hello to {name}{punctuation}"

examples/o1.py

@@ -0,0 +1,36 @@
+import ell
+
+@ell.simple(model="o1-preview")
+def solve_complex_math_problem(equation: str, variables: dict, constraints: list, optimization_goal: str):
+    return f"""You are an expert mathematician and problem solver. Please solve the following complex mathematical problem:
+
+Equation: {equation}
+Variables: {variables}
+Constraints: {constraints}
+Optimization Goal: {optimization_goal}"""
+
+@ell.simple(model="o1-preview")
+def write_plot_code_for_problem_and_solution(solution :str):
+    return f"""You are an expert programmer and problem solver. 
+Please write code in python with matplotlib to plot the solution to the following problem: It should work in the terminal. Full script with imports.
+IMPORTANT: Do not include any other text only the code.
+Solution to plot: {solution}"""
+
+def solve_and_plot(**kwargs):
+    solution = solve_complex_math_problem(**kwargs)
+    plot_code = write_plot_code_for_problem_and_solution(solution)
+    # remove backticks and ```python
+    plot_code = plot_code.replace("```python", "").replace("```", "").strip()
+    exec(plot_code)
+    return solution
+
+if __name__ == "__main__":
+
+    ell.init(store='./logdir', autocommit=True, verbose=True)
+    result = solve_and_plot(
+        equation="y = ax^2 + bx + c",
+        variables={"a": 1, "b": -5, "c": 6},
+        constraints=["x >= 0", "x <= 10"],
+        optimization_goal="Find the minimum value of y within the given constraints"
+    )
+    print(result)

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "ell-ai"
-version = "0.0.4"
+version = "0.0.5"
 description = "ell - the language model programming library"
 authors = ["William Guss <will@lrsys.xyz>"]
 license = "MIT"

src/ell/configurator.py

@@ -19,7 +19,6 @@ class Config(BaseModel):
     autocommit: bool = Field(default=False, description="If True, enables automatic committing of changes to the store.")
     lazy_versioning: bool = Field(default=True, description="If True, enables lazy versioning for improved performance.")
     default_lm_params: Dict[str, Any] = Field(default_factory=dict, description="Default parameters for language models.")
-    default_system_prompt: str = Field(default="You are a helpful AI assistant.", description="The default system prompt used for AI interactions.")
     default_client: Optional[openai.Client] = Field(default=None, description="The default OpenAI client used when a specific model client is not found.")
     providers: Dict[Type, Type[Provider]] = Field(default_factory=dict, description="A dictionary mapping client types to provider classes.")
 
@@ -138,14 +137,7 @@ class Config(BaseModel):
         """
         self.default_lm_params = params
     
-    def set_default_system_prompt(self, prompt: str) -> None:
-        """
-        Set the default system prompt.
 
-        :param prompt: The default system prompt to set.
-        :type prompt: str
-        """
-        self.default_system_prompt = prompt
 
     def set_default_client(self, client: openai.Client) -> None:
         """
@@ -186,7 +178,6 @@ def init(
     autocommit: bool = True,
     lazy_versioning: bool = True,
     default_lm_params: Optional[Dict[str, Any]] = None,
-    default_system_prompt: Optional[str] = None,
     default_openai_client: Optional[openai.Client] = None
 ) -> None:
     """
@@ -202,8 +193,6 @@ def init(
     :type lazy_versioning: bool
     :param default_lm_params: Set default parameters for language models.
     :type default_lm_params: Dict[str, Any], optional
-    :param default_system_prompt: Set the default system prompt.
-    :type default_system_prompt: str, optional
     :param default_openai_client: Set the default OpenAI client.
     :type default_openai_client: openai.Client, optional
     """
@@ -216,8 +205,7 @@ def init(
     if default_lm_params is not None:
         config.set_default_lm_params(**default_lm_params)
 
-    if default_system_prompt is not None:
-        config.set_default_system_prompt(default_system_prompt)
+
 
     if default_openai_client is not None:
         config.set_default_client(default_openai_client)
@@ -235,10 +223,6 @@ def set_store(*args, **kwargs) -> None:
 def set_default_lm_params(*args, **kwargs) -> None:
     return config.set_default_lm_params(*args, **kwargs)
 
-@wraps(config.set_default_system_prompt)
-def set_default_system_prompt(*args, **kwargs) -> None:
-    return config.set_default_system_prompt(*args, **kwargs)
-
 # You can add more helper functions here if needed
 @wraps(config.register_provider)
 def register_provider(*args, **kwargs) -> None:

src/ell/lmp/complex.py

@@ -266,9 +266,10 @@ def _get_messages(prompt_ret: Union[str, list[MessageOrDict]], prompt: LMP) -> l
     Helper function to convert the output of an LMP into a list of Messages.
     """
     if isinstance(prompt_ret, str):
-        return [
-            Message(role="system", content=[ContentBlock(text=_lstr(prompt.__doc__) or config.default_system_prompt)]),
-            Message(role="user", content=[ContentBlock(text=prompt_ret)]),
+        has_system_prompt = prompt.__doc__ is not None and prompt.__doc__.strip() != ""
+        messages =     [Message(role="system", content=[ContentBlock(text=_lstr(prompt.__doc__) )])] if has_system_prompt else []
+        return messages + [
+            Message(role="user", content=[ContentBlock(text=prompt_ret)])
         ]
     else:
         assert isinstance(

src/ell/models/openai.py

@@ -73,7 +73,9 @@ def register(client: openai.Client):
         ('gpt-3.5-turbo-instruct', 'system'),
         ('gpt-4-0613', 'openai'),
         ('gpt-4', 'openai'),
-        ('gpt-4-0314', 'openai')
+        ('gpt-4-0314', 'openai'),
+        ('o1-preview', 'system'),
+        ('o1-mini', 'system'),
     ]
     for model_id, owned_by in model_data:
         config.register_model(model_id, client)

src/ell/providers/openai.py

@@ -87,7 +87,16 @@ try:
             final_call_params["model"] = model
             final_call_params["messages"] = openai_messages
 
-            if final_call_params.get("response_format"):
+            if model == "o1-preview" or model == "o1-mini":
+                # Ensure no system messages are present
+                assert all(msg['role'] != 'system' for msg in final_call_params['messages']), "System messages are not allowed for o1-preview or o1-mini models"
+                
+                response = client.chat.completions.create(**final_call_params)
+                final_call_params.pop("stream", None)
+                final_call_params.pop("stream_options", None)
+
+
+            elif final_call_params.get("response_format"):
                 final_call_params.pop("stream", None)
                 final_call_params.pop("stream_options", None)
                 response = client.beta.chat.completions.parse(**final_call_params)

tests/test_lmp_to_prompt.py

@@ -10,10 +10,6 @@
 
 
 
-# @lm(model="gpt-4-turbo", temperature=0.1, max_tokens=5)
-# def lmp_with_default_system_prompt(*args, **kwargs):
-#     return "Test user prompt"
-
 
 # @lm(model="gpt-4-turbo", temperature=0.1, max_tokens=5)
 # def lmp_with_docstring_system_prompt(*args, **kwargs):