Comet ml updates 17 04 2023 (#3074)

I made a couple of improvements to the Comet tracker:

* The Comet project name is configurable in various ways (code,
environment variable or file), having a default value in code meant that
users couldn't set the project name in an environment variable or in a
file.
* I added error catching when the `flush_tracker` is called in order to
avoid crashing the whole process. Instead we are gonna display a warning
or error log message (`extra={"show_traceback": True}` is an internal
convention to force the display of the traceback when using our own
logger).

I decided to add the error catching after seeing the following error in
the third example of the notebook:
```
COMET ERROR: Failed to export agent or LLM to Comet
Traceback (most recent call last):
  File "/home/lothiraldan/project/cometml/langchain/langchain/callbacks/comet_ml_callback.py", line 484, in _log_model
    langchain_asset.save(langchain_asset_path)
  File "/home/lothiraldan/project/cometml/langchain/langchain/agents/agent.py", line 591, in save
    raise ValueError(
ValueError: Saving not supported for agent executors. If you are trying to save the agent, please use the `.save_agent(...)`

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/lothiraldan/project/cometml/langchain/langchain/callbacks/comet_ml_callback.py", line 449, in flush_tracker
    self._log_model(langchain_asset)
  File "/home/lothiraldan/project/cometml/langchain/langchain/callbacks/comet_ml_callback.py", line 488, in _log_model
    langchain_asset.save_agent(langchain_asset_path)
  File "/home/lothiraldan/project/cometml/langchain/langchain/agents/agent.py", line 599, in save_agent
    return self.agent.save(file_path)
  File "/home/lothiraldan/project/cometml/langchain/langchain/agents/agent.py", line 145, in save
    agent_dict = self.dict()
  File "/home/lothiraldan/project/cometml/langchain/langchain/agents/agent.py", line 119, in dict
    _dict = super().dict()
  File "pydantic/main.py", line 449, in pydantic.main.BaseModel.dict
  File "pydantic/main.py", line 868, in _iter
  File "pydantic/main.py", line 743, in pydantic.main.BaseModel._get_value
  File "/home/lothiraldan/project/cometml/langchain/langchain/schema.py", line 381, in dict
    output_parser_dict["_type"] = self._type
  File "/home/lothiraldan/project/cometml/langchain/langchain/schema.py", line 376, in _type
    raise NotImplementedError
NotImplementedError
```

I still need to investigate and try to fix it, it looks related to
saving an agent to a file.

langchain/callbacks/comet_ml_callback.py

@@ -96,7 +96,7 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
         self,
         task_type: Optional[str] = "inference",
         workspace: Optional[str] = None,
-        project_name: Optional[str] = "comet-langchain-demo",
+        project_name: Optional[str] = None,
         tags: Optional[Sequence] = None,
         name: Optional[str] = None,
         visualizations: Optional[List[str]] = None,
@@ -106,7 +106,7 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
     ) -> None:
         """Initialize callback handler."""
 
-        comet_ml = import_comet_ml()
+        self.comet_ml = import_comet_ml()
         super().__init__()
 
         self.task_type = task_type
@@ -133,7 +133,7 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
             "https://github.com/comet-ml/issue_tracking/issues with the tag "
             "`langchain`."
         )
-        comet_ml.LOGGER.warning(warning)
+        self.comet_ml.LOGGER.warning(warning)
 
         self.callback_columns: list = []
         self.action_records: list = []
@@ -242,8 +242,6 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
         resp.update(flatten_dict(serialized))
         resp.update(self.get_custom_callback_meta())
 
-        comet_ml = import_comet_ml()
-
         for chain_input_key, chain_input_val in inputs.items():
             if isinstance(chain_input_val, str):
                 input_resp = deepcopy(resp)
@@ -253,7 +251,7 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
                 self.action_records.append(input_resp)
 
             else:
-                comet_ml.LOGGER.warning(
+                self.comet_ml.LOGGER.warning(
                     f"Unexpected data format provided! "
                     f"Input Value for {chain_input_key} will not be logged"
                 )
@@ -268,8 +266,6 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
         resp.update({"action": "on_chain_end"})
         resp.update(self.get_custom_callback_meta())
 
-        comet_ml = import_comet_ml()
-
         for chain_output_key, chain_output_val in outputs.items():
             if isinstance(chain_output_val, str):
                 output_resp = deepcopy(resp)
@@ -278,7 +274,7 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
                 output_resp.update({chain_output_key: chain_output_val})
                 self.action_records.append(output_resp)
             else:
-                comet_ml.LOGGER.warning(
+                self.comet_ml.LOGGER.warning(
                     f"Unexpected data format provided! "
                     f"Output Value for {chain_output_key} will not be logged"
                 )
@@ -449,7 +445,14 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
         self._log_session(langchain_asset)
 
         if langchain_asset:
+            try:
                 self._log_model(langchain_asset)
+            except Exception:
+                self.comet_ml.LOGGER.error(
+                    "Failed to export agent or LLM to Comet",
+                    exc_info=True,
+                    extra={"show_traceback": True},
+                )
 
         if finish:
             self.experiment.end()
@@ -470,8 +473,6 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
         self.experiment.log_text(prompt, metadata=metadata, step=step)
 
     def _log_model(self, langchain_asset: Any) -> None:
-        comet_ml = import_comet_ml()
-
         model_parameters = self._get_llm_parameters(langchain_asset)
         self.experiment.log_parameters(model_parameters, prefix="model")
 
@@ -487,24 +488,45 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
                 langchain_asset.save_agent(langchain_asset_path)
                 self.experiment.log_model(model_name, str(langchain_asset_path))
             else:
-                comet_ml.LOGGER.warning(
+                self.comet_ml.LOGGER.error(
                     f"{e}"
                     " Could not save Langchain Asset "
                     f"for {langchain_asset.__class__.__name__}"
                 )
 
     def _log_session(self, langchain_asset: Optional[Any] = None) -> None:
+        try:
             llm_session_df = self._create_session_analysis_dataframe(langchain_asset)
             # Log the cleaned dataframe as a table
             self.experiment.log_table("langchain-llm-session.csv", llm_session_df)
+        except Exception:
+            self.comet_ml.LOGGER.warning(
+                "Failed to log session data to Comet",
+                exc_info=True,
+                extra={"show_traceback": True},
+            )
 
+        try:
             metadata = {"langchain_version": str(langchain.__version__)}
             # Log the langchain low-level records as a JSON file directly
             self.experiment.log_asset_data(
                 self.action_records, "langchain-action_records.json", metadata=metadata
             )
+        except Exception:
+            self.comet_ml.LOGGER.warning(
+                "Failed to log session data to Comet",
+                exc_info=True,
+                extra={"show_traceback": True},
+            )
 
+        try:
             self._log_visualizations(llm_session_df)
+        except Exception:
+            self.comet_ml.LOGGER.warning(
+                "Failed to log visualizations to Comet",
+                exc_info=True,
+                extra={"show_traceback": True},
+            )
 
     def _log_text_metrics(self, metrics: Sequence[dict], step: int) -> None:
         if not metrics:
@@ -519,7 +541,6 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
             return
 
         spacy = import_spacy()
-        comet_ml = import_comet_ml()
 
         prompts = session_df["prompts"].tolist()
         outputs = session_df["text"].tolist()
@@ -544,7 +565,9 @@ class CometCallbackHandler(BaseMetadataCallbackHandler, BaseCallbackHandler):
                         step=idx,
                     )
                 except Exception as e:
-                    comet_ml.LOGGER.warning(e)
+                    self.comet_ml.LOGGER.warning(
+                        e, exc_info=True, extra={"show_traceback": True}
+                    )
 
         return