Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Llm bot id classifier corrected #1646

Closed
wants to merge 2 commits into from
Closed

Llm bot id classifier corrected #1646

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
8b54a96
added secret and bot_id in config
sfahad1414 Dec 12, 2024
7daa2f9
config corrected
sfahad1414 Dec 12, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 5 additions & 5 deletions kairon/nlu/classifiers/llm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -82,17 +82,17 @@ def load_api_key(self, bot_id: Text):
if bot_id:
from kairon.shared.admin.processor import Sysadmin
llm_secret = Sysadmin.get_llm_secret("openai", bot_id)
self.api_key = llm_secret.get('api_key')
self.secret = llm_secret.get('api_key')
elif os.environ.get("LLM_API_KEY"):
self.api_key = os.environ.get("LLM_API_KEY")
self.secret = {'api_key': os.environ.get("LLM_API_KEY")}
Comment on lines +85 to +87
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🛠️ Refactor suggestion

Add validation for API key and type hints

The API key handling could be more robust with proper validation and type hints.

+    secret: Dict[str, str]
+
     def load_api_key(self, bot_id: Text):
         if bot_id:
             from kairon.shared.admin.processor import Sysadmin
             llm_secret = Sysadmin.get_llm_secret("openai", bot_id)
-            self.secret = llm_secret.get('api_key')
+            api_key = llm_secret.get('api_key')
+            if not api_key:
+                raise ValueError("API key not found in llm_secret")
+            self.secret = {'api_key': api_key}
         elif os.environ.get("LLM_API_KEY"):
             self.secret = {'api_key': os.environ.get("LLM_API_KEY")}
         else:
             raise KeyError(
                 "either set bot_id in LLMClassifier config or set LLM_API_KEY in environment variables"
             )
📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change
self.secret = llm_secret.get('api_key')
elif os.environ.get("LLM_API_KEY"):
self.api_key = os.environ.get("LLM_API_KEY")
self.secret = {'api_key': os.environ.get("LLM_API_KEY")}
secret: Dict[str, str]
def load_api_key(self, bot_id: Text):
if bot_id:
from kairon.shared.admin.processor import Sysadmin
llm_secret = Sysadmin.get_llm_secret("openai", bot_id)
api_key = llm_secret.get('api_key')
if not api_key:
raise ValueError("API key not found in llm_secret")
self.secret = {'api_key': api_key}
elif os.environ.get("LLM_API_KEY"):
self.secret = {'api_key': os.environ.get("LLM_API_KEY")}
else:
raise KeyError(
"either set bot_id in LLMClassifier config or set LLM_API_KEY in environment variables"
)

else:
raise KeyError(
f"either set bot_id'in LLMClassifier config or set LLM_API_KEY in environment variables"
)

def get_embeddings(self, text):
embeddings = litellm.embedding(
model="text-embedding-3-small", input=text, api_key=self.api_key, max_retries=3
model="text-embedding-3-small", input=text, max_retries=3, **self.secret
)
Comment on lines +95 to 96
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🛠️ Refactor suggestion

Use configuration for embedding model

The embedding model name is hardcoded despite being available in the component configuration.

         embeddings = litellm.embedding(
-            model="text-embedding-3-small", input=text, max_retries=3, **self.secret
+            model=self.component_config.get("embedding_model", "text-embedding-3-small"),
+            input=text,
+            max_retries=self.component_config.get("retry", 3),
+            **self.secret
         )
📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change
model="text-embedding-3-small", input=text, max_retries=3, **self.secret
)
model=self.component_config.get("embedding_model", "text-embedding-3-small"),
input=text,
max_retries=self.component_config.get("retry", 3),
**self.secret
)

return [ embedding['embedding'] for embedding in embeddings['data']]

Expand Down Expand Up @@ -188,8 +188,8 @@ def predict(self, text):
top_p=1,
frequency_penalty=0,
presence_penalty=0,
api_key=self.api_key,
max_retries=3
max_retries=3,
**self.secret
Comment on lines +191 to +192
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🛠️ Refactor suggestion

Improve error handling in predict method

The error handling could be more specific and informative.

             max_retries=3,
             **self.secret
         )
         logger.debug(response)
         responses = json.loads(response.choices[0]["message"]["content"])
         intent = responses["intent"] if "intent" in responses.keys() else "nlu_fallback"
         explanation = responses["explanation"] if "explanation" in responses.keys() else None
         entities = responses["entities"]if "entities" in responses.keys() else []
-    except Exception as e:
-        logger.error(e)
+    except json.JSONDecodeError as e:
+        logger.error(f"Failed to parse LLM response: {e}")
+    except (KeyError, IndexError) as e:
+        logger.error(f"Unexpected response format from LLM: {e}")
+    except Exception as e:
+        logger.error(f"Error during prediction: {str(e)}")

Committable suggestion skipped: line range outside the PR's diff.

)
logger.debug(response)
responses = json.loads(response.choices[0]["message"]["content"])
Expand Down
7 changes: 7 additions & 0 deletions kairon/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,6 +38,11 @@ def train_model_for_bot(bot: str):
stories = stories.merge(multiflow_stories[0])
config = processor.load_config(bot)
config['assistant_id'] = bot

index = next((index for (index, d) in enumerate(config['pipeline']) if d["name"] == "kairon.nlu.LLMClassifier"), None)
if index:
config['pipeline'][index]['bot_id']= bot
Comment on lines +42 to +44
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🛠️ Refactor suggestion

Add error handling for LLMClassifier configuration

While the logic is correct, consider adding error handling to make the code more robust:

-        index = next((index for (index, d) in enumerate(config['pipeline']) if d["name"] == "kairon.nlu.LLMClassifier"), None)
-        if index:
-            config['pipeline'][index]['bot_id']= bot
+        if not config.get('pipeline'):
+            raise ValueError("Pipeline configuration is missing")
+        
+        index = next((index for (index, d) in enumerate(config['pipeline']) if d["name"] == "kairon.nlu.LLMClassifier"), None)
+        if index is not None:
+            config['pipeline'][index]['bot_id'] = bot
+        else:
+            logging.debug("LLMClassifier not found in pipeline configuration")
📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change
index = next((index for (index, d) in enumerate(config['pipeline']) if d["name"] == "kairon.nlu.LLMClassifier"), None)
if index:
config['pipeline'][index]['bot_id']= bot
if not config.get('pipeline'):
raise ValueError("Pipeline configuration is missing")
index = next((index for (index, d) in enumerate(config['pipeline']) if d["name"] == "kairon.nlu.LLMClassifier"), None)
if index is not None:
config['pipeline'][index]['bot_id'] = bot
else:
logging.debug("LLMClassifier not found in pipeline configuration")


rules = processor.get_rules_for_training(bot)
rules = rules.merge(multiflow_stories[1])

Expand Down Expand Up @@ -124,3 +129,5 @@ def start_training(bot: str, user: str, token: str = None):
exception=exception,
)
return model_file