reasoning output #9
Conversation
src/nuggetizer/core/llm.py
Outdated
| # print(f"🔍 DEBUG LLM: API call completed successfully") # not | ||
| # removed because it's very helpful for debugging |
There was a problem hiding this comment.
- I think we should remove these even if it is helpful for debugging, coding practice wise it is for the developer to debug.
- We could have a --debug mode where we expose such things, but that is a separate PR. And we shouldn't be doing print within function methods, logging is a better alternative.
There was a problem hiding this comment.
I see let me remove this then. We can do debug mode in another PR if needed
src/nuggetizer/core/llm.py
Outdated
| # print(f"🔍 DEBUG LLM: Full response: {completion}") # not | ||
| # removed because it's very helpful for debugging |
src/nuggetizer/core/llm.py
Outdated
| response = reasoning_content if reasoning_content else "" | ||
| reasoning_content = message["reasoning_content"] | ||
| else: | ||
| print(f"No reasoning found in response from {self.model}") |
There was a problem hiding this comment.
should this be printed? Feels spammy. if the model is a reasoning model and we hit here, we reach some error state and should warn. else this should not be printing
There was a problem hiding this comment.
this is also debugging print. removed
src/nuggetizer/core/llm.py
Outdated
| "qwen" in self.model.lower() | ||
| or "qwen2" in self.model.lower() | ||
| or "qwen3" in self.model.lower() | ||
| ): | ||
| # Use cl100k_base for Qwen models as they typically use | ||
| # similar tokenization | ||
| encoding = tiktoken.get_encoding("cl100k_base") |
There was a problem hiding this comment.
let's not conflate this into this PR, if we don't know the model just don't include the encoding
| if self.log_level >= 1: | ||
| self.logger.info( | ||
| f"Initialized Nuggetizer with models: {creator_model}, {scorer_model}, {assigner_model}" | ||
| ) | ||
| if log_level >= 1: | ||
| self.logger.setLevel(logging.INFO) | ||
| if log_level >= 2: | ||
| self.logger.setLevel(logging.DEBUG) |
There was a problem hiding this comment.
what's this code doing, we don't update log_level?
There was a problem hiding this comment.
we can specify log level in the input. same debugging logs appear only if we specify log level to be 2
I see thanks for pointing out. I will break them into smaller PRs |
2 participants
This PR enables reasoning output.