Fix uninitialized prompts bug

CHANGELOG.md

@@ -25,6 +25,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Added support for Python 3.8.
 - Added code to throw an error if `output_attentions` is set to `True` in forward call to `OLMoForCausalLM`. This functionality hasn't been implemented yet.
 - Fixed running with data loading workers on LUMI
+- Minor bug fix: uninitialized prompts variable
 
 ### Added
 - Added `output_hidden_states` argument and associated functionality to `OLMo` and `OLMoForCausalLM` to return model intermediate hidden states.

olmo/eval/downstream.py

@@ -165,7 +165,7 @@ def __init__(
         self.model_ctx_len = model_ctx_len
         self.prompts = prompts
         self.current_prompt = None
-        self.log_instances = 5  # Log the first few instances as a sanity check
+        self.log_instances = 0  # Set to > 0 to log the first few instances as a sanity check
 
         self.samples: List[Dict[str, Any]] = []
         dataset_names: Sequence[Optional[str]]
@@ -1099,6 +1099,7 @@ def __init__(
                 if dataset_name in cats:
                     dataset_names.append(name)
         self.dev_set = {}
+        prompts: List[Union[None, str]] = [None]
         if prompt_variations == 1:
             prompts = [None, "inst", "inst+1", "inst+2", "inst+3", "inst+4", "inst+5"]
             # Need to grab the dev set for the few-shot prompts