Skip to content

armgit / dspy

Public Imported from GitHub
0

DSPy: The framework for programming—not prompting—language models

Python

fix(dspy): add metric_threshold to MIPROv2

William Horton7cd4011400ae
diff --git a/dspy/teleprompt/mipro_optimizer_v2.py b/dspy/teleprompt/mipro_optimizer_v2.py
index fbd5cbb67..82d8e50f6 100644
--- a/dspy/teleprompt/mipro_optimizer_v2.py
+++ b/dspy/teleprompt/mipro_optimizer_v2.py
@@ -67,6 +67,7 @@ Note that this teleprompter takes in the following parameters:
* view_data_batch_size: The number of examples to view in the data batch when producing the dataset summary. Default=10.
* minibatch_size: The size of the minibatch to use when evaluating the program if using minibatched evaluations. Default=25.
* minibatch_full_eval_steps: The number of steps to take before doing a full evaluation of the program if using minibatched evaluations. Default=10.
+* metric_threshold: If the metric yields a numerical value, then check it against this threshold when deciding whether or not to accept a bootstrap example.
"""
BOOTSTRAPPED_FEWSHOT_EXAMPLES_IN_CONTEXT = 3
@@ -90,6 +91,7 @@ class MIPROv2(Teleprompter):
view_data_batch_size=10,
minibatch_size=MINIBATCH_SIZE,
minibatch_full_eval_steps=MB_FULL_EVAL_STEPS,
+ metric_threshold=None,
):
self.n = num_candidates
self.metric = metric
@@ -105,6 +107,7 @@ class MIPROv2(Teleprompter):
self.total_calls = 0
self.minibatch_size = minibatch_size
self.minibatch_full_eval_steps = minibatch_full_eval_steps
+ self.metric_threshold = None
# Check if WANDB_RUN_ID is set in the environment
self.wandb_run_id = None
@@ -273,6 +276,7 @@ class MIPROv2(Teleprompter):
metric=self.metric,
teacher_settings=self.teacher_settings,
seed=seed,
+ metric_threshold=metric_threshold,
)
except Exception as e:
print(f"Error generating fewshot examples: {e}")