formatting

dspy/teleprompt/bootstrap.py

@@ -84,13 +84,13 @@ def _prepare_predictor_mappings(self):
         student, teacher = self.student, self.teacher
 
         assert len(student.predictors()) == len(
-            teacher.predictors()
+            teacher.predictors(),
         ), "Student and teacher must have the same number of predictors."
 
         for (name1, predictor1), (name2, predictor2) in zip(student.named_predictors(), teacher.named_predictors()):
             assert name1 == name2, "Student and teacher must have the same program structure."
             assert predictor1.signature.equals(
-                predictor2.signature
+                predictor2.signature,
             ), f"Student and teacher must have the same signatures. {type(predictor1.signature)} != {type(predictor2.signature)}"
             assert id(predictor1) != id(predictor2), "Student and teacher must be different objects."
 
@@ -195,11 +195,11 @@ def _bootstrap_one_example(self, example, round_idx=0):
                     # TODO: Look closer into this. It's a bit tricky to reproduce.
                     print(f"Failed to find predictor {predictor} in {self.predictor2name}.")
                     print(
-                        "Are you doing this in a notebook (Jupyter)? This might be caused by redefining values by rerunning cells."
+                        "Are you doing this in a notebook (Jupyter)? This might be caused by redefining values by rerunning cells.",
                     )
                     print("Try restarting the notebook, or open an issue.")
                     raise KeyError(
-                        f"Failed to find predictor {id(predictor)} {predictor} in {self.predictor2name}."
+                        f"Failed to find predictor {id(predictor)} {predictor} in {self.predictor2name}.",
                     ) from e
 
                 name2traces[predictor_name].append(demo)

dspy/teleprompt/copro_optimizer.py

@@ -39,7 +39,7 @@ class BasicGenerateInstruction(Signature):
     basic_instruction = dspy.InputField(desc="The initial instructions before optimization")
     proposed_instruction = dspy.OutputField(desc="The improved instructions for the language model")
     proposed_prefix_for_output_field = dspy.OutputField(
-        desc="The string at the end of the prompt, which will help the model start solving the task"
+        desc="The string at the end of the prompt, which will help the model start solving the task",
     )
 
 
@@ -51,7 +51,7 @@ class GenerateInstructionGivenAttempts(dspy.Signature):
     attempted_instructions = dspy.InputField(format=dsp.passages2text)
     proposed_instruction = dspy.OutputField(desc="The improved instructions for the language model")
     proposed_prefix_for_output_field = dspy.OutputField(
-        desc="The string at the end of the prompt, which will help the model start solving the task"
+        desc="The string at the end of the prompt, which will help the model start solving the task",
     )
 
 
@@ -153,11 +153,15 @@ def compile(self, student, *, trainset, eval_kwargs):
             if self.prompt_model:
                 with dspy.settings.context(lm=self.prompt_model):
                     instruct = dspy.Predict(
-                        BasicGenerateInstruction, n=self.breadth - 1, temperature=self.init_temperature
+                        BasicGenerateInstruction,
+                        n=self.breadth - 1,
+                        temperature=self.init_temperature,
                     )(basic_instruction=basic_instruction)
             else:
                 instruct = dspy.Predict(
-                    BasicGenerateInstruction, n=self.breadth - 1, temperature=self.init_temperature
+                    BasicGenerateInstruction,
+                    n=self.breadth - 1,
+                    temperature=self.init_temperature,
                 )(basic_instruction=basic_instruction)
             # Add in our initial prompt as a candidate as well
             instruct.completions.proposed_instruction.append(basic_instruction)
@@ -175,7 +179,7 @@ def compile(self, student, *, trainset, eval_kwargs):
 
         # For each iteration in depth...
         for d in range(
-            self.depth
+            self.depth,
         ):  # TODO: fix this so that we eval the new batch of predictors with the new best followoing predictors
             print(f"Iteration Depth: {d+1}/{self.depth}.")
 
@@ -214,7 +218,7 @@ def compile(self, student, *, trainset, eval_kwargs):
                             print(f"Predictor {i+1}")
                         self._print_signature(predictor)
                     print(
-                        f"At Depth {d+1}/{self.depth}, Evaluating Prompt Candidate #{c_i+1}/{len(candidates_)} for Predictor {p_i+1} of {len(module.predictors())}."
+                        f"At Depth {d+1}/{self.depth}, Evaluating Prompt Candidate #{c_i+1}/{len(candidates_)} for Predictor {p_i+1} of {len(module.predictors())}.",
                     )
                     score = evaluate(module_clone, devset=trainset, **eval_kwargs)
                     if self.verbose and self.prompt_model:
@@ -264,7 +268,7 @@ def compile(self, student, *, trainset, eval_kwargs):
                 self._set_signature(p_new, updated_signature)
                 if self.verbose:
                     print(
-                        f"Updating Predictor {id(p_old)} to:\ni: {best_candidate['instruction']}\np: {best_candidate['prefix']}"
+                        f"Updating Predictor {id(p_old)} to:\ni: {best_candidate['instruction']}\np: {best_candidate['prefix']}",
                     )
                 if self.verbose:
                     print("Full predictor with update: ")
@@ -305,11 +309,15 @@ def compile(self, student, *, trainset, eval_kwargs):
                 if self.prompt_model:
                     with dspy.settings.context(lm=self.prompt_model):
                         instr = dspy.Predict(
-                            GenerateInstructionGivenAttempts, n=self.breadth, temperature=self.init_temperature
+                            GenerateInstructionGivenAttempts,
+                            n=self.breadth,
+                            temperature=self.init_temperature,
                         )(attempted_instructions=attempts)
                 else:
                     instr = dspy.Predict(
-                        GenerateInstructionGivenAttempts, n=self.breadth, temperature=self.init_temperature
+                        GenerateInstructionGivenAttempts,
+                        n=self.breadth,
+                        temperature=self.init_temperature,
                     )(attempted_instructions=attempts)
 
                 if self.verbose and self.prompt_model:
@@ -318,7 +326,7 @@ def compile(self, student, *, trainset, eval_kwargs):
                 new_candidates[id(p_base)] = instr.completions
                 all_candidates[id(p_base)].proposed_instruction.extend(instr.completions.proposed_instruction)
                 all_candidates[id(p_base)].proposed_prefix_for_output_field.extend(
-                    instr.completions.proposed_prefix_for_output_field
+                    instr.completions.proposed_prefix_for_output_field,
                 )
 
             if self.verbose and self.prompt_model:

dspy/teleprompt/finetune.py

@@ -82,7 +82,7 @@ def compile(
         if teacher is None:
             print(
                 "WARNING: Using a vanilla teacher. "
-                "Are you sure you want to use BootstrapFinetune without a compiled teacher?"
+                "Are you sure you want to use BootstrapFinetune without a compiled teacher?",
             )
 
         teachers = teacher if isinstance(teacher, list) else [teacher]
@@ -136,7 +136,7 @@ def compile(
 
         compiler_config = {
             "save": "".join(
-                random.Random(time.time()).choices(string.ascii_uppercase + string.digits, k=13)
+                random.Random(time.time()).choices(string.ascii_uppercase + string.digits, k=13),
             ),  # https://stackoverflow.com/a/2257449/1493011
             "peft": peft,
             "fp16": False,

dspy/teleprompt/knn_fewshot.py

@@ -18,7 +18,10 @@ def forward_pass(*args, **kwargs):
             knn_trainset = self.KNN(**kwargs)
             few_shot_bootstrap = BootstrapFewShot()
             compiled_program = few_shot_bootstrap.compile(
-                student, teacher=teacher, trainset=knn_trainset, valset=valset
+                student,
+                teacher=teacher,
+                trainset=knn_trainset,
+                valset=valset,
             )
             return compiled_program(**kwargs)
 

dspy/teleprompt/mipro_optimizer.py

@@ -51,7 +51,7 @@ class BasicGenerateInstruction(Signature):
     basic_instruction = dspy.InputField(desc="The initial instructions before optimization")
     proposed_instruction = dspy.OutputField(desc="The improved instructions for the language model")
     proposed_prefix_for_output_field = dspy.OutputField(
-        desc="The string at the end of the prompt, which will help the model start solving the task"
+        desc="The string at the end of the prompt, which will help the model start solving the task",
     )
 
 
@@ -62,7 +62,7 @@ class BasicGenerateInstructionWithDataObservations(Signature):
     observations = dspy.InputField(desc="Observations about the dataset and task")
     proposed_instruction = dspy.OutputField(desc="The improved instructions for the language model")
     proposed_prefix_for_output_field = dspy.OutputField(
-        desc="The string at the end of the prompt, which will help the model start solving the task"
+        desc="The string at the end of the prompt, which will help the model start solving the task",
     )
 
 
@@ -77,7 +77,7 @@ class BasicGenerateInstructionWithExamples(dspy.Signature):
     examples = dspy.InputField(format=dsp.passages2text, desc="Example(s) of the task")
     proposed_instruction = dspy.OutputField(desc="The improved instructions for the language model")
     proposed_prefix_for_output_field = dspy.OutputField(
-        desc="The string at the end of the prompt, which will help the model start solving the task"
+        desc="The string at the end of the prompt, which will help the model start solving the task",
     )
 
 
@@ -91,7 +91,7 @@ class BasicGenerateInstructionWithExamplesAndDataObservations(dspy.Signature):
     basic_instruction = dspy.InputField(desc="The initial instructions before optimization")
     proposed_instruction = dspy.OutputField(desc="The improved instructions for the language model")
     proposed_prefix_for_output_field = dspy.OutputField(
-        desc="The string at the end of the prompt, which will help the model start solving the task"
+        desc="The string at the end of the prompt, which will help the model start solving the task",
     )
 
 
@@ -100,7 +100,7 @@ class ObservationSummarizer(dspy.Signature):
 
     observations = dspy.InputField(desc="Observations I have made about my dataset")
     summary = dspy.OutputField(
-        desc="Two to Three sentence summary of only the most significant highlights of my observations"
+        desc="Two to Three sentence summary of only the most significant highlights of my observations",
     )
 
 
@@ -126,7 +126,7 @@ class DatasetDescriptorWithPriorObservations(dspy.Signature):
     examples = dspy.InputField(desc="Sample data points from the dataset")
     prior_observations = dspy.InputField(desc="Some prior observations I made about the data")
     observations = dspy.OutputField(
-        desc="Somethings that holds true for most or all of the data you observed or COMPLETE if you have nothing to add"
+        desc="Somethings that holds true for most or all of the data you observed or COMPLETE if you have nothing to add",
     )
 
 
@@ -180,7 +180,8 @@ def _observe_data(self, trainset, max_iterations=10):
         for b in range(self.view_data_batch_size, len(trainset), self.view_data_batch_size):
             upper_lim = min(len(trainset), b + self.view_data_batch_size)
             output = dspy.Predict(DatasetDescriptorWithPriorObservations, n=1, temperature=1.0)(
-                prior_observations=observations, examples=(trainset[b:upper_lim].__repr__())
+                prior_observations=observations,
+                examples=(trainset[b:upper_lim].__repr__()),
             )
             iterations += 1
             if len(output["observations"]) >= 8 and output["observations"][:8].upper() == "COMPLETE":
@@ -293,15 +294,17 @@ def _generate_first_N_candidates(  # noqa: N802
                             instruct = new_instruct
                         else:
                             instruct.completions.proposed_instruction.extend(
-                                new_instruct.completions.proposed_instruction
+                                new_instruct.completions.proposed_instruction,
                             )
                             instruct.completions.proposed_prefix_for_output_field.extend(
-                                new_instruct.completions.proposed_prefix_for_output_field
+                                new_instruct.completions.proposed_prefix_for_output_field,
                             )
                 # Just data
                 elif view_data:
                     instruct = dspy.Predict(
-                        BasicGenerateInstructionWithDataObservations, n=N - 1, temperature=self.init_temperature
+                        BasicGenerateInstructionWithDataObservations,
+                        n=N - 1,
+                        temperature=self.init_temperature,
                     )(basic_instruction=basic_instruction, observations=self.observations)
                 # Just examples
                 elif view_examples:
@@ -327,7 +330,7 @@ def _generate_first_N_candidates(  # noqa: N802
                 # Neither
                 else:
                     instruct = dspy.Predict(BasicGenerateInstruction, n=N - 1, temperature=self.init_temperature)(
-                        basic_instruction=basic_instruction
+                        basic_instruction=basic_instruction,
                     )
 
             # Add in our initial prompt as a candidate as well
@@ -365,7 +368,7 @@ def compile(
 
         estimated_task_model_calls_wo_module_calls = len(trainset) * num_trials  # M * T * P
         estimated_prompt_model_calls = 10 + self.num_candidates * len(
-            student.predictors()
+            student.predictors(),
         )  # num data summary calls + N * P
 
         user_message = textwrap.dedent(f"""\
@@ -454,7 +457,12 @@ def compile(
 
             # Generate N candidate prompts
             instruction_candidates, _ = self._generate_first_N_candidates(
-                module, self.num_candidates, view_data, view_examples, demo_candidates, trainset
+                module,
+                self.num_candidates,
+                view_data,
+                view_examples,
+                demo_candidates,
+                trainset,
             )
 
             # Reset demo_candidates to None for our optimization if the user asked for no fewshot examples
@@ -486,11 +494,13 @@ def objective(trial):
 
                         # Suggest the index of the instruction candidate to use in our trial
                         instruction_idx = trial.suggest_categorical(
-                            f"{id(p_old)}_predictor_instruction", range(len(p_instruction_candidates))
+                            f"{id(p_old)}_predictor_instruction",
+                            range(len(p_instruction_candidates)),
                         )
                         if demo_candidates:
                             demos_idx = trial.suggest_categorical(
-                                f"{id(p_old)}_predictor_demos", range(len(p_demo_candidates))
+                                f"{id(p_old)}_predictor_demos",
+                                range(len(p_demo_candidates)),
                             )
                         trial_logs[trial_num][f"{id(p_old)}_predictor_instruction"] = instruction_idx
                         if demo_candidates:

dspy/teleprompt/signature_opt.py

@@ -44,7 +44,7 @@ def __init__(
         track_stats=False,
     ):
         print(
-            "\u001b[31m[WARNING] SignatureOptimizer has been deprecated and replaced with COPRO.  SignatureOptimizer will be removed in a future release. \u001b[31m"
+            "\u001b[31m[WARNING] SignatureOptimizer has been deprecated and replaced with COPRO.  SignatureOptimizer will be removed in a future release. \u001b[31m",
         )
         super().__init__(prompt_model, metric, breadth, depth, init_temperature, verbose, track_stats)
 

dspy/teleprompt/signature_opt_bayesian.py

@@ -49,7 +49,7 @@ def __init__(
         view_data_batch_size=10,
     ):
         print(
-            "\u001b[31m[WARNING] BayesianSignatureOptimizer has been deprecated and replaced with MIPRO.  BayesianSignatureOptimizer will be removed in a future release. \u001b[31m"
+            "\u001b[31m[WARNING] BayesianSignatureOptimizer has been deprecated and replaced with MIPRO.  BayesianSignatureOptimizer will be removed in a future release. \u001b[31m",
         )
 
         super().__init__(

dspy/teleprompt/teleprompt_optuna.py

@@ -39,7 +39,7 @@ def __init__(
     def objective(self, trial):
         program2 = self.student.reset_copy()
         for (name, compiled_predictor), (_, program2_predictor) in zip(
-            self.compiled_teleprompter.named_predictors(), program2.named_predictors()
+            self.compiled_teleprompter.named_predictors(), program2.named_predictors(),
         ):
             all_demos = compiled_predictor.demos
             demo_index = trial.suggest_int(f"demo_index_for_{name}", 0, len(all_demos) - 1)
@@ -69,7 +69,7 @@ def compile(self, student, *, teacher=None, max_demos, trainset, valset=None):
             max_rounds=self.max_rounds,
         )
         self.compiled_teleprompter = teleprompter_optimize.compile(
-            self.student, teacher=self.teacher, trainset=self.trainset
+            self.student, teacher=self.teacher, trainset=self.trainset,
         )
         study = optuna.create_study(direction="maximize")
         study.optimize(self.objective, n_trials=self.num_candidate_sets)