Print results as they are produced

Signed-off-by: David Gageot <david.gageot@docker.com>

Author: dgageot

cmd/root/eval.go

@@ -52,16 +52,11 @@ func (f *evalFlags) runEvalCommand(cmd *cobra.Command, args []string) error {
 		return err
 	}
 
-	results, err := evaluation.Evaluate(ctx, agents, evalsDir)
-	if err != nil {
-		return err
-	}
-
-	for _, result := range results {
+	_, err = evaluation.Evaluate(ctx, agents, evalsDir, func(result evaluation.Result) {
 		out.Printf("Eval file: %s\n", result.EvalFile)
 		out.Printf("Tool trajectory score: %f\n", result.Score.ToolTrajectoryScore)
 		out.Printf("Rouge-1 score: %f\n", result.Score.Rouge1Score)
-	}
 
-	return nil
+	})
+	return err
 }

pkg/evaluation/evaluation.go

@@ -22,7 +22,7 @@ type Result struct {
 	EvalFile string
 }
 
-func Evaluate(ctx context.Context, t *team.Team, evalsDir string) ([]Result, error) {
+func Evaluate(ctx context.Context, t *team.Team, evalsDir string, onResult func(Result)) ([]Result, error) {
 	evalFiles, err := os.ReadDir(evalsDir)
 	if err != nil {
 		return nil, err
@@ -56,11 +56,13 @@ func Evaluate(ctx context.Context, t *team.Team, evalsDir string) ([]Result, err
 		}
 
 		score := score(evals[i].GetAllMessages(), actualMessages)
-
-		results = append(results, Result{
+		result := Result{
 			Score:    score,
 			EvalFile: evals[i].ID,
-		})
+		}
+		onResult(result)
+
+		results = append(results, result)
 	}
 
 	return results, nil