Commit ·
bc8710c
1
Parent(s): 55a5567
Fix score bounds for validator
Browse files- env/environment.py +2 -2
env/environment.py
CHANGED
|
@@ -49,7 +49,7 @@ class CustomerSupportEnv:
|
|
| 49 |
max_steps = self.current_task.get("max_steps", 10)
|
| 50 |
if self.step_count >= max_steps and not self.done:
|
| 51 |
self.done = True
|
| 52 |
-
new_score = max(0.
|
| 53 |
reward = Reward(
|
| 54 |
score=new_score,
|
| 55 |
feedback=reward.feedback + " | time limit hit, -0.05",
|
|
@@ -117,7 +117,7 @@ class CustomerSupportEnv:
|
|
| 117 |
score += bonus
|
| 118 |
breakdown["close_bonus"] = bonus
|
| 119 |
|
| 120 |
-
score = round(max(0.
|
| 121 |
feedback = self._make_feedback(action, breakdown, correct)
|
| 122 |
|
| 123 |
return Reward(score=score, feedback=feedback, breakdown=breakdown)
|
|
|
|
| 49 |
max_steps = self.current_task.get("max_steps", 10)
|
| 50 |
if self.step_count >= max_steps and not self.done:
|
| 51 |
self.done = True
|
| 52 |
+
new_score = max(0.001, reward.score - 0.05)
|
| 53 |
reward = Reward(
|
| 54 |
score=new_score,
|
| 55 |
feedback=reward.feedback + " | time limit hit, -0.05",
|
|
|
|
| 117 |
score += bonus
|
| 118 |
breakdown["close_bonus"] = bonus
|
| 119 |
|
| 120 |
+
score = round(max(0.001, min(0.999, score)), 4)
|
| 121 |
feedback = self._make_feedback(action, breakdown, correct)
|
| 122 |
|
| 123 |
return Reward(score=score, feedback=feedback, breakdown=breakdown)
|