Add reward for correct letter no matter if it has been rewarded before

wordle_env/state.py

@@ -148,8 +148,7 @@ def update(state: WordleState, word: str, goal_word: str) -> Tuple[WordleState,
         offset = 1 + cint * WORDLE_N * 3
         if goal_word[i] == c:
             # char at position i = yes, all other chars at position i == no
-            if state[offset + 3 * i:offset + 3 * i + 3][2] == 0:
-                reward += CHAR_REWARD
+            reward += CHAR_REWARD
             state[offset + 3 * i:offset + 3 * i + 3] = [0, 0, 1]
             for ocint in range(len(WORDLE_CHARS)):
                 if ocint != cint:
@@ -165,6 +164,7 @@ def update(state: WordleState, word: str, goal_word: str) -> Tuple[WordleState,
                 # Char at position i = no, and in other positions maybe, other chars stay as they are
                 state[offset:offset + 3 * WORDLE_N] = [0, 1, 0] * WORDLE_N
                 state[offset + 3 * i:offset + 3 * i + 3] = [1, 0, 0]
+                reward += CHAR_REWARD * 0.1
             else:
                 # Char at all positions = no
                 state[offset:offset + 3 * WORDLE_N] = [1, 0, 0] * WORDLE_N