[rllib] Added vanilla policy gradient (#1497)

2026-08-05 13:21:03 +08:00 · 2018-02-10 13:54:51 -08:00
parent 1ab2e63dbd
commit 81a4be8f65
9 changed files with 258 additions and 2 deletions
@@ -120,6 +120,10 @@ class ModelSupportedSpaces(unittest.TestCase):
            {"num_workers": 1, "noise_size": 10000000,
             "episodes_per_batch": 1, "timesteps_per_batch": 1},
            stats)
+        check_support(
+            "PG",
+            {"num_workers": 1, "optimizer": {"grads_per_step": 1}},
+            stats)
        num_unexpected_errors = 0
        num_unexpected_success = 0
        for (alg, a_name, o_name), stat in sorted(stats.items()):