vanilla_policy_gradient: fix typo

Former-commit-id: fc5e3eb
EffiSciencesResearch · Aug 28, 2023 · ab14e18 · ab14e18
1 parent 201be50
commit ab14e18
Show file tree

Hide file tree

Showing 2 changed files with 2 additions and 2 deletions.
diff --git a/days/w1d5/1_simple_pg_correction.py b/days/w1d5/1_simple_pg_correction.py
@@ -38,7 +38,7 @@ def mlp(sizes, activation=nn.Tanh, output_activation=nn.Identity):
         layers += [nn.Linear(sizes[j], sizes[j+1]), act()]
     return nn.Sequential(*layers)
 
-def train(env_name='CartPole-v0', hidden_sizes=[32], lr=1e-2, 
+def train(env_name='CartPole-v0', hidden_sizes=[32], lr=1e-2,
           epochs=50, batch_size=5000, render=False):
 
     # make environment, check spaces, get obs / act dims

diff --git a/days/w1d5/vanilla_policy_gradient.ipynb b/days/w1d5/vanilla_policy_gradient.ipynb
@@ -97,7 +97,7 @@
     "    # What is the shape of obs?\n",
     "    @typechecked # To be typed\n",
     "    def get_action(obs):\n",
-    "        return get_policy(obs.unsquezze(0)).sample().item()\n",
+    "        return get_policy(obs.unsqueeze(0)).sample().item()\n",
     "\n",
     "    # make loss function whose gradient, for the right data, is policy gradient\n",
     "    # What is the shape of obs?\n",