Merge pull request #51 from vincentzhang:main

copybara-github · copybara-github · commit f7782209e75c · 2025-02-10T15:35:12.000-08:00
PiperOrigin-RevId: 725371773
Change-Id: I83058caf6457e2b96b7f52f900f5fb62e018de1e
diff --git a/learning/notebooks/locomotion.ipynb b/learning/notebooks/locomotion.ipynb
@@ -970,7 +970,7 @@
         "id": "v5p0Z3PPSRik"
       },
       "source": [
-        "Notive that the above policy looks jittery and unlikely to transfer on the robot. The max power outut is also quite high.\n",
+        "Notice that the above policy looks jittery and unlikely to transfer on the robot. The max power output is also quite high.\n",
         "\n",
         "The sim-to-real deployment of the handstand policy was trained using a curriculum on the `energy_termination_threshold`, `energy` and `dof_acc`, which are config values that penalize high torques and high power output. Let's finetune the above policy with a decreased  `energy_termination_threshold`, as well as non-zero values for `energy` and `dof_acc` rewards to get a smoother policy."
       ]

Original file line number	Diff line number	Diff line change
`@@ -970,7 +970,7 @@`
`970`	`970`	`"id": "v5p0Z3PPSRik"`
`971`	`971`	`},`
`972`	`972`	`"source": [`
`973`		`- "Notive that the above policy looks jittery and unlikely to transfer on the robot. The max power outut is also quite high.\n",`
	`973`	`+ "Notice that the above policy looks jittery and unlikely to transfer on the robot. The max power output is also quite high.\n",`
`974`	`974`	`"\n",`
`975`	`975`	"The sim-to-real deployment of the handstand policy was trained using a curriculum on the `energy_termination_threshold`, `energy` and `dof_acc`, which are config values that penalize high torques and high power output. Let's finetune the above policy with a decreased `energy_termination_threshold`, as well as non-zero values for `energy` and `dof_acc` rewards to get a smoother policy."
`976`	`976`	`]`