aPonza
aPonza
I'm temporarily working on another thing for a couple weeks, I was editing `franka_control_wrappers/src/franka_control_wrappers/panda_commander.py` like this: ``` diff --git a/franka_control_wrappers/src/franka_control_wrappers/panda_commander.py b/franka_control_wrappers/src/franka_control_wrappers/panda_commander.py index 6abe106..3e606ee 100644 --- a/franka_control_wrappers/src/franka_control_wrappers/panda_commander.py +++ b/franka_control_wrappers/src/franka_control_wrappers/panda_commander.py @@ -22,6...
Hey, just got a moment to continue here, you're right in that `self` is needed, and in fact this is what I currently have ``` diff --git a/franka_control_wrappers/src/franka_control_wrappers/panda_commander.py b/franka_control_wrappers/src/franka_control_wrappers/panda_commander.py index...
I can see where my error was: I wasn't including the two joint values for the fingers. Anyways it's weird, because it goes to "ready" but not to "grip_ready"... and...
I checked `panda_moveit_config` and I only see trajopt changes and a couple extra default poses, but nothing that would bother a missing end effector. Similarly, for `franka_ros`, I can see...
Mmm you might actually want to change this line https://github.com/dougsm/mvp_grasp/blob/48330d45a135a5cf1bb058909ed1323be200069f/mvp_grasping/src/mvp_grasping/panda_base_grasping_controller.py#L170 to say `'panda_arm'`, instead of the other two edits you made. It does do more than before, but since method...
``` $ git diff ggcnn/src/ggcnn/ggcnn.py diff --git a/ggcnn/src/ggcnn/ggcnn.py b/ggcnn/src/ggcnn/ggcnn.py index 396e627..8a49dc5 100644 --- a/ggcnn/src/ggcnn/ggcnn.py +++ b/ggcnn/src/ggcnn/ggcnn.py @@ -8,10 +8,17 @@ import tensorflow as tf from tensorflow.keras.models import load_model from dougsm_helpers.timeit...
Ok, actually it seems to almost work, I did not have any pickable under the camera, but adding one started moving the robot and the eef went some centimeters off...
I don't have a gpu on my workbench at the moment (hence > ~~This leads to tensorflow issues, which is what I suspected would happen installing tensorflow instead of tensorflow-gpu~~...
> with tensorflow-cpu I don't get the error as above anymore You might have an incompatible cuDNN, maybe?
Cheers! I should probably push a PR with the changes before you start then.