mat kelcey
mat kelcey
the entire point of me starting this project was to try to train a grasping model but i haven't got there yet o_O as things are i get the feeling...
everything about this training has focussed on the arm, how do we anneal in hard/easy triples w.r.t the objects? one way would be to start with a smaller number (say...
the above talks about hard negatives and hard positives but i only trained hard negatives; i should do some more work on the mining of explicit hard positives.
additional to the offline mining another win would be to borrow an idea from from offline reinforcement learning; the replay buffer. if we mine triples we can use them to...
if our main goal is to keep a training loop busy doing useful work (i.e. minimising zero loss cases) we can farm out the checking of triples to a fleet...