def dump_weights_file(env, fpath, trainer=None):
fpath = io.assert_extension(fpath, __weights_ext__)
weights = env.network.fetch_all_variables_dict()
if trainer is not None:
trainer.trigger_event('plugin:weights:dump', weights)
io.dump(fpath, weights)
return fpath
def dump_snapshot_on_epoch_after(trainer):
if trainer.epoch % save_interval != 0:
return
snapshot_dir = get_snapshot_dir()
snapshot = trainer.dump_snapshot()
fpath = osp.join(snapshot_dir, 'epoch_{}'.format(trainer.epoch) + __snapshot_ext__)
io.mkdir(osp.dirname(fpath))
io.dump(fpath, snapshot)
fpath_aliased = []
fpath_last = osp.join(snapshot_dir, 'last_epoch' + __snapshot_ext__)
fpath_aliased.append(fpath_last)
best_loss = trainer.runtime.get('best_loss', np.inf)
current_loss = trainer.runtime.get('loss', None)
if current_loss is not None and best_loss > current_loss:
trainer.runtime['best_loss'] = current_loss
fpath_best_loss = osp.join(snapshot_dir, 'best_loss' + __snapshot_ext__)
fpath_aliased.append(fpath_best_loss)
best_error = trainer.runtime.get('best_error', np.inf)
current_error = trainer.runtime.get('error', None)
if current_error is not None and best_error > current_error:
trainer.runtime['best_error'] = current_error
fpath_best_error = osp.join(snapshot_dir, 'best_error' + __snapshot_ext__)
fpath_aliased.append(fpath_best_error)
io.link(fpath, *fpath_aliased)
logger.info('Model at epoch {} dumped to {}.\n(Also alias: {}).'.format(
trainer.epoch, fpath, ', '.join(fpath_aliased)))
def post_preference(self, uid, pref):
dirname = _compose_dir(uid)
pair = io.load(osp.join(dirname, 'pair.pkl'))
io.dump(osp.join(dirname, 'pref.txt'), str(pref))
logger.info('Post preference uid={}, pref={}.'.format(uid, pref))
data = TrainingData(pair.t1_state, pair.t1_action, pair.t2_state, pair.t2_action, pref)
# Lock acquired inside this function call.
self._rpredictor.add_training_data(data)
def _process(self, pair):
uid = uuid.uuid4().hex
dirname = _compose_dir(uid)
io.mkdir(dirname)
# dump the file for displaying
if self._tformat == 'GIF':
_save_gif(pair.t1_observation, osp.join(dirname, '1.gif'))
_save_gif(pair.t2_observation, osp.join(dirname, '2.gif'))
else:
raise ValueError('Unknown trajectory format: {}'.format(self._tformat))
# cleanup
pair = TrajectoryPair(pair.t1_state, None, pair.t1_action, pair.t2_state, None, pair.t2_action)
# dump the raw pair
io.dump(osp.join(dirname, 'pair.pkl'), pair)
return uid, pair
def dump_weights_file(env, fpath):
fpath = io.assert_extension(fpath, __weights_ext__)
weights = env.network.fetch_all_variables_dict()
io.dump(fpath, weights)
return fpath