Summary: 8 instances, 8 unique Text Count mlen: TODO Lysandre didn't fill 1 # TODO: This is not implemented 1 pred_mask[0] = 0 # TODO: remove 1 encoder = TransformerModel(params, dico, is_encoder=True, with_output=False) # TODO: only output when necessary - len(params.clm_steps + params.mlm_steps) > 0 1 # TODO: add extra layer norm here? 1 if False: # AMP checkpoint reloading is buggy, we cannot do that - TODO: fix - https://github.com/NVIDIA/apex/issues/250 1 qlen: TODO Lysandre didn't fill 1 # TODO: make sure we are using `xlm-mlm-enro-1024`, since XLM-100 doesn't have this step 1