Saving wav2vec2 model results in serialization error

Hi,

I am trying to save the compiled model using the code below:

w2v = torch.load(model_path)
model = Wav2VecCtc.build_model(args, target_dict)
model.load_state_dict(w2v["model"], strict=True)
torch.save(model, 'test_again.pt' )

But whenever I execute this code, the model is built successfully but it throws an error:

This error I did not encounter in non hydra version of fairseq. Can you please let me know what can be done??

Thanks!

[minghao-wu]

@villmow @harveenchadha Hi there,

Thank you very much for your solution, but it still doesn’t work on my code.

I insert your snippt into line 460 at fairseq/fairseq/dataclass/utils.py and my function look like this now:

def merge_with_parent(dc: FairseqDataclass, cfg: FairseqDataclass):
    cfg = OmegaConf.merge(
        OmegaConf.structured(
            dc
        ),
        OmegaConf.create(
            OmegaConf.to_yaml(cfg, resolve=True)
        )
    )
    merged_cfg = OmegaConf.merge(dc, cfg)
    merged_cfg.__dict__["_parent"] = cfg.__dict__["_parent"]
    OmegaConf.set_struct(merged_cfg, True)
    return merged_cfg

Is this the right way to use your snippt?

I am training my custom task using multiple GPUs and encountered a similar error.

More details can be found at https://github.com/pytorch/fairseq/issues/3634

Thanks again.

[villmow]

I receive the same error when training a model with fairseq-hydra-train on multiple GPUs. Single GPU works.