mm 10 pre-trained model cannot be restored properly

[hanyangii]

When running methylbert with -p hanyangii/methylbert_mm10_4l option

Restore the pretrained model hanyangii/methylbert_mm10_4l
Cross entropy loss assigned
Some weights of MethylBertEmbeddedDMR were not initialized from the model checkpoint at hanyangii/methylbert_mm10_4l and are newly initialized: ['dmr_encoder.0.weight', 'read_classifier.0.bias', 'read_classifier.0.weight', 'read_classifier.3.bias', 'read_classifier.3.weight', 'read_classifier.4.bias', 'read_classifier.4.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
Total Parameters: 46941186
MethylBertEmbeddedDMR(
  (classification_loss_fct): CrossEntropyLoss()
  (bert): BertModel(
    (embeddings): BertEmbeddings(
      (word_embeddings): Embedding(69, 768, padding_idx=0)
      (position_embeddings): Embedding(512, 768)
      (token_type_embeddings): Embedding(2, 768)
      (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
      (dropout): Dropout(p=0.1, inplace=False)
    )
    (enco

token type embeddings must be (3, 768)