Using convert_tacotron2_torch_to_tf.py on Vanilla Tacotron2 returns an error

[KakoozaJerry]

I run this command

!python /content/TTS/TTS/bin/convert_tacotron2_torch_to_tf.py --torch_model_path /content/drive/MyDrive/Emergent/train/Day23/Vanilla_Tacotron2-February-07-2022_10+46AM-0000000/checkpoint_52000.pth.tar --config_path /content/drive/MyDrive/Emergent/train/Day23/Vanilla_Tacotron2-February-07-2022_10+46AM-0000000/config.json --output_path /content/

And I get the error below

 > Using model: tacotron2
(1, None, 80)
(1, None, 80)
[('embedding/embeddings:0', 'embedding.weight'),
 ('encoder/lstm/forward_lstm/lstm_cell_1/kernel:0',
  'encoder.lstm.weight_ih_l0'),
 ('encoder/lstm/forward_lstm/lstm_cell_1/recurrent_kernel:0',
  'encoder.lstm.weight_hh_l0'),
 ('encoder/lstm/backward_lstm/lstm_cell_2/kernel:0',
  'encoder.lstm.weight_ih_l0_reverse'),
 ('encoder/lstm/backward_lstm/lstm_cell_2/recurrent_kernel:0',
  'encoder.lstm.weight_hh_l0_reverse'),
 ('encoder/lstm/forward_lstm/lstm_cell_1/bias:0',
  ('encoder.lstm.bias_ih_l0', 'encoder.lstm.bias_hh_l0')),
 ('encoder/lstm/backward_lstm/lstm_cell_2/bias:0',
  ('encoder.lstm.bias_ih_l0_reverse', 'encoder.lstm.bias_hh_l0_reverse')),
 ('attention/v/kernel:0', 'decoder.attention.v.linear_layer.weight'),
 ('decoder/linear_projection/kernel:0',
  'decoder.linear_projection.linear_layer.weight'),
 ('decoder/stopnet/kernel:0', 'decoder.stopnet.1.linear_layer.weight'),
 ('encoder/convolutions_0/convolution1d/kernel:0',
  'encoder.convolutions.0.convolution1d.weight'),
 ('encoder/convolutions_0/convolution1d/bias:0',
  'encoder.convolutions.0.convolution1d.bias'),
 ('encoder/convolutions_0/batch_normalization/gamma:0',
  'encoder.convolutions.0.batch_normalization.weight'),
 ('encoder/convolutions_0/batch_normalization/beta:0',
  'encoder.convolutions.0.batch_normalization.bias'),
 ('encoder/convolutions_1/convolution1d/kernel:0',
  'encoder.convolutions.1.convolution1d.weight'),
 ('encoder/convolutions_1/convolution1d/bias:0',
  'encoder.convolutions.1.convolution1d.bias'),
 ('encoder/convolutions_1/batch_normalization/gamma:0',
  'encoder.convolutions.1.batch_normalization.weight'),
 ('encoder/convolutions_1/batch_normalization/beta:0',
  'encoder.convolutions.1.batch_normalization.bias'),
 ('encoder/convolutions_2/convolution1d/kernel:0',
  'encoder.convolutions.2.convolution1d.weight'),
 ('encoder/convolutions_2/convolution1d/bias:0',
  'encoder.convolutions.2.convolution1d.bias'),
 ('encoder/convolutions_2/batch_normalization/gamma:0',
  'encoder.convolutions.2.batch_normalization.weight'),
 ('encoder/convolutions_2/batch_normalization/beta:0',
  'encoder.convolutions.2.batch_normalization.bias'),
 ('encoder/convolutions_0/batch_normalization/moving_mean:0',
  'encoder.convolutions.0.batch_normalization.running_mean'),
 ('encoder/convolutions_0/batch_normalization/moving_variance:0',
  'encoder.convolutions.0.batch_normalization.running_var'),
 ('encoder/convolutions_1/batch_normalization/moving_mean:0',
  'encoder.convolutions.1.batch_normalization.running_mean'),
 ('encoder/convolutions_1/batch_normalization/moving_variance:0',
  'encoder.convolutions.1.batch_normalization.running_var'),
 ('encoder/convolutions_2/batch_normalization/moving_mean:0',
  'encoder.convolutions.2.batch_normalization.running_mean'),
 ('encoder/convolutions_2/batch_normalization/moving_variance:0',
  'encoder.convolutions.2.batch_normalization.running_var'),
 ('decoder/while/prenet/linear_layer_0/linear_layer/kernel:0',
  'decoder.prenet.linear_layers.0.linear_layer.weight'),
 ('decoder/while/prenet/linear_layer_1/linear_layer/kernel:0',
  'decoder.prenet.linear_layers.1.linear_layer.weight'),
 ('decoder/while/attention_rnn/kernel:0', 'decoder.attention_rnn.weight_ih'),
 ('decoder/while/attention_rnn/recurrent_kernel:0',
  'decoder.attention_rnn.weight_hh'),
 ('decoder/while/attention_rnn/bias:0', 'decoder.attention_rnn.bias_ih'),
 ('decoder/while/attention/query_layer/linear_layer/kernel:0',
  'decoder.attention.query_layer.linear_layer.weight'),
 ('decoder/attention/inputs_layer/linear_layer/kernel:0',
  'decoder.attention.inputs_layer.linear_layer.weight'),
 ('decoder/while/attention/v/linear_layer/kernel:0',
  'decoder.attention.v.linear_layer.weight'),
 ('decoder/while/attention/v/linear_layer/bias:0',
  'decoder.attention.v.linear_layer.bias'),
 ('decoder/while/attention/location_layer/location_conv1d/kernel:0',
  'decoder.attention.location_layer.location_conv1d.weight'),
 ('decoder/while/attention/location_layer/location_dense/kernel:0',
  'decoder.attention.location_layer.location_dense.linear_layer.weight'),
 ('decoder/while/decoder_rnn/kernel:0', 'decoder.decoder_rnn.weight_ih'),
 ('decoder/while/decoder_rnn/recurrent_kernel:0',
  'decoder.decoder_rnn.weight_hh'),
 ('decoder/while/decoder_rnn/bias:0', 'decoder.decoder_rnn.bias_ih'),
 ('decoder/while/linear_projection/linear_layer/kernel:0',
  'decoder.linear_projection.linear_layer.weight'),
 ('decoder/while/linear_projection/linear_layer/bias:0',
  'decoder.linear_projection.linear_layer.bias'),
 ('decoder/while/stopnet/linear_layer/kernel:0',
  'decoder.stopnet.1.linear_layer.weight'),
 ('decoder/while/stopnet/linear_layer/bias:0',
  'decoder.stopnet.1.linear_layer.bias'),
 ('postnet/convolutions_0/convolution1d/kernel:0',
  'postnet.convolutions.0.convolution1d.weight'),
 ('postnet/convolutions_0/convolution1d/bias:0',
  'postnet.convolutions.0.convolution1d.bias'),
 ('postnet/convolutions_0/batch_normalization/gamma:0',
  'postnet.convolutions.0.batch_normalization.weight'),
 ('postnet/convolutions_0/batch_normalization/beta:0',
  'postnet.convolutions.0.batch_normalization.bias'),
 ('postnet/convolutions_1/convolution1d/kernel:0',
  'postnet.convolutions.1.convolution1d.weight'),
 ('postnet/convolutions_1/convolution1d/bias:0',
  'postnet.convolutions.1.convolution1d.bias'),
 ('postnet/convolutions_1/batch_normalization/gamma:0',
  'postnet.convolutions.1.batch_normalization.weight'),
 ('postnet/convolutions_1/batch_normalization/beta:0',
  'postnet.convolutions.1.batch_normalization.bias'),
 ('postnet/convolutions_2/convolution1d/kernel:0',
  'postnet.convolutions.2.convolution1d.weight'),
 ('postnet/convolutions_2/convolution1d/bias:0',
  'postnet.convolutions.2.convolution1d.bias'),
 ('postnet/convolutions_2/batch_normalization/gamma:0',
  'postnet.convolutions.2.batch_normalization.weight'),
 ('postnet/convolutions_2/batch_normalization/beta:0',
  'postnet.convolutions.2.batch_normalization.bias'),
 ('postnet/convolutions_3/convolution1d/kernel:0',
  'postnet.convolutions.3.convolution1d.weight'),
 ('postnet/convolutions_3/convolution1d/bias:0',
  'postnet.convolutions.3.convolution1d.bias'),
 ('postnet/convolutions_3/batch_normalization/gamma:0',
  'postnet.convolutions.3.batch_normalization.weight'),
 ('postnet/convolutions_3/batch_normalization/beta:0',
  'postnet.convolutions.3.batch_normalization.bias'),
 ('postnet/convolutions_4/convolution1d/kernel:0',
  'postnet.convolutions.4.convolution1d.weight'),
 ('postnet/convolutions_4/convolution1d/bias:0',
  'postnet.convolutions.4.convolution1d.bias'),
 ('postnet/convolutions_4/batch_normalization/gamma:0',
  'postnet.convolutions.4.batch_normalization.weight'),
 ('postnet/convolutions_4/batch_normalization/beta:0',
  'postnet.convolutions.4.batch_normalization.bias'),
 ('postnet/convolutions_0/batch_normalization/moving_mean:0',
  'postnet.convolutions.0.batch_normalization.running_mean'),
 ('postnet/convolutions_0/batch_normalization/moving_variance:0',
  'postnet.convolutions.0.batch_normalization.running_var'),
 ('postnet/convolutions_1/batch_normalization/moving_mean:0',
  'postnet.convolutions.1.batch_normalization.running_mean'),
 ('postnet/convolutions_1/batch_normalization/moving_variance:0',
  'postnet.convolutions.1.batch_normalization.running_var'),
 ('postnet/convolutions_2/batch_normalization/moving_mean:0',
  'postnet.convolutions.2.batch_normalization.running_mean'),
 ('postnet/convolutions_2/batch_normalization/moving_variance:0',
  'postnet.convolutions.2.batch_normalization.running_var'),
 ('postnet/convolutions_3/batch_normalization/moving_mean:0',
  'postnet.convolutions.3.batch_normalization.running_mean'),
 ('postnet/convolutions_3/batch_normalization/moving_variance:0',
  'postnet.convolutions.3.batch_normalization.running_var'),
 ('postnet/convolutions_4/batch_normalization/moving_mean:0',
  'postnet.convolutions.4.batch_normalization.running_mean'),
 ('postnet/convolutions_4/batch_normalization/moving_variance:0',
  'postnet.convolutions.4.batch_normalization.running_var')]
['embedding.weight',
 'encoder.convolutions.0.batch_normalization.num_batches_tracked',
 'encoder.convolutions.1.batch_normalization.num_batches_tracked',
 'encoder.convolutions.2.batch_normalization.num_batches_tracked',
 'encoder.lstm.weight_ih_l0',
 'encoder.lstm.weight_hh_l0',
 'encoder.lstm.bias_ih_l0',
 'encoder.lstm.bias_hh_l0',
 'encoder.lstm.weight_ih_l0_reverse',
 'encoder.lstm.weight_hh_l0_reverse',
 'encoder.lstm.bias_ih_l0_reverse',
 'encoder.lstm.bias_hh_l0_reverse',
 'decoder.attention_rnn.bias_hh',
 'decoder.decoder_rnn.bias_hh',
 'postnet.convolutions.0.batch_normalization.num_batches_tracked',
 'postnet.convolutions.1.batch_normalization.num_batches_tracked',
 'postnet.convolutions.2.batch_normalization.num_batches_tracked',
 'postnet.convolutions.3.batch_normalization.num_batches_tracked',
 'postnet.convolutions.4.batch_normalization.num_batches_tracked']
 > Passing weights from Torch to TF ...
 | > embedding/embeddings:0 <-- embedding.weight
Traceback (most recent call last):
  File "/content/TTS/TTS/bin/convert_tacotron2_torch_to_tf.py", line 106, in <module>
    tf_vars = transfer_weights_torch_to_tf(tf_vars, dict(var_map), state_dict)
  File "/usr/local/lib/python3.7/dist-packages/TTS-0.5.0-py3.7-linux-x86_64.egg/TTS/tts/tf/utils/convert_torch_to_tf_utils.py", line 79, in transfer_weights_torch_to_tf
    ), f" [!] weight shapes does not match: {tf_var.name} vs {torch_var_name} --> {tf_var.shape} vs {numpy_weight.shape}"
AssertionError:  [!] weight shapes does not match: embedding/embeddings:0 vs embedding.weight --> (66, 512) vs (67, 512)

What could be the cause of the error.

Below is my config.json file

{
    "model": "tacotron2",
    "run_name": "Vanilla_Tacotron2",
    "run_description": "",
    "epochs": 10000,
    "batch_size": 32,
    "eval_batch_size": 16,
    "mixed_precision": false,
    "scheduler_after_epoch": false,
    "run_eval": true,
    "test_delay_epochs": -1,
    "print_eval": true,
    "dashboard_logger": "tensorboard",
    "print_step": 25,
    "plot_step": 100,
    "model_param_stats": false,
    "project_name": null,
    "log_model_step": null,
    "wandb_entity": null,
    "save_step": 1000,
    "checkpoint": true,
    "keep_all_best": false,
    "keep_after": 10000,
    "num_loader_workers": 4,
    "num_eval_loader_workers": 4,
    "use_noise_augment": false,
    "use_language_weighted_sampler": false,
    "output_path": "/content/drive/MyDrive/Emergent/train/Day23/",
    "distributed_backend": "nccl",
    "distributed_url": "tcp://localhost:54321",
    "audio": {
        "fft_size": 1024,
        "win_length": 1024,
        "hop_length": 256,
        "frame_shift_ms": null,
        "frame_length_ms": null,
        "stft_pad_mode": "reflect",
        "sample_rate": 22050,
        "resample": false,
        "preemphasis": 0.98,
        "ref_level_db": 20,
        "do_sound_norm": false,
        "log_func": "np.log",
        "do_trim_silence": false,
        "trim_db": 60.0,
        "do_rms_norm": false,
        "db_level": null,
        "power": 1.5,
        "griffin_lim_iters": 60,
        "num_mels": 80,
        "mel_fmin": 0.0,
        "mel_fmax": 8000,
        "spec_gain": 1.0,
        "do_amp_to_db_linear": true,
        "do_amp_to_db_mel": true,
        "signal_norm": false,
        "min_level_db": -100,
        "symmetric_norm": true,
        "max_norm": 4.0,
        "clip_norm": true,
        "stats_path": null
    },
    "use_phonemes": false,
    "use_espeak_phonemes": true,
    "phoneme_language": "en-us",
    "compute_input_seq_cache": false,
    "text_cleaner": "basic_cleaners",
    "enable_eos_bos_chars": false,
    "test_sentences_file": "",
    "phoneme_cache_path": "/content/drive/MyDrive/Emergent/train/Day11_newData/phoneme_cache",
    "characters": {
        "pad": "_",
        "eos": "~",
        "bos": "^",
        "characters": "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\u014b!'(),-.:;? ",
        "punctuations": "!'(),-.:;? ",
        "phonemes": "iy\u0268\u0289\u026fu\u026a\u028f\u028ae\u00f8\u0258\u0259\u0275\u0264o\u025b\u0153\u025c\u025e\u028c\u0254\u00e6\u0250a\u0276\u0251\u0252\u1d7b\u0298\u0253\u01c0\u0257\u01c3\u0284\u01c2\u0260\u01c1\u029bpbtd\u0288\u0256c\u025fk\u0261q\u0262\u0294\u0274\u014b\u0272\u0273n\u0271m\u0299r\u0280\u2c71\u027e\u027d\u0278\u03b2fv\u03b8\u00f0sz\u0283\u0292\u0282\u0290\u00e7\u029dx\u0263\u03c7\u0281\u0127\u0295h\u0266\u026c\u026e\u028b\u0279\u027bj\u0270l\u026d\u028e\u029f\u02c8\u02cc\u02d0\u02d1\u028dw\u0265\u029c\u02a2\u02a1\u0255\u0291\u027a\u0267\u02b2\u025a\u02de\u026b",
        "unique": true
    },
    "batch_group_size": 0,
    "loss_masking": true,
    "sort_by_audio_len": false,
    "min_seq_len": 50,
    "max_seq_len": 500000,
    "compute_f0": false,
    "compute_linear_spec": false,
    "add_blank": false,
    "datasets": [
        {
            "name": "ljspeech",
            "path": "/content/data/",
            "meta_file_train": "metadata.csv",
            "ignored_speakers": null,
            "language": "",
            "meta_file_val": "",
            "meta_file_attn_mask": ""
        }
    ],
    "optimizer": "RAdam",
    "optimizer_params": {
        "betas": [
            0.9,
            0.998
        ],
        "weight_decay": 1e-06
    },
    "lr_scheduler": "NoamLR",
    "lr_scheduler_params": {
        "warmup_steps": 4000
    },
    "test_sentences": [
        "Abaaliwo bonna baasooka ne basiriikirira nga beewuunya omulanga oguvudde mu musajja ayogereza mwannyinaabwe.",
        "Wuuno eyakubye bbomu mu bbaasi.",
        "Engeri omupangisa gy\u2019atutte omugagga nabukeera mu kkomera.",
        "Pulezidenti awadde ekiragiro ku takisi.",
        "Aba bodaboda abasse ow\u2019emmotoka babalina."
    ],
    "use_gst": false,
    "gst": null,
    "gst_style_input": null,
    "num_speakers": 1,
    "num_chars": 67,
    "r": 2,
    "gradual_training": null,
    "memory_size": -1,
    "prenet_type": "original",
    "prenet_dropout": true,
    "prenet_dropout_at_inference": false,
    "stopnet": true,
    "separate_stopnet": true,
    "stopnet_pos_weight": 10.0,
    "max_decoder_steps": 2000,
    "encoder_in_features": 512,
    "decoder_in_features": 512,
    "decoder_output_dim": 80,
    "out_channels": 80,
    "attention_type": "original",
    "attention_heads": null,
    "attention_norm": "sigmoid",
    "attention_win": false,
    "windowing": false,
    "use_forward_attn": false,
    "forward_attn_mask": false,
    "transition_agent": false,
    "location_attn": true,
    "bidirectional_decoder": false,
    "double_decoder_consistency": false,
    "ddc_r": 6,
    "use_speaker_embedding": false,
    "speaker_embedding_dim": 512,
    "use_d_vector_file": false,
    "d_vector_file": false,
    "d_vector_dim": null,
    "lr": 0.0001,
    "grad_clip": 5.0,
    "seq_len_norm": false,
    "decoder_loss_alpha": 0.25,
    "postnet_loss_alpha": 0.25,
    "postnet_diff_spec_alpha": 0.25,
    "decoder_diff_spec_alpha": 0.25,
    "decoder_ssim_alpha": 0.25,
    "postnet_ssim_alpha": 0.25,
    "ga_alpha": 5.0
}

[erogol]

TF-related stuff is totally dumped in the dev branch for the next version. So this error will not be relatable soon.