Tensor shapes mismatch during running lm optimizer py

[JRMeyer]

>>> sangjeedondrub
[December 10, 2020, 5:43am]

The error is like:

(np_val.shape, subfeed_t.name, str(subfeed_t.get_shape())))
ValueError: Cannot feed value of shape (400,) for Tensor 'cudnn_lstm/rnn/multi_rnn_cell/cell_0/cudnn_compatible_lstm_cell/bias/Initializer/Const:0', which has shape '(8192,)'e

Here is how I train the models:

# Train kenlm

python $DSPEECH/data/lm/generate_lm.py slash
--input_txt data/text.txt --output_dir models/ slash
--top_k 500000 --kenlm_bins $DSPEECH/kenlm/build/bin/ slash
--arpa_order 5 --max_arpa_memory '85%' --arpa_prune '0|0|1' slash
--binary_a_bits 255 --binary_q_bits 8 --binary_type trie

# Train STT model

python $DSPEECH/DeepSpeech.py slash
--alphabet_config_path data/alphabet.txt slash
--train_files data/train.csv slash
--dev_files data/dev.csv slash
--test_files data/test.csv slash
--audio_sample_rate 16000 slash
--epochs 300 slash
--train_batch_size 12 slash
--test_batch_size 12 slash
--n_hidden 2048 slash
--checkpoint_dir checkpoints/1st-run/ slash
--export_dir models/1st-run/ slash
--scorer models/kenlm.scorer slash
--report_count 20 slash
--log_level 0
'$'

# Optimize LM

python $DSPEECH/lm_optimizer.py slash
--test_files data/dev.csv slash
--alphabet_config_path data/alphabet.txt slash
--scorer_path models/kenlm.scorer slash
--checkpoint_dir checkpoints/1st-run slash
--log_level 0

This is the full log produced by running lm_optimizer.py

log.txt

Please HELP, thank you!

[This is an archived TTS discussion thread from discourse.mozilla.org/t/tensor-shapes-mismatch-during-running-lm-optimizer-py]

[JRMeyer]

>>> lissyx
[December 10, 2020, 8:39am]

> python $DSPEECH/lm_optimizer.py slash
> --test_files data/dev.csv slash
> --alphabet_config_path data/alphabet.txt slash
> --scorer_path models/kenlm.scorer slash
> --checkpoint_dir checkpoints/1st-run slash
> --log_level 0

you pass no lm_alpha_max, no lm_beta_max, no model shape n_hidden and no
n_trials.

> Please HELP , thank you!

please follow the guidelines and share more infos on what you do, and
read the docs.

> ValueError: Cannot feed value of shape (400,) for Tensor 'cudnn_lstm/rnn/multi_rnn_cell/cell_0/cudnn_compatible_lstm_cell/bias/Initializer/Const:0', which has shape '(8192,)'e

400=4*100 ; 8192=4*2048 are you certain you are running on the
checkpoint you think you are?

[Archived Post]

[JRMeyer]

>>> sangjeedondrub
[December 11, 2020, 2:43am]

. I passed a wrong
checkpoints.

[Archived Post]