Update script for ngram rnnt and hat beam search decoding (#6370)

* add rnnt ngram beamsearch script

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* add return encoding embedding option

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* update script

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* add rnnt and hat ngram decoding script

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* add some parameters

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* add return_encoder_embeddings parameter to RNNTDecodingConfig

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* replace return_encoder_embeddings parameter

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* generalization of scipt behavior

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* remove return_encoder_embeddings parameter

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* remove return_encoder_embeddings parameter

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* add manual encoder_embeddings calculation

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* fix beam_width value to 8

Signed-off-by: Andrei Andrusenko <52885736+andrusenkoau@users.noreply.github.com>

* fix rescoring description

Signed-off-by: Andrei Andrusenko <52885736+andrusenkoau@users.noreply.github.com>

---------

Signed-off-by: andrusenkoau <andrusenkoau@gmail.com>
Signed-off-by: Andrei Andrusenko <52885736+andrusenkoau@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Somshubra Majumdar <titu1994@gmail.com>

docs/source/asr/asr_language_modeling.rst

@@ -281,6 +281,29 @@ For instance, the following set of parameters would results in 2*1*2=4 beam sear
                         beam_beta=[1.0,0.5]
 
 
+Beam search ngram decoding for Transducer models (RNNT and HAT)
+===============================================================
+
+The similar script to evaluate an RNNT/HAT model with beam search decoding and N-gram models can be found at
+`scripts/asr_language_modeling/ngram_lm/eval_beamsearch_ngram_transducer.py <https://github.com/NVIDIA/NeMo/blob/stable/scripts/asr_language_modeling/ngram_lm/eval_beamsearch_ngram_transducer.py>`_
+
+.. code-block::
+
+    python eval_beamsearch_ngram_transducer.py nemo_model_file=<path to the .nemo file of the model> \
+            input_manifest=<path to the evaluation JSON manifest file \
+            kenlm_model_file=<path to the binary KenLM model> \
+            beam_width=[<list of the beam widths, separated with commas>] \
+            beam_alpha=[<list of the beam alphas, separated with commas>] \
+            preds_output_folder=<optional folder to store the predictions> \
+            probs_cache_file=null \
+            decoding_strategy=<greedy_batch or maes decoding>
+            maes_prefix_alpha=[<list of the maes prefix alphas, separated with commas>] \
+            maes_expansion_gamma=[<list of the maes expansion gammas, separated with commas>] \
+            hat_subtract_ilm=<in case of HAT model: subtract internal LM or not (True/False)> \
+            hat_ilm_weight=[<in case of HAT model: list of the HAT internal LM weights, separated with commas>] \
+           
+
+
 .. _neural_rescoring:
 
 ****************

nemo/collections/asr/models/rnnt_models.py

@@ -242,6 +242,7 @@ def transcribe(
         """
         if paths2audio_files is None or len(paths2audio_files) == 0:
             return {}
+
         # We will store transcriptions here
         hypotheses = []
         all_hypotheses = []