Hi,
I’ve been trying to build my own scorer, but all the information I found on the internet was suitable for version 0.6.0 and unfortunately the scripts are not compatible with my version. Also, I found some details about building my own scorer on this webpage: https://deepspeech.readthedocs.io/en/v0.8.0/Scorer.html
The main idea is that I do not really know what to write in this script:
python3 generate_lm.py --input_txt librispeech-lm-norm.txt.gz --output_dir .
–top_k 500000 --kenlm_bins path/to/kenlm/build/bin/
–arpa_order 5 --max_arpa_memory “85%” --arpa_prune “0|0|1”
–binary_a_bits 255 --binary_q_bits 8 --binary_type trie
Thank you !!!