Is it a possible to convert fastconformer CTC model to stt_xx_fastconformer_hybrid_large_streaming_multi.nemo?

[ican24]

Is your feature request related to a problem? Please describe.

I am trying to generate new cache_aware_streaming model from my fastconformer CTC model (developed with 25k hours dataset). Unfortunately all my efforts are failed.

Describe the solution you'd like

I tried asr_hybrid_transducer_ctc/speech_to_text_hybrid_rnnt_ctc_bpe.py script with next configuration and options

nohup python -u examples/asr/asr_hybrid_transducer_ctc/speech_to_text_hybrid_rnnt_ctc_bpe.py \
    --config-path="../conf/fastconformer/hybrid_cache_aware_streaming" \
    --config-name="fastconformer_hybrid_transducer_ctc_bpe_streaming" \
    model.train_ds.manifest_filepath=train \
    model.validation_ds.manifest_filepath=test \
    model.tokenizer.dir=datanew/tokenize/tokenizer_spe_bpe_v1024 \
    model.tokenizer.type="bpe" \
    trainer.max_epochs=100 \
    model.optim.name="adamw" \
    model.optim.weight_decay=0.0001 \
    model.optim.sched.warmup_steps=2000 \
    model.aux_ctc.ctc_loss_weight=0.3 \
    model.optim.lr=0.005 \
    model.optim.betas=[0.9,0.999] \
    +init_from_nemo_model="ASR-Char-Model-Language-hy.nemo" > _log.txt 2>&1 &

Also I had tried self-made scripts with different parameters. No any progress.

Is it a possible or not!

Thank you in advance, you may save a hug time.