Merge pull request #63 from huggingface/add_parakeet-v2

Add parakeet v2

Author: nithinraok

nemo_asr/run_fast_conformer_rnnt.sh

@@ -2,9 +2,11 @@
 
 export PYTHONPATH="..":$PYTHONPATH
 
-#considering FC-L, FC-XL, FC-XXL, C-L and C-S RNNT models
-MODEL_IDs=("nvidia/parakeet-tdt-1.1b" "nvidia/parakeet-rnnt-1.1b" "nvidia/parakeet-rnnt-0.6b" "nvidia/stt_en_fastconformer_transducer_large" "nvidia/stt_en_conformer_transducer_large" "stt_en_conformer_transducer_small")
-BATCH_SIZE=64
+#considering latest model
+MODEL_IDs=("nvidia/parakeet-tdt-0.6b-v2")
+# For previous parakeet models: FC-L, FC-XL, FC-XXL, C-L and C-S RNNT models
+#  ("nvidia/parakeet-tdt-1.1b" "nvidia/parakeet-rnnt-1.1b" "nvidia/parakeet-rnnt-0.6b" "nvidia/stt_en_fastconformer_transducer_large" "nvidia/stt_en_conformer_transducer_large" "stt_en_conformer_transducer_small")
+BATCH_SIZE=128
 DEVICE_ID=0
 
 num_models=${#MODEL_IDs[@]}

requirements/requirements_nemo.txt

@@ -1,6 +1,6 @@
-git+https://github.com/NVIDIA/NeMo.git@208e0da28e2ada8da84d8f7ddff8623efe1ff01c#egg=nemo_toolkit[asr] # This commit hash is a recent version of main at the time of testing.
+git+https://github.com/NVIDIA/NeMo.git@r2.3.0#egg=nemo_toolkit[asr]
 tqdm
 soundfile
 librosa
 IPython # Workaround for https://github.com/NVIDIA/NeMo/pull/9890#discussion_r1701028427
-cuda-python>=12.4 # Used for fast TDT and RNN-T inference
+cuda-python>=12.4 # Used for fast TDT and RNN-T inference