-
Notifications
You must be signed in to change notification settings - Fork 8
/
Copy pathevaluate_asr.sh
37 lines (32 loc) · 2.3 KB
/
evaluate_asr.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
# hifphg
mkdir output_asr
mkdir output_asr/hifphg
python3 -m TTS.bin.synthesize --text "/nlsasfs/home/ai4bharat/praveens/ttsteam/datasets/indictts/hi/metadata_test_30.csv" \
--model_path /nlsasfs/home/ai4bharat/praveens/ttsteam/repos/text2speech-asr/output/hi_fastpitch_indictts_all_8_ae:0_ema:False_pretrained_asrloss0.5-October-19-2022_11+47PM-08fc70d/best_model.pth \
--config_path /nlsasfs/home/ai4bharat/praveens/ttsteam/repos/text2speech-asr/output/hi_fastpitch_indictts_all_8_ae:0_ema:False_pretrained_asrloss0.5-October-19-2022_11+47PM-08fc70d/config_v2.json \
--vocoder_path /nlsasfs/home/ai4bharat/praveens/ttsteam/repos/text2speech-ta/saved_models/hifigan/v1/hi/best_model.pth \
--vocoder_config_path /nlsasfs/home/ai4bharat/praveens/ttsteam/repos/text2speech-ta/saved_models/hifigan/v1/hi/config.json \
--out_path output_asr/hifphg \
--use_cuda t
python3 scripts/evaluate_mcd.py \
output_asr/hifphg \
/nlsasfs/home/ai4bharat/praveens/ttsteam/datasets/indictts/hi/wavs-22k/
python3 scripts/evaluate_f0.py \
output_asr/hifphg \
/nlsasfs/home/ai4bharat/praveens/ttsteam/datasets/indictts/hi/wavs-22k/
# # tafphg
# mkdir output_asr
# mkdir output_asr/tafphg
# python3 -m TTS.bin.synthesize --text "/nlsasfs/home/ai4bharat/praveens/ttsteam/datasets/indictts/ta/metadata_test_30.csv" \
# --model_path /nlsasfs/home/ai4bharat/praveens/ttsteam/repos/text2speech-asr/output_ta/output_ta/ta_fastpitch_indictts_all_8_ae:0_ema:False_pretrained_asrloss0.5-October-19-2022_11+51PM-08fc70d/best_model.pth \
# --config_path /nlsasfs/home/ai4bharat/praveens/ttsteam/repos/text2speech-asr/output_ta/output_ta/ta_fastpitch_indictts_all_8_ae:0_ema:False_pretrained_asrloss0.5-October-19-2022_11+51PM-08fc70d/config.json \
# --vocoder_path /nlsasfs/home/ai4bharat/praveens/ttsteam/repos/text2speech-ta/saved_models/hifigan/v1/ta/best_model.pth \
# --vocoder_config_path /nlsasfs/home/ai4bharat/praveens/ttsteam/repos/text2speech-ta/saved_models/hifigan/v1/ta/config.json \
# --out_path output_asr/tafphg \
# --use_cuda t
# python3 scripts/evaluate_mcd.py \
# output_asr/tafphg \
# /nlsasfs/home/ai4bharat/praveens/ttsteam/datasets/indictts/ta/wavs-22k/
# python3 scripts/evaluate_f0.py \
# output_asr/tafphg \
# /nlsasfs/home/ai4bharat/praveens/ttsteam/datasets/indictts/ta/wavs-22k/