 Kaushik Ram Sadagopan
|
521a374213
Online feature extractor SimulEval agent. (#107)
|
2 years ago |
 Ruslan Mavlyutov
|
ca1ebf90ea
* Training recipees for M4T-nano/-micro\n* Adjustments to fairseq2/sc updates\n* Fixing MyPy warnings (#59)
|
2 years ago |
 Can Balioglu
|
00b066c6f8
Fix breaking fairseq2 API changes (#119)
|
2 years ago |
 Yilin Yang
|
0cc7bc610a
Add integrated test for ProsodyUnitY/seamless_expressivity model (#99)
|
2 years ago |
|
Can Balioglu
|
0bdc7b60ac
Revise, clean up MinTox implementation. Part 1 (#96)
|
2 years ago |
|
Can Balioglu
|
2393016090
Move to generic loaders (#111)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
5198e0586c
Add seamless_streaming assets. (#106)
|
2 years ago |
|
Yilin Yang
|
1a91d39931
Rename gcmvn_fbank to prosody_encoder_input (#105)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
e3c40244e1
Fix bug in eval script for S2ST, T2ST tasks. (#102)
|
2 years ago |
|
Yilin Yang
|
ed18e69190
Implement PretsselModel & its inference (#89)
|
2 years ago |
|
Can Balioglu
|
29935570cd
Start using fairseq2's NGramRepeatBlockProcessor (#100)
|
2 years ago |
|
Can Balioglu
|
a618cd43f0
Update Shaw attention init (#101)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
4e93254fa5
Skip loading text_encoder for S2X tasks, skip loading T2U model for X2T tasks. (#95)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
fcaf953981
Clean up M4T v2 and vocoder v2 checkpoints. (#94)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
d6ff19b20b
Introduce monotonic_decoder. (#73)
|
2 years ago |
 Changhan Wang
|
1cead95b37
Add PRETSSEL HiFiGAN Vocoder (#78)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
a11376477b
Add use_text_decoder parameter to UnitYConfig. (#88)
|
2 years ago |
|
Can Balioglu
|
05419775be
Introduce Prosody encoder (#87)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
bfcdf3ba4f
Skips loading the vocoder for X2T eval. (#84)
|
2 years ago |
|
Can Balioglu
|
9daf4692ef
Revise project structure (#82)
|
2 years ago |
 spopuri
|
a1646750f1
Add evaluation pipeline to compute all metrics (#79)
|
2 years ago |
|
Can Balioglu
|
74364ff4db
Derive UnitYT2UModel from EncoderDecoderModel (#77)
|
2 years ago |
|
Can Balioglu
|
e83a4de3af
Start using vocab_info instead of pad_idx (#76)
|
2 years ago |
|
Can Balioglu
|
04d9a49462
Update LogitsProcessor to StepProcessor (#74)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
45974cd2f4
Disintegrate UnitYNART2UModel's decoder and decoder_layer from TransformerDecoder, TransformerDecoderLayer. (#71)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
4d679cb67e
Separate T2U and NAR T2U builders for UnitY T2U model. (#70)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
c60f92c86f
Rename 'commercial' to 'v2' and condition duration prediction on T2U model type for vocoder. (#72)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
f4dffda0f8
Translator integration tests and unity unittests. (#63)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
3b34f6defa
Don't mandate audio_root_dir for text input eval. (#69)
|
2 years ago |
|
Kaushik Ram Sadagopan
|
eba940ace3
m4t evaluation script (#57)
|
2 years ago |
