Skip to content

Releases: k2-fsa/sherpa

doc

11 Apr 08:57
85200aa
Compare
Choose a tag to compare
add a mirror for the pdf doc (#572)

Release v1.2

10 Mar 13:54
cac92c8
Compare
Choose a tag to compare
  • Support pre-trained CTC models from NeMo

Release v1.0

04 Nov 08:27
078c709
Compare
Choose a tag to compare
Release v1.0 (#186)

* Release v1.0

* Fix building on windows

Release v0.9.1

19 Sep 22:48
Compare
Choose a tag to compare

Support LSTM transducer.

Release v0.8

15 Sep 16:26
be1a0ce
Compare
Choose a tag to compare

Support endpointing.

Release v0.7

21 Aug 04:07
f22cf4c
Compare
Choose a tag to compare

Add c++ frontend for non-streaming ASR.

See https://k2-fsa.github.io/sherpa/cpp/installation/index.html for installation.

See https://k2-fsa.github.io/sherpa/cpp/offline_asr/index.html for usage.

Note: It supports conda install on Linux/macOS/Windows.

Release v0.6

03 Jul 01:32
6e596dc
Compare
Choose a tag to compare

Bugfixes.

Release v0.5

11 Jun 01:41
01d9e17
Compare
Choose a tag to compare
  • Add @torch.no_grad() to other threads in addition to the main thread to reduce memory consumption since torch.no_grad() is thread-local

Release v0.4

07 Jun 05:05
522884e
Compare
Choose a tag to compare
  • Support torch >= 1.6.0
  • Support specifying the max number of clients that can be handled by the server

Release v0.3

04 Jun 11:02
aff9cd0
Compare
Choose a tag to compare
  1. Support streaming ASR with pruned stateless Emformer RNN-T
  2. Add checks for torch version during import time
  3. Support Windows