-
Hello,
I am fine-tuning an ASR model (the English Conformer Transducer) and encountering the following error:
```log
2023-02-03 14:46:28 | numba.cuda.cudadrv.driver.CudaAPIError: [1] Call to cu…
-
What does it look like to send data to another transducer? How can it be done in an ergonomic way? Currently it's done relatively externally and then fed into channels which feed into a hydroflow prog…
zzlk updated
7 months ago
-
**Describe the bug**
As shown in
(https://github.com/NVIDIA/NeMo/blob/main/nemo/collections/asr/metrics/wer.py#L349-L350), the new scores and words will be assigned to the object and the pervious sc…
-
2023-10-03 14:09:50,632 INFO [decode.py:917] Calculating the averaged model over epoch range from 10 (excluded) to 20
Traceback (most recent call last):
File "./pruned_transducer_stateless7/deco…
-
Hi, i'm trying to train conformer_ctc on timit and training seems to work, but during the decoding with ctc-decoding method during get_lattice() call in decode_one_batch() function whole process hangs…
-
Hi,
i want to know how to train a phone-based RNNT in icefall.
Thanks
-
I'm trying to use the HLG as the decoding graph for conv_emformer_transducer_stateless2. In the streaming_decode.py I have:
hlg_filename = "data/lang_bpe_500/HLG.pt"
decoding_graph = k2.Fsa.from_dic…
-
**Describe the bug 🐞**
when I use the default alg, I get a long error. all of my variables are `Float32` and the problem appeared recently (two weeks or so).
**Minimal Reproducible Example 👇**
…
-
[log_prepare_stage14_maxDur500_1GPU_01.txt](https://github.com/k2-fsa/icefall/files/9635055/log_prepare_stage14_maxDur500_1GPU_01.txt)
[train.py.txt](https://github.com/k2-fsa/icefall/files/9635057…
-
In the pruned transducer recipes which use auxiliary CTC loss, such as the following:
https://github.com/k2-fsa/icefall/blob/45c13e90e42d0f6ff190d69acb18f4e868bfa954/egs/librispeech/ASR/pruned_tran…