-
Good evening! I have 2 questions...
1. Is it possible to write your own dataloader and embed it in the architecture of the QuartzNet model for training?
It seems to me that I saw somewhere how the d…
-
Hi, I trained a conformer model before. Now I wanna quantize that model and convert it to TRT. But when I run the **speech_to_text_quant_infer_trt.py** an error occur. The error is:
`TypeError: Err…
-
```cs
builder.Services.AddQuartz(q =>
{
q.UsePersistentStore(storeOptions =>
{
storeOptions.UseRavenDb(options =>
{
var configuration = new ConfigurationBuil…
-
Code:
```
var timeOfDay = new TimeOfDay(11 - DateTimeOffset.UtcNow.Offset.Hours, 50);
var trigger = TriggerBuilder.Create()
.WithDailyTimeIntervalS…
-
Hi
I'm using the script https://github.com/NVIDIA/NeMo/blob/main/tutorials/speaker_tasks/ASR_with_SpeakerDiarization.ipynb in Colab and I came across a problem. When trying to transcribe an audio th…
-
I'm working on a [QuartzNet](https://arxiv.org/abs/1910.10261)-based model (includes depthwise convolutions) with Pytorch, and it is successfully deployed on `ZCU102` by modifying the [tutorial](https…
-
Hi, thanks for your priceless tutorial,
I know the audio sample rate for fine-tuning the Nemo models should be 16khz. Is the sample rate should be 16khz for the audios to transcribe with Nemo models?
-
Hi all,
I am using the PyTorch API to compile the Quartznet model and deploy the model on KV260. The compilation is successful but I found that there is too much accuracy loss when I ran the model w…
-
Hi
With a supposed Colab update yesterday, I have been having the following issue during training.
I say supposed update because I noticed changes in the Colab interface, but I didn't find any ver…
-
Hello!
I am trying to replicate Quartznet 5x5 results to get around 11/12% WER on the test set.
So far my last experiment was using 4 gpus, with batch 64, using speed perturbed data, specaugmenta…