-
Hello All,
we are training a zip former model for about 3400 hours of Tamil data.
We were facing this issue:
RuntimeError:
grad_scale is too small, exiting: 1.4901161193847656e-08
…
-
Hi I tested timit data using eesen, but the result is not good as follows:
# training process
```
EPOCH 11 RUNNING ... ENDS [2016-Jun-6 17:02:47]: lrate 4e-05, TRAIN ACCURACY 23.4300%, VALID ACCURACY…
-
NOTE: See also the [tracker](https://github.com/pytorch/pytorch/issues/75432) for forward-over-reverse formulas: Sometimes implementing a forward AD formula will also implicitly support its forward-o…
-
### Environment
* **Tesseract Version**: tesseract 5.0.0-alpha-20210401-123-g5eb2e8
* **Commit Number**:
* **Platform**: either linux ubuntu 18.4
### Current Behavior:
Training command :…
-
https://arxiv.org/abs/1801.01671
-
Dear all:
Recently I tries to adopt LF-MMI in training AISHELL dataset inside model "tdnn-lstm", and I found that validation error is always infinity, while training-loss is perfectly normal.
In…
-
Thanks for a great toolkit!
I am working on ESPNET2 Streaming Conformer model (based on espnet2 Librispeech setup) adapted to my own data set and get very good results for non-streaming model. For…
-
Hello,
I am trying to run the TensorFlow based EESEN setup for Switchboard. More specifically, I am using the `tf_clean` branch and trying to run the `asr_egs/swbd/v1-tf/run_ctc_char.sh` script. I am…
-
Can u tell me exact steps to train the model?
with all the datasets and upto what extent it should be trained along with learning rates and all...plz help me put brother
-
非常感谢楼主的分享，我运行测试了一下，测试结果却是：训练accuracy一直是0，而且decoded出来只有一个字符？ 请问该哪里出问题了，怎么调整？
seq 0: origin: [52, 23, 35, 62, 24, 33] decoded:[26]
seq 1: origin: [54, 49, 2, 40, 26, 38] decoded:[26]
seq 2: ori…
rolai updated
6 years ago