Rajeev
Rajeev
there are 4 GPU on my device as it can be seen from this command ``` >>nvidia-smi -L GPU 0: GeForce GTX 1080 Ti (UUID: GPU-c249ebc7-3e15-08ee-4d23-2f6dfcd5d97a) GPU 1: GeForce GTX...
getting below error while running test.py when **.arpa l**m file of size 2.5GB is used Loading the LM will be faster if you build a binary file. Reading /home/rajeev/Documents/agent-lm/agent_lm_44_updated.arpa ----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100...
I have trained a transformer encoder-decoder model by replacing the encoder with some pre-trained model and putting decoder-related code (Tutorial 6 Attention is all you need) on top of the...
I can understand masking is necessary for decoder during training to hide future tokens from decoder but why do we need masking during inference in decoder ?
Traceback (most recent call last): File "./lib/python/VAD_test.py", line 60, in pred, label = graph_test.do_test(graph_list[-1], data_dir, norm_dir, data_len, is_default, mode) File "/home/rajeev/pb/VAD/lib/python/graph_test.py", line 89, in do_test valid_inputs, valid_labels = valid_data_set.next_batch(valid_batch_size) File...
## 🐛 Bug ### To Reproduce Steps to reproduce the behavior (**always include the command you ran**): #### 1. Run cmd : python fairseq-hydra-train task.data=/mnt/rajeev/speech/finetuning/750_hour model.w2v_path=/mnt/rajeev/speech/checkpoints/pretraining/hindi_pretrained_4kh.pt distributed_training.distributed_world_size=4 optimization.update_freq=[6] common.tensorboard_logdir=/mnt/rajeev/speech/logs/finetuning/750_hour/tensorboard_2021-07-07_15-48-19 checkpoint.save_dir=/mnt/rajeev/speech/checkpoints/finetuning/750_hour...
Please add the Squeezeformer model architecture for ASR as it is very efficient compared to Conformer.
Please add multi gpu training support to the code
Hi have exported the espnet model trained on my custom dataset using espnet_onnx. Model fails to work properly on some audios. Below is the error which i am getting ```...