Stack Overflow
Stack Overflow

Questions tagged [fairseq]

40 questions
4
votes
1 answer

Not able to execute sample code provided in Hugging faces Models card

When i am trying sample code from Hugging face i get below error. the code can be found from https://huggingface.co/facebook/tts_transformer-en-ljspeech Code: from fairseq.checkpoint_utils import load_model_ensemble_and_task_from_hf_hub from…
user2478236
  • 691
  • 12
  • 32
3
votes
1 answer

Fairseq Transform model not working (Float can't be cast to long)

I've installed python 3.8, pytorch 1.7, and fairseq 0.10.1, on a new machine, then copied in a script and model from a machine with python 3.6, pytorch 1.4 and fairseq 0.9.0, where it is working. The model is loaded and prepared with: model =…
Darren Cook
  • 27,837
  • 13
  • 117
  • 217
3
votes
2 answers

How to use fairseq interactive.py non-interactively?

I am trying to translate from English to Arabic using Fairseq. But the interactive.py script translate pieces of text fragment on-the-fly. But I need to use it as reading an input text file and writing output text file write. I referred this GitHub…
Ramraj Chandradevan
  • 141
  • 2
  • 10
3
votes
2 answers

install and run fairseq in google colab (ModuleNotFoundError: No module named 'fairseq')

Until yesterday, we installed fairseq normally and executed it. An error called "ModuleNotFoundError: No module named 'fairseq'" suddenly occurred today. Does anyone know how to solve this? install fairseq... !git clone…
JEONGHYEOK PARK
  • 31
  • 1
  • 2
2
votes
1 answer

How to inspect values in binarized FairSeq datasets?

Running the fairseq-preprocess script produces binary files with integer indices corresponding to token ids in a dictionary. When I no longer have the original tokenized texts, what is the simplest way to explore the binarized dataset? The…
Jindřich
  • 10,270
  • 2
  • 23
  • 44
2
votes
1 answer

Cannot find some fairseq parameters in docs

I am using fairseq (version: 1.0.0a0+14c5bd0) to fine-tune a model as per this link. However, there are lots of parameters used that I cannot find in the docs nor when I run fairseq-train --help. Examples…
M.A.G
  • 559
  • 2
  • 6
  • 21
2
votes
0 answers

Should we not expect wav2vec to outperform Microsoft STT

I compared the accuracy levels of both wav2vec and Microsoft STT on a few Ted talks. Accuracy levels are word error rates |Video |Wav2Vec|MicrosftSTT| |:---: |:-----:|:---------:| |[1][1]|8.57 |3.7 | |[2][2]|13.83 |5.8 …
Vishaal MK
  • 21
  • 1
2
votes
0 answers

How to log the graph from fairseq to tensorboard

I got the impression from reading tutorials that this would be easy: slap in a call to add_graph(). First tricky part was working out where to put it. I've gone with fairseq_cli/train.py at the very top of the enumerate() loop in train(): for i,…
Darren Cook
  • 27,837
  • 13
  • 117
  • 217
2
votes
1 answer

generating segment labels for a Tensor given a value indicating segment boundaries

Does anyone know of a way to generate a 'segment label' for a Tensor, given a unique value that represents segment boundaries within the Tensor? For example, given a 1D input tensor where the value 1 represents a segment boundary, x =…
tannonk
  • 23
  • 4
2
votes
1 answer

RPC failed with status = "Unavailable: Socket closed" Error when training FairSeq RoBERTa on Cloud TPU using PyTorch

I followed the tutorials "Pre-training FairSeq RoBERTa on Cloud TPU using Pytorch" to setup a Preemptible (v2-8) TPU env and train my RoBERTa model. The PyTorch env is based on torch-xla-1.6 as instructed by the document. However, it does not output…
user3786340
  • 190
  • 1
  • 8
1
vote
1 answer

What are the return values from fairseq WMT19 machine translation model's .generate() function?

I am trying to play around with the Fairseq machine translation model using en2de = torch.hub.load('pytorch/fairseq', 'transformer.wmt19.en-de', checkpoint_file='model1.pt:model2.pt:model3.pt:model4.pt', …
Deergh Singh
  • 11
  • 1
1
vote
1 answer

Early stopping based on BLEU in FairSeq

My goal is to use BLEU as early stopping metric while training a translation model in FairSeq. Following the documentation, I am adding the following arguments to my training script: --eval-bleu --eval-bleu-args --eval-bleu-detok…
Nikhil Jaiswal
  • 13
  • 3
1
vote
2 answers

How to select fairseq option `--ddp-backend`

I'm learning how to use fairseq to implement a simple translation model based on Transformer. I would like to use 2 GeForce RTX 3090 GPUs on my lab server. Which option should I select for --ddp-backend of fairseq-train? Furthermore, could you…
Shunchi Zhang
  • 49
  • 7
1
vote
1 answer

MT: Calculating METEOR Score of two different files

For an Machine Translation evaluation I need to calculate the METEOR score between the translation output file and the reference file. I already found this question How can I implement meteor score when evaluating a model when using the meteor_score…
Ramón Wilhelm
  • 967
  • 2
  • 18
  • 39
1
vote
1 answer

Train Wav2Vec-U for a custom dataset

I found the github repo of Wav2Vec-U, but it is not well-documented. I wonder if there is any train procedure for it. I'm trying to train it on common voice audios. But it needs wrd, ltr, and phn files, which I don't have access to.
Soroush
  • 83
  • 8
1
2 3