fairseq

fairseq copied to clipboard

checkpoint_wrapper at the first block disabling gradients

1

## 🐛 Bug If I use checkpoint_wrapper at the first block, the gradients of their parameters are None after backward. ### Code sample ``` import os import sys import torch...

YingChenlu

## ❓ Questions and Help ### Before asking: 1. search the issues. 2. search the docs. #### What is your question? I noticed fairseq can train with fp16 but the...

SefaZeng

How can I use fairseq on Apple M1 chip with GPU?

4

## ❓ Questions and Help I want to use fairseq on Apple M1 chip for BART model. I checked the document and optional arguments but I could not figure out...

sataketatsuya

Gradient mismatch when using fp16 and freezing part parameters of model

1

## ❓ Questions and Help ### Before asking: 1. search the issues. 2. search the docs. #### What is your question? I try to freeze part parameters of a model...

SefaZeng

Use absolute path in scripts

1

The script with relative paths raises an error of "No such file or directory: data-bin/wikitext-103/dict.txt" in `fairseq-hydra-train`. # Before submitting - [ ] Was this discussed/approved via a Github issue?...

gonglinyuan

Finetuning NLLB models with error "ValueError: --share-all-embeddings requires a joined dictionary", need help!

3

## ❓ Questions and Help I want to test finetuning nllb models(3.3B) , I followed the doc in [Finetuning NLLB models](https://github.com/facebookresearch/fairseq/tree/nllb/examples/nllb/modeling) with this command: `python fairseq/examples/nllb/modeling/train/train_script.py \ cfg=nllb200_dense3.3B_finetune_on_fbseed \ cfg/dataset=fairseq/examples/nllb/modeling/train/conf/cfg/dataset/fbseed_chat.yaml...

cokuehuang

I implemented a function in fairseq_cli/train.py, to freeze the parameters, ``` def freeze_param_grad_zero(model): for name, param in model.named_parameters(): if "fc1" in name or "fc2" in name: print("========= start freezing =========")...

robotsp

How can I save the checkpoint, of which the weights has been altered?

27

## ❓ Questions and Help ### Before asking: 1. search the issues. 2. search the docs. #### What is your question? #### Code #### What have you tried? #### What's...

robotsp

How to use fairseq-hydra-train with multi-nodes?

2

## ❓ Questions and Help I do not plan to use srun and just start the training on two machines by hands. But how to use fairseq-hydra-train with multi-nodes? Configure...

Snowdar

# Before submitting - [ ] Was this discussed/approved via a Github issue? (no need for typos, doc improvements) - [x] Did you read the [contributor guideline](https://github.com/pytorch/fairseq/blob/main/CONTRIBUTING.md)? - [ ]...

hirofumi0810