Can't reproduce pretraining results for Wav2vec2 using LibriSpeech recipe about speechbrain HOT 9 OPEN

Hi, thanks @TParcollet for the explanation, it's clearer now. Thanks @Adel-Moumen for pointing out the flag.

I am currently pretraining with --grad_accumulation_factor=2 and max_batch_length=400 on 8 gpus yielding 2 * 400 * 8 = 6400 (~1.8h).

Here's the logs for the first epoch: epoch: 1, steps: 4611, lr: 7.68e-05 - train loss: 4.84e+04 - valid loss: 2.86e+03, valid accuracy: 0.26230588555336

Seems to be similar to our model checkpoint. Note that now you have done during your first epoch "only" 4611 steps meaning that the training will go for much longer. I do expect that you'll get better results.

BTW, are you using --precision=fp16 for the pre-training?

from speechbrain.

Hello @GasserElbanna, thanks a lot for opening this issue!

Could you please @TParcollet and/or @salah-zaiem have a look? Thanks a lot :)

from speechbrain.

Hi, it's important that the total batch size corresponds to roughly 1.6h. By changing the gradient accumulation factor your can adjust this.

from speechbrain.

Hello, thank you for the quick response. I used the default config file for pre-training. So, I am assuming these are the parameters below I need to adjust?

Dynamic Batching parameters:
max_batch_length: 200 # Fits in a 32GB GPUs (V100)
num_buckets: 70
shuffle: True # if true re-creates batches at each epoch shuffling examples.
batch_ordering: random

from speechbrain.

@Adel-Moumen i see that the gradient accumulation factor is missing on this recipe. Could you add it? (No need to PR imho push directly to develop).

@GasserElbanna have a look at any other yaml for asr in the libri folder, you will find the gradient accumulation factor param. Just copy and past it in this yaml, anywhere. Then play with grad accum / max batch len to make sure that you have 1.2-1.6h of speech per batch. Grad_accum * max_batch_len * nb gpu = 1.6h.

Also, your A100 must certainly be able to accommodate more than 200s.

from speechbrain.

@Adel-Moumen i see that the gradient accumulation factor is missing on this recipe. Could you add it? (No need to PR imho push directly to develop).

Why would it be missing? By default, grad_accumulation_factor is set to 1 (see: https://github.com/speechbrain/speechbrain/blob/develop/speechbrain/core.py#L84). The var is called in each fit_batch call (see: https://github.com/speechbrain/speechbrain/blob/develop/speechbrain/core.py#L1199). As grad_accumulation_factor can also be set through a flag (see: https://github.com/speechbrain/speechbrain/blob/develop/speechbrain/core.py#L422-L426) the recipe is technically not missing from this feature. You just need to play with --grad_accumulation_factor=N where N is the grad acc steps.

from speechbrain.

Hi, thanks @TParcollet for the explanation, it's clearer now.
Thanks @Adel-Moumen for pointing out the flag.

I am currently pretraining with --grad_accumulation_factor=2 and max_batch_length=400 on 8 gpus yielding 2 * 400 * 8 = 6400 (~1.8h).

Here's the logs for the first epoch:
epoch: 1, steps: 4611, lr: 7.68e-05 - train loss: 4.84e+04 - valid loss: 2.86e+03, valid accuracy: 0.26230588555336

from speechbrain.

BTW, are you using --precision=fp16 for the pre-training?

I am using fp32 now.

from speechbrain.

fp16 or bf16 would make the training much faster if you have a compatible GPU.

from speechbrain.