diff options
| author | Gustaf Rydholm <gustaf.rydholm@gmail.com> | 2022-06-20 00:09:20 +0200 |
|---|---|---|
| committer | Gustaf Rydholm <gustaf.rydholm@gmail.com> | 2022-06-20 00:09:20 +0200 |
| commit | 521f3bbbba9f04f48e81d78033c6e1c29a08e515 (patch) | |
| tree | 1ffb50acde358fb151d114f63b760e67c77e3274 /training/conf/experiment | |
| parent | ce3f63801013aba2f05cfb92f1a3a87393610d27 (diff) | |
Update eff net config
Diffstat (limited to 'training/conf/experiment')
| -rw-r--r-- | training/conf/experiment/conv_transformer_paragraphs.yaml | 13 |
1 files changed, 6 insertions, 7 deletions
diff --git a/training/conf/experiment/conv_transformer_paragraphs.yaml b/training/conf/experiment/conv_transformer_paragraphs.yaml index 7c6e231..c8db485 100644 --- a/training/conf/experiment/conv_transformer_paragraphs.yaml +++ b/training/conf/experiment/conv_transformer_paragraphs.yaml @@ -34,15 +34,14 @@ optimizer: betas: [0.9, 0.999] weight_decay: 0 eps: 1.0e-8 - parameters: network lr_scheduler: _target_: torch.optim.lr_scheduler.OneCycleLR max_lr: 3.0e-4 total_steps: null epochs: *epochs - steps_per_epoch: 3201 - pct_start: 0.3 + steps_per_epoch: 5037 + pct_start: 0.15 anneal_strategy: cos cycle_momentum: true base_momentum: 0.85 @@ -56,7 +55,7 @@ lr_scheduler: monitor: val/cer datamodule: - batch_size: 6 + batch_size: 4 train_fraction: 0.95 network: @@ -66,9 +65,9 @@ network: encoder: depth: 5 decoder: - depth: 6 + depth: 4 pixel_embedding: - shape: [18, 78] + shape: [17, 79] model: max_output_len: *max_output_len @@ -76,4 +75,4 @@ model: trainer: gradient_clip_val: 0.5 max_epochs: *epochs - accumulate_grad_batches: 1 + accumulate_grad_batches: 2 |