123456789101112131415161718192021222324252627282930313233343536373839 |
- default: &default
- use_tqdm: true
- random_seed: 42
- base_save_path: /home/msadraei/trained_final
- model_name: google/t5-base-lm-adapt
- project_name_prefix: iclr_softmax_effect
- experiment_name_suffix: null
- train_batch_size: 24
- valid_batch_size: 24
- remove_dropout: true
- learning_rate: 0.01
- weight_decay: 0.01
- num_epochs: 20
- peft_params: null # no mutation
- hot_modules:
- - sadcl
- best_finder:
- save: True
- metric: valid_mean
- higher_better: true
- tasks:
- - glue:qqp
- - glue:qnli
- - glue:mnli
- - glue:sst2
-
- run_configs:
- - <<: *default
- peft_params:
- kind: combine
- n_tokens: 10
- n_comb_tokens: 128
- softmax: true
- # - <<: *default
- # learning_rate: 0.3
- # peft_params:
- # kind: residual
- # n_tokens: 10
- # mlp_size: 128
|