# python3 -m espnet2.bin.s2t_train --use_preprocessor true --bpemodel data/token_list/bpe_unigram50000/bpe.model --token_type bpe --token_list data/token_list/bpe_unigram50000/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --valid_data_path_and_name_and_type dump/raw/cgn_valid/wav.scp,speech,kaldi_ark --valid_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/speech_shape --resume true --fold_length 80000 --output_dir exp/s2t_lora_2 --config conf/lora_medium_2.yaml --frontend_conf fs=16k --normalize=global_mvn --normalize_conf stats_file=/ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/feats_stats.npz --train_data_path_and_name_and_type dump/raw/cgn_train/wav.scp,speech,kaldi_ark --train_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/speech_shape --fold_length 150 --train_data_path_and_name_and_type dump/raw/cgn_train/text.prev,text_prev,text --train_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/text_prev_shape.bpe --fold_length 150 --train_data_path_and_name_and_type dump/raw/cgn_train/text.ctc,text_ctc,text --train_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/text_ctc_shape.bpe --fold_length 150 --train_data_path_and_name_and_type dump/raw/cgn_train/text,text,text --train_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/text_shape.bpe --valid_data_path_and_name_and_type dump/raw/cgn_valid/text.prev,text_prev,text --valid_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/text_prev_shape.bpe --valid_data_path_and_name_and_type dump/raw/cgn_valid/text.ctc,text_ctc,text --valid_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/text_ctc_shape.bpe --valid_data_path_and_name_and_type dump/raw/cgn_valid/text,text,text --valid_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/text_shape.bpe --model_conf extract_feats_in_collect_stats=false --ngpu 1 --multiprocessing_distributed True 
# Started at Wed May 21 06:58:11 EDT 2025
#
/ocean/projects/cis210027p/pwang10/miniconda3/envs/rota/bin/python3 /ocean/projects/cis210027p/pwang10/rota/espnet/espnet2/bin/s2t_train.py --use_preprocessor true --bpemodel data/token_list/bpe_unigram50000/bpe.model --token_type bpe --token_list data/token_list/bpe_unigram50000/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --valid_data_path_and_name_and_type dump/raw/cgn_valid/wav.scp,speech,kaldi_ark --valid_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/speech_shape --resume true --fold_length 80000 --output_dir exp/s2t_lora_2 --config conf/lora_medium_2.yaml --frontend_conf fs=16k --normalize=global_mvn --normalize_conf stats_file=/ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/feats_stats.npz --train_data_path_and_name_and_type dump/raw/cgn_train/wav.scp,speech,kaldi_ark --train_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/speech_shape --fold_length 150 --train_data_path_and_name_and_type dump/raw/cgn_train/text.prev,text_prev,text --train_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/text_prev_shape.bpe --fold_length 150 --train_data_path_and_name_and_type dump/raw/cgn_train/text.ctc,text_ctc,text --train_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/text_ctc_shape.bpe --fold_length 150 --train_data_path_and_name_and_type dump/raw/cgn_train/text,text,text --train_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/text_shape.bpe --valid_data_path_and_name_and_type dump/raw/cgn_valid/text.prev,text_prev,text --valid_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/text_prev_shape.bpe --valid_data_path_and_name_and_type dump/raw/cgn_valid/text.ctc,text_ctc,text --valid_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/text_ctc_shape.bpe --valid_data_path_and_name_and_type dump/raw/cgn_valid/text,text,text --valid_shape_file /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/text_shape.bpe --model_conf extract_feats_in_collect_stats=false --ngpu 1 --multiprocessing_distributed True
[w010] 2025-05-21 07:00:59,207 (s2t:462) INFO: Vocabulary size: 50002
[w010] 2025-05-21 07:01:04,088 (e_branchformer_encoder:445) INFO: Gradient checkpoint layers: []
[w010] 2025-05-21 07:01:04,595 (transformer_decoder:103) INFO: Gradient checkpoint layers: []
[w010] 2025-05-21 07:01:07,045 (abs_task:1370) INFO: Setting encoder.embed.conv.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,045 (abs_task:1370) INFO: Setting encoder.embed.conv.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,045 (abs_task:1370) INFO: Setting encoder.embed.conv.2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,045 (abs_task:1370) INFO: Setting encoder.embed.conv.2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,045 (abs_task:1370) INFO: Setting encoder.embed.out.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,045 (abs_task:1370) INFO: Setting encoder.embed.out.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,045 (abs_task:1370) INFO: Setting encoder.encoders.0.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,045 (abs_task:1370) INFO: Setting encoder.encoders.0.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.0.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,046 (abs_task:1370) INFO: Setting encoder.encoders.1.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.1.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.2.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.3.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.3.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.3.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.3.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.3.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.3.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.3.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,047 (abs_task:1370) INFO: Setting encoder.encoders.3.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.3.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,048 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.4.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.4.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.4.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.4.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.4.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.5.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,049 (abs_task:1370) INFO: Setting encoder.encoders.6.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.6.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.7.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.8.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.8.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,050 (abs_task:1370) INFO: Setting encoder.encoders.8.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.8.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,051 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.9.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.10.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,052 (abs_task:1370) INFO: Setting encoder.encoders.11.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.11.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,053 (abs_task:1370) INFO: Setting encoder.encoders.12.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.12.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.12.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.13.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,054 (abs_task:1370) INFO: Setting encoder.encoders.14.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.14.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.15.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.16.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.16.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.16.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.16.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.16.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.16.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.16.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,055 (abs_task:1370) INFO: Setting encoder.encoders.16.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.16.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.cgmlp.channel_proj1.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.cgmlp.channel_proj1.0.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.cgmlp.csgu.norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.cgmlp.csgu.norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.cgmlp.csgu.conv.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.cgmlp.csgu.conv.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.cgmlp.channel_proj2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.cgmlp.channel_proj2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.feed_forward_macaron.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.feed_forward_macaron.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.feed_forward_macaron.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.feed_forward_macaron.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_ff.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_ff.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_ff_macaron.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_ff_macaron.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_mha.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_mha.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_mlp.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,056 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_mlp.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,057 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_final.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,057 (abs_task:1370) INFO: Setting encoder.encoders.17.norm_final.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,057 (abs_task:1370) INFO: Setting encoder.encoders.17.depthwise_conv_fusion.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,057 (abs_task:1370) INFO: Setting encoder.encoders.17.depthwise_conv_fusion.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,057 (abs_task:1370) INFO: Setting encoder.encoders.17.merge_proj.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,057 (abs_task:1370) INFO: Setting encoder.encoders.17.merge_proj.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,057 (abs_task:1370) INFO: Setting encoder.after_norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,057 (abs_task:1370) INFO: Setting encoder.after_norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.embed.0.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.after_norm.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.after_norm.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.output_layer.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.output_layer.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,058 (abs_task:1370) INFO: Setting decoder.decoders.0.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.0.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.0.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.1.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.2.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.3.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.3.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.3.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.3.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.3.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.3.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.3.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,059 (abs_task:1370) INFO: Setting decoder.decoders.3.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.3.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.4.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,060 (abs_task:1370) INFO: Setting decoder.decoders.5.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.5.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.5.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.5.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.5.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.5.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.5.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.6.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.7.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.8.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.8.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.8.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.8.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,061 (abs_task:1370) INFO: Setting decoder.decoders.8.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.8.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.9.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,062 (abs_task:1370) INFO: Setting decoder.decoders.10.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.10.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.11.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,063 (abs_task:1370) INFO: Setting decoder.decoders.12.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.12.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.12.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.12.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.13.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.14.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,064 (abs_task:1370) INFO: Setting decoder.decoders.15.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.15.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.16.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.self_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.self_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.self_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.self_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.self_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.self_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.self_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.self_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.src_attn.linear_q.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.src_attn.linear_q.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.src_attn.linear_k.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.src_attn.linear_k.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.src_attn.linear_v.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.src_attn.linear_v.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.src_attn.linear_out.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.src_attn.linear_out.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.feed_forward.w_1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.feed_forward.w_1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.feed_forward.w_2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.feed_forward.w_2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.norm1.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,065 (abs_task:1370) INFO: Setting decoder.decoders.17.norm1.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,066 (abs_task:1370) INFO: Setting decoder.decoders.17.norm2.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,066 (abs_task:1370) INFO: Setting decoder.decoders.17.norm2.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,066 (abs_task:1370) INFO: Setting decoder.decoders.17.norm3.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,066 (abs_task:1370) INFO: Setting decoder.decoders.17.norm3.bias.requires_grad = False
[w010] 2025-05-21 07:01:07,068 (abs_task:1370) INFO: Setting ctc.ctc_lo.weight.requires_grad = False
[w010] 2025-05-21 07:01:07,068 (abs_task:1370) INFO: Setting ctc.ctc_lo.bias.requires_grad = False
[w010] 2025-05-21 07:01:22,366 (abs_task:1398) INFO: pytorch.version=2.1.2, cuda.available=True, cudnn.version=8902, cudnn.benchmark=False, cudnn.deterministic=True
[w010] 2025-05-21 07:01:22,377 (abs_task:1399) INFO: Model structure:
ESPnetS2TModel(
  (frontend): DefaultFrontend(
    (stft): Stft(n_fft=512, win_length=400, hop_length=160, center=True, normalized=False, onesided=True)
    (frontend): Frontend()
    (logmel): LogMel(sr=16000, n_fft=512, n_mels=80, fmin=0, fmax=8000.0, htk=False)
  )
  (specaug): SpecAug(
    (freq_mask): MaskAlongAxis(mask_width_range=[0, 27], num_mask=2, axis=freq)
    (time_mask): MaskAlongAxisVariableMaxWidth(mask_width_ratio_range=[0.0, 0.05], num_mask=10, axis=time)
  )
  (normalize): GlobalMVN(stats_file=/ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium/train/feats_stats.npz, norm_means=True, norm_vars=True)
  (encoder): EBranchformerEncoder(
    (embed): Conv2dSubsampling(
      (conv): Sequential(
        (0): Conv2d(1, 1024, kernel_size=(3, 3), stride=(2, 2))
        (1): ReLU()
        (2): Conv2d(1024, 1024, kernel_size=(3, 3), stride=(2, 2))
        (3): ReLU()
      )
      (out): Sequential(
        (0): Linear(in_features=19456, out_features=1024, bias=True)
        (1): PositionalEncoding(
          (dropout): Dropout(p=0.1, inplace=False)
        )
      )
    )
    (encoders): MultiSequential(
      (0): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (1): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (2): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (3): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (4): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (5): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (6): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (7): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (8): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (9): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (10): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (11): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (12): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (13): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (14): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (15): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (16): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
      (17): EBranchformerEncoderLayer(
        (attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (cgmlp): ConvolutionalGatingMLP(
          (channel_proj1): Sequential(
            (0): Linear(in_features=1024, out_features=4096, bias=True)
            (1): GELU(approximate='none')
          )
          (csgu): ConvolutionalSpatialGatingUnit(
            (norm): LayerNorm((2048,), eps=1e-12, elementwise_affine=True)
            (conv): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
            (act): Identity()
            (dropout): Dropout(p=0.1, inplace=False)
          )
          (channel_proj2): Linear(in_features=2048, out_features=1024, bias=True)
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (feed_forward_macaron): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): Swish()
        )
        (norm_ff): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_ff_macaron): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mha): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_mlp): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm_final): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
        (depthwise_conv_fusion): Conv1d(2048, 2048, kernel_size=(31,), stride=(1,), padding=(15,), groups=2048)
        (merge_proj): Linear(in_features=2048, out_features=1024, bias=True)
      )
    )
    (after_norm): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
  )
  (decoder): TransformerDecoder(
    (embed): Sequential(
      (0): Embedding(50002, 1024)
      (1): PositionalEncoding(
        (dropout): Dropout(p=0.1, inplace=False)
      )
    )
    (after_norm): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
    (output_layer): Linear(in_features=1024, out_features=50002, bias=True)
    (decoders): MultiSequential(
      (0): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (1): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (2): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (3): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (4): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (5): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (6): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (7): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (8): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (9): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (10): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (11): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (12): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (13): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (14): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (15): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (16): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
      (17): DecoderLayer(
        (self_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (src_attn): MultiHeadedAttention(
          (linear_q): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_k): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_v): Linear(in_features=1024, out_features=1024, bias=True)
          (linear_out): Linear(in_features=1024, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (q_norm): Identity()
          (k_norm): Identity()
        )
        (feed_forward): PositionwiseFeedForward(
          (w_1): Linear(in_features=1024, out_features=4096, bias=True)
          (w_2): Linear(in_features=4096, out_features=1024, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (activation): ReLU()
        )
        (norm1): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm2): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (norm3): LayerNorm((1024,), eps=1e-12, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
    )
  )
  (criterion_att): LabelSmoothingLoss(
    (criterion): KLDivLoss()
  )
  (ctc): CTC(
    (ctc_lo): Linear(in_features=1024, out_features=50002, bias=True)
    (ctc_loss): CTCLoss()
  )
)

Model summary:
    Class Name: ESPnetS2TModel
    Total Number of model parameters: 1.04 B
    Number of trainable parameters: 21.13 M (2.0%)
    Size: 84.52 MB
    Type: torch.float32
[w010] 2025-05-21 07:01:22,377 (abs_task:1402) INFO: Optimizer:
AdamW (
Parameter Group 0
    amsgrad: False
    betas: [0.9, 0.98]
    capturable: False
    differentiable: False
    eps: 1e-06
    foreach: None
    fused: None
    initial_lr: 0.0001
    lr: 2.5e-08
    maximize: False
    weight_decay: 0.0
)
[w010] 2025-05-21 07:01:22,377 (abs_task:1403) INFO: Scheduler: PiecewiseLinearWarmupLR(warmup_steps_list=[0, 10000, 50000], warmup_lr_list=[0.0, 0.00025, 0.0001])
[w010] 2025-05-21 07:01:22,416 (abs_task:1412) INFO: Saving the configuration in exp/s2t_lora_2/config.yaml
[w010] 2025-05-21 07:01:22,944 (abs_task:1477) INFO: Loading pretrained params from /ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/owsm_v3.1_ebf/valid.total_count.ave_5best.till45epoch.pth
[w010] 2025-05-21 07:01:42,612 (abs_task:1826) INFO: [train] Batch sampler: UnsortedBatchSampler(N-batch=22430, batch_size=8, key_file=/ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//train/speech_shape, 
[w010] 2025-05-21 07:01:42,616 (abs_task:1827) INFO: [train] mini-batch sizes summary: N-batch=22430, mean=8.0, min=8, max=8
[w010] 2025-05-21 07:01:43,000 (read_text:31) INFO: keys_to_load is not None, only loading 179440 keys from dump/raw/cgn_train/text.prev
[w010] 2025-05-21 07:01:43,118 (read_text:31) INFO: keys_to_load is not None, only loading 179440 keys from dump/raw/cgn_train/text.ctc
[w010] 2025-05-21 07:01:43,628 (read_text:31) INFO: keys_to_load is not None, only loading 179440 keys from dump/raw/cgn_train/text
[w010] 2025-05-21 07:01:44,307 (s2t:444) INFO: Optional Data Names: ('text_prev', 'text_ctc', 'text_spk2', 'text_spk3', 'text_spk4')
[w010] 2025-05-21 07:01:44,310 (abs_task:1851) INFO: [train] dataset:
ESPnetDataset(
  speech: {"path": "dump/raw/cgn_train/wav.scp", "type": "kaldi_ark"}
  text_prev: {"path": "dump/raw/cgn_train/text.prev", "type": "text"}
  text_ctc: {"path": "dump/raw/cgn_train/text.ctc", "type": "text"}
  text: {"path": "dump/raw/cgn_train/text", "type": "text"}
  preprocess: <espnet2.train.preprocessor.S2TPreprocessor object at 0x14a0c0171b40>)
[w010] 2025-05-21 07:01:44,334 (abs_task:1826) INFO: [valid] Batch sampler: UnsortedBatchSampler(N-batch=3220, batch_size=8, key_file=/ocean/projects/cis210027p/pwang10/rota/espnet/egs2/owsm_v3.1/s2t1/exp/s2t_stats_raw_bpe50000_medium//valid/speech_shape, 
[w010] 2025-05-21 07:01:44,335 (abs_task:1827) INFO: [valid] mini-batch sizes summary: N-batch=3220, mean=8.0, min=8, max=9
[w010] 2025-05-21 07:01:44,359 (read_text:31) INFO: keys_to_load is not None, only loading 25766 keys from dump/raw/cgn_valid/text.prev
[w010] 2025-05-21 07:01:44,373 (read_text:31) INFO: keys_to_load is not None, only loading 25766 keys from dump/raw/cgn_valid/text.ctc
[w010] 2025-05-21 07:01:44,405 (read_text:31) INFO: keys_to_load is not None, only loading 25766 keys from dump/raw/cgn_valid/text
[w010] 2025-05-21 07:01:44,461 (s2t:444) INFO: Optional Data Names: ('text_prev', 'text_ctc', 'text_spk2', 'text_spk3', 'text_spk4')
[w010] 2025-05-21 07:01:44,462 (abs_task:1851) INFO: [valid] dataset:
ESPnetDataset(
  speech: {"path": "dump/raw/cgn_valid/wav.scp", "type": "kaldi_ark"}
  text_prev: {"path": "dump/raw/cgn_valid/text.prev", "type": "text"}
  text_ctc: {"path": "dump/raw/cgn_valid/text.ctc", "type": "text"}
  text: {"path": "dump/raw/cgn_valid/text", "type": "text"}
  preprocess: <espnet2.train.preprocessor.S2TPreprocessor object at 0x14a0b89f3d30>)
[w010] 2025-05-21 07:02:20,090 (trainer:177) INFO: The training was resumed using exp/s2t_lora_2/checkpoint.pth
[w010] 2025-05-21 07:02:31,887 (trainer:336) INFO: 5/5epoch started
Failed to import Flash Attention, using ESPnet default: /ocean/projects/cis210027p/pwang10/miniconda3/envs/rota/lib/python3.10/site-packages/flash_attn_2_cuda.cpython-310-x86_64-linux-gnu.so: undefined symbol: _ZNK3c105Error4whatEv
[w010] 2025-05-21 07:07:28,526 (trainer:811) INFO: 5epoch:train:1-750batch: iter_time=6.343e-04, forward_time=0.181, loss_ctc=58.558, loss_att=31.069, acc=0.766, loss=9.829, backward_time=0.179, grad_norm=28.215, clip=100.000, loss_scale=8.389e+06, optim_step_time=0.023, optim0_lr0=2.309e-04, train_time=1.582
[w010] 2025-05-21 07:12:16,988 (trainer:811) INFO: 5epoch:train:751-1500batch: iter_time=1.191e-04, forward_time=0.172, loss_ctc=57.402, loss_att=30.559, acc=0.766, loss=9.653, backward_time=0.178, grad_norm=29.352, clip=100.000, loss_scale=8.389e+06, optim_step_time=0.022, optim0_lr0=2.302e-04, train_time=1.538
[w010] 2025-05-21 07:17:05,533 (trainer:811) INFO: 5epoch:train:1501-2250batch: iter_time=1.163e-04, forward_time=0.172, loss_ctc=57.634, loss_att=30.616, acc=0.771, loss=9.680, backward_time=0.178, grad_norm=26.738, clip=100.000, loss_scale=8.389e+06, optim_step_time=0.022, optim0_lr0=2.295e-04, train_time=1.539
[w010] 2025-05-21 07:21:53,850 (trainer:811) INFO: 5epoch:train:2251-3000batch: iter_time=1.155e-04, forward_time=0.172, loss_ctc=57.076, loss_att=30.148, acc=0.773, loss=9.557, backward_time=0.178, grad_norm=28.065, clip=100.000, loss_scale=8.389e+06, optim_step_time=0.023, optim0_lr0=2.288e-04, train_time=1.537
[w010] 2025-05-21 07:26:42,321 (trainer:811) INFO: 5epoch:train:3001-3750batch: iter_time=1.164e-04, forward_time=0.172, loss_ctc=58.260, loss_att=30.986, acc=0.768, loss=9.792, backward_time=0.178, grad_norm=28.414, clip=100.000, loss_scale=8.389e+06, optim_step_time=0.022, optim0_lr0=2.281e-04, train_time=1.539
[w010] 2025-05-21 07:31:30,882 (trainer:811) INFO: 5epoch:train:3751-4500batch: iter_time=1.176e-04, forward_time=0.172, loss_ctc=56.601, loss_att=30.084, acc=0.770, loss=9.510, backward_time=0.178, grad_norm=29.613, clip=100.000, loss_scale=1.397e+07, optim_step_time=0.022, optim0_lr0=2.274e-04, train_time=1.539
[w010] 2025-05-21 07:36:19,403 (trainer:811) INFO: 5epoch:train:4501-5250batch: iter_time=1.192e-04, forward_time=0.172, loss_ctc=56.591, loss_att=30.002, acc=0.771, loss=9.495, backward_time=0.178, grad_norm=27.209, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.267e-04, train_time=1.539
[w010] 2025-05-21 07:41:07,895 (trainer:811) INFO: 5epoch:train:5251-6000batch: iter_time=1.172e-04, forward_time=0.172, loss_ctc=55.958, loss_att=29.780, acc=0.772, loss=9.408, backward_time=0.178, grad_norm=28.080, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.260e-04, train_time=1.538
[w010] 2025-05-21 07:45:56,174 (trainer:811) INFO: 5epoch:train:6001-6750batch: iter_time=1.151e-04, forward_time=0.172, loss_ctc=56.932, loss_att=30.304, acc=0.769, loss=9.573, backward_time=0.178, grad_norm=28.886, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.253e-04, train_time=1.538
[w010] 2025-05-21 07:50:44,799 (trainer:811) INFO: 5epoch:train:6751-7500batch: iter_time=1.156e-04, forward_time=0.172, loss_ctc=56.285, loss_att=29.821, acc=0.775, loss=9.440, backward_time=0.178, grad_norm=28.730, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.246e-04, train_time=1.539
[w010] 2025-05-21 07:55:34,412 (trainer:811) INFO: 5epoch:train:7501-8250batch: iter_time=1.204e-04, forward_time=0.173, loss_ctc=56.829, loss_att=30.000, acc=0.773, loss=9.512, backward_time=0.179, grad_norm=27.846, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.239e-04, train_time=1.545
[w010] 2025-05-21 08:00:23,042 (trainer:811) INFO: 5epoch:train:8251-9000batch: iter_time=1.189e-04, forward_time=0.172, loss_ctc=56.729, loss_att=29.773, acc=0.774, loss=9.465, backward_time=0.178, grad_norm=29.727, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.232e-04, train_time=1.539
[w010] 2025-05-21 08:05:12,640 (trainer:811) INFO: 5epoch:train:9001-9750batch: iter_time=1.179e-04, forward_time=0.172, loss_ctc=56.605, loss_att=29.687, acc=0.774, loss=9.441, backward_time=0.178, grad_norm=27.438, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.225e-04, train_time=1.544
[w010] 2025-05-21 08:10:13,774 (trainer:811) INFO: 5epoch:train:9751-10500batch: iter_time=3.602e-04, forward_time=0.177, loss_ctc=55.526, loss_att=29.238, acc=0.775, loss=9.281, backward_time=0.184, grad_norm=30.426, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.042, optim0_lr0=2.218e-04, train_time=1.606
[w010] 2025-05-21 08:15:02,469 (trainer:811) INFO: 5epoch:train:10501-11250batch: iter_time=1.199e-04, forward_time=0.172, loss_ctc=56.103, loss_att=29.551, acc=0.776, loss=9.379, backward_time=0.178, grad_norm=28.532, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.210e-04, train_time=1.540
[w010] 2025-05-21 08:19:50,469 (trainer:811) INFO: 5epoch:train:11251-12000batch: iter_time=1.163e-04, forward_time=0.172, loss_ctc=55.003, loss_att=29.101, acc=0.779, loss=9.218, backward_time=0.177, grad_norm=29.368, clip=100.000, loss_scale=1.678e+07, optim_step_time=0.022, optim0_lr0=2.203e-04, train_time=1.536
[w010] 2025-05-21 08:24:39,169 (trainer:811) INFO: 5epoch:train:12001-12750batch: iter_time=1.135e-04, forward_time=0.173, loss_ctc=56.282, loss_att=29.668, acc=0.775, loss=9.413, backward_time=0.178, grad_norm=28.710, clip=100.000, loss_scale=3.355e+07, optim_step_time=0.023, optim0_lr0=2.196e-04, train_time=1.540
[w010] 2025-05-21 08:29:27,102 (trainer:811) INFO: 5epoch:train:12751-13500batch: iter_time=1.157e-04, forward_time=0.172, loss_ctc=55.146, loss_att=29.062, acc=0.776, loss=9.222, backward_time=0.177, grad_norm=31.067, clip=100.000, loss_scale=3.355e+07, optim_step_time=0.022, optim0_lr0=2.189e-04, train_time=1.535
[w010] 2025-05-21 08:34:15,088 (trainer:811) INFO: 5epoch:train:13501-14250batch: iter_time=1.146e-04, forward_time=0.172, loss_ctc=56.991, loss_att=29.821, acc=0.776, loss=9.493, backward_time=0.178, grad_norm=29.138, clip=100.000, loss_scale=3.355e+07, optim_step_time=0.022, optim0_lr0=2.182e-04, train_time=1.536
[w010] 2025-05-21 08:39:03,288 (trainer:811) INFO: 5epoch:train:14251-15000batch: iter_time=1.122e-04, forward_time=0.172, loss_ctc=57.320, loss_att=30.220, acc=0.772, loss=9.588, backward_time=0.178, grad_norm=29.030, clip=100.000, loss_scale=3.355e+07, optim_step_time=0.022, optim0_lr0=2.175e-04, train_time=1.537
[w010] 2025-05-21 08:46:45,861 (trainer:394) INFO: 5epoch results: [train] iter_time=1.548e-04, forward_time=0.173, loss_ctc=56.692, loss_att=29.975, acc=0.773, loss=9.497, backward_time=0.178, grad_norm=28.731, clip=100.000, loss_scale=1.790e+07, optim_step_time=0.023, optim0_lr0=2.242e-04, train_time=1.544, time=1 hour, 36 minutes and 31.58 seconds, total_count=75000, gpu_max_cached_mem_GB=43.420, [valid] loss_ctc=22.228, cer_ctc=0.118, loss_att=10.853, acc=0.914, cer=0.065, wer=0.809, loss=14.266, time=7 minutes and 42.15 seconds, total_count=16100, gpu_max_cached_mem_GB=43.420
[w010] 2025-05-21 08:47:11,206 (trainer:462) INFO: The best model has been updated: valid.acc, valid.total_count
[w010] 2025-05-21 08:47:11,216 (trainer:516) INFO: The model files were removed: exp/s2t_lora_2/4epoch.pth
[w010] 2025-05-21 08:47:11,216 (trainer:534) INFO: The training was finished at 5 epochs 
# Accounting: time=6563 threads=1
# Ended (code 0) at Wed May 21 08:47:34 EDT 2025, elapsed time 6563 seconds