JohnnyZeppelin commited on
Commit
c78c96b
·
verified ·
1 Parent(s): bed4c06

Upload folder using huggingface_hub

Browse files
Files changed (48) hide show
  1. planB_polish/ckpt.pt +3 -0
  2. planB_polish/config.json +38 -0
  3. planB_polish/diag.jsonl +2 -0
  4. planB_polish/diag_full.json +19 -0
  5. planB_polish/logs.jsonl +50 -0
  6. planB_polish/probe_leakage.json +35 -0
  7. planB_polish_robust_dec/ckpt.pt +3 -0
  8. planB_polish_robust_dec/config.json +40 -0
  9. planB_polish_robust_dec/diag.jsonl +1 -0
  10. planB_polish_robust_dec/diag_full.json +19 -0
  11. planB_polish_robust_dec/logs.jsonl +25 -0
  12. planB_polish_robust_dec/probe_leakage.json +35 -0
  13. planB_polish_robust_dec_long/ckpt.pt +3 -0
  14. planB_polish_robust_dec_long/config.json +24 -0
  15. planB_polish_robust_dec_long/diag.jsonl +38 -0
  16. planB_polish_robust_dec_long/diag_full.json +19 -0
  17. planB_polish_robust_dec_long/logs.jsonl +0 -0
  18. planB_polish_robust_dec_long/probe_leakage.json +35 -0
  19. planB_polish_robust_dec_noise02/ckpt.pt +3 -0
  20. planB_polish_robust_dec_noise02/config.json +24 -0
  21. planB_polish_robust_dec_noise02/diag.jsonl +22 -0
  22. planB_polish_robust_dec_noise02/diag_full.json +19 -0
  23. planB_polish_robust_dec_noise02/logs.jsonl +448 -0
  24. planB_polish_robust_dec_noise02/probe_leakage.json +35 -0
  25. planB_polish_robust_dec_noise05/ckpt.pt +3 -0
  26. planB_polish_robust_dec_noise05/config.json +24 -0
  27. planB_polish_robust_dec_noise05/diag.jsonl +22 -0
  28. planB_polish_robust_dec_noise05/diag_full.json +19 -0
  29. planB_polish_robust_dec_noise05/logs.jsonl +448 -0
  30. planB_polish_robust_dec_noise05/probe_leakage.json +35 -0
  31. planB_polish_robust_dec_noise07_longwarm/ckpt.pt +3 -0
  32. planB_polish_robust_dec_noise07_longwarm/config.json +24 -0
  33. planB_polish_robust_dec_noise07_longwarm/diag.jsonl +22 -0
  34. planB_polish_robust_dec_noise07_longwarm/diag_full.json +19 -0
  35. planB_polish_robust_dec_noise07_longwarm/logs.jsonl +448 -0
  36. planB_polish_robust_dec_noise07_longwarm/probe_leakage.json +35 -0
  37. planB_polish_robust_dec_noise10/ckpt.pt +3 -0
  38. planB_polish_robust_dec_noise10/config.json +24 -0
  39. planB_polish_robust_dec_noise10/diag.jsonl +22 -0
  40. planB_polish_robust_dec_noise10/diag_full.json +19 -0
  41. planB_polish_robust_dec_noise10/logs.jsonl +448 -0
  42. planB_polish_robust_dec_noise10/probe_leakage.json +35 -0
  43. planB_robust/ckpt.pt +3 -0
  44. planB_robust/config.json +38 -0
  45. planB_robust/diag.jsonl +1 -0
  46. planB_robust/diag_full.json +19 -0
  47. planB_robust/logs.jsonl +25 -0
  48. planB_robust/probe_leakage.json +35 -0
planB_polish/ckpt.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d521da49f0c3ec3362d3e017b559edcfe7ed87d3fa505963cd52c9ef65e80eb9
3
+ size 3646627185
planB_polish/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_jsonl": "data/groups_train.jsonl",
3
+ "valid_jsonl": "data/groups_valid.jsonl",
4
+ "run_dir": "runs/planB_polish",
5
+ "resume": "runs/planB_ramp_adv/ckpt.pt",
6
+ "backbone": "google/mt5-small",
7
+ "num_latents": 16,
8
+ "latent_dropout": 0.0,
9
+ "latent_noise_std": 0.0,
10
+ "batch_size": 4,
11
+ "grad_accum": 8,
12
+ "epochs": 2,
13
+ "max_doc_len": 256,
14
+ "max_sum_len": 64,
15
+ "eval_every": 400,
16
+ "max_train_examples": 2000,
17
+ "max_valid_examples": 200,
18
+ "lambda_align": 1.0,
19
+ "tau": 0.07,
20
+ "lambda_varcov": 0.1,
21
+ "var_target_std": 0.05,
22
+ "lambda_mean": 0.5,
23
+ "lambda_mean_diff": 0.1,
24
+ "lambda_pair": 0.005,
25
+ "lambda_lang": 0.5,
26
+ "lambda_len": 1.0,
27
+ "adv_start_step": 0,
28
+ "grl_alpha": 1.0,
29
+ "grl_warmup": 200,
30
+ "lr_model": 0.0003,
31
+ "lr_lang": 0.005,
32
+ "lr_len": 0.001,
33
+ "adv_clf_steps": 5,
34
+ "adv_clf_weight_decay": 0.0,
35
+ "adv_queue_size": 4096,
36
+ "adv_clf_batch": 256,
37
+ "adv_mix_current": 0.5
38
+ }
planB_polish/diag.jsonl ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ {"step": 2400, "nll_en": 5.047823696136475, "nll_zh": 5.0464230728149415, "nll": 5.047123384475708, "lang_acc_valid_evalmode": 0.49, "len_acc_valid_evalmode": 0.895}
2
+ {"step": 2800, "nll_en": 4.876935539245605, "nll_zh": 4.888795242309571, "nll": 4.882865390777588, "lang_acc_valid_evalmode": 0.515, "len_acc_valid_evalmode": 0.8875}
planB_polish/diag_full.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish/ckpt.pt",
3
+ "nll_en": 4.856141153904146,
4
+ "nll_zh": 4.867885981755327,
5
+ "swap_delta_en": 0.4700156619130431,
6
+ "swap_delta_zh": 0.4502284098377187,
7
+ "ablate_zero_delta_en": 0.713076062232697,
8
+ "ablate_mean_delta_en": 0.19092984028144822,
9
+ "ablate_noise_delta_en": 3.3194047441946277,
10
+ "ablate_zero_delta_zh": 0.7013312343815165,
11
+ "ablate_mean_delta_zh": 0.17876631512954422,
12
+ "ablate_noise_delta_zh": 3.354815980092668,
13
+ "inv_top1_full": 0.060253698378801346,
14
+ "inv_top5_full": 0.20401690900325775,
15
+ "diag_sim_mean": 0.965337336063385,
16
+ "offdiag_sim_mean": 0.4495629370212555,
17
+ "sim_margin": 0.5157743990421295,
18
+ "n_valid": 946
19
+ }
planB_polish/logs.jsonl ADDED
@@ -0,0 +1,50 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 2020, "epoch": 0, "loss": 8.432245254516602, "loss_render": 6.179324150085449, "loss_align": 0.5220547318458557, "loss_varcov": 0.05400950089097023, "loss_mean": 0.6911861300468445, "loss_mean_diff": 0.012336799874901772, "mean_diff_l2": 0.1110711470246315, "loss_pair": 5.848134605912492e-05, "loss_lang": 2.509106159210205, "loss_len": 0.12408515810966492, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.5654761791229248, "len_acc_clf_detached_mix": 0.8690476417541504, "adv_buf_fill": 160, "elapsed_sec": 8.348091125488281}
2
+ {"step": 2040, "epoch": 0, "loss": 8.454788208007812, "loss_render": 5.9538655281066895, "loss_align": 0.7937119007110596, "loss_varcov": 0.06152995675802231, "loss_mean": 0.8048664331436157, "loss_mean_diff": 0.0014553789515048265, "mean_diff_l2": 0.038149427622556686, "loss_pair": 2.8349248168524355e-05, "loss_lang": 1.0162914991378784, "loss_len": 0.790332555770874, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.64453125, "len_acc_clf_detached_mix": 0.90625, "adv_buf_fill": 320, "elapsed_sec": 16.114933490753174}
3
+ {"step": 2060, "epoch": 0, "loss": 9.439117431640625, "loss_render": 6.358119010925293, "loss_align": 0.32713401317596436, "loss_varcov": 0.05208443105220795, "loss_mean": 0.6751819849014282, "loss_mean_diff": 0.02081209048628807, "mean_diff_l2": 0.14426396787166595, "loss_pair": 0.00018489659123588353, "loss_lang": 3.977900743484497, "loss_len": 0.420031875371933, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.875, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.875, "lang_acc_clf_detached_mix": 0.64453125, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 480, "elapsed_sec": 23.738660097122192}
4
+ {"step": 2080, "epoch": 0, "loss": 8.303224563598633, "loss_render": 6.696105003356934, "loss_align": 0.6665774583816528, "loss_varcov": 0.05075671523809433, "loss_mean": 0.5891607999801636, "loss_mean_diff": 0.020316867157816887, "mean_diff_l2": 0.14253725111484528, "loss_pair": 0.0001137369399657473, "loss_lang": 1.0339231491088867, "loss_len": 0.12189283967018127, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.61328125, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 640, "elapsed_sec": 31.59323477745056}
5
+ {"step": 2100, "epoch": 0, "loss": 8.138189315795898, "loss_render": 5.982770919799805, "loss_align": 0.36127209663391113, "loss_varcov": 0.042167551815509796, "loss_mean": 0.44467693567276, "loss_mean_diff": 0.03013492189347744, "mean_diff_l2": 0.173594132065773, "loss_pair": 0.00026721874019131064, "loss_lang": 1.20796799659729, "loss_len": 0.96059250831604, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 0.625, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 0.625, "lang_acc_clf_detached_mix": 0.58984375, "len_acc_clf_detached_mix": 0.93359375, "adv_buf_fill": 800, "elapsed_sec": 38.75260877609253}
6
+ {"step": 2120, "epoch": 0, "loss": 6.913006782531738, "loss_render": 5.4898176193237305, "loss_align": 0.4929021894931793, "loss_varcov": 0.048337798565626144, "loss_mean": 0.5435984134674072, "loss_mean_diff": 0.026746824383735657, "mean_diff_l2": 0.16354456543922424, "loss_pair": 0.00011604805331444368, "loss_lang": 1.0475423336029053, "loss_len": 0.1272079348564148, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.25, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.25, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.578125, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 960, "elapsed_sec": 46.786195278167725}
7
+ {"step": 2140, "epoch": 0, "loss": 7.455108165740967, "loss_render": 5.4744415283203125, "loss_align": 0.5238228440284729, "loss_varcov": 0.05664694309234619, "loss_mean": 0.7510408163070679, "loss_mean_diff": 0.004368389956653118, "mean_diff_l2": 0.0660937950015068, "loss_pair": 0.00012321726535446942, "loss_lang": 2.1047840118408203, "loss_len": 0.022829536348581314, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.55078125, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 1120, "elapsed_sec": 54.0298285484314}
8
+ {"step": 2160, "epoch": 0, "loss": 7.8427958488464355, "loss_render": 5.697396755218506, "loss_align": 1.0171148777008057, "loss_varcov": 0.047145865857601166, "loss_mean": 0.5258265733718872, "loss_mean_diff": 0.06748493015766144, "mean_diff_l2": 0.25977861881256104, "loss_pair": 0.0002962924772873521, "loss_lang": 0.8799240589141846, "loss_len": 0.41394445300102234, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.60546875, "len_acc_clf_detached_mix": 0.92578125, "adv_buf_fill": 1280, "elapsed_sec": 61.65147662162781}
9
+ {"step": 2180, "epoch": 0, "loss": 6.799259662628174, "loss_render": 5.672228813171387, "loss_align": 0.4496840834617615, "loss_varcov": 0.04664335772395134, "loss_mean": 0.5405099391937256, "loss_mean_diff": 0.04637732729315758, "mean_diff_l2": 0.21535395085811615, "loss_pair": 0.0001891774300020188, "loss_lang": 0.7469799518585205, "loss_len": 0.02429870143532753, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.75, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.75, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.69140625, "len_acc_clf_detached_mix": 0.93359375, "adv_buf_fill": 1440, "elapsed_sec": 68.54602360725403}
10
+ {"step": 2200, "epoch": 0, "loss": 7.60024356842041, "loss_render": 5.131892204284668, "loss_align": 0.8027728199958801, "loss_varcov": 0.051322925835847855, "loss_mean": 0.6060923337936401, "loss_mean_diff": 0.021392300724983215, "mean_diff_l2": 0.146261066198349, "loss_pair": 0.0001292075467063114, "loss_lang": 1.5172581672668457, "loss_len": 0.5966311097145081, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.59375, "len_acc_clf_detached_mix": 0.921875, "adv_buf_fill": 1600, "elapsed_sec": 75.70219159126282}
11
+ {"step": 2220, "epoch": 0, "loss": 7.661475658416748, "loss_render": 5.847012519836426, "loss_align": 0.07052421569824219, "loss_varcov": 0.04100217670202255, "loss_mean": 0.3689679205417633, "loss_mean_diff": 0.009925302118062973, "mean_diff_l2": 0.09962581098079681, "loss_pair": 7.943976379465312e-05, "loss_lang": 2.947748899459839, "loss_len": 0.08048711717128754, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.70703125, "len_acc_clf_detached_mix": 0.92578125, "adv_buf_fill": 1760, "elapsed_sec": 83.8407096862793}
12
+ {"step": 2240, "epoch": 0, "loss": 8.128841400146484, "loss_render": 5.740845680236816, "loss_align": 0.35049304366111755, "loss_varcov": 0.040174249559640884, "loss_mean": 0.39410164952278137, "loss_mean_diff": 0.024943649768829346, "mean_diff_l2": 0.15793558955192566, "loss_pair": 0.00012857439287472516, "loss_lang": 0.8992084860801697, "loss_len": 1.384335994720459, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.5, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.5, "lang_acc_clf_detached_mix": 0.63671875, "len_acc_clf_detached_mix": 0.91015625, "adv_buf_fill": 1920, "elapsed_sec": 92.09641432762146}
13
+ {"step": 2260, "epoch": 0, "loss": 8.754077911376953, "loss_render": 6.074049949645996, "loss_align": 0.846519947052002, "loss_varcov": 0.05600658804178238, "loss_mean": 0.7005522847175598, "loss_mean_diff": 0.022828727960586548, "mean_diff_l2": 0.1510917842388153, "loss_pair": 0.00043868302600458264, "loss_lang": 2.894388437271118, "loss_len": 0.02815117873251438, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6796875, "len_acc_clf_detached_mix": 0.92578125, "adv_buf_fill": 2080, "elapsed_sec": 100.66716074943542}
14
+ {"step": 2280, "epoch": 0, "loss": 7.055285453796387, "loss_render": 5.0745930671691895, "loss_align": 0.7090506553649902, "loss_varcov": 0.046681538224220276, "loss_mean": 0.47311341762542725, "loss_mean_diff": 0.06792638450860977, "mean_diff_l2": 0.2606269121170044, "loss_pair": 0.0003356885281391442, "loss_lang": 1.9712920188903809, "loss_len": 0.037976283580064774, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.58984375, "len_acc_clf_detached_mix": 0.9375, "adv_buf_fill": 2240, "elapsed_sec": 108.90034437179565}
15
+ {"step": 2300, "epoch": 0, "loss": 7.308385848999023, "loss_render": 5.55955696105957, "loss_align": 0.5478301048278809, "loss_varcov": 0.052591294050216675, "loss_mean": 0.6486594676971436, "loss_mean_diff": 0.03394661843776703, "mean_diff_l2": 0.18424607813358307, "loss_pair": 0.0004092962190043181, "loss_lang": 1.6372103691101074, "loss_len": 0.049408070743083954, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6171875, "len_acc_clf_detached_mix": 0.94140625, "adv_buf_fill": 2400, "elapsed_sec": 116.85040402412415}
16
+ {"step": 2320, "epoch": 0, "loss": 6.9919209480285645, "loss_render": 5.368945121765137, "loss_align": 0.8128483295440674, "loss_varcov": 0.054375723004341125, "loss_mean": 0.6507823467254639, "loss_mean_diff": 0.01333271712064743, "mean_diff_l2": 0.1154673844575882, "loss_pair": 4.8895995860220864e-05, "loss_lang": 0.870909571647644, "loss_len": 0.04251052439212799, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.640625, "len_acc_clf_detached_mix": 0.94921875, "adv_buf_fill": 2560, "elapsed_sec": 124.74422240257263}
17
+ {"step": 2340, "epoch": 0, "loss": 7.382444381713867, "loss_render": 5.411495685577393, "loss_align": 0.7850296497344971, "loss_varcov": 0.05425461381673813, "loss_mean": 0.6286738514900208, "loss_mean_diff": 0.004980956204235554, "mean_diff_l2": 0.07057589292526245, "loss_pair": 2.9360424377955496e-05, "loss_lang": 1.5415778160095215, "loss_len": 0.09486955404281616, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.62109375, "len_acc_clf_detached_mix": 0.88671875, "adv_buf_fill": 2720, "elapsed_sec": 133.04614639282227}
18
+ {"step": 2360, "epoch": 0, "loss": 6.910673141479492, "loss_render": 5.616985321044922, "loss_align": 0.5049238204956055, "loss_varcov": 0.05236847698688507, "loss_mean": 0.6127978563308716, "loss_mean_diff": 0.001157041871920228, "mean_diff_l2": 0.034015316516160965, "loss_pair": 2.223800947831478e-05, "loss_lang": 0.7309256792068481, "loss_len": 0.11154971271753311, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.65234375, "len_acc_clf_detached_mix": 0.91015625, "adv_buf_fill": 2880, "elapsed_sec": 141.20867109298706}
19
+ {"step": 2380, "epoch": 0, "loss": 6.987215042114258, "loss_render": 5.1249847412109375, "loss_align": 0.7023035287857056, "loss_varcov": 0.049918174743652344, "loss_mean": 0.4639621376991272, "loss_mean_diff": 0.0050551872700452805, "mean_diff_l2": 0.07109984010457993, "loss_pair": 5.8491230447543785e-05, "loss_lang": 1.088517427444458, "loss_len": 0.37818920612335205, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.875, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.875, "lang_acc_clf_detached_mix": 0.6796875, "len_acc_clf_detached_mix": 0.9375, "adv_buf_fill": 3040, "elapsed_sec": 149.57889676094055}
20
+ {"step": 2400, "epoch": 0, "loss": 6.946876049041748, "loss_render": 5.560789108276367, "loss_align": 0.4598637521266937, "loss_varcov": 0.04878554865717888, "loss_mean": 0.5330585241317749, "loss_mean_diff": 0.012456746771931648, "mean_diff_l2": 0.1116097941994667, "loss_pair": 6.580528861377388e-05, "loss_lang": 1.237382173538208, "loss_len": 0.03487871587276459, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.66796875, "len_acc_clf_detached_mix": 0.87890625, "adv_buf_fill": 3200, "elapsed_sec": 158.19630217552185}
21
+ {"step": 2420, "epoch": 0, "loss": 8.124646186828613, "loss_render": 5.217735290527344, "loss_align": 1.2843619585037231, "loss_varcov": 0.07266464829444885, "loss_mean": 0.9399237632751465, "loss_mean_diff": 0.014321889728307724, "mean_diff_l2": 0.11967410147190094, "loss_pair": 0.0001567260769661516, "loss_lang": 2.17319917678833, "loss_len": 0.05728795751929283, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6484375, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 3360, "elapsed_sec": 177.55492305755615}
22
+ {"step": 2440, "epoch": 0, "loss": 6.378420829772949, "loss_render": 5.374331474304199, "loss_align": 0.3013423681259155, "loss_varcov": 0.04462404549121857, "loss_mean": 0.4438955783843994, "loss_mean_diff": 0.002731515094637871, "mean_diff_l2": 0.05226389691233635, "loss_pair": 3.855187969747931e-05, "loss_lang": 0.6885992288589478, "loss_len": 0.13176411390304565, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.64453125, "len_acc_clf_detached_mix": 0.91796875, "adv_buf_fill": 3520, "elapsed_sec": 185.9959578514099}
23
+ {"step": 2460, "epoch": 0, "loss": 7.462551593780518, "loss_render": 5.180554389953613, "loss_align": 0.4856002628803253, "loss_varcov": 0.05582088604569435, "loss_mean": 0.7025367617607117, "loss_mean_diff": 0.04949355870485306, "mean_diff_l2": 0.2224714756011963, "loss_pair": 0.00023962598061189055, "loss_lang": 2.7334885597229004, "loss_len": 0.06785154342651367, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.25, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.25, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6015625, "len_acc_clf_detached_mix": 0.9375, "adv_buf_fill": 3680, "elapsed_sec": 194.64414381980896}
24
+ {"step": 2480, "epoch": 0, "loss": 8.03963851928711, "loss_render": 5.666196823120117, "loss_align": 0.24791663885116577, "loss_varcov": 0.04851426184177399, "loss_mean": 0.5308257341384888, "loss_mean_diff": 0.031466543674468994, "mean_diff_l2": 0.17738811671733856, "loss_pair": 0.00012165836960775778, "loss_lang": 1.9852598905563354, "loss_len": 0.8594835996627808, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.62109375, "len_acc_clf_detached_mix": 0.8984375, "adv_buf_fill": 3840, "elapsed_sec": 203.3888545036316}
25
+ {"step": 2500, "epoch": 0, "loss": 8.485847473144531, "loss_render": 5.429369926452637, "loss_align": 0.0666179209947586, "loss_varcov": 0.05050269886851311, "loss_mean": 0.5938384532928467, "loss_mean_diff": 0.017245417460799217, "mean_diff_l2": 0.13132180273532867, "loss_pair": 0.00010503844532649964, "loss_lang": 3.3776168823242188, "loss_len": 0.9973562955856323, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.61328125, "len_acc_clf_detached_mix": 0.90625, "adv_buf_fill": 4000, "elapsed_sec": 211.8817949295044}
26
+ {"step": 2520, "epoch": 1, "loss": 6.599848747253418, "loss_render": 5.131601333618164, "loss_align": 0.3769349753856659, "loss_varcov": 0.05128224939107895, "loss_mean": 0.6264356374740601, "loss_mean_diff": 0.03924765810370445, "mean_diff_l2": 0.19811022281646729, "loss_pair": 0.00014803450903855264, "loss_lang": 1.4644688367843628, "loss_len": 0.03680605813860893, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.58984375, "len_acc_clf_detached_mix": 0.91796875, "adv_buf_fill": 4096, "elapsed_sec": 220.52107977867126}
27
+ {"step": 2540, "epoch": 1, "loss": 6.262823581695557, "loss_render": 5.061046600341797, "loss_align": 0.08988380432128906, "loss_varcov": 0.0478525385260582, "loss_mean": 0.543313205242157, "loss_mean_diff": 0.011552654206752777, "mean_diff_l2": 0.10748327523469925, "loss_pair": 9.796085942070931e-05, "loss_lang": 1.495964527130127, "loss_len": 0.08631375432014465, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.66015625, "len_acc_clf_detached_mix": 0.94140625, "adv_buf_fill": 4096, "elapsed_sec": 228.28239917755127}
28
+ {"step": 2560, "epoch": 1, "loss": 6.790939807891846, "loss_render": 4.889214515686035, "loss_align": 0.721789538860321, "loss_varcov": 0.06119423732161522, "loss_mean": 0.7951768636703491, "loss_mean_diff": 0.013758774846792221, "mean_diff_l2": 0.11729780584573746, "loss_pair": 0.00010747999476734549, "loss_lang": 1.4322483539581299, "loss_len": 0.05872827023267746, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.625, "len_acc_clf_detached_mix": 0.9296875, "adv_buf_fill": 4096, "elapsed_sec": 236.4033453464508}
29
+ {"step": 2580, "epoch": 1, "loss": 7.487343788146973, "loss_render": 5.490227222442627, "loss_align": 0.11694781482219696, "loss_varcov": 0.046837322413921356, "loss_mean": 0.57428377866745, "loss_mean_diff": 0.03555521368980408, "mean_diff_l2": 0.18856090307235718, "loss_pair": 0.00018275401089340448, "loss_lang": 3.0386157035827637, "loss_len": 0.06547920405864716, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.640625, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 4096, "elapsed_sec": 243.95104908943176}
30
+ {"step": 2600, "epoch": 1, "loss": 6.279510021209717, "loss_render": 5.079255104064941, "loss_align": 0.2688593566417694, "loss_varcov": 0.051914528012275696, "loss_mean": 0.6674661636352539, "loss_mean_diff": 0.019260544329881668, "mean_diff_l2": 0.13878236711025238, "loss_pair": 0.00013908441178500652, "loss_lang": 0.997173011302948, "loss_len": 0.09195810556411743, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6015625, "len_acc_clf_detached_mix": 0.93359375, "adv_buf_fill": 4096, "elapsed_sec": 252.20418524742126}
31
+ {"step": 2620, "epoch": 1, "loss": 8.2630615234375, "loss_render": 5.047208786010742, "loss_align": 0.752984881401062, "loss_varcov": 0.04717573523521423, "loss_mean": 0.5315133333206177, "loss_mean_diff": 0.02449980564415455, "mean_diff_l2": 0.15652413666248322, "loss_pair": 0.00023642193991690874, "loss_lang": 1.4204857349395752, "loss_len": 1.4797002077102661, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.5, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.5, "lang_acc_clf_detached_mix": 0.6328125, "len_acc_clf_detached_mix": 0.91796875, "adv_buf_fill": 4096, "elapsed_sec": 260.25734210014343}
32
+ {"step": 2640, "epoch": 1, "loss": 7.792613983154297, "loss_render": 5.511990547180176, "loss_align": 0.15171398222446442, "loss_varcov": 0.04937530308961868, "loss_mean": 0.6251848936080933, "loss_mean_diff": 0.0333007276058197, "mean_diff_l2": 0.18248486518859863, "loss_pair": 0.00022945625823922455, "loss_lang": 1.5273370742797852, "loss_len": 1.0443795919418335, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.59765625, "len_acc_clf_detached_mix": 0.90234375, "adv_buf_fill": 4096, "elapsed_sec": 268.59954619407654}
33
+ {"step": 2660, "epoch": 1, "loss": 5.796278953552246, "loss_render": 5.093378067016602, "loss_align": 0.03718061000108719, "loss_varcov": 0.04019087553024292, "loss_mean": 0.41506779193878174, "loss_mean_diff": 0.010585382580757141, "mean_diff_l2": 0.10288529098033905, "loss_pair": 8.889130549505353e-05, "loss_lang": 0.646791934967041, "loss_len": 0.12971191108226776, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.609375, "len_acc_clf_detached_mix": 0.9453125, "adv_buf_fill": 4096, "elapsed_sec": 275.9460287094116}
34
+ {"step": 2680, "epoch": 1, "loss": 8.264688491821289, "loss_render": 5.657985687255859, "loss_align": 0.49703875184059143, "loss_varcov": 0.04953797534108162, "loss_mean": 0.5910893082618713, "loss_mean_diff": 0.056294284760951996, "mean_diff_l2": 0.23726417124271393, "loss_pair": 0.00022839047596789896, "loss_lang": 1.7251404523849487, "loss_len": 0.9409642219543457, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.62890625, "len_acc_clf_detached_mix": 0.91015625, "adv_buf_fill": 4096, "elapsed_sec": 283.90607833862305}
35
+ {"step": 2700, "epoch": 1, "loss": 6.907713413238525, "loss_render": 5.405125617980957, "loss_align": 0.1331290453672409, "loss_varcov": 0.04752184450626373, "loss_mean": 0.5648655891418457, "loss_mean_diff": 0.04137600213289261, "mean_diff_l2": 0.20341092348098755, "loss_pair": 0.00020081066759303212, "loss_lang": 2.013303518295288, "loss_len": 0.07148371636867523, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.25, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.25, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.609375, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 4096, "elapsed_sec": 292.2394049167633}
36
+ {"step": 2720, "epoch": 1, "loss": 6.705268383026123, "loss_render": 5.078215599060059, "loss_align": 0.43562719225883484, "loss_varcov": 0.05211036279797554, "loss_mean": 0.6539069414138794, "loss_mean_diff": 0.005899499170482159, "mean_diff_l2": 0.07680819928646088, "loss_pair": 0.00012971702381037176, "loss_lang": 1.6448332071304321, "loss_len": 0.0362544059753418, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.5703125, "len_acc_clf_detached_mix": 0.94921875, "adv_buf_fill": 4096, "elapsed_sec": 300.398241519928}
37
+ {"step": 2740, "epoch": 1, "loss": 6.60788106918335, "loss_render": 5.094322204589844, "loss_align": 0.32772189378738403, "loss_varcov": 0.05520971864461899, "loss_mean": 0.7002456188201904, "loss_mean_diff": 0.007205964997410774, "mean_diff_l2": 0.08488795906305313, "loss_pair": 4.675781019614078e-05, "loss_lang": 1.4783399105072021, "loss_len": 0.09030262380838394, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.609375, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 4096, "elapsed_sec": 307.54498052597046}
38
+ {"step": 2760, "epoch": 1, "loss": 6.0640387535095215, "loss_render": 4.802050590515137, "loss_align": 0.06428647041320801, "loss_varcov": 0.046824343502521515, "loss_mean": 0.5707664489746094, "loss_mean_diff": 0.018675241619348526, "mean_diff_l2": 0.13665738701820374, "loss_pair": 7.991994789335877e-05, "loss_lang": 1.574737787246704, "loss_len": 0.11839956045150757, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6171875, "len_acc_clf_detached_mix": 0.890625, "adv_buf_fill": 4096, "elapsed_sec": 315.0056092739105}
39
+ {"step": 2780, "epoch": 1, "loss": 6.4692511558532715, "loss_render": 5.474024772644043, "loss_align": 0.12821653485298157, "loss_varcov": 0.04507732018828392, "loss_mean": 0.5349506735801697, "loss_mean_diff": 0.01526686456054449, "mean_diff_l2": 0.12355915457010269, "loss_pair": 6.491704698419198e-05, "loss_lang": 1.0484888553619385, "loss_len": 0.06925535947084427, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.62890625, "len_acc_clf_detached_mix": 0.9453125, "adv_buf_fill": 4096, "elapsed_sec": 322.7376163005829}
40
+ {"step": 2800, "epoch": 1, "loss": 6.309009552001953, "loss_render": 5.223487854003906, "loss_align": 0.3440673053264618, "loss_varcov": 0.04397273808717728, "loss_mean": 0.49700912833213806, "loss_mean_diff": 0.013312110677361488, "mean_diff_l2": 0.11537811905145645, "loss_pair": 9.047871571965516e-05, "loss_lang": 0.8018254041671753, "loss_len": 0.08630776405334473, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.58984375, "len_acc_clf_detached_mix": 0.94140625, "adv_buf_fill": 4096, "elapsed_sec": 330.51809191703796}
41
+ {"step": 2820, "epoch": 1, "loss": 6.631684303283691, "loss_render": 4.79958438873291, "loss_align": 0.803274393081665, "loss_varcov": 0.05635480955243111, "loss_mean": 0.7139475345611572, "loss_mean_diff": 0.009432838298380375, "mean_diff_l2": 0.09712280333042145, "loss_pair": 3.4327866160310805e-05, "loss_lang": 1.22541344165802, "loss_len": 0.05256691202521324, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6015625, "len_acc_clf_detached_mix": 0.93359375, "adv_buf_fill": 4096, "elapsed_sec": 349.6525809764862}
42
+ {"step": 2840, "epoch": 1, "loss": 6.570076942443848, "loss_render": 5.102474689483643, "loss_align": 0.5217831134796143, "loss_varcov": 0.059790194034576416, "loss_mean": 0.7851623296737671, "loss_mean_diff": 0.011718355119228363, "mean_diff_l2": 0.10825134813785553, "loss_pair": 9.19812882784754e-05, "loss_lang": 1.0448167324066162, "loss_len": 0.023677609860897064, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.62890625, "len_acc_clf_detached_mix": 0.9453125, "adv_buf_fill": 4096, "elapsed_sec": 357.8014802932739}
43
+ {"step": 2860, "epoch": 1, "loss": 7.620456218719482, "loss_render": 5.694301605224609, "loss_align": 0.2016521394252777, "loss_varcov": 0.04285199195146561, "loss_mean": 0.45632511377334595, "loss_mean_diff": 0.0028114947490394115, "mean_diff_l2": 0.053023528307676315, "loss_pair": 3.0967334168963134e-05, "loss_lang": 2.2187979221343994, "loss_len": 0.38237425684928894, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.875, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.875, "lang_acc_clf_detached_mix": 0.60546875, "len_acc_clf_detached_mix": 0.90625, "adv_buf_fill": 4096, "elapsed_sec": 365.1789894104004}
44
+ {"step": 2880, "epoch": 1, "loss": 8.542922973632812, "loss_render": 5.394974708557129, "loss_align": 1.2641007900238037, "loss_varcov": 0.05115702748298645, "loss_mean": 0.6198285818099976, "loss_mean_diff": 0.13154421746730804, "mean_diff_l2": 0.3626902401447296, "loss_pair": 0.0006327533628791571, "loss_lang": 1.3350632190704346, "loss_len": 0.8881285786628723, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.6171875, "len_acc_clf_detached_mix": 0.9296875, "adv_buf_fill": 4096, "elapsed_sec": 373.69257283210754}
45
+ {"step": 2900, "epoch": 1, "loss": 6.601009845733643, "loss_render": 4.989941120147705, "loss_align": 0.23544928431510925, "loss_varcov": 0.05719364434480667, "loss_mean": 0.7462212443351746, "loss_mean_diff": 0.006663067266345024, "mean_diff_l2": 0.08162761479616165, "loss_pair": 3.4855256672017276e-05, "loss_lang": 0.7852693200111389, "loss_len": 0.603488564491272, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.57421875, "len_acc_clf_detached_mix": 0.92578125, "adv_buf_fill": 4096, "elapsed_sec": 381.6709158420563}
46
+ {"step": 2920, "epoch": 1, "loss": 6.384342193603516, "loss_render": 5.365021228790283, "loss_align": 0.09196941554546356, "loss_varcov": 0.043040093034505844, "loss_mean": 0.48713672161102295, "loss_mean_diff": 0.006877743639051914, "mean_diff_l2": 0.08293215930461884, "loss_pair": 5.3803691116627306e-05, "loss_lang": 0.5152645111083984, "loss_len": 0.4211586117744446, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.875, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.875, "lang_acc_clf_detached_mix": 0.62109375, "len_acc_clf_detached_mix": 0.921875, "adv_buf_fill": 4096, "elapsed_sec": 389.54567074775696}
47
+ {"step": 2940, "epoch": 1, "loss": 7.175053119659424, "loss_render": 4.871315002441406, "loss_align": 0.8545704483985901, "loss_varcov": 0.054677318781614304, "loss_mean": 0.6831730604171753, "loss_mean_diff": 0.02581837587058544, "mean_diff_l2": 0.16068097949028015, "loss_pair": 0.00020444518304429948, "loss_lang": 0.8429899215698242, "loss_len": 0.6780354380607605, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.66015625, "len_acc_clf_detached_mix": 0.92578125, "adv_buf_fill": 4096, "elapsed_sec": 397.55750370025635}
48
+ {"step": 2960, "epoch": 1, "loss": 6.455283164978027, "loss_render": 5.187819957733154, "loss_align": 0.24699977040290833, "loss_varcov": 0.04577154293656349, "loss_mean": 0.5125292539596558, "loss_mean_diff": 0.015523126348853111, "mean_diff_l2": 0.12459184229373932, "loss_pair": 0.00011908051237696782, "loss_lang": 1.318716287612915, "loss_len": 0.09871095418930054, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.61328125, "len_acc_clf_detached_mix": 0.90234375, "adv_buf_fill": 4096, "elapsed_sec": 405.3400025367737}
49
+ {"step": 2980, "epoch": 1, "loss": 6.27493953704834, "loss_render": 5.295500755310059, "loss_align": 0.3945789039134979, "loss_varcov": 0.050448641180992126, "loss_mean": 0.6098806858062744, "loss_mean_diff": 0.025134362280368805, "mean_diff_l2": 0.15853820741176605, "loss_pair": 0.00016448323731310666, "loss_lang": 0.40522223711013794, "loss_len": 0.06974945217370987, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.75, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.75, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.640625, "len_acc_clf_detached_mix": 0.91796875, "adv_buf_fill": 4096, "elapsed_sec": 413.14878726005554}
50
+ {"step": 3000, "epoch": 1, "loss": 7.421226501464844, "loss_render": 5.300811767578125, "loss_align": 0.014522181823849678, "loss_varcov": 0.04150805622339249, "loss_mean": 0.45071732997894287, "loss_mean_diff": 0.025674935430288315, "mean_diff_l2": 0.1602340042591095, "loss_pair": 0.0001647394528845325, "loss_lang": 2.5260496139526367, "loss_len": 0.6107903122901917, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.609375, "len_acc_clf_detached_mix": 0.94921875, "adv_buf_fill": 4096, "elapsed_sec": 420.91780281066895}
planB_polish/probe_leakage.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish/ckpt.pt",
3
+ "max_train_groups": 8000,
4
+ "max_valid_groups": 946,
5
+ "probe_hidden": 256,
6
+ "epochs": 20,
7
+ "lr": 0.01,
8
+ "baseline_majority_lang_acc": 0.5,
9
+ "baseline_majority_len_acc": 0.9271249771118164,
10
+ "lang_probe": {
11
+ "train_acc": 0.5943125,
12
+ "valid_acc": 0.5813953488372093,
13
+ "best_valid_acc": 0.5887949260042283
14
+ },
15
+ "len_probe": {
16
+ "train_acc": 0.92875,
17
+ "valid_acc": 0.9011627906976745,
18
+ "best_valid_acc": 0.9011627906976745
19
+ },
20
+ "lang_probe_length_matched": {
21
+ "train_acc": 0.5951952723535457,
22
+ "valid_acc": 0.5859977949283351,
23
+ "best_valid_acc": 0.601433296582139
24
+ },
25
+ "lang_bucketwise_valid_acc": {
26
+ "bucket_0_acc": 0.40625,
27
+ "bucket_1_acc": 0.6000000238418579,
28
+ "bucket_2_acc": 0.5565217137336731,
29
+ "bucket_3_acc": 0.5859237909317017
30
+ },
31
+ "length_matched_sizes": {
32
+ "train": 15568,
33
+ "valid": 1814
34
+ }
35
+ }
planB_polish_robust_dec/ckpt.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:569ebe88c92cc95a760251489ef7d91912d9c6773060b67634c4806d80d0cd65
3
+ size 3465979330
planB_polish_robust_dec/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_jsonl": "data/groups_train.jsonl",
3
+ "valid_jsonl": "data/groups_valid.jsonl",
4
+ "run_dir": "runs/planB_polish_robust_dec",
5
+ "resume": "runs/planB_polish/ckpt.pt",
6
+ "backbone": "google/mt5-small",
7
+ "num_latents": 16,
8
+ "latent_dropout": 0.1,
9
+ "latent_noise_std": 0.01,
10
+ "batch_size": 4,
11
+ "grad_accum": 8,
12
+ "epochs": 1,
13
+ "max_doc_len": 256,
14
+ "max_sum_len": 64,
15
+ "eval_every": 400,
16
+ "max_train_examples": 2000,
17
+ "max_valid_examples": 200,
18
+ "lambda_align": 0.0,
19
+ "tau": 0.07,
20
+ "lambda_varcov": 0.0,
21
+ "var_target_std": 0.05,
22
+ "lambda_mean": 0.0,
23
+ "lambda_mean_diff": 0.0,
24
+ "lambda_pair": 0.0,
25
+ "lambda_lang": 0.0,
26
+ "lambda_len": 0.0,
27
+ "adv_start_step": 999999,
28
+ "grl_alpha": 1.0,
29
+ "grl_warmup": 200,
30
+ "lr_model": 0.0001,
31
+ "lr_lang": 0.001,
32
+ "lr_len": 0.001,
33
+ "adv_clf_steps": 4,
34
+ "adv_clf_weight_decay": 0.0,
35
+ "adv_queue_size": 4096,
36
+ "adv_clf_batch": 256,
37
+ "adv_mix_current": 0.5,
38
+ "decoder_only": true,
39
+ "resume_model_only": true
40
+ }
planB_polish_robust_dec/diag.jsonl ADDED
@@ -0,0 +1 @@
 
 
1
+ {"step": 3200, "nll_en": 4.768383893966675, "nll_zh": 4.778253097534179, "nll": 4.773318495750427, "lang_acc_valid_evalmode": 0.5, "len_acc_valid_evalmode": 0.895}
planB_polish_robust_dec/diag_full.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec/ckpt.pt",
3
+ "nll_en": 4.762824895265995,
4
+ "nll_zh": 4.769128660318968,
5
+ "swap_delta_en": 0.49594723699460824,
6
+ "swap_delta_zh": 0.4821954438852709,
7
+ "ablate_zero_delta_en": 0.7210078239440918,
8
+ "ablate_mean_delta_en": 0.22742677491008606,
9
+ "ablate_noise_delta_en": 2.953899601297197,
10
+ "ablate_zero_delta_zh": 0.714704058891119,
11
+ "ablate_mean_delta_zh": 0.2206332567630308,
12
+ "ablate_noise_delta_zh": 3.031246384908987,
13
+ "inv_top1_full": 0.0613107830286026,
14
+ "inv_top5_full": 0.2050739973783493,
15
+ "diag_sim_mean": 0.9653781652450562,
16
+ "offdiag_sim_mean": 0.4495665431022644,
17
+ "sim_margin": 0.5158116221427917,
18
+ "n_valid": 946
19
+ }
planB_polish_robust_dec/logs.jsonl ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3020, "epoch": 0, "loss": 5.631680965423584, "loss_render": 5.631680965423584, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.6150376796722412, "loss_mean_diff": 0.011596184223890305, "mean_diff_l2": 0.10768558084964752, "loss_pair": 0.00011448925943113863, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 160, "elapsed_sec": 3.6211800575256348}
2
+ {"step": 3040, "epoch": 0, "loss": 5.410102844238281, "loss_render": 5.410102844238281, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.6298912763595581, "loss_mean_diff": 0.06626003980636597, "mean_diff_l2": 0.2574102580547333, "loss_pair": 0.0003280025557614863, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 320, "elapsed_sec": 6.6114888191223145}
3
+ {"step": 3060, "epoch": 0, "loss": 5.64467716217041, "loss_render": 5.64467716217041, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.47115084528923035, "loss_mean_diff": 0.011376983486115932, "mean_diff_l2": 0.10666294395923615, "loss_pair": 7.776531856507063e-05, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 480, "elapsed_sec": 9.741488933563232}
4
+ {"step": 3080, "epoch": 0, "loss": 6.101323127746582, "loss_render": 6.101323127746582, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.4650561809539795, "loss_mean_diff": 0.006715109571814537, "mean_diff_l2": 0.08194576948881149, "loss_pair": 6.295090133789927e-05, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 640, "elapsed_sec": 12.95272970199585}
5
+ {"step": 3100, "epoch": 0, "loss": 5.176516056060791, "loss_render": 5.176516056060791, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.5547142028808594, "loss_mean_diff": 0.009514890611171722, "mean_diff_l2": 0.09754429757595062, "loss_pair": 9.414715896127746e-05, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 800, "elapsed_sec": 16.11647129058838}
6
+ {"step": 3120, "epoch": 0, "loss": 4.753480911254883, "loss_render": 4.753480911254883, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.5141536593437195, "loss_mean_diff": 0.03376246988773346, "mean_diff_l2": 0.18374566733837128, "loss_pair": 0.00023208693892229348, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 960, "elapsed_sec": 19.344083786010742}
7
+ {"step": 3140, "epoch": 0, "loss": 4.908468246459961, "loss_render": 4.908468246459961, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.6493403911590576, "loss_mean_diff": 0.01209488045424223, "mean_diff_l2": 0.10997672379016876, "loss_pair": 0.00011903270205948502, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 1120, "elapsed_sec": 22.361193895339966}
8
+ {"step": 3160, "epoch": 0, "loss": 4.953296661376953, "loss_render": 4.953296661376953, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.5684100389480591, "loss_mean_diff": 0.009280805476009846, "mean_diff_l2": 0.09633693844079971, "loss_pair": 7.268520857905969e-05, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 1280, "elapsed_sec": 25.544042825698853}
9
+ {"step": 3180, "epoch": 0, "loss": 5.151930332183838, "loss_render": 5.151930332183838, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.7015585899353027, "loss_mean_diff": 0.016968771815299988, "mean_diff_l2": 0.13026423752307892, "loss_pair": 0.00017023770487867296, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 1440, "elapsed_sec": 28.683798789978027}
10
+ {"step": 3200, "epoch": 0, "loss": 4.695162773132324, "loss_render": 4.695162773132324, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.6107100248336792, "loss_mean_diff": 0.01986495405435562, "mean_diff_l2": 0.14094309508800507, "loss_pair": 0.00010516037582419813, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 1600, "elapsed_sec": 31.837865352630615}
11
+ {"step": 3220, "epoch": 0, "loss": 5.15194034576416, "loss_render": 5.15194034576416, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.4896160960197449, "loss_mean_diff": 0.04858623817563057, "mean_diff_l2": 0.2204228639602661, "loss_pair": 0.00023253000108525157, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 1760, "elapsed_sec": 45.492348194122314}
12
+ {"step": 3240, "epoch": 0, "loss": 4.98980712890625, "loss_render": 4.98980712890625, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.5042635798454285, "loss_mean_diff": 0.03962373360991478, "mean_diff_l2": 0.19905711710453033, "loss_pair": 0.00028840411687269807, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 1920, "elapsed_sec": 48.80889177322388}
13
+ {"step": 3260, "epoch": 0, "loss": 5.4674296379089355, "loss_render": 5.4674296379089355, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.6475939154624939, "loss_mean_diff": 0.009448235854506493, "mean_diff_l2": 0.09720203280448914, "loss_pair": 0.00011519801046233624, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 2080, "elapsed_sec": 51.9842643737793}
14
+ {"step": 3280, "epoch": 0, "loss": 4.422120094299316, "loss_render": 4.422120094299316, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.7420861721038818, "loss_mean_diff": 0.01552241574972868, "mean_diff_l2": 0.1245889887213707, "loss_pair": 9.256125485990196e-05, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 2240, "elapsed_sec": 55.44223189353943}
15
+ {"step": 3300, "epoch": 0, "loss": 5.029345512390137, "loss_render": 5.029345512390137, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.5954892635345459, "loss_mean_diff": 0.026383597403764725, "mean_diff_l2": 0.1624302864074707, "loss_pair": 0.00018176088633481413, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 2400, "elapsed_sec": 58.97809457778931}
16
+ {"step": 3320, "epoch": 0, "loss": 4.823001861572266, "loss_render": 4.823001861572266, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.769568920135498, "loss_mean_diff": 0.02784150466322899, "mean_diff_l2": 0.1668577343225479, "loss_pair": 0.00022472930140793324, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 2560, "elapsed_sec": 62.59023070335388}
17
+ {"step": 3340, "epoch": 0, "loss": 4.812470436096191, "loss_render": 4.812470436096191, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.5672426819801331, "loss_mean_diff": 0.010172193869948387, "mean_diff_l2": 0.10085729509592056, "loss_pair": 8.846040873322636e-05, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 2720, "elapsed_sec": 66.0759379863739}
18
+ {"step": 3360, "epoch": 0, "loss": 5.188187122344971, "loss_render": 5.188187122344971, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.5040833353996277, "loss_mean_diff": 0.007734925020486116, "mean_diff_l2": 0.08794841915369034, "loss_pair": 7.332015957217664e-05, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 2880, "elapsed_sec": 69.63828778266907}
19
+ {"step": 3380, "epoch": 0, "loss": 4.697408199310303, "loss_render": 4.697408199310303, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.4333137273788452, "loss_mean_diff": 0.03503121808171272, "mean_diff_l2": 0.18716628849506378, "loss_pair": 0.00027789041632786393, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 3040, "elapsed_sec": 73.0626335144043}
20
+ {"step": 3400, "epoch": 0, "loss": 5.154892921447754, "loss_render": 5.154892921447754, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.4505867063999176, "loss_mean_diff": 0.00972425565123558, "mean_diff_l2": 0.09861163794994354, "loss_pair": 7.382083276752383e-05, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 3200, "elapsed_sec": 76.61041808128357}
21
+ {"step": 3420, "epoch": 0, "loss": 4.900773048400879, "loss_render": 4.900773048400879, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.7373412847518921, "loss_mean_diff": 0.16094563901424408, "mean_diff_l2": 0.40118029713630676, "loss_pair": 0.0005682703922502697, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 3360, "elapsed_sec": 80.04734826087952}
22
+ {"step": 3440, "epoch": 0, "loss": 4.860998153686523, "loss_render": 4.860998153686523, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.4975969195365906, "loss_mean_diff": 0.05051785334944725, "mean_diff_l2": 0.22476176917552948, "loss_pair": 0.00029846897814422846, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 3520, "elapsed_sec": 83.51609563827515}
23
+ {"step": 3460, "epoch": 0, "loss": 4.804428577423096, "loss_render": 4.804428577423096, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.46031278371810913, "loss_mean_diff": 0.08767025917768478, "mean_diff_l2": 0.29609164595603943, "loss_pair": 0.0003665403928607702, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 3680, "elapsed_sec": 86.96110343933105}
24
+ {"step": 3480, "epoch": 0, "loss": 5.394003391265869, "loss_render": 5.394003391265869, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.44830983877182007, "loss_mean_diff": 0.09439471364021301, "mean_diff_l2": 0.3072372376918793, "loss_pair": 0.0006847942713648081, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 3840, "elapsed_sec": 90.47929072380066}
25
+ {"step": 3500, "epoch": 0, "loss": 5.047396659851074, "loss_render": 5.047396659851074, "loss_align": 0.0, "loss_varcov": 0.0, "loss_mean": 0.5032382011413574, "loss_mean_diff": 0.022277912124991417, "mean_diff_l2": 0.1492578685283661, "loss_pair": 0.00020781921921297908, "loss_lang": 0.0, "loss_len": 0.0, "grl_alpha_eff": 0.0, "lang_acc_batch": 0.0, "len_acc_batch": 0.0, "lang_acc_cur_detached": 0.0, "len_acc_cur_detached": 0.0, "lang_acc_clf_detached_mix": 0.0, "len_acc_clf_detached_mix": 0.0, "adv_buf_fill": 4000, "elapsed_sec": 93.89007544517517}
planB_polish_robust_dec/probe_leakage.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec/ckpt.pt",
3
+ "max_train_groups": 8000,
4
+ "max_valid_groups": 946,
5
+ "probe_hidden": 256,
6
+ "epochs": 20,
7
+ "lr": 0.01,
8
+ "baseline_majority_lang_acc": 0.5,
9
+ "baseline_majority_len_acc": 0.9271249771118164,
10
+ "lang_probe": {
11
+ "train_acc": 0.588125,
12
+ "valid_acc": 0.587737843551797,
13
+ "best_valid_acc": 0.5887949260042283
14
+ },
15
+ "len_probe": {
16
+ "train_acc": 0.9288125,
17
+ "valid_acc": 0.9011627906976745,
18
+ "best_valid_acc": 0.9011627906976745
19
+ },
20
+ "lang_probe_length_matched": {
21
+ "train_acc": 0.5892857142857143,
22
+ "valid_acc": 0.5821389195148843,
23
+ "best_valid_acc": 0.5970231532524807
24
+ },
25
+ "lang_bucketwise_valid_acc": {
26
+ "bucket_0_acc": 0.4375,
27
+ "bucket_1_acc": 0.6500000357627869,
28
+ "bucket_2_acc": 0.5999999642372131,
29
+ "bucket_3_acc": 0.588269829750061
30
+ },
31
+ "length_matched_sizes": {
32
+ "train": 15568,
33
+ "valid": 1814
34
+ }
35
+ }
planB_polish_robust_dec_long/ckpt.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:163ab2ec23a27194d109582378a85338c4d418a8b67c5a0637325438081202e6
3
+ size 2439403193
planB_polish_robust_dec_long/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_jsonl": "data/groups_train.jsonl",
3
+ "valid_jsonl": "data/groups_valid.jsonl",
4
+ "run_dir": "runs/planB_polish_robust_dec_long",
5
+ "resume": "runs/planB_polish/ckpt.pt",
6
+ "backbone": "google/mt5-small",
7
+ "num_latents": 16,
8
+ "batch_size": 4,
9
+ "grad_accum": 8,
10
+ "epochs": 3,
11
+ "max_doc_len": 256,
12
+ "max_sum_len": 64,
13
+ "eval_every": 400,
14
+ "max_train_examples": 0,
15
+ "max_valid_examples": 0,
16
+ "lr": 0.0003,
17
+ "weight_decay": 0.0,
18
+ "latent_dropout": 0.1,
19
+ "latent_noise_std": 0.01,
20
+ "noise_warmup_steps": 500,
21
+ "noise_warmup_start_step": -1,
22
+ "untie_lm_head": true,
23
+ "seed": 42
24
+ }
planB_polish_robust_dec_long/diag.jsonl ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3200, "nll_en": 4.738007259368897, "nll_zh": 4.747281007766723, "nll": 4.7426441335678104, "lang_acc_valid_evalmode": 0.5, "len_acc_valid_evalmode": 0.025}
2
+ {"step": 3600, "nll_en": 4.631216464042663, "nll_zh": 4.639870862960816, "nll": 4.635543663501739, "lang_acc_valid_evalmode": 0.5, "len_acc_valid_evalmode": 0.025}
3
+ {"step": 4000, "nll_en": 4.5431574010849, "nll_zh": 4.552404751777649, "nll": 4.547781076431274, "lang_acc_valid_evalmode": 0.5, "len_acc_valid_evalmode": 0.025}
4
+ {"step": 4400, "nll_en": 4.454369354248047, "nll_zh": 4.4649154615402225, "nll": 4.459642407894134, "lang_acc_valid_evalmode": 0.5, "len_acc_valid_evalmode": 0.025}
5
+ {"step": 3200, "nll_en": 4.726770976633149, "nll_zh": 4.732470808775178, "nll": 4.729620892704164}
6
+ {"step": 3600, "nll_en": 4.3129979188044025, "nll_zh": 4.315242195734262, "nll": 4.314120057269332}
7
+ {"step": 4000, "nll_en": 4.090272497173092, "nll_zh": 4.092214378695659, "nll": 4.091243437934375}
8
+ {"step": 4400, "nll_en": 3.9515462909642034, "nll_zh": 3.960411111093727, "nll": 3.955978701028965}
9
+ {"step": 4800, "nll_en": 3.8644339347742576, "nll_zh": 3.8680522003335125, "nll": 3.866243067553885}
10
+ {"step": 5200, "nll_en": 3.771670523960041, "nll_zh": 3.775045036009452, "nll": 3.7733577799847464}
11
+ {"step": 5600, "nll_en": 3.712917871253435, "nll_zh": 3.714967165134422, "nll": 3.7139425181939285}
12
+ {"step": 6000, "nll_en": 3.6640999705292456, "nll_zh": 3.6683099668071337, "nll": 3.6662049686681897}
13
+ {"step": 6400, "nll_en": 3.617421502038742, "nll_zh": 3.6206912177813733, "nll": 3.619056359910058}
14
+ {"step": 6800, "nll_en": 3.5835763294147385, "nll_zh": 3.582717639447267, "nll": 3.5831469844310027}
15
+ {"step": 7200, "nll_en": 3.5504482372114574, "nll_zh": 3.5543588503218606, "nll": 3.552403543766659}
16
+ {"step": 7600, "nll_en": 3.772425605177123, "nll_zh": 3.7710336318227755, "nll": 3.7717296184999496}
17
+ {"step": 8000, "nll_en": 3.5049120185239118, "nll_zh": 3.5125938016314837, "nll": 3.508752910077698}
18
+ {"step": 8400, "nll_en": 3.471799067160544, "nll_zh": 3.4740455528646392, "nll": 3.4729223100125917}
19
+ {"step": 8800, "nll_en": 3.4548070738229892, "nll_zh": 3.458574728532271, "nll": 3.45669090117763}
20
+ {"step": 9200, "nll_en": 3.439204857162895, "nll_zh": 3.444061423456946, "nll": 3.4416331403099205}
21
+ {"step": 9600, "nll_en": 3.4242512277518466, "nll_zh": 3.425472506509019, "nll": 3.4248618671304327}
22
+ {"step": 10000, "nll_en": 3.402887813132611, "nll_zh": 3.4049878422856077, "nll": 3.4039378277091092}
23
+ {"step": 10400, "nll_en": 3.3928586494090944, "nll_zh": 3.3954227671310715, "nll": 3.394140708270083}
24
+ {"step": 10800, "nll_en": 3.382973665918911, "nll_zh": 3.3827065604937756, "nll": 3.3828401132063433}
25
+ {"step": 11200, "nll_en": 3.3614858260366427, "nll_zh": 3.3621233714307337, "nll": 3.361804598733688}
26
+ {"step": 11600, "nll_en": 3.3495029636971796, "nll_zh": 3.352324916242797, "nll": 3.3509139399699883}
27
+ {"step": 12000, "nll_en": 3.3784578234650366, "nll_zh": 3.3760307596299413, "nll": 3.377244291547489}
28
+ {"step": 12400, "nll_en": 3.32743300040731, "nll_zh": 3.33091990388221, "nll": 3.3291764521447598}
29
+ {"step": 12800, "nll_en": 3.326908253716112, "nll_zh": 3.3272379780420542, "nll": 3.327073115879083}
30
+ {"step": 13200, "nll_en": 3.3118652299393054, "nll_zh": 3.3161710845995653, "nll": 3.3140181572694356}
31
+ {"step": 13600, "nll_en": 3.30334316701264, "nll_zh": 3.3069664610364726, "nll": 3.305154814024556}
32
+ {"step": 14000, "nll_en": 3.292484433656011, "nll_zh": 3.296051383270522, "nll": 3.2942679084632664}
33
+ {"step": 14400, "nll_en": 3.2919716270479036, "nll_zh": 3.292308284162719, "nll": 3.2921399556053115}
34
+ {"step": 14800, "nll_en": 3.2743541433241603, "nll_zh": 3.2755244357893383, "nll": 3.274939289556749}
35
+ {"step": 15200, "nll_en": 3.2750622061795966, "nll_zh": 3.2758719724530145, "nll": 3.275467089316306}
36
+ {"step": 15600, "nll_en": 3.2596704027365426, "nll_zh": 3.262144302464943, "nll": 3.260907352600743}
37
+ {"step": 16000, "nll_en": 3.257569935054658, "nll_zh": 3.2588827625137555, "nll": 3.258226348784207}
38
+ {"step": 16400, "nll_en": 3.2444708402766738, "nll_zh": 3.2454178025808194, "nll": 3.2449443214287466}
planB_polish_robust_dec_long/diag_full.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_long/ckpt.pt",
3
+ "nll_en": 3.244555667893327,
4
+ "nll_zh": 3.2475476688360816,
5
+ "swap_delta_en": 0.5506592724338394,
6
+ "swap_delta_zh": 0.5321295639425201,
7
+ "ablate_zero_delta_en": 2.462635235857006,
8
+ "ablate_mean_delta_en": 0.21426737484921865,
9
+ "ablate_noise_delta_en": 7.151853417241296,
10
+ "ablate_zero_delta_zh": 2.4596432349142514,
11
+ "ablate_mean_delta_zh": 0.20696192009504452,
12
+ "ablate_noise_delta_zh": 7.240511431502489,
13
+ "inv_top1_full": 0.060253698378801346,
14
+ "inv_top5_full": 0.20401690900325775,
15
+ "diag_sim_mean": 0.965337336063385,
16
+ "offdiag_sim_mean": 0.4495629370212555,
17
+ "sim_margin": 0.5157743990421295,
18
+ "n_valid": 946
19
+ }
planB_polish_robust_dec_long/logs.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
planB_polish_robust_dec_long/probe_leakage.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_long/ckpt.pt",
3
+ "max_train_groups": 8000,
4
+ "max_valid_groups": 946,
5
+ "probe_hidden": 256,
6
+ "epochs": 20,
7
+ "lr": 0.01,
8
+ "baseline_majority_lang_acc": 0.5,
9
+ "baseline_majority_len_acc": 0.9271249771118164,
10
+ "lang_probe": {
11
+ "train_acc": 0.5870625,
12
+ "valid_acc": 0.5824524312896406,
13
+ "best_valid_acc": 0.6014799154334038
14
+ },
15
+ "len_probe": {
16
+ "train_acc": 0.92875,
17
+ "valid_acc": 0.9011627906976745,
18
+ "best_valid_acc": 0.9011627906976745
19
+ },
20
+ "lang_probe_length_matched": {
21
+ "train_acc": 0.5869732785200411,
22
+ "valid_acc": 0.5854465270121278,
23
+ "best_valid_acc": 0.5959206174200662
24
+ },
25
+ "lang_bucketwise_valid_acc": {
26
+ "bucket_0_acc": 0.34375,
27
+ "bucket_1_acc": 0.6000000238418579,
28
+ "bucket_2_acc": 0.530434787273407,
29
+ "bucket_3_acc": 0.5900293588638306
30
+ },
31
+ "length_matched_sizes": {
32
+ "train": 15568,
33
+ "valid": 1814
34
+ }
35
+ }
planB_polish_robust_dec_noise02/ckpt.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:403aedb48676cb25dc625a5d460d68744ad354b74d154b15d3aaf6f4127c09db
3
+ size 2439403193
planB_polish_robust_dec_noise02/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_jsonl": "data/groups_train.jsonl",
3
+ "valid_jsonl": "data/groups_valid.jsonl",
4
+ "run_dir": "runs/planB_polish_robust_dec_noise02",
5
+ "resume": "runs/planB_polish/ckpt.pt",
6
+ "backbone": "google/mt5-small",
7
+ "num_latents": 16,
8
+ "batch_size": 4,
9
+ "grad_accum": 8,
10
+ "epochs": 2,
11
+ "max_doc_len": 256,
12
+ "max_sum_len": 64,
13
+ "eval_every": 400,
14
+ "max_train_examples": 0,
15
+ "max_valid_examples": 0,
16
+ "lr": 0.0003,
17
+ "weight_decay": 0.0,
18
+ "latent_dropout": 0.1,
19
+ "latent_noise_std": 0.02,
20
+ "noise_warmup_steps": 800,
21
+ "noise_warmup_start_step": -1,
22
+ "untie_lm_head": true,
23
+ "seed": 42
24
+ }
planB_polish_robust_dec_noise02/diag.jsonl ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3200, "nll_en": 4.726687601706442, "nll_zh": 4.732378584637955, "nll": 4.729533093172198}
2
+ {"step": 3600, "nll_en": 4.314645891209738, "nll_zh": 4.315904144504862, "nll": 4.315275017857299}
3
+ {"step": 4000, "nll_en": 4.094815027386139, "nll_zh": 4.096858743381299, "nll": 4.095836885383719}
4
+ {"step": 4400, "nll_en": 3.9570389590354096, "nll_zh": 3.96577752316981, "nll": 3.96140824110261}
5
+ {"step": 4800, "nll_en": 3.8651220460774782, "nll_zh": 3.8686307066338763, "nll": 3.8668763763556773}
6
+ {"step": 5200, "nll_en": 3.771388160753956, "nll_zh": 3.7728675989469567, "nll": 3.7721278798504563}
7
+ {"step": 5600, "nll_en": 3.717093608848136, "nll_zh": 3.7174127762211806, "nll": 3.7172531925346584}
8
+ {"step": 6000, "nll_en": 3.660282932426662, "nll_zh": 3.66561883095707, "nll": 3.662950881691866}
9
+ {"step": 6400, "nll_en": 3.614995673141318, "nll_zh": 3.6189114810548424, "nll": 3.61695357709808}
10
+ {"step": 6800, "nll_en": 3.5807273443355117, "nll_zh": 3.579956094003883, "nll": 3.5803417191696973}
11
+ {"step": 7200, "nll_en": 3.5464490473144403, "nll_zh": 3.549751218227201, "nll": 3.548100132770821}
12
+ {"step": 7600, "nll_en": 3.756626025314815, "nll_zh": 3.7557674512802675, "nll": 3.756196738297541}
13
+ {"step": 8000, "nll_en": 3.50723839511831, "nll_zh": 3.51329014013988, "nll": 3.510264267629095}
14
+ {"step": 8400, "nll_en": 3.476454969692432, "nll_zh": 3.476457744017716, "nll": 3.4764563568550737}
15
+ {"step": 8800, "nll_en": 3.4499150703623735, "nll_zh": 3.4546108941447153, "nll": 3.4522629822535444}
16
+ {"step": 9200, "nll_en": 3.4374389315760414, "nll_zh": 3.4416958010473917, "nll": 3.4395673663117163}
17
+ {"step": 9600, "nll_en": 3.422439192112606, "nll_zh": 3.4235046308085986, "nll": 3.422971911460602}
18
+ {"step": 10000, "nll_en": 3.4043741165709545, "nll_zh": 3.4067958989052642, "nll": 3.4055850077381096}
19
+ {"step": 10400, "nll_en": 3.390590481223818, "nll_zh": 3.393908268805538, "nll": 3.392249375014678}
20
+ {"step": 10800, "nll_en": 3.380558049199949, "nll_zh": 3.380652252261785, "nll": 3.380605150730867}
21
+ {"step": 11200, "nll_en": 3.364188308191602, "nll_zh": 3.3654494083961017, "nll": 3.364818858293852}
22
+ {"step": 11600, "nll_en": 3.3440420713283547, "nll_zh": 3.3479478162640497, "nll": 3.345994943796202}
planB_polish_robust_dec_noise02/diag_full.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_noise02/ckpt.pt",
3
+ "nll_en": 3.341824184764515,
4
+ "nll_zh": 3.3433680100874468,
5
+ "swap_delta_en": 0.5562419346976734,
6
+ "swap_delta_zh": 0.5396175626468457,
7
+ "ablate_zero_delta_en": 1.5730591435261054,
8
+ "ablate_mean_delta_en": 0.22105616696541203,
9
+ "ablate_noise_delta_en": 6.751106149298444,
10
+ "ablate_zero_delta_zh": 1.571515318203174,
11
+ "ablate_mean_delta_zh": 0.21421958528159787,
12
+ "ablate_noise_delta_zh": 6.816207936147807,
13
+ "inv_top1_full": 0.060253698378801346,
14
+ "inv_top5_full": 0.20401690900325775,
15
+ "diag_sim_mean": 0.965337336063385,
16
+ "offdiag_sim_mean": 0.4495629370212555,
17
+ "sim_margin": 0.5157743990421295,
18
+ "n_valid": 946
19
+ }
planB_polish_robust_dec_noise02/logs.jsonl ADDED
@@ -0,0 +1,448 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3020, "epoch": 0, "loss": 5.796799659729004, "loss_en": 5.743430137634277, "loss_zh": 5.850168704986572, "noise_std_eff": 0.0005, "elapsed_sec": 3.4203007221221924}
2
+ {"step": 3040, "epoch": 0, "loss": 5.2285051345825195, "loss_en": 5.177862644195557, "loss_zh": 5.279147624969482, "noise_std_eff": 0.001, "elapsed_sec": 6.412633419036865}
3
+ {"step": 3060, "epoch": 0, "loss": 5.252622604370117, "loss_en": 5.334871292114258, "loss_zh": 5.170373439788818, "noise_std_eff": 0.0015, "elapsed_sec": 9.435014724731445}
4
+ {"step": 3080, "epoch": 0, "loss": 5.561959266662598, "loss_en": 5.54374361038208, "loss_zh": 5.580174922943115, "noise_std_eff": 0.002, "elapsed_sec": 12.517982244491577}
5
+ {"step": 3100, "epoch": 0, "loss": 5.2716546058654785, "loss_en": 5.259927272796631, "loss_zh": 5.283381938934326, "noise_std_eff": 0.0025, "elapsed_sec": 15.546177864074707}
6
+ {"step": 3120, "epoch": 0, "loss": 5.621967315673828, "loss_en": 5.688499927520752, "loss_zh": 5.5554351806640625, "noise_std_eff": 0.003, "elapsed_sec": 18.55265736579895}
7
+ {"step": 3140, "epoch": 0, "loss": 5.172842025756836, "loss_en": 5.133390426635742, "loss_zh": 5.212294101715088, "noise_std_eff": 0.0034999999999999996, "elapsed_sec": 21.545820474624634}
8
+ {"step": 3160, "epoch": 0, "loss": 5.827211380004883, "loss_en": 5.787778377532959, "loss_zh": 5.866644382476807, "noise_std_eff": 0.004, "elapsed_sec": 24.571044445037842}
9
+ {"step": 3180, "epoch": 0, "loss": 5.132044792175293, "loss_en": 5.168938636779785, "loss_zh": 5.095150947570801, "noise_std_eff": 0.0045000000000000005, "elapsed_sec": 27.544373512268066}
10
+ {"step": 3200, "epoch": 0, "loss": 5.196186065673828, "loss_en": 5.2775750160217285, "loss_zh": 5.1147966384887695, "noise_std_eff": 0.005, "elapsed_sec": 30.58199715614319}
11
+ {"step": 3220, "epoch": 0, "loss": 5.000565052032471, "loss_en": 4.988380432128906, "loss_zh": 5.012749671936035, "noise_std_eff": 0.0055000000000000005, "elapsed_sec": 50.32763361930847}
12
+ {"step": 3240, "epoch": 0, "loss": 4.9852495193481445, "loss_en": 4.977023601531982, "loss_zh": 4.993475437164307, "noise_std_eff": 0.006, "elapsed_sec": 53.25921678543091}
13
+ {"step": 3260, "epoch": 0, "loss": 4.761623382568359, "loss_en": 4.745314121246338, "loss_zh": 4.777932167053223, "noise_std_eff": 0.006500000000000001, "elapsed_sec": 56.15965437889099}
14
+ {"step": 3280, "epoch": 0, "loss": 5.302664279937744, "loss_en": 5.288912296295166, "loss_zh": 5.316416263580322, "noise_std_eff": 0.006999999999999999, "elapsed_sec": 59.13148641586304}
15
+ {"step": 3300, "epoch": 0, "loss": 4.5848588943481445, "loss_en": 4.574359893798828, "loss_zh": 4.595358371734619, "noise_std_eff": 0.0075, "elapsed_sec": 62.05224251747131}
16
+ {"step": 3320, "epoch": 0, "loss": 4.253688335418701, "loss_en": 4.261787414550781, "loss_zh": 4.245589256286621, "noise_std_eff": 0.008, "elapsed_sec": 64.99401092529297}
17
+ {"step": 3340, "epoch": 0, "loss": 4.887582302093506, "loss_en": 4.914684295654297, "loss_zh": 4.860480308532715, "noise_std_eff": 0.0085, "elapsed_sec": 67.99308729171753}
18
+ {"step": 3360, "epoch": 0, "loss": 4.343026161193848, "loss_en": 4.3106489181518555, "loss_zh": 4.375402927398682, "noise_std_eff": 0.009000000000000001, "elapsed_sec": 70.93724727630615}
19
+ {"step": 3380, "epoch": 0, "loss": 4.401309490203857, "loss_en": 4.402198314666748, "loss_zh": 4.400420665740967, "noise_std_eff": 0.0095, "elapsed_sec": 73.78100943565369}
20
+ {"step": 3400, "epoch": 0, "loss": 4.156264305114746, "loss_en": 4.116456508636475, "loss_zh": 4.196072101593018, "noise_std_eff": 0.01, "elapsed_sec": 76.71688103675842}
21
+ {"step": 3420, "epoch": 0, "loss": 3.369266986846924, "loss_en": 3.3708746433258057, "loss_zh": 3.367659568786621, "noise_std_eff": 0.0105, "elapsed_sec": 79.60879135131836}
22
+ {"step": 3440, "epoch": 0, "loss": 4.12277889251709, "loss_en": 4.07906436920166, "loss_zh": 4.166492938995361, "noise_std_eff": 0.011000000000000001, "elapsed_sec": 82.54479193687439}
23
+ {"step": 3460, "epoch": 0, "loss": 4.643231391906738, "loss_en": 4.596972465515137, "loss_zh": 4.68949031829834, "noise_std_eff": 0.0115, "elapsed_sec": 85.4506824016571}
24
+ {"step": 3480, "epoch": 0, "loss": 4.815507411956787, "loss_en": 4.778378486633301, "loss_zh": 4.852636337280273, "noise_std_eff": 0.012, "elapsed_sec": 88.40130996704102}
25
+ {"step": 3500, "epoch": 0, "loss": 4.43985652923584, "loss_en": 4.422976970672607, "loss_zh": 4.4567365646362305, "noise_std_eff": 0.0125, "elapsed_sec": 91.32398319244385}
26
+ {"step": 3520, "epoch": 0, "loss": 4.350632667541504, "loss_en": 4.332754611968994, "loss_zh": 4.368511199951172, "noise_std_eff": 0.013000000000000001, "elapsed_sec": 94.30030250549316}
27
+ {"step": 3540, "epoch": 0, "loss": 4.301290035247803, "loss_en": 4.286069869995117, "loss_zh": 4.316510200500488, "noise_std_eff": 0.013500000000000002, "elapsed_sec": 97.23365688323975}
28
+ {"step": 3560, "epoch": 0, "loss": 4.174412727355957, "loss_en": 4.2063164710998535, "loss_zh": 4.142509460449219, "noise_std_eff": 0.013999999999999999, "elapsed_sec": 100.23308205604553}
29
+ {"step": 3580, "epoch": 0, "loss": 4.282077312469482, "loss_en": 4.340432643890381, "loss_zh": 4.223721981048584, "noise_std_eff": 0.014499999999999999, "elapsed_sec": 103.18482875823975}
30
+ {"step": 3600, "epoch": 0, "loss": 4.189810752868652, "loss_en": 4.1842732429504395, "loss_zh": 4.195347785949707, "noise_std_eff": 0.015, "elapsed_sec": 106.19124221801758}
31
+ {"step": 3620, "epoch": 0, "loss": 4.78132963180542, "loss_en": 4.74851655960083, "loss_zh": 4.81414270401001, "noise_std_eff": 0.015500000000000002, "elapsed_sec": 133.30765342712402}
32
+ {"step": 3640, "epoch": 0, "loss": 4.401313304901123, "loss_en": 4.3971943855285645, "loss_zh": 4.405432224273682, "noise_std_eff": 0.016, "elapsed_sec": 136.19244074821472}
33
+ {"step": 3660, "epoch": 0, "loss": 3.957441568374634, "loss_en": 4.006048202514648, "loss_zh": 3.908834934234619, "noise_std_eff": 0.0165, "elapsed_sec": 139.0878632068634}
34
+ {"step": 3680, "epoch": 0, "loss": 3.96955943107605, "loss_en": 4.014604091644287, "loss_zh": 3.9245147705078125, "noise_std_eff": 0.017, "elapsed_sec": 142.00446963310242}
35
+ {"step": 3700, "epoch": 0, "loss": 4.57257604598999, "loss_en": 4.5815300941467285, "loss_zh": 4.563621997833252, "noise_std_eff": 0.0175, "elapsed_sec": 144.8811740875244}
36
+ {"step": 3720, "epoch": 0, "loss": 4.258642673492432, "loss_en": 4.241209983825684, "loss_zh": 4.27607536315918, "noise_std_eff": 0.018000000000000002, "elapsed_sec": 147.79147672653198}
37
+ {"step": 3740, "epoch": 0, "loss": 4.131010055541992, "loss_en": 4.127406597137451, "loss_zh": 4.134613513946533, "noise_std_eff": 0.018500000000000003, "elapsed_sec": 150.67635655403137}
38
+ {"step": 3760, "epoch": 0, "loss": 4.572167873382568, "loss_en": 4.576511383056641, "loss_zh": 4.567824363708496, "noise_std_eff": 0.019, "elapsed_sec": 153.60872435569763}
39
+ {"step": 3780, "epoch": 0, "loss": 4.194507598876953, "loss_en": 4.2233805656433105, "loss_zh": 4.1656341552734375, "noise_std_eff": 0.0195, "elapsed_sec": 156.51680994033813}
40
+ {"step": 3800, "epoch": 0, "loss": 4.101620197296143, "loss_en": 4.117923736572266, "loss_zh": 4.0853166580200195, "noise_std_eff": 0.02, "elapsed_sec": 159.46836161613464}
41
+ {"step": 3820, "epoch": 0, "loss": 5.150385856628418, "loss_en": 5.17932653427124, "loss_zh": 5.1214447021484375, "noise_std_eff": 0.02, "elapsed_sec": 162.3683590888977}
42
+ {"step": 3840, "epoch": 0, "loss": 4.561019420623779, "loss_en": 4.539376735687256, "loss_zh": 4.582662105560303, "noise_std_eff": 0.02, "elapsed_sec": 165.30115151405334}
43
+ {"step": 3860, "epoch": 0, "loss": 4.36818790435791, "loss_en": 4.345452785491943, "loss_zh": 4.390923023223877, "noise_std_eff": 0.02, "elapsed_sec": 168.193377494812}
44
+ {"step": 3880, "epoch": 0, "loss": 4.169336318969727, "loss_en": 4.158626556396484, "loss_zh": 4.180046558380127, "noise_std_eff": 0.02, "elapsed_sec": 171.13644361495972}
45
+ {"step": 3900, "epoch": 0, "loss": 3.977576732635498, "loss_en": 3.9763498306274414, "loss_zh": 3.9788033962249756, "noise_std_eff": 0.02, "elapsed_sec": 174.00696444511414}
46
+ {"step": 3920, "epoch": 0, "loss": 3.534553050994873, "loss_en": 3.5189168453216553, "loss_zh": 3.550189256668091, "noise_std_eff": 0.02, "elapsed_sec": 176.92445373535156}
47
+ {"step": 3940, "epoch": 0, "loss": 5.084809303283691, "loss_en": 5.057297706604004, "loss_zh": 5.112320423126221, "noise_std_eff": 0.02, "elapsed_sec": 179.84544944763184}
48
+ {"step": 3960, "epoch": 0, "loss": 3.9994382858276367, "loss_en": 4.004201412200928, "loss_zh": 3.9946751594543457, "noise_std_eff": 0.02, "elapsed_sec": 182.78539752960205}
49
+ {"step": 3980, "epoch": 0, "loss": 4.585289001464844, "loss_en": 4.587390899658203, "loss_zh": 4.583187103271484, "noise_std_eff": 0.02, "elapsed_sec": 185.74418258666992}
50
+ {"step": 4000, "epoch": 0, "loss": 4.038907051086426, "loss_en": 4.060107231140137, "loss_zh": 4.017706871032715, "noise_std_eff": 0.02, "elapsed_sec": 188.71499466896057}
51
+ {"step": 4020, "epoch": 0, "loss": 3.9919495582580566, "loss_en": 4.004234313964844, "loss_zh": 3.9796650409698486, "noise_std_eff": 0.02, "elapsed_sec": 215.74451065063477}
52
+ {"step": 4040, "epoch": 0, "loss": 3.6986446380615234, "loss_en": 3.698375940322876, "loss_zh": 3.69891357421875, "noise_std_eff": 0.02, "elapsed_sec": 218.70512008666992}
53
+ {"step": 4060, "epoch": 0, "loss": 4.538440704345703, "loss_en": 4.560594081878662, "loss_zh": 4.516287326812744, "noise_std_eff": 0.02, "elapsed_sec": 221.5925235748291}
54
+ {"step": 4080, "epoch": 0, "loss": 3.873448133468628, "loss_en": 3.8352856636047363, "loss_zh": 3.9116106033325195, "noise_std_eff": 0.02, "elapsed_sec": 224.52728652954102}
55
+ {"step": 4100, "epoch": 0, "loss": 4.096792697906494, "loss_en": 4.076803684234619, "loss_zh": 4.116781711578369, "noise_std_eff": 0.02, "elapsed_sec": 227.4037082195282}
56
+ {"step": 4120, "epoch": 0, "loss": 4.063114643096924, "loss_en": 4.064176082611084, "loss_zh": 4.062053203582764, "noise_std_eff": 0.02, "elapsed_sec": 230.33871245384216}
57
+ {"step": 4140, "epoch": 0, "loss": 4.317628860473633, "loss_en": 4.2960896492004395, "loss_zh": 4.339168548583984, "noise_std_eff": 0.02, "elapsed_sec": 233.21676802635193}
58
+ {"step": 4160, "epoch": 0, "loss": 3.904561996459961, "loss_en": 3.8752007484436035, "loss_zh": 3.9339234828948975, "noise_std_eff": 0.02, "elapsed_sec": 236.13298058509827}
59
+ {"step": 4180, "epoch": 0, "loss": 4.132402420043945, "loss_en": 4.146928787231445, "loss_zh": 4.117876052856445, "noise_std_eff": 0.02, "elapsed_sec": 239.019469499588}
60
+ {"step": 4200, "epoch": 0, "loss": 4.067115783691406, "loss_en": 4.070797920227051, "loss_zh": 4.063433647155762, "noise_std_eff": 0.02, "elapsed_sec": 241.96397805213928}
61
+ {"step": 4220, "epoch": 0, "loss": 3.8904075622558594, "loss_en": 3.8756463527679443, "loss_zh": 3.9051690101623535, "noise_std_eff": 0.02, "elapsed_sec": 244.86853432655334}
62
+ {"step": 4240, "epoch": 0, "loss": 4.179715633392334, "loss_en": 4.182239532470703, "loss_zh": 4.177191734313965, "noise_std_eff": 0.02, "elapsed_sec": 247.80825352668762}
63
+ {"step": 4260, "epoch": 0, "loss": 4.069424629211426, "loss_en": 4.088862419128418, "loss_zh": 4.049987316131592, "noise_std_eff": 0.02, "elapsed_sec": 250.6765661239624}
64
+ {"step": 4280, "epoch": 0, "loss": 3.998659610748291, "loss_en": 4.02363395690918, "loss_zh": 3.9736855030059814, "noise_std_eff": 0.02, "elapsed_sec": 253.6005504131317}
65
+ {"step": 4300, "epoch": 0, "loss": 4.0146589279174805, "loss_en": 3.994990587234497, "loss_zh": 4.034327507019043, "noise_std_eff": 0.02, "elapsed_sec": 256.1392283439636}
66
+ {"step": 4320, "epoch": 0, "loss": 4.066336631774902, "loss_en": 4.085277557373047, "loss_zh": 4.047395706176758, "noise_std_eff": 0.02, "elapsed_sec": 259.06590151786804}
67
+ {"step": 4340, "epoch": 0, "loss": 4.058927536010742, "loss_en": 4.052394866943359, "loss_zh": 4.065460205078125, "noise_std_eff": 0.02, "elapsed_sec": 261.94484281539917}
68
+ {"step": 4360, "epoch": 0, "loss": 3.7302846908569336, "loss_en": 3.768963098526001, "loss_zh": 3.6916065216064453, "noise_std_eff": 0.02, "elapsed_sec": 264.8867268562317}
69
+ {"step": 4380, "epoch": 0, "loss": 3.81063175201416, "loss_en": 3.817863941192627, "loss_zh": 3.8033995628356934, "noise_std_eff": 0.02, "elapsed_sec": 267.7608280181885}
70
+ {"step": 4400, "epoch": 0, "loss": 4.148674011230469, "loss_en": 4.148589611053467, "loss_zh": 4.1487579345703125, "noise_std_eff": 0.02, "elapsed_sec": 270.67489528656006}
71
+ {"step": 4420, "epoch": 0, "loss": 4.165624618530273, "loss_en": 4.165587425231934, "loss_zh": 4.165661811828613, "noise_std_eff": 0.02, "elapsed_sec": 297.7199823856354}
72
+ {"step": 4440, "epoch": 0, "loss": 4.240630149841309, "loss_en": 4.269030570983887, "loss_zh": 4.212230205535889, "noise_std_eff": 0.02, "elapsed_sec": 300.6559920310974}
73
+ {"step": 4460, "epoch": 0, "loss": 4.075514316558838, "loss_en": 4.035861015319824, "loss_zh": 4.115167617797852, "noise_std_eff": 0.02, "elapsed_sec": 303.5721607208252}
74
+ {"step": 4480, "epoch": 0, "loss": 3.700826406478882, "loss_en": 3.6593399047851562, "loss_zh": 3.7423129081726074, "noise_std_eff": 0.02, "elapsed_sec": 306.56488275527954}
75
+ {"step": 4500, "epoch": 0, "loss": 4.462973594665527, "loss_en": 4.437343597412109, "loss_zh": 4.4886040687561035, "noise_std_eff": 0.02, "elapsed_sec": 309.4721345901489}
76
+ {"step": 4520, "epoch": 0, "loss": 4.019312858581543, "loss_en": 4.028932094573975, "loss_zh": 4.0096940994262695, "noise_std_eff": 0.02, "elapsed_sec": 312.4160883426666}
77
+ {"step": 4540, "epoch": 0, "loss": 3.8015005588531494, "loss_en": 3.8127076625823975, "loss_zh": 3.7902934551239014, "noise_std_eff": 0.02, "elapsed_sec": 315.29841113090515}
78
+ {"step": 4560, "epoch": 0, "loss": 3.5797152519226074, "loss_en": 3.5848820209503174, "loss_zh": 3.5745484828948975, "noise_std_eff": 0.02, "elapsed_sec": 318.2553708553314}
79
+ {"step": 4580, "epoch": 0, "loss": 3.859844207763672, "loss_en": 3.8580880165100098, "loss_zh": 3.861600160598755, "noise_std_eff": 0.02, "elapsed_sec": 321.1505379676819}
80
+ {"step": 4600, "epoch": 0, "loss": 3.9847888946533203, "loss_en": 4.004213333129883, "loss_zh": 3.965364694595337, "noise_std_eff": 0.02, "elapsed_sec": 324.0881562232971}
81
+ {"step": 4620, "epoch": 0, "loss": 4.846346855163574, "loss_en": 4.900197505950928, "loss_zh": 4.7924957275390625, "noise_std_eff": 0.02, "elapsed_sec": 326.97873067855835}
82
+ {"step": 4640, "epoch": 0, "loss": 3.823059558868408, "loss_en": 3.7704246044158936, "loss_zh": 3.8756942749023438, "noise_std_eff": 0.02, "elapsed_sec": 329.944304227829}
83
+ {"step": 4660, "epoch": 0, "loss": 3.6971662044525146, "loss_en": 3.682452917098999, "loss_zh": 3.7118794918060303, "noise_std_eff": 0.02, "elapsed_sec": 332.82297682762146}
84
+ {"step": 4680, "epoch": 0, "loss": 3.5278072357177734, "loss_en": 3.608076333999634, "loss_zh": 3.447538137435913, "noise_std_eff": 0.02, "elapsed_sec": 335.75528931617737}
85
+ {"step": 4700, "epoch": 0, "loss": 3.9364376068115234, "loss_en": 3.9585213661193848, "loss_zh": 3.914353847503662, "noise_std_eff": 0.02, "elapsed_sec": 338.63195610046387}
86
+ {"step": 4720, "epoch": 0, "loss": 4.119994640350342, "loss_en": 4.157469749450684, "loss_zh": 4.08251953125, "noise_std_eff": 0.02, "elapsed_sec": 341.5594952106476}
87
+ {"step": 4740, "epoch": 0, "loss": 4.29338264465332, "loss_en": 4.288270950317383, "loss_zh": 4.298494338989258, "noise_std_eff": 0.02, "elapsed_sec": 344.4589328765869}
88
+ {"step": 4760, "epoch": 0, "loss": 3.9264659881591797, "loss_en": 3.9424924850463867, "loss_zh": 3.9104392528533936, "noise_std_eff": 0.02, "elapsed_sec": 347.3684720993042}
89
+ {"step": 4780, "epoch": 0, "loss": 4.0017805099487305, "loss_en": 4.001039505004883, "loss_zh": 4.00252103805542, "noise_std_eff": 0.02, "elapsed_sec": 350.23991680145264}
90
+ {"step": 4800, "epoch": 0, "loss": 3.74092960357666, "loss_en": 3.715859889984131, "loss_zh": 3.7659993171691895, "noise_std_eff": 0.02, "elapsed_sec": 353.19563364982605}
91
+ {"step": 4820, "epoch": 0, "loss": 3.626040458679199, "loss_en": 3.604022741317749, "loss_zh": 3.6480581760406494, "noise_std_eff": 0.02, "elapsed_sec": 380.35524225234985}
92
+ {"step": 4840, "epoch": 0, "loss": 3.870488166809082, "loss_en": 3.87906813621521, "loss_zh": 3.861908435821533, "noise_std_eff": 0.02, "elapsed_sec": 383.33828830718994}
93
+ {"step": 4860, "epoch": 0, "loss": 3.3858518600463867, "loss_en": 3.40728759765625, "loss_zh": 3.3644161224365234, "noise_std_eff": 0.02, "elapsed_sec": 386.20845890045166}
94
+ {"step": 4880, "epoch": 0, "loss": 3.5379257202148438, "loss_en": 3.627880811691284, "loss_zh": 3.4479708671569824, "noise_std_eff": 0.02, "elapsed_sec": 389.13145422935486}
95
+ {"step": 4900, "epoch": 0, "loss": 3.6775360107421875, "loss_en": 3.631058692932129, "loss_zh": 3.724013090133667, "noise_std_eff": 0.02, "elapsed_sec": 392.04403495788574}
96
+ {"step": 4920, "epoch": 0, "loss": 4.092677593231201, "loss_en": 4.0803303718566895, "loss_zh": 4.105024814605713, "noise_std_eff": 0.02, "elapsed_sec": 395.011239528656}
97
+ {"step": 4940, "epoch": 0, "loss": 4.292415142059326, "loss_en": 4.30841064453125, "loss_zh": 4.276419639587402, "noise_std_eff": 0.02, "elapsed_sec": 397.92793679237366}
98
+ {"step": 4960, "epoch": 0, "loss": 4.2843217849731445, "loss_en": 4.292796611785889, "loss_zh": 4.275846481323242, "noise_std_eff": 0.02, "elapsed_sec": 400.8466536998749}
99
+ {"step": 4980, "epoch": 0, "loss": 3.5295498371124268, "loss_en": 3.5543694496154785, "loss_zh": 3.504730224609375, "noise_std_eff": 0.02, "elapsed_sec": 403.7525591850281}
100
+ {"step": 5000, "epoch": 0, "loss": 3.6541073322296143, "loss_en": 3.6523799896240234, "loss_zh": 3.655834674835205, "noise_std_eff": 0.02, "elapsed_sec": 406.67465901374817}
101
+ {"step": 5020, "epoch": 0, "loss": 3.77231502532959, "loss_en": 3.768822431564331, "loss_zh": 3.7758073806762695, "noise_std_eff": 0.02, "elapsed_sec": 409.5549120903015}
102
+ {"step": 5040, "epoch": 0, "loss": 3.9452855587005615, "loss_en": 3.895145893096924, "loss_zh": 3.995425224304199, "noise_std_eff": 0.02, "elapsed_sec": 412.4966313838959}
103
+ {"step": 5060, "epoch": 0, "loss": 3.5932726860046387, "loss_en": 3.5898330211639404, "loss_zh": 3.596712112426758, "noise_std_eff": 0.02, "elapsed_sec": 415.3876316547394}
104
+ {"step": 5080, "epoch": 0, "loss": 3.416210651397705, "loss_en": 3.4127931594848633, "loss_zh": 3.419628143310547, "noise_std_eff": 0.02, "elapsed_sec": 418.0707845687866}
105
+ {"step": 5100, "epoch": 0, "loss": 3.2218947410583496, "loss_en": 3.2889206409454346, "loss_zh": 3.1548690795898438, "noise_std_eff": 0.02, "elapsed_sec": 420.19298696517944}
106
+ {"step": 5120, "epoch": 0, "loss": 3.8911614418029785, "loss_en": 3.8935701847076416, "loss_zh": 3.8887529373168945, "noise_std_eff": 0.02, "elapsed_sec": 422.36180090904236}
107
+ {"step": 5140, "epoch": 0, "loss": 5.0474653244018555, "loss_en": 4.928060531616211, "loss_zh": 5.166869640350342, "noise_std_eff": 0.02, "elapsed_sec": 424.83674788475037}
108
+ {"step": 5160, "epoch": 0, "loss": 3.67500376701355, "loss_en": 3.663123846054077, "loss_zh": 3.6868836879730225, "noise_std_eff": 0.02, "elapsed_sec": 427.7758939266205}
109
+ {"step": 5180, "epoch": 0, "loss": 3.316762924194336, "loss_en": 3.3236985206604004, "loss_zh": 3.3098273277282715, "noise_std_eff": 0.02, "elapsed_sec": 430.60152316093445}
110
+ {"step": 5200, "epoch": 0, "loss": 3.4854507446289062, "loss_en": 3.5013186931610107, "loss_zh": 3.4695825576782227, "noise_std_eff": 0.02, "elapsed_sec": 433.5656690597534}
111
+ {"step": 5220, "epoch": 0, "loss": 3.534630537033081, "loss_en": 3.5198800563812256, "loss_zh": 3.5493810176849365, "noise_std_eff": 0.02, "elapsed_sec": 460.8794219493866}
112
+ {"step": 5240, "epoch": 0, "loss": 4.166019439697266, "loss_en": 4.1882452964782715, "loss_zh": 4.143793106079102, "noise_std_eff": 0.02, "elapsed_sec": 463.77184677124023}
113
+ {"step": 5260, "epoch": 0, "loss": 3.6728596687316895, "loss_en": 3.7101011276245117, "loss_zh": 3.635618209838867, "noise_std_eff": 0.02, "elapsed_sec": 466.60845398902893}
114
+ {"step": 5280, "epoch": 0, "loss": 3.8518102169036865, "loss_en": 3.8757686614990234, "loss_zh": 3.8278517723083496, "noise_std_eff": 0.02, "elapsed_sec": 469.5081148147583}
115
+ {"step": 5300, "epoch": 0, "loss": 4.104333877563477, "loss_en": 4.1132588386535645, "loss_zh": 4.0954084396362305, "noise_std_eff": 0.02, "elapsed_sec": 472.3833374977112}
116
+ {"step": 5320, "epoch": 0, "loss": 3.914928674697876, "loss_en": 3.9231607913970947, "loss_zh": 3.9066965579986572, "noise_std_eff": 0.02, "elapsed_sec": 475.30500769615173}
117
+ {"step": 5340, "epoch": 0, "loss": 4.292413711547852, "loss_en": 4.303460121154785, "loss_zh": 4.281367301940918, "noise_std_eff": 0.02, "elapsed_sec": 478.1721365451813}
118
+ {"step": 5360, "epoch": 0, "loss": 3.673227548599243, "loss_en": 3.6785459518432617, "loss_zh": 3.6679091453552246, "noise_std_eff": 0.02, "elapsed_sec": 481.10505628585815}
119
+ {"step": 5380, "epoch": 0, "loss": 3.8096556663513184, "loss_en": 3.7794196605682373, "loss_zh": 3.8398914337158203, "noise_std_eff": 0.02, "elapsed_sec": 484.007794380188}
120
+ {"step": 5400, "epoch": 0, "loss": 4.038647651672363, "loss_en": 4.043335437774658, "loss_zh": 4.033960342407227, "noise_std_eff": 0.02, "elapsed_sec": 486.9206907749176}
121
+ {"step": 5420, "epoch": 0, "loss": 4.1362504959106445, "loss_en": 4.165543079376221, "loss_zh": 4.106958389282227, "noise_std_eff": 0.02, "elapsed_sec": 489.7676639556885}
122
+ {"step": 5440, "epoch": 0, "loss": 3.668729543685913, "loss_en": 3.7004542350769043, "loss_zh": 3.637004852294922, "noise_std_eff": 0.02, "elapsed_sec": 492.68808031082153}
123
+ {"step": 5460, "epoch": 0, "loss": 3.796924591064453, "loss_en": 3.797349214553833, "loss_zh": 3.796499729156494, "noise_std_eff": 0.02, "elapsed_sec": 495.5597004890442}
124
+ {"step": 5480, "epoch": 0, "loss": 3.564260959625244, "loss_en": 3.5694987773895264, "loss_zh": 3.559023141860962, "noise_std_eff": 0.02, "elapsed_sec": 498.4791646003723}
125
+ {"step": 5500, "epoch": 0, "loss": 3.8647680282592773, "loss_en": 3.870699644088745, "loss_zh": 3.8588366508483887, "noise_std_eff": 0.02, "elapsed_sec": 501.35238003730774}
126
+ {"step": 5520, "epoch": 0, "loss": 4.11713981628418, "loss_en": 4.147911548614502, "loss_zh": 4.086368083953857, "noise_std_eff": 0.02, "elapsed_sec": 504.2528917789459}
127
+ {"step": 5540, "epoch": 0, "loss": 3.9031643867492676, "loss_en": 3.8851206302642822, "loss_zh": 3.921208143234253, "noise_std_eff": 0.02, "elapsed_sec": 507.1017379760742}
128
+ {"step": 5560, "epoch": 0, "loss": 3.69441556930542, "loss_en": 3.7000644207000732, "loss_zh": 3.6887669563293457, "noise_std_eff": 0.02, "elapsed_sec": 509.9932997226715}
129
+ {"step": 5580, "epoch": 0, "loss": 3.5463638305664062, "loss_en": 3.681211471557617, "loss_zh": 3.4115161895751953, "noise_std_eff": 0.02, "elapsed_sec": 512.8551452159882}
130
+ {"step": 5600, "epoch": 0, "loss": 3.9938302040100098, "loss_en": 4.022293567657471, "loss_zh": 3.9653666019439697, "noise_std_eff": 0.02, "elapsed_sec": 515.7723672389984}
131
+ {"step": 5620, "epoch": 0, "loss": 4.128993511199951, "loss_en": 4.0896172523498535, "loss_zh": 4.168369770050049, "noise_std_eff": 0.02, "elapsed_sec": 542.750715970993}
132
+ {"step": 5640, "epoch": 0, "loss": 4.382177352905273, "loss_en": 4.38604211807251, "loss_zh": 4.378313064575195, "noise_std_eff": 0.02, "elapsed_sec": 545.671037197113}
133
+ {"step": 5660, "epoch": 0, "loss": 3.8740344047546387, "loss_en": 3.86179518699646, "loss_zh": 3.8862738609313965, "noise_std_eff": 0.02, "elapsed_sec": 548.5469696521759}
134
+ {"step": 5680, "epoch": 0, "loss": 3.8769025802612305, "loss_en": 3.8119659423828125, "loss_zh": 3.9418392181396484, "noise_std_eff": 0.02, "elapsed_sec": 551.4476516246796}
135
+ {"step": 5700, "epoch": 0, "loss": 3.414339542388916, "loss_en": 3.3810808658599854, "loss_zh": 3.4475979804992676, "noise_std_eff": 0.02, "elapsed_sec": 554.2841639518738}
136
+ {"step": 5720, "epoch": 0, "loss": 3.8795158863067627, "loss_en": 3.850627899169922, "loss_zh": 3.9084038734436035, "noise_std_eff": 0.02, "elapsed_sec": 557.1875705718994}
137
+ {"step": 5740, "epoch": 0, "loss": 3.769127130508423, "loss_en": 3.741734743118286, "loss_zh": 3.7965195178985596, "noise_std_eff": 0.02, "elapsed_sec": 560.0551371574402}
138
+ {"step": 5760, "epoch": 0, "loss": 3.9478745460510254, "loss_en": 3.9505677223205566, "loss_zh": 3.9451816082000732, "noise_std_eff": 0.02, "elapsed_sec": 562.9363191127777}
139
+ {"step": 5780, "epoch": 0, "loss": 3.9216928482055664, "loss_en": 3.915788412094116, "loss_zh": 3.9275972843170166, "noise_std_eff": 0.02, "elapsed_sec": 565.8031296730042}
140
+ {"step": 5800, "epoch": 0, "loss": 4.127882957458496, "loss_en": 4.085864067077637, "loss_zh": 4.169902324676514, "noise_std_eff": 0.02, "elapsed_sec": 568.6844398975372}
141
+ {"step": 5820, "epoch": 0, "loss": 3.923311233520508, "loss_en": 3.9869370460510254, "loss_zh": 3.8596856594085693, "noise_std_eff": 0.02, "elapsed_sec": 571.5160772800446}
142
+ {"step": 5840, "epoch": 0, "loss": 3.2523560523986816, "loss_en": 3.289869785308838, "loss_zh": 3.2148420810699463, "noise_std_eff": 0.02, "elapsed_sec": 574.4198310375214}
143
+ {"step": 5860, "epoch": 0, "loss": 3.8681704998016357, "loss_en": 3.817075252532959, "loss_zh": 3.9192657470703125, "noise_std_eff": 0.02, "elapsed_sec": 577.2759959697723}
144
+ {"step": 5880, "epoch": 0, "loss": 3.702848196029663, "loss_en": 3.720301389694214, "loss_zh": 3.6853950023651123, "noise_std_eff": 0.02, "elapsed_sec": 580.1563704013824}
145
+ {"step": 5900, "epoch": 0, "loss": 4.038462162017822, "loss_en": 3.9866321086883545, "loss_zh": 4.090291976928711, "noise_std_eff": 0.02, "elapsed_sec": 582.9797692298889}
146
+ {"step": 5920, "epoch": 0, "loss": 4.02570915222168, "loss_en": 4.0268874168396, "loss_zh": 4.02453088760376, "noise_std_eff": 0.02, "elapsed_sec": 585.8754191398621}
147
+ {"step": 5940, "epoch": 0, "loss": 3.5127899646759033, "loss_en": 3.5142579078674316, "loss_zh": 3.511322021484375, "noise_std_eff": 0.02, "elapsed_sec": 588.7400302886963}
148
+ {"step": 5960, "epoch": 0, "loss": 4.157806396484375, "loss_en": 4.20422887802124, "loss_zh": 4.111384391784668, "noise_std_eff": 0.02, "elapsed_sec": 591.6329216957092}
149
+ {"step": 5980, "epoch": 0, "loss": 3.57485294342041, "loss_en": 3.5185163021087646, "loss_zh": 3.6311893463134766, "noise_std_eff": 0.02, "elapsed_sec": 594.4889953136444}
150
+ {"step": 6000, "epoch": 0, "loss": 3.6896231174468994, "loss_en": 3.699902296066284, "loss_zh": 3.6793439388275146, "noise_std_eff": 0.02, "elapsed_sec": 597.3971717357635}
151
+ {"step": 6020, "epoch": 0, "loss": 3.608567953109741, "loss_en": 3.613818883895874, "loss_zh": 3.6033170223236084, "noise_std_eff": 0.02, "elapsed_sec": 624.5930035114288}
152
+ {"step": 6040, "epoch": 0, "loss": 2.9154112339019775, "loss_en": 2.953857898712158, "loss_zh": 2.876964569091797, "noise_std_eff": 0.02, "elapsed_sec": 627.5411050319672}
153
+ {"step": 6060, "epoch": 0, "loss": 3.7785398960113525, "loss_en": 3.7738468647003174, "loss_zh": 3.7832329273223877, "noise_std_eff": 0.02, "elapsed_sec": 630.4732174873352}
154
+ {"step": 6080, "epoch": 0, "loss": 3.7332491874694824, "loss_en": 3.7691848278045654, "loss_zh": 3.6973133087158203, "noise_std_eff": 0.02, "elapsed_sec": 633.3884131908417}
155
+ {"step": 6100, "epoch": 0, "loss": 3.779677152633667, "loss_en": 3.833911180496216, "loss_zh": 3.725443124771118, "noise_std_eff": 0.02, "elapsed_sec": 636.2804703712463}
156
+ {"step": 6120, "epoch": 0, "loss": 4.098194122314453, "loss_en": 4.0941572189331055, "loss_zh": 4.102231502532959, "noise_std_eff": 0.02, "elapsed_sec": 639.1802923679352}
157
+ {"step": 6140, "epoch": 0, "loss": 3.6344199180603027, "loss_en": 3.6321449279785156, "loss_zh": 3.63669490814209, "noise_std_eff": 0.02, "elapsed_sec": 642.0558433532715}
158
+ {"step": 6160, "epoch": 0, "loss": 3.7824628353118896, "loss_en": 3.7820396423339844, "loss_zh": 3.782886028289795, "noise_std_eff": 0.02, "elapsed_sec": 644.9561243057251}
159
+ {"step": 6180, "epoch": 0, "loss": 3.8292531967163086, "loss_en": 3.8112633228302, "loss_zh": 3.847243309020996, "noise_std_eff": 0.02, "elapsed_sec": 647.8356459140778}
160
+ {"step": 6200, "epoch": 0, "loss": 3.3210864067077637, "loss_en": 3.2922909259796143, "loss_zh": 3.349881887435913, "noise_std_eff": 0.02, "elapsed_sec": 650.7520558834076}
161
+ {"step": 6220, "epoch": 0, "loss": 3.438384532928467, "loss_en": 3.4046268463134766, "loss_zh": 3.472141981124878, "noise_std_eff": 0.02, "elapsed_sec": 653.6228361129761}
162
+ {"step": 6240, "epoch": 0, "loss": 3.541086196899414, "loss_en": 3.5472264289855957, "loss_zh": 3.5349457263946533, "noise_std_eff": 0.02, "elapsed_sec": 656.5755879878998}
163
+ {"step": 6260, "epoch": 0, "loss": 3.740877389907837, "loss_en": 3.7295799255371094, "loss_zh": 3.7521748542785645, "noise_std_eff": 0.02, "elapsed_sec": 659.4709498882294}
164
+ {"step": 6280, "epoch": 0, "loss": 3.9978256225585938, "loss_en": 3.995620012283325, "loss_zh": 4.000031471252441, "noise_std_eff": 0.02, "elapsed_sec": 662.4147791862488}
165
+ {"step": 6300, "epoch": 0, "loss": 3.5377326011657715, "loss_en": 3.4919843673706055, "loss_zh": 3.5834805965423584, "noise_std_eff": 0.02, "elapsed_sec": 665.3113799095154}
166
+ {"step": 6320, "epoch": 0, "loss": 3.413328170776367, "loss_en": 3.3853251934051514, "loss_zh": 3.441331148147583, "noise_std_eff": 0.02, "elapsed_sec": 668.2599773406982}
167
+ {"step": 6340, "epoch": 0, "loss": 3.0589609146118164, "loss_en": 3.032975912094116, "loss_zh": 3.0849461555480957, "noise_std_eff": 0.02, "elapsed_sec": 671.1164364814758}
168
+ {"step": 6360, "epoch": 0, "loss": 3.766186237335205, "loss_en": 3.780578136444092, "loss_zh": 3.7517940998077393, "noise_std_eff": 0.02, "elapsed_sec": 674.0373666286469}
169
+ {"step": 6380, "epoch": 0, "loss": 3.416584014892578, "loss_en": 3.409764051437378, "loss_zh": 3.4234042167663574, "noise_std_eff": 0.02, "elapsed_sec": 676.9193823337555}
170
+ {"step": 6400, "epoch": 0, "loss": 3.7960715293884277, "loss_en": 3.7688820362091064, "loss_zh": 3.823261022567749, "noise_std_eff": 0.02, "elapsed_sec": 679.8359611034393}
171
+ {"step": 6420, "epoch": 0, "loss": 3.8930869102478027, "loss_en": 3.889995813369751, "loss_zh": 3.8961782455444336, "noise_std_eff": 0.02, "elapsed_sec": 706.829293012619}
172
+ {"step": 6440, "epoch": 0, "loss": 3.76351261138916, "loss_en": 3.7785394191741943, "loss_zh": 3.748485565185547, "noise_std_eff": 0.02, "elapsed_sec": 709.7691946029663}
173
+ {"step": 6460, "epoch": 0, "loss": 3.875340223312378, "loss_en": 3.856825351715088, "loss_zh": 3.893855094909668, "noise_std_eff": 0.02, "elapsed_sec": 712.652220249176}
174
+ {"step": 6480, "epoch": 0, "loss": 3.745967388153076, "loss_en": 3.7595996856689453, "loss_zh": 3.732334852218628, "noise_std_eff": 0.02, "elapsed_sec": 715.5771191120148}
175
+ {"step": 6500, "epoch": 0, "loss": 3.3174383640289307, "loss_en": 3.314971446990967, "loss_zh": 3.3199052810668945, "noise_std_eff": 0.02, "elapsed_sec": 718.4527070522308}
176
+ {"step": 6520, "epoch": 0, "loss": 3.492746353149414, "loss_en": 3.5076584815979004, "loss_zh": 3.4778339862823486, "noise_std_eff": 0.02, "elapsed_sec": 721.3689742088318}
177
+ {"step": 6540, "epoch": 0, "loss": 3.3924813270568848, "loss_en": 3.4048337936401367, "loss_zh": 3.380129098892212, "noise_std_eff": 0.02, "elapsed_sec": 724.2609286308289}
178
+ {"step": 6560, "epoch": 0, "loss": 4.219328880310059, "loss_en": 4.214672088623047, "loss_zh": 4.22398567199707, "noise_std_eff": 0.02, "elapsed_sec": 727.2237906455994}
179
+ {"step": 6580, "epoch": 0, "loss": 3.5383241176605225, "loss_en": 3.522059917449951, "loss_zh": 3.5545883178710938, "noise_std_eff": 0.02, "elapsed_sec": 730.109160900116}
180
+ {"step": 6600, "epoch": 0, "loss": 3.8895368576049805, "loss_en": 3.8672754764556885, "loss_zh": 3.9117984771728516, "noise_std_eff": 0.02, "elapsed_sec": 733.0486204624176}
181
+ {"step": 6620, "epoch": 0, "loss": 3.5488204956054688, "loss_en": 3.528615951538086, "loss_zh": 3.5690252780914307, "noise_std_eff": 0.02, "elapsed_sec": 735.9567012786865}
182
+ {"step": 6640, "epoch": 0, "loss": 3.6277666091918945, "loss_en": 3.697904109954834, "loss_zh": 3.557629108428955, "noise_std_eff": 0.02, "elapsed_sec": 738.8530476093292}
183
+ {"step": 6660, "epoch": 0, "loss": 3.8970727920532227, "loss_en": 3.867506265640259, "loss_zh": 3.9266393184661865, "noise_std_eff": 0.02, "elapsed_sec": 741.7475941181183}
184
+ {"step": 6680, "epoch": 0, "loss": 3.133471727371216, "loss_en": 3.0945889949798584, "loss_zh": 3.1723544597625732, "noise_std_eff": 0.02, "elapsed_sec": 744.6440396308899}
185
+ {"step": 6700, "epoch": 0, "loss": 3.137392520904541, "loss_en": 3.1175990104675293, "loss_zh": 3.157186269760132, "noise_std_eff": 0.02, "elapsed_sec": 747.5240478515625}
186
+ {"step": 6720, "epoch": 0, "loss": 3.587867259979248, "loss_en": 3.5770068168640137, "loss_zh": 3.5987277030944824, "noise_std_eff": 0.02, "elapsed_sec": 750.4285686016083}
187
+ {"step": 6740, "epoch": 0, "loss": 3.7425918579101562, "loss_en": 3.742727756500244, "loss_zh": 3.7424561977386475, "noise_std_eff": 0.02, "elapsed_sec": 753.3193919658661}
188
+ {"step": 6760, "epoch": 0, "loss": 3.865396022796631, "loss_en": 3.776120901107788, "loss_zh": 3.9546709060668945, "noise_std_eff": 0.02, "elapsed_sec": 756.2202024459839}
189
+ {"step": 6780, "epoch": 0, "loss": 3.5667991638183594, "loss_en": 3.5806782245635986, "loss_zh": 3.552920341491699, "noise_std_eff": 0.02, "elapsed_sec": 759.1078772544861}
190
+ {"step": 6800, "epoch": 0, "loss": 3.5284337997436523, "loss_en": 3.528395891189575, "loss_zh": 3.5284714698791504, "noise_std_eff": 0.02, "elapsed_sec": 762.0355463027954}
191
+ {"step": 6820, "epoch": 0, "loss": 4.060451507568359, "loss_en": 4.053103923797607, "loss_zh": 4.067799091339111, "noise_std_eff": 0.02, "elapsed_sec": 788.9690334796906}
192
+ {"step": 6840, "epoch": 0, "loss": 4.035606384277344, "loss_en": 4.1642985343933105, "loss_zh": 3.906914234161377, "noise_std_eff": 0.02, "elapsed_sec": 791.8683431148529}
193
+ {"step": 6860, "epoch": 0, "loss": 3.4901304244995117, "loss_en": 3.5065948963165283, "loss_zh": 3.473665714263916, "noise_std_eff": 0.02, "elapsed_sec": 794.8134546279907}
194
+ {"step": 6880, "epoch": 0, "loss": 3.6528663635253906, "loss_en": 3.611201286315918, "loss_zh": 3.6945316791534424, "noise_std_eff": 0.02, "elapsed_sec": 797.772842168808}
195
+ {"step": 6900, "epoch": 0, "loss": 3.677323579788208, "loss_en": 3.7068119049072266, "loss_zh": 3.6478352546691895, "noise_std_eff": 0.02, "elapsed_sec": 800.669793844223}
196
+ {"step": 6920, "epoch": 0, "loss": 3.251107692718506, "loss_en": 3.21789288520813, "loss_zh": 3.284322738647461, "noise_std_eff": 0.02, "elapsed_sec": 803.6216349601746}
197
+ {"step": 6940, "epoch": 0, "loss": 3.47029447555542, "loss_en": 3.4781980514526367, "loss_zh": 3.462390661239624, "noise_std_eff": 0.02, "elapsed_sec": 806.5213634967804}
198
+ {"step": 6960, "epoch": 0, "loss": 3.6703343391418457, "loss_en": 3.676535129547119, "loss_zh": 3.6641335487365723, "noise_std_eff": 0.02, "elapsed_sec": 809.4367823600769}
199
+ {"step": 6980, "epoch": 0, "loss": 3.325054883956909, "loss_en": 3.352646827697754, "loss_zh": 3.2974629402160645, "noise_std_eff": 0.02, "elapsed_sec": 812.3288514614105}
200
+ {"step": 7000, "epoch": 0, "loss": 3.681408405303955, "loss_en": 3.6620640754699707, "loss_zh": 3.7007527351379395, "noise_std_eff": 0.02, "elapsed_sec": 815.2137904167175}
201
+ {"step": 7020, "epoch": 0, "loss": 3.7730603218078613, "loss_en": 3.7513785362243652, "loss_zh": 3.7947423458099365, "noise_std_eff": 0.02, "elapsed_sec": 818.0813884735107}
202
+ {"step": 7040, "epoch": 0, "loss": 3.913329839706421, "loss_en": 3.900984048843384, "loss_zh": 3.925675630569458, "noise_std_eff": 0.02, "elapsed_sec": 821.0125999450684}
203
+ {"step": 7060, "epoch": 0, "loss": 3.6699204444885254, "loss_en": 3.6837756633758545, "loss_zh": 3.656064987182617, "noise_std_eff": 0.02, "elapsed_sec": 823.8608524799347}
204
+ {"step": 7080, "epoch": 0, "loss": 4.0173845291137695, "loss_en": 3.9802956581115723, "loss_zh": 4.054473876953125, "noise_std_eff": 0.02, "elapsed_sec": 826.7640540599823}
205
+ {"step": 7100, "epoch": 0, "loss": 3.742976188659668, "loss_en": 3.74698543548584, "loss_zh": 3.738966703414917, "noise_std_eff": 0.02, "elapsed_sec": 829.6279067993164}
206
+ {"step": 7120, "epoch": 0, "loss": 3.7594637870788574, "loss_en": 3.7588582038879395, "loss_zh": 3.7600691318511963, "noise_std_eff": 0.02, "elapsed_sec": 832.50923371315}
207
+ {"step": 7140, "epoch": 0, "loss": 4.221591472625732, "loss_en": 4.176703929901123, "loss_zh": 4.266479015350342, "noise_std_eff": 0.02, "elapsed_sec": 835.370099067688}
208
+ {"step": 7160, "epoch": 0, "loss": 3.7905540466308594, "loss_en": 3.8026297092437744, "loss_zh": 3.7784781455993652, "noise_std_eff": 0.02, "elapsed_sec": 838.2529766559601}
209
+ {"step": 7180, "epoch": 0, "loss": 3.2527008056640625, "loss_en": 3.2615461349487305, "loss_zh": 3.2438557147979736, "noise_std_eff": 0.02, "elapsed_sec": 841.0976459980011}
210
+ {"step": 7200, "epoch": 0, "loss": 3.426815986633301, "loss_en": 3.445966958999634, "loss_zh": 3.4076647758483887, "noise_std_eff": 0.02, "elapsed_sec": 843.9530072212219}
211
+ {"step": 7220, "epoch": 0, "loss": 3.736701726913452, "loss_en": 3.73248291015625, "loss_zh": 3.7409205436706543, "noise_std_eff": 0.02, "elapsed_sec": 870.9916756153107}
212
+ {"step": 7240, "epoch": 0, "loss": 3.545321464538574, "loss_en": 3.5367321968078613, "loss_zh": 3.553910970687866, "noise_std_eff": 0.02, "elapsed_sec": 873.9002864360809}
213
+ {"step": 7260, "epoch": 0, "loss": 3.3181979656219482, "loss_en": 3.3293330669403076, "loss_zh": 3.307062864303589, "noise_std_eff": 0.02, "elapsed_sec": 876.743528842926}
214
+ {"step": 7280, "epoch": 0, "loss": 3.2949955463409424, "loss_en": 3.282806634902954, "loss_zh": 3.3071844577789307, "noise_std_eff": 0.02, "elapsed_sec": 879.6165194511414}
215
+ {"step": 7300, "epoch": 0, "loss": 3.252169609069824, "loss_en": 3.243985652923584, "loss_zh": 3.2603533267974854, "noise_std_eff": 0.02, "elapsed_sec": 882.4567179679871}
216
+ {"step": 7320, "epoch": 0, "loss": 3.4599108695983887, "loss_en": 3.4481372833251953, "loss_zh": 3.471684217453003, "noise_std_eff": 0.02, "elapsed_sec": 885.3479962348938}
217
+ {"step": 7340, "epoch": 0, "loss": 4.276022434234619, "loss_en": 4.222749710083008, "loss_zh": 4.3292951583862305, "noise_std_eff": 0.02, "elapsed_sec": 888.2270729541779}
218
+ {"step": 7360, "epoch": 0, "loss": 3.243515968322754, "loss_en": 3.230480670928955, "loss_zh": 3.256551504135132, "noise_std_eff": 0.02, "elapsed_sec": 891.1314601898193}
219
+ {"step": 7380, "epoch": 0, "loss": 3.4895682334899902, "loss_en": 3.474442958831787, "loss_zh": 3.5046937465667725, "noise_std_eff": 0.02, "elapsed_sec": 894.0243000984192}
220
+ {"step": 7400, "epoch": 0, "loss": 3.2328858375549316, "loss_en": 3.241330146789551, "loss_zh": 3.2244412899017334, "noise_std_eff": 0.02, "elapsed_sec": 896.9075148105621}
221
+ {"step": 7420, "epoch": 0, "loss": 3.721323251724243, "loss_en": 3.6910977363586426, "loss_zh": 3.7515487670898438, "noise_std_eff": 0.02, "elapsed_sec": 899.7794966697693}
222
+ {"step": 7440, "epoch": 0, "loss": 3.4444024562835693, "loss_en": 3.4363200664520264, "loss_zh": 3.4524848461151123, "noise_std_eff": 0.02, "elapsed_sec": 902.6881232261658}
223
+ {"step": 7460, "epoch": 0, "loss": 3.374791383743286, "loss_en": 3.3803656101226807, "loss_zh": 3.3692171573638916, "noise_std_eff": 0.02, "elapsed_sec": 905.5444192886353}
224
+ {"step": 7480, "epoch": 0, "loss": 3.919816493988037, "loss_en": 3.927016496658325, "loss_zh": 3.912616729736328, "noise_std_eff": 0.02, "elapsed_sec": 908.4162163734436}
225
+ {"step": 7500, "epoch": 1, "loss": 5.215428352355957, "loss_en": 5.19863224029541, "loss_zh": 5.232223987579346, "noise_std_eff": 0.02, "elapsed_sec": 911.3702261447906}
226
+ {"step": 7520, "epoch": 1, "loss": 4.764598846435547, "loss_en": 4.651387691497803, "loss_zh": 4.877810478210449, "noise_std_eff": 0.02, "elapsed_sec": 914.3704018592834}
227
+ {"step": 7540, "epoch": 1, "loss": 4.540335655212402, "loss_en": 4.532250881195068, "loss_zh": 4.548420429229736, "noise_std_eff": 0.02, "elapsed_sec": 917.367819070816}
228
+ {"step": 7560, "epoch": 1, "loss": 4.468413829803467, "loss_en": 4.428526401519775, "loss_zh": 4.508301258087158, "noise_std_eff": 0.02, "elapsed_sec": 920.3782587051392}
229
+ {"step": 7580, "epoch": 1, "loss": 4.811252593994141, "loss_en": 4.828670501708984, "loss_zh": 4.793834686279297, "noise_std_eff": 0.02, "elapsed_sec": 923.3601007461548}
230
+ {"step": 7600, "epoch": 1, "loss": 5.262836933135986, "loss_en": 5.260496139526367, "loss_zh": 5.2651777267456055, "noise_std_eff": 0.02, "elapsed_sec": 926.3617615699768}
231
+ {"step": 7620, "epoch": 1, "loss": 3.850417137145996, "loss_en": 3.854146718978882, "loss_zh": 3.8466877937316895, "noise_std_eff": 0.02, "elapsed_sec": 953.2877442836761}
232
+ {"step": 7640, "epoch": 1, "loss": 3.4732179641723633, "loss_en": 3.465517044067383, "loss_zh": 3.4809188842773438, "noise_std_eff": 0.02, "elapsed_sec": 956.2043190002441}
233
+ {"step": 7660, "epoch": 1, "loss": 3.8484997749328613, "loss_en": 3.847137212753296, "loss_zh": 3.8498623371124268, "noise_std_eff": 0.02, "elapsed_sec": 959.0436363220215}
234
+ {"step": 7680, "epoch": 1, "loss": 3.2980403900146484, "loss_en": 3.3398098945617676, "loss_zh": 3.25627064704895, "noise_std_eff": 0.02, "elapsed_sec": 961.9480049610138}
235
+ {"step": 7700, "epoch": 1, "loss": 2.9972312450408936, "loss_en": 3.0035831928253174, "loss_zh": 2.9908792972564697, "noise_std_eff": 0.02, "elapsed_sec": 964.772173166275}
236
+ {"step": 7720, "epoch": 1, "loss": 3.4250502586364746, "loss_en": 3.421103000640869, "loss_zh": 3.42899751663208, "noise_std_eff": 0.02, "elapsed_sec": 967.6836354732513}
237
+ {"step": 7740, "epoch": 1, "loss": 3.7050094604492188, "loss_en": 3.683140277862549, "loss_zh": 3.7268786430358887, "noise_std_eff": 0.02, "elapsed_sec": 970.5436758995056}
238
+ {"step": 7760, "epoch": 1, "loss": 3.5109944343566895, "loss_en": 3.5002341270446777, "loss_zh": 3.5217549800872803, "noise_std_eff": 0.02, "elapsed_sec": 973.4433629512787}
239
+ {"step": 7780, "epoch": 1, "loss": 3.506405830383301, "loss_en": 3.5124223232269287, "loss_zh": 3.500389337539673, "noise_std_eff": 0.02, "elapsed_sec": 976.2698609828949}
240
+ {"step": 7800, "epoch": 1, "loss": 4.306281089782715, "loss_en": 4.328876495361328, "loss_zh": 4.283685207366943, "noise_std_eff": 0.02, "elapsed_sec": 979.1642422676086}
241
+ {"step": 7820, "epoch": 1, "loss": 3.164196491241455, "loss_en": 3.1818766593933105, "loss_zh": 3.1465163230895996, "noise_std_eff": 0.02, "elapsed_sec": 982.0309069156647}
242
+ {"step": 7840, "epoch": 1, "loss": 3.812518358230591, "loss_en": 3.8279669284820557, "loss_zh": 3.797069787979126, "noise_std_eff": 0.02, "elapsed_sec": 984.9353976249695}
243
+ {"step": 7860, "epoch": 1, "loss": 3.288508892059326, "loss_en": 3.3202669620513916, "loss_zh": 3.2567508220672607, "noise_std_eff": 0.02, "elapsed_sec": 987.791187286377}
244
+ {"step": 7880, "epoch": 1, "loss": 3.652590751647949, "loss_en": 3.684175968170166, "loss_zh": 3.6210057735443115, "noise_std_eff": 0.02, "elapsed_sec": 990.6641449928284}
245
+ {"step": 7900, "epoch": 1, "loss": 3.5625052452087402, "loss_en": 3.531550168991089, "loss_zh": 3.5934600830078125, "noise_std_eff": 0.02, "elapsed_sec": 993.4877374172211}
246
+ {"step": 7920, "epoch": 1, "loss": 3.444242000579834, "loss_en": 3.4790608882904053, "loss_zh": 3.409423351287842, "noise_std_eff": 0.02, "elapsed_sec": 996.3875455856323}
247
+ {"step": 7940, "epoch": 1, "loss": 3.4204506874084473, "loss_en": 3.4261863231658936, "loss_zh": 3.414714813232422, "noise_std_eff": 0.02, "elapsed_sec": 999.2710621356964}
248
+ {"step": 7960, "epoch": 1, "loss": 3.689709186553955, "loss_en": 3.6906657218933105, "loss_zh": 3.6887526512145996, "noise_std_eff": 0.02, "elapsed_sec": 1002.2394046783447}
249
+ {"step": 7980, "epoch": 1, "loss": 3.496372699737549, "loss_en": 3.4862139225006104, "loss_zh": 3.506531238555908, "noise_std_eff": 0.02, "elapsed_sec": 1005.1121127605438}
250
+ {"step": 8000, "epoch": 1, "loss": 3.485344886779785, "loss_en": 3.4906973838806152, "loss_zh": 3.479992389678955, "noise_std_eff": 0.02, "elapsed_sec": 1008.0443832874298}
251
+ {"step": 8020, "epoch": 1, "loss": 3.200690269470215, "loss_en": 3.2239179611206055, "loss_zh": 3.177462577819824, "noise_std_eff": 0.02, "elapsed_sec": 1035.1196022033691}
252
+ {"step": 8040, "epoch": 1, "loss": 3.272982120513916, "loss_en": 3.2622265815734863, "loss_zh": 3.2837376594543457, "noise_std_eff": 0.02, "elapsed_sec": 1038.0894968509674}
253
+ {"step": 8060, "epoch": 1, "loss": 4.086793899536133, "loss_en": 4.091405391693115, "loss_zh": 4.082181930541992, "noise_std_eff": 0.02, "elapsed_sec": 1040.9604659080505}
254
+ {"step": 8080, "epoch": 1, "loss": 3.7384650707244873, "loss_en": 3.7190463542938232, "loss_zh": 3.7578837871551514, "noise_std_eff": 0.02, "elapsed_sec": 1043.9126796722412}
255
+ {"step": 8100, "epoch": 1, "loss": 3.4217188358306885, "loss_en": 3.4506642818450928, "loss_zh": 3.392773389816284, "noise_std_eff": 0.02, "elapsed_sec": 1046.7924072742462}
256
+ {"step": 8120, "epoch": 1, "loss": 3.5707311630249023, "loss_en": 3.585923671722412, "loss_zh": 3.5555384159088135, "noise_std_eff": 0.02, "elapsed_sec": 1049.7361137866974}
257
+ {"step": 8140, "epoch": 1, "loss": 3.352121114730835, "loss_en": 3.326279640197754, "loss_zh": 3.377962589263916, "noise_std_eff": 0.02, "elapsed_sec": 1052.5837264060974}
258
+ {"step": 8160, "epoch": 1, "loss": 3.2608909606933594, "loss_en": 3.2818074226379395, "loss_zh": 3.2399747371673584, "noise_std_eff": 0.02, "elapsed_sec": 1055.4920551776886}
259
+ {"step": 8180, "epoch": 1, "loss": 3.5951290130615234, "loss_en": 3.545450448989868, "loss_zh": 3.644807815551758, "noise_std_eff": 0.02, "elapsed_sec": 1058.387770652771}
260
+ {"step": 8200, "epoch": 1, "loss": 3.743579387664795, "loss_en": 3.7299463748931885, "loss_zh": 3.7572121620178223, "noise_std_eff": 0.02, "elapsed_sec": 1061.3232250213623}
261
+ {"step": 8220, "epoch": 1, "loss": 3.3866443634033203, "loss_en": 3.3735830783843994, "loss_zh": 3.399705410003662, "noise_std_eff": 0.02, "elapsed_sec": 1064.214375257492}
262
+ {"step": 8240, "epoch": 1, "loss": 3.3458545207977295, "loss_en": 3.431715726852417, "loss_zh": 3.259993314743042, "noise_std_eff": 0.02, "elapsed_sec": 1067.1481394767761}
263
+ {"step": 8260, "epoch": 1, "loss": 3.454857349395752, "loss_en": 3.4530868530273438, "loss_zh": 3.4566280841827393, "noise_std_eff": 0.02, "elapsed_sec": 1070.0756442546844}
264
+ {"step": 8280, "epoch": 1, "loss": 3.019078493118286, "loss_en": 2.9960851669311523, "loss_zh": 3.04207181930542, "noise_std_eff": 0.02, "elapsed_sec": 1073.007611989975}
265
+ {"step": 8300, "epoch": 1, "loss": 3.73828125, "loss_en": 3.7485272884368896, "loss_zh": 3.7280352115631104, "noise_std_eff": 0.02, "elapsed_sec": 1075.894719839096}
266
+ {"step": 8320, "epoch": 1, "loss": 3.352726936340332, "loss_en": 3.3575615882873535, "loss_zh": 3.3478925228118896, "noise_std_eff": 0.02, "elapsed_sec": 1078.8901267051697}
267
+ {"step": 8340, "epoch": 1, "loss": 3.4358201026916504, "loss_en": 3.4330332279205322, "loss_zh": 3.4386067390441895, "noise_std_eff": 0.02, "elapsed_sec": 1081.7475299835205}
268
+ {"step": 8360, "epoch": 1, "loss": 3.097078800201416, "loss_en": 3.0986437797546387, "loss_zh": 3.0955140590667725, "noise_std_eff": 0.02, "elapsed_sec": 1084.6914646625519}
269
+ {"step": 8380, "epoch": 1, "loss": 3.4404096603393555, "loss_en": 3.413649797439575, "loss_zh": 3.4671692848205566, "noise_std_eff": 0.02, "elapsed_sec": 1087.5948841571808}
270
+ {"step": 8400, "epoch": 1, "loss": 3.5366153717041016, "loss_en": 3.5093560218811035, "loss_zh": 3.5638749599456787, "noise_std_eff": 0.02, "elapsed_sec": 1090.528380870819}
271
+ {"step": 8420, "epoch": 1, "loss": 3.4133524894714355, "loss_en": 3.3595352172851562, "loss_zh": 3.467170000076294, "noise_std_eff": 0.02, "elapsed_sec": 1117.5877361297607}
272
+ {"step": 8440, "epoch": 1, "loss": 3.0394740104675293, "loss_en": 3.0268328189849854, "loss_zh": 3.052114963531494, "noise_std_eff": 0.02, "elapsed_sec": 1120.512377023697}
273
+ {"step": 8460, "epoch": 1, "loss": 2.958902359008789, "loss_en": 2.959085464477539, "loss_zh": 2.95871901512146, "noise_std_eff": 0.02, "elapsed_sec": 1123.3559465408325}
274
+ {"step": 8480, "epoch": 1, "loss": 3.3659212589263916, "loss_en": 3.3488755226135254, "loss_zh": 3.382966995239258, "noise_std_eff": 0.02, "elapsed_sec": 1126.3148927688599}
275
+ {"step": 8500, "epoch": 1, "loss": 3.158687114715576, "loss_en": 3.149724006652832, "loss_zh": 3.167649984359741, "noise_std_eff": 0.02, "elapsed_sec": 1129.1962540149689}
276
+ {"step": 8520, "epoch": 1, "loss": 3.9505133628845215, "loss_en": 3.9929697513580322, "loss_zh": 3.9080569744110107, "noise_std_eff": 0.02, "elapsed_sec": 1132.1446695327759}
277
+ {"step": 8540, "epoch": 1, "loss": 3.3564956188201904, "loss_en": 3.37091326713562, "loss_zh": 3.3420779705047607, "noise_std_eff": 0.02, "elapsed_sec": 1135.0435736179352}
278
+ {"step": 8560, "epoch": 1, "loss": 3.6094040870666504, "loss_en": 3.6714670658111572, "loss_zh": 3.5473413467407227, "noise_std_eff": 0.02, "elapsed_sec": 1137.968647480011}
279
+ {"step": 8580, "epoch": 1, "loss": 3.651052236557007, "loss_en": 3.646430253982544, "loss_zh": 3.6556742191314697, "noise_std_eff": 0.02, "elapsed_sec": 1140.8361649513245}
280
+ {"step": 8600, "epoch": 1, "loss": 3.3115620613098145, "loss_en": 3.296211004257202, "loss_zh": 3.326913356781006, "noise_std_eff": 0.02, "elapsed_sec": 1143.7879920005798}
281
+ {"step": 8620, "epoch": 1, "loss": 3.713923931121826, "loss_en": 3.711111307144165, "loss_zh": 3.7167367935180664, "noise_std_eff": 0.02, "elapsed_sec": 1146.7081079483032}
282
+ {"step": 8640, "epoch": 1, "loss": 3.4937076568603516, "loss_en": 3.5095033645629883, "loss_zh": 3.477912187576294, "noise_std_eff": 0.02, "elapsed_sec": 1149.6316614151}
283
+ {"step": 8660, "epoch": 1, "loss": 3.2244043350219727, "loss_en": 3.206418991088867, "loss_zh": 3.242389440536499, "noise_std_eff": 0.02, "elapsed_sec": 1152.4923062324524}
284
+ {"step": 8680, "epoch": 1, "loss": 3.1785969734191895, "loss_en": 3.1892311573028564, "loss_zh": 3.1679630279541016, "noise_std_eff": 0.02, "elapsed_sec": 1155.3986847400665}
285
+ {"step": 8700, "epoch": 1, "loss": 2.6505517959594727, "loss_en": 2.621917247772217, "loss_zh": 2.6791865825653076, "noise_std_eff": 0.02, "elapsed_sec": 1158.2839970588684}
286
+ {"step": 8720, "epoch": 1, "loss": 3.404956340789795, "loss_en": 3.401346445083618, "loss_zh": 3.4085659980773926, "noise_std_eff": 0.02, "elapsed_sec": 1161.2030453681946}
287
+ {"step": 8740, "epoch": 1, "loss": 3.3213014602661133, "loss_en": 3.3192429542541504, "loss_zh": 3.323359727859497, "noise_std_eff": 0.02, "elapsed_sec": 1164.0754783153534}
288
+ {"step": 8760, "epoch": 1, "loss": 3.7199950218200684, "loss_en": 3.719074249267578, "loss_zh": 3.7209157943725586, "noise_std_eff": 0.02, "elapsed_sec": 1166.996488571167}
289
+ {"step": 8780, "epoch": 1, "loss": 3.6704421043395996, "loss_en": 3.6891250610351562, "loss_zh": 3.651758909225464, "noise_std_eff": 0.02, "elapsed_sec": 1169.8717694282532}
290
+ {"step": 8800, "epoch": 1, "loss": 3.4365949630737305, "loss_en": 3.4213976860046387, "loss_zh": 3.4517924785614014, "noise_std_eff": 0.02, "elapsed_sec": 1172.7913575172424}
291
+ {"step": 8820, "epoch": 1, "loss": 2.971341609954834, "loss_en": 2.971869707107544, "loss_zh": 2.970813751220703, "noise_std_eff": 0.02, "elapsed_sec": 1199.8678228855133}
292
+ {"step": 8840, "epoch": 1, "loss": 3.606365442276001, "loss_en": 3.561248302459717, "loss_zh": 3.651482582092285, "noise_std_eff": 0.02, "elapsed_sec": 1202.804641008377}
293
+ {"step": 8860, "epoch": 1, "loss": 3.2079081535339355, "loss_en": 3.2184481620788574, "loss_zh": 3.1973683834075928, "noise_std_eff": 0.02, "elapsed_sec": 1205.7081112861633}
294
+ {"step": 8880, "epoch": 1, "loss": 3.8009376525878906, "loss_en": 3.797619581222534, "loss_zh": 3.804255723953247, "noise_std_eff": 0.02, "elapsed_sec": 1208.6641430854797}
295
+ {"step": 8900, "epoch": 1, "loss": 3.3399109840393066, "loss_en": 3.3142800331115723, "loss_zh": 3.365541934967041, "noise_std_eff": 0.02, "elapsed_sec": 1211.6119482517242}
296
+ {"step": 8920, "epoch": 1, "loss": 4.0910539627075195, "loss_en": 4.141756057739258, "loss_zh": 4.040351390838623, "noise_std_eff": 0.02, "elapsed_sec": 1214.5522129535675}
297
+ {"step": 8940, "epoch": 1, "loss": 3.4923481941223145, "loss_en": 3.4846460819244385, "loss_zh": 3.5000503063201904, "noise_std_eff": 0.02, "elapsed_sec": 1217.4438154697418}
298
+ {"step": 8960, "epoch": 1, "loss": 3.4046425819396973, "loss_en": 3.3718087673187256, "loss_zh": 3.437476634979248, "noise_std_eff": 0.02, "elapsed_sec": 1220.3921942710876}
299
+ {"step": 8980, "epoch": 1, "loss": 3.573052406311035, "loss_en": 3.5687620639801025, "loss_zh": 3.5773425102233887, "noise_std_eff": 0.02, "elapsed_sec": 1223.341092824936}
300
+ {"step": 9000, "epoch": 1, "loss": 3.334259510040283, "loss_en": 3.333658218383789, "loss_zh": 3.3348605632781982, "noise_std_eff": 0.02, "elapsed_sec": 1226.2571430206299}
301
+ {"step": 9020, "epoch": 1, "loss": 3.245018720626831, "loss_en": 3.2383265495300293, "loss_zh": 3.251710891723633, "noise_std_eff": 0.02, "elapsed_sec": 1229.1725916862488}
302
+ {"step": 9040, "epoch": 1, "loss": 3.126258373260498, "loss_en": 3.1324641704559326, "loss_zh": 3.1200525760650635, "noise_std_eff": 0.02, "elapsed_sec": 1232.112975358963}
303
+ {"step": 9060, "epoch": 1, "loss": 3.359930992126465, "loss_en": 3.368892192840576, "loss_zh": 3.3509697914123535, "noise_std_eff": 0.02, "elapsed_sec": 1235.0170240402222}
304
+ {"step": 9080, "epoch": 1, "loss": 2.962691307067871, "loss_en": 2.962587833404541, "loss_zh": 2.962794542312622, "noise_std_eff": 0.02, "elapsed_sec": 1237.960550069809}
305
+ {"step": 9100, "epoch": 1, "loss": 3.4086079597473145, "loss_en": 3.482219934463501, "loss_zh": 3.334995985031128, "noise_std_eff": 0.02, "elapsed_sec": 1240.8608355522156}
306
+ {"step": 9120, "epoch": 1, "loss": 3.5445656776428223, "loss_en": 3.560969829559326, "loss_zh": 3.5281617641448975, "noise_std_eff": 0.02, "elapsed_sec": 1243.8135418891907}
307
+ {"step": 9140, "epoch": 1, "loss": 3.5026891231536865, "loss_en": 3.514756679534912, "loss_zh": 3.490621566772461, "noise_std_eff": 0.02, "elapsed_sec": 1246.7206304073334}
308
+ {"step": 9160, "epoch": 1, "loss": 3.265190362930298, "loss_en": 3.26944637298584, "loss_zh": 3.260934352874756, "noise_std_eff": 0.02, "elapsed_sec": 1249.676432132721}
309
+ {"step": 9180, "epoch": 1, "loss": 3.518078327178955, "loss_en": 3.4922256469726562, "loss_zh": 3.543931245803833, "noise_std_eff": 0.02, "elapsed_sec": 1252.588725566864}
310
+ {"step": 9200, "epoch": 1, "loss": 3.2692909240722656, "loss_en": 3.293487310409546, "loss_zh": 3.2450942993164062, "noise_std_eff": 0.02, "elapsed_sec": 1255.5234806537628}
311
+ {"step": 9220, "epoch": 1, "loss": 3.210805654525757, "loss_en": 3.213243246078491, "loss_zh": 3.2083680629730225, "noise_std_eff": 0.02, "elapsed_sec": 1282.4681005477905}
312
+ {"step": 9240, "epoch": 1, "loss": 3.1785707473754883, "loss_en": 3.2230684757232666, "loss_zh": 3.13407301902771, "noise_std_eff": 0.02, "elapsed_sec": 1285.3618721961975}
313
+ {"step": 9260, "epoch": 1, "loss": 2.8499364852905273, "loss_en": 2.8215818405151367, "loss_zh": 2.878290891647339, "noise_std_eff": 0.02, "elapsed_sec": 1288.2200109958649}
314
+ {"step": 9280, "epoch": 1, "loss": 3.563366413116455, "loss_en": 3.568563222885132, "loss_zh": 3.558169364929199, "noise_std_eff": 0.02, "elapsed_sec": 1291.1033000946045}
315
+ {"step": 9300, "epoch": 1, "loss": 3.6791634559631348, "loss_en": 3.6620192527770996, "loss_zh": 3.696307897567749, "noise_std_eff": 0.02, "elapsed_sec": 1293.9755144119263}
316
+ {"step": 9320, "epoch": 1, "loss": 2.810952663421631, "loss_en": 2.801631450653076, "loss_zh": 2.8202738761901855, "noise_std_eff": 0.02, "elapsed_sec": 1296.827265739441}
317
+ {"step": 9340, "epoch": 1, "loss": 3.255434513092041, "loss_en": 3.2595055103302, "loss_zh": 3.251363515853882, "noise_std_eff": 0.02, "elapsed_sec": 1299.68718957901}
318
+ {"step": 9360, "epoch": 1, "loss": 2.93479061126709, "loss_en": 2.94862961769104, "loss_zh": 2.9209513664245605, "noise_std_eff": 0.02, "elapsed_sec": 1302.5712883472443}
319
+ {"step": 9380, "epoch": 1, "loss": 3.7437868118286133, "loss_en": 3.7519898414611816, "loss_zh": 3.735583782196045, "noise_std_eff": 0.02, "elapsed_sec": 1305.4493420124054}
320
+ {"step": 9400, "epoch": 1, "loss": 3.6078150272369385, "loss_en": 3.6135714054107666, "loss_zh": 3.6020586490631104, "noise_std_eff": 0.02, "elapsed_sec": 1308.3822400569916}
321
+ {"step": 9420, "epoch": 1, "loss": 3.5729849338531494, "loss_en": 3.524242639541626, "loss_zh": 3.621727228164673, "noise_std_eff": 0.02, "elapsed_sec": 1311.275319814682}
322
+ {"step": 9440, "epoch": 1, "loss": 2.776149272918701, "loss_en": 2.8104965686798096, "loss_zh": 2.7418017387390137, "noise_std_eff": 0.02, "elapsed_sec": 1314.3346138000488}
323
+ {"step": 9460, "epoch": 1, "loss": 3.0402424335479736, "loss_en": 3.0420305728912354, "loss_zh": 3.038454294204712, "noise_std_eff": 0.02, "elapsed_sec": 1317.3010857105255}
324
+ {"step": 9480, "epoch": 1, "loss": 3.2326786518096924, "loss_en": 3.227860450744629, "loss_zh": 3.237496852874756, "noise_std_eff": 0.02, "elapsed_sec": 1320.2720623016357}
325
+ {"step": 9500, "epoch": 1, "loss": 2.909536600112915, "loss_en": 2.938655138015747, "loss_zh": 2.880418062210083, "noise_std_eff": 0.02, "elapsed_sec": 1323.1719889640808}
326
+ {"step": 9520, "epoch": 1, "loss": 3.4055352210998535, "loss_en": 3.3942902088165283, "loss_zh": 3.416780471801758, "noise_std_eff": 0.02, "elapsed_sec": 1326.1386404037476}
327
+ {"step": 9540, "epoch": 1, "loss": 3.7033846378326416, "loss_en": 3.70375919342041, "loss_zh": 3.703010082244873, "noise_std_eff": 0.02, "elapsed_sec": 1329.0491733551025}
328
+ {"step": 9560, "epoch": 1, "loss": 3.179379463195801, "loss_en": 3.183546304702759, "loss_zh": 3.175212860107422, "noise_std_eff": 0.02, "elapsed_sec": 1331.7150106430054}
329
+ {"step": 9580, "epoch": 1, "loss": 3.0954372882843018, "loss_en": 3.1280670166015625, "loss_zh": 3.062807559967041, "noise_std_eff": 0.02, "elapsed_sec": 1333.7978579998016}
330
+ {"step": 9600, "epoch": 1, "loss": 3.893620491027832, "loss_en": 3.835369348526001, "loss_zh": 3.951871395111084, "noise_std_eff": 0.02, "elapsed_sec": 1335.8966264724731}
331
+ {"step": 9620, "epoch": 1, "loss": 3.5566582679748535, "loss_en": 3.5635273456573486, "loss_zh": 3.5497891902923584, "noise_std_eff": 0.02, "elapsed_sec": 1362.8931465148926}
332
+ {"step": 9640, "epoch": 1, "loss": 3.5439484119415283, "loss_en": 3.510058641433716, "loss_zh": 3.577838182449341, "noise_std_eff": 0.02, "elapsed_sec": 1365.8385214805603}
333
+ {"step": 9660, "epoch": 1, "loss": 3.4915950298309326, "loss_en": 3.4985568523406982, "loss_zh": 3.484633207321167, "noise_std_eff": 0.02, "elapsed_sec": 1368.7177572250366}
334
+ {"step": 9680, "epoch": 1, "loss": 3.0241312980651855, "loss_en": 3.0553929805755615, "loss_zh": 2.9928698539733887, "noise_std_eff": 0.02, "elapsed_sec": 1371.6647934913635}
335
+ {"step": 9700, "epoch": 1, "loss": 2.6974692344665527, "loss_en": 2.6507272720336914, "loss_zh": 2.744211196899414, "noise_std_eff": 0.02, "elapsed_sec": 1374.568106174469}
336
+ {"step": 9720, "epoch": 1, "loss": 3.075819969177246, "loss_en": 3.0270819664001465, "loss_zh": 3.1245579719543457, "noise_std_eff": 0.02, "elapsed_sec": 1377.4806926250458}
337
+ {"step": 9740, "epoch": 1, "loss": 3.212629556655884, "loss_en": 3.2195212841033936, "loss_zh": 3.205737829208374, "noise_std_eff": 0.02, "elapsed_sec": 1380.3438513278961}
338
+ {"step": 9760, "epoch": 1, "loss": 2.8335018157958984, "loss_en": 2.8092432022094727, "loss_zh": 2.857760190963745, "noise_std_eff": 0.02, "elapsed_sec": 1383.2406747341156}
339
+ {"step": 9780, "epoch": 1, "loss": 2.8960909843444824, "loss_en": 2.8648221492767334, "loss_zh": 2.9273600578308105, "noise_std_eff": 0.02, "elapsed_sec": 1386.1158356666565}
340
+ {"step": 9800, "epoch": 1, "loss": 3.1359925270080566, "loss_en": 3.1007654666900635, "loss_zh": 3.17121958732605, "noise_std_eff": 0.02, "elapsed_sec": 1389.0317578315735}
341
+ {"step": 9820, "epoch": 1, "loss": 3.6886916160583496, "loss_en": 3.684340000152588, "loss_zh": 3.6930434703826904, "noise_std_eff": 0.02, "elapsed_sec": 1391.8751993179321}
342
+ {"step": 9840, "epoch": 1, "loss": 3.5567309856414795, "loss_en": 3.54868483543396, "loss_zh": 3.564777135848999, "noise_std_eff": 0.02, "elapsed_sec": 1394.7629871368408}
343
+ {"step": 9860, "epoch": 1, "loss": 4.016868591308594, "loss_en": 4.038916110992432, "loss_zh": 3.994821310043335, "noise_std_eff": 0.02, "elapsed_sec": 1397.6397442817688}
344
+ {"step": 9880, "epoch": 1, "loss": 3.74774169921875, "loss_en": 3.748748779296875, "loss_zh": 3.746734619140625, "noise_std_eff": 0.02, "elapsed_sec": 1400.5333745479584}
345
+ {"step": 9900, "epoch": 1, "loss": 2.917865753173828, "loss_en": 2.9187116622924805, "loss_zh": 2.9170196056365967, "noise_std_eff": 0.02, "elapsed_sec": 1403.3958008289337}
346
+ {"step": 9920, "epoch": 1, "loss": 3.552086353302002, "loss_en": 3.5330448150634766, "loss_zh": 3.5711281299591064, "noise_std_eff": 0.02, "elapsed_sec": 1406.2962036132812}
347
+ {"step": 9940, "epoch": 1, "loss": 3.4637551307678223, "loss_en": 3.447381019592285, "loss_zh": 3.4801294803619385, "noise_std_eff": 0.02, "elapsed_sec": 1409.136174917221}
348
+ {"step": 9960, "epoch": 1, "loss": 2.9836459159851074, "loss_en": 2.9518063068389893, "loss_zh": 3.0154855251312256, "noise_std_eff": 0.02, "elapsed_sec": 1412.067943096161}
349
+ {"step": 9980, "epoch": 1, "loss": 3.2497966289520264, "loss_en": 3.2372090816497803, "loss_zh": 3.2623841762542725, "noise_std_eff": 0.02, "elapsed_sec": 1414.9364228248596}
350
+ {"step": 10000, "epoch": 1, "loss": 2.668426275253296, "loss_en": 2.709928274154663, "loss_zh": 2.6269242763519287, "noise_std_eff": 0.02, "elapsed_sec": 1417.8408153057098}
351
+ {"step": 10020, "epoch": 1, "loss": 3.9161906242370605, "loss_en": 3.8886733055114746, "loss_zh": 3.9437077045440674, "noise_std_eff": 0.02, "elapsed_sec": 1444.752149105072}
352
+ {"step": 10040, "epoch": 1, "loss": 3.011127471923828, "loss_en": 3.0138704776763916, "loss_zh": 3.0083847045898438, "noise_std_eff": 0.02, "elapsed_sec": 1447.660686969757}
353
+ {"step": 10060, "epoch": 1, "loss": 3.4956982135772705, "loss_en": 3.4414138793945312, "loss_zh": 3.5499825477600098, "noise_std_eff": 0.02, "elapsed_sec": 1450.4963133335114}
354
+ {"step": 10080, "epoch": 1, "loss": 3.062112331390381, "loss_en": 3.022447109222412, "loss_zh": 3.1017777919769287, "noise_std_eff": 0.02, "elapsed_sec": 1453.448617219925}
355
+ {"step": 10100, "epoch": 1, "loss": 3.0378146171569824, "loss_en": 3.019439697265625, "loss_zh": 3.05618953704834, "noise_std_eff": 0.02, "elapsed_sec": 1456.3235743045807}
356
+ {"step": 10120, "epoch": 1, "loss": 3.1066527366638184, "loss_en": 3.12550950050354, "loss_zh": 3.0877957344055176, "noise_std_eff": 0.02, "elapsed_sec": 1459.243254184723}
357
+ {"step": 10140, "epoch": 1, "loss": 3.5539989471435547, "loss_en": 3.535261631011963, "loss_zh": 3.5727365016937256, "noise_std_eff": 0.02, "elapsed_sec": 1462.111934185028}
358
+ {"step": 10160, "epoch": 1, "loss": 3.399738311767578, "loss_en": 3.360159158706665, "loss_zh": 3.439317464828491, "noise_std_eff": 0.02, "elapsed_sec": 1465.0560624599457}
359
+ {"step": 10180, "epoch": 1, "loss": 3.06882905960083, "loss_en": 3.0633111000061035, "loss_zh": 3.0743467807769775, "noise_std_eff": 0.02, "elapsed_sec": 1467.904254436493}
360
+ {"step": 10200, "epoch": 1, "loss": 3.332425117492676, "loss_en": 3.302155017852783, "loss_zh": 3.3626954555511475, "noise_std_eff": 0.02, "elapsed_sec": 1470.7603452205658}
361
+ {"step": 10220, "epoch": 1, "loss": 3.4747915267944336, "loss_en": 3.460792064666748, "loss_zh": 3.488790988922119, "noise_std_eff": 0.02, "elapsed_sec": 1473.6036977767944}
362
+ {"step": 10240, "epoch": 1, "loss": 2.9717283248901367, "loss_en": 2.97711443901062, "loss_zh": 2.9663424491882324, "noise_std_eff": 0.02, "elapsed_sec": 1476.515370130539}
363
+ {"step": 10260, "epoch": 1, "loss": 3.4915051460266113, "loss_en": 3.494748830795288, "loss_zh": 3.4882616996765137, "noise_std_eff": 0.02, "elapsed_sec": 1479.3752303123474}
364
+ {"step": 10280, "epoch": 1, "loss": 3.8219423294067383, "loss_en": 3.7869279384613037, "loss_zh": 3.856956720352173, "noise_std_eff": 0.02, "elapsed_sec": 1482.2713205814362}
365
+ {"step": 10300, "epoch": 1, "loss": 3.7315545082092285, "loss_en": 3.6890387535095215, "loss_zh": 3.7740705013275146, "noise_std_eff": 0.02, "elapsed_sec": 1485.1236820220947}
366
+ {"step": 10320, "epoch": 1, "loss": 3.174440383911133, "loss_en": 3.203169822692871, "loss_zh": 3.1457111835479736, "noise_std_eff": 0.02, "elapsed_sec": 1488.0388824939728}
367
+ {"step": 10340, "epoch": 1, "loss": 3.183246612548828, "loss_en": 3.198936700820923, "loss_zh": 3.1675567626953125, "noise_std_eff": 0.02, "elapsed_sec": 1490.9244680404663}
368
+ {"step": 10360, "epoch": 1, "loss": 3.2914037704467773, "loss_en": 3.2209668159484863, "loss_zh": 3.3618407249450684, "noise_std_eff": 0.02, "elapsed_sec": 1493.8283281326294}
369
+ {"step": 10380, "epoch": 1, "loss": 3.3127784729003906, "loss_en": 3.347393035888672, "loss_zh": 3.2781639099121094, "noise_std_eff": 0.02, "elapsed_sec": 1496.656255722046}
370
+ {"step": 10400, "epoch": 1, "loss": 3.3279590606689453, "loss_en": 3.3263771533966064, "loss_zh": 3.329540967941284, "noise_std_eff": 0.02, "elapsed_sec": 1499.5640444755554}
371
+ {"step": 10420, "epoch": 1, "loss": 4.0600996017456055, "loss_en": 4.040959358215332, "loss_zh": 4.079239845275879, "noise_std_eff": 0.02, "elapsed_sec": 1526.9682829380035}
372
+ {"step": 10440, "epoch": 1, "loss": 3.3214802742004395, "loss_en": 3.2924246788024902, "loss_zh": 3.3505358695983887, "noise_std_eff": 0.02, "elapsed_sec": 1529.8689625263214}
373
+ {"step": 10460, "epoch": 1, "loss": 3.3753581047058105, "loss_en": 3.395740032196045, "loss_zh": 3.354976177215576, "noise_std_eff": 0.02, "elapsed_sec": 1532.7328288555145}
374
+ {"step": 10480, "epoch": 1, "loss": 3.2100467681884766, "loss_en": 3.1962575912475586, "loss_zh": 3.2238361835479736, "noise_std_eff": 0.02, "elapsed_sec": 1535.6641738414764}
375
+ {"step": 10500, "epoch": 1, "loss": 3.681504726409912, "loss_en": 3.6808178424835205, "loss_zh": 3.6821916103363037, "noise_std_eff": 0.02, "elapsed_sec": 1538.5843787193298}
376
+ {"step": 10520, "epoch": 1, "loss": 3.6703927516937256, "loss_en": 3.6370577812194824, "loss_zh": 3.7037277221679688, "noise_std_eff": 0.02, "elapsed_sec": 1541.5044856071472}
377
+ {"step": 10540, "epoch": 1, "loss": 3.2247982025146484, "loss_en": 3.2170755863189697, "loss_zh": 3.232520580291748, "noise_std_eff": 0.02, "elapsed_sec": 1544.4119610786438}
378
+ {"step": 10560, "epoch": 1, "loss": 3.815093994140625, "loss_en": 3.800166130065918, "loss_zh": 3.830021619796753, "noise_std_eff": 0.02, "elapsed_sec": 1547.3040671348572}
379
+ {"step": 10580, "epoch": 1, "loss": 2.9970760345458984, "loss_en": 3.0048232078552246, "loss_zh": 2.989328622817993, "noise_std_eff": 0.02, "elapsed_sec": 1550.1689848899841}
380
+ {"step": 10600, "epoch": 1, "loss": 3.504970073699951, "loss_en": 3.498091697692871, "loss_zh": 3.511848211288452, "noise_std_eff": 0.02, "elapsed_sec": 1553.1180567741394}
381
+ {"step": 10620, "epoch": 1, "loss": 2.835435152053833, "loss_en": 2.855597496032715, "loss_zh": 2.815272808074951, "noise_std_eff": 0.02, "elapsed_sec": 1556.3395318984985}
382
+ {"step": 10640, "epoch": 1, "loss": 3.4992551803588867, "loss_en": 3.504215955734253, "loss_zh": 3.4942946434020996, "noise_std_eff": 0.02, "elapsed_sec": 1559.4191091060638}
383
+ {"step": 10660, "epoch": 1, "loss": 3.5471363067626953, "loss_en": 3.556978702545166, "loss_zh": 3.5372941493988037, "noise_std_eff": 0.02, "elapsed_sec": 1562.5398936271667}
384
+ {"step": 10680, "epoch": 1, "loss": 3.7108540534973145, "loss_en": 3.694578170776367, "loss_zh": 3.727130174636841, "noise_std_eff": 0.02, "elapsed_sec": 1565.7807734012604}
385
+ {"step": 10700, "epoch": 1, "loss": 3.247375965118408, "loss_en": 3.237027883529663, "loss_zh": 3.2577242851257324, "noise_std_eff": 0.02, "elapsed_sec": 1568.7601079940796}
386
+ {"step": 10720, "epoch": 1, "loss": 3.5087337493896484, "loss_en": 3.479976177215576, "loss_zh": 3.5374913215637207, "noise_std_eff": 0.02, "elapsed_sec": 1571.8509001731873}
387
+ {"step": 10740, "epoch": 1, "loss": 3.406184673309326, "loss_en": 3.4171764850616455, "loss_zh": 3.3951926231384277, "noise_std_eff": 0.02, "elapsed_sec": 1574.7766845226288}
388
+ {"step": 10760, "epoch": 1, "loss": 3.450247049331665, "loss_en": 3.485442638397217, "loss_zh": 3.4150514602661133, "noise_std_eff": 0.02, "elapsed_sec": 1577.7636210918427}
389
+ {"step": 10780, "epoch": 1, "loss": 3.6479225158691406, "loss_en": 3.659047842025757, "loss_zh": 3.6367974281311035, "noise_std_eff": 0.02, "elapsed_sec": 1580.6572058200836}
390
+ {"step": 10800, "epoch": 1, "loss": 3.024441957473755, "loss_en": 3.000196695327759, "loss_zh": 3.048687219619751, "noise_std_eff": 0.02, "elapsed_sec": 1583.5672419071198}
391
+ {"step": 10820, "epoch": 1, "loss": 3.494462013244629, "loss_en": 3.537336826324463, "loss_zh": 3.451587200164795, "noise_std_eff": 0.02, "elapsed_sec": 1610.7109906673431}
392
+ {"step": 10840, "epoch": 1, "loss": 3.244837760925293, "loss_en": 3.2315120697021484, "loss_zh": 3.2581632137298584, "noise_std_eff": 0.02, "elapsed_sec": 1613.6680808067322}
393
+ {"step": 10860, "epoch": 1, "loss": 3.6488757133483887, "loss_en": 3.6631064414978027, "loss_zh": 3.6346449851989746, "noise_std_eff": 0.02, "elapsed_sec": 1616.5796146392822}
394
+ {"step": 10880, "epoch": 1, "loss": 3.3179469108581543, "loss_en": 3.334946870803833, "loss_zh": 3.3009469509124756, "noise_std_eff": 0.02, "elapsed_sec": 1619.511928319931}
395
+ {"step": 10900, "epoch": 1, "loss": 3.2986836433410645, "loss_en": 3.2868969440460205, "loss_zh": 3.3104703426361084, "noise_std_eff": 0.02, "elapsed_sec": 1622.4275875091553}
396
+ {"step": 10920, "epoch": 1, "loss": 3.4335131645202637, "loss_en": 3.4208872318267822, "loss_zh": 3.446139335632324, "noise_std_eff": 0.02, "elapsed_sec": 1624.579290151596}
397
+ {"step": 10940, "epoch": 1, "loss": 3.2822647094726562, "loss_en": 3.2569310665130615, "loss_zh": 3.30759859085083, "noise_std_eff": 0.02, "elapsed_sec": 1626.7205007076263}
398
+ {"step": 10960, "epoch": 1, "loss": 3.261011838912964, "loss_en": 3.2152743339538574, "loss_zh": 3.3067493438720703, "noise_std_eff": 0.02, "elapsed_sec": 1628.8535211086273}
399
+ {"step": 10980, "epoch": 1, "loss": 3.5365254878997803, "loss_en": 3.519507884979248, "loss_zh": 3.5535430908203125, "noise_std_eff": 0.02, "elapsed_sec": 1631.471880197525}
400
+ {"step": 11000, "epoch": 1, "loss": 3.2532992362976074, "loss_en": 3.2197210788726807, "loss_zh": 3.286877393722534, "noise_std_eff": 0.02, "elapsed_sec": 1634.3685557842255}
401
+ {"step": 11020, "epoch": 1, "loss": 3.431036949157715, "loss_en": 3.425062894821167, "loss_zh": 3.437011241912842, "noise_std_eff": 0.02, "elapsed_sec": 1637.2642376422882}
402
+ {"step": 11040, "epoch": 1, "loss": 2.8710808753967285, "loss_en": 2.8554575443267822, "loss_zh": 2.8867039680480957, "noise_std_eff": 0.02, "elapsed_sec": 1640.1676456928253}
403
+ {"step": 11060, "epoch": 1, "loss": 3.1430740356445312, "loss_en": 3.145709753036499, "loss_zh": 3.1404383182525635, "noise_std_eff": 0.02, "elapsed_sec": 1643.04105758667}
404
+ {"step": 11080, "epoch": 1, "loss": 3.298856258392334, "loss_en": 3.290581703186035, "loss_zh": 3.307131052017212, "noise_std_eff": 0.02, "elapsed_sec": 1645.928132534027}
405
+ {"step": 11100, "epoch": 1, "loss": 3.7378334999084473, "loss_en": 3.7240560054779053, "loss_zh": 3.7516112327575684, "noise_std_eff": 0.02, "elapsed_sec": 1648.7849504947662}
406
+ {"step": 11120, "epoch": 1, "loss": 3.617753505706787, "loss_en": 3.639986038208008, "loss_zh": 3.5955212116241455, "noise_std_eff": 0.02, "elapsed_sec": 1651.6992135047913}
407
+ {"step": 11140, "epoch": 1, "loss": 3.3081412315368652, "loss_en": 3.3074071407318115, "loss_zh": 3.308875322341919, "noise_std_eff": 0.02, "elapsed_sec": 1654.5876348018646}
408
+ {"step": 11160, "epoch": 1, "loss": 3.6678991317749023, "loss_en": 3.630315065383911, "loss_zh": 3.7054829597473145, "noise_std_eff": 0.02, "elapsed_sec": 1657.5091943740845}
409
+ {"step": 11180, "epoch": 1, "loss": 3.463718891143799, "loss_en": 3.4727795124053955, "loss_zh": 3.4546585083007812, "noise_std_eff": 0.02, "elapsed_sec": 1660.436321735382}
410
+ {"step": 11200, "epoch": 1, "loss": 3.217634916305542, "loss_en": 3.1734352111816406, "loss_zh": 3.2618346214294434, "noise_std_eff": 0.02, "elapsed_sec": 1663.3619244098663}
411
+ {"step": 11220, "epoch": 1, "loss": 3.524181365966797, "loss_en": 3.5006556510925293, "loss_zh": 3.5477070808410645, "noise_std_eff": 0.02, "elapsed_sec": 1690.4711685180664}
412
+ {"step": 11240, "epoch": 1, "loss": 3.391510486602783, "loss_en": 3.3562753200531006, "loss_zh": 3.426745653152466, "noise_std_eff": 0.02, "elapsed_sec": 1693.384967803955}
413
+ {"step": 11260, "epoch": 1, "loss": 3.1506600379943848, "loss_en": 3.1622154712677, "loss_zh": 3.1391048431396484, "noise_std_eff": 0.02, "elapsed_sec": 1696.2786130905151}
414
+ {"step": 11280, "epoch": 1, "loss": 3.369452714920044, "loss_en": 3.3520138263702393, "loss_zh": 3.3868916034698486, "noise_std_eff": 0.02, "elapsed_sec": 1699.2073512077332}
415
+ {"step": 11300, "epoch": 1, "loss": 4.0016913414001465, "loss_en": 3.946560859680176, "loss_zh": 4.056821823120117, "noise_std_eff": 0.02, "elapsed_sec": 1702.052677154541}
416
+ {"step": 11320, "epoch": 1, "loss": 3.3226590156555176, "loss_en": 3.3696305751800537, "loss_zh": 3.2756876945495605, "noise_std_eff": 0.02, "elapsed_sec": 1705.012306213379}
417
+ {"step": 11340, "epoch": 1, "loss": 3.496060609817505, "loss_en": 3.4134819507598877, "loss_zh": 3.578639268875122, "noise_std_eff": 0.02, "elapsed_sec": 1707.8959321975708}
418
+ {"step": 11360, "epoch": 1, "loss": 3.527528762817383, "loss_en": 3.5109546184539795, "loss_zh": 3.544102907180786, "noise_std_eff": 0.02, "elapsed_sec": 1710.8617084026337}
419
+ {"step": 11380, "epoch": 1, "loss": 3.2562499046325684, "loss_en": 3.2774598598480225, "loss_zh": 3.2350399494171143, "noise_std_eff": 0.02, "elapsed_sec": 1713.8047201633453}
420
+ {"step": 11400, "epoch": 1, "loss": 3.1304750442504883, "loss_en": 3.0880942344665527, "loss_zh": 3.172855854034424, "noise_std_eff": 0.02, "elapsed_sec": 1716.7574934959412}
421
+ {"step": 11420, "epoch": 1, "loss": 3.362283229827881, "loss_en": 3.3526294231414795, "loss_zh": 3.3719370365142822, "noise_std_eff": 0.02, "elapsed_sec": 1719.6438579559326}
422
+ {"step": 11440, "epoch": 1, "loss": 2.9460349082946777, "loss_en": 2.943371057510376, "loss_zh": 2.9486987590789795, "noise_std_eff": 0.02, "elapsed_sec": 1722.5371112823486}
423
+ {"step": 11460, "epoch": 1, "loss": 3.084167957305908, "loss_en": 3.0857737064361572, "loss_zh": 3.0825624465942383, "noise_std_eff": 0.02, "elapsed_sec": 1725.4287686347961}
424
+ {"step": 11480, "epoch": 1, "loss": 3.1885719299316406, "loss_en": 3.1640894412994385, "loss_zh": 3.2130541801452637, "noise_std_eff": 0.02, "elapsed_sec": 1728.3444428443909}
425
+ {"step": 11500, "epoch": 1, "loss": 3.3997201919555664, "loss_en": 3.3886094093322754, "loss_zh": 3.4108312129974365, "noise_std_eff": 0.02, "elapsed_sec": 1731.2170462608337}
426
+ {"step": 11520, "epoch": 1, "loss": 3.751380443572998, "loss_en": 3.740572929382324, "loss_zh": 3.762187957763672, "noise_std_eff": 0.02, "elapsed_sec": 1734.1297779083252}
427
+ {"step": 11540, "epoch": 1, "loss": 3.2481963634490967, "loss_en": 3.2742762565612793, "loss_zh": 3.222116470336914, "noise_std_eff": 0.02, "elapsed_sec": 1737.0084586143494}
428
+ {"step": 11560, "epoch": 1, "loss": 3.2114992141723633, "loss_en": 3.210498094558716, "loss_zh": 3.2125003337860107, "noise_std_eff": 0.02, "elapsed_sec": 1739.9376184940338}
429
+ {"step": 11580, "epoch": 1, "loss": 3.3716084957122803, "loss_en": 3.362886428833008, "loss_zh": 3.3803305625915527, "noise_std_eff": 0.02, "elapsed_sec": 1742.8330080509186}
430
+ {"step": 11600, "epoch": 1, "loss": 3.311739444732666, "loss_en": 3.380521297454834, "loss_zh": 3.242957353591919, "noise_std_eff": 0.02, "elapsed_sec": 1745.7604794502258}
431
+ {"step": 11620, "epoch": 1, "loss": 3.3163981437683105, "loss_en": 3.3573226928710938, "loss_zh": 3.2754733562469482, "noise_std_eff": 0.02, "elapsed_sec": 1772.747963666916}
432
+ {"step": 11640, "epoch": 1, "loss": 4.454227924346924, "loss_en": 4.475417137145996, "loss_zh": 4.433038711547852, "noise_std_eff": 0.02, "elapsed_sec": 1775.6794142723083}
433
+ {"step": 11660, "epoch": 1, "loss": 3.398017406463623, "loss_en": 3.371140241622925, "loss_zh": 3.4248945713043213, "noise_std_eff": 0.02, "elapsed_sec": 1778.5637896060944}
434
+ {"step": 11680, "epoch": 1, "loss": 3.1912455558776855, "loss_en": 3.1869046688079834, "loss_zh": 3.1955864429473877, "noise_std_eff": 0.02, "elapsed_sec": 1781.4829857349396}
435
+ {"step": 11700, "epoch": 1, "loss": 2.860600709915161, "loss_en": 2.859208583831787, "loss_zh": 2.861992835998535, "noise_std_eff": 0.02, "elapsed_sec": 1784.34414768219}
436
+ {"step": 11720, "epoch": 1, "loss": 3.481335163116455, "loss_en": 3.50213623046875, "loss_zh": 3.460533857345581, "noise_std_eff": 0.02, "elapsed_sec": 1787.2593185901642}
437
+ {"step": 11740, "epoch": 1, "loss": 3.7146332263946533, "loss_en": 3.706204652786255, "loss_zh": 3.7230618000030518, "noise_std_eff": 0.02, "elapsed_sec": 1790.143584728241}
438
+ {"step": 11760, "epoch": 1, "loss": 2.7980828285217285, "loss_en": 2.762160301208496, "loss_zh": 2.834005117416382, "noise_std_eff": 0.02, "elapsed_sec": 1793.0393986701965}
439
+ {"step": 11780, "epoch": 1, "loss": 3.8022875785827637, "loss_en": 3.868492364883423, "loss_zh": 3.7360830307006836, "noise_std_eff": 0.02, "elapsed_sec": 1795.9314260482788}
440
+ {"step": 11800, "epoch": 1, "loss": 2.6502444744110107, "loss_en": 2.6344656944274902, "loss_zh": 2.6660232543945312, "noise_std_eff": 0.02, "elapsed_sec": 1798.8534326553345}
441
+ {"step": 11820, "epoch": 1, "loss": 3.3534786701202393, "loss_en": 3.32834792137146, "loss_zh": 3.3786094188690186, "noise_std_eff": 0.02, "elapsed_sec": 1801.7366552352905}
442
+ {"step": 11840, "epoch": 1, "loss": 3.5683703422546387, "loss_en": 3.5581061840057373, "loss_zh": 3.57863450050354, "noise_std_eff": 0.02, "elapsed_sec": 1804.652081489563}
443
+ {"step": 11860, "epoch": 1, "loss": 3.3547487258911133, "loss_en": 3.3586676120758057, "loss_zh": 3.350830078125, "noise_std_eff": 0.02, "elapsed_sec": 1807.5234742164612}
444
+ {"step": 11880, "epoch": 1, "loss": 3.3271148204803467, "loss_en": 3.321084499359131, "loss_zh": 3.3331451416015625, "noise_std_eff": 0.02, "elapsed_sec": 1810.483604669571}
445
+ {"step": 11900, "epoch": 1, "loss": 3.3949475288391113, "loss_en": 3.346667766571045, "loss_zh": 3.4432272911071777, "noise_std_eff": 0.02, "elapsed_sec": 1813.3753745555878}
446
+ {"step": 11920, "epoch": 1, "loss": 3.1699318885803223, "loss_en": 3.1746206283569336, "loss_zh": 3.165242910385132, "noise_std_eff": 0.02, "elapsed_sec": 1816.3364942073822}
447
+ {"step": 11940, "epoch": 1, "loss": 3.177377223968506, "loss_en": 3.191964626312256, "loss_zh": 3.162789821624756, "noise_std_eff": 0.02, "elapsed_sec": 1819.2201108932495}
448
+ {"step": 11960, "epoch": 1, "loss": 3.6673755645751953, "loss_en": 3.6868767738342285, "loss_zh": 3.647874593734741, "noise_std_eff": 0.02, "elapsed_sec": 1822.1436755657196}
planB_polish_robust_dec_noise02/probe_leakage.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_noise02/ckpt.pt",
3
+ "max_train_groups": 8000,
4
+ "max_valid_groups": 946,
5
+ "probe_hidden": 256,
6
+ "epochs": 20,
7
+ "lr": 0.01,
8
+ "baseline_majority_lang_acc": 0.5,
9
+ "baseline_majority_len_acc": 0.9271249771118164,
10
+ "lang_probe": {
11
+ "train_acc": 0.5859375,
12
+ "valid_acc": 0.5808668076109936,
13
+ "best_valid_acc": 0.5924947145877378
14
+ },
15
+ "len_probe": {
16
+ "train_acc": 0.9290625,
17
+ "valid_acc": 0.9011627906976745,
18
+ "best_valid_acc": 0.9011627906976745
19
+ },
20
+ "lang_probe_length_matched": {
21
+ "train_acc": 0.5941032887975334,
22
+ "valid_acc": 0.5992282249173098,
23
+ "best_valid_acc": 0.6025358324145534
24
+ },
25
+ "lang_bucketwise_valid_acc": {
26
+ "bucket_0_acc": 0.34375,
27
+ "bucket_1_acc": 0.625,
28
+ "bucket_2_acc": 0.5826086401939392,
29
+ "bucket_3_acc": 0.5841642618179321
30
+ },
31
+ "length_matched_sizes": {
32
+ "train": 15568,
33
+ "valid": 1814
34
+ }
35
+ }
planB_polish_robust_dec_noise05/ckpt.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ecc5b07aa8fc56f6cd5e951b3cea2637c90ab81fde9642dc9a2f738ef4e5796
3
+ size 2439403193
planB_polish_robust_dec_noise05/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_jsonl": "data/groups_train.jsonl",
3
+ "valid_jsonl": "data/groups_valid.jsonl",
4
+ "run_dir": "runs/planB_polish_robust_dec_noise05",
5
+ "resume": "runs/planB_polish/ckpt.pt",
6
+ "backbone": "google/mt5-small",
7
+ "num_latents": 16,
8
+ "batch_size": 4,
9
+ "grad_accum": 8,
10
+ "epochs": 2,
11
+ "max_doc_len": 256,
12
+ "max_sum_len": 64,
13
+ "eval_every": 400,
14
+ "max_train_examples": 0,
15
+ "max_valid_examples": 0,
16
+ "lr": 0.0002,
17
+ "weight_decay": 0.0,
18
+ "latent_dropout": 0.2,
19
+ "latent_noise_std": 0.05,
20
+ "noise_warmup_steps": 1200,
21
+ "noise_warmup_start_step": -1,
22
+ "untie_lm_head": true,
23
+ "seed": 42
24
+ }
planB_polish_robust_dec_noise05/diag.jsonl ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3200, "nll_en": 4.782749432085937, "nll_zh": 4.7874990396721415, "nll": 4.78512423587904}
2
+ {"step": 3600, "nll_en": 4.483191290567087, "nll_zh": 4.486125991409979, "nll": 4.484658640988533}
3
+ {"step": 4000, "nll_en": 4.2837101581484776, "nll_zh": 4.286820159654033, "nll": 4.285265158901255}
4
+ {"step": 4400, "nll_en": 4.157541248813492, "nll_zh": 4.162395651698365, "nll": 4.159968450255929}
5
+ {"step": 4800, "nll_en": 4.055126797069203, "nll_zh": 4.059964848371187, "nll": 4.057545822720195}
6
+ {"step": 5200, "nll_en": 3.9836253515005615, "nll_zh": 3.986341521805487, "nll": 3.984983436653024}
7
+ {"step": 5600, "nll_en": 3.916977587008325, "nll_zh": 3.917651007089756, "nll": 3.9173142970490407}
8
+ {"step": 6000, "nll_en": 3.8548827584651764, "nll_zh": 3.857142798240291, "nll": 3.8560127783527336}
9
+ {"step": 6400, "nll_en": 3.8102813674329954, "nll_zh": 3.8142959501980225, "nll": 3.812288658815509}
10
+ {"step": 6800, "nll_en": 3.771744955921778, "nll_zh": 3.773628197563123, "nll": 3.7726865767424504}
11
+ {"step": 7200, "nll_en": 3.7292228158382232, "nll_zh": 3.7323989928650807, "nll": 3.7308109043516517}
12
+ {"step": 7600, "nll_en": 3.940199995141705, "nll_zh": 3.9392004567523333, "nll": 3.939700225947019}
13
+ {"step": 8000, "nll_en": 3.6888686345445176, "nll_zh": 3.6935126524394963, "nll": 3.691190643492007}
14
+ {"step": 8400, "nll_en": 3.6482749186913006, "nll_zh": 3.651625295522097, "nll": 3.6499501071066986}
15
+ {"step": 8800, "nll_en": 3.6288658136041168, "nll_zh": 3.634234065485303, "nll": 3.63154993954471}
16
+ {"step": 9200, "nll_en": 3.603973396736774, "nll_zh": 3.6096010319022245, "nll": 3.6067872143194992}
17
+ {"step": 9600, "nll_en": 3.5898498591610544, "nll_zh": 3.591751549259049, "nll": 3.5908007042100514}
18
+ {"step": 10000, "nll_en": 3.562294121272468, "nll_zh": 3.568615640192657, "nll": 3.5654548807325623}
19
+ {"step": 10400, "nll_en": 3.5467458093897233, "nll_zh": 3.5529534267320693, "nll": 3.5498496180608963}
20
+ {"step": 10800, "nll_en": 3.5316718355545786, "nll_zh": 3.5352415292510018, "nll": 3.5334566824027904}
21
+ {"step": 11200, "nll_en": 3.5158646474680992, "nll_zh": 3.5186522818518995, "nll": 3.5172584646599994}
22
+ {"step": 11600, "nll_en": 3.499430883258392, "nll_zh": 3.5046755718126357, "nll": 3.502053227535514}
planB_polish_robust_dec_noise05/diag_full.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_noise05/ckpt.pt",
3
+ "nll_en": 3.4998955807272525,
4
+ "nll_zh": 3.502026866404753,
5
+ "swap_delta_en": 0.5621915478534981,
6
+ "swap_delta_zh": 0.5458833702522907,
7
+ "ablate_zero_delta_en": 1.5891318159930046,
8
+ "ablate_mean_delta_en": 0.24496816530288149,
9
+ "ablate_noise_delta_en": 5.406745256379593,
10
+ "ablate_zero_delta_zh": 1.587000530315504,
11
+ "ablate_mean_delta_zh": 0.23715047654843482,
12
+ "ablate_noise_delta_zh": 5.408196035953707,
13
+ "inv_top1_full": 0.060253698378801346,
14
+ "inv_top5_full": 0.20401690900325775,
15
+ "diag_sim_mean": 0.965337336063385,
16
+ "offdiag_sim_mean": 0.4495629370212555,
17
+ "sim_margin": 0.5157743990421295,
18
+ "n_valid": 946
19
+ }
planB_polish_robust_dec_noise05/logs.jsonl ADDED
@@ -0,0 +1,448 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3020, "epoch": 0, "loss": 5.887317180633545, "loss_en": 5.860618591308594, "loss_zh": 5.914015769958496, "noise_std_eff": 0.00011694790174563726, "elapsed_sec": 2.8009402751922607}
2
+ {"step": 3040, "epoch": 0, "loss": 5.037763595581055, "loss_en": 5.014706611633301, "loss_zh": 5.060820579528809, "noise_std_eff": 0.001462951203187307, "elapsed_sec": 5.006574869155884}
3
+ {"step": 3060, "epoch": 0, "loss": 5.405976295471191, "loss_en": 5.385055065155029, "loss_zh": 5.426897048950195, "noise_std_eff": 0.0024880659580230717, "elapsed_sec": 7.344042778015137}
4
+ {"step": 3080, "epoch": 0, "loss": 5.63817024230957, "loss_en": 5.696621417999268, "loss_zh": 5.579719543457031, "noise_std_eff": 0.0028210546573003135, "elapsed_sec": 9.66260027885437}
5
+ {"step": 3100, "epoch": 0, "loss": 5.363603591918945, "loss_en": 5.2590742111206055, "loss_zh": 5.468132972717285, "noise_std_eff": 0.001801346739133199, "elapsed_sec": 11.93059754371643}
6
+ {"step": 3120, "epoch": 0, "loss": 5.772175312042236, "loss_en": 5.838954925537109, "loss_zh": 5.705395698547363, "noise_std_eff": 0.00010210481472313406, "elapsed_sec": 14.328672409057617}
7
+ {"step": 3140, "epoch": 0, "loss": 5.058171272277832, "loss_en": 5.0176615715026855, "loss_zh": 5.098681449890137, "noise_std_eff": 0.0005156687026222548, "elapsed_sec": 16.494669437408447}
8
+ {"step": 3160, "epoch": 0, "loss": 5.743980407714844, "loss_en": 5.923311233520508, "loss_zh": 5.56464958190918, "noise_std_eff": 0.0021588399012883504, "elapsed_sec": 18.909644842147827}
9
+ {"step": 3180, "epoch": 0, "loss": 5.129791736602783, "loss_en": 5.213996410369873, "loss_zh": 5.045587062835693, "noise_std_eff": 0.0003853753488510847, "elapsed_sec": 21.079617738723755}
10
+ {"step": 3200, "epoch": 0, "loss": 5.271066665649414, "loss_en": 5.373752117156982, "loss_zh": 5.168381214141846, "noise_std_eff": 0.002987680087486903, "elapsed_sec": 23.460620403289795}
11
+ {"step": 3220, "epoch": 0, "loss": 5.0961480140686035, "loss_en": 5.089468955993652, "loss_zh": 5.102827072143555, "noise_std_eff": 0.001563828798631827, "elapsed_sec": 42.3618700504303}
12
+ {"step": 3240, "epoch": 0, "loss": 5.1359100341796875, "loss_en": 5.114530086517334, "loss_zh": 5.157290458679199, "noise_std_eff": 0.008970178365707399, "elapsed_sec": 44.50108289718628}
13
+ {"step": 3260, "epoch": 0, "loss": 4.854958534240723, "loss_en": 4.848080635070801, "loss_zh": 4.861836910247803, "noise_std_eff": 0.002166225835680962, "elapsed_sec": 46.77178168296814}
14
+ {"step": 3280, "epoch": 0, "loss": 5.436870574951172, "loss_en": 5.440090656280518, "loss_zh": 5.433650970458984, "noise_std_eff": 0.006722756922245027, "elapsed_sec": 48.92801070213318}
15
+ {"step": 3300, "epoch": 0, "loss": 4.695265769958496, "loss_en": 4.680610179901123, "loss_zh": 4.709921836853027, "noise_std_eff": 0.0011112839914858342, "elapsed_sec": 51.235702991485596}
16
+ {"step": 3320, "epoch": 0, "loss": 4.350317001342773, "loss_en": 4.388892650604248, "loss_zh": 4.311741352081299, "noise_std_eff": 0.010961025555928549, "elapsed_sec": 53.352341175079346}
17
+ {"step": 3340, "epoch": 0, "loss": 5.050292015075684, "loss_en": 5.063292026519775, "loss_zh": 5.037292003631592, "noise_std_eff": 0.00868033359448115, "elapsed_sec": 55.655781745910645}
18
+ {"step": 3360, "epoch": 0, "loss": 4.458961486816406, "loss_en": 4.453353404998779, "loss_zh": 4.464569568634033, "noise_std_eff": 0.013688298761844636, "elapsed_sec": 57.98276233673096}
19
+ {"step": 3380, "epoch": 0, "loss": 4.5106096267700195, "loss_en": 4.541741371154785, "loss_zh": 4.479477405548096, "noise_std_eff": 0.010083726346492768, "elapsed_sec": 60.02154040336609}
20
+ {"step": 3400, "epoch": 0, "loss": 4.3357672691345215, "loss_en": 4.275353908538818, "loss_zh": 4.396180629730225, "noise_std_eff": 0.00916672945022583, "elapsed_sec": 62.32534337043762}
21
+ {"step": 3420, "epoch": 0, "loss": 3.56770658493042, "loss_en": 3.5412895679473877, "loss_zh": 3.594123601913452, "noise_std_eff": 0.016735621690750123, "elapsed_sec": 64.43766832351685}
22
+ {"step": 3440, "epoch": 0, "loss": 4.389113426208496, "loss_en": 4.421744346618652, "loss_zh": 4.35648250579834, "noise_std_eff": 0.007008370906114578, "elapsed_sec": 66.7641088962555}
23
+ {"step": 3460, "epoch": 0, "loss": 4.894651412963867, "loss_en": 4.901114463806152, "loss_zh": 4.888188362121582, "noise_std_eff": 0.005285058890779814, "elapsed_sec": 68.85138845443726}
24
+ {"step": 3480, "epoch": 0, "loss": 5.003623962402344, "loss_en": 5.0163164138793945, "loss_zh": 4.990931987762451, "noise_std_eff": 0.006692734956741334, "elapsed_sec": 71.17815899848938}
25
+ {"step": 3500, "epoch": 0, "loss": 4.480271339416504, "loss_en": 4.45639181137085, "loss_zh": 4.504150867462158, "noise_std_eff": 0.016246169805526737, "elapsed_sec": 73.27680659294128}
26
+ {"step": 3520, "epoch": 0, "loss": 4.612211227416992, "loss_en": 4.616052150726318, "loss_zh": 4.608370780944824, "noise_std_eff": 0.0034252699961264932, "elapsed_sec": 75.60608959197998}
27
+ {"step": 3540, "epoch": 0, "loss": 4.547619819641113, "loss_en": 4.5396904945373535, "loss_zh": 4.555548667907715, "noise_std_eff": 0.022219526767730712, "elapsed_sec": 77.87156581878662}
28
+ {"step": 3560, "epoch": 0, "loss": 4.3792829513549805, "loss_en": 4.3917951583862305, "loss_zh": 4.366770267486572, "noise_std_eff": 0.02048250734806061, "elapsed_sec": 80.0518889427185}
29
+ {"step": 3580, "epoch": 0, "loss": 4.453993797302246, "loss_en": 4.5005364418029785, "loss_zh": 4.407451152801514, "noise_std_eff": 0.016896864573160808, "elapsed_sec": 82.34190225601196}
30
+ {"step": 3600, "epoch": 0, "loss": 4.424088478088379, "loss_en": 4.398327350616455, "loss_zh": 4.449849605560303, "noise_std_eff": 0.017511460185050964, "elapsed_sec": 84.49793720245361}
31
+ {"step": 3620, "epoch": 0, "loss": 4.985874652862549, "loss_en": 4.980703353881836, "loss_zh": 4.991045951843262, "noise_std_eff": 0.014950167437394463, "elapsed_sec": 110.73890376091003}
32
+ {"step": 3640, "epoch": 0, "loss": 4.572558403015137, "loss_en": 4.572470188140869, "loss_zh": 4.572646617889404, "noise_std_eff": 0.014565763473510743, "elapsed_sec": 113.12388730049133}
33
+ {"step": 3660, "epoch": 0, "loss": 4.217889308929443, "loss_en": 4.226870536804199, "loss_zh": 4.2089080810546875, "noise_std_eff": 0.014285709559917452, "elapsed_sec": 115.62936806678772}
34
+ {"step": 3680, "epoch": 0, "loss": 4.101467132568359, "loss_en": 4.1401686668396, "loss_zh": 4.062765598297119, "noise_std_eff": 0.009924769351879754, "elapsed_sec": 118.07572913169861}
35
+ {"step": 3700, "epoch": 0, "loss": 4.79701042175293, "loss_en": 4.828144073486328, "loss_zh": 4.765876293182373, "noise_std_eff": 0.002655213822921117, "elapsed_sec": 120.48040318489075}
36
+ {"step": 3720, "epoch": 0, "loss": 4.421460151672363, "loss_en": 4.354667663574219, "loss_zh": 4.488253116607666, "noise_std_eff": 0.024552510380744935, "elapsed_sec": 122.99142956733704}
37
+ {"step": 3740, "epoch": 0, "loss": 4.2594499588012695, "loss_en": 4.235957622528076, "loss_zh": 4.282941818237305, "noise_std_eff": 0.010514335458477339, "elapsed_sec": 125.23352360725403}
38
+ {"step": 3760, "epoch": 0, "loss": 4.8698015213012695, "loss_en": 4.877110958099365, "loss_zh": 4.862492561340332, "noise_std_eff": 0.0073848662277062735, "elapsed_sec": 127.75867199897766}
39
+ {"step": 3780, "epoch": 0, "loss": 4.4639081954956055, "loss_en": 4.4492316246032715, "loss_zh": 4.478585243225098, "noise_std_eff": 0.014330706670880319, "elapsed_sec": 130.14009165763855}
40
+ {"step": 3800, "epoch": 0, "loss": 4.277736663818359, "loss_en": 4.267457008361816, "loss_zh": 4.2880167961120605, "noise_std_eff": 0.027696428696314494, "elapsed_sec": 132.6832880973816}
41
+ {"step": 3820, "epoch": 0, "loss": 5.3356428146362305, "loss_en": 5.325977802276611, "loss_zh": 5.34530782699585, "noise_std_eff": 0.030452363838752113, "elapsed_sec": 135.17497062683105}
42
+ {"step": 3840, "epoch": 0, "loss": 4.809118270874023, "loss_en": 4.784048557281494, "loss_zh": 4.8341875076293945, "noise_std_eff": 0.02199331879615784, "elapsed_sec": 137.71881699562073}
43
+ {"step": 3860, "epoch": 0, "loss": 4.534021377563477, "loss_en": 4.53450870513916, "loss_zh": 4.533534526824951, "noise_std_eff": 0.028197076419989267, "elapsed_sec": 140.1067352294922}
44
+ {"step": 3880, "epoch": 0, "loss": 4.401373863220215, "loss_en": 4.428153038024902, "loss_zh": 4.374594688415527, "noise_std_eff": 0.014788263936837515, "elapsed_sec": 142.62418961524963}
45
+ {"step": 3900, "epoch": 0, "loss": 4.198102951049805, "loss_en": 4.20689058303833, "loss_zh": 4.1893157958984375, "noise_std_eff": 0.01456107795238495, "elapsed_sec": 144.97760438919067}
46
+ {"step": 3920, "epoch": 0, "loss": 3.825593948364258, "loss_en": 3.832230567932129, "loss_zh": 3.818957567214966, "noise_std_eff": 0.01840376387039821, "elapsed_sec": 147.45324873924255}
47
+ {"step": 3940, "epoch": 0, "loss": 5.292659759521484, "loss_en": 5.2697248458862305, "loss_zh": 5.3155951499938965, "noise_std_eff": 0.03144733985265096, "elapsed_sec": 149.92403626441956}
48
+ {"step": 3960, "epoch": 0, "loss": 4.2374677658081055, "loss_en": 4.269098281860352, "loss_zh": 4.205837726593018, "noise_std_eff": 0.014115155935287477, "elapsed_sec": 152.39162635803223}
49
+ {"step": 3980, "epoch": 0, "loss": 4.861170768737793, "loss_en": 4.861700534820557, "loss_zh": 4.860640525817871, "noise_std_eff": 0.03759172146519025, "elapsed_sec": 154.83789777755737}
50
+ {"step": 4000, "epoch": 0, "loss": 4.267380714416504, "loss_en": 4.259169578552246, "loss_zh": 4.275591850280762, "noise_std_eff": 0.04161318143208822, "elapsed_sec": 157.26625061035156}
51
+ {"step": 4020, "epoch": 0, "loss": 4.259418487548828, "loss_en": 4.285063743591309, "loss_zh": 4.233773231506348, "noise_std_eff": 0.009841327518224717, "elapsed_sec": 183.91878962516785}
52
+ {"step": 4040, "epoch": 0, "loss": 3.990335464477539, "loss_en": 3.9937796592712402, "loss_zh": 3.986891031265259, "noise_std_eff": 0.03393629411856334, "elapsed_sec": 186.88455653190613}
53
+ {"step": 4060, "epoch": 0, "loss": 4.764108657836914, "loss_en": 4.753994464874268, "loss_zh": 4.774222373962402, "noise_std_eff": 0.0026816208412249886, "elapsed_sec": 189.75519967079163}
54
+ {"step": 4080, "epoch": 0, "loss": 4.111395835876465, "loss_en": 4.0457763671875, "loss_zh": 4.177015781402588, "noise_std_eff": 0.012574264258146287, "elapsed_sec": 192.67570519447327}
55
+ {"step": 4100, "epoch": 0, "loss": 4.337811470031738, "loss_en": 4.340029716491699, "loss_zh": 4.3355937004089355, "noise_std_eff": 0.008514103417595228, "elapsed_sec": 195.55141520500183}
56
+ {"step": 4120, "epoch": 0, "loss": 4.321137428283691, "loss_en": 4.337745189666748, "loss_zh": 4.304530143737793, "noise_std_eff": 0.03781647245089213, "elapsed_sec": 198.47979140281677}
57
+ {"step": 4140, "epoch": 0, "loss": 4.516325950622559, "loss_en": 4.512447357177734, "loss_zh": 4.520205020904541, "noise_std_eff": 0.03676105558872223, "elapsed_sec": 201.34826231002808}
58
+ {"step": 4160, "epoch": 0, "loss": 4.061432838439941, "loss_en": 4.061733245849609, "loss_zh": 4.061132907867432, "noise_std_eff": 0.03555741359790166, "elapsed_sec": 204.26460099220276}
59
+ {"step": 4180, "epoch": 0, "loss": 4.546334266662598, "loss_en": 4.5269575119018555, "loss_zh": 4.565710544586182, "noise_std_eff": 0.010968140748639901, "elapsed_sec": 207.1437110900879}
60
+ {"step": 4200, "epoch": 0, "loss": 4.357025146484375, "loss_en": 4.324619293212891, "loss_zh": 4.389430522918701, "noise_std_eff": 0.01725379973649979, "elapsed_sec": 210.09144759178162}
61
+ {"step": 4220, "epoch": 0, "loss": 4.1334228515625, "loss_en": 4.1252923011779785, "loss_zh": 4.14155387878418, "noise_std_eff": 0.020482084155082705, "elapsed_sec": 212.97537207603455}
62
+ {"step": 4240, "epoch": 0, "loss": 4.532323837280273, "loss_en": 4.501669883728027, "loss_zh": 4.562977313995361, "noise_std_eff": 0.0268917590379715, "elapsed_sec": 215.8959493637085}
63
+ {"step": 4260, "epoch": 0, "loss": 4.2880096435546875, "loss_en": 4.271827697753906, "loss_zh": 4.304191589355469, "noise_std_eff": 0.044440874457359315, "elapsed_sec": 218.76005506515503}
64
+ {"step": 4280, "epoch": 0, "loss": 4.205747604370117, "loss_en": 4.19696044921875, "loss_zh": 4.214535236358643, "noise_std_eff": 0.03889036178588867, "elapsed_sec": 221.68843722343445}
65
+ {"step": 4300, "epoch": 0, "loss": 4.4839091300964355, "loss_en": 4.4407429695129395, "loss_zh": 4.527075290679932, "noise_std_eff": 0.015397365391254426, "elapsed_sec": 224.56398916244507}
66
+ {"step": 4320, "epoch": 0, "loss": 4.274324893951416, "loss_en": 4.294376373291016, "loss_zh": 4.254273414611816, "noise_std_eff": 0.02458661049604416, "elapsed_sec": 227.49686455726624}
67
+ {"step": 4340, "epoch": 0, "loss": 4.357532501220703, "loss_en": 4.347196102142334, "loss_zh": 4.367868423461914, "noise_std_eff": 0.04169567823410034, "elapsed_sec": 230.38338828086853}
68
+ {"step": 4360, "epoch": 0, "loss": 3.972776412963867, "loss_en": 4.0236897468566895, "loss_zh": 3.921863317489624, "noise_std_eff": 0.005916262045502663, "elapsed_sec": 233.3288276195526}
69
+ {"step": 4380, "epoch": 0, "loss": 4.036138534545898, "loss_en": 4.028265953063965, "loss_zh": 4.04401159286499, "noise_std_eff": 0.021372631192207336, "elapsed_sec": 236.2194263935089}
70
+ {"step": 4400, "epoch": 0, "loss": 4.393045425415039, "loss_en": 4.387284755706787, "loss_zh": 4.398806571960449, "noise_std_eff": 0.0017392415553331376, "elapsed_sec": 239.144225358963}
71
+ {"step": 4420, "epoch": 0, "loss": 4.412775993347168, "loss_en": 4.478152275085449, "loss_zh": 4.347399711608887, "noise_std_eff": 0.0036361634731292727, "elapsed_sec": 265.8676908016205}
72
+ {"step": 4440, "epoch": 0, "loss": 4.481074810028076, "loss_en": 4.483570575714111, "loss_zh": 4.478579044342041, "noise_std_eff": 0.04918486773967743, "elapsed_sec": 268.77248883247375}
73
+ {"step": 4460, "epoch": 0, "loss": 4.209804058074951, "loss_en": 4.217731952667236, "loss_zh": 4.201876163482666, "noise_std_eff": 0.014452986419200897, "elapsed_sec": 271.64434146881104}
74
+ {"step": 4480, "epoch": 0, "loss": 3.999527931213379, "loss_en": 4.019500732421875, "loss_zh": 3.979555130004883, "noise_std_eff": 0.006855607777833939, "elapsed_sec": 274.58368468284607}
75
+ {"step": 4500, "epoch": 0, "loss": 4.770263671875, "loss_en": 4.723423004150391, "loss_zh": 4.817103862762451, "noise_std_eff": 0.013144934177398683, "elapsed_sec": 277.45921897888184}
76
+ {"step": 4520, "epoch": 0, "loss": 4.29276704788208, "loss_en": 4.309821605682373, "loss_zh": 4.275712490081787, "noise_std_eff": 0.02909006476402283, "elapsed_sec": 280.3796761035919}
77
+ {"step": 4540, "epoch": 0, "loss": 3.920625686645508, "loss_en": 3.879408359527588, "loss_zh": 3.961843252182007, "noise_std_eff": 0.023235349357128145, "elapsed_sec": 283.24396204948425}
78
+ {"step": 4560, "epoch": 0, "loss": 3.840697765350342, "loss_en": 3.8300628662109375, "loss_zh": 3.851332902908325, "noise_std_eff": 0.04582820534706116, "elapsed_sec": 286.19993567466736}
79
+ {"step": 4580, "epoch": 0, "loss": 4.094263553619385, "loss_en": 4.0542073249816895, "loss_zh": 4.13431978225708, "noise_std_eff": 0.019132551550865174, "elapsed_sec": 289.0750708580017}
80
+ {"step": 4600, "epoch": 0, "loss": 4.2900190353393555, "loss_en": 4.281344890594482, "loss_zh": 4.298693656921387, "noise_std_eff": 0.011371226608753206, "elapsed_sec": 291.97153210639954}
81
+ {"step": 4620, "epoch": 0, "loss": 4.997722148895264, "loss_en": 5.002646446228027, "loss_zh": 4.9927978515625, "noise_std_eff": 0.030644485354423524, "elapsed_sec": 294.8482162952423}
82
+ {"step": 4640, "epoch": 0, "loss": 4.062626838684082, "loss_en": 4.0019307136535645, "loss_zh": 4.1233229637146, "noise_std_eff": 0.04077326059341431, "elapsed_sec": 297.8003740310669}
83
+ {"step": 4660, "epoch": 0, "loss": 3.871702194213867, "loss_en": 3.9071738719940186, "loss_zh": 3.836230754852295, "noise_std_eff": 0.010094421356916428, "elapsed_sec": 300.65575194358826}
84
+ {"step": 4680, "epoch": 0, "loss": 3.822016477584839, "loss_en": 3.8485238552093506, "loss_zh": 3.795509099960327, "noise_std_eff": 0.03890139758586884, "elapsed_sec": 303.57220935821533}
85
+ {"step": 4700, "epoch": 0, "loss": 4.130426406860352, "loss_en": 4.158168315887451, "loss_zh": 4.102684020996094, "noise_std_eff": 0.04123497009277344, "elapsed_sec": 306.45236802101135}
86
+ {"step": 4720, "epoch": 0, "loss": 4.409292221069336, "loss_en": 4.3750386238098145, "loss_zh": 4.443545341491699, "noise_std_eff": 0.048613378405570985, "elapsed_sec": 309.3675138950348}
87
+ {"step": 4740, "epoch": 0, "loss": 4.433681488037109, "loss_en": 4.401307582855225, "loss_zh": 4.466055393218994, "noise_std_eff": 0.026228994131088257, "elapsed_sec": 312.2531225681305}
88
+ {"step": 4760, "epoch": 0, "loss": 4.187422752380371, "loss_en": 4.229672431945801, "loss_zh": 4.145172595977783, "noise_std_eff": 0.034271764755249026, "elapsed_sec": 315.1486768722534}
89
+ {"step": 4780, "epoch": 0, "loss": 4.216985702514648, "loss_en": 4.215451717376709, "loss_zh": 4.21851921081543, "noise_std_eff": 0.023592670261859894, "elapsed_sec": 318.02304315567017}
90
+ {"step": 4800, "epoch": 0, "loss": 3.9147984981536865, "loss_en": 3.9085543155670166, "loss_zh": 3.9210426807403564, "noise_std_eff": 0.0018436618149280549, "elapsed_sec": 320.96454429626465}
91
+ {"step": 4820, "epoch": 0, "loss": 3.678637742996216, "loss_en": 3.6789355278015137, "loss_zh": 3.678339958190918, "noise_std_eff": 0.047632825374603276, "elapsed_sec": 347.7240524291992}
92
+ {"step": 4840, "epoch": 0, "loss": 4.222301959991455, "loss_en": 4.160618782043457, "loss_zh": 4.283985137939453, "noise_std_eff": 4.6915950952097775e-05, "elapsed_sec": 350.6767244338989}
93
+ {"step": 4860, "epoch": 0, "loss": 3.6971216201782227, "loss_en": 3.68487811088562, "loss_zh": 3.709365129470825, "noise_std_eff": 0.009467327594757081, "elapsed_sec": 353.5645272731781}
94
+ {"step": 4880, "epoch": 0, "loss": 3.981830596923828, "loss_en": 3.988579750061035, "loss_zh": 3.975081443786621, "noise_std_eff": 0.02757089138031006, "elapsed_sec": 356.47709107398987}
95
+ {"step": 4900, "epoch": 0, "loss": 3.9918148517608643, "loss_en": 3.9355311393737793, "loss_zh": 4.048098564147949, "noise_std_eff": 0.029019078612327578, "elapsed_sec": 359.3597846031189}
96
+ {"step": 4920, "epoch": 0, "loss": 4.348459243774414, "loss_en": 4.316079139709473, "loss_zh": 4.3808393478393555, "noise_std_eff": 0.026078817248344422, "elapsed_sec": 362.2991786003113}
97
+ {"step": 4940, "epoch": 0, "loss": 4.5170183181762695, "loss_en": 4.519837856292725, "loss_zh": 4.514199256896973, "noise_std_eff": 0.02543299496173859, "elapsed_sec": 365.16747522354126}
98
+ {"step": 4960, "epoch": 0, "loss": 4.562209129333496, "loss_en": 4.561681270599365, "loss_zh": 4.562736511230469, "noise_std_eff": 0.005972002074122429, "elapsed_sec": 368.0438942909241}
99
+ {"step": 4980, "epoch": 0, "loss": 3.767737865447998, "loss_en": 3.743577480316162, "loss_zh": 3.791898250579834, "noise_std_eff": 0.018418696522712708, "elapsed_sec": 370.89555191993713}
100
+ {"step": 5000, "epoch": 0, "loss": 3.9058616161346436, "loss_en": 3.882869005203247, "loss_zh": 3.92885422706604, "noise_std_eff": 0.0077800251543521885, "elapsed_sec": 373.79946184158325}
101
+ {"step": 5020, "epoch": 0, "loss": 4.099826335906982, "loss_en": 4.115495681762695, "loss_zh": 4.0841569900512695, "noise_std_eff": 0.04255070686340332, "elapsed_sec": 376.6796259880066}
102
+ {"step": 5040, "epoch": 0, "loss": 4.261631965637207, "loss_en": 4.294739246368408, "loss_zh": 4.228525161743164, "noise_std_eff": 0.034471288323402405, "elapsed_sec": 379.6155869960785}
103
+ {"step": 5060, "epoch": 0, "loss": 3.836536169052124, "loss_en": 3.8139662742614746, "loss_zh": 3.8591060638427734, "noise_std_eff": 0.03448677659034729, "elapsed_sec": 382.50740098953247}
104
+ {"step": 5080, "epoch": 0, "loss": 3.7750589847564697, "loss_en": 3.7372312545776367, "loss_zh": 3.8128867149353027, "noise_std_eff": 0.005004080757498742, "elapsed_sec": 385.41983342170715}
105
+ {"step": 5100, "epoch": 0, "loss": 3.556515693664551, "loss_en": 3.5569844245910645, "loss_zh": 3.556046724319458, "noise_std_eff": 0.024627293646335605, "elapsed_sec": 388.2722587585449}
106
+ {"step": 5120, "epoch": 0, "loss": 4.222677707672119, "loss_en": 4.211542129516602, "loss_zh": 4.233813285827637, "noise_std_eff": 0.04487544298171997, "elapsed_sec": 391.1992988586426}
107
+ {"step": 5140, "epoch": 0, "loss": 5.2874627113342285, "loss_en": 5.248538494110107, "loss_zh": 5.32638692855835, "noise_std_eff": 0.015883372724056245, "elapsed_sec": 394.09907150268555}
108
+ {"step": 5160, "epoch": 0, "loss": 3.8419113159179688, "loss_en": 3.8147032260894775, "loss_zh": 3.869119644165039, "noise_std_eff": 0.005109912529587746, "elapsed_sec": 397.0355033874512}
109
+ {"step": 5180, "epoch": 0, "loss": 3.5395402908325195, "loss_en": 3.5148658752441406, "loss_zh": 3.5642144680023193, "noise_std_eff": 0.012736141681671143, "elapsed_sec": 399.8719172477722}
110
+ {"step": 5200, "epoch": 0, "loss": 3.797029972076416, "loss_en": 3.7594659328460693, "loss_zh": 3.8345937728881836, "noise_std_eff": 0.008543850481510162, "elapsed_sec": 402.79158067703247}
111
+ {"step": 5220, "epoch": 0, "loss": 3.742814302444458, "loss_en": 3.6923487186431885, "loss_zh": 3.7932798862457275, "noise_std_eff": 0.006269926577806473, "elapsed_sec": 429.3713300228119}
112
+ {"step": 5240, "epoch": 0, "loss": 4.422159194946289, "loss_en": 4.429410934448242, "loss_zh": 4.414907455444336, "noise_std_eff": 0.011917611956596375, "elapsed_sec": 432.2957720756531}
113
+ {"step": 5260, "epoch": 0, "loss": 4.137088775634766, "loss_en": 4.090996742248535, "loss_zh": 4.183181285858154, "noise_std_eff": 0.028017982840538025, "elapsed_sec": 435.1598265171051}
114
+ {"step": 5280, "epoch": 0, "loss": 4.099159240722656, "loss_en": 4.17031717300415, "loss_zh": 4.028000831604004, "noise_std_eff": 0.03226532340049744, "elapsed_sec": 438.0763018131256}
115
+ {"step": 5300, "epoch": 0, "loss": 4.431474685668945, "loss_en": 4.434946537017822, "loss_zh": 4.428002834320068, "noise_std_eff": 0.017531448602676393, "elapsed_sec": 440.9560775756836}
116
+ {"step": 5320, "epoch": 0, "loss": 4.108392715454102, "loss_en": 4.156833648681641, "loss_zh": 4.0599517822265625, "noise_std_eff": 0.04646025598049164, "elapsed_sec": 443.8838312625885}
117
+ {"step": 5340, "epoch": 0, "loss": 4.619003772735596, "loss_en": 4.551603317260742, "loss_zh": 4.686404228210449, "noise_std_eff": 0.041479778289794926, "elapsed_sec": 446.78100323677063}
118
+ {"step": 5360, "epoch": 0, "loss": 3.907876968383789, "loss_en": 3.862260103225708, "loss_zh": 3.953494071960449, "noise_std_eff": 0.026798468828201295, "elapsed_sec": 449.77907395362854}
119
+ {"step": 5380, "epoch": 0, "loss": 3.956868886947632, "loss_en": 4.008194923400879, "loss_zh": 3.9055428504943848, "noise_std_eff": 0.03557422757148743, "elapsed_sec": 452.76738929748535}
120
+ {"step": 5400, "epoch": 0, "loss": 4.257427215576172, "loss_en": 4.232550621032715, "loss_zh": 4.282303810119629, "noise_std_eff": 0.01875133216381073, "elapsed_sec": 455.7043776512146}
121
+ {"step": 5420, "epoch": 0, "loss": 4.43051815032959, "loss_en": 4.408164024353027, "loss_zh": 4.4528727531433105, "noise_std_eff": 0.0267765074968338, "elapsed_sec": 458.5914571285248}
122
+ {"step": 5440, "epoch": 0, "loss": 3.8976221084594727, "loss_en": 3.862091302871704, "loss_zh": 3.9331531524658203, "noise_std_eff": 0.024345313012599946, "elapsed_sec": 461.53166818618774}
123
+ {"step": 5460, "epoch": 0, "loss": 4.009732246398926, "loss_en": 3.9710769653320312, "loss_zh": 4.04838752746582, "noise_std_eff": 0.04056338667869568, "elapsed_sec": 464.41539216041565}
124
+ {"step": 5480, "epoch": 0, "loss": 3.9032974243164062, "loss_en": 3.929635524749756, "loss_zh": 3.8769590854644775, "noise_std_eff": 0.03233688175678253, "elapsed_sec": 467.3474135398865}
125
+ {"step": 5500, "epoch": 0, "loss": 4.050577163696289, "loss_en": 4.068861484527588, "loss_zh": 4.032293319702148, "noise_std_eff": 0.04165820777416229, "elapsed_sec": 470.2237551212311}
126
+ {"step": 5520, "epoch": 0, "loss": 4.283064365386963, "loss_en": 4.311123847961426, "loss_zh": 4.2550048828125, "noise_std_eff": 0.03482182621955872, "elapsed_sec": 473.1559262275696}
127
+ {"step": 5540, "epoch": 0, "loss": 4.077127456665039, "loss_en": 4.05006742477417, "loss_zh": 4.104187965393066, "noise_std_eff": 0.026557326316833496, "elapsed_sec": 476.0108571052551}
128
+ {"step": 5560, "epoch": 0, "loss": 3.9940614700317383, "loss_en": 3.9862523078918457, "loss_zh": 4.001870632171631, "noise_std_eff": 0.0037008680403232577, "elapsed_sec": 478.92328810691833}
129
+ {"step": 5580, "epoch": 0, "loss": 3.811429023742676, "loss_en": 3.9471263885498047, "loss_zh": 3.675731658935547, "noise_std_eff": 0.011912795156240464, "elapsed_sec": 481.7877597808838}
130
+ {"step": 5600, "epoch": 0, "loss": 4.193597793579102, "loss_en": 4.250016689300537, "loss_zh": 4.137179374694824, "noise_std_eff": 0.016778282821178436, "elapsed_sec": 484.70454621315}
131
+ {"step": 5620, "epoch": 0, "loss": 4.402604579925537, "loss_en": 4.380900859832764, "loss_zh": 4.4243083000183105, "noise_std_eff": 0.015143832564353943, "elapsed_sec": 511.4162654876709}
132
+ {"step": 5640, "epoch": 0, "loss": 4.7627153396606445, "loss_en": 4.78287935256958, "loss_zh": 4.742550849914551, "noise_std_eff": 0.029278215765953065, "elapsed_sec": 514.384064912796}
133
+ {"step": 5660, "epoch": 0, "loss": 4.055654525756836, "loss_en": 4.026522159576416, "loss_zh": 4.084786415100098, "noise_std_eff": 0.04027127027511597, "elapsed_sec": 517.2993221282959}
134
+ {"step": 5680, "epoch": 0, "loss": 4.0969133377075195, "loss_en": 4.094614505767822, "loss_zh": 4.099212646484375, "noise_std_eff": 0.026942387223243713, "elapsed_sec": 520.2423241138458}
135
+ {"step": 5700, "epoch": 0, "loss": 3.7386248111724854, "loss_en": 3.720862627029419, "loss_zh": 3.7563869953155518, "noise_std_eff": 0.010243351757526399, "elapsed_sec": 523.1229722499847}
136
+ {"step": 5720, "epoch": 0, "loss": 4.135359764099121, "loss_en": 4.166758060455322, "loss_zh": 4.103961944580078, "noise_std_eff": 0.018437974154949188, "elapsed_sec": 526.0597939491272}
137
+ {"step": 5740, "epoch": 0, "loss": 4.02495813369751, "loss_en": 4.0228729248046875, "loss_zh": 4.027043342590332, "noise_std_eff": 0.002311651222407818, "elapsed_sec": 528.9660036563873}
138
+ {"step": 5760, "epoch": 0, "loss": 4.082837104797363, "loss_en": 4.029473304748535, "loss_zh": 4.13620138168335, "noise_std_eff": 0.010472676157951355, "elapsed_sec": 531.8911173343658}
139
+ {"step": 5780, "epoch": 0, "loss": 4.065978050231934, "loss_en": 4.034485816955566, "loss_zh": 4.097469806671143, "noise_std_eff": 0.011796709895133973, "elapsed_sec": 534.7875986099243}
140
+ {"step": 5800, "epoch": 0, "loss": 4.42013692855835, "loss_en": 4.431214809417725, "loss_zh": 4.409059047698975, "noise_std_eff": 0.00579148679971695, "elapsed_sec": 537.7113883495331}
141
+ {"step": 5820, "epoch": 0, "loss": 4.179900169372559, "loss_en": 4.146240234375, "loss_zh": 4.213559627532959, "noise_std_eff": 0.04526429176330567, "elapsed_sec": 540.5800590515137}
142
+ {"step": 5840, "epoch": 0, "loss": 3.5114822387695312, "loss_en": 3.486912488937378, "loss_zh": 3.5360519886016846, "noise_std_eff": 0.011288046836853027, "elapsed_sec": 543.52703166008}
143
+ {"step": 5860, "epoch": 0, "loss": 4.120896339416504, "loss_en": 4.102701187133789, "loss_zh": 4.139091968536377, "noise_std_eff": 0.007959413528442384, "elapsed_sec": 546.4202287197113}
144
+ {"step": 5880, "epoch": 0, "loss": 3.966170310974121, "loss_en": 4.0076189041137695, "loss_zh": 3.9247217178344727, "noise_std_eff": 0.001066320389509201, "elapsed_sec": 549.3315687179565}
145
+ {"step": 5900, "epoch": 0, "loss": 4.2143073081970215, "loss_en": 4.20936393737793, "loss_zh": 4.219250679016113, "noise_std_eff": 0.03469061255455017, "elapsed_sec": 552.1994907855988}
146
+ {"step": 5920, "epoch": 0, "loss": 4.373056411743164, "loss_en": 4.337910175323486, "loss_zh": 4.408202648162842, "noise_std_eff": 0.003442540764808655, "elapsed_sec": 555.1275944709778}
147
+ {"step": 5940, "epoch": 0, "loss": 3.8273913860321045, "loss_en": 3.8262572288513184, "loss_zh": 3.8285255432128906, "noise_std_eff": 0.020580483973026278, "elapsed_sec": 558.0075633525848}
148
+ {"step": 5960, "epoch": 0, "loss": 4.346657752990723, "loss_en": 4.255321502685547, "loss_zh": 4.43799352645874, "noise_std_eff": 0.007645596563816071, "elapsed_sec": 560.9319887161255}
149
+ {"step": 5980, "epoch": 0, "loss": 3.8269622325897217, "loss_en": 3.8608782291412354, "loss_zh": 3.793046236038208, "noise_std_eff": 0.013611638545989992, "elapsed_sec": 563.8323810100555}
150
+ {"step": 6000, "epoch": 0, "loss": 3.9123311042785645, "loss_en": 3.8569135665893555, "loss_zh": 3.9677486419677734, "noise_std_eff": 0.004711689800024033, "elapsed_sec": 566.7608902454376}
151
+ {"step": 6020, "epoch": 0, "loss": 3.88358211517334, "loss_en": 3.8633241653442383, "loss_zh": 3.9038403034210205, "noise_std_eff": 0.04106196165084839, "elapsed_sec": 593.7396523952484}
152
+ {"step": 6040, "epoch": 0, "loss": 3.241354465484619, "loss_en": 3.2763185501098633, "loss_zh": 3.206390380859375, "noise_std_eff": 0.033592617511749266, "elapsed_sec": 596.6960282325745}
153
+ {"step": 6060, "epoch": 0, "loss": 3.9189939498901367, "loss_en": 3.8739945888519287, "loss_zh": 3.963993549346924, "noise_std_eff": 0.03192891776561737, "elapsed_sec": 599.6321921348572}
154
+ {"step": 6080, "epoch": 0, "loss": 4.010003566741943, "loss_en": 4.051098346710205, "loss_zh": 3.9689090251922607, "noise_std_eff": 0.04488615989685059, "elapsed_sec": 602.5650074481964}
155
+ {"step": 6100, "epoch": 0, "loss": 3.956653118133545, "loss_en": 4.0054731369018555, "loss_zh": 3.9078328609466553, "noise_std_eff": 0.000138250133022666, "elapsed_sec": 605.4873371124268}
156
+ {"step": 6120, "epoch": 0, "loss": 4.308806896209717, "loss_en": 4.2937493324279785, "loss_zh": 4.323864459991455, "noise_std_eff": 0.016041190922260286, "elapsed_sec": 608.4237143993378}
157
+ {"step": 6140, "epoch": 0, "loss": 3.8412961959838867, "loss_en": 3.817380666732788, "loss_zh": 3.8652114868164062, "noise_std_eff": 0.001118827611207962, "elapsed_sec": 611.3322832584381}
158
+ {"step": 6160, "epoch": 0, "loss": 3.973630428314209, "loss_en": 3.95296573638916, "loss_zh": 3.9942948818206787, "noise_std_eff": 0.04797394573688507, "elapsed_sec": 614.2683086395264}
159
+ {"step": 6180, "epoch": 0, "loss": 3.988788366317749, "loss_en": 3.9814844131469727, "loss_zh": 3.9960923194885254, "noise_std_eff": 0.038688451051712036, "elapsed_sec": 617.1923096179962}
160
+ {"step": 6200, "epoch": 0, "loss": 3.58038330078125, "loss_en": 3.584564685821533, "loss_zh": 3.5762016773223877, "noise_std_eff": 0.020053969323635103, "elapsed_sec": 620.1679027080536}
161
+ {"step": 6220, "epoch": 0, "loss": 3.756779670715332, "loss_en": 3.746522903442383, "loss_zh": 3.7670364379882812, "noise_std_eff": 0.009458570182323456, "elapsed_sec": 623.0602824687958}
162
+ {"step": 6240, "epoch": 0, "loss": 3.741820812225342, "loss_en": 3.752469062805176, "loss_zh": 3.731172800064087, "noise_std_eff": 0.009951309859752656, "elapsed_sec": 626.0165646076202}
163
+ {"step": 6260, "epoch": 0, "loss": 3.989173412322998, "loss_en": 3.9792704582214355, "loss_zh": 3.9990761280059814, "noise_std_eff": 0.045049652457237244, "elapsed_sec": 628.928346157074}
164
+ {"step": 6280, "epoch": 0, "loss": 4.271501064300537, "loss_en": 4.252818584442139, "loss_zh": 4.2901835441589355, "noise_std_eff": 0.028634029626846316, "elapsed_sec": 631.8807361125946}
165
+ {"step": 6300, "epoch": 0, "loss": 3.808932304382324, "loss_en": 3.8690474033355713, "loss_zh": 3.7488174438476562, "noise_std_eff": 0.040100824832916264, "elapsed_sec": 634.8155946731567}
166
+ {"step": 6320, "epoch": 0, "loss": 3.656662940979004, "loss_en": 3.6691603660583496, "loss_zh": 3.644165515899658, "noise_std_eff": 0.016835784912109377, "elapsed_sec": 637.7848131656647}
167
+ {"step": 6340, "epoch": 0, "loss": 3.3739137649536133, "loss_en": 3.311976671218872, "loss_zh": 3.4358508586883545, "noise_std_eff": 0.011707479506731034, "elapsed_sec": 640.6964359283447}
168
+ {"step": 6360, "epoch": 0, "loss": 4.008577346801758, "loss_en": 4.014779567718506, "loss_zh": 4.002375602722168, "noise_std_eff": 0.013113583624362947, "elapsed_sec": 643.6714918613434}
169
+ {"step": 6380, "epoch": 0, "loss": 3.6951074600219727, "loss_en": 3.6191821098327637, "loss_zh": 3.7710328102111816, "noise_std_eff": 0.02405712902545929, "elapsed_sec": 646.5999128818512}
170
+ {"step": 6400, "epoch": 0, "loss": 4.005957126617432, "loss_en": 4.008077144622803, "loss_zh": 4.0038371086120605, "noise_std_eff": 0.03915860354900361, "elapsed_sec": 649.564596414566}
171
+ {"step": 6420, "epoch": 0, "loss": 4.104935646057129, "loss_en": 4.077354907989502, "loss_zh": 4.132516384124756, "noise_std_eff": 0.027941069006919863, "elapsed_sec": 676.620932340622}
172
+ {"step": 6440, "epoch": 0, "loss": 3.957949638366699, "loss_en": 3.9640181064605713, "loss_zh": 3.951880931854248, "noise_std_eff": 0.0040153563022613525, "elapsed_sec": 679.6208806037903}
173
+ {"step": 6460, "epoch": 0, "loss": 3.996973991394043, "loss_en": 4.021021842956543, "loss_zh": 3.972926378250122, "noise_std_eff": 0.03551915585994721, "elapsed_sec": 682.517431974411}
174
+ {"step": 6480, "epoch": 0, "loss": 4.0193257331848145, "loss_en": 3.9695475101470947, "loss_zh": 4.069104194641113, "noise_std_eff": 0.008283404260873794, "elapsed_sec": 685.4653260707855}
175
+ {"step": 6500, "epoch": 0, "loss": 3.523172378540039, "loss_en": 3.539706230163574, "loss_zh": 3.506638765335083, "noise_std_eff": 0.03808864951133728, "elapsed_sec": 688.3403108119965}
176
+ {"step": 6520, "epoch": 0, "loss": 3.7153067588806152, "loss_en": 3.742598295211792, "loss_zh": 3.6880149841308594, "noise_std_eff": 0.04781303703784943, "elapsed_sec": 691.2579007148743}
177
+ {"step": 6540, "epoch": 0, "loss": 3.7088470458984375, "loss_en": 3.6712005138397217, "loss_zh": 3.746493339538574, "noise_std_eff": 0.001677979715168476, "elapsed_sec": 694.1414837837219}
178
+ {"step": 6560, "epoch": 0, "loss": 4.513547897338867, "loss_en": 4.466270446777344, "loss_zh": 4.560824871063232, "noise_std_eff": 0.04981692135334015, "elapsed_sec": 697.082489490509}
179
+ {"step": 6580, "epoch": 0, "loss": 3.7147979736328125, "loss_en": 3.7330069541931152, "loss_zh": 3.6965887546539307, "noise_std_eff": 0.02579861581325531, "elapsed_sec": 699.9464995861053}
180
+ {"step": 6600, "epoch": 0, "loss": 3.99265193939209, "loss_en": 3.9742448329925537, "loss_zh": 4.011058807373047, "noise_std_eff": 0.03419179320335388, "elapsed_sec": 702.883287191391}
181
+ {"step": 6620, "epoch": 0, "loss": 3.78513240814209, "loss_en": 3.7659389972686768, "loss_zh": 3.804325580596924, "noise_std_eff": 0.04741639792919159, "elapsed_sec": 705.8296732902527}
182
+ {"step": 6640, "epoch": 0, "loss": 3.778742790222168, "loss_en": 3.744555950164795, "loss_zh": 3.812929391860962, "noise_std_eff": 0.04141362011432648, "elapsed_sec": 708.725879907608}
183
+ {"step": 6660, "epoch": 0, "loss": 4.122109413146973, "loss_en": 4.118069648742676, "loss_zh": 4.126149654388428, "noise_std_eff": 0.009326586872339249, "elapsed_sec": 711.7428138256073}
184
+ {"step": 6680, "epoch": 0, "loss": 3.3787448406219482, "loss_en": 3.4031717777252197, "loss_zh": 3.3543179035186768, "noise_std_eff": 0.01838470995426178, "elapsed_sec": 714.7099833488464}
185
+ {"step": 6700, "epoch": 0, "loss": 3.3306376934051514, "loss_en": 3.333961009979248, "loss_zh": 3.3273143768310547, "noise_std_eff": 0.044530007243156436, "elapsed_sec": 717.6901774406433}
186
+ {"step": 6720, "epoch": 0, "loss": 3.925285816192627, "loss_en": 3.929147720336914, "loss_zh": 3.9214236736297607, "noise_std_eff": 0.008571862429380418, "elapsed_sec": 720.6089084148407}
187
+ {"step": 6740, "epoch": 0, "loss": 3.882829189300537, "loss_en": 3.918975830078125, "loss_zh": 3.846682548522949, "noise_std_eff": 0.020411913096904755, "elapsed_sec": 723.5200703144073}
188
+ {"step": 6760, "epoch": 0, "loss": 4.031590461730957, "loss_en": 4.036078453063965, "loss_zh": 4.027102947235107, "noise_std_eff": 0.028190630674362185, "elapsed_sec": 726.4431707859039}
189
+ {"step": 6780, "epoch": 0, "loss": 3.759880542755127, "loss_en": 3.7490060329437256, "loss_zh": 3.770754814147949, "noise_std_eff": 0.013142268359661103, "elapsed_sec": 729.3932058811188}
190
+ {"step": 6800, "epoch": 0, "loss": 3.740797996520996, "loss_en": 3.7455663681030273, "loss_zh": 3.7360293865203857, "noise_std_eff": 0.021390584111213685, "elapsed_sec": 732.4666578769684}
191
+ {"step": 6820, "epoch": 0, "loss": 4.23007869720459, "loss_en": 4.259039878845215, "loss_zh": 4.201117515563965, "noise_std_eff": 0.02403313219547272, "elapsed_sec": 759.4765675067902}
192
+ {"step": 6840, "epoch": 0, "loss": 4.240195274353027, "loss_en": 4.3503923416137695, "loss_zh": 4.129998207092285, "noise_std_eff": 0.018801775574684144, "elapsed_sec": 762.4231116771698}
193
+ {"step": 6860, "epoch": 0, "loss": 3.738898277282715, "loss_en": 3.7126882076263428, "loss_zh": 3.765108346939087, "noise_std_eff": 0.04540146589279175, "elapsed_sec": 765.3796243667603}
194
+ {"step": 6880, "epoch": 0, "loss": 3.860219955444336, "loss_en": 3.8288049697875977, "loss_zh": 3.891634941101074, "noise_std_eff": 0.01939539760351181, "elapsed_sec": 768.3434629440308}
195
+ {"step": 6900, "epoch": 0, "loss": 4.010709762573242, "loss_en": 4.007661819458008, "loss_zh": 4.013758182525635, "noise_std_eff": 0.017452868819236755, "elapsed_sec": 771.2510826587677}
196
+ {"step": 6920, "epoch": 0, "loss": 3.448009490966797, "loss_en": 3.4104197025299072, "loss_zh": 3.4855990409851074, "noise_std_eff": 0.005819241330027581, "elapsed_sec": 774.2196555137634}
197
+ {"step": 6940, "epoch": 0, "loss": 3.74509859085083, "loss_en": 3.777174472808838, "loss_zh": 3.7130229473114014, "noise_std_eff": 0.020782294869422915, "elapsed_sec": 777.1427373886108}
198
+ {"step": 6960, "epoch": 0, "loss": 3.7812232971191406, "loss_en": 3.806692361831665, "loss_zh": 3.755753993988037, "noise_std_eff": 0.028284960985183717, "elapsed_sec": 780.0911428928375}
199
+ {"step": 6980, "epoch": 0, "loss": 3.485426425933838, "loss_en": 3.4795260429382324, "loss_zh": 3.4913268089294434, "noise_std_eff": 0.016542184352874755, "elapsed_sec": 783.0258541107178}
200
+ {"step": 7000, "epoch": 0, "loss": 3.9790332317352295, "loss_en": 3.938354015350342, "loss_zh": 4.019712448120117, "noise_std_eff": 0.004634381458163262, "elapsed_sec": 785.9640884399414}
201
+ {"step": 7020, "epoch": 0, "loss": 4.003347396850586, "loss_en": 4.031236171722412, "loss_zh": 3.9754586219787598, "noise_std_eff": 0.005290938541293144, "elapsed_sec": 788.8778579235077}
202
+ {"step": 7040, "epoch": 0, "loss": 4.035655498504639, "loss_en": 4.051894664764404, "loss_zh": 4.019416332244873, "noise_std_eff": 0.02673775851726532, "elapsed_sec": 791.8455286026001}
203
+ {"step": 7060, "epoch": 0, "loss": 3.8161306381225586, "loss_en": 3.8011105060577393, "loss_zh": 3.831150770187378, "noise_std_eff": 0.015489403903484345, "elapsed_sec": 794.7453577518463}
204
+ {"step": 7080, "epoch": 0, "loss": 4.313494682312012, "loss_en": 4.308594226837158, "loss_zh": 4.318395137786865, "noise_std_eff": 0.030091676115989688, "elapsed_sec": 797.691596031189}
205
+ {"step": 7100, "epoch": 0, "loss": 3.9991869926452637, "loss_en": 3.986801862716675, "loss_zh": 4.011571884155273, "noise_std_eff": 0.003237058222293854, "elapsed_sec": 800.6161167621613}
206
+ {"step": 7120, "epoch": 0, "loss": 3.992353916168213, "loss_en": 3.931715965270996, "loss_zh": 4.05299186706543, "noise_std_eff": 0.0427046686410904, "elapsed_sec": 803.5589323043823}
207
+ {"step": 7140, "epoch": 0, "loss": 4.395349502563477, "loss_en": 4.427481174468994, "loss_zh": 4.363217830657959, "noise_std_eff": 0.03557673692703247, "elapsed_sec": 806.4668641090393}
208
+ {"step": 7160, "epoch": 0, "loss": 4.030540466308594, "loss_en": 4.034021377563477, "loss_zh": 4.027060031890869, "noise_std_eff": 0.026088935136795045, "elapsed_sec": 809.4075493812561}
209
+ {"step": 7180, "epoch": 0, "loss": 3.542818784713745, "loss_en": 3.566392183303833, "loss_zh": 3.5192453861236572, "noise_std_eff": 0.04464989602565766, "elapsed_sec": 812.3050463199615}
210
+ {"step": 7200, "epoch": 0, "loss": 3.7140378952026367, "loss_en": 3.742389678955078, "loss_zh": 3.6856863498687744, "noise_std_eff": 0.027711448073387147, "elapsed_sec": 815.2315940856934}
211
+ {"step": 7220, "epoch": 0, "loss": 3.887662410736084, "loss_en": 3.879380464553833, "loss_zh": 3.895944595336914, "noise_std_eff": 0.01027733236551285, "elapsed_sec": 841.8573112487793}
212
+ {"step": 7240, "epoch": 0, "loss": 3.723039150238037, "loss_en": 3.6965999603271484, "loss_zh": 3.749478578567505, "noise_std_eff": 0.031850627064704894, "elapsed_sec": 844.5393187999725}
213
+ {"step": 7260, "epoch": 0, "loss": 3.4822380542755127, "loss_en": 3.491230010986328, "loss_zh": 3.4732460975646973, "noise_std_eff": 0.007239495217800141, "elapsed_sec": 847.3717877864838}
214
+ {"step": 7280, "epoch": 0, "loss": 3.534421920776367, "loss_en": 3.532118558883667, "loss_zh": 3.5367252826690674, "noise_std_eff": 0.03818592429161072, "elapsed_sec": 850.3326051235199}
215
+ {"step": 7300, "epoch": 0, "loss": 3.470635414123535, "loss_en": 3.476064682006836, "loss_zh": 3.4652063846588135, "noise_std_eff": 0.04534485936164856, "elapsed_sec": 853.2896676063538}
216
+ {"step": 7320, "epoch": 0, "loss": 3.6555886268615723, "loss_en": 3.6437036991119385, "loss_zh": 3.667473793029785, "noise_std_eff": 0.005062760040163994, "elapsed_sec": 856.2538068294525}
217
+ {"step": 7340, "epoch": 0, "loss": 4.4881720542907715, "loss_en": 4.5027008056640625, "loss_zh": 4.4736433029174805, "noise_std_eff": 0.03338879942893982, "elapsed_sec": 859.2087025642395}
218
+ {"step": 7360, "epoch": 0, "loss": 3.5608091354370117, "loss_en": 3.55218243598938, "loss_zh": 3.5694360733032227, "noise_std_eff": 0.03392603695392609, "elapsed_sec": 862.1531319618225}
219
+ {"step": 7380, "epoch": 0, "loss": 3.7765989303588867, "loss_en": 3.793030261993408, "loss_zh": 3.7601675987243652, "noise_std_eff": 0.011719396710395813, "elapsed_sec": 865.113062620163}
220
+ {"step": 7400, "epoch": 0, "loss": 3.3990225791931152, "loss_en": 3.4266276359558105, "loss_zh": 3.37141752243042, "noise_std_eff": 0.007443138957023621, "elapsed_sec": 868.036158323288}
221
+ {"step": 7420, "epoch": 0, "loss": 3.815894842147827, "loss_en": 3.825815439224243, "loss_zh": 3.805974245071411, "noise_std_eff": 0.017661020159721375, "elapsed_sec": 870.9933044910431}
222
+ {"step": 7440, "epoch": 0, "loss": 3.7771968841552734, "loss_en": 3.7921507358551025, "loss_zh": 3.7622432708740234, "noise_std_eff": 0.011693891882896424, "elapsed_sec": 873.9575803279877}
223
+ {"step": 7460, "epoch": 0, "loss": 3.579653263092041, "loss_en": 3.6056082248687744, "loss_zh": 3.5536980628967285, "noise_std_eff": 0.010788413137197495, "elapsed_sec": 876.8943531513214}
224
+ {"step": 7480, "epoch": 0, "loss": 4.13328742980957, "loss_en": 4.134308815002441, "loss_zh": 4.132265567779541, "noise_std_eff": 0.010074599087238312, "elapsed_sec": 879.841222524643}
225
+ {"step": 7500, "epoch": 1, "loss": 5.135199546813965, "loss_en": 5.1660685539245605, "loss_zh": 5.104330062866211, "noise_std_eff": 0.014417831599712373, "elapsed_sec": 882.865864276886}
226
+ {"step": 7520, "epoch": 1, "loss": 4.883364677429199, "loss_en": 4.837099552154541, "loss_zh": 4.929629325866699, "noise_std_eff": 0.0037095423787832263, "elapsed_sec": 885.9220151901245}
227
+ {"step": 7540, "epoch": 1, "loss": 4.915366172790527, "loss_en": 4.906091690063477, "loss_zh": 4.924641132354736, "noise_std_eff": 0.02082539498806, "elapsed_sec": 888.9383053779602}
228
+ {"step": 7560, "epoch": 1, "loss": 4.924857139587402, "loss_en": 5.051235198974609, "loss_zh": 4.7984795570373535, "noise_std_eff": 0.04262905418872834, "elapsed_sec": 891.9712953567505}
229
+ {"step": 7580, "epoch": 1, "loss": 5.036630630493164, "loss_en": 5.084940433502197, "loss_zh": 4.988320350646973, "noise_std_eff": 0.02065652906894684, "elapsed_sec": 894.9863085746765}
230
+ {"step": 7600, "epoch": 1, "loss": 5.576292991638184, "loss_en": 5.7405924797058105, "loss_zh": 5.411993026733398, "noise_std_eff": 0.021828652918338777, "elapsed_sec": 898.0584745407104}
231
+ {"step": 7620, "epoch": 1, "loss": 4.100864410400391, "loss_en": 4.084290981292725, "loss_zh": 4.117437362670898, "noise_std_eff": 0.0008811875246465206, "elapsed_sec": 925.0611398220062}
232
+ {"step": 7640, "epoch": 1, "loss": 3.641568183898926, "loss_en": 3.6244759559631348, "loss_zh": 3.6586601734161377, "noise_std_eff": 0.027755507826805116, "elapsed_sec": 928.0240740776062}
233
+ {"step": 7660, "epoch": 1, "loss": 4.243793487548828, "loss_en": 4.246615409851074, "loss_zh": 4.24097204208374, "noise_std_eff": 0.008135295659303666, "elapsed_sec": 930.907876253128}
234
+ {"step": 7680, "epoch": 1, "loss": 3.5692691802978516, "loss_en": 3.5380852222442627, "loss_zh": 3.6004528999328613, "noise_std_eff": 0.03398455083370209, "elapsed_sec": 933.8532841205597}
235
+ {"step": 7700, "epoch": 1, "loss": 3.273013114929199, "loss_en": 3.2930123805999756, "loss_zh": 3.253013849258423, "noise_std_eff": 0.021998998522758485, "elapsed_sec": 936.7404363155365}
236
+ {"step": 7720, "epoch": 1, "loss": 3.583723545074463, "loss_en": 3.562610387802124, "loss_zh": 3.6048367023468018, "noise_std_eff": 0.02417653501033783, "elapsed_sec": 939.6920976638794}
237
+ {"step": 7740, "epoch": 1, "loss": 3.9012255668640137, "loss_en": 3.983731269836426, "loss_zh": 3.8187201023101807, "noise_std_eff": 0.007552389800548554, "elapsed_sec": 942.5962862968445}
238
+ {"step": 7760, "epoch": 1, "loss": 3.8050808906555176, "loss_en": 3.7952141761779785, "loss_zh": 3.8149473667144775, "noise_std_eff": 0.012762583792209625, "elapsed_sec": 945.5364458560944}
239
+ {"step": 7780, "epoch": 1, "loss": 3.729802370071411, "loss_en": 3.702371835708618, "loss_zh": 3.757232904434204, "noise_std_eff": 0.026765289902687076, "elapsed_sec": 948.42458319664}
240
+ {"step": 7800, "epoch": 1, "loss": 4.474738597869873, "loss_en": 4.504068374633789, "loss_zh": 4.445408821105957, "noise_std_eff": 0.042252713441848756, "elapsed_sec": 951.3678824901581}
241
+ {"step": 7820, "epoch": 1, "loss": 3.368265390396118, "loss_en": 3.375924825668335, "loss_zh": 3.3606059551239014, "noise_std_eff": 0.0417623907327652, "elapsed_sec": 954.2918920516968}
242
+ {"step": 7840, "epoch": 1, "loss": 4.029252052307129, "loss_en": 4.043735027313232, "loss_zh": 4.014768600463867, "noise_std_eff": 0.03639289140701294, "elapsed_sec": 957.2372779846191}
243
+ {"step": 7860, "epoch": 1, "loss": 3.47847318649292, "loss_en": 3.487011432647705, "loss_zh": 3.4699347019195557, "noise_std_eff": 0.04232113063335419, "elapsed_sec": 960.1487655639648}
244
+ {"step": 7880, "epoch": 1, "loss": 3.891781806945801, "loss_en": 3.8911333084106445, "loss_zh": 3.892430067062378, "noise_std_eff": 0.008039617538452148, "elapsed_sec": 963.0885453224182}
245
+ {"step": 7900, "epoch": 1, "loss": 3.836940050125122, "loss_en": 3.8043909072875977, "loss_zh": 3.8694891929626465, "noise_std_eff": 0.021854472160339356, "elapsed_sec": 965.9956803321838}
246
+ {"step": 7920, "epoch": 1, "loss": 3.6983227729797363, "loss_en": 3.653210163116455, "loss_zh": 3.7434353828430176, "noise_std_eff": 0.02104296386241913, "elapsed_sec": 968.9403486251831}
247
+ {"step": 7940, "epoch": 1, "loss": 3.664337396621704, "loss_en": 3.6167051792144775, "loss_zh": 3.7119696140289307, "noise_std_eff": 0.006981739401817322, "elapsed_sec": 971.8644499778748}
248
+ {"step": 7960, "epoch": 1, "loss": 3.984259605407715, "loss_en": 4.026962757110596, "loss_zh": 3.941556215286255, "noise_std_eff": 0.02277447283267975, "elapsed_sec": 974.8123891353607}
249
+ {"step": 7980, "epoch": 1, "loss": 3.7811012268066406, "loss_en": 3.7913248538970947, "loss_zh": 3.7708773612976074, "noise_std_eff": 0.02715825140476227, "elapsed_sec": 977.7361788749695}
250
+ {"step": 8000, "epoch": 1, "loss": 3.699310779571533, "loss_en": 3.732560873031616, "loss_zh": 3.66606068611145, "noise_std_eff": 0.015409141778945923, "elapsed_sec": 980.7246656417847}
251
+ {"step": 8020, "epoch": 1, "loss": 3.4879918098449707, "loss_en": 3.501492977142334, "loss_zh": 3.4744906425476074, "noise_std_eff": 0.01795426905155182, "elapsed_sec": 1007.760614156723}
252
+ {"step": 8040, "epoch": 1, "loss": 3.5645461082458496, "loss_en": 3.5454318523406982, "loss_zh": 3.583660364151001, "noise_std_eff": 0.008861547708511353, "elapsed_sec": 1010.7422707080841}
253
+ {"step": 8060, "epoch": 1, "loss": 4.417862892150879, "loss_en": 4.373734951019287, "loss_zh": 4.461990833282471, "noise_std_eff": 0.03611513078212738, "elapsed_sec": 1013.6395576000214}
254
+ {"step": 8080, "epoch": 1, "loss": 4.1204514503479, "loss_en": 4.14041805267334, "loss_zh": 4.100484848022461, "noise_std_eff": 0.0013213700614869595, "elapsed_sec": 1016.659063577652}
255
+ {"step": 8100, "epoch": 1, "loss": 3.651078939437866, "loss_en": 3.624309539794922, "loss_zh": 3.6778483390808105, "noise_std_eff": 0.009402955323457718, "elapsed_sec": 1019.572781085968}
256
+ {"step": 8120, "epoch": 1, "loss": 3.703688859939575, "loss_en": 3.6685049533843994, "loss_zh": 3.738872766494751, "noise_std_eff": 0.044343435764312746, "elapsed_sec": 1022.5481479167938}
257
+ {"step": 8140, "epoch": 1, "loss": 3.5205130577087402, "loss_en": 3.497479200363159, "loss_zh": 3.5435469150543213, "noise_std_eff": 0.031256559491157535, "elapsed_sec": 1025.4451534748077}
258
+ {"step": 8160, "epoch": 1, "loss": 3.6006200313568115, "loss_en": 3.6179697513580322, "loss_zh": 3.583270311355591, "noise_std_eff": 0.024594978988170625, "elapsed_sec": 1028.388057231903}
259
+ {"step": 8180, "epoch": 1, "loss": 3.7307474613189697, "loss_en": 3.773582696914673, "loss_zh": 3.6879122257232666, "noise_std_eff": 0.011475202441215516, "elapsed_sec": 1031.3197038173676}
260
+ {"step": 8200, "epoch": 1, "loss": 3.971768379211426, "loss_en": 3.96054744720459, "loss_zh": 3.9829890727996826, "noise_std_eff": 0.02403607964515686, "elapsed_sec": 1034.2812807559967}
261
+ {"step": 8220, "epoch": 1, "loss": 3.601177215576172, "loss_en": 3.5660126209259033, "loss_zh": 3.6363415718078613, "noise_std_eff": 0.020324657857418063, "elapsed_sec": 1037.2024366855621}
262
+ {"step": 8240, "epoch": 1, "loss": 3.698112726211548, "loss_en": 3.7172632217407227, "loss_zh": 3.678962230682373, "noise_std_eff": 0.04413625299930573, "elapsed_sec": 1040.1651446819305}
263
+ {"step": 8260, "epoch": 1, "loss": 3.6533141136169434, "loss_en": 3.6725332736968994, "loss_zh": 3.634094715118408, "noise_std_eff": 0.007615183293819428, "elapsed_sec": 1043.122573852539}
264
+ {"step": 8280, "epoch": 1, "loss": 3.355541467666626, "loss_en": 3.3674423694610596, "loss_zh": 3.3436405658721924, "noise_std_eff": 0.015793979167938232, "elapsed_sec": 1046.0918197631836}
265
+ {"step": 8300, "epoch": 1, "loss": 4.008236408233643, "loss_en": 3.9826443195343018, "loss_zh": 4.033828258514404, "noise_std_eff": 0.03494682610034943, "elapsed_sec": 1049.0226764678955}
266
+ {"step": 8320, "epoch": 1, "loss": 3.5986547470092773, "loss_en": 3.562638998031616, "loss_zh": 3.6346707344055176, "noise_std_eff": 0.0447754979133606, "elapsed_sec": 1052.0542001724243}
267
+ {"step": 8340, "epoch": 1, "loss": 3.6956701278686523, "loss_en": 3.690675973892212, "loss_zh": 3.7006642818450928, "noise_std_eff": 0.03175995349884033, "elapsed_sec": 1054.923740386963}
268
+ {"step": 8360, "epoch": 1, "loss": 3.259152889251709, "loss_en": 3.2727015018463135, "loss_zh": 3.2456045150756836, "noise_std_eff": 0.01511847972869873, "elapsed_sec": 1057.8697583675385}
269
+ {"step": 8380, "epoch": 1, "loss": 3.6323251724243164, "loss_en": 3.6309735774993896, "loss_zh": 3.633676767349243, "noise_std_eff": 0.007022546231746674, "elapsed_sec": 1060.7718901634216}
270
+ {"step": 8400, "epoch": 1, "loss": 3.7355270385742188, "loss_en": 3.6756603717803955, "loss_zh": 3.795393705368042, "noise_std_eff": 0.041516417264938356, "elapsed_sec": 1063.7168250083923}
271
+ {"step": 8420, "epoch": 1, "loss": 3.648087978363037, "loss_en": 3.6209018230438232, "loss_zh": 3.675274133682251, "noise_std_eff": 0.03269751965999603, "elapsed_sec": 1090.5054533481598}
272
+ {"step": 8440, "epoch": 1, "loss": 3.2718639373779297, "loss_en": 3.2471539974212646, "loss_zh": 3.2965738773345947, "noise_std_eff": 0.00042952634394168856, "elapsed_sec": 1093.452564716339}
273
+ {"step": 8460, "epoch": 1, "loss": 3.0616774559020996, "loss_en": 3.076586961746216, "loss_zh": 3.0467681884765625, "noise_std_eff": 0.004791062697768212, "elapsed_sec": 1096.3128473758698}
274
+ {"step": 8480, "epoch": 1, "loss": 3.595426082611084, "loss_en": 3.613029956817627, "loss_zh": 3.577821969985962, "noise_std_eff": 0.01647102236747742, "elapsed_sec": 1099.2475543022156}
275
+ {"step": 8500, "epoch": 1, "loss": 3.419044017791748, "loss_en": 3.4492785930633545, "loss_zh": 3.3888092041015625, "noise_std_eff": 0.023859368264675142, "elapsed_sec": 1102.2149300575256}
276
+ {"step": 8520, "epoch": 1, "loss": 4.138917922973633, "loss_en": 4.148807525634766, "loss_zh": 4.1290283203125, "noise_std_eff": 0.01033836528658867, "elapsed_sec": 1105.1929168701172}
277
+ {"step": 8540, "epoch": 1, "loss": 3.577939987182617, "loss_en": 3.5786314010620117, "loss_zh": 3.5772488117218018, "noise_std_eff": 0.024605084955692292, "elapsed_sec": 1108.131962299347}
278
+ {"step": 8560, "epoch": 1, "loss": 3.985077381134033, "loss_en": 4.0236406326293945, "loss_zh": 3.946514129638672, "noise_std_eff": 0.01663366109132767, "elapsed_sec": 1111.1396205425262}
279
+ {"step": 8580, "epoch": 1, "loss": 3.953045129776001, "loss_en": 3.9389333724975586, "loss_zh": 3.9671568870544434, "noise_std_eff": 0.018054112792015076, "elapsed_sec": 1114.1112270355225}
280
+ {"step": 8600, "epoch": 1, "loss": 3.5460901260375977, "loss_en": 3.5710465908050537, "loss_zh": 3.5211336612701416, "noise_std_eff": 0.03936219811439515, "elapsed_sec": 1117.1202738285065}
281
+ {"step": 8620, "epoch": 1, "loss": 3.9544410705566406, "loss_en": 3.9303243160247803, "loss_zh": 3.978557825088501, "noise_std_eff": 0.009929468482732774, "elapsed_sec": 1120.215829372406}
282
+ {"step": 8640, "epoch": 1, "loss": 3.5831072330474854, "loss_en": 3.600017786026001, "loss_zh": 3.5661966800689697, "noise_std_eff": 0.02010280787944794, "elapsed_sec": 1123.1996846199036}
283
+ {"step": 8660, "epoch": 1, "loss": 3.4538164138793945, "loss_en": 3.4364593029022217, "loss_zh": 3.4711735248565674, "noise_std_eff": 0.026942625641822815, "elapsed_sec": 1126.0841085910797}
284
+ {"step": 8680, "epoch": 1, "loss": 3.434324264526367, "loss_en": 3.420112133026123, "loss_zh": 3.4485361576080322, "noise_std_eff": 0.04996708333492279, "elapsed_sec": 1129.0160405635834}
285
+ {"step": 8700, "epoch": 1, "loss": 2.830437660217285, "loss_en": 2.8394100666046143, "loss_zh": 2.821465492248535, "noise_std_eff": 0.047151246666908266, "elapsed_sec": 1131.924852848053}
286
+ {"step": 8720, "epoch": 1, "loss": 3.655069351196289, "loss_en": 3.640246629714966, "loss_zh": 3.6698923110961914, "noise_std_eff": 0.0008839353919029236, "elapsed_sec": 1134.8641147613525}
287
+ {"step": 8740, "epoch": 1, "loss": 3.625901699066162, "loss_en": 3.5771870613098145, "loss_zh": 3.674616575241089, "noise_std_eff": 0.00693020224571228, "elapsed_sec": 1137.7808468341827}
288
+ {"step": 8760, "epoch": 1, "loss": 3.968712329864502, "loss_en": 3.940535068511963, "loss_zh": 3.99688982963562, "noise_std_eff": 0.002604796551167965, "elapsed_sec": 1140.7279362678528}
289
+ {"step": 8780, "epoch": 1, "loss": 3.8590457439422607, "loss_en": 3.890671491622925, "loss_zh": 3.8274199962615967, "noise_std_eff": 0.02092678397893906, "elapsed_sec": 1143.6475088596344}
290
+ {"step": 8800, "epoch": 1, "loss": 3.6596364974975586, "loss_en": 3.631901264190674, "loss_zh": 3.6873714923858643, "noise_std_eff": 0.02977631688117981, "elapsed_sec": 1146.599483013153}
291
+ {"step": 8820, "epoch": 1, "loss": 3.1942977905273438, "loss_en": 3.1484930515289307, "loss_zh": 3.240102767944336, "noise_std_eff": 0.03950628638267517, "elapsed_sec": 1173.6565301418304}
292
+ {"step": 8840, "epoch": 1, "loss": 3.9980762004852295, "loss_en": 4.027106761932373, "loss_zh": 3.969045639038086, "noise_std_eff": 0.006246259063482285, "elapsed_sec": 1176.588562965393}
293
+ {"step": 8860, "epoch": 1, "loss": 3.4671812057495117, "loss_en": 3.4899537563323975, "loss_zh": 3.444408893585205, "noise_std_eff": 0.048634308576583865, "elapsed_sec": 1179.5157916545868}
294
+ {"step": 8880, "epoch": 1, "loss": 3.991605520248413, "loss_en": 3.9873011112213135, "loss_zh": 3.9959099292755127, "noise_std_eff": 0.01436508446931839, "elapsed_sec": 1182.4714438915253}
295
+ {"step": 8900, "epoch": 1, "loss": 3.5540146827697754, "loss_en": 3.5577385425567627, "loss_zh": 3.550290822982788, "noise_std_eff": 0.019667986035346988, "elapsed_sec": 1185.3884961605072}
296
+ {"step": 8920, "epoch": 1, "loss": 4.484706878662109, "loss_en": 4.4934983253479, "loss_zh": 4.475915431976318, "noise_std_eff": 0.046718171238899237, "elapsed_sec": 1188.3238167762756}
297
+ {"step": 8940, "epoch": 1, "loss": 3.7526395320892334, "loss_en": 3.7706029415130615, "loss_zh": 3.7346761226654053, "noise_std_eff": 0.02770509421825409, "elapsed_sec": 1191.1874117851257}
298
+ {"step": 8960, "epoch": 1, "loss": 3.711737632751465, "loss_en": 3.6980414390563965, "loss_zh": 3.725433588027954, "noise_std_eff": 0.022089605033397675, "elapsed_sec": 1194.103131532669}
299
+ {"step": 8980, "epoch": 1, "loss": 3.7886924743652344, "loss_en": 3.7402236461639404, "loss_zh": 3.8371615409851074, "noise_std_eff": 0.008841568976640702, "elapsed_sec": 1197.03093957901}
300
+ {"step": 9000, "epoch": 1, "loss": 3.6279263496398926, "loss_en": 3.6045455932617188, "loss_zh": 3.6513073444366455, "noise_std_eff": 0.029611095786094666, "elapsed_sec": 1199.9243133068085}
301
+ {"step": 9020, "epoch": 1, "loss": 3.5495786666870117, "loss_en": 3.516599178314209, "loss_zh": 3.5825579166412354, "noise_std_eff": 0.02188233733177185, "elapsed_sec": 1202.804907798767}
302
+ {"step": 9040, "epoch": 1, "loss": 3.345776081085205, "loss_en": 3.3633759021759033, "loss_zh": 3.3281760215759277, "noise_std_eff": 0.0389743059873581, "elapsed_sec": 1205.7237157821655}
303
+ {"step": 9060, "epoch": 1, "loss": 3.568939447402954, "loss_en": 3.5642354488372803, "loss_zh": 3.573643445968628, "noise_std_eff": 0.02349703013896942, "elapsed_sec": 1208.6007568836212}
304
+ {"step": 9080, "epoch": 1, "loss": 3.2146451473236084, "loss_en": 3.213230848312378, "loss_zh": 3.216059446334839, "noise_std_eff": 0.018546168506145478, "elapsed_sec": 1211.5123674869537}
305
+ {"step": 9100, "epoch": 1, "loss": 3.7169923782348633, "loss_en": 3.767765998840332, "loss_zh": 3.6662187576293945, "noise_std_eff": 0.03926067054271698, "elapsed_sec": 1214.3836297988892}
306
+ {"step": 9120, "epoch": 1, "loss": 3.597135543823242, "loss_en": 3.607264757156372, "loss_zh": 3.5870065689086914, "noise_std_eff": 0.043200200796127325, "elapsed_sec": 1217.3083097934723}
307
+ {"step": 9140, "epoch": 1, "loss": 3.731299638748169, "loss_en": 3.7156081199645996, "loss_zh": 3.7469911575317383, "noise_std_eff": 0.01897257119417191, "elapsed_sec": 1220.1833341121674}
308
+ {"step": 9160, "epoch": 1, "loss": 3.611342430114746, "loss_en": 3.6081366539001465, "loss_zh": 3.6145482063293457, "noise_std_eff": 0.0060176152735948564, "elapsed_sec": 1223.1196792125702}
309
+ {"step": 9180, "epoch": 1, "loss": 3.780742645263672, "loss_en": 3.828747034072876, "loss_zh": 3.7327380180358887, "noise_std_eff": 0.025772345066070557, "elapsed_sec": 1226.0117223262787}
310
+ {"step": 9200, "epoch": 1, "loss": 3.517475128173828, "loss_en": 3.569464683532715, "loss_zh": 3.4654855728149414, "noise_std_eff": 0.009178172796964645, "elapsed_sec": 1228.9470844268799}
311
+ {"step": 9220, "epoch": 1, "loss": 3.441586971282959, "loss_en": 3.4321768283843994, "loss_zh": 3.4509973526000977, "noise_std_eff": 0.001447801571339369, "elapsed_sec": 1254.983098745346}
312
+ {"step": 9240, "epoch": 1, "loss": 3.3575057983398438, "loss_en": 3.3718342781066895, "loss_zh": 3.343177318572998, "noise_std_eff": 0.04363033771514893, "elapsed_sec": 1257.3367764949799}
313
+ {"step": 9260, "epoch": 1, "loss": 3.093778133392334, "loss_en": 3.1411943435668945, "loss_zh": 3.0463621616363525, "noise_std_eff": 0.009370112419128418, "elapsed_sec": 1259.7495758533478}
314
+ {"step": 9280, "epoch": 1, "loss": 3.8525938987731934, "loss_en": 3.8293981552124023, "loss_zh": 3.8757898807525635, "noise_std_eff": 0.005071293190121651, "elapsed_sec": 1262.1404435634613}
315
+ {"step": 9300, "epoch": 1, "loss": 3.743671417236328, "loss_en": 3.7551629543304443, "loss_zh": 3.732180118560791, "noise_std_eff": 0.0005030341912060976, "elapsed_sec": 1264.5297076702118}
316
+ {"step": 9320, "epoch": 1, "loss": 3.054743766784668, "loss_en": 3.0634632110595703, "loss_zh": 3.0460245609283447, "noise_std_eff": 0.025141239166259766, "elapsed_sec": 1266.8528966903687}
317
+ {"step": 9340, "epoch": 1, "loss": 3.557755947113037, "loss_en": 3.541110038757324, "loss_zh": 3.574402093887329, "noise_std_eff": 0.025561740994453432, "elapsed_sec": 1269.1844727993011}
318
+ {"step": 9360, "epoch": 1, "loss": 3.2811193466186523, "loss_en": 3.321833372116089, "loss_zh": 3.2404050827026367, "noise_std_eff": 0.04097640216350556, "elapsed_sec": 1271.5301988124847}
319
+ {"step": 9380, "epoch": 1, "loss": 4.03757381439209, "loss_en": 4.020235061645508, "loss_zh": 4.054912567138672, "noise_std_eff": 0.0031452402472496033, "elapsed_sec": 1273.8532543182373}
320
+ {"step": 9400, "epoch": 1, "loss": 3.867042064666748, "loss_en": 3.8557658195495605, "loss_zh": 3.8783183097839355, "noise_std_eff": 0.009496573358774185, "elapsed_sec": 1276.2468020915985}
321
+ {"step": 9420, "epoch": 1, "loss": 3.68877911567688, "loss_en": 3.630033016204834, "loss_zh": 3.747525215148926, "noise_std_eff": 0.02874661386013031, "elapsed_sec": 1278.6106934547424}
322
+ {"step": 9440, "epoch": 1, "loss": 3.006490707397461, "loss_en": 3.030698537826538, "loss_zh": 2.982283115386963, "noise_std_eff": 0.025831565260887146, "elapsed_sec": 1280.9639971256256}
323
+ {"step": 9460, "epoch": 1, "loss": 3.3294878005981445, "loss_en": 3.3170573711395264, "loss_zh": 3.3419179916381836, "noise_std_eff": 0.024274566769599916, "elapsed_sec": 1283.4178366661072}
324
+ {"step": 9480, "epoch": 1, "loss": 3.4388070106506348, "loss_en": 3.4432694911956787, "loss_zh": 3.434344530105591, "noise_std_eff": 0.03504686653614044, "elapsed_sec": 1285.7779865264893}
325
+ {"step": 9500, "epoch": 1, "loss": 3.1421003341674805, "loss_en": 3.190375804901123, "loss_zh": 3.093824863433838, "noise_std_eff": 0.010310503095388413, "elapsed_sec": 1288.1047310829163}
326
+ {"step": 9520, "epoch": 1, "loss": 3.5274243354797363, "loss_en": 3.4931561946868896, "loss_zh": 3.561692714691162, "noise_std_eff": 0.03359776437282563, "elapsed_sec": 1290.4823896884918}
327
+ {"step": 9540, "epoch": 1, "loss": 3.9323081970214844, "loss_en": 3.9227800369262695, "loss_zh": 3.9418365955352783, "noise_std_eff": 0.018743479251861574, "elapsed_sec": 1292.8081369400024}
328
+ {"step": 9560, "epoch": 1, "loss": 3.452800750732422, "loss_en": 3.4725637435913086, "loss_zh": 3.433037519454956, "noise_std_eff": 0.03169397115707397, "elapsed_sec": 1295.1851046085358}
329
+ {"step": 9580, "epoch": 1, "loss": 3.3509020805358887, "loss_en": 3.3410871028900146, "loss_zh": 3.360717296600342, "noise_std_eff": 0.04130745828151703, "elapsed_sec": 1297.4776918888092}
330
+ {"step": 9600, "epoch": 1, "loss": 4.148366928100586, "loss_en": 4.146603584289551, "loss_zh": 4.150129795074463, "noise_std_eff": 0.021206916868686677, "elapsed_sec": 1299.8253526687622}
331
+ {"step": 9620, "epoch": 1, "loss": 3.7113189697265625, "loss_en": 3.6756339073181152, "loss_zh": 3.7470037937164307, "noise_std_eff": 0.01635858863592148, "elapsed_sec": 1326.7439315319061}
332
+ {"step": 9640, "epoch": 1, "loss": 3.863720655441284, "loss_en": 3.8832528591156006, "loss_zh": 3.8441884517669678, "noise_std_eff": 0.002192802354693413, "elapsed_sec": 1329.685105085373}
333
+ {"step": 9660, "epoch": 1, "loss": 3.647785186767578, "loss_en": 3.639492988586426, "loss_zh": 3.6560771465301514, "noise_std_eff": 0.0002966349944472313, "elapsed_sec": 1332.5435664653778}
334
+ {"step": 9680, "epoch": 1, "loss": 3.2514748573303223, "loss_en": 3.2441797256469727, "loss_zh": 3.2587697505950928, "noise_std_eff": 0.014625000953674317, "elapsed_sec": 1335.4635004997253}
335
+ {"step": 9700, "epoch": 1, "loss": 2.854292392730713, "loss_en": 2.8173162937164307, "loss_zh": 2.891268730163574, "noise_std_eff": 0.019093209505081178, "elapsed_sec": 1338.3656075000763}
336
+ {"step": 9720, "epoch": 1, "loss": 3.2534852027893066, "loss_en": 3.221675157546997, "loss_zh": 3.2852954864501953, "noise_std_eff": 0.027387967705726626, "elapsed_sec": 1341.3970935344696}
337
+ {"step": 9740, "epoch": 1, "loss": 3.4182419776916504, "loss_en": 3.364600658416748, "loss_zh": 3.471883535385132, "noise_std_eff": 0.035036563873291016, "elapsed_sec": 1344.2721118927002}
338
+ {"step": 9760, "epoch": 1, "loss": 3.1465845108032227, "loss_en": 3.2070696353912354, "loss_zh": 3.086099147796631, "noise_std_eff": 0.007874303311109544, "elapsed_sec": 1347.1858723163605}
339
+ {"step": 9780, "epoch": 1, "loss": 3.1862552165985107, "loss_en": 3.2007668018341064, "loss_zh": 3.171743631362915, "noise_std_eff": 0.01973601132631302, "elapsed_sec": 1350.0842044353485}
340
+ {"step": 9800, "epoch": 1, "loss": 3.4898085594177246, "loss_en": 3.5098681449890137, "loss_zh": 3.4697492122650146, "noise_std_eff": 0.04905408322811127, "elapsed_sec": 1352.9968514442444}
341
+ {"step": 9820, "epoch": 1, "loss": 3.9632182121276855, "loss_en": 4.003405570983887, "loss_zh": 3.9230308532714844, "noise_std_eff": 0.012661474943161012, "elapsed_sec": 1355.8597161769867}
342
+ {"step": 9840, "epoch": 1, "loss": 3.7702183723449707, "loss_en": 3.780783176422119, "loss_zh": 3.7596538066864014, "noise_std_eff": 0.018382781744003297, "elapsed_sec": 1358.767734527588}
343
+ {"step": 9860, "epoch": 1, "loss": 4.127605438232422, "loss_en": 4.084735870361328, "loss_zh": 4.170474529266357, "noise_std_eff": 0.04068087935447693, "elapsed_sec": 1361.6767568588257}
344
+ {"step": 9880, "epoch": 1, "loss": 4.049582004547119, "loss_en": 4.089395523071289, "loss_zh": 4.009768486022949, "noise_std_eff": 0.0058359611779451375, "elapsed_sec": 1364.597999572754}
345
+ {"step": 9900, "epoch": 1, "loss": 3.151254177093506, "loss_en": 3.1190974712371826, "loss_zh": 3.183411121368408, "noise_std_eff": 0.013611856102943421, "elapsed_sec": 1367.4763798713684}
346
+ {"step": 9920, "epoch": 1, "loss": 3.7762346267700195, "loss_en": 3.813556671142578, "loss_zh": 3.738912582397461, "noise_std_eff": 0.032689303159713745, "elapsed_sec": 1370.40389585495}
347
+ {"step": 9940, "epoch": 1, "loss": 3.631178379058838, "loss_en": 3.633667469024658, "loss_zh": 3.6286892890930176, "noise_std_eff": 0.015365666151046753, "elapsed_sec": 1373.2716794013977}
348
+ {"step": 9960, "epoch": 1, "loss": 3.319643020629883, "loss_en": 3.3460960388183594, "loss_zh": 3.293189764022827, "noise_std_eff": 0.0028592441231012344, "elapsed_sec": 1376.2362315654755}
349
+ {"step": 9980, "epoch": 1, "loss": 3.495232582092285, "loss_en": 3.4843051433563232, "loss_zh": 3.506159782409668, "noise_std_eff": 0.004797963052988053, "elapsed_sec": 1379.1321737766266}
350
+ {"step": 10000, "epoch": 1, "loss": 2.9684224128723145, "loss_en": 2.9842941761016846, "loss_zh": 2.9525506496429443, "noise_std_eff": 0.023487502336502077, "elapsed_sec": 1382.0677452087402}
351
+ {"step": 10020, "epoch": 1, "loss": 4.060390472412109, "loss_en": 4.104584217071533, "loss_zh": 4.016196250915527, "noise_std_eff": 0.035269328951835634, "elapsed_sec": 1409.0446944236755}
352
+ {"step": 10040, "epoch": 1, "loss": 3.1888012886047363, "loss_en": 3.160996198654175, "loss_zh": 3.2166061401367188, "noise_std_eff": 0.018447910249233247, "elapsed_sec": 1412.0198712348938}
353
+ {"step": 10060, "epoch": 1, "loss": 3.7069013118743896, "loss_en": 3.6758744716644287, "loss_zh": 3.7379281520843506, "noise_std_eff": 0.04288336336612702, "elapsed_sec": 1414.9233703613281}
354
+ {"step": 10080, "epoch": 1, "loss": 3.271408796310425, "loss_en": 3.2642383575439453, "loss_zh": 3.2785792350769043, "noise_std_eff": 0.01855098158121109, "elapsed_sec": 1417.880051612854}
355
+ {"step": 10100, "epoch": 1, "loss": 3.2222654819488525, "loss_en": 3.2352328300476074, "loss_zh": 3.2092981338500977, "noise_std_eff": 0.015457212924957275, "elapsed_sec": 1420.776494026184}
356
+ {"step": 10120, "epoch": 1, "loss": 3.237520217895508, "loss_en": 3.234347343444824, "loss_zh": 3.2406928539276123, "noise_std_eff": 0.031058731675148013, "elapsed_sec": 1423.726938009262}
357
+ {"step": 10140, "epoch": 1, "loss": 3.8400514125823975, "loss_en": 3.887299060821533, "loss_zh": 3.7928037643432617, "noise_std_eff": 0.011641462892293931, "elapsed_sec": 1426.6351668834686}
358
+ {"step": 10160, "epoch": 1, "loss": 3.558549642562866, "loss_en": 3.4386703968048096, "loss_zh": 3.678428888320923, "noise_std_eff": 0.01074550822377205, "elapsed_sec": 1429.5803534984589}
359
+ {"step": 10180, "epoch": 1, "loss": 3.358921527862549, "loss_en": 3.3137192726135254, "loss_zh": 3.4041237831115723, "noise_std_eff": 0.010412432998418809, "elapsed_sec": 1432.4593951702118}
360
+ {"step": 10200, "epoch": 1, "loss": 3.482496738433838, "loss_en": 3.4633543491363525, "loss_zh": 3.5016393661499023, "noise_std_eff": 0.0025862095877528194, "elapsed_sec": 1435.3843767642975}
361
+ {"step": 10220, "epoch": 1, "loss": 3.6577138900756836, "loss_en": 3.635566234588623, "loss_zh": 3.6798617839813232, "noise_std_eff": 0.002125917933881283, "elapsed_sec": 1438.251941204071}
362
+ {"step": 10240, "epoch": 1, "loss": 3.098104953765869, "loss_en": 3.0873475074768066, "loss_zh": 3.1088624000549316, "noise_std_eff": 0.021428272128105164, "elapsed_sec": 1441.1650066375732}
363
+ {"step": 10260, "epoch": 1, "loss": 3.5883028507232666, "loss_en": 3.5801632404327393, "loss_zh": 3.596442461013794, "noise_std_eff": 0.03073742985725403, "elapsed_sec": 1444.0593831539154}
364
+ {"step": 10280, "epoch": 1, "loss": 4.077725410461426, "loss_en": 4.031756401062012, "loss_zh": 4.123693943023682, "noise_std_eff": 0.017945468425750732, "elapsed_sec": 1446.9756422042847}
365
+ {"step": 10300, "epoch": 1, "loss": 4.081348896026611, "loss_en": 4.108808517456055, "loss_zh": 4.053889274597168, "noise_std_eff": 0.043926295638084416, "elapsed_sec": 1449.8560388088226}
366
+ {"step": 10320, "epoch": 1, "loss": 3.3594229221343994, "loss_en": 3.3452980518341064, "loss_zh": 3.3735477924346924, "noise_std_eff": 0.03011934459209442, "elapsed_sec": 1452.779819726944}
367
+ {"step": 10340, "epoch": 1, "loss": 3.40291166305542, "loss_en": 3.4164326190948486, "loss_zh": 3.389390707015991, "noise_std_eff": 0.014891549944877625, "elapsed_sec": 1455.6601054668427}
368
+ {"step": 10360, "epoch": 1, "loss": 3.414872169494629, "loss_en": 3.423412561416626, "loss_zh": 3.406332015991211, "noise_std_eff": 0.005747144296765328, "elapsed_sec": 1458.6000821590424}
369
+ {"step": 10380, "epoch": 1, "loss": 3.5656566619873047, "loss_en": 3.5898995399475098, "loss_zh": 3.5414140224456787, "noise_std_eff": 0.04665699601173401, "elapsed_sec": 1461.464779138565}
370
+ {"step": 10400, "epoch": 1, "loss": 3.520392417907715, "loss_en": 3.4951775074005127, "loss_zh": 3.545607566833496, "noise_std_eff": 0.036817204952239994, "elapsed_sec": 1464.4207346439362}
371
+ {"step": 10420, "epoch": 1, "loss": 4.33503532409668, "loss_en": 4.274796962738037, "loss_zh": 4.3952741622924805, "noise_std_eff": 0.03025960922241211, "elapsed_sec": 1491.4349703788757}
372
+ {"step": 10440, "epoch": 1, "loss": 3.6360440254211426, "loss_en": 3.663865566253662, "loss_zh": 3.608222723007202, "noise_std_eff": 0.04737118184566498, "elapsed_sec": 1494.3233370780945}
373
+ {"step": 10460, "epoch": 1, "loss": 3.695004463195801, "loss_en": 3.721672773361206, "loss_zh": 3.6683361530303955, "noise_std_eff": 0.018961432576179507, "elapsed_sec": 1497.1954236030579}
374
+ {"step": 10480, "epoch": 1, "loss": 3.3425352573394775, "loss_en": 3.3026115894317627, "loss_zh": 3.3824589252471924, "noise_std_eff": 0.0018351640552282335, "elapsed_sec": 1500.1353998184204}
375
+ {"step": 10500, "epoch": 1, "loss": 3.8857200145721436, "loss_en": 3.914262294769287, "loss_zh": 3.857177734375, "noise_std_eff": 0.04130769670009613, "elapsed_sec": 1503.0160336494446}
376
+ {"step": 10520, "epoch": 1, "loss": 3.9557719230651855, "loss_en": 3.896888494491577, "loss_zh": 4.014655113220215, "noise_std_eff": 0.006075077503919602, "elapsed_sec": 1505.9249041080475}
377
+ {"step": 10540, "epoch": 1, "loss": 3.436661720275879, "loss_en": 3.421475887298584, "loss_zh": 3.451847791671753, "noise_std_eff": 0.0373053252696991, "elapsed_sec": 1508.8029181957245}
378
+ {"step": 10560, "epoch": 1, "loss": 4.100600242614746, "loss_en": 4.115294933319092, "loss_zh": 4.085906028747559, "noise_std_eff": 0.030404901504516604, "elapsed_sec": 1511.7190420627594}
379
+ {"step": 10580, "epoch": 1, "loss": 3.2508621215820312, "loss_en": 3.232055187225342, "loss_zh": 3.2696692943573, "noise_std_eff": 0.03379752933979035, "elapsed_sec": 1514.563381910324}
380
+ {"step": 10600, "epoch": 1, "loss": 3.637875556945801, "loss_en": 3.6468002796173096, "loss_zh": 3.628951072692871, "noise_std_eff": 0.039081916213035583, "elapsed_sec": 1517.4684872627258}
381
+ {"step": 10620, "epoch": 1, "loss": 3.1027474403381348, "loss_en": 3.0340023040771484, "loss_zh": 3.171492576599121, "noise_std_eff": 0.006078991293907166, "elapsed_sec": 1520.3767910003662}
382
+ {"step": 10640, "epoch": 1, "loss": 3.5895166397094727, "loss_en": 3.6053476333618164, "loss_zh": 3.57368540763855, "noise_std_eff": 0.031148651242256166, "elapsed_sec": 1523.285352230072}
383
+ {"step": 10660, "epoch": 1, "loss": 3.7786433696746826, "loss_en": 3.8009824752807617, "loss_zh": 3.7563042640686035, "noise_std_eff": 0.037338155508041385, "elapsed_sec": 1526.139268398285}
384
+ {"step": 10680, "epoch": 1, "loss": 3.941129684448242, "loss_en": 3.9494686126708984, "loss_zh": 3.932790756225586, "noise_std_eff": 0.001971634291112423, "elapsed_sec": 1529.1282305717468}
385
+ {"step": 10700, "epoch": 1, "loss": 3.4783294200897217, "loss_en": 3.488173484802246, "loss_zh": 3.4684853553771973, "noise_std_eff": 0.020042827725410464, "elapsed_sec": 1532.0151889324188}
386
+ {"step": 10720, "epoch": 1, "loss": 3.7155466079711914, "loss_en": 3.7340924739837646, "loss_zh": 3.697000741958618, "noise_std_eff": 0.01061624363064766, "elapsed_sec": 1534.936681985855}
387
+ {"step": 10740, "epoch": 1, "loss": 3.6281790733337402, "loss_en": 3.6210649013519287, "loss_zh": 3.635293483734131, "noise_std_eff": 0.048870992660522465, "elapsed_sec": 1537.8038058280945}
388
+ {"step": 10760, "epoch": 1, "loss": 3.699425220489502, "loss_en": 3.695544958114624, "loss_zh": 3.703305721282959, "noise_std_eff": 0.04694351255893708, "elapsed_sec": 1540.7089869976044}
389
+ {"step": 10780, "epoch": 1, "loss": 3.882625102996826, "loss_en": 3.8513050079345703, "loss_zh": 3.913944959640503, "noise_std_eff": 0.021112699806690217, "elapsed_sec": 1543.5756287574768}
390
+ {"step": 10800, "epoch": 1, "loss": 3.2098255157470703, "loss_en": 3.2295334339141846, "loss_zh": 3.190117597579956, "noise_std_eff": 0.0042612053453922275, "elapsed_sec": 1546.4849944114685}
391
+ {"step": 10820, "epoch": 1, "loss": 3.8496932983398438, "loss_en": 3.87634539604187, "loss_zh": 3.8230414390563965, "noise_std_eff": 0.02060060501098633, "elapsed_sec": 1573.3441588878632}
392
+ {"step": 10840, "epoch": 1, "loss": 3.4010562896728516, "loss_en": 3.369433641433716, "loss_zh": 3.432678699493408, "noise_std_eff": 0.04608264267444611, "elapsed_sec": 1576.2685675621033}
393
+ {"step": 10860, "epoch": 1, "loss": 3.809673309326172, "loss_en": 3.837045907974243, "loss_zh": 3.7823009490966797, "noise_std_eff": 0.013610725104808808, "elapsed_sec": 1579.1564271450043}
394
+ {"step": 10880, "epoch": 1, "loss": 3.5436530113220215, "loss_en": 3.546095848083496, "loss_zh": 3.541210412979126, "noise_std_eff": 0.0239156112074852, "elapsed_sec": 1582.0720417499542}
395
+ {"step": 10900, "epoch": 1, "loss": 3.487440347671509, "loss_en": 3.482983112335205, "loss_zh": 3.4918975830078125, "noise_std_eff": 0.005411064624786377, "elapsed_sec": 1584.944376707077}
396
+ {"step": 10920, "epoch": 1, "loss": 3.650326728820801, "loss_en": 3.6469547748565674, "loss_zh": 3.653698444366455, "noise_std_eff": 0.026540791988372805, "elapsed_sec": 1587.8683729171753}
397
+ {"step": 10940, "epoch": 1, "loss": 3.386251449584961, "loss_en": 3.333308696746826, "loss_zh": 3.4391942024230957, "noise_std_eff": 0.0062054269015789035, "elapsed_sec": 1590.7476780414581}
398
+ {"step": 10960, "epoch": 1, "loss": 3.4226388931274414, "loss_en": 3.3675456047058105, "loss_zh": 3.4777324199676514, "noise_std_eff": 0.01970397084951401, "elapsed_sec": 1593.6445314884186}
399
+ {"step": 10980, "epoch": 1, "loss": 3.708883047103882, "loss_en": 3.7208118438720703, "loss_zh": 3.6969542503356934, "noise_std_eff": 0.03941194415092469, "elapsed_sec": 1596.5072729587555}
400
+ {"step": 11000, "epoch": 1, "loss": 3.4471187591552734, "loss_en": 3.4652440547943115, "loss_zh": 3.4289932250976562, "noise_std_eff": 0.025443610548973084, "elapsed_sec": 1599.3807232379913}
401
+ {"step": 11020, "epoch": 1, "loss": 3.6898186206817627, "loss_en": 3.6924257278442383, "loss_zh": 3.687211513519287, "noise_std_eff": 0.04845547676086426, "elapsed_sec": 1602.2507753372192}
402
+ {"step": 11040, "epoch": 1, "loss": 3.3750386238098145, "loss_en": 3.3695831298828125, "loss_zh": 3.3804941177368164, "noise_std_eff": 0.047816455364227295, "elapsed_sec": 1605.129049539566}
403
+ {"step": 11060, "epoch": 1, "loss": 3.351186990737915, "loss_en": 3.351118326187134, "loss_zh": 3.3512556552886963, "noise_std_eff": 0.02339136004447937, "elapsed_sec": 1607.9823620319366}
404
+ {"step": 11080, "epoch": 1, "loss": 3.4688069820404053, "loss_en": 3.4484448432922363, "loss_zh": 3.489169120788574, "noise_std_eff": 0.0010768943466246128, "elapsed_sec": 1610.844673871994}
405
+ {"step": 11100, "epoch": 1, "loss": 4.098025321960449, "loss_en": 4.127313613891602, "loss_zh": 4.068737030029297, "noise_std_eff": 0.026802179217338563, "elapsed_sec": 1613.6797442436218}
406
+ {"step": 11120, "epoch": 1, "loss": 3.9455928802490234, "loss_en": 3.950082540512085, "loss_zh": 3.941103219985962, "noise_std_eff": 0.04509324431419373, "elapsed_sec": 1616.5725922584534}
407
+ {"step": 11140, "epoch": 1, "loss": 3.665066957473755, "loss_en": 3.721132516860962, "loss_zh": 3.609001398086548, "noise_std_eff": 0.015746809542179108, "elapsed_sec": 1619.4278948307037}
408
+ {"step": 11160, "epoch": 1, "loss": 3.9083709716796875, "loss_en": 3.9140288829803467, "loss_zh": 3.902712821960449, "noise_std_eff": 0.031537976861000065, "elapsed_sec": 1622.3332142829895}
409
+ {"step": 11180, "epoch": 1, "loss": 3.6941146850585938, "loss_en": 3.6421053409576416, "loss_zh": 3.746124029159546, "noise_std_eff": 0.018126986920833588, "elapsed_sec": 1625.2353568077087}
410
+ {"step": 11200, "epoch": 1, "loss": 3.473452091217041, "loss_en": 3.4536921977996826, "loss_zh": 3.4932122230529785, "noise_std_eff": 0.0035472605377435684, "elapsed_sec": 1628.1626329421997}
411
+ {"step": 11220, "epoch": 1, "loss": 3.6325576305389404, "loss_en": 3.6177518367767334, "loss_zh": 3.6473634243011475, "noise_std_eff": 0.01330697685480118, "elapsed_sec": 1654.7458097934723}
412
+ {"step": 11240, "epoch": 1, "loss": 3.483551025390625, "loss_en": 3.5028207302093506, "loss_zh": 3.4642813205718994, "noise_std_eff": 0.04436547160148621, "elapsed_sec": 1657.2865164279938}
413
+ {"step": 11260, "epoch": 1, "loss": 3.33858323097229, "loss_en": 3.3246374130249023, "loss_zh": 3.3525290489196777, "noise_std_eff": 0.04297734797000885, "elapsed_sec": 1659.7956941127777}
414
+ {"step": 11280, "epoch": 1, "loss": 3.632173538208008, "loss_en": 3.6676816940307617, "loss_zh": 3.596665143966675, "noise_std_eff": 0.034887203574180604, "elapsed_sec": 1662.354707956314}
415
+ {"step": 11300, "epoch": 1, "loss": 4.0609893798828125, "loss_en": 4.012557029724121, "loss_zh": 4.109421253204346, "noise_std_eff": 0.018542878329753876, "elapsed_sec": 1664.7364852428436}
416
+ {"step": 11320, "epoch": 1, "loss": 3.565112590789795, "loss_en": 3.4960992336273193, "loss_zh": 3.6341257095336914, "noise_std_eff": 0.024906384944915774, "elapsed_sec": 1667.2950024604797}
417
+ {"step": 11340, "epoch": 1, "loss": 3.6573405265808105, "loss_en": 3.6187527179718018, "loss_zh": 3.6959280967712402, "noise_std_eff": 0.03338271081447602, "elapsed_sec": 1669.7794797420502}
418
+ {"step": 11360, "epoch": 1, "loss": 3.779296875, "loss_en": 3.797266721725464, "loss_zh": 3.761326789855957, "noise_std_eff": 0.005613534152507782, "elapsed_sec": 1672.322598695755}
419
+ {"step": 11380, "epoch": 1, "loss": 3.47786283493042, "loss_en": 3.516798973083496, "loss_zh": 3.4389266967773438, "noise_std_eff": 0.013544566929340363, "elapsed_sec": 1674.8237166404724}
420
+ {"step": 11400, "epoch": 1, "loss": 3.3243613243103027, "loss_en": 3.321024179458618, "loss_zh": 3.3276984691619873, "noise_std_eff": 0.047597756981849676, "elapsed_sec": 1677.369820356369}
421
+ {"step": 11420, "epoch": 1, "loss": 3.501508951187134, "loss_en": 3.5036957263946533, "loss_zh": 3.4993221759796143, "noise_std_eff": 0.013394245505332948, "elapsed_sec": 1679.8643207550049}
422
+ {"step": 11440, "epoch": 1, "loss": 3.1637747287750244, "loss_en": 3.201995372772217, "loss_zh": 3.125554084777832, "noise_std_eff": 0.03984796404838562, "elapsed_sec": 1682.3543903827667}
423
+ {"step": 11460, "epoch": 1, "loss": 3.3321962356567383, "loss_en": 3.3008077144622803, "loss_zh": 3.363584518432617, "noise_std_eff": 0.005066724121570588, "elapsed_sec": 1684.8183472156525}
424
+ {"step": 11480, "epoch": 1, "loss": 3.429652690887451, "loss_en": 3.4411067962646484, "loss_zh": 3.418198823928833, "noise_std_eff": 0.01966173499822617, "elapsed_sec": 1687.2821004390717}
425
+ {"step": 11500, "epoch": 1, "loss": 3.5586843490600586, "loss_en": 3.554434061050415, "loss_zh": 3.562934637069702, "noise_std_eff": 0.03848623931407929, "elapsed_sec": 1689.756507396698}
426
+ {"step": 11520, "epoch": 1, "loss": 3.979015350341797, "loss_en": 4.01014518737793, "loss_zh": 3.947885513305664, "noise_std_eff": 0.026105797290802004, "elapsed_sec": 1692.278268814087}
427
+ {"step": 11540, "epoch": 1, "loss": 3.4547836780548096, "loss_en": 3.457902669906616, "loss_zh": 3.451664686203003, "noise_std_eff": 0.022950759530067446, "elapsed_sec": 1694.7442774772644}
428
+ {"step": 11560, "epoch": 1, "loss": 3.460193157196045, "loss_en": 3.441429615020752, "loss_zh": 3.478956460952759, "noise_std_eff": 0.012760822474956513, "elapsed_sec": 1697.2544553279877}
429
+ {"step": 11580, "epoch": 1, "loss": 3.5966222286224365, "loss_en": 3.6108508110046387, "loss_zh": 3.5823936462402344, "noise_std_eff": 0.0496271163225174, "elapsed_sec": 1699.749891281128}
430
+ {"step": 11600, "epoch": 1, "loss": 3.467602252960205, "loss_en": 3.551372528076172, "loss_zh": 3.3838322162628174, "noise_std_eff": 0.023271626234054568, "elapsed_sec": 1702.2333536148071}
431
+ {"step": 11620, "epoch": 1, "loss": 3.4516587257385254, "loss_en": 3.4535632133483887, "loss_zh": 3.449754238128662, "noise_std_eff": 0.03407792150974274, "elapsed_sec": 1729.1916601657867}
432
+ {"step": 11640, "epoch": 1, "loss": 4.808208465576172, "loss_en": 4.85988712310791, "loss_zh": 4.756529808044434, "noise_std_eff": 0.015318542718887329, "elapsed_sec": 1732.128095626831}
433
+ {"step": 11660, "epoch": 1, "loss": 3.583662509918213, "loss_en": 3.568666696548462, "loss_zh": 3.598658561706543, "noise_std_eff": 0.0018954390659928324, "elapsed_sec": 1735.0075759887695}
434
+ {"step": 11680, "epoch": 1, "loss": 3.4409360885620117, "loss_en": 3.430630922317505, "loss_zh": 3.4512410163879395, "noise_std_eff": 0.0094189390540123, "elapsed_sec": 1737.908096075058}
435
+ {"step": 11700, "epoch": 1, "loss": 2.9458885192871094, "loss_en": 2.9565188884735107, "loss_zh": 2.935258150100708, "noise_std_eff": 0.013586974143981934, "elapsed_sec": 1740.7635915279388}
436
+ {"step": 11720, "epoch": 1, "loss": 3.6882410049438477, "loss_en": 3.6866345405578613, "loss_zh": 3.689847469329834, "noise_std_eff": 0.003909290209412575, "elapsed_sec": 1743.6354489326477}
437
+ {"step": 11740, "epoch": 1, "loss": 3.907468318939209, "loss_en": 3.9116036891937256, "loss_zh": 3.9033329486846924, "noise_std_eff": 0.04914937913417816, "elapsed_sec": 1746.523773908615}
438
+ {"step": 11760, "epoch": 1, "loss": 2.9683847427368164, "loss_en": 2.948590040206909, "loss_zh": 2.9881794452667236, "noise_std_eff": 0.04478629231452942, "elapsed_sec": 1749.415760755539}
439
+ {"step": 11780, "epoch": 1, "loss": 3.9938602447509766, "loss_en": 4.043455600738525, "loss_zh": 3.944265127182007, "noise_std_eff": 0.04709245264530182, "elapsed_sec": 1752.2996520996094}
440
+ {"step": 11800, "epoch": 1, "loss": 2.930112838745117, "loss_en": 2.908519744873047, "loss_zh": 2.9517056941986084, "noise_std_eff": 0.017623503506183625, "elapsed_sec": 1755.217353105545}
441
+ {"step": 11820, "epoch": 1, "loss": 3.538057327270508, "loss_en": 3.453259229660034, "loss_zh": 3.6228556632995605, "noise_std_eff": 0.008376017212867737, "elapsed_sec": 1758.0800595283508}
442
+ {"step": 11840, "epoch": 1, "loss": 3.8042635917663574, "loss_en": 3.857928991317749, "loss_zh": 3.7505979537963867, "noise_std_eff": 0.02187556624412537, "elapsed_sec": 1760.9951906204224}
443
+ {"step": 11860, "epoch": 1, "loss": 3.4953155517578125, "loss_en": 3.531205177307129, "loss_zh": 3.459426164627075, "noise_std_eff": 0.0001003007171675563, "elapsed_sec": 1763.8595788478851}
444
+ {"step": 11880, "epoch": 1, "loss": 3.512098789215088, "loss_en": 3.500511407852173, "loss_zh": 3.523686408996582, "noise_std_eff": 0.008581289649009704, "elapsed_sec": 1766.8195798397064}
445
+ {"step": 11900, "epoch": 1, "loss": 3.6118478775024414, "loss_en": 3.617539882659912, "loss_zh": 3.6061558723449707, "noise_std_eff": 0.025606232881546023, "elapsed_sec": 1769.7149665355682}
446
+ {"step": 11920, "epoch": 1, "loss": 3.428551197052002, "loss_en": 3.404353380203247, "loss_zh": 3.4527487754821777, "noise_std_eff": 0.02237551659345627, "elapsed_sec": 1772.6731173992157}
447
+ {"step": 11940, "epoch": 1, "loss": 3.314603328704834, "loss_en": 3.2756969928741455, "loss_zh": 3.3535099029541016, "noise_std_eff": 0.011239050328731537, "elapsed_sec": 1775.5572776794434}
448
+ {"step": 11960, "epoch": 1, "loss": 3.9177210330963135, "loss_en": 3.9281728267669678, "loss_zh": 3.907269239425659, "noise_std_eff": 0.037756341695785525, "elapsed_sec": 1778.4760353565216}
planB_polish_robust_dec_noise05/probe_leakage.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_noise05/ckpt.pt",
3
+ "max_train_groups": 8000,
4
+ "max_valid_groups": 946,
5
+ "probe_hidden": 256,
6
+ "epochs": 20,
7
+ "lr": 0.01,
8
+ "baseline_majority_lang_acc": 0.5,
9
+ "baseline_majority_len_acc": 0.9271249771118164,
10
+ "lang_probe": {
11
+ "train_acc": 0.5860625,
12
+ "valid_acc": 0.5882663847780126,
13
+ "best_valid_acc": 0.596723044397463
14
+ },
15
+ "len_probe": {
16
+ "train_acc": 0.9288125,
17
+ "valid_acc": 0.9011627906976745,
18
+ "best_valid_acc": 0.9011627906976745
19
+ },
20
+ "lang_probe_length_matched": {
21
+ "train_acc": 0.5971865364850977,
22
+ "valid_acc": 0.6097023153252481,
23
+ "best_valid_acc": 0.6097023153252481
24
+ },
25
+ "lang_bucketwise_valid_acc": {
26
+ "bucket_0_acc": 0.34375,
27
+ "bucket_1_acc": 0.574999988079071,
28
+ "bucket_2_acc": 0.5999999642372131,
29
+ "bucket_3_acc": 0.5923753976821899
30
+ },
31
+ "length_matched_sizes": {
32
+ "train": 15568,
33
+ "valid": 1814
34
+ }
35
+ }
planB_polish_robust_dec_noise07_longwarm/ckpt.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b372d96b9bc4c4fea1170f1f88af81df80328945eedf67c3201dab054b353c22
3
+ size 2439403193
planB_polish_robust_dec_noise07_longwarm/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_jsonl": "data/groups_train.jsonl",
3
+ "valid_jsonl": "data/groups_valid.jsonl",
4
+ "run_dir": "runs/planB_polish_robust_dec_noise07_longwarm",
5
+ "resume": "runs/planB_polish/ckpt.pt",
6
+ "backbone": "google/mt5-small",
7
+ "num_latents": 16,
8
+ "batch_size": 4,
9
+ "grad_accum": 8,
10
+ "epochs": 2,
11
+ "max_doc_len": 256,
12
+ "max_sum_len": 64,
13
+ "eval_every": 400,
14
+ "max_train_examples": 0,
15
+ "max_valid_examples": 0,
16
+ "lr": 0.0002,
17
+ "weight_decay": 0.0,
18
+ "latent_dropout": 0.25,
19
+ "latent_noise_std": 0.07,
20
+ "noise_warmup_steps": 2500,
21
+ "noise_warmup_start_step": -1,
22
+ "untie_lm_head": true,
23
+ "seed": 42
24
+ }
planB_polish_robust_dec_noise07_longwarm/diag.jsonl ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3200, "nll_en": 4.794706048219451, "nll_zh": 4.800433078225521, "nll": 4.797569563222486}
2
+ {"step": 3600, "nll_en": 4.521231632152017, "nll_zh": 4.523748138742265, "nll": 4.522489885447142}
3
+ {"step": 4000, "nll_en": 4.346325073887379, "nll_zh": 4.348451669826064, "nll": 4.347388371856722}
4
+ {"step": 4400, "nll_en": 4.220105996847657, "nll_zh": 4.223607321370228, "nll": 4.221856659108942}
5
+ {"step": 4800, "nll_en": 4.11841867640457, "nll_zh": 4.123507970483308, "nll": 4.120963323443939}
6
+ {"step": 5200, "nll_en": 4.044372755175665, "nll_zh": 4.048080023953073, "nll": 4.046226389564369}
7
+ {"step": 5600, "nll_en": 3.9795823399662718, "nll_zh": 3.981269772914701, "nll": 3.9804260564404865}
8
+ {"step": 6000, "nll_en": 3.9281736146067967, "nll_zh": 3.9298997923385266, "nll": 3.929036703472662}
9
+ {"step": 6400, "nll_en": 3.8792620562095723, "nll_zh": 3.8827758137821897, "nll": 3.8810189349958812}
10
+ {"step": 6800, "nll_en": 3.847525051229852, "nll_zh": 3.8496814036218105, "nll": 3.8486032274258313}
11
+ {"step": 7200, "nll_en": 3.7993753152972296, "nll_zh": 3.8009545031864094, "nll": 3.8001649092418193}
12
+ {"step": 7600, "nll_en": 3.990521487423532, "nll_zh": 3.9883836883319104, "nll": 3.989452587877721}
13
+ {"step": 8000, "nll_en": 3.7546059576199875, "nll_zh": 3.7572081315845063, "nll": 3.755907044602247}
14
+ {"step": 8400, "nll_en": 3.7142923488173376, "nll_zh": 3.716958363514874, "nll": 3.7156253561661057}
15
+ {"step": 8800, "nll_en": 3.689040029779801, "nll_zh": 3.693409402577116, "nll": 3.6912247161784584}
16
+ {"step": 9200, "nll_en": 3.660560117967537, "nll_zh": 3.6644423678359823, "nll": 3.66250124290176}
17
+ {"step": 9600, "nll_en": 3.646389051925304, "nll_zh": 3.649315898564603, "nll": 3.6478524752449535}
18
+ {"step": 10000, "nll_en": 3.624204314032266, "nll_zh": 3.628514831212308, "nll": 3.6263595726222873}
19
+ {"step": 10400, "nll_en": 3.6072151363526035, "nll_zh": 3.613118091042904, "nll": 3.6101666136977535}
20
+ {"step": 10800, "nll_en": 3.5928711387118124, "nll_zh": 3.594432375143749, "nll": 3.5936517569277804}
21
+ {"step": 11200, "nll_en": 3.5673929026968634, "nll_zh": 3.5703728768588623, "nll": 3.568882889777863}
22
+ {"step": 11600, "nll_en": 3.551901196324548, "nll_zh": 3.557215574679869, "nll": 3.5545583855022085}
planB_polish_robust_dec_noise07_longwarm/diag_full.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_noise07_longwarm/ckpt.pt",
3
+ "nll_en": 3.5448289734112537,
4
+ "nll_zh": 3.548646854295791,
5
+ "swap_delta_en": 0.5619245311422529,
6
+ "swap_delta_zh": 0.5452998334711249,
7
+ "ablate_zero_delta_en": 1.5234333546418721,
8
+ "ablate_mean_delta_en": 0.2538326688851161,
9
+ "ablate_noise_delta_en": 4.565984893298804,
10
+ "ablate_zero_delta_zh": 1.5196154737573346,
11
+ "ablate_mean_delta_zh": 0.2438206541613847,
12
+ "ablate_noise_delta_zh": 4.6161616952424325,
13
+ "inv_top1_full": 0.060253698378801346,
14
+ "inv_top5_full": 0.20401690900325775,
15
+ "diag_sim_mean": 0.965337336063385,
16
+ "offdiag_sim_mean": 0.4495629370212555,
17
+ "sim_margin": 0.5157743990421295,
18
+ "n_valid": 946
19
+ }
planB_polish_robust_dec_noise07_longwarm/logs.jsonl ADDED
@@ -0,0 +1,448 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3020, "epoch": 0, "loss": 5.986845016479492, "loss_en": 5.8926215171813965, "loss_zh": 6.081068992614746, "noise_std_eff": 7.858898997306826e-05, "elapsed_sec": 2.781815767288208}
2
+ {"step": 3040, "epoch": 0, "loss": 5.110728740692139, "loss_en": 5.050684928894043, "loss_zh": 5.170772552490234, "noise_std_eff": 0.0009831032085418701, "elapsed_sec": 5.26045823097229}
3
+ {"step": 3060, "epoch": 0, "loss": 5.480625152587891, "loss_en": 5.507300853729248, "loss_zh": 5.453949451446533, "noise_std_eff": 0.0016719803237915042, "elapsed_sec": 7.410889625549316}
4
+ {"step": 3080, "epoch": 0, "loss": 5.712712287902832, "loss_en": 5.758455276489258, "loss_zh": 5.6669697761535645, "noise_std_eff": 0.0018957487297058108, "elapsed_sec": 9.88645625114441}
5
+ {"step": 3100, "epoch": 0, "loss": 5.422070503234863, "loss_en": 5.30646276473999, "loss_zh": 5.537678241729736, "noise_std_eff": 0.00121050500869751, "elapsed_sec": 12.351501941680908}
6
+ {"step": 3120, "epoch": 0, "loss": 5.856568336486816, "loss_en": 5.900378227233887, "loss_zh": 5.812757968902588, "noise_std_eff": 6.861443549394608e-05, "elapsed_sec": 15.00914454460144}
7
+ {"step": 3140, "epoch": 0, "loss": 5.048142433166504, "loss_en": 5.028130054473877, "loss_zh": 5.068154335021973, "noise_std_eff": 0.0003465293681621552, "elapsed_sec": 17.44188141822815}
8
+ {"step": 3160, "epoch": 0, "loss": 5.770446300506592, "loss_en": 5.968389987945557, "loss_zh": 5.572502613067627, "noise_std_eff": 0.0014507404136657716, "elapsed_sec": 19.629440784454346}
9
+ {"step": 3180, "epoch": 0, "loss": 5.1043701171875, "loss_en": 5.17366886138916, "loss_zh": 5.035070896148682, "noise_std_eff": 0.00025897223442792895, "elapsed_sec": 22.04859495162964}
10
+ {"step": 3200, "epoch": 0, "loss": 5.28609561920166, "loss_en": 5.372126579284668, "loss_zh": 5.2000651359558105, "noise_std_eff": 0.002007721018791199, "elapsed_sec": 24.30393671989441}
11
+ {"step": 3220, "epoch": 0, "loss": 5.128954887390137, "loss_en": 5.122016429901123, "loss_zh": 5.135892868041992, "noise_std_eff": 0.0010508929526805877, "elapsed_sec": 43.3745641708374}
12
+ {"step": 3240, "epoch": 0, "loss": 5.130999565124512, "loss_en": 5.1230926513671875, "loss_zh": 5.138906955718994, "noise_std_eff": 0.006027959861755371, "elapsed_sec": 45.795499086380005}
13
+ {"step": 3260, "epoch": 0, "loss": 4.877739906311035, "loss_en": 4.838327407836914, "loss_zh": 4.917151927947998, "noise_std_eff": 0.0014557037615776061, "elapsed_sec": 48.057175636291504}
14
+ {"step": 3280, "epoch": 0, "loss": 5.460089683532715, "loss_en": 5.439770698547363, "loss_zh": 5.480409145355225, "noise_std_eff": 0.004517692651748658, "elapsed_sec": 50.376293659210205}
15
+ {"step": 3300, "epoch": 0, "loss": 4.749672889709473, "loss_en": 4.740137100219727, "loss_zh": 4.759208679199219, "noise_std_eff": 0.0007467828422784806, "elapsed_sec": 52.67057967185974}
16
+ {"step": 3320, "epoch": 0, "loss": 4.3949127197265625, "loss_en": 4.420964241027832, "loss_zh": 4.368860721588135, "noise_std_eff": 0.007365809173583985, "elapsed_sec": 54.865506649017334}
17
+ {"step": 3340, "epoch": 0, "loss": 5.103445053100586, "loss_en": 5.1590752601623535, "loss_zh": 5.04781436920166, "noise_std_eff": 0.005833184175491334, "elapsed_sec": 57.22643709182739}
18
+ {"step": 3360, "epoch": 0, "loss": 4.470244407653809, "loss_en": 4.476426124572754, "loss_zh": 4.464062213897705, "noise_std_eff": 0.009198536767959595, "elapsed_sec": 59.332112073898315}
19
+ {"step": 3380, "epoch": 0, "loss": 4.570621490478516, "loss_en": 4.592737197875977, "loss_zh": 4.548505783081055, "noise_std_eff": 0.00677626410484314, "elapsed_sec": 61.65464973449707}
20
+ {"step": 3400, "epoch": 0, "loss": 4.349213600158691, "loss_en": 4.308261871337891, "loss_zh": 4.390165328979492, "noise_std_eff": 0.006160042190551759, "elapsed_sec": 63.7518253326416}
21
+ {"step": 3420, "epoch": 0, "loss": 3.6338348388671875, "loss_en": 3.5891623497009277, "loss_zh": 3.6785073280334473, "noise_std_eff": 0.011246337776184083, "elapsed_sec": 66.06710910797119}
22
+ {"step": 3440, "epoch": 0, "loss": 4.473969459533691, "loss_en": 4.449126243591309, "loss_zh": 4.498812198638916, "noise_std_eff": 0.004709625248908996, "elapsed_sec": 68.22406530380249}
23
+ {"step": 3460, "epoch": 0, "loss": 4.942563533782959, "loss_en": 4.952454566955566, "loss_zh": 4.932672500610352, "noise_std_eff": 0.003551559574604035, "elapsed_sec": 70.57203888893127}
24
+ {"step": 3480, "epoch": 0, "loss": 5.005748748779297, "loss_en": 5.004595756530762, "loss_zh": 5.006901264190674, "noise_std_eff": 0.004497517890930176, "elapsed_sec": 73.06695652008057}
25
+ {"step": 3500, "epoch": 0, "loss": 4.537383079528809, "loss_en": 4.523228168487549, "loss_zh": 4.55153751373291, "noise_std_eff": 0.010917426109313967, "elapsed_sec": 75.44050526618958}
26
+ {"step": 3520, "epoch": 0, "loss": 4.679610729217529, "loss_en": 4.690051555633545, "loss_zh": 4.669169902801514, "noise_std_eff": 0.0023017814373970033, "elapsed_sec": 77.78409123420715}
27
+ {"step": 3540, "epoch": 0, "loss": 4.661086082458496, "loss_en": 4.656205177307129, "loss_zh": 4.665966510772705, "noise_std_eff": 0.01493152198791504, "elapsed_sec": 79.8902063369751}
28
+ {"step": 3560, "epoch": 0, "loss": 4.41204833984375, "loss_en": 4.437816619873047, "loss_zh": 4.386279582977295, "noise_std_eff": 0.013764244937896731, "elapsed_sec": 82.19855833053589}
29
+ {"step": 3580, "epoch": 0, "loss": 4.521488189697266, "loss_en": 4.534658432006836, "loss_zh": 4.5083184242248535, "noise_std_eff": 0.011354692993164063, "elapsed_sec": 84.37242221832275}
30
+ {"step": 3600, "epoch": 0, "loss": 4.479526519775391, "loss_en": 4.42856502532959, "loss_zh": 4.53048849105835, "noise_std_eff": 0.011767701244354249, "elapsed_sec": 86.70983171463013}
31
+ {"step": 3620, "epoch": 0, "loss": 5.077143669128418, "loss_en": 5.078821182250977, "loss_zh": 5.075466156005859, "noise_std_eff": 0.010046512517929079, "elapsed_sec": 113.90172100067139}
32
+ {"step": 3640, "epoch": 0, "loss": 4.672732353210449, "loss_en": 4.662669658660889, "loss_zh": 4.68279504776001, "noise_std_eff": 0.009788193054199219, "elapsed_sec": 116.83793568611145}
33
+ {"step": 3660, "epoch": 0, "loss": 4.251448154449463, "loss_en": 4.251012325286865, "loss_zh": 4.2518839836120605, "noise_std_eff": 0.009599996824264528, "elapsed_sec": 119.76158046722412}
34
+ {"step": 3680, "epoch": 0, "loss": 4.153682708740234, "loss_en": 4.182936191558838, "loss_zh": 4.124429702758789, "noise_std_eff": 0.006669445004463197, "elapsed_sec": 122.6862325668335}
35
+ {"step": 3700, "epoch": 0, "loss": 4.848084449768066, "loss_en": 4.894137859344482, "loss_zh": 4.802030563354492, "noise_std_eff": 0.001784303689002991, "elapsed_sec": 125.61640906333923}
36
+ {"step": 3720, "epoch": 0, "loss": 4.472397804260254, "loss_en": 4.496124744415283, "loss_zh": 4.448671340942383, "noise_std_eff": 0.016499286975860596, "elapsed_sec": 128.55487203598022}
37
+ {"step": 3740, "epoch": 0, "loss": 4.315127849578857, "loss_en": 4.3045148849487305, "loss_zh": 4.325740814208984, "noise_std_eff": 0.007065633428096771, "elapsed_sec": 131.4330735206604}
38
+ {"step": 3760, "epoch": 0, "loss": 4.938814640045166, "loss_en": 4.9007568359375, "loss_zh": 4.976872444152832, "noise_std_eff": 0.004962630105018616, "elapsed_sec": 134.37521815299988}
39
+ {"step": 3780, "epoch": 0, "loss": 4.532111167907715, "loss_en": 4.502161979675293, "loss_zh": 4.5620598793029785, "noise_std_eff": 0.009630234882831575, "elapsed_sec": 137.2808518409729}
40
+ {"step": 3800, "epoch": 0, "loss": 4.3323822021484375, "loss_en": 4.343652725219727, "loss_zh": 4.321111679077148, "noise_std_eff": 0.01861200008392334, "elapsed_sec": 140.24570488929749}
41
+ {"step": 3820, "epoch": 0, "loss": 5.357076644897461, "loss_en": 5.356358528137207, "loss_zh": 5.357794284820557, "noise_std_eff": 0.02046398849964142, "elapsed_sec": 143.1416347026825}
42
+ {"step": 3840, "epoch": 0, "loss": 4.860408306121826, "loss_en": 4.809084415435791, "loss_zh": 4.911732196807861, "noise_std_eff": 0.014779510231018068, "elapsed_sec": 146.0823094844818}
43
+ {"step": 3860, "epoch": 0, "loss": 4.573185920715332, "loss_en": 4.561126708984375, "loss_zh": 4.585245609283447, "noise_std_eff": 0.01894843535423279, "elapsed_sec": 148.94411659240723}
44
+ {"step": 3880, "epoch": 0, "loss": 4.488958358764648, "loss_en": 4.508881568908691, "loss_zh": 4.469035625457764, "noise_std_eff": 0.00993771336555481, "elapsed_sec": 151.8971769809723}
45
+ {"step": 3900, "epoch": 0, "loss": 4.263396263122559, "loss_en": 4.295041561126709, "loss_zh": 4.23175048828125, "noise_std_eff": 0.009785044384002687, "elapsed_sec": 154.7843222618103}
46
+ {"step": 3920, "epoch": 0, "loss": 3.867588758468628, "loss_en": 3.8540427684783936, "loss_zh": 3.8811347484588623, "noise_std_eff": 0.012367329320907594, "elapsed_sec": 157.70870804786682}
47
+ {"step": 3940, "epoch": 0, "loss": 5.389166831970215, "loss_en": 5.355116844177246, "loss_zh": 5.423217296600342, "noise_std_eff": 0.021132612380981446, "elapsed_sec": 160.6159188747406}
48
+ {"step": 3960, "epoch": 0, "loss": 4.2297210693359375, "loss_en": 4.2145795822143555, "loss_zh": 4.244863033294678, "noise_std_eff": 0.009485384788513185, "elapsed_sec": 163.55691742897034}
49
+ {"step": 3980, "epoch": 0, "loss": 4.882489204406738, "loss_en": 4.861359119415283, "loss_zh": 4.903619289398193, "noise_std_eff": 0.02526163682460785, "elapsed_sec": 166.49854373931885}
50
+ {"step": 4000, "epoch": 0, "loss": 4.344159126281738, "loss_en": 4.361159324645996, "loss_zh": 4.327159404754639, "noise_std_eff": 0.027964057922363283, "elapsed_sec": 169.4417314529419}
51
+ {"step": 4020, "epoch": 0, "loss": 4.366164207458496, "loss_en": 4.385794162750244, "loss_zh": 4.346534252166748, "noise_std_eff": 0.006613372092247009, "elapsed_sec": 196.38436675071716}
52
+ {"step": 4040, "epoch": 0, "loss": 4.052545547485352, "loss_en": 4.088082790374756, "loss_zh": 4.017007827758789, "noise_std_eff": 0.02280518964767456, "elapsed_sec": 199.38572764396667}
53
+ {"step": 4060, "epoch": 0, "loss": 4.859553337097168, "loss_en": 4.834718227386475, "loss_zh": 4.884387969970703, "noise_std_eff": 0.0018020492053031922, "elapsed_sec": 202.3231508731842}
54
+ {"step": 4080, "epoch": 0, "loss": 4.1688714027404785, "loss_en": 4.106045246124268, "loss_zh": 4.2316975593566895, "noise_std_eff": 0.008449905581474305, "elapsed_sec": 205.2891902923584}
55
+ {"step": 4100, "epoch": 0, "loss": 4.409572124481201, "loss_en": 4.460554599761963, "loss_zh": 4.3585896492004395, "noise_std_eff": 0.005721477496623994, "elapsed_sec": 208.2057065963745}
56
+ {"step": 4120, "epoch": 0, "loss": 4.405026435852051, "loss_en": 4.425232887268066, "loss_zh": 4.384820461273193, "noise_std_eff": 0.025412669486999512, "elapsed_sec": 211.17340731620789}
57
+ {"step": 4140, "epoch": 0, "loss": 4.568731307983398, "loss_en": 4.5432963371276855, "loss_zh": 4.594166278839111, "noise_std_eff": 0.02470342935562134, "elapsed_sec": 214.09143161773682}
58
+ {"step": 4160, "epoch": 0, "loss": 4.1185078620910645, "loss_en": 4.118035793304443, "loss_zh": 4.1189799308776855, "noise_std_eff": 0.02389458193778992, "elapsed_sec": 217.03756427764893}
59
+ {"step": 4180, "epoch": 0, "loss": 4.644909858703613, "loss_en": 4.6294941902160645, "loss_zh": 4.660325527191162, "noise_std_eff": 0.007370590583086014, "elapsed_sec": 219.97336554527283}
60
+ {"step": 4200, "epoch": 0, "loss": 4.487537860870361, "loss_en": 4.438126087188721, "loss_zh": 4.536949634552002, "noise_std_eff": 0.011594553422927856, "elapsed_sec": 222.9346899986267}
61
+ {"step": 4220, "epoch": 0, "loss": 4.202093124389648, "loss_en": 4.173723220825195, "loss_zh": 4.230463027954102, "noise_std_eff": 0.013993359894752503, "elapsed_sec": 225.8573169708252}
62
+ {"step": 4240, "epoch": 0, "loss": 4.604063034057617, "loss_en": 4.55974817276001, "loss_zh": 4.648378372192383, "noise_std_eff": 0.018673637475967408, "elapsed_sec": 228.79629755020142}
63
+ {"step": 4260, "epoch": 0, "loss": 4.362873077392578, "loss_en": 4.348756313323975, "loss_zh": 4.376989841461182, "noise_std_eff": 0.031357481017112734, "elapsed_sec": 231.68327617645264}
64
+ {"step": 4280, "epoch": 0, "loss": 4.334543704986572, "loss_en": 4.286980628967285, "loss_zh": 4.382106781005859, "noise_std_eff": 0.027876611328125002, "elapsed_sec": 234.62679266929626}
65
+ {"step": 4300, "epoch": 0, "loss": 4.572750091552734, "loss_en": 4.5562825202941895, "loss_zh": 4.589217185974121, "noise_std_eff": 0.011209282004833222, "elapsed_sec": 237.52139616012573}
66
+ {"step": 4320, "epoch": 0, "loss": 4.376204490661621, "loss_en": 4.393830299377441, "loss_zh": 4.358578681945801, "noise_std_eff": 0.018174422478675845, "elapsed_sec": 240.48984289169312}
67
+ {"step": 4340, "epoch": 0, "loss": 4.456394195556641, "loss_en": 4.49788236618042, "loss_zh": 4.414905548095703, "noise_std_eff": 0.0312884369468689, "elapsed_sec": 243.3905611038208}
68
+ {"step": 4360, "epoch": 0, "loss": 4.119892120361328, "loss_en": 4.10729455947876, "loss_zh": 4.132489204406738, "noise_std_eff": 0.004505825173854829, "elapsed_sec": 246.35371375083923}
69
+ {"step": 4380, "epoch": 0, "loss": 4.1202497482299805, "loss_en": 4.119073390960693, "loss_zh": 4.121426582336426, "noise_std_eff": 0.01651676938533783, "elapsed_sec": 249.25478267669678}
70
+ {"step": 4400, "epoch": 0, "loss": 4.443902015686035, "loss_en": 4.384776592254639, "loss_zh": 4.50302791595459, "noise_std_eff": 0.00136356537938118, "elapsed_sec": 252.1777422428131}
71
+ {"step": 4420, "epoch": 0, "loss": 4.494271755218506, "loss_en": 4.5214033126831055, "loss_zh": 4.467140197753906, "noise_std_eff": 0.002891477193832397, "elapsed_sec": 279.18291664123535}
72
+ {"step": 4440, "epoch": 0, "loss": 4.560628890991211, "loss_en": 4.571142196655273, "loss_zh": 4.550116062164307, "noise_std_eff": 0.039662677345275885, "elapsed_sec": 282.3539206981659}
73
+ {"step": 4460, "epoch": 0, "loss": 4.263164520263672, "loss_en": 4.246719837188721, "loss_zh": 4.279608726501465, "noise_std_eff": 0.011816761696338654, "elapsed_sec": 285.23495221138}
74
+ {"step": 4480, "epoch": 0, "loss": 4.072495460510254, "loss_en": 4.069312572479248, "loss_zh": 4.07567834854126, "noise_std_eff": 0.005681927726268769, "elapsed_sec": 288.1976590156555}
75
+ {"step": 4500, "epoch": 0, "loss": 4.857064247131348, "loss_en": 4.879358291625977, "loss_zh": 4.834770679473877, "noise_std_eff": 0.011041744709014895, "elapsed_sec": 291.0817320346832}
76
+ {"step": 4520, "epoch": 0, "loss": 4.411576747894287, "loss_en": 4.412557601928711, "loss_zh": 4.410595893859863, "noise_std_eff": 0.024761463127136233, "elapsed_sec": 294.03734278678894}
77
+ {"step": 4540, "epoch": 0, "loss": 3.9592056274414062, "loss_en": 3.9349591732025146, "loss_zh": 3.983452081680298, "noise_std_eff": 0.02003816528558731, "elapsed_sec": 296.92961287498474}
78
+ {"step": 4560, "epoch": 0, "loss": 3.9450647830963135, "loss_en": 3.955389976501465, "loss_zh": 3.934739589691162, "noise_std_eff": 0.04003552019119263, "elapsed_sec": 299.9497277736664}
79
+ {"step": 4580, "epoch": 0, "loss": 4.187555313110352, "loss_en": 4.197013854980469, "loss_zh": 4.178096771240234, "noise_std_eff": 0.016928481612205507, "elapsed_sec": 302.85052967071533}
80
+ {"step": 4600, "epoch": 0, "loss": 4.409819602966309, "loss_en": 4.378664970397949, "loss_zh": 4.440974235534668, "noise_std_eff": 0.010188619041442873, "elapsed_sec": 305.8092083930969}
81
+ {"step": 4620, "epoch": 0, "loss": 5.133465766906738, "loss_en": 5.149107933044434, "loss_zh": 5.117823123931885, "noise_std_eff": 0.027800677113533023, "elapsed_sec": 308.7129852771759}
82
+ {"step": 4640, "epoch": 0, "loss": 4.14644718170166, "loss_en": 4.12014627456665, "loss_zh": 4.172747611999512, "noise_std_eff": 0.0374461625289917, "elapsed_sec": 311.6817162036896}
83
+ {"step": 4660, "epoch": 0, "loss": 3.930422782897949, "loss_en": 4.001120090484619, "loss_zh": 3.8597254753112793, "noise_std_eff": 0.009383774093389512, "elapsed_sec": 314.54599714279175}
84
+ {"step": 4680, "epoch": 0, "loss": 3.9129419326782227, "loss_en": 3.9549429416656494, "loss_zh": 3.870940685272217, "noise_std_eff": 0.03659843484878541, "elapsed_sec": 317.47362780570984}
85
+ {"step": 4700, "epoch": 0, "loss": 4.279188632965088, "loss_en": 4.33473539352417, "loss_zh": 4.223641872406006, "noise_std_eff": 0.03925569152832032, "elapsed_sec": 320.3549807071686}
86
+ {"step": 4720, "epoch": 0, "loss": 4.450679779052734, "loss_en": 4.465057849884033, "loss_zh": 4.436302185058594, "noise_std_eff": 0.04682440608024597, "elapsed_sec": 323.26088666915894}
87
+ {"step": 4740, "epoch": 0, "loss": 4.579409599304199, "loss_en": 4.52665901184082, "loss_zh": 4.63215970993042, "noise_std_eff": 0.025557531881332397, "elapsed_sec": 326.15421509742737}
88
+ {"step": 4760, "epoch": 0, "loss": 4.173710823059082, "loss_en": 4.197210311889648, "loss_zh": 4.150211334228516, "noise_std_eff": 0.03377825134277344, "elapsed_sec": 329.0343120098114}
89
+ {"step": 4780, "epoch": 0, "loss": 4.328391075134277, "loss_en": 4.325702667236328, "loss_zh": 4.331079006195068, "noise_std_eff": 0.023517173717021944, "elapsed_sec": 331.8902897834778}
90
+ {"step": 4800, "epoch": 0, "loss": 4.028041839599609, "loss_en": 4.080708026885986, "loss_zh": 3.975375175476074, "noise_std_eff": 0.0018584111094474794, "elapsed_sec": 334.8984408378601}
91
+ {"step": 4820, "epoch": 0, "loss": 3.8175220489501953, "loss_en": 3.821080207824707, "loss_zh": 3.8139641284942627, "noise_std_eff": 0.04854737562179566, "elapsed_sec": 362.1697783470154}
92
+ {"step": 4840, "epoch": 0, "loss": 4.354378700256348, "loss_en": 4.322177886962891, "loss_zh": 4.386579513549805, "noise_std_eff": 4.834219586104155e-05, "elapsed_sec": 365.18159222602844}
93
+ {"step": 4860, "epoch": 0, "loss": 3.829887866973877, "loss_en": 3.786850690841675, "loss_zh": 3.872925281524658, "noise_std_eff": 0.009861168422698977, "elapsed_sec": 368.1052665710449}
94
+ {"step": 4880, "epoch": 0, "loss": 4.048504829406738, "loss_en": 4.0134758949279785, "loss_zh": 4.083534240722656, "noise_std_eff": 0.029026634445190432, "elapsed_sec": 371.06097078323364}
95
+ {"step": 4900, "epoch": 0, "loss": 4.045708656311035, "loss_en": 4.030165195465088, "loss_zh": 4.061251640319824, "noise_std_eff": 0.030876299643516545, "elapsed_sec": 373.98658871650696}
96
+ {"step": 4920, "epoch": 0, "loss": 4.493852138519287, "loss_en": 4.483942031860352, "loss_zh": 4.503762245178223, "noise_std_eff": 0.028039944305419926, "elapsed_sec": 376.969051361084}
97
+ {"step": 4940, "epoch": 0, "loss": 4.501288414001465, "loss_en": 4.507800579071045, "loss_zh": 4.494776248931885, "noise_std_eff": 0.027630405726432803, "elapsed_sec": 379.8697717189789}
98
+ {"step": 4960, "epoch": 0, "loss": 4.750514030456543, "loss_en": 4.766746997833252, "loss_zh": 4.734281063079834, "noise_std_eff": 0.00655486947655678, "elapsed_sec": 382.7783143520355}
99
+ {"step": 4980, "epoch": 0, "loss": 3.8351478576660156, "loss_en": 3.743594169616699, "loss_zh": 3.926701784133911, "noise_std_eff": 0.020422650704383854, "elapsed_sec": 385.6575753688812}
100
+ {"step": 5000, "epoch": 0, "loss": 3.977119207382202, "loss_en": 3.9535417556762695, "loss_zh": 4.000696659088135, "noise_std_eff": 0.008713628172874452, "elapsed_sec": 388.5980751514435}
101
+ {"step": 5020, "epoch": 0, "loss": 4.146731376647949, "loss_en": 4.1421051025390625, "loss_zh": 4.151358127593994, "noise_std_eff": 0.04813335960388184, "elapsed_sec": 391.4974699020386}
102
+ {"step": 5040, "epoch": 0, "loss": 4.400363445281982, "loss_en": 4.383790493011475, "loss_zh": 4.41693639755249, "noise_std_eff": 0.03937999978065491, "elapsed_sec": 394.4704556465149}
103
+ {"step": 5060, "epoch": 0, "loss": 3.985252857208252, "loss_en": 4.006101608276367, "loss_zh": 3.9644041061401367, "noise_std_eff": 0.03978394547462463, "elapsed_sec": 397.3857491016388}
104
+ {"step": 5080, "epoch": 0, "loss": 3.8949356079101562, "loss_en": 3.835376262664795, "loss_zh": 3.9544951915740967, "noise_std_eff": 0.005828753266334534, "elapsed_sec": 400.35742449760437}
105
+ {"step": 5100, "epoch": 0, "loss": 3.6243839263916016, "loss_en": 3.5618367195129395, "loss_zh": 3.6869313716888428, "noise_std_eff": 0.028961697328090667, "elapsed_sec": 403.25404930114746}
106
+ {"step": 5120, "epoch": 0, "loss": 4.217538833618164, "loss_en": 4.19871711730957, "loss_zh": 4.236360549926758, "noise_std_eff": 0.053276125907897945, "elapsed_sec": 406.20203471183777}
107
+ {"step": 5140, "epoch": 0, "loss": 5.3510823249816895, "loss_en": 5.320361137390137, "loss_zh": 5.381803512573242, "noise_std_eff": 0.019034633872509005, "elapsed_sec": 409.1090941429138}
108
+ {"step": 5160, "epoch": 0, "loss": 3.9096765518188477, "loss_en": 3.9293746948242188, "loss_zh": 3.8899786472320557, "noise_std_eff": 0.006180950195789337, "elapsed_sec": 412.05773973464966}
109
+ {"step": 5180, "epoch": 0, "loss": 3.620575428009033, "loss_en": 3.6050171852111816, "loss_zh": 3.6361334323883057, "noise_std_eff": 0.015548281764984134, "elapsed_sec": 414.91856241226196}
110
+ {"step": 5200, "epoch": 0, "loss": 3.830875873565674, "loss_en": 3.7847704887390137, "loss_zh": 3.876981258392334, "noise_std_eff": 0.010526023793220521, "elapsed_sec": 417.8486783504486}
111
+ {"step": 5220, "epoch": 0, "loss": 3.7946486473083496, "loss_en": 3.777726411819458, "loss_zh": 3.811570644378662, "noise_std_eff": 0.007794772721529008, "elapsed_sec": 446.38242197036743}
112
+ {"step": 5240, "epoch": 0, "loss": 4.507078170776367, "loss_en": 4.534722328186035, "loss_zh": 4.479433536529541, "noise_std_eff": 0.014949452438354494, "elapsed_sec": 449.3352527618408}
113
+ {"step": 5260, "epoch": 0, "loss": 4.1660966873168945, "loss_en": 4.179179668426514, "loss_zh": 4.153013706207275, "noise_std_eff": 0.03545955908298493, "elapsed_sec": 452.0081603527069}
114
+ {"step": 5280, "epoch": 0, "loss": 4.198241233825684, "loss_en": 4.196399688720703, "loss_zh": 4.200082302093506, "noise_std_eff": 0.041196364917755135, "elapsed_sec": 454.431853055954}
115
+ {"step": 5300, "epoch": 0, "loss": 4.504748344421387, "loss_en": 4.474172592163086, "loss_zh": 4.535323619842529, "noise_std_eff": 0.022580505800247196, "elapsed_sec": 456.88240003585815}
116
+ {"step": 5320, "epoch": 0, "loss": 4.203388214111328, "loss_en": 4.223842620849609, "loss_zh": 4.182933807373047, "noise_std_eff": 0.06036116456985474, "elapsed_sec": 459.285124540329}
117
+ {"step": 5340, "epoch": 0, "loss": 4.73929500579834, "loss_en": 4.7213134765625, "loss_zh": 4.75727653503418, "noise_std_eff": 0.05435510147094728, "elapsed_sec": 461.8900954723358}
118
+ {"step": 5360, "epoch": 0, "loss": 3.9858193397521973, "loss_en": 3.96771502494812, "loss_zh": 4.003923416137695, "noise_std_eff": 0.03541685640335083, "elapsed_sec": 464.3275001049042}
119
+ {"step": 5380, "epoch": 0, "loss": 4.051323890686035, "loss_en": 4.102203369140625, "loss_zh": 4.000443935394287, "noise_std_eff": 0.047413330507278445, "elapsed_sec": 466.74207758903503}
120
+ {"step": 5400, "epoch": 0, "loss": 4.342013835906982, "loss_en": 4.342574119567871, "loss_zh": 4.341453552246094, "noise_std_eff": 0.025201790428161622, "elapsed_sec": 469.02672481536865}
121
+ {"step": 5420, "epoch": 0, "loss": 4.559568405151367, "loss_en": 4.523775577545166, "loss_zh": 4.595361709594727, "noise_std_eff": 0.03628752295970917, "elapsed_sec": 471.2060401439667}
122
+ {"step": 5440, "epoch": 0, "loss": 3.9663968086242676, "loss_en": 3.908762216567993, "loss_zh": 4.024031639099121, "noise_std_eff": 0.03326543570041657, "elapsed_sec": 473.56421184539795}
123
+ {"step": 5460, "epoch": 0, "loss": 4.148341178894043, "loss_en": 4.144768238067627, "loss_zh": 4.151914596557617, "noise_std_eff": 0.055880121488571174, "elapsed_sec": 476.1413540840149}
124
+ {"step": 5480, "epoch": 0, "loss": 4.0661797523498535, "loss_en": 4.001491069793701, "loss_zh": 4.130868434906006, "noise_std_eff": 0.044909461383819585, "elapsed_sec": 478.367431640625}
125
+ {"step": 5500, "epoch": 0, "loss": 4.140560150146484, "loss_en": 4.117710113525391, "loss_zh": 4.163410663604736, "noise_std_eff": 0.058321490883827216, "elapsed_sec": 480.66103196144104}
126
+ {"step": 5520, "epoch": 0, "loss": 4.390240669250488, "loss_en": 4.395776271820068, "loss_zh": 4.384705543518066, "noise_std_eff": 0.04875055670738221, "elapsed_sec": 483.030725479126}
127
+ {"step": 5540, "epoch": 0, "loss": 4.186663627624512, "loss_en": 4.1804022789001465, "loss_zh": 4.192925453186035, "noise_std_eff": 0.0371802568435669, "elapsed_sec": 485.26995277404785}
128
+ {"step": 5560, "epoch": 0, "loss": 4.121676445007324, "loss_en": 4.064952850341797, "loss_zh": 4.178400039672852, "noise_std_eff": 0.005181215256452561, "elapsed_sec": 487.49271178245544}
129
+ {"step": 5580, "epoch": 0, "loss": 3.8065671920776367, "loss_en": 3.9331440925598145, "loss_zh": 3.67999005317688, "noise_std_eff": 0.01667791321873665, "elapsed_sec": 490.01214838027954}
130
+ {"step": 5600, "epoch": 0, "loss": 4.23543643951416, "loss_en": 4.283048152923584, "loss_zh": 4.1878252029418945, "noise_std_eff": 0.023489595949649812, "elapsed_sec": 492.3196849822998}
131
+ {"step": 5620, "epoch": 0, "loss": 4.493361949920654, "loss_en": 4.406258583068848, "loss_zh": 4.580465316772461, "noise_std_eff": 0.02120136559009552, "elapsed_sec": 519.237035036087}
132
+ {"step": 5640, "epoch": 0, "loss": 4.87975549697876, "loss_en": 4.868552207946777, "loss_zh": 4.890958786010742, "noise_std_eff": 0.040989502072334295, "elapsed_sec": 522.2129235267639}
133
+ {"step": 5660, "epoch": 0, "loss": 4.086063385009766, "loss_en": 4.043007850646973, "loss_zh": 4.1291184425354, "noise_std_eff": 0.05637977838516236, "elapsed_sec": 525.1208906173706}
134
+ {"step": 5680, "epoch": 0, "loss": 4.169104099273682, "loss_en": 4.145082473754883, "loss_zh": 4.1931257247924805, "noise_std_eff": 0.0377193421125412, "elapsed_sec": 528.0418429374695}
135
+ {"step": 5700, "epoch": 0, "loss": 3.8147125244140625, "loss_en": 3.800499439239502, "loss_zh": 3.828925371170044, "noise_std_eff": 0.014340692460536958, "elapsed_sec": 530.8968467712402}
136
+ {"step": 5720, "epoch": 0, "loss": 4.265463829040527, "loss_en": 4.290288925170898, "loss_zh": 4.240638732910156, "noise_std_eff": 0.025813163816928865, "elapsed_sec": 533.8484218120575}
137
+ {"step": 5740, "epoch": 0, "loss": 4.1363677978515625, "loss_en": 4.104730606079102, "loss_zh": 4.168004512786865, "noise_std_eff": 0.003236311711370945, "elapsed_sec": 536.7860491275787}
138
+ {"step": 5760, "epoch": 0, "loss": 4.204039573669434, "loss_en": 4.145918369293213, "loss_zh": 4.262160778045654, "noise_std_eff": 0.0146617466211319, "elapsed_sec": 539.7173008918762}
139
+ {"step": 5780, "epoch": 0, "loss": 4.096221923828125, "loss_en": 4.068639755249023, "loss_zh": 4.123804569244385, "noise_std_eff": 0.016515393853187562, "elapsed_sec": 542.6489109992981}
140
+ {"step": 5800, "epoch": 0, "loss": 4.521763801574707, "loss_en": 4.5101542472839355, "loss_zh": 4.53337287902832, "noise_std_eff": 0.00810808151960373, "elapsed_sec": 545.6100914478302}
141
+ {"step": 5820, "epoch": 0, "loss": 4.20004940032959, "loss_en": 4.1521897315979, "loss_zh": 4.247909069061279, "noise_std_eff": 0.06337000846862793, "elapsed_sec": 548.4899616241455}
142
+ {"step": 5840, "epoch": 0, "loss": 3.5480873584747314, "loss_en": 3.5189874172210693, "loss_zh": 3.5771872997283936, "noise_std_eff": 0.01580326557159424, "elapsed_sec": 551.4734673500061}
143
+ {"step": 5860, "epoch": 0, "loss": 4.267176628112793, "loss_en": 4.273794174194336, "loss_zh": 4.260558605194092, "noise_std_eff": 0.011143178939819337, "elapsed_sec": 554.3538508415222}
144
+ {"step": 5880, "epoch": 0, "loss": 4.050136566162109, "loss_en": 4.093631744384766, "loss_zh": 4.006641387939453, "noise_std_eff": 0.0014928485453128816, "elapsed_sec": 557.2734303474426}
145
+ {"step": 5900, "epoch": 0, "loss": 4.295360565185547, "loss_en": 4.279545783996582, "loss_zh": 4.3111748695373535, "noise_std_eff": 0.04856685757637024, "elapsed_sec": 560.1173288822174}
146
+ {"step": 5920, "epoch": 0, "loss": 4.433374404907227, "loss_en": 4.405166149139404, "loss_zh": 4.461582183837891, "noise_std_eff": 0.004819557070732118, "elapsed_sec": 563.0092144012451}
147
+ {"step": 5940, "epoch": 0, "loss": 3.9733777046203613, "loss_en": 3.9468395709991455, "loss_zh": 3.9999160766601562, "noise_std_eff": 0.02881267756223679, "elapsed_sec": 565.8648431301117}
148
+ {"step": 5960, "epoch": 0, "loss": 4.471737861633301, "loss_en": 4.426492214202881, "loss_zh": 4.516983985900879, "noise_std_eff": 0.0107038351893425, "elapsed_sec": 568.765052318573}
149
+ {"step": 5980, "epoch": 0, "loss": 3.9136838912963867, "loss_en": 3.900977849960327, "loss_zh": 3.926389694213867, "noise_std_eff": 0.01905629396438599, "elapsed_sec": 571.6205358505249}
150
+ {"step": 6000, "epoch": 0, "loss": 4.016775131225586, "loss_en": 3.992125988006592, "loss_zh": 4.04142427444458, "noise_std_eff": 0.006596365720033646, "elapsed_sec": 574.5222299098969}
151
+ {"step": 6020, "epoch": 0, "loss": 3.989560127258301, "loss_en": 3.9308879375457764, "loss_zh": 4.048232078552246, "noise_std_eff": 0.05748674631118775, "elapsed_sec": 601.2926602363586}
152
+ {"step": 6040, "epoch": 0, "loss": 3.376065969467163, "loss_en": 3.368990421295166, "loss_zh": 3.38314151763916, "noise_std_eff": 0.04702966451644898, "elapsed_sec": 604.205082654953}
153
+ {"step": 6060, "epoch": 0, "loss": 3.952813148498535, "loss_en": 3.9166178703308105, "loss_zh": 3.9890084266662598, "noise_std_eff": 0.04470048487186432, "elapsed_sec": 607.088583946228}
154
+ {"step": 6080, "epoch": 0, "loss": 4.101181983947754, "loss_en": 4.146279335021973, "loss_zh": 4.056085109710693, "noise_std_eff": 0.06284062385559083, "elapsed_sec": 609.9765648841858}
155
+ {"step": 6100, "epoch": 0, "loss": 4.069345474243164, "loss_en": 4.08486795425415, "loss_zh": 4.053822994232178, "noise_std_eff": 0.0001935501862317324, "elapsed_sec": 612.8489093780518}
156
+ {"step": 6120, "epoch": 0, "loss": 4.393695831298828, "loss_en": 4.356238842010498, "loss_zh": 4.431153297424316, "noise_std_eff": 0.0224576672911644, "elapsed_sec": 615.7489862442017}
157
+ {"step": 6140, "epoch": 0, "loss": 3.9573731422424316, "loss_en": 3.972172975540161, "loss_zh": 3.942573308944702, "noise_std_eff": 0.001566358655691147, "elapsed_sec": 618.6096096038818}
158
+ {"step": 6160, "epoch": 0, "loss": 4.114423751831055, "loss_en": 4.141165256500244, "loss_zh": 4.087681770324707, "noise_std_eff": 0.06716352403163911, "elapsed_sec": 621.5013420581818}
159
+ {"step": 6180, "epoch": 0, "loss": 4.079555511474609, "loss_en": 4.038172245025635, "loss_zh": 4.120938301086426, "noise_std_eff": 0.05416383147239685, "elapsed_sec": 624.3803544044495}
160
+ {"step": 6200, "epoch": 0, "loss": 3.6254124641418457, "loss_en": 3.6141750812530518, "loss_zh": 3.6366496086120605, "noise_std_eff": 0.028075557053089145, "elapsed_sec": 627.3173379898071}
161
+ {"step": 6220, "epoch": 0, "loss": 3.862109661102295, "loss_en": 3.873476028442383, "loss_zh": 3.850743532180786, "noise_std_eff": 0.01324199825525284, "elapsed_sec": 630.1772742271423}
162
+ {"step": 6240, "epoch": 0, "loss": 3.779439926147461, "loss_en": 3.7666163444519043, "loss_zh": 3.7922637462615967, "noise_std_eff": 0.013931833803653718, "elapsed_sec": 633.0980422496796}
163
+ {"step": 6260, "epoch": 0, "loss": 4.074182510375977, "loss_en": 4.078605651855469, "loss_zh": 4.069758892059326, "noise_std_eff": 0.06306951344013215, "elapsed_sec": 635.9770836830139}
164
+ {"step": 6280, "epoch": 0, "loss": 4.313384056091309, "loss_en": 4.294676303863525, "loss_zh": 4.332091808319092, "noise_std_eff": 0.04008764147758484, "elapsed_sec": 638.9018838405609}
165
+ {"step": 6300, "epoch": 0, "loss": 3.891913890838623, "loss_en": 3.9231348037719727, "loss_zh": 3.8606932163238525, "noise_std_eff": 0.05614115476608277, "elapsed_sec": 641.7934110164642}
166
+ {"step": 6320, "epoch": 0, "loss": 3.7147061824798584, "loss_en": 3.7026147842407227, "loss_zh": 3.726797580718994, "noise_std_eff": 0.023570098876953126, "elapsed_sec": 644.7298879623413}
167
+ {"step": 6340, "epoch": 0, "loss": 3.3980984687805176, "loss_en": 3.3654356002807617, "loss_zh": 3.4307615756988525, "noise_std_eff": 0.01639047130942345, "elapsed_sec": 647.6011493206024}
168
+ {"step": 6360, "epoch": 0, "loss": 4.1283369064331055, "loss_en": 4.1105756759643555, "loss_zh": 4.146097660064697, "noise_std_eff": 0.018359017074108124, "elapsed_sec": 650.5421738624573}
169
+ {"step": 6380, "epoch": 0, "loss": 3.766204833984375, "loss_en": 3.703035593032837, "loss_zh": 3.829373836517334, "noise_std_eff": 0.03367998063564301, "elapsed_sec": 653.4377176761627}
170
+ {"step": 6400, "epoch": 0, "loss": 4.061345100402832, "loss_en": 4.121177673339844, "loss_zh": 4.0015130043029785, "noise_std_eff": 0.054822044968605044, "elapsed_sec": 656.3739910125732}
171
+ {"step": 6420, "epoch": 0, "loss": 4.096801280975342, "loss_en": 4.056675434112549, "loss_zh": 4.136927127838135, "noise_std_eff": 0.03911749660968781, "elapsed_sec": 683.1971969604492}
172
+ {"step": 6440, "epoch": 0, "loss": 4.061692714691162, "loss_en": 4.084464073181152, "loss_zh": 4.038921356201172, "noise_std_eff": 0.005621498823165894, "elapsed_sec": 686.1290557384491}
173
+ {"step": 6460, "epoch": 0, "loss": 4.054237365722656, "loss_en": 4.031150817871094, "loss_zh": 4.077324390411377, "noise_std_eff": 0.04972681820392609, "elapsed_sec": 688.988603591919}
174
+ {"step": 6480, "epoch": 0, "loss": 4.110766887664795, "loss_en": 4.103132724761963, "loss_zh": 4.118401050567627, "noise_std_eff": 0.011596765965223313, "elapsed_sec": 691.8931307792664}
175
+ {"step": 6500, "epoch": 0, "loss": 3.64078426361084, "loss_en": 3.6560754776000977, "loss_zh": 3.625493049621582, "noise_std_eff": 0.0533241093158722, "elapsed_sec": 694.746062040329}
176
+ {"step": 6520, "epoch": 0, "loss": 3.8462471961975098, "loss_en": 3.8571696281433105, "loss_zh": 3.83532452583313, "noise_std_eff": 0.0669382518529892, "elapsed_sec": 697.625461101532}
177
+ {"step": 6540, "epoch": 0, "loss": 3.797457695007324, "loss_en": 3.793140411376953, "loss_zh": 3.801774740219116, "noise_std_eff": 0.002349171601235867, "elapsed_sec": 700.4961698055267}
178
+ {"step": 6560, "epoch": 0, "loss": 4.667041778564453, "loss_en": 4.690582752227783, "loss_zh": 4.643500804901123, "noise_std_eff": 0.06974368989467622, "elapsed_sec": 703.417165517807}
179
+ {"step": 6580, "epoch": 0, "loss": 3.789231300354004, "loss_en": 3.7608349323272705, "loss_zh": 3.817627429962158, "noise_std_eff": 0.036118062138557436, "elapsed_sec": 706.2562501430511}
180
+ {"step": 6600, "epoch": 0, "loss": 4.075246810913086, "loss_en": 4.043938159942627, "loss_zh": 4.106555461883545, "noise_std_eff": 0.04786851048469544, "elapsed_sec": 709.1768815517426}
181
+ {"step": 6620, "epoch": 0, "loss": 3.844151258468628, "loss_en": 3.798180103302002, "loss_zh": 3.890122413635254, "noise_std_eff": 0.06638295710086824, "elapsed_sec": 712.0688843727112}
182
+ {"step": 6640, "epoch": 0, "loss": 3.8183140754699707, "loss_en": 3.809563159942627, "loss_zh": 3.8270647525787354, "noise_std_eff": 0.05797906816005707, "elapsed_sec": 714.9526464939117}
183
+ {"step": 6660, "epoch": 0, "loss": 4.246709823608398, "loss_en": 4.336408615112305, "loss_zh": 4.157011032104492, "noise_std_eff": 0.013057221621274949, "elapsed_sec": 717.834853887558}
184
+ {"step": 6680, "epoch": 0, "loss": 3.4187474250793457, "loss_en": 3.4457709789276123, "loss_zh": 3.3917236328125, "noise_std_eff": 0.025738593935966496, "elapsed_sec": 720.7248909473419}
185
+ {"step": 6700, "epoch": 0, "loss": 3.4114413261413574, "loss_en": 3.44327449798584, "loss_zh": 3.379607915878296, "noise_std_eff": 0.06234201014041901, "elapsed_sec": 723.6608238220215}
186
+ {"step": 6720, "epoch": 0, "loss": 3.9864964485168457, "loss_en": 3.975645065307617, "loss_zh": 3.997347831726074, "noise_std_eff": 0.012000607401132585, "elapsed_sec": 726.5745306015015}
187
+ {"step": 6740, "epoch": 0, "loss": 3.9362738132476807, "loss_en": 4.013734340667725, "loss_zh": 3.8588132858276367, "noise_std_eff": 0.02857667833566666, "elapsed_sec": 729.4810001850128}
188
+ {"step": 6760, "epoch": 0, "loss": 4.072355270385742, "loss_en": 4.034003257751465, "loss_zh": 4.110706806182861, "noise_std_eff": 0.03946688294410706, "elapsed_sec": 732.3823385238647}
189
+ {"step": 6780, "epoch": 0, "loss": 3.818265199661255, "loss_en": 3.8159921169281006, "loss_zh": 3.820538282394409, "noise_std_eff": 0.018399175703525544, "elapsed_sec": 735.2734482288361}
190
+ {"step": 6800, "epoch": 0, "loss": 3.8685832023620605, "loss_en": 3.8515450954437256, "loss_zh": 3.8856213092803955, "noise_std_eff": 0.02994681775569916, "elapsed_sec": 738.1989650726318}
191
+ {"step": 6820, "epoch": 0, "loss": 4.330073356628418, "loss_en": 4.376039505004883, "loss_zh": 4.284107208251953, "noise_std_eff": 0.03364638507366181, "elapsed_sec": 764.8791084289551}
192
+ {"step": 6840, "epoch": 0, "loss": 4.364051818847656, "loss_en": 4.472923278808594, "loss_zh": 4.2551798820495605, "noise_std_eff": 0.026322485804557802, "elapsed_sec": 767.7777717113495}
193
+ {"step": 6860, "epoch": 0, "loss": 3.8336262702941895, "loss_en": 3.8184642791748047, "loss_zh": 3.848788261413574, "noise_std_eff": 0.06356205224990845, "elapsed_sec": 770.7193183898926}
194
+ {"step": 6880, "epoch": 0, "loss": 3.9960453510284424, "loss_en": 3.9352827072143555, "loss_zh": 4.056807994842529, "noise_std_eff": 0.027153556644916536, "elapsed_sec": 773.5932686328888}
195
+ {"step": 6900, "epoch": 0, "loss": 4.018328666687012, "loss_en": 4.008445739746094, "loss_zh": 4.028212070465088, "noise_std_eff": 0.02443401634693146, "elapsed_sec": 776.4770534038544}
196
+ {"step": 6920, "epoch": 0, "loss": 3.5783591270446777, "loss_en": 3.531785726547241, "loss_zh": 3.6249327659606934, "noise_std_eff": 0.008146937862038612, "elapsed_sec": 779.3774099349976}
197
+ {"step": 6940, "epoch": 0, "loss": 3.82969331741333, "loss_en": 3.8265249729156494, "loss_zh": 3.8328616619110107, "noise_std_eff": 0.02909521281719208, "elapsed_sec": 782.2662854194641}
198
+ {"step": 6960, "epoch": 0, "loss": 3.877725839614868, "loss_en": 3.924959659576416, "loss_zh": 3.8304920196533203, "noise_std_eff": 0.039598945379257205, "elapsed_sec": 785.1902811527252}
199
+ {"step": 6980, "epoch": 0, "loss": 3.6051692962646484, "loss_en": 3.6234474182128906, "loss_zh": 3.586890935897827, "noise_std_eff": 0.02315905809402466, "elapsed_sec": 788.1031622886658}
200
+ {"step": 7000, "epoch": 0, "loss": 4.080613136291504, "loss_en": 4.041696071624756, "loss_zh": 4.119529724121094, "noise_std_eff": 0.006488134041428567, "elapsed_sec": 791.0024464130402}
201
+ {"step": 7020, "epoch": 0, "loss": 4.07177734375, "loss_en": 4.057391166687012, "loss_zh": 4.086163520812988, "noise_std_eff": 0.007407313957810403, "elapsed_sec": 793.8744170665741}
202
+ {"step": 7040, "epoch": 0, "loss": 4.082152366638184, "loss_en": 4.130748748779297, "loss_zh": 4.03355598449707, "noise_std_eff": 0.03743286192417145, "elapsed_sec": 796.7935018539429}
203
+ {"step": 7060, "epoch": 0, "loss": 3.8793020248413086, "loss_en": 3.876802682876587, "loss_zh": 3.8818013668060303, "noise_std_eff": 0.021685165464878083, "elapsed_sec": 799.6499280929565}
204
+ {"step": 7080, "epoch": 0, "loss": 4.430710315704346, "loss_en": 4.391374588012695, "loss_zh": 4.470046043395996, "noise_std_eff": 0.042128346562385566, "elapsed_sec": 802.5573332309723}
205
+ {"step": 7100, "epoch": 0, "loss": 4.0966997146606445, "loss_en": 4.112567901611328, "loss_zh": 4.080832004547119, "noise_std_eff": 0.004531881511211396, "elapsed_sec": 805.4340074062347}
206
+ {"step": 7120, "epoch": 0, "loss": 4.076416015625, "loss_en": 4.029940605163574, "loss_zh": 4.122891426086426, "noise_std_eff": 0.05978653609752656, "elapsed_sec": 808.3336918354034}
207
+ {"step": 7140, "epoch": 0, "loss": 4.440066337585449, "loss_en": 4.422963619232178, "loss_zh": 4.4571685791015625, "noise_std_eff": 0.04980743169784546, "elapsed_sec": 811.2253954410553}
208
+ {"step": 7160, "epoch": 0, "loss": 4.104240417480469, "loss_en": 4.078061103820801, "loss_zh": 4.130420207977295, "noise_std_eff": 0.036524509191513066, "elapsed_sec": 814.1304216384888}
209
+ {"step": 7180, "epoch": 0, "loss": 3.6985864639282227, "loss_en": 3.6902191638946533, "loss_zh": 3.706954002380371, "noise_std_eff": 0.06250985443592072, "elapsed_sec": 817.0014085769653}
210
+ {"step": 7200, "epoch": 0, "loss": 3.9039993286132812, "loss_en": 3.884816884994507, "loss_zh": 3.9231815338134766, "noise_std_eff": 0.03879602730274201, "elapsed_sec": 819.8930389881134}
211
+ {"step": 7220, "epoch": 0, "loss": 3.9880356788635254, "loss_en": 3.977970838546753, "loss_zh": 3.998100519180298, "noise_std_eff": 0.014388265311717989, "elapsed_sec": 846.6866085529327}
212
+ {"step": 7240, "epoch": 0, "loss": 3.8995232582092285, "loss_en": 3.8689637184143066, "loss_zh": 3.9300830364227295, "noise_std_eff": 0.044590877890586855, "elapsed_sec": 849.6167802810669}
213
+ {"step": 7260, "epoch": 0, "loss": 3.583409309387207, "loss_en": 3.6288580894470215, "loss_zh": 3.5379605293273926, "noise_std_eff": 0.010135293304920197, "elapsed_sec": 852.4821782112122}
214
+ {"step": 7280, "epoch": 0, "loss": 3.6406266689300537, "loss_en": 3.6157724857330322, "loss_zh": 3.665480852127075, "noise_std_eff": 0.05346029400825501, "elapsed_sec": 855.3622696399689}
215
+ {"step": 7300, "epoch": 0, "loss": 3.552309036254883, "loss_en": 3.538177013397217, "loss_zh": 3.566441297531128, "noise_std_eff": 0.063482803106308, "elapsed_sec": 858.2251954078674}
216
+ {"step": 7320, "epoch": 0, "loss": 3.792827606201172, "loss_en": 3.7562241554260254, "loss_zh": 3.8294312953948975, "noise_std_eff": 0.007087864056229592, "elapsed_sec": 861.1374163627625}
217
+ {"step": 7340, "epoch": 0, "loss": 4.523853302001953, "loss_en": 4.547335624694824, "loss_zh": 4.50037145614624, "noise_std_eff": 0.04674431920051575, "elapsed_sec": 864.0191600322723}
218
+ {"step": 7360, "epoch": 0, "loss": 3.6855688095092773, "loss_en": 3.590322971343994, "loss_zh": 3.7808148860931396, "noise_std_eff": 0.04749645173549653, "elapsed_sec": 866.9215741157532}
219
+ {"step": 7380, "epoch": 0, "loss": 3.8639278411865234, "loss_en": 3.92759108543396, "loss_zh": 3.800264358520508, "noise_std_eff": 0.01640715539455414, "elapsed_sec": 869.8338711261749}
220
+ {"step": 7400, "epoch": 0, "loss": 3.497527599334717, "loss_en": 3.498539447784424, "loss_zh": 3.4965157508850098, "noise_std_eff": 0.01042039453983307, "elapsed_sec": 872.7369840145111}
221
+ {"step": 7420, "epoch": 0, "loss": 3.9212489128112793, "loss_en": 3.9070398807525635, "loss_zh": 3.935458183288574, "noise_std_eff": 0.024725428223609926, "elapsed_sec": 875.6208071708679}
222
+ {"step": 7440, "epoch": 0, "loss": 3.7914085388183594, "loss_en": 3.761810064315796, "loss_zh": 3.8210067749023438, "noise_std_eff": 0.016371448636054994, "elapsed_sec": 878.5241119861603}
223
+ {"step": 7460, "epoch": 0, "loss": 3.6157941818237305, "loss_en": 3.6661298274993896, "loss_zh": 3.5654585361480713, "noise_std_eff": 0.015103778392076495, "elapsed_sec": 881.3890326023102}
224
+ {"step": 7480, "epoch": 0, "loss": 4.174813270568848, "loss_en": 4.161431312561035, "loss_zh": 4.188195705413818, "noise_std_eff": 0.014104438722133638, "elapsed_sec": 884.268972158432}
225
+ {"step": 7500, "epoch": 1, "loss": 5.129813194274902, "loss_en": 5.115854740142822, "loss_zh": 5.143771171569824, "noise_std_eff": 0.02018496423959732, "elapsed_sec": 887.2208771705627}
226
+ {"step": 7520, "epoch": 1, "loss": 4.814685821533203, "loss_en": 4.70751428604126, "loss_zh": 4.921856880187988, "noise_std_eff": 0.0051933593302965165, "elapsed_sec": 890.2424519062042}
227
+ {"step": 7540, "epoch": 1, "loss": 4.986798286437988, "loss_en": 5.088174819946289, "loss_zh": 4.885422229766846, "noise_std_eff": 0.029155552983284, "elapsed_sec": 893.2230224609375}
228
+ {"step": 7560, "epoch": 1, "loss": 4.921220302581787, "loss_en": 5.067561626434326, "loss_zh": 4.774878978729248, "noise_std_eff": 0.05968067586421967, "elapsed_sec": 896.2206981182098}
229
+ {"step": 7580, "epoch": 1, "loss": 5.036901473999023, "loss_en": 5.101630687713623, "loss_zh": 4.972171783447266, "noise_std_eff": 0.028919140696525576, "elapsed_sec": 899.2097716331482}
230
+ {"step": 7600, "epoch": 1, "loss": 5.658048629760742, "loss_en": 5.805049419403076, "loss_zh": 5.511047840118408, "noise_std_eff": 0.03056011408567429, "elapsed_sec": 902.2413969039917}
231
+ {"step": 7620, "epoch": 1, "loss": 4.185967445373535, "loss_en": 4.157230854034424, "loss_zh": 4.214704513549805, "noise_std_eff": 0.001233662534505129, "elapsed_sec": 929.1349158287048}
232
+ {"step": 7640, "epoch": 1, "loss": 3.694725513458252, "loss_en": 3.677361011505127, "loss_zh": 3.712090253829956, "noise_std_eff": 0.038857710957527165, "elapsed_sec": 932.0814046859741}
233
+ {"step": 7660, "epoch": 1, "loss": 4.3311309814453125, "loss_en": 4.355731964111328, "loss_zh": 4.306530475616455, "noise_std_eff": 0.011389413923025133, "elapsed_sec": 934.957056760788}
234
+ {"step": 7680, "epoch": 1, "loss": 3.6648881435394287, "loss_en": 3.6372485160827637, "loss_zh": 3.6925277709960938, "noise_std_eff": 0.04757837116718293, "elapsed_sec": 937.8901169300079}
235
+ {"step": 7700, "epoch": 1, "loss": 3.381178855895996, "loss_en": 3.4533729553222656, "loss_zh": 3.3089849948883057, "noise_std_eff": 0.03079859793186188, "elapsed_sec": 940.7585244178772}
236
+ {"step": 7720, "epoch": 1, "loss": 3.591848850250244, "loss_en": 3.6065738201141357, "loss_zh": 3.5771236419677734, "noise_std_eff": 0.033847149014472964, "elapsed_sec": 943.7058787345886}
237
+ {"step": 7740, "epoch": 1, "loss": 3.9747705459594727, "loss_en": 3.9966118335723877, "loss_zh": 3.9529294967651367, "noise_std_eff": 0.010573345720767976, "elapsed_sec": 946.6012692451477}
238
+ {"step": 7760, "epoch": 1, "loss": 3.8743207454681396, "loss_en": 3.8306922912597656, "loss_zh": 3.9179491996765137, "noise_std_eff": 0.017867617309093475, "elapsed_sec": 949.5415480136871}
239
+ {"step": 7780, "epoch": 1, "loss": 3.8267722129821777, "loss_en": 3.806365966796875, "loss_zh": 3.8471786975860596, "noise_std_eff": 0.03747140586376191, "elapsed_sec": 952.4099853038788}
240
+ {"step": 7800, "epoch": 1, "loss": 4.5251970291137695, "loss_en": 4.5354132652282715, "loss_zh": 4.514980316162109, "noise_std_eff": 0.05915379881858826, "elapsed_sec": 955.3536787033081}
241
+ {"step": 7820, "epoch": 1, "loss": 3.4222850799560547, "loss_en": 3.453307867050171, "loss_zh": 3.3912625312805176, "noise_std_eff": 0.05846734702587128, "elapsed_sec": 958.2656893730164}
242
+ {"step": 7840, "epoch": 1, "loss": 4.14748477935791, "loss_en": 4.179890155792236, "loss_zh": 4.115078926086426, "noise_std_eff": 0.05095004796981812, "elapsed_sec": 961.2187922000885}
243
+ {"step": 7860, "epoch": 1, "loss": 3.5616402626037598, "loss_en": 3.544250011444092, "loss_zh": 3.5790305137634277, "noise_std_eff": 0.059249582886695865, "elapsed_sec": 964.1296944618225}
244
+ {"step": 7880, "epoch": 1, "loss": 3.9447898864746094, "loss_en": 3.959148645401001, "loss_zh": 3.9304308891296387, "noise_std_eff": 0.011255464553833009, "elapsed_sec": 967.0498433113098}
245
+ {"step": 7900, "epoch": 1, "loss": 4.0155744552612305, "loss_en": 3.9772214889526367, "loss_zh": 4.053926944732666, "noise_std_eff": 0.030596261024475102, "elapsed_sec": 969.9377348423004}
246
+ {"step": 7920, "epoch": 1, "loss": 3.7736401557922363, "loss_en": 3.7432620525360107, "loss_zh": 3.804018020629883, "noise_std_eff": 0.029460149407386782, "elapsed_sec": 972.8778023719788}
247
+ {"step": 7940, "epoch": 1, "loss": 3.7235851287841797, "loss_en": 3.6845755577087402, "loss_zh": 3.76259446144104, "noise_std_eff": 0.009774435162544252, "elapsed_sec": 975.7684874534607}
248
+ {"step": 7960, "epoch": 1, "loss": 4.091702461242676, "loss_en": 4.099332332611084, "loss_zh": 4.084073066711426, "noise_std_eff": 0.03188426196575165, "elapsed_sec": 978.7009103298187}
249
+ {"step": 7980, "epoch": 1, "loss": 3.8505611419677734, "loss_en": 3.9061055183410645, "loss_zh": 3.7950170040130615, "noise_std_eff": 0.03802155196666718, "elapsed_sec": 981.5895192623138}
250
+ {"step": 8000, "epoch": 1, "loss": 3.7536020278930664, "loss_en": 3.781914472579956, "loss_zh": 3.7252893447875977, "noise_std_eff": 0.021572798490524295, "elapsed_sec": 984.5263411998749}
251
+ {"step": 8020, "epoch": 1, "loss": 3.5879693031311035, "loss_en": 3.587350606918335, "loss_zh": 3.588587760925293, "noise_std_eff": 0.025135976672172548, "elapsed_sec": 1011.2808740139008}
252
+ {"step": 8040, "epoch": 1, "loss": 3.5817489624023438, "loss_en": 3.567861557006836, "loss_zh": 3.5956363677978516, "noise_std_eff": 0.012406166791915895, "elapsed_sec": 1014.2416923046112}
253
+ {"step": 8060, "epoch": 1, "loss": 4.561095237731934, "loss_en": 4.555644512176514, "loss_zh": 4.566545486450195, "noise_std_eff": 0.05056118309497834, "elapsed_sec": 1017.0981168746948}
254
+ {"step": 8080, "epoch": 1, "loss": 4.187359809875488, "loss_en": 4.17650032043457, "loss_zh": 4.1982197761535645, "noise_std_eff": 0.0018499180860817434, "elapsed_sec": 1020.0775961875916}
255
+ {"step": 8100, "epoch": 1, "loss": 3.7850680351257324, "loss_en": 3.793292284011841, "loss_zh": 3.776843786239624, "noise_std_eff": 0.013164137452840806, "elapsed_sec": 1023.0202898979187}
256
+ {"step": 8120, "epoch": 1, "loss": 3.8470935821533203, "loss_en": 3.7990119457244873, "loss_zh": 3.895174980163574, "noise_std_eff": 0.06208081007003785, "elapsed_sec": 1026.0328323841095}
257
+ {"step": 8140, "epoch": 1, "loss": 3.565396785736084, "loss_en": 3.577052354812622, "loss_zh": 3.553741455078125, "noise_std_eff": 0.043759183287620546, "elapsed_sec": 1028.8743948936462}
258
+ {"step": 8160, "epoch": 1, "loss": 3.5945701599121094, "loss_en": 3.6250438690185547, "loss_zh": 3.564096450805664, "noise_std_eff": 0.034432970583438874, "elapsed_sec": 1031.714823961258}
259
+ {"step": 8180, "epoch": 1, "loss": 3.7932300567626953, "loss_en": 3.82331919670105, "loss_zh": 3.76314115524292, "noise_std_eff": 0.016065283417701724, "elapsed_sec": 1034.6014227867126}
260
+ {"step": 8200, "epoch": 1, "loss": 4.000020980834961, "loss_en": 4.001836776733398, "loss_zh": 3.9982047080993652, "noise_std_eff": 0.033650511503219606, "elapsed_sec": 1037.528370141983}
261
+ {"step": 8220, "epoch": 1, "loss": 3.625856876373291, "loss_en": 3.6519105434417725, "loss_zh": 3.5998029708862305, "noise_std_eff": 0.028454521000385286, "elapsed_sec": 1040.4124627113342}
262
+ {"step": 8240, "epoch": 1, "loss": 3.8423776626586914, "loss_en": 3.883971929550171, "loss_zh": 3.800783395767212, "noise_std_eff": 0.06179075419902802, "elapsed_sec": 1043.3293702602386}
263
+ {"step": 8260, "epoch": 1, "loss": 3.6242780685424805, "loss_en": 3.638453722000122, "loss_zh": 3.610102653503418, "noise_std_eff": 0.0106612566113472, "elapsed_sec": 1046.2485740184784}
264
+ {"step": 8280, "epoch": 1, "loss": 3.423201084136963, "loss_en": 3.396090507507324, "loss_zh": 3.4503114223480225, "noise_std_eff": 0.022111570835113527, "elapsed_sec": 1049.181539773941}
265
+ {"step": 8300, "epoch": 1, "loss": 4.081489562988281, "loss_en": 4.0426177978515625, "loss_zh": 4.120360851287842, "noise_std_eff": 0.048925556540489204, "elapsed_sec": 1052.060827255249}
266
+ {"step": 8320, "epoch": 1, "loss": 3.7114980220794678, "loss_en": 3.7117135524749756, "loss_zh": 3.71128249168396, "noise_std_eff": 0.06268569707870485, "elapsed_sec": 1055.0512907505035}
267
+ {"step": 8340, "epoch": 1, "loss": 3.734038829803467, "loss_en": 3.749236583709717, "loss_zh": 3.718841314315796, "noise_std_eff": 0.04446393489837647, "elapsed_sec": 1057.913342475891}
268
+ {"step": 8360, "epoch": 1, "loss": 3.3448269367218018, "loss_en": 3.3577945232391357, "loss_zh": 3.3318593502044678, "noise_std_eff": 0.021165871620178224, "elapsed_sec": 1060.8490612506866}
269
+ {"step": 8380, "epoch": 1, "loss": 3.798072338104248, "loss_en": 3.828296422958374, "loss_zh": 3.767848491668701, "noise_std_eff": 0.009831564724445343, "elapsed_sec": 1063.7367269992828}
270
+ {"step": 8400, "epoch": 1, "loss": 3.759174346923828, "loss_en": 3.734694480895996, "loss_zh": 3.78365421295166, "noise_std_eff": 0.0581229841709137, "elapsed_sec": 1066.666030406952}
271
+ {"step": 8420, "epoch": 1, "loss": 3.7231597900390625, "loss_en": 3.730506658554077, "loss_zh": 3.715813159942627, "noise_std_eff": 0.04577652752399445, "elapsed_sec": 1093.174322605133}
272
+ {"step": 8440, "epoch": 1, "loss": 3.356609344482422, "loss_en": 3.4017748832702637, "loss_zh": 3.311443567276001, "noise_std_eff": 0.000601336881518364, "elapsed_sec": 1096.0860028266907}
273
+ {"step": 8460, "epoch": 1, "loss": 3.1302456855773926, "loss_en": 3.118943929672241, "loss_zh": 3.141547679901123, "noise_std_eff": 0.006707487776875497, "elapsed_sec": 1098.9387130737305}
274
+ {"step": 8480, "epoch": 1, "loss": 3.6860456466674805, "loss_en": 3.671452522277832, "loss_zh": 3.700638771057129, "noise_std_eff": 0.023059431314468387, "elapsed_sec": 1101.8818600177765}
275
+ {"step": 8500, "epoch": 1, "loss": 3.541658878326416, "loss_en": 3.500821828842163, "loss_zh": 3.582496166229248, "noise_std_eff": 0.0334031155705452, "elapsed_sec": 1104.7666504383087}
276
+ {"step": 8520, "epoch": 1, "loss": 4.147528171539307, "loss_en": 4.151949882507324, "loss_zh": 4.143106460571289, "noise_std_eff": 0.014473711401224137, "elapsed_sec": 1107.701189994812}
277
+ {"step": 8540, "epoch": 1, "loss": 3.635350227355957, "loss_en": 3.66399884223938, "loss_zh": 3.6067018508911133, "noise_std_eff": 0.03444711893796921, "elapsed_sec": 1110.5930697917938}
278
+ {"step": 8560, "epoch": 1, "loss": 4.077966690063477, "loss_en": 4.177407741546631, "loss_zh": 3.9785258769989014, "noise_std_eff": 0.023287125527858735, "elapsed_sec": 1113.513948917389}
279
+ {"step": 8580, "epoch": 1, "loss": 4.083631992340088, "loss_en": 4.0335187911987305, "loss_zh": 4.133745193481445, "noise_std_eff": 0.025275757908821108, "elapsed_sec": 1116.3853628635406}
280
+ {"step": 8600, "epoch": 1, "loss": 3.583705425262451, "loss_en": 3.643807888031006, "loss_zh": 3.5236032009124756, "noise_std_eff": 0.0551070773601532, "elapsed_sec": 1119.326508283615}
281
+ {"step": 8620, "epoch": 1, "loss": 3.999880790710449, "loss_en": 3.943011522293091, "loss_zh": 4.056750297546387, "noise_std_eff": 0.013901255875825884, "elapsed_sec": 1122.6103382110596}
282
+ {"step": 8640, "epoch": 1, "loss": 3.611473560333252, "loss_en": 3.6437551975250244, "loss_zh": 3.5791921615600586, "noise_std_eff": 0.028143931031227115, "elapsed_sec": 1125.669893026352}
283
+ {"step": 8660, "epoch": 1, "loss": 3.4747772216796875, "loss_en": 3.484264373779297, "loss_zh": 3.465289831161499, "noise_std_eff": 0.03771967589855194, "elapsed_sec": 1128.57119846344}
284
+ {"step": 8680, "epoch": 1, "loss": 3.454908609390259, "loss_en": 3.397165060043335, "loss_zh": 3.5126521587371826, "noise_std_eff": 0.06995391666889192, "elapsed_sec": 1131.8626546859741}
285
+ {"step": 8700, "epoch": 1, "loss": 2.961801052093506, "loss_en": 2.9240779876708984, "loss_zh": 2.9995243549346924, "noise_std_eff": 0.06601174533367157, "elapsed_sec": 1135.0434596538544}
286
+ {"step": 8720, "epoch": 1, "loss": 3.78658390045166, "loss_en": 3.8028461933135986, "loss_zh": 3.7703213691711426, "noise_std_eff": 0.0012375095486640932, "elapsed_sec": 1138.2669496536255}
287
+ {"step": 8740, "epoch": 1, "loss": 3.7214062213897705, "loss_en": 3.7046854496002197, "loss_zh": 3.7381269931793213, "noise_std_eff": 0.009702283143997193, "elapsed_sec": 1141.1793174743652}
288
+ {"step": 8760, "epoch": 1, "loss": 4.049659252166748, "loss_en": 4.045126914978027, "loss_zh": 4.054191589355469, "noise_std_eff": 0.0036467151716351514, "elapsed_sec": 1144.4119884967804}
289
+ {"step": 8780, "epoch": 1, "loss": 3.875173807144165, "loss_en": 3.8734843730926514, "loss_zh": 3.8768632411956787, "noise_std_eff": 0.029297497570514683, "elapsed_sec": 1147.6020879745483}
290
+ {"step": 8800, "epoch": 1, "loss": 3.6668782234191895, "loss_en": 3.687248706817627, "loss_zh": 3.646507740020752, "noise_std_eff": 0.041686843633651736, "elapsed_sec": 1150.8225252628326}
291
+ {"step": 8820, "epoch": 1, "loss": 3.240551471710205, "loss_en": 3.2073960304260254, "loss_zh": 3.2737069129943848, "noise_std_eff": 0.05530880093574524, "elapsed_sec": 1177.7470879554749}
292
+ {"step": 8840, "epoch": 1, "loss": 4.077431678771973, "loss_en": 4.070234298706055, "loss_zh": 4.084628582000732, "noise_std_eff": 0.0087447626888752, "elapsed_sec": 1180.6854507923126}
293
+ {"step": 8860, "epoch": 1, "loss": 3.5101821422576904, "loss_en": 3.482571840286255, "loss_zh": 3.537792444229126, "noise_std_eff": 0.06808803200721741, "elapsed_sec": 1183.5779023170471}
294
+ {"step": 8880, "epoch": 1, "loss": 4.0372419357299805, "loss_en": 4.033965110778809, "loss_zh": 4.040518760681152, "noise_std_eff": 0.020111118257045747, "elapsed_sec": 1186.51669383049}
295
+ {"step": 8900, "epoch": 1, "loss": 3.619503974914551, "loss_en": 3.640195608139038, "loss_zh": 3.5988121032714844, "noise_std_eff": 0.02753518044948578, "elapsed_sec": 1189.4173765182495}
296
+ {"step": 8920, "epoch": 1, "loss": 4.6316375732421875, "loss_en": 4.676180839538574, "loss_zh": 4.587094783782959, "noise_std_eff": 0.06540543973445893, "elapsed_sec": 1192.3334777355194}
297
+ {"step": 8940, "epoch": 1, "loss": 3.84000301361084, "loss_en": 3.842266321182251, "loss_zh": 3.8377397060394287, "noise_std_eff": 0.03878713190555573, "elapsed_sec": 1195.1965119838715}
298
+ {"step": 8960, "epoch": 1, "loss": 3.7987751960754395, "loss_en": 3.775074005126953, "loss_zh": 3.822476625442505, "noise_std_eff": 0.030925447046756746, "elapsed_sec": 1198.1288845539093}
299
+ {"step": 8980, "epoch": 1, "loss": 3.8345720767974854, "loss_en": 3.771024227142334, "loss_zh": 3.8981199264526367, "noise_std_eff": 0.012378196567296983, "elapsed_sec": 1201.0567710399628}
300
+ {"step": 9000, "epoch": 1, "loss": 3.7641000747680664, "loss_en": 3.8023412227630615, "loss_zh": 3.725858688354492, "noise_std_eff": 0.041455534100532536, "elapsed_sec": 1203.942349433899}
301
+ {"step": 9020, "epoch": 1, "loss": 3.6514296531677246, "loss_en": 3.665299892425537, "loss_zh": 3.637559652328491, "noise_std_eff": 0.030635272264480595, "elapsed_sec": 1206.8180627822876}
302
+ {"step": 9040, "epoch": 1, "loss": 3.4289467334747314, "loss_en": 3.4201858043670654, "loss_zh": 3.4377076625823975, "noise_std_eff": 0.05456402838230134, "elapsed_sec": 1209.7298471927643}
303
+ {"step": 9060, "epoch": 1, "loss": 3.5769925117492676, "loss_en": 3.567678689956665, "loss_zh": 3.58630633354187, "noise_std_eff": 0.03289584219455719, "elapsed_sec": 1212.6209590435028}
304
+ {"step": 9080, "epoch": 1, "loss": 3.2683749198913574, "loss_en": 3.3148367404937744, "loss_zh": 3.2219130992889404, "noise_std_eff": 0.02596463590860367, "elapsed_sec": 1215.5339069366455}
305
+ {"step": 9100, "epoch": 1, "loss": 3.788963794708252, "loss_en": 3.868879795074463, "loss_zh": 3.709047555923462, "noise_std_eff": 0.05496493875980378, "elapsed_sec": 1218.4012072086334}
306
+ {"step": 9120, "epoch": 1, "loss": 3.708611488342285, "loss_en": 3.6796469688415527, "loss_zh": 3.7375760078430176, "noise_std_eff": 0.06048028111457825, "elapsed_sec": 1221.321902513504}
307
+ {"step": 9140, "epoch": 1, "loss": 3.894097328186035, "loss_en": 3.910144090652466, "loss_zh": 3.8780505657196045, "noise_std_eff": 0.02656159967184067, "elapsed_sec": 1224.2013154029846}
308
+ {"step": 9160, "epoch": 1, "loss": 3.581012487411499, "loss_en": 3.557344675064087, "loss_zh": 3.604680299758911, "noise_std_eff": 0.0084246613830328, "elapsed_sec": 1227.125548839569}
309
+ {"step": 9180, "epoch": 1, "loss": 3.823411464691162, "loss_en": 3.8672261238098145, "loss_zh": 3.779597043991089, "noise_std_eff": 0.036081283092498785, "elapsed_sec": 1230.013823747635}
310
+ {"step": 9200, "epoch": 1, "loss": 3.6851165294647217, "loss_en": 3.7224040031433105, "loss_zh": 3.647829055786133, "noise_std_eff": 0.012849441915750505, "elapsed_sec": 1232.9732694625854}
311
+ {"step": 9220, "epoch": 1, "loss": 3.4548704624176025, "loss_en": 3.4330554008483887, "loss_zh": 3.4766855239868164, "noise_std_eff": 0.0020269221998751165, "elapsed_sec": 1259.6857998371124}
312
+ {"step": 9240, "epoch": 1, "loss": 3.555129051208496, "loss_en": 3.5863256454467773, "loss_zh": 3.5239322185516357, "noise_std_eff": 0.0610824728012085, "elapsed_sec": 1262.628357410431}
313
+ {"step": 9260, "epoch": 1, "loss": 3.1628899574279785, "loss_en": 3.14251708984375, "loss_zh": 3.183263063430786, "noise_std_eff": 0.013118157386779787, "elapsed_sec": 1265.6105003356934}
314
+ {"step": 9280, "epoch": 1, "loss": 3.9014313220977783, "loss_en": 3.821946144104004, "loss_zh": 3.9809165000915527, "noise_std_eff": 0.007099810466170311, "elapsed_sec": 1268.5490114688873}
315
+ {"step": 9300, "epoch": 1, "loss": 3.904665946960449, "loss_en": 3.9004414081573486, "loss_zh": 3.90889048576355, "noise_std_eff": 0.0007042478676885367, "elapsed_sec": 1271.4811873435974}
316
+ {"step": 9320, "epoch": 1, "loss": 3.0825531482696533, "loss_en": 3.077857732772827, "loss_zh": 3.0872485637664795, "noise_std_eff": 0.03519773483276367, "elapsed_sec": 1274.4110448360443}
317
+ {"step": 9340, "epoch": 1, "loss": 3.6327900886535645, "loss_en": 3.64884614944458, "loss_zh": 3.616734266281128, "noise_std_eff": 0.035786437392234804, "elapsed_sec": 1277.329327583313}
318
+ {"step": 9360, "epoch": 1, "loss": 3.2979400157928467, "loss_en": 3.301621913909912, "loss_zh": 3.2942581176757812, "noise_std_eff": 0.05736696302890778, "elapsed_sec": 1280.268695116043}
319
+ {"step": 9380, "epoch": 1, "loss": 4.046286582946777, "loss_en": 4.020415782928467, "loss_zh": 4.07215690612793, "noise_std_eff": 0.004403336346149445, "elapsed_sec": 1283.1489343643188}
320
+ {"step": 9400, "epoch": 1, "loss": 3.9452381134033203, "loss_en": 3.9457056522369385, "loss_zh": 3.9447708129882812, "noise_std_eff": 0.01329520270228386, "elapsed_sec": 1286.1233878135681}
321
+ {"step": 9420, "epoch": 1, "loss": 3.75687313079834, "loss_en": 3.7285714149475098, "loss_zh": 3.785175085067749, "noise_std_eff": 0.040245259404182436, "elapsed_sec": 1289.0399236679077}
322
+ {"step": 9440, "epoch": 1, "loss": 3.168163299560547, "loss_en": 3.2281110286712646, "loss_zh": 3.10821533203125, "noise_std_eff": 0.03616419136524201, "elapsed_sec": 1291.9730062484741}
323
+ {"step": 9460, "epoch": 1, "loss": 3.3307576179504395, "loss_en": 3.323993682861328, "loss_zh": 3.337521553039551, "noise_std_eff": 0.03398439347743988, "elapsed_sec": 1294.8879997730255}
324
+ {"step": 9480, "epoch": 1, "loss": 3.524385929107666, "loss_en": 3.5201289653778076, "loss_zh": 3.5286431312561035, "noise_std_eff": 0.049065613150596626, "elapsed_sec": 1297.817039489746}
325
+ {"step": 9500, "epoch": 1, "loss": 3.1903181076049805, "loss_en": 3.239450454711914, "loss_zh": 3.141185998916626, "noise_std_eff": 0.014434704333543779, "elapsed_sec": 1300.7172451019287}
326
+ {"step": 9520, "epoch": 1, "loss": 3.6575639247894287, "loss_en": 3.6079251766204834, "loss_zh": 3.707202672958374, "noise_std_eff": 0.04703687012195588, "elapsed_sec": 1303.6577208042145}
327
+ {"step": 9540, "epoch": 1, "loss": 3.98217511177063, "loss_en": 3.9475979804992676, "loss_zh": 4.016752243041992, "noise_std_eff": 0.026240870952606205, "elapsed_sec": 1306.564114332199}
328
+ {"step": 9560, "epoch": 1, "loss": 3.4815452098846436, "loss_en": 3.461238145828247, "loss_zh": 3.50185227394104, "noise_std_eff": 0.04437155961990357, "elapsed_sec": 1309.4946048259735}
329
+ {"step": 9580, "epoch": 1, "loss": 3.379286289215088, "loss_en": 3.348485231399536, "loss_zh": 3.4100875854492188, "noise_std_eff": 0.057830441594123844, "elapsed_sec": 1312.3525187969208}
330
+ {"step": 9600, "epoch": 1, "loss": 4.233738899230957, "loss_en": 4.231003761291504, "loss_zh": 4.236474514007568, "noise_std_eff": 0.02968968361616135, "elapsed_sec": 1315.273057460785}
331
+ {"step": 9620, "epoch": 1, "loss": 3.752070426940918, "loss_en": 3.7487077713012695, "loss_zh": 3.7554328441619873, "noise_std_eff": 0.02290202409029007, "elapsed_sec": 1342.0406517982483}
332
+ {"step": 9640, "epoch": 1, "loss": 3.931684970855713, "loss_en": 3.9652130603790283, "loss_zh": 3.8981571197509766, "noise_std_eff": 0.0030699232965707784, "elapsed_sec": 1345.013055562973}
333
+ {"step": 9660, "epoch": 1, "loss": 3.7164294719696045, "loss_en": 3.6447298526763916, "loss_zh": 3.7881290912628174, "noise_std_eff": 0.00041528899222612384, "elapsed_sec": 1347.8973655700684}
334
+ {"step": 9680, "epoch": 1, "loss": 3.4084959030151367, "loss_en": 3.389052152633667, "loss_zh": 3.4279398918151855, "noise_std_eff": 0.020475001335144044, "elapsed_sec": 1350.853880405426}
335
+ {"step": 9700, "epoch": 1, "loss": 2.984304428100586, "loss_en": 2.974724769592285, "loss_zh": 2.9938840866088867, "noise_std_eff": 0.02673049330711365, "elapsed_sec": 1353.7369771003723}
336
+ {"step": 9720, "epoch": 1, "loss": 3.410398006439209, "loss_en": 3.4042084217071533, "loss_zh": 3.4165873527526855, "noise_std_eff": 0.038343154788017274, "elapsed_sec": 1356.6733512878418}
337
+ {"step": 9740, "epoch": 1, "loss": 3.5642247200012207, "loss_en": 3.5423336029052734, "loss_zh": 3.586115598678589, "noise_std_eff": 0.04905118942260742, "elapsed_sec": 1359.5615637302399}
338
+ {"step": 9760, "epoch": 1, "loss": 3.1441447734832764, "loss_en": 3.1255297660827637, "loss_zh": 3.162759780883789, "noise_std_eff": 0.01102402463555336, "elapsed_sec": 1362.4951610565186}
339
+ {"step": 9780, "epoch": 1, "loss": 3.205967426300049, "loss_en": 3.156578779220581, "loss_zh": 3.2553563117980957, "noise_std_eff": 0.027630415856838227, "elapsed_sec": 1365.4051823616028}
340
+ {"step": 9800, "epoch": 1, "loss": 3.5806899070739746, "loss_en": 3.5488903522491455, "loss_zh": 3.612489700317383, "noise_std_eff": 0.06867571651935578, "elapsed_sec": 1368.330642938614}
341
+ {"step": 9820, "epoch": 1, "loss": 3.986466407775879, "loss_en": 4.006691932678223, "loss_zh": 3.9662411212921143, "noise_std_eff": 0.017726064920425416, "elapsed_sec": 1371.2033836841583}
342
+ {"step": 9840, "epoch": 1, "loss": 3.856687307357788, "loss_en": 3.8894288539886475, "loss_zh": 3.8239457607269287, "noise_std_eff": 0.025735894441604616, "elapsed_sec": 1374.1239902973175}
343
+ {"step": 9860, "epoch": 1, "loss": 4.146584987640381, "loss_en": 4.102758884429932, "loss_zh": 4.19041109085083, "noise_std_eff": 0.05695323109626771, "elapsed_sec": 1377.044190645218}
344
+ {"step": 9880, "epoch": 1, "loss": 3.9981629848480225, "loss_en": 3.977396011352539, "loss_zh": 4.018929958343506, "noise_std_eff": 0.008170345649123193, "elapsed_sec": 1379.9795079231262}
345
+ {"step": 9900, "epoch": 1, "loss": 3.317429542541504, "loss_en": 3.311742067337036, "loss_zh": 3.3231170177459717, "noise_std_eff": 0.01905659854412079, "elapsed_sec": 1382.8809390068054}
346
+ {"step": 9920, "epoch": 1, "loss": 3.899156332015991, "loss_en": 3.9162404537200928, "loss_zh": 3.8820722103118896, "noise_std_eff": 0.04576502442359925, "elapsed_sec": 1385.7805953025818}
347
+ {"step": 9940, "epoch": 1, "loss": 3.7226366996765137, "loss_en": 3.6947641372680664, "loss_zh": 3.75050950050354, "noise_std_eff": 0.021511932611465456, "elapsed_sec": 1388.6444506645203}
348
+ {"step": 9960, "epoch": 1, "loss": 3.3194808959960938, "loss_en": 3.3262131214141846, "loss_zh": 3.312748432159424, "noise_std_eff": 0.004002941772341728, "elapsed_sec": 1391.604867696762}
349
+ {"step": 9980, "epoch": 1, "loss": 3.604569911956787, "loss_en": 3.579087972640991, "loss_zh": 3.630052089691162, "noise_std_eff": 0.006717148274183274, "elapsed_sec": 1394.5005896091461}
350
+ {"step": 10000, "epoch": 1, "loss": 3.001830577850342, "loss_en": 2.976077079772949, "loss_zh": 3.0275843143463135, "noise_std_eff": 0.03288250327110291, "elapsed_sec": 1397.43310379982}
351
+ {"step": 10020, "epoch": 1, "loss": 4.098254680633545, "loss_en": 4.085581302642822, "loss_zh": 4.110928058624268, "noise_std_eff": 0.04937706053256989, "elapsed_sec": 1424.2938342094421}
352
+ {"step": 10040, "epoch": 1, "loss": 3.246692657470703, "loss_en": 3.2024941444396973, "loss_zh": 3.290891408920288, "noise_std_eff": 0.025827074348926546, "elapsed_sec": 1427.260530948639}
353
+ {"step": 10060, "epoch": 1, "loss": 3.803104877471924, "loss_en": 3.8344132900238037, "loss_zh": 3.771796464920044, "noise_std_eff": 0.060036708712577824, "elapsed_sec": 1430.1432628631592}
354
+ {"step": 10080, "epoch": 1, "loss": 3.308727264404297, "loss_en": 3.2837603092193604, "loss_zh": 3.3336942195892334, "noise_std_eff": 0.02597137421369553, "elapsed_sec": 1433.0807344913483}
355
+ {"step": 10100, "epoch": 1, "loss": 3.3795855045318604, "loss_en": 3.4067599773406982, "loss_zh": 3.3524110317230225, "noise_std_eff": 0.021640098094940188, "elapsed_sec": 1435.977212190628}
356
+ {"step": 10120, "epoch": 1, "loss": 3.3032498359680176, "loss_en": 3.3014678955078125, "loss_zh": 3.3050315380096436, "noise_std_eff": 0.04348222434520722, "elapsed_sec": 1438.9449141025543}
357
+ {"step": 10140, "epoch": 1, "loss": 3.9683241844177246, "loss_en": 4.061522006988525, "loss_zh": 3.875126361846924, "noise_std_eff": 0.016298048049211502, "elapsed_sec": 1441.8644971847534}
358
+ {"step": 10160, "epoch": 1, "loss": 3.6534178256988525, "loss_en": 3.5648093223571777, "loss_zh": 3.7420263290405273, "noise_std_eff": 0.01504371151328087, "elapsed_sec": 1444.850240945816}
359
+ {"step": 10180, "epoch": 1, "loss": 3.4621362686157227, "loss_en": 3.477999687194824, "loss_zh": 3.446272611618042, "noise_std_eff": 0.014577406197786332, "elapsed_sec": 1447.757889509201}
360
+ {"step": 10200, "epoch": 1, "loss": 3.6502091884613037, "loss_en": 3.6341946125030518, "loss_zh": 3.6662237644195557, "noise_std_eff": 0.003620693422853947, "elapsed_sec": 1450.6993427276611}
361
+ {"step": 10220, "epoch": 1, "loss": 3.7391860485076904, "loss_en": 3.738046884536743, "loss_zh": 3.7403252124786377, "noise_std_eff": 0.002976285107433796, "elapsed_sec": 1453.5734612941742}
362
+ {"step": 10240, "epoch": 1, "loss": 3.137702226638794, "loss_en": 3.1201517581939697, "loss_zh": 3.155252695083618, "noise_std_eff": 0.029999580979347233, "elapsed_sec": 1456.5060865879059}
363
+ {"step": 10260, "epoch": 1, "loss": 3.6223902702331543, "loss_en": 3.6065914630889893, "loss_zh": 3.6381890773773193, "noise_std_eff": 0.04303240180015564, "elapsed_sec": 1459.3978152275085}
364
+ {"step": 10280, "epoch": 1, "loss": 4.177611351013184, "loss_en": 4.165832042694092, "loss_zh": 4.189390659332275, "noise_std_eff": 0.02512365579605103, "elapsed_sec": 1462.3378479480743}
365
+ {"step": 10300, "epoch": 1, "loss": 4.121638298034668, "loss_en": 4.165673732757568, "loss_zh": 4.077603340148926, "noise_std_eff": 0.06149681389331818, "elapsed_sec": 1465.23100066185}
366
+ {"step": 10320, "epoch": 1, "loss": 3.403554677963257, "loss_en": 3.413905143737793, "loss_zh": 3.3932042121887207, "noise_std_eff": 0.0421670824289322, "elapsed_sec": 1468.1417932510376}
367
+ {"step": 10340, "epoch": 1, "loss": 3.4829859733581543, "loss_en": 3.551116704940796, "loss_zh": 3.4148552417755127, "noise_std_eff": 0.020848169922828678, "elapsed_sec": 1471.0418899059296}
368
+ {"step": 10360, "epoch": 1, "loss": 3.450015068054199, "loss_en": 3.476140022277832, "loss_zh": 3.4238898754119873, "noise_std_eff": 0.00804600201547146, "elapsed_sec": 1474.0013666152954}
369
+ {"step": 10380, "epoch": 1, "loss": 3.5918636322021484, "loss_en": 3.5870232582092285, "loss_zh": 3.5967037677764893, "noise_std_eff": 0.06531979441642762, "elapsed_sec": 1476.8977088928223}
370
+ {"step": 10400, "epoch": 1, "loss": 3.589787483215332, "loss_en": 3.5875322818756104, "loss_zh": 3.592042922973633, "noise_std_eff": 0.05154408693313599, "elapsed_sec": 1479.8115503787994}
371
+ {"step": 10420, "epoch": 1, "loss": 4.380226135253906, "loss_en": 4.392199516296387, "loss_zh": 4.368252277374268, "noise_std_eff": 0.042363452911376956, "elapsed_sec": 1506.741643667221}
372
+ {"step": 10440, "epoch": 1, "loss": 3.603817939758301, "loss_en": 3.6002845764160156, "loss_zh": 3.607351303100586, "noise_std_eff": 0.06631965458393098, "elapsed_sec": 1509.6827266216278}
373
+ {"step": 10460, "epoch": 1, "loss": 3.814709186553955, "loss_en": 3.8733432292938232, "loss_zh": 3.756075143814087, "noise_std_eff": 0.02654600560665131, "elapsed_sec": 1512.5656719207764}
374
+ {"step": 10480, "epoch": 1, "loss": 3.4529809951782227, "loss_en": 3.468132734298706, "loss_zh": 3.43782901763916, "noise_std_eff": 0.002569229677319527, "elapsed_sec": 1515.5299453735352}
375
+ {"step": 10500, "epoch": 1, "loss": 4.032916069030762, "loss_en": 4.050409317016602, "loss_zh": 4.015422821044922, "noise_std_eff": 0.05783077538013459, "elapsed_sec": 1518.4509284496307}
376
+ {"step": 10520, "epoch": 1, "loss": 4.023543357849121, "loss_en": 4.014459609985352, "loss_zh": 4.032626628875732, "noise_std_eff": 0.008505108505487443, "elapsed_sec": 1521.4073932170868}
377
+ {"step": 10540, "epoch": 1, "loss": 3.5014665126800537, "loss_en": 3.5159380435943604, "loss_zh": 3.486994981765747, "noise_std_eff": 0.05222745537757874, "elapsed_sec": 1524.3182864189148}
378
+ {"step": 10560, "epoch": 1, "loss": 4.20599365234375, "loss_en": 4.18655252456665, "loss_zh": 4.225435256958008, "noise_std_eff": 0.042566862106323246, "elapsed_sec": 1527.277202129364}
379
+ {"step": 10580, "epoch": 1, "loss": 3.3136086463928223, "loss_en": 3.3437788486480713, "loss_zh": 3.283438205718994, "noise_std_eff": 0.047316541075706486, "elapsed_sec": 1530.182272195816}
380
+ {"step": 10600, "epoch": 1, "loss": 3.677377223968506, "loss_en": 3.715360164642334, "loss_zh": 3.639394521713257, "noise_std_eff": 0.05471468269824982, "elapsed_sec": 1533.1384630203247}
381
+ {"step": 10620, "epoch": 1, "loss": 3.170358180999756, "loss_en": 3.192119836807251, "loss_zh": 3.1485962867736816, "noise_std_eff": 0.008510587811470032, "elapsed_sec": 1536.0939741134644}
382
+ {"step": 10640, "epoch": 1, "loss": 3.6909079551696777, "loss_en": 3.698413610458374, "loss_zh": 3.6834025382995605, "noise_std_eff": 0.043608111739158635, "elapsed_sec": 1539.0623109340668}
383
+ {"step": 10660, "epoch": 1, "loss": 3.8177108764648438, "loss_en": 3.8283214569091797, "loss_zh": 3.8071000576019287, "noise_std_eff": 0.05227341771125794, "elapsed_sec": 1541.954184770584}
384
+ {"step": 10680, "epoch": 1, "loss": 3.9782614707946777, "loss_en": 3.99245548248291, "loss_zh": 3.9640674591064453, "noise_std_eff": 0.0027602880075573923, "elapsed_sec": 1544.9544882774353}
385
+ {"step": 10700, "epoch": 1, "loss": 3.464297294616699, "loss_en": 3.4634485244750977, "loss_zh": 3.465146064758301, "noise_std_eff": 0.02805995881557465, "elapsed_sec": 1547.88152551651}
386
+ {"step": 10720, "epoch": 1, "loss": 3.7755422592163086, "loss_en": 3.7687833309173584, "loss_zh": 3.7823009490966797, "noise_std_eff": 0.014862741082906724, "elapsed_sec": 1550.8346405029297}
387
+ {"step": 10740, "epoch": 1, "loss": 3.6877496242523193, "loss_en": 3.69036602973938, "loss_zh": 3.685133218765259, "noise_std_eff": 0.06841938972473145, "elapsed_sec": 1553.7617123126984}
388
+ {"step": 10760, "epoch": 1, "loss": 3.8362784385681152, "loss_en": 3.8295669555664062, "loss_zh": 3.842989921569824, "noise_std_eff": 0.0657209175825119, "elapsed_sec": 1556.7140154838562}
389
+ {"step": 10780, "epoch": 1, "loss": 3.9145500659942627, "loss_en": 3.826456069946289, "loss_zh": 4.002644062042236, "noise_std_eff": 0.029557779729366306, "elapsed_sec": 1559.6214497089386}
390
+ {"step": 10800, "epoch": 1, "loss": 3.316155433654785, "loss_en": 3.3038411140441895, "loss_zh": 3.32846999168396, "noise_std_eff": 0.005965687483549118, "elapsed_sec": 1562.5619881153107}
391
+ {"step": 10820, "epoch": 1, "loss": 3.8616366386413574, "loss_en": 3.9147980213165283, "loss_zh": 3.8084752559661865, "noise_std_eff": 0.028840847015380863, "elapsed_sec": 1589.4801914691925}
392
+ {"step": 10840, "epoch": 1, "loss": 3.5424094200134277, "loss_en": 3.529083013534546, "loss_zh": 3.5557360649108887, "noise_std_eff": 0.06451569974422455, "elapsed_sec": 1592.4509840011597}
393
+ {"step": 10860, "epoch": 1, "loss": 3.8874258995056152, "loss_en": 3.952164649963379, "loss_zh": 3.8226871490478516, "noise_std_eff": 0.019055015146732333, "elapsed_sec": 1595.3616008758545}
394
+ {"step": 10880, "epoch": 1, "loss": 3.6259872913360596, "loss_en": 3.6350393295288086, "loss_zh": 3.6169352531433105, "noise_std_eff": 0.03348185569047928, "elapsed_sec": 1598.2853944301605}
395
+ {"step": 10900, "epoch": 1, "loss": 3.590559482574463, "loss_en": 3.55912446975708, "loss_zh": 3.6219942569732666, "noise_std_eff": 0.007575490474700929, "elapsed_sec": 1601.1964135169983}
396
+ {"step": 10920, "epoch": 1, "loss": 3.7713565826416016, "loss_en": 3.809825897216797, "loss_zh": 3.7328875064849854, "noise_std_eff": 0.03715710878372193, "elapsed_sec": 1604.1528611183167}
397
+ {"step": 10940, "epoch": 1, "loss": 3.4838619232177734, "loss_en": 3.5027010440826416, "loss_zh": 3.4650230407714844, "noise_std_eff": 0.008687597662210465, "elapsed_sec": 1607.0451457500458}
398
+ {"step": 10960, "epoch": 1, "loss": 3.469714879989624, "loss_en": 3.3988702297210693, "loss_zh": 3.5405595302581787, "noise_std_eff": 0.027585559189319615, "elapsed_sec": 1609.9818563461304}
399
+ {"step": 10980, "epoch": 1, "loss": 3.8454947471618652, "loss_en": 3.8483047485351562, "loss_zh": 3.842684507369995, "noise_std_eff": 0.055176721811294564, "elapsed_sec": 1612.8812429904938}
400
+ {"step": 11000, "epoch": 1, "loss": 3.42763090133667, "loss_en": 3.4446868896484375, "loss_zh": 3.4105751514434814, "noise_std_eff": 0.03562105476856232, "elapsed_sec": 1615.8378195762634}
401
+ {"step": 11020, "epoch": 1, "loss": 3.743319034576416, "loss_en": 3.730748176574707, "loss_zh": 3.755889892578125, "noise_std_eff": 0.06783766746520997, "elapsed_sec": 1618.7771110534668}
402
+ {"step": 11040, "epoch": 1, "loss": 3.500842332839966, "loss_en": 3.422349691390991, "loss_zh": 3.5793349742889404, "noise_std_eff": 0.06694303750991822, "elapsed_sec": 1621.7177829742432}
403
+ {"step": 11060, "epoch": 1, "loss": 3.4515433311462402, "loss_en": 3.429628372192383, "loss_zh": 3.4734585285186768, "noise_std_eff": 0.03274790406227112, "elapsed_sec": 1624.6449587345123}
404
+ {"step": 11080, "epoch": 1, "loss": 3.5262269973754883, "loss_en": 3.5658559799194336, "loss_zh": 3.486598014831543, "noise_std_eff": 0.001507652085274458, "elapsed_sec": 1627.59308719635}
405
+ {"step": 11100, "epoch": 1, "loss": 4.1233415603637695, "loss_en": 4.155997276306152, "loss_zh": 4.090686321258545, "noise_std_eff": 0.03752305090427399, "elapsed_sec": 1630.498071193695}
406
+ {"step": 11120, "epoch": 1, "loss": 4.001456260681152, "loss_en": 3.974731922149658, "loss_zh": 4.028180122375488, "noise_std_eff": 0.06313054203987122, "elapsed_sec": 1633.4649894237518}
407
+ {"step": 11140, "epoch": 1, "loss": 3.7606053352355957, "loss_en": 3.793689727783203, "loss_zh": 3.7275209426879883, "noise_std_eff": 0.022045533359050754, "elapsed_sec": 1636.4053361415863}
408
+ {"step": 11160, "epoch": 1, "loss": 3.8976457118988037, "loss_en": 3.9360945224761963, "loss_zh": 3.859196901321411, "noise_std_eff": 0.04415316760540009, "elapsed_sec": 1639.3780992031097}
409
+ {"step": 11180, "epoch": 1, "loss": 3.771331310272217, "loss_en": 3.749528408050537, "loss_zh": 3.7931342124938965, "noise_std_eff": 0.025377781689167024, "elapsed_sec": 1642.3225135803223}
410
+ {"step": 11200, "epoch": 1, "loss": 3.52390193939209, "loss_en": 3.5134308338165283, "loss_zh": 3.5343730449676514, "noise_std_eff": 0.0049661647528409965, "elapsed_sec": 1645.2793486118317}
411
+ {"step": 11220, "epoch": 1, "loss": 3.616971969604492, "loss_en": 3.5992074012756348, "loss_zh": 3.6347365379333496, "noise_std_eff": 0.01862976759672165, "elapsed_sec": 1672.3355889320374}
412
+ {"step": 11240, "epoch": 1, "loss": 3.5727505683898926, "loss_en": 3.579829216003418, "loss_zh": 3.565671682357788, "noise_std_eff": 0.062111660242080695, "elapsed_sec": 1675.661743402481}
413
+ {"step": 11260, "epoch": 1, "loss": 3.390645980834961, "loss_en": 3.3889167308807373, "loss_zh": 3.3923752307891846, "noise_std_eff": 0.060168287158012396, "elapsed_sec": 1678.8722133636475}
414
+ {"step": 11280, "epoch": 1, "loss": 3.7041027545928955, "loss_en": 3.7013700008392334, "loss_zh": 3.7068355083465576, "noise_std_eff": 0.04884208500385285, "elapsed_sec": 1682.0706822872162}
415
+ {"step": 11300, "epoch": 1, "loss": 4.13513708114624, "loss_en": 4.089017391204834, "loss_zh": 4.1812567710876465, "noise_std_eff": 0.025960029661655427, "elapsed_sec": 1684.9714770317078}
416
+ {"step": 11320, "epoch": 1, "loss": 3.629939317703247, "loss_en": 3.6283199787139893, "loss_zh": 3.631558656692505, "noise_std_eff": 0.034868938922882085, "elapsed_sec": 1688.2932000160217}
417
+ {"step": 11340, "epoch": 1, "loss": 3.7146878242492676, "loss_en": 3.6197733879089355, "loss_zh": 3.8096020221710205, "noise_std_eff": 0.04673579514026642, "elapsed_sec": 1691.4713888168335}
418
+ {"step": 11360, "epoch": 1, "loss": 3.826880931854248, "loss_en": 3.785945415496826, "loss_zh": 3.86781644821167, "noise_std_eff": 0.007858947813510896, "elapsed_sec": 1694.6496741771698}
419
+ {"step": 11380, "epoch": 1, "loss": 3.559591293334961, "loss_en": 3.6278538703918457, "loss_zh": 3.4913289546966553, "noise_std_eff": 0.01896239370107651, "elapsed_sec": 1697.7438826560974}
420
+ {"step": 11400, "epoch": 1, "loss": 3.4674296379089355, "loss_en": 3.4630467891693115, "loss_zh": 3.4718122482299805, "noise_std_eff": 0.06663685977458954, "elapsed_sec": 1701.06791806221}
421
+ {"step": 11420, "epoch": 1, "loss": 3.490813970565796, "loss_en": 3.49389386177063, "loss_zh": 3.487734079360962, "noise_std_eff": 0.018751943707466127, "elapsed_sec": 1704.074909210205}
422
+ {"step": 11440, "epoch": 1, "loss": 3.1292810440063477, "loss_en": 3.1363649368286133, "loss_zh": 3.122196912765503, "noise_std_eff": 0.05578714966773987, "elapsed_sec": 1707.175662279129}
423
+ {"step": 11460, "epoch": 1, "loss": 3.4297571182250977, "loss_en": 3.4468941688537598, "loss_zh": 3.4126203060150146, "noise_std_eff": 0.007093413770198823, "elapsed_sec": 1710.0994408130646}
424
+ {"step": 11480, "epoch": 1, "loss": 3.4554598331451416, "loss_en": 3.4951987266540527, "loss_zh": 3.4157209396362305, "noise_std_eff": 0.027526428997516634, "elapsed_sec": 1713.2469701766968}
425
+ {"step": 11500, "epoch": 1, "loss": 3.654444694519043, "loss_en": 3.6183767318725586, "loss_zh": 3.6905126571655273, "noise_std_eff": 0.053880735039711004, "elapsed_sec": 1716.134860277176}
426
+ {"step": 11520, "epoch": 1, "loss": 4.099204063415527, "loss_en": 4.053974151611328, "loss_zh": 4.144433975219727, "noise_std_eff": 0.036548116207122804, "elapsed_sec": 1719.0869991779327}
427
+ {"step": 11540, "epoch": 1, "loss": 3.576498031616211, "loss_en": 3.560814142227173, "loss_zh": 3.59218168258667, "noise_std_eff": 0.032131063342094424, "elapsed_sec": 1722.1574277877808}
428
+ {"step": 11560, "epoch": 1, "loss": 3.5073189735412598, "loss_en": 3.4584336280822754, "loss_zh": 3.5562045574188232, "noise_std_eff": 0.017865151464939118, "elapsed_sec": 1725.3593218326569}
429
+ {"step": 11580, "epoch": 1, "loss": 3.616623878479004, "loss_en": 3.6317858695983887, "loss_zh": 3.6014621257781982, "noise_std_eff": 0.06947796285152436, "elapsed_sec": 1728.359967470169}
430
+ {"step": 11600, "epoch": 1, "loss": 3.5290608406066895, "loss_en": 3.5935940742492676, "loss_zh": 3.4645273685455322, "noise_std_eff": 0.0325802767276764, "elapsed_sec": 1731.2754056453705}
431
+ {"step": 11620, "epoch": 1, "loss": 3.5651073455810547, "loss_en": 3.5302937030792236, "loss_zh": 3.5999209880828857, "noise_std_eff": 0.047709090113639836, "elapsed_sec": 1758.134094953537}
432
+ {"step": 11640, "epoch": 1, "loss": 4.8825602531433105, "loss_en": 4.914834976196289, "loss_zh": 4.850285530090332, "noise_std_eff": 0.021445959806442264, "elapsed_sec": 1761.0869245529175}
433
+ {"step": 11660, "epoch": 1, "loss": 3.6725945472717285, "loss_en": 3.659254550933838, "loss_zh": 3.685934543609619, "noise_std_eff": 0.002653614692389965, "elapsed_sec": 1763.9781093597412}
434
+ {"step": 11680, "epoch": 1, "loss": 3.5076653957366943, "loss_en": 3.5372376441955566, "loss_zh": 3.478093147277832, "noise_std_eff": 0.01318651467561722, "elapsed_sec": 1766.922393321991}
435
+ {"step": 11700, "epoch": 1, "loss": 3.034637928009033, "loss_en": 3.0417869091033936, "loss_zh": 3.0274887084960938, "noise_std_eff": 0.01902176380157471, "elapsed_sec": 1769.803363084793}
436
+ {"step": 11720, "epoch": 1, "loss": 3.726069211959839, "loss_en": 3.7060842514038086, "loss_zh": 3.746054172515869, "noise_std_eff": 0.005473006293177605, "elapsed_sec": 1772.7266182899475}
437
+ {"step": 11740, "epoch": 1, "loss": 3.980027437210083, "loss_en": 3.9661777019500732, "loss_zh": 3.9938771724700928, "noise_std_eff": 0.06880913078784943, "elapsed_sec": 1775.6436970233917}
438
+ {"step": 11760, "epoch": 1, "loss": 3.015962600708008, "loss_en": 2.9106361865997314, "loss_zh": 3.121288776397705, "noise_std_eff": 0.06270080924034119, "elapsed_sec": 1778.5698544979095}
439
+ {"step": 11780, "epoch": 1, "loss": 3.922905921936035, "loss_en": 3.9663360118865967, "loss_zh": 3.8794755935668945, "noise_std_eff": 0.06592943370342255, "elapsed_sec": 1781.4829368591309}
440
+ {"step": 11800, "epoch": 1, "loss": 2.9410078525543213, "loss_en": 2.9116337299346924, "loss_zh": 2.97038197517395, "noise_std_eff": 0.024672904908657076, "elapsed_sec": 1784.4357511997223}
441
+ {"step": 11820, "epoch": 1, "loss": 3.5769667625427246, "loss_en": 3.5379374027252197, "loss_zh": 3.6159958839416504, "noise_std_eff": 0.011726424098014833, "elapsed_sec": 1787.3430330753326}
442
+ {"step": 11840, "epoch": 1, "loss": 3.8567733764648438, "loss_en": 3.871105432510376, "loss_zh": 3.8424415588378906, "noise_std_eff": 0.030625792741775515, "elapsed_sec": 1790.2625644207}
443
+ {"step": 11860, "epoch": 1, "loss": 3.563915252685547, "loss_en": 3.59377384185791, "loss_zh": 3.5340566635131836, "noise_std_eff": 0.00014042100403457883, "elapsed_sec": 1793.1418278217316}
444
+ {"step": 11880, "epoch": 1, "loss": 3.603508949279785, "loss_en": 3.621264696121216, "loss_zh": 3.5857529640197754, "noise_std_eff": 0.012013805508613588, "elapsed_sec": 1796.1157672405243}
445
+ {"step": 11900, "epoch": 1, "loss": 3.71522855758667, "loss_en": 3.7012898921966553, "loss_zh": 3.7291672229766846, "noise_std_eff": 0.03584872603416443, "elapsed_sec": 1799.0608081817627}
446
+ {"step": 11920, "epoch": 1, "loss": 3.4803595542907715, "loss_en": 3.4768121242523193, "loss_zh": 3.4839072227478027, "noise_std_eff": 0.03132572323083878, "elapsed_sec": 1802.0252945423126}
447
+ {"step": 11940, "epoch": 1, "loss": 3.3438339233398438, "loss_en": 3.289041042327881, "loss_zh": 3.3986270427703857, "noise_std_eff": 0.015734670460224153, "elapsed_sec": 1804.9100754261017}
448
+ {"step": 11960, "epoch": 1, "loss": 3.9370226860046387, "loss_en": 3.9571309089660645, "loss_zh": 3.916914701461792, "noise_std_eff": 0.052858878374099735, "elapsed_sec": 1807.8267526626587}
planB_polish_robust_dec_noise07_longwarm/probe_leakage.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_noise07_longwarm/ckpt.pt",
3
+ "max_train_groups": 8000,
4
+ "max_valid_groups": 946,
5
+ "probe_hidden": 256,
6
+ "epochs": 20,
7
+ "lr": 0.01,
8
+ "baseline_majority_lang_acc": 0.5,
9
+ "baseline_majority_len_acc": 0.9271249771118164,
10
+ "lang_probe": {
11
+ "train_acc": 0.5880625,
12
+ "valid_acc": 0.5819238900634249,
13
+ "best_valid_acc": 0.6009513742071881
14
+ },
15
+ "len_probe": {
16
+ "train_acc": 0.929,
17
+ "valid_acc": 0.9011627906976745,
18
+ "best_valid_acc": 0.9011627906976745
19
+ },
20
+ "lang_probe_length_matched": {
21
+ "train_acc": 0.5971865364850977,
22
+ "valid_acc": 0.5887541345093715,
23
+ "best_valid_acc": 0.5959206174200662
24
+ },
25
+ "lang_bucketwise_valid_acc": {
26
+ "bucket_0_acc": 0.5625,
27
+ "bucket_1_acc": 0.574999988079071,
28
+ "bucket_2_acc": 0.5913043022155762,
29
+ "bucket_3_acc": 0.5818182229995728
30
+ },
31
+ "length_matched_sizes": {
32
+ "train": 15568,
33
+ "valid": 1814
34
+ }
35
+ }
planB_polish_robust_dec_noise10/ckpt.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9df34815e2fb03b342db572e58a3e37a63f79d677fd6d4f8c665c53c3122c5e8
3
+ size 2439403193
planB_polish_robust_dec_noise10/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_jsonl": "data/groups_train.jsonl",
3
+ "valid_jsonl": "data/groups_valid.jsonl",
4
+ "run_dir": "runs/planB_polish_robust_dec_noise10",
5
+ "resume": "runs/planB_polish/ckpt.pt",
6
+ "backbone": "google/mt5-small",
7
+ "num_latents": 16,
8
+ "batch_size": 4,
9
+ "grad_accum": 8,
10
+ "epochs": 2,
11
+ "max_doc_len": 256,
12
+ "max_sum_len": 64,
13
+ "eval_every": 400,
14
+ "max_train_examples": 0,
15
+ "max_valid_examples": 0,
16
+ "lr": 0.0002,
17
+ "weight_decay": 0.0,
18
+ "latent_dropout": 0.3,
19
+ "latent_noise_std": 0.1,
20
+ "noise_warmup_steps": 1600,
21
+ "noise_warmup_start_step": -1,
22
+ "untie_lm_head": true,
23
+ "seed": 42
24
+ }
planB_polish_robust_dec_noise10/diag.jsonl ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3200, "nll_en": 4.818785853920225, "nll_zh": 4.823528621463897, "nll": 4.821157237692061}
2
+ {"step": 3600, "nll_en": 4.558684722061641, "nll_zh": 4.561774813347589, "nll": 4.560229767704615}
3
+ {"step": 4000, "nll_en": 4.3977198237848585, "nll_zh": 4.398492805045453, "nll": 4.3981063144151555}
4
+ {"step": 4400, "nll_en": 4.289761447503249, "nll_zh": 4.29273828643573, "nll": 4.29124986696949}
5
+ {"step": 4800, "nll_en": 4.1880216860619965, "nll_zh": 4.1916203760949555, "nll": 4.189821031078476}
6
+ {"step": 5200, "nll_en": 4.119209863166728, "nll_zh": 4.12193023326785, "nll": 4.12057004821729}
7
+ {"step": 5600, "nll_en": 4.052169113280153, "nll_zh": 4.055341969584814, "nll": 4.0537555414324835}
8
+ {"step": 6000, "nll_en": 4.001367066929759, "nll_zh": 4.003301262603502, "nll": 4.00233416476663}
9
+ {"step": 6400, "nll_en": 3.9502853480252353, "nll_zh": 3.9534024930151523, "nll": 3.9518439205201936}
10
+ {"step": 6800, "nll_en": 3.915163915202683, "nll_zh": 3.919316881554827, "nll": 3.9172403983787554}
11
+ {"step": 7200, "nll_en": 3.8680995332262227, "nll_zh": 3.87081017605094, "nll": 3.8694548546385814}
12
+ {"step": 7600, "nll_en": 4.035178409318339, "nll_zh": 4.034775654306875, "nll": 4.034977031812607}
13
+ {"step": 8000, "nll_en": 3.827928181178474, "nll_zh": 3.8317373731423636, "nll": 3.829832777160419}
14
+ {"step": 8400, "nll_en": 3.7941851172336314, "nll_zh": 3.7951166332398105, "nll": 3.7946508752367207}
15
+ {"step": 8800, "nll_en": 3.759905180013709, "nll_zh": 3.7659378535651755, "nll": 3.7629215167894423}
16
+ {"step": 9200, "nll_en": 3.7378642695146684, "nll_zh": 3.743671048267195, "nll": 3.740767658890932}
17
+ {"step": 9600, "nll_en": 3.718832838610917, "nll_zh": 3.721077516769506, "nll": 3.7199551776902116}
18
+ {"step": 10000, "nll_en": 3.6934871935693203, "nll_zh": 3.6977316199095003, "nll": 3.6956094067394103}
19
+ {"step": 10400, "nll_en": 3.6692928229527544, "nll_zh": 3.674749111272316, "nll": 3.6720209671125352}
20
+ {"step": 10800, "nll_en": 3.648681319037149, "nll_zh": 3.6525892505686075, "nll": 3.6506352848028785}
21
+ {"step": 11200, "nll_en": 3.630272066870401, "nll_zh": 3.632752947777069, "nll": 3.631512507323735}
22
+ {"step": 11600, "nll_en": 3.6134926783862125, "nll_zh": 3.6195825209829318, "nll": 3.6165375996845723}
planB_polish_robust_dec_noise10/diag_full.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_noise10/ckpt.pt",
3
+ "nll_en": 3.6022131921877065,
4
+ "nll_zh": 3.6058931108761034,
5
+ "swap_delta_en": 0.5930466732565495,
6
+ "swap_delta_zh": 0.5758072649449677,
7
+ "ablate_zero_delta_en": 1.4273097026171695,
8
+ "ablate_mean_delta_en": 0.27035850803141365,
9
+ "ablate_noise_delta_en": 3.804356868372911,
10
+ "ablate_zero_delta_zh": 1.4236297839287724,
11
+ "ablate_mean_delta_zh": 0.2602228029585792,
12
+ "ablate_noise_delta_zh": 3.837372825211249,
13
+ "inv_top1_full": 0.060253698378801346,
14
+ "inv_top5_full": 0.20401690900325775,
15
+ "diag_sim_mean": 0.965337336063385,
16
+ "offdiag_sim_mean": 0.4495629370212555,
17
+ "sim_margin": 0.5157743990421295,
18
+ "n_valid": 946
19
+ }
planB_polish_robust_dec_noise10/logs.jsonl ADDED
@@ -0,0 +1,448 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3020, "epoch": 0, "loss": 6.061827182769775, "loss_en": 5.989404201507568, "loss_zh": 6.134250164031982, "noise_std_eff": 0.0001754218526184559, "elapsed_sec": 2.691404104232788}
2
+ {"step": 3040, "epoch": 0, "loss": 5.1357316970825195, "loss_en": 5.040549278259277, "loss_zh": 5.2309136390686035, "noise_std_eff": 0.0021944268047809604, "elapsed_sec": 4.911317586898804}
3
+ {"step": 3060, "epoch": 0, "loss": 5.58241081237793, "loss_en": 5.631616115570068, "loss_zh": 5.533205032348633, "noise_std_eff": 0.0037320989370346067, "elapsed_sec": 7.0614635944366455}
4
+ {"step": 3080, "epoch": 0, "loss": 5.780708312988281, "loss_en": 5.848476886749268, "loss_zh": 5.712939262390137, "noise_std_eff": 0.0042315819859504705, "elapsed_sec": 9.29016399383545}
5
+ {"step": 3100, "epoch": 0, "loss": 5.43476676940918, "loss_en": 5.302699565887451, "loss_zh": 5.566834449768066, "noise_std_eff": 0.0027020201086997986, "elapsed_sec": 11.506059885025024}
6
+ {"step": 3120, "epoch": 0, "loss": 5.94249963760376, "loss_en": 5.97524356842041, "loss_zh": 5.909755706787109, "noise_std_eff": 0.00015315722208470106, "elapsed_sec": 13.83811092376709}
7
+ {"step": 3140, "epoch": 0, "loss": 5.147428512573242, "loss_en": 5.094768047332764, "loss_zh": 5.200089454650879, "noise_std_eff": 0.0007735030539333821, "elapsed_sec": 16.059319496154785}
8
+ {"step": 3160, "epoch": 0, "loss": 5.877725601196289, "loss_en": 6.044073104858398, "loss_zh": 5.7113776206970215, "noise_std_eff": 0.0032382598519325257, "elapsed_sec": 18.24372148513794}
9
+ {"step": 3180, "epoch": 0, "loss": 5.191889762878418, "loss_en": 5.217616558074951, "loss_zh": 5.166162490844727, "noise_std_eff": 0.0005780630232766271, "elapsed_sec": 20.474750518798828}
10
+ {"step": 3200, "epoch": 0, "loss": 5.353506088256836, "loss_en": 5.388092517852783, "loss_zh": 5.318920135498047, "noise_std_eff": 0.004481520131230355, "elapsed_sec": 22.679996728897095}
11
+ {"step": 3220, "epoch": 0, "loss": 5.1997833251953125, "loss_en": 5.181720733642578, "loss_zh": 5.217845439910889, "noise_std_eff": 0.002345743197947741, "elapsed_sec": 41.33049821853638}
12
+ {"step": 3240, "epoch": 0, "loss": 5.2040910720825195, "loss_en": 5.168893814086914, "loss_zh": 5.239288806915283, "noise_std_eff": 0.013455267548561098, "elapsed_sec": 43.4530816078186}
13
+ {"step": 3260, "epoch": 0, "loss": 4.925349235534668, "loss_en": 4.907140731811523, "loss_zh": 4.9435577392578125, "noise_std_eff": 0.0032493387535214428, "elapsed_sec": 45.52519655227661}
14
+ {"step": 3280, "epoch": 0, "loss": 5.516790390014648, "loss_en": 5.511528491973877, "loss_zh": 5.522051811218262, "noise_std_eff": 0.010084135383367538, "elapsed_sec": 47.66542673110962}
15
+ {"step": 3300, "epoch": 0, "loss": 4.765016555786133, "loss_en": 4.791459560394287, "loss_zh": 4.73857307434082, "noise_std_eff": 0.0016669259872287513, "elapsed_sec": 49.789955377578735}
16
+ {"step": 3320, "epoch": 0, "loss": 4.446735382080078, "loss_en": 4.487561225891113, "loss_zh": 4.405909061431885, "noise_std_eff": 0.016441538333892825, "elapsed_sec": 51.88322186470032}
17
+ {"step": 3340, "epoch": 0, "loss": 5.186830520629883, "loss_en": 5.1862287521362305, "loss_zh": 5.187431812286377, "noise_std_eff": 0.013020500391721725, "elapsed_sec": 54.03527116775513}
18
+ {"step": 3360, "epoch": 0, "loss": 4.560796737670898, "loss_en": 4.571335315704346, "loss_zh": 4.550257682800293, "noise_std_eff": 0.020532448142766953, "elapsed_sec": 56.237412452697754}
19
+ {"step": 3380, "epoch": 0, "loss": 4.597558975219727, "loss_en": 4.581031322479248, "loss_zh": 4.614086627960205, "noise_std_eff": 0.015125589519739153, "elapsed_sec": 58.412195444107056}
20
+ {"step": 3400, "epoch": 0, "loss": 4.379913806915283, "loss_en": 4.334409236907959, "loss_zh": 4.425418376922607, "noise_std_eff": 0.013750094175338746, "elapsed_sec": 60.900331258773804}
21
+ {"step": 3420, "epoch": 0, "loss": 3.7098727226257324, "loss_en": 3.6711108684539795, "loss_zh": 3.7486348152160645, "noise_std_eff": 0.025103432536125184, "elapsed_sec": 63.00894045829773}
22
+ {"step": 3440, "epoch": 0, "loss": 4.515671730041504, "loss_en": 4.48799991607666, "loss_zh": 4.5433430671691895, "noise_std_eff": 0.010512556359171869, "elapsed_sec": 65.42993831634521}
23
+ {"step": 3460, "epoch": 0, "loss": 4.978893756866455, "loss_en": 4.991073131561279, "loss_zh": 4.966714382171631, "noise_std_eff": 0.00792758833616972, "elapsed_sec": 68.60483384132385}
24
+ {"step": 3480, "epoch": 0, "loss": 5.069268226623535, "loss_en": 5.103120803833008, "loss_zh": 5.0354156494140625, "noise_std_eff": 0.010039102435112, "elapsed_sec": 71.74601244926453}
25
+ {"step": 3500, "epoch": 0, "loss": 4.560745716094971, "loss_en": 4.561920166015625, "loss_zh": 4.559571266174316, "noise_std_eff": 0.024369254708290104, "elapsed_sec": 74.64721965789795}
26
+ {"step": 3520, "epoch": 0, "loss": 4.750950813293457, "loss_en": 4.7258830070495605, "loss_zh": 4.776018142700195, "noise_std_eff": 0.0051379049941897405, "elapsed_sec": 77.8263144493103}
27
+ {"step": 3540, "epoch": 0, "loss": 4.70989465713501, "loss_en": 4.729620933532715, "loss_zh": 4.690168380737305, "noise_std_eff": 0.033329290151596073, "elapsed_sec": 81.08179688453674}
28
+ {"step": 3560, "epoch": 0, "loss": 4.44553804397583, "loss_en": 4.4536919593811035, "loss_zh": 4.437384128570557, "noise_std_eff": 0.03072376102209091, "elapsed_sec": 84.32127618789673}
29
+ {"step": 3580, "epoch": 0, "loss": 4.603093147277832, "loss_en": 4.602278709411621, "loss_zh": 4.603907108306885, "noise_std_eff": 0.025345296859741212, "elapsed_sec": 87.56486344337463}
30
+ {"step": 3600, "epoch": 0, "loss": 4.522439956665039, "loss_en": 4.481605052947998, "loss_zh": 4.56327486038208, "noise_std_eff": 0.026267190277576444, "elapsed_sec": 90.82866287231445}
31
+ {"step": 3620, "epoch": 0, "loss": 5.12519645690918, "loss_en": 5.145527362823486, "loss_zh": 5.104865074157715, "noise_std_eff": 0.022425251156091692, "elapsed_sec": 117.89149069786072}
32
+ {"step": 3640, "epoch": 0, "loss": 4.665097236633301, "loss_en": 4.6562371253967285, "loss_zh": 4.673956871032715, "noise_std_eff": 0.021848645210266117, "elapsed_sec": 120.880535364151}
33
+ {"step": 3660, "epoch": 0, "loss": 4.29643440246582, "loss_en": 4.29066801071167, "loss_zh": 4.3022003173828125, "noise_std_eff": 0.021428564339876173, "elapsed_sec": 123.78127121925354}
34
+ {"step": 3680, "epoch": 0, "loss": 4.256502628326416, "loss_en": 4.304238796234131, "loss_zh": 4.208766460418701, "noise_std_eff": 0.014887154027819632, "elapsed_sec": 126.69687032699585}
35
+ {"step": 3700, "epoch": 0, "loss": 4.874455451965332, "loss_en": 4.926660537719727, "loss_zh": 4.822249889373779, "noise_std_eff": 0.003982820734381676, "elapsed_sec": 129.61153626441956}
36
+ {"step": 3720, "epoch": 0, "loss": 4.471463203430176, "loss_en": 4.460190773010254, "loss_zh": 4.482735633850098, "noise_std_eff": 0.0368287655711174, "elapsed_sec": 132.5317108631134}
37
+ {"step": 3740, "epoch": 0, "loss": 4.357461452484131, "loss_en": 4.3554606437683105, "loss_zh": 4.359462261199951, "noise_std_eff": 0.01577150318771601, "elapsed_sec": 135.4170904159546}
38
+ {"step": 3760, "epoch": 0, "loss": 5.059028625488281, "loss_en": 5.043327808380127, "loss_zh": 5.074729919433594, "noise_std_eff": 0.01107729934155941, "elapsed_sec": 138.34944367408752}
39
+ {"step": 3780, "epoch": 0, "loss": 4.567619323730469, "loss_en": 4.558236122131348, "loss_zh": 4.57700252532959, "noise_std_eff": 0.021496060006320476, "elapsed_sec": 141.2349991798401}
40
+ {"step": 3800, "epoch": 0, "loss": 4.400505065917969, "loss_en": 4.348328113555908, "loss_zh": 4.452682018280029, "noise_std_eff": 0.041544643044471745, "elapsed_sec": 144.173273563385}
41
+ {"step": 3820, "epoch": 0, "loss": 5.435439109802246, "loss_en": 5.368415355682373, "loss_zh": 5.502462863922119, "noise_std_eff": 0.045678545758128165, "elapsed_sec": 147.05094861984253}
42
+ {"step": 3840, "epoch": 0, "loss": 4.9073686599731445, "loss_en": 4.847806930541992, "loss_zh": 4.966930866241455, "noise_std_eff": 0.03298997819423676, "elapsed_sec": 149.96792817115784}
43
+ {"step": 3860, "epoch": 0, "loss": 4.617994785308838, "loss_en": 4.6216278076171875, "loss_zh": 4.614361763000488, "noise_std_eff": 0.0422956146299839, "elapsed_sec": 152.83531546592712}
44
+ {"step": 3880, "epoch": 0, "loss": 4.57366943359375, "loss_en": 4.558657646179199, "loss_zh": 4.588681697845459, "noise_std_eff": 0.022182395905256277, "elapsed_sec": 155.75528645515442}
45
+ {"step": 3900, "epoch": 0, "loss": 4.331183433532715, "loss_en": 4.321146488189697, "loss_zh": 4.341219902038574, "noise_std_eff": 0.021841616928577424, "elapsed_sec": 158.63942122459412}
46
+ {"step": 3920, "epoch": 0, "loss": 3.978022813796997, "loss_en": 3.977266550064087, "loss_zh": 3.9787790775299072, "noise_std_eff": 0.027605645805597306, "elapsed_sec": 161.55711960792542}
47
+ {"step": 3940, "epoch": 0, "loss": 5.491033554077148, "loss_en": 5.493699073791504, "loss_zh": 5.488368034362793, "noise_std_eff": 0.047171009778976446, "elapsed_sec": 164.43523597717285}
48
+ {"step": 3960, "epoch": 0, "loss": 4.354855537414551, "loss_en": 4.363328456878662, "loss_zh": 4.346383094787598, "noise_std_eff": 0.021172733902931215, "elapsed_sec": 167.33712339401245}
49
+ {"step": 3980, "epoch": 0, "loss": 5.007299423217773, "loss_en": 4.967280864715576, "loss_zh": 5.0473175048828125, "noise_std_eff": 0.056387582197785384, "elapsed_sec": 170.22817659378052}
50
+ {"step": 4000, "epoch": 0, "loss": 4.469239711761475, "loss_en": 4.477897644042969, "loss_zh": 4.4605817794799805, "noise_std_eff": 0.062419772148132324, "elapsed_sec": 173.16201996803284}
51
+ {"step": 4020, "epoch": 0, "loss": 4.371463775634766, "loss_en": 4.402125835418701, "loss_zh": 4.340801239013672, "noise_std_eff": 0.014761991277337076, "elapsed_sec": 199.97957038879395}
52
+ {"step": 4040, "epoch": 0, "loss": 4.106801986694336, "loss_en": 4.1465020179748535, "loss_zh": 4.06710147857666, "noise_std_eff": 0.05090444117784501, "elapsed_sec": 202.90673422813416}
53
+ {"step": 4060, "epoch": 0, "loss": 4.943704605102539, "loss_en": 4.854328632354736, "loss_zh": 5.033080577850342, "noise_std_eff": 0.004022431261837483, "elapsed_sec": 205.77223014831543}
54
+ {"step": 4080, "epoch": 0, "loss": 4.281999588012695, "loss_en": 4.249302387237549, "loss_zh": 4.314697265625, "noise_std_eff": 0.018861396387219433, "elapsed_sec": 208.71960926055908}
55
+ {"step": 4100, "epoch": 0, "loss": 4.459500312805176, "loss_en": 4.520665168762207, "loss_zh": 4.3983354568481445, "noise_std_eff": 0.012771155126392843, "elapsed_sec": 211.61965656280518}
56
+ {"step": 4120, "epoch": 0, "loss": 4.475579738616943, "loss_en": 4.507450103759766, "loss_zh": 4.443709373474121, "noise_std_eff": 0.0567247086763382, "elapsed_sec": 214.55217504501343}
57
+ {"step": 4140, "epoch": 0, "loss": 4.582004547119141, "loss_en": 4.550520420074463, "loss_zh": 4.613489151000977, "noise_std_eff": 0.05514158338308335, "elapsed_sec": 217.42614340782166}
58
+ {"step": 4160, "epoch": 0, "loss": 4.190288543701172, "loss_en": 4.148235321044922, "loss_zh": 4.23234224319458, "noise_std_eff": 0.05333612039685249, "elapsed_sec": 220.33857440948486}
59
+ {"step": 4180, "epoch": 0, "loss": 4.796027660369873, "loss_en": 4.801837921142578, "loss_zh": 4.790217399597168, "noise_std_eff": 0.016452211122959855, "elapsed_sec": 223.23171877861023}
60
+ {"step": 4200, "epoch": 0, "loss": 4.579216003417969, "loss_en": 4.532209396362305, "loss_zh": 4.626222133636475, "noise_std_eff": 0.02588069960474968, "elapsed_sec": 226.1604027748108}
61
+ {"step": 4220, "epoch": 0, "loss": 4.3371734619140625, "loss_en": 4.27286434173584, "loss_zh": 4.401482582092285, "noise_std_eff": 0.031235178336501124, "elapsed_sec": 229.03630423545837}
62
+ {"step": 4240, "epoch": 0, "loss": 4.656195640563965, "loss_en": 4.601750373840332, "loss_zh": 4.710640907287598, "noise_std_eff": 0.04168222650885582, "elapsed_sec": 231.93481540679932}
63
+ {"step": 4260, "epoch": 0, "loss": 4.44574499130249, "loss_en": 4.422077655792236, "loss_zh": 4.469412326812744, "noise_std_eff": 0.06999437727034091, "elapsed_sec": 234.77698349952698}
64
+ {"step": 4280, "epoch": 0, "loss": 4.451455593109131, "loss_en": 4.44136381149292, "loss_zh": 4.461547374725342, "noise_std_eff": 0.06222457885742188, "elapsed_sec": 237.68048644065857}
65
+ {"step": 4300, "epoch": 0, "loss": 4.712305068969727, "loss_en": 4.6619157791137695, "loss_zh": 4.762694835662842, "noise_std_eff": 0.025020718760788443, "elapsed_sec": 240.54064011573792}
66
+ {"step": 4320, "epoch": 0, "loss": 4.502044677734375, "loss_en": 4.475196838378906, "loss_zh": 4.528892993927002, "noise_std_eff": 0.04056790731847286, "elapsed_sec": 243.46527647972107}
67
+ {"step": 4340, "epoch": 0, "loss": 4.544926643371582, "loss_en": 4.553666114807129, "loss_zh": 4.536187648773193, "noise_std_eff": 0.06984026104211807, "elapsed_sec": 246.32426476478577}
68
+ {"step": 4360, "epoch": 0, "loss": 4.243021011352539, "loss_en": 4.294402122497559, "loss_zh": 4.191640377044678, "noise_std_eff": 0.010057645477354526, "elapsed_sec": 249.2580165863037}
69
+ {"step": 4380, "epoch": 0, "loss": 4.195290565490723, "loss_en": 4.203419208526611, "loss_zh": 4.187162399291992, "noise_std_eff": 0.03686778880655766, "elapsed_sec": 252.1515772342682}
70
+ {"step": 4400, "epoch": 0, "loss": 4.548520088195801, "loss_en": 4.528177261352539, "loss_zh": 4.568862438201904, "noise_std_eff": 0.003043672721832991, "elapsed_sec": 255.05918049812317}
71
+ {"step": 4420, "epoch": 0, "loss": 4.575366497039795, "loss_en": 4.636046409606934, "loss_zh": 4.514686584472656, "noise_std_eff": 0.0064541901648044585, "elapsed_sec": 281.7931456565857}
72
+ {"step": 4440, "epoch": 0, "loss": 4.619319915771484, "loss_en": 4.637815952301025, "loss_zh": 4.600823402404785, "noise_std_eff": 0.08853276193141937, "elapsed_sec": 284.7221646308899}
73
+ {"step": 4460, "epoch": 0, "loss": 4.386967182159424, "loss_en": 4.368485927581787, "loss_zh": 4.4054484367370605, "noise_std_eff": 0.026376700215041637, "elapsed_sec": 287.6049838066101}
74
+ {"step": 4480, "epoch": 0, "loss": 4.204720497131348, "loss_en": 4.250717639923096, "loss_zh": 4.158723831176758, "noise_std_eff": 0.012682874388992788, "elapsed_sec": 290.5332281589508}
75
+ {"step": 4500, "epoch": 0, "loss": 4.996125221252441, "loss_en": 4.950571537017822, "loss_zh": 5.0416789054870605, "noise_std_eff": 0.02464675158262253, "elapsed_sec": 293.4093644618988}
76
+ {"step": 4520, "epoch": 0, "loss": 4.479645729064941, "loss_en": 4.490284442901611, "loss_zh": 4.469006538391113, "noise_std_eff": 0.055271123051643375, "elapsed_sec": 296.33127641677856}
77
+ {"step": 4540, "epoch": 0, "loss": 4.022144794464111, "loss_en": 3.981396436691284, "loss_zh": 4.062893390655518, "noise_std_eff": 0.04472804751247168, "elapsed_sec": 299.1972532272339}
78
+ {"step": 4560, "epoch": 0, "loss": 4.030787467956543, "loss_en": 4.035321235656738, "loss_zh": 4.026254177093506, "noise_std_eff": 0.08936500042676926, "elapsed_sec": 302.1331932544708}
79
+ {"step": 4580, "epoch": 0, "loss": 4.262275695800781, "loss_en": 4.294863700866699, "loss_zh": 4.229687690734863, "noise_std_eff": 0.03778678931295872, "elapsed_sec": 305.0188603401184}
80
+ {"step": 4600, "epoch": 0, "loss": 4.471635818481445, "loss_en": 4.479032516479492, "loss_zh": 4.46423864364624, "noise_std_eff": 0.02274245321750641, "elapsed_sec": 307.9332745075226}
81
+ {"step": 4620, "epoch": 0, "loss": 5.226893424987793, "loss_en": 5.248307704925537, "loss_zh": 5.205478668212891, "noise_std_eff": 0.06128897070884705, "elapsed_sec": 310.8007423877716}
82
+ {"step": 4640, "epoch": 0, "loss": 4.234832763671875, "loss_en": 4.181459903717041, "loss_zh": 4.288205623626709, "noise_std_eff": 0.08154652118682862, "elapsed_sec": 313.75256848335266}
83
+ {"step": 4660, "epoch": 0, "loss": 3.967008590698242, "loss_en": 3.923243284225464, "loss_zh": 4.010773658752441, "noise_std_eff": 0.020188842713832856, "elapsed_sec": 316.60918164253235}
84
+ {"step": 4680, "epoch": 0, "loss": 4.001281261444092, "loss_en": 4.025635719299316, "loss_zh": 3.9769270420074463, "noise_std_eff": 0.07780279517173767, "elapsed_sec": 319.55068707466125}
85
+ {"step": 4700, "epoch": 0, "loss": 4.310303211212158, "loss_en": 4.317502021789551, "loss_zh": 4.303104400634766, "noise_std_eff": 0.08246994018554688, "elapsed_sec": 322.4305577278137}
86
+ {"step": 4720, "epoch": 0, "loss": 4.4592604637146, "loss_en": 4.468562126159668, "loss_zh": 4.449958801269531, "noise_std_eff": 0.09722675681114197, "elapsed_sec": 325.39892768859863}
87
+ {"step": 4740, "epoch": 0, "loss": 4.610126495361328, "loss_en": 4.633407115936279, "loss_zh": 4.586845397949219, "noise_std_eff": 0.052457988262176514, "elapsed_sec": 328.33379316329956}
88
+ {"step": 4760, "epoch": 0, "loss": 4.225986480712891, "loss_en": 4.251156806945801, "loss_zh": 4.200815677642822, "noise_std_eff": 0.06854352951049805, "elapsed_sec": 331.2571234703064}
89
+ {"step": 4780, "epoch": 0, "loss": 4.431462287902832, "loss_en": 4.397999286651611, "loss_zh": 4.4649248123168945, "noise_std_eff": 0.04718534052371979, "elapsed_sec": 334.12468910217285}
90
+ {"step": 4800, "epoch": 0, "loss": 4.085810661315918, "loss_en": 4.078141212463379, "loss_zh": 4.093480587005615, "noise_std_eff": 0.0036873236298561097, "elapsed_sec": 337.0636262893677}
91
+ {"step": 4820, "epoch": 0, "loss": 3.8538384437561035, "loss_en": 3.8378307819366455, "loss_zh": 3.8698458671569824, "noise_std_eff": 0.09526565074920655, "elapsed_sec": 364.0766770839691}
92
+ {"step": 4840, "epoch": 0, "loss": 4.504721641540527, "loss_en": 4.545465469360352, "loss_zh": 4.463977336883545, "noise_std_eff": 9.383190190419555e-05, "elapsed_sec": 367.0287582874298}
93
+ {"step": 4860, "epoch": 0, "loss": 3.8393807411193848, "loss_en": 3.809673547744751, "loss_zh": 3.8690876960754395, "noise_std_eff": 0.018934655189514163, "elapsed_sec": 369.9062457084656}
94
+ {"step": 4880, "epoch": 0, "loss": 4.263975143432617, "loss_en": 4.238808631896973, "loss_zh": 4.28914213180542, "noise_std_eff": 0.05514178276062012, "elapsed_sec": 372.8154582977295}
95
+ {"step": 4900, "epoch": 0, "loss": 4.181395530700684, "loss_en": 4.173470973968506, "loss_zh": 4.189319610595703, "noise_std_eff": 0.058038157224655156, "elapsed_sec": 375.78391194343567}
96
+ {"step": 4920, "epoch": 0, "loss": 4.571406841278076, "loss_en": 4.5607428550720215, "loss_zh": 4.582070827484131, "noise_std_eff": 0.052157634496688844, "elapsed_sec": 378.72505950927734}
97
+ {"step": 4940, "epoch": 0, "loss": 4.6387128829956055, "loss_en": 4.648740768432617, "loss_zh": 4.6286845207214355, "noise_std_eff": 0.05086598992347718, "elapsed_sec": 381.6084051132202}
98
+ {"step": 4960, "epoch": 0, "loss": 4.792810440063477, "loss_en": 4.804121494293213, "loss_zh": 4.78149938583374, "noise_std_eff": 0.011944004148244858, "elapsed_sec": 384.5005795955658}
99
+ {"step": 4980, "epoch": 0, "loss": 3.8941612243652344, "loss_en": 3.8406779766082764, "loss_zh": 3.9476442337036133, "noise_std_eff": 0.036837393045425416, "elapsed_sec": 387.36436104774475}
100
+ {"step": 5000, "epoch": 0, "loss": 4.054884910583496, "loss_en": 4.052116394042969, "loss_zh": 4.057653903961182, "noise_std_eff": 0.015560050308704377, "elapsed_sec": 390.2806315422058}
101
+ {"step": 5020, "epoch": 0, "loss": 4.26114559173584, "loss_en": 4.291949272155762, "loss_zh": 4.230341911315918, "noise_std_eff": 0.08510141372680664, "elapsed_sec": 393.17673659324646}
102
+ {"step": 5040, "epoch": 0, "loss": 4.532694339752197, "loss_en": 4.506054401397705, "loss_zh": 4.5593342781066895, "noise_std_eff": 0.06894257664680481, "elapsed_sec": 396.10626554489136}
103
+ {"step": 5060, "epoch": 0, "loss": 4.023819446563721, "loss_en": 4.006338119506836, "loss_zh": 4.0413007736206055, "noise_std_eff": 0.06897355318069458, "elapsed_sec": 398.996737241745}
104
+ {"step": 5080, "epoch": 0, "loss": 4.0130696296691895, "loss_en": 4.017292022705078, "loss_zh": 4.008847236633301, "noise_std_eff": 0.010008161514997484, "elapsed_sec": 401.93711018562317}
105
+ {"step": 5100, "epoch": 0, "loss": 3.842155933380127, "loss_en": 3.851787805557251, "loss_zh": 3.832524299621582, "noise_std_eff": 0.04925458729267121, "elapsed_sec": 404.8063087463379}
106
+ {"step": 5120, "epoch": 0, "loss": 4.412898063659668, "loss_en": 4.363254070281982, "loss_zh": 4.4625420570373535, "noise_std_eff": 0.08975088596343994, "elapsed_sec": 407.74877977371216}
107
+ {"step": 5140, "epoch": 0, "loss": 5.400996208190918, "loss_en": 5.374048709869385, "loss_zh": 5.427943706512451, "noise_std_eff": 0.03176674544811249, "elapsed_sec": 410.6685857772827}
108
+ {"step": 5160, "epoch": 0, "loss": 3.9936490058898926, "loss_en": 3.999016761779785, "loss_zh": 3.98828125, "noise_std_eff": 0.010219825059175493, "elapsed_sec": 413.6093876361847}
109
+ {"step": 5180, "epoch": 0, "loss": 3.711966037750244, "loss_en": 3.7631402015686035, "loss_zh": 3.6607916355133057, "noise_std_eff": 0.025472283363342285, "elapsed_sec": 416.46273851394653}
110
+ {"step": 5200, "epoch": 0, "loss": 3.961444139480591, "loss_en": 3.9297659397125244, "loss_zh": 3.9931223392486572, "noise_std_eff": 0.017087700963020324, "elapsed_sec": 419.400315284729}
111
+ {"step": 5220, "epoch": 0, "loss": 3.828847646713257, "loss_en": 3.815558433532715, "loss_zh": 3.842136859893799, "noise_std_eff": 0.012539853155612946, "elapsed_sec": 445.8402280807495}
112
+ {"step": 5240, "epoch": 0, "loss": 4.6580634117126465, "loss_en": 4.6382527351379395, "loss_zh": 4.6778740882873535, "noise_std_eff": 0.02383522391319275, "elapsed_sec": 448.3515405654907}
113
+ {"step": 5260, "epoch": 0, "loss": 4.281237602233887, "loss_en": 4.308562278747559, "loss_zh": 4.253912925720215, "noise_std_eff": 0.05603596568107605, "elapsed_sec": 450.9584574699402}
114
+ {"step": 5280, "epoch": 0, "loss": 4.230630397796631, "loss_en": 4.283751487731934, "loss_zh": 4.177509307861328, "noise_std_eff": 0.06453064680099488, "elapsed_sec": 453.93230986595154}
115
+ {"step": 5300, "epoch": 0, "loss": 4.60191535949707, "loss_en": 4.60800313949585, "loss_zh": 4.595827579498291, "noise_std_eff": 0.035062897205352786, "elapsed_sec": 456.84421491622925}
116
+ {"step": 5320, "epoch": 0, "loss": 4.238175392150879, "loss_en": 4.227699279785156, "loss_zh": 4.24865198135376, "noise_std_eff": 0.09292051196098328, "elapsed_sec": 459.8399238586426}
117
+ {"step": 5340, "epoch": 0, "loss": 4.74442195892334, "loss_en": 4.708089828491211, "loss_zh": 4.780754089355469, "noise_std_eff": 0.08295955657958985, "elapsed_sec": 462.7432839870453}
118
+ {"step": 5360, "epoch": 0, "loss": 4.136745452880859, "loss_en": 4.111550807952881, "loss_zh": 4.161940574645996, "noise_std_eff": 0.05359693765640259, "elapsed_sec": 465.75999999046326}
119
+ {"step": 5380, "epoch": 0, "loss": 4.085380554199219, "loss_en": 4.1161885261535645, "loss_zh": 4.054572582244873, "noise_std_eff": 0.07114845514297485, "elapsed_sec": 468.72514057159424}
120
+ {"step": 5400, "epoch": 0, "loss": 4.395012855529785, "loss_en": 4.368309020996094, "loss_zh": 4.421716213226318, "noise_std_eff": 0.03750266432762146, "elapsed_sec": 471.68015027046204}
121
+ {"step": 5420, "epoch": 0, "loss": 4.635402679443359, "loss_en": 4.637923240661621, "loss_zh": 4.632882595062256, "noise_std_eff": 0.0535530149936676, "elapsed_sec": 474.560045003891}
122
+ {"step": 5440, "epoch": 0, "loss": 4.030348777770996, "loss_en": 4.018098831176758, "loss_zh": 4.042599201202393, "noise_std_eff": 0.04869062602519989, "elapsed_sec": 477.4817204475403}
123
+ {"step": 5460, "epoch": 0, "loss": 4.260231018066406, "loss_en": 4.208887577056885, "loss_zh": 4.3115739822387695, "noise_std_eff": 0.08112677335739137, "elapsed_sec": 480.4244222640991}
124
+ {"step": 5480, "epoch": 0, "loss": 4.096585273742676, "loss_en": 4.086747646331787, "loss_zh": 4.106423377990723, "noise_std_eff": 0.06467376351356506, "elapsed_sec": 483.40055656433105}
125
+ {"step": 5500, "epoch": 0, "loss": 4.276568412780762, "loss_en": 4.2440314292907715, "loss_zh": 4.30910587310791, "noise_std_eff": 0.08331641554832458, "elapsed_sec": 486.29651403427124}
126
+ {"step": 5520, "epoch": 0, "loss": 4.519710540771484, "loss_en": 4.577369213104248, "loss_zh": 4.4620513916015625, "noise_std_eff": 0.06964365243911744, "elapsed_sec": 489.2460889816284}
127
+ {"step": 5540, "epoch": 0, "loss": 4.2184062004089355, "loss_en": 4.242162704467773, "loss_zh": 4.194649696350098, "noise_std_eff": 0.05311465263366699, "elapsed_sec": 492.13639855384827}
128
+ {"step": 5560, "epoch": 0, "loss": 4.171419143676758, "loss_en": 4.1646647453308105, "loss_zh": 4.178173065185547, "noise_std_eff": 0.007401736080646515, "elapsed_sec": 495.0654182434082}
129
+ {"step": 5580, "epoch": 0, "loss": 3.9546961784362793, "loss_en": 4.087588310241699, "loss_zh": 3.8218038082122803, "noise_std_eff": 0.023825590312480927, "elapsed_sec": 498.0006470680237}
130
+ {"step": 5600, "epoch": 0, "loss": 4.406975746154785, "loss_en": 4.4342570304870605, "loss_zh": 4.379694938659668, "noise_std_eff": 0.03355656564235687, "elapsed_sec": 500.9256889820099}
131
+ {"step": 5620, "epoch": 0, "loss": 4.577333927154541, "loss_en": 4.525482654571533, "loss_zh": 4.629185199737549, "noise_std_eff": 0.030287665128707886, "elapsed_sec": 527.5326397418976}
132
+ {"step": 5640, "epoch": 0, "loss": 5.033511638641357, "loss_en": 5.032321929931641, "loss_zh": 5.034701347351074, "noise_std_eff": 0.05855643153190613, "elapsed_sec": 530.4690229892731}
133
+ {"step": 5660, "epoch": 0, "loss": 4.233713150024414, "loss_en": 4.203347206115723, "loss_zh": 4.2640790939331055, "noise_std_eff": 0.08054254055023194, "elapsed_sec": 533.4225924015045}
134
+ {"step": 5680, "epoch": 0, "loss": 4.245596885681152, "loss_en": 4.234531402587891, "loss_zh": 4.256662368774414, "noise_std_eff": 0.05388477444648743, "elapsed_sec": 536.3470313549042}
135
+ {"step": 5700, "epoch": 0, "loss": 3.936054229736328, "loss_en": 3.9345626831054688, "loss_zh": 3.9375460147857666, "noise_std_eff": 0.020486703515052797, "elapsed_sec": 539.2179975509644}
136
+ {"step": 5720, "epoch": 0, "loss": 4.351818561553955, "loss_en": 4.376986503601074, "loss_zh": 4.326650619506836, "noise_std_eff": 0.036875948309898376, "elapsed_sec": 542.1305804252625}
137
+ {"step": 5740, "epoch": 0, "loss": 4.182753086090088, "loss_en": 4.152784824371338, "loss_zh": 4.212721347808838, "noise_std_eff": 0.004623302444815636, "elapsed_sec": 545.0395390987396}
138
+ {"step": 5760, "epoch": 0, "loss": 4.293022632598877, "loss_en": 4.269146919250488, "loss_zh": 4.316898345947266, "noise_std_eff": 0.02094535231590271, "elapsed_sec": 547.9926905632019}
139
+ {"step": 5780, "epoch": 0, "loss": 4.113334655761719, "loss_en": 4.088308811187744, "loss_zh": 4.138360500335693, "noise_std_eff": 0.023593419790267946, "elapsed_sec": 550.8910765647888}
140
+ {"step": 5800, "epoch": 0, "loss": 4.635810852050781, "loss_en": 4.6894636154174805, "loss_zh": 4.582157611846924, "noise_std_eff": 0.0115829735994339, "elapsed_sec": 553.8166151046753}
141
+ {"step": 5820, "epoch": 0, "loss": 4.283271789550781, "loss_en": 4.237304210662842, "loss_zh": 4.329239845275879, "noise_std_eff": 0.09052858352661133, "elapsed_sec": 556.6757640838623}
142
+ {"step": 5840, "epoch": 0, "loss": 3.5856356620788574, "loss_en": 3.5743942260742188, "loss_zh": 3.596877098083496, "noise_std_eff": 0.022576093673706055, "elapsed_sec": 559.6080889701843}
143
+ {"step": 5860, "epoch": 0, "loss": 4.376083850860596, "loss_en": 4.331719398498535, "loss_zh": 4.420448303222656, "noise_std_eff": 0.015918827056884768, "elapsed_sec": 562.4484748840332}
144
+ {"step": 5880, "epoch": 0, "loss": 4.21596622467041, "loss_en": 4.229339599609375, "loss_zh": 4.202592372894287, "noise_std_eff": 0.002132640779018402, "elapsed_sec": 565.3419609069824}
145
+ {"step": 5900, "epoch": 0, "loss": 4.397330284118652, "loss_en": 4.427899360656738, "loss_zh": 4.366761684417725, "noise_std_eff": 0.06938122510910034, "elapsed_sec": 568.2043700218201}
146
+ {"step": 5920, "epoch": 0, "loss": 4.540289402008057, "loss_en": 4.506087303161621, "loss_zh": 4.574491500854492, "noise_std_eff": 0.00688508152961731, "elapsed_sec": 571.0943627357483}
147
+ {"step": 5940, "epoch": 0, "loss": 4.082942962646484, "loss_en": 4.072699546813965, "loss_zh": 4.093185901641846, "noise_std_eff": 0.041160967946052555, "elapsed_sec": 574.0409877300262}
148
+ {"step": 5960, "epoch": 0, "loss": 4.569228649139404, "loss_en": 4.554348468780518, "loss_zh": 4.584108829498291, "noise_std_eff": 0.015291193127632143, "elapsed_sec": 577.0273435115814}
149
+ {"step": 5980, "epoch": 0, "loss": 4.0443115234375, "loss_en": 4.06700325012207, "loss_zh": 4.0216193199157715, "noise_std_eff": 0.027223277091979983, "elapsed_sec": 580.0273313522339}
150
+ {"step": 6000, "epoch": 0, "loss": 4.029739856719971, "loss_en": 3.980633497238159, "loss_zh": 4.078845977783203, "noise_std_eff": 0.009423379600048066, "elapsed_sec": 582.9510703086853}
151
+ {"step": 6020, "epoch": 0, "loss": 4.134744644165039, "loss_en": 4.09615421295166, "loss_zh": 4.17333459854126, "noise_std_eff": 0.08212392330169678, "elapsed_sec": 609.9941916465759}
152
+ {"step": 6040, "epoch": 0, "loss": 3.467245101928711, "loss_en": 3.4094460010528564, "loss_zh": 3.5250444412231445, "noise_std_eff": 0.06718523502349853, "elapsed_sec": 612.8901162147522}
153
+ {"step": 6060, "epoch": 0, "loss": 4.056488037109375, "loss_en": 4.03832483291626, "loss_zh": 4.07465124130249, "noise_std_eff": 0.06385783553123474, "elapsed_sec": 615.7945296764374}
154
+ {"step": 6080, "epoch": 0, "loss": 4.211997985839844, "loss_en": 4.241757869720459, "loss_zh": 4.18223762512207, "noise_std_eff": 0.08977231979370118, "elapsed_sec": 618.7098536491394}
155
+ {"step": 6100, "epoch": 0, "loss": 4.190611839294434, "loss_en": 4.209909439086914, "loss_zh": 4.171313762664795, "noise_std_eff": 0.000276500266045332, "elapsed_sec": 621.5812065601349}
156
+ {"step": 6120, "epoch": 0, "loss": 4.491664886474609, "loss_en": 4.495882511138916, "loss_zh": 4.487447261810303, "noise_std_eff": 0.03208238184452057, "elapsed_sec": 624.4745342731476}
157
+ {"step": 6140, "epoch": 0, "loss": 4.059443473815918, "loss_en": 4.047981262207031, "loss_zh": 4.0709052085876465, "noise_std_eff": 0.002237655222415924, "elapsed_sec": 627.3418185710907}
158
+ {"step": 6160, "epoch": 0, "loss": 4.152255058288574, "loss_en": 4.1693220138549805, "loss_zh": 4.13518762588501, "noise_std_eff": 0.09594789147377014, "elapsed_sec": 630.242760181427}
159
+ {"step": 6180, "epoch": 0, "loss": 4.172747611999512, "loss_en": 4.073394298553467, "loss_zh": 4.272100448608398, "noise_std_eff": 0.07737690210342407, "elapsed_sec": 633.1060574054718}
160
+ {"step": 6200, "epoch": 0, "loss": 3.7712676525115967, "loss_en": 3.7905642986297607, "loss_zh": 3.7519710063934326, "noise_std_eff": 0.040107938647270205, "elapsed_sec": 636.0069246292114}
161
+ {"step": 6220, "epoch": 0, "loss": 3.9523279666900635, "loss_en": 3.9742045402526855, "loss_zh": 3.9304513931274414, "noise_std_eff": 0.018917140364646912, "elapsed_sec": 638.8391931056976}
162
+ {"step": 6240, "epoch": 0, "loss": 3.8871917724609375, "loss_en": 3.832688808441162, "loss_zh": 3.941694498062134, "noise_std_eff": 0.019902619719505313, "elapsed_sec": 641.725145816803}
163
+ {"step": 6260, "epoch": 0, "loss": 4.199700355529785, "loss_en": 4.169081687927246, "loss_zh": 4.230318546295166, "noise_std_eff": 0.09009930491447449, "elapsed_sec": 644.5500719547272}
164
+ {"step": 6280, "epoch": 0, "loss": 4.432948112487793, "loss_en": 4.441094398498535, "loss_zh": 4.424802303314209, "noise_std_eff": 0.05726805925369263, "elapsed_sec": 647.43004322052}
165
+ {"step": 6300, "epoch": 0, "loss": 3.9303789138793945, "loss_en": 3.953932762145996, "loss_zh": 3.906824827194214, "noise_std_eff": 0.08020164966583253, "elapsed_sec": 650.2728524208069}
166
+ {"step": 6320, "epoch": 0, "loss": 3.848454475402832, "loss_en": 3.8669214248657227, "loss_zh": 3.8299877643585205, "noise_std_eff": 0.033671569824218754, "elapsed_sec": 653.1543838977814}
167
+ {"step": 6340, "epoch": 0, "loss": 3.5823607444763184, "loss_en": 3.5157837867736816, "loss_zh": 3.648937702178955, "noise_std_eff": 0.023414959013462067, "elapsed_sec": 655.9860484600067}
168
+ {"step": 6360, "epoch": 0, "loss": 4.295183181762695, "loss_en": 4.3642964363098145, "loss_zh": 4.226070404052734, "noise_std_eff": 0.026227167248725893, "elapsed_sec": 658.8669021129608}
169
+ {"step": 6380, "epoch": 0, "loss": 3.789916515350342, "loss_en": 3.8096814155578613, "loss_zh": 3.7701516151428223, "noise_std_eff": 0.04811425805091858, "elapsed_sec": 661.7098371982574}
170
+ {"step": 6400, "epoch": 0, "loss": 4.133044242858887, "loss_en": 4.120359897613525, "loss_zh": 4.14572811126709, "noise_std_eff": 0.07831720709800721, "elapsed_sec": 664.5858752727509}
171
+ {"step": 6420, "epoch": 0, "loss": 4.177938461303711, "loss_en": 4.234639644622803, "loss_zh": 4.121237277984619, "noise_std_eff": 0.055882138013839726, "elapsed_sec": 691.3496301174164}
172
+ {"step": 6440, "epoch": 0, "loss": 4.06437873840332, "loss_en": 4.085473537445068, "loss_zh": 4.0432844161987305, "noise_std_eff": 0.008030712604522705, "elapsed_sec": 694.294111251831}
173
+ {"step": 6460, "epoch": 0, "loss": 4.078014373779297, "loss_en": 4.075385093688965, "loss_zh": 4.080644130706787, "noise_std_eff": 0.07103831171989441, "elapsed_sec": 697.1559875011444}
174
+ {"step": 6480, "epoch": 0, "loss": 4.269293785095215, "loss_en": 4.28524112701416, "loss_zh": 4.253346920013428, "noise_std_eff": 0.01656680852174759, "elapsed_sec": 700.061859369278}
175
+ {"step": 6500, "epoch": 0, "loss": 3.7130026817321777, "loss_en": 3.721132516860962, "loss_zh": 3.7048726081848145, "noise_std_eff": 0.07617729902267456, "elapsed_sec": 702.9139330387115}
176
+ {"step": 6520, "epoch": 0, "loss": 3.876997947692871, "loss_en": 3.9038078784942627, "loss_zh": 3.8501877784729004, "noise_std_eff": 0.09562607407569886, "elapsed_sec": 705.809298992157}
177
+ {"step": 6540, "epoch": 0, "loss": 3.833609104156494, "loss_en": 3.847975969314575, "loss_zh": 3.819242238998413, "noise_std_eff": 0.003355959430336952, "elapsed_sec": 708.6762547492981}
178
+ {"step": 6560, "epoch": 0, "loss": 4.797603607177734, "loss_en": 4.7618303298950195, "loss_zh": 4.833376407623291, "noise_std_eff": 0.0996338427066803, "elapsed_sec": 711.5931944847107}
179
+ {"step": 6580, "epoch": 0, "loss": 3.8778791427612305, "loss_en": 3.865009307861328, "loss_zh": 3.890748977661133, "noise_std_eff": 0.05159723162651062, "elapsed_sec": 714.4501783847809}
180
+ {"step": 6600, "epoch": 0, "loss": 4.108623504638672, "loss_en": 4.102266311645508, "loss_zh": 4.114981174468994, "noise_std_eff": 0.06838358640670776, "elapsed_sec": 717.3688991069794}
181
+ {"step": 6620, "epoch": 0, "loss": 3.9339466094970703, "loss_en": 3.898282766342163, "loss_zh": 3.9696104526519775, "noise_std_eff": 0.09483279585838318, "elapsed_sec": 720.2564918994904}
182
+ {"step": 6640, "epoch": 0, "loss": 3.8542513847351074, "loss_en": 3.810046672821045, "loss_zh": 3.89845609664917, "noise_std_eff": 0.08282724022865295, "elapsed_sec": 723.1417257785797}
183
+ {"step": 6660, "epoch": 0, "loss": 4.311872959136963, "loss_en": 4.360252857208252, "loss_zh": 4.263493061065674, "noise_std_eff": 0.018653173744678498, "elapsed_sec": 726.0226426124573}
184
+ {"step": 6680, "epoch": 0, "loss": 3.556107521057129, "loss_en": 3.5734200477600098, "loss_zh": 3.538795232772827, "noise_std_eff": 0.03676941990852356, "elapsed_sec": 728.9095544815063}
185
+ {"step": 6700, "epoch": 0, "loss": 3.5899102687835693, "loss_en": 3.6318535804748535, "loss_zh": 3.547966957092285, "noise_std_eff": 0.08906001448631287, "elapsed_sec": 731.7744069099426}
186
+ {"step": 6720, "epoch": 0, "loss": 4.124873161315918, "loss_en": 4.1362528800964355, "loss_zh": 4.113492965698242, "noise_std_eff": 0.017143724858760836, "elapsed_sec": 734.6712365150452}
187
+ {"step": 6740, "epoch": 0, "loss": 4.0562591552734375, "loss_en": 4.092789649963379, "loss_zh": 4.019728183746338, "noise_std_eff": 0.04082382619380951, "elapsed_sec": 737.5458686351776}
188
+ {"step": 6760, "epoch": 0, "loss": 4.170685768127441, "loss_en": 4.19294548034668, "loss_zh": 4.148426532745361, "noise_std_eff": 0.05638126134872437, "elapsed_sec": 740.4507954120636}
189
+ {"step": 6780, "epoch": 0, "loss": 3.939495801925659, "loss_en": 3.9464519023895264, "loss_zh": 3.932539701461792, "noise_std_eff": 0.026284536719322207, "elapsed_sec": 743.3312237262726}
190
+ {"step": 6800, "epoch": 0, "loss": 3.962501049041748, "loss_en": 3.9412834644317627, "loss_zh": 3.9837188720703125, "noise_std_eff": 0.04278116822242737, "elapsed_sec": 746.2458267211914}
191
+ {"step": 6820, "epoch": 0, "loss": 4.450061798095703, "loss_en": 4.456645965576172, "loss_zh": 4.443478107452393, "noise_std_eff": 0.04806626439094544, "elapsed_sec": 773.0418446063995}
192
+ {"step": 6840, "epoch": 0, "loss": 4.465451240539551, "loss_en": 4.589513301849365, "loss_zh": 4.3413896560668945, "noise_std_eff": 0.03760355114936829, "elapsed_sec": 775.9220004081726}
193
+ {"step": 6860, "epoch": 0, "loss": 3.909973382949829, "loss_en": 3.884326219558716, "loss_zh": 3.9356205463409424, "noise_std_eff": 0.0908029317855835, "elapsed_sec": 779.0909714698792}
194
+ {"step": 6880, "epoch": 0, "loss": 4.045411109924316, "loss_en": 4.051043510437012, "loss_zh": 4.039778709411621, "noise_std_eff": 0.03879079520702362, "elapsed_sec": 782.352786064148}
195
+ {"step": 6900, "epoch": 0, "loss": 4.139927864074707, "loss_en": 4.170746803283691, "loss_zh": 4.1091084480285645, "noise_std_eff": 0.03490573763847351, "elapsed_sec": 785.5532615184784}
196
+ {"step": 6920, "epoch": 0, "loss": 3.598865509033203, "loss_en": 3.589495897293091, "loss_zh": 3.6082348823547363, "noise_std_eff": 0.011638482660055162, "elapsed_sec": 788.8148698806763}
197
+ {"step": 6940, "epoch": 0, "loss": 3.884692668914795, "loss_en": 3.8867430686950684, "loss_zh": 3.8826420307159424, "noise_std_eff": 0.04156458973884583, "elapsed_sec": 792.0392384529114}
198
+ {"step": 6960, "epoch": 0, "loss": 3.9390006065368652, "loss_en": 3.9536819458007812, "loss_zh": 3.9243195056915283, "noise_std_eff": 0.056569921970367434, "elapsed_sec": 795.2949714660645}
199
+ {"step": 6980, "epoch": 0, "loss": 3.7227649688720703, "loss_en": 3.712475538253784, "loss_zh": 3.7330541610717773, "noise_std_eff": 0.03308436870574951, "elapsed_sec": 798.539514541626}
200
+ {"step": 7000, "epoch": 0, "loss": 4.115448951721191, "loss_en": 4.071257591247559, "loss_zh": 4.159640312194824, "noise_std_eff": 0.009268762916326523, "elapsed_sec": 801.7895607948303}
201
+ {"step": 7020, "epoch": 0, "loss": 4.10771369934082, "loss_en": 4.114213943481445, "loss_zh": 4.1012139320373535, "noise_std_eff": 0.010581877082586288, "elapsed_sec": 804.9853208065033}
202
+ {"step": 7040, "epoch": 0, "loss": 4.25765323638916, "loss_en": 4.27769660949707, "loss_zh": 4.237609386444092, "noise_std_eff": 0.05347551703453064, "elapsed_sec": 808.2707033157349}
203
+ {"step": 7060, "epoch": 0, "loss": 4.037903785705566, "loss_en": 4.083699703216553, "loss_zh": 3.992108106613159, "noise_std_eff": 0.03097880780696869, "elapsed_sec": 811.4591870307922}
204
+ {"step": 7080, "epoch": 0, "loss": 4.5530524253845215, "loss_en": 4.4880900382995605, "loss_zh": 4.618014812469482, "noise_std_eff": 0.060183352231979376, "elapsed_sec": 814.7296390533447}
205
+ {"step": 7100, "epoch": 0, "loss": 4.089316368103027, "loss_en": 4.11367654800415, "loss_zh": 4.064956188201904, "noise_std_eff": 0.006474116444587708, "elapsed_sec": 817.8432805538177}
206
+ {"step": 7120, "epoch": 0, "loss": 4.113682746887207, "loss_en": 4.079658031463623, "loss_zh": 4.147707462310791, "noise_std_eff": 0.0854093372821808, "elapsed_sec": 820.8552994728088}
207
+ {"step": 7140, "epoch": 0, "loss": 4.55783224105835, "loss_en": 4.521496772766113, "loss_zh": 4.594167709350586, "noise_std_eff": 0.07115347385406494, "elapsed_sec": 823.7331023216248}
208
+ {"step": 7160, "epoch": 0, "loss": 4.2585368156433105, "loss_en": 4.2870402336120605, "loss_zh": 4.2300333976745605, "noise_std_eff": 0.05217787027359009, "elapsed_sec": 826.7107381820679}
209
+ {"step": 7180, "epoch": 0, "loss": 3.7725749015808105, "loss_en": 3.826319932937622, "loss_zh": 3.718829870223999, "noise_std_eff": 0.08929979205131532, "elapsed_sec": 829.606701374054}
210
+ {"step": 7200, "epoch": 0, "loss": 3.986398220062256, "loss_en": 4.009593486785889, "loss_zh": 3.963202714920044, "noise_std_eff": 0.05542289614677429, "elapsed_sec": 832.5068190097809}
211
+ {"step": 7220, "epoch": 0, "loss": 4.0705790519714355, "loss_en": 4.0179901123046875, "loss_zh": 4.123167991638184, "noise_std_eff": 0.0205546647310257, "elapsed_sec": 859.3341867923737}
212
+ {"step": 7240, "epoch": 0, "loss": 3.8302597999572754, "loss_en": 3.7919249534606934, "loss_zh": 3.8685948848724365, "noise_std_eff": 0.06370125412940979, "elapsed_sec": 862.249589920044}
213
+ {"step": 7260, "epoch": 0, "loss": 3.733952522277832, "loss_en": 3.766879081726074, "loss_zh": 3.7010257244110107, "noise_std_eff": 0.014478990435600282, "elapsed_sec": 865.1134378910065}
214
+ {"step": 7280, "epoch": 0, "loss": 3.703648567199707, "loss_en": 3.7054502964019775, "loss_zh": 3.7018470764160156, "noise_std_eff": 0.07637184858322144, "elapsed_sec": 868.022876739502}
215
+ {"step": 7300, "epoch": 0, "loss": 3.621786117553711, "loss_en": 3.601999282836914, "loss_zh": 3.6415727138519287, "noise_std_eff": 0.09068971872329712, "elapsed_sec": 870.8941357135773}
216
+ {"step": 7320, "epoch": 0, "loss": 3.876955509185791, "loss_en": 3.8873064517974854, "loss_zh": 3.8666043281555176, "noise_std_eff": 0.010125520080327988, "elapsed_sec": 873.8118677139282}
217
+ {"step": 7340, "epoch": 0, "loss": 4.607444763183594, "loss_en": 4.619564056396484, "loss_zh": 4.595325469970703, "noise_std_eff": 0.06677759885787964, "elapsed_sec": 876.70215010643}
218
+ {"step": 7360, "epoch": 0, "loss": 3.7698521614074707, "loss_en": 3.7248096466064453, "loss_zh": 3.814894437789917, "noise_std_eff": 0.06785207390785218, "elapsed_sec": 879.5901503562927}
219
+ {"step": 7380, "epoch": 0, "loss": 3.9577698707580566, "loss_en": 4.031519412994385, "loss_zh": 3.8840205669403076, "noise_std_eff": 0.023438793420791627, "elapsed_sec": 882.4787290096283}
220
+ {"step": 7400, "epoch": 0, "loss": 3.6157331466674805, "loss_en": 3.6547746658325195, "loss_zh": 3.5766916275024414, "noise_std_eff": 0.014886277914047242, "elapsed_sec": 885.3545544147491}
221
+ {"step": 7420, "epoch": 0, "loss": 4.002717971801758, "loss_en": 4.019887447357178, "loss_zh": 3.985548734664917, "noise_std_eff": 0.03532204031944275, "elapsed_sec": 888.2185068130493}
222
+ {"step": 7440, "epoch": 0, "loss": 3.909323215484619, "loss_en": 3.876458168029785, "loss_zh": 3.942188024520874, "noise_std_eff": 0.023387783765792848, "elapsed_sec": 891.106428861618}
223
+ {"step": 7460, "epoch": 0, "loss": 3.7151761054992676, "loss_en": 3.7133846282958984, "loss_zh": 3.7169673442840576, "noise_std_eff": 0.02157682627439499, "elapsed_sec": 893.9546258449554}
224
+ {"step": 7480, "epoch": 0, "loss": 4.292357444763184, "loss_en": 4.247401237487793, "loss_zh": 4.337313175201416, "noise_std_eff": 0.020149198174476624, "elapsed_sec": 896.8223004341125}
225
+ {"step": 7500, "epoch": 1, "loss": 5.172224044799805, "loss_en": 5.102956295013428, "loss_zh": 5.241491794586182, "noise_std_eff": 0.028835663199424745, "elapsed_sec": 899.7758679389954}
226
+ {"step": 7520, "epoch": 1, "loss": 4.944433689117432, "loss_en": 4.782658100128174, "loss_zh": 5.1062092781066895, "noise_std_eff": 0.0074190847575664525, "elapsed_sec": 902.7721893787384}
227
+ {"step": 7540, "epoch": 1, "loss": 5.029420852661133, "loss_en": 5.01795768737793, "loss_zh": 5.040883541107178, "noise_std_eff": 0.04165078997612, "elapsed_sec": 905.7358586788177}
228
+ {"step": 7560, "epoch": 1, "loss": 4.98674201965332, "loss_en": 5.066978931427002, "loss_zh": 4.906505584716797, "noise_std_eff": 0.08525810837745668, "elapsed_sec": 908.7080602645874}
229
+ {"step": 7580, "epoch": 1, "loss": 5.079340934753418, "loss_en": 5.082379341125488, "loss_zh": 5.0763020515441895, "noise_std_eff": 0.04131305813789368, "elapsed_sec": 911.6806995868683}
230
+ {"step": 7600, "epoch": 1, "loss": 5.643168926239014, "loss_en": 5.784675598144531, "loss_zh": 5.501662254333496, "noise_std_eff": 0.043657305836677554, "elapsed_sec": 914.7056775093079}
231
+ {"step": 7620, "epoch": 1, "loss": 4.264737129211426, "loss_en": 4.250869274139404, "loss_zh": 4.278604507446289, "noise_std_eff": 0.0017623750492930413, "elapsed_sec": 941.4754440784454}
232
+ {"step": 7640, "epoch": 1, "loss": 3.737342357635498, "loss_en": 3.692646026611328, "loss_zh": 3.782038927078247, "noise_std_eff": 0.05551101565361023, "elapsed_sec": 944.4517149925232}
233
+ {"step": 7660, "epoch": 1, "loss": 4.415035247802734, "loss_en": 4.418068885803223, "loss_zh": 4.412002086639404, "noise_std_eff": 0.016270591318607332, "elapsed_sec": 947.3394300937653}
234
+ {"step": 7680, "epoch": 1, "loss": 3.761282444000244, "loss_en": 3.7548673152923584, "loss_zh": 3.767697811126709, "noise_std_eff": 0.06796910166740418, "elapsed_sec": 950.2795341014862}
235
+ {"step": 7700, "epoch": 1, "loss": 3.5565366744995117, "loss_en": 3.539860725402832, "loss_zh": 3.5732126235961914, "noise_std_eff": 0.04399799704551697, "elapsed_sec": 953.1591491699219}
236
+ {"step": 7720, "epoch": 1, "loss": 3.744513750076294, "loss_en": 3.764516830444336, "loss_zh": 3.724510669708252, "noise_std_eff": 0.04835307002067566, "elapsed_sec": 956.0907409191132}
237
+ {"step": 7740, "epoch": 1, "loss": 4.0253777503967285, "loss_en": 4.0352654457092285, "loss_zh": 4.0154900550842285, "noise_std_eff": 0.015104779601097108, "elapsed_sec": 958.9812626838684}
238
+ {"step": 7760, "epoch": 1, "loss": 3.99008846282959, "loss_en": 3.9764528274536133, "loss_zh": 4.003724098205566, "noise_std_eff": 0.02552516758441925, "elapsed_sec": 961.9156219959259}
239
+ {"step": 7780, "epoch": 1, "loss": 3.876162528991699, "loss_en": 3.8855934143066406, "loss_zh": 3.866731882095337, "noise_std_eff": 0.05353057980537415, "elapsed_sec": 964.7908802032471}
240
+ {"step": 7800, "epoch": 1, "loss": 4.587019920349121, "loss_en": 4.548828125, "loss_zh": 4.6252121925354, "noise_std_eff": 0.08450542688369751, "elapsed_sec": 967.7197606563568}
241
+ {"step": 7820, "epoch": 1, "loss": 3.4906039237976074, "loss_en": 3.4500346183776855, "loss_zh": 3.5311732292175293, "noise_std_eff": 0.0835247814655304, "elapsed_sec": 970.6183977127075}
242
+ {"step": 7840, "epoch": 1, "loss": 4.211109161376953, "loss_en": 4.182861804962158, "loss_zh": 4.23935604095459, "noise_std_eff": 0.07278578281402588, "elapsed_sec": 973.5591566562653}
243
+ {"step": 7860, "epoch": 1, "loss": 3.642085552215576, "loss_en": 3.6663131713867188, "loss_zh": 3.6178581714630127, "noise_std_eff": 0.08464226126670837, "elapsed_sec": 976.4588334560394}
244
+ {"step": 7880, "epoch": 1, "loss": 3.9898629188537598, "loss_en": 3.990724563598633, "loss_zh": 3.989001512527466, "noise_std_eff": 0.016079235076904296, "elapsed_sec": 979.3635721206665}
245
+ {"step": 7900, "epoch": 1, "loss": 4.076443672180176, "loss_en": 4.046926021575928, "loss_zh": 4.105961799621582, "noise_std_eff": 0.04370894432067871, "elapsed_sec": 982.2344431877136}
246
+ {"step": 7920, "epoch": 1, "loss": 3.89158296585083, "loss_en": 3.8590283393859863, "loss_zh": 3.924137830734253, "noise_std_eff": 0.04208592772483826, "elapsed_sec": 985.1513299942017}
247
+ {"step": 7940, "epoch": 1, "loss": 3.7294299602508545, "loss_en": 3.705791711807251, "loss_zh": 3.753068208694458, "noise_std_eff": 0.013963478803634645, "elapsed_sec": 988.0466220378876}
248
+ {"step": 7960, "epoch": 1, "loss": 4.193987846374512, "loss_en": 4.158372402191162, "loss_zh": 4.229603290557861, "noise_std_eff": 0.0455489456653595, "elapsed_sec": 990.9670741558075}
249
+ {"step": 7980, "epoch": 1, "loss": 3.9706473350524902, "loss_en": 4.01150369644165, "loss_zh": 3.929790735244751, "noise_std_eff": 0.05431650280952454, "elapsed_sec": 993.8511574268341}
250
+ {"step": 8000, "epoch": 1, "loss": 3.825986385345459, "loss_en": 3.824303150177002, "loss_zh": 3.827669382095337, "noise_std_eff": 0.030818283557891846, "elapsed_sec": 996.7842202186584}
251
+ {"step": 8020, "epoch": 1, "loss": 3.6920676231384277, "loss_en": 3.721858501434326, "loss_zh": 3.6622767448425293, "noise_std_eff": 0.03590853810310364, "elapsed_sec": 1023.9057343006134}
252
+ {"step": 8040, "epoch": 1, "loss": 3.666266918182373, "loss_en": 3.6794281005859375, "loss_zh": 3.6531059741973877, "noise_std_eff": 0.017723095417022706, "elapsed_sec": 1026.863554239273}
253
+ {"step": 8060, "epoch": 1, "loss": 4.567615509033203, "loss_en": 4.531426906585693, "loss_zh": 4.603804588317871, "noise_std_eff": 0.07223026156425476, "elapsed_sec": 1029.7179825305939}
254
+ {"step": 8080, "epoch": 1, "loss": 4.259527683258057, "loss_en": 4.243832111358643, "loss_zh": 4.275223255157471, "noise_std_eff": 0.002642740122973919, "elapsed_sec": 1032.6543607711792}
255
+ {"step": 8100, "epoch": 1, "loss": 3.8594465255737305, "loss_en": 3.8772411346435547, "loss_zh": 3.841651678085327, "noise_std_eff": 0.018805910646915436, "elapsed_sec": 1035.5180640220642}
256
+ {"step": 8120, "epoch": 1, "loss": 3.9967434406280518, "loss_en": 3.9649338722229004, "loss_zh": 4.028553009033203, "noise_std_eff": 0.08868687152862549, "elapsed_sec": 1038.4543843269348}
257
+ {"step": 8140, "epoch": 1, "loss": 3.56304931640625, "loss_en": 3.5865705013275146, "loss_zh": 3.5395281314849854, "noise_std_eff": 0.06251311898231507, "elapsed_sec": 1041.3062152862549}
258
+ {"step": 8160, "epoch": 1, "loss": 3.621493101119995, "loss_en": 3.6374218463897705, "loss_zh": 3.6055643558502197, "noise_std_eff": 0.04918995797634125, "elapsed_sec": 1044.1984505653381}
259
+ {"step": 8180, "epoch": 1, "loss": 3.9343626499176025, "loss_en": 3.934054374694824, "loss_zh": 3.934670925140381, "noise_std_eff": 0.02295040488243103, "elapsed_sec": 1047.078145980835}
260
+ {"step": 8200, "epoch": 1, "loss": 4.1464128494262695, "loss_en": 4.17581033706665, "loss_zh": 4.1170148849487305, "noise_std_eff": 0.04807215929031372, "elapsed_sec": 1049.9908652305603}
261
+ {"step": 8220, "epoch": 1, "loss": 3.8188304901123047, "loss_en": 3.8419878482818604, "loss_zh": 3.79567289352417, "noise_std_eff": 0.040649315714836126, "elapsed_sec": 1052.8643200397491}
262
+ {"step": 8240, "epoch": 1, "loss": 3.9244351387023926, "loss_en": 3.943943977355957, "loss_zh": 3.904926300048828, "noise_std_eff": 0.08827250599861146, "elapsed_sec": 1055.7857525348663}
263
+ {"step": 8260, "epoch": 1, "loss": 3.7378129959106445, "loss_en": 3.7694530487060547, "loss_zh": 3.7061731815338135, "noise_std_eff": 0.015230366587638856, "elapsed_sec": 1058.6925683021545}
264
+ {"step": 8280, "epoch": 1, "loss": 3.544168472290039, "loss_en": 3.515439033508301, "loss_zh": 3.5728981494903564, "noise_std_eff": 0.031587958335876465, "elapsed_sec": 1061.605580329895}
265
+ {"step": 8300, "epoch": 1, "loss": 4.111591339111328, "loss_en": 4.093311309814453, "loss_zh": 4.129871368408203, "noise_std_eff": 0.06989365220069886, "elapsed_sec": 1064.4658949375153}
266
+ {"step": 8320, "epoch": 1, "loss": 3.7736406326293945, "loss_en": 3.721524238586426, "loss_zh": 3.825756788253784, "noise_std_eff": 0.0895509958267212, "elapsed_sec": 1067.4040668010712}
267
+ {"step": 8340, "epoch": 1, "loss": 3.8625388145446777, "loss_en": 3.841007947921753, "loss_zh": 3.8840696811676025, "noise_std_eff": 0.06351990699768066, "elapsed_sec": 1070.2540085315704}
268
+ {"step": 8360, "epoch": 1, "loss": 3.4412453174591064, "loss_en": 3.4303534030914307, "loss_zh": 3.4521372318267822, "noise_std_eff": 0.03023695945739746, "elapsed_sec": 1073.1868071556091}
269
+ {"step": 8380, "epoch": 1, "loss": 3.808943271636963, "loss_en": 3.835609197616577, "loss_zh": 3.7822775840759277, "noise_std_eff": 0.014045092463493348, "elapsed_sec": 1076.1067621707916}
270
+ {"step": 8400, "epoch": 1, "loss": 3.7418713569641113, "loss_en": 3.733072519302368, "loss_zh": 3.7506704330444336, "noise_std_eff": 0.08303283452987671, "elapsed_sec": 1079.0336380004883}
271
+ {"step": 8420, "epoch": 1, "loss": 3.8356804847717285, "loss_en": 3.7841949462890625, "loss_zh": 3.8871660232543945, "noise_std_eff": 0.06539503931999206, "elapsed_sec": 1106.1023359298706}
272
+ {"step": 8440, "epoch": 1, "loss": 3.4886515140533447, "loss_en": 3.4772942066192627, "loss_zh": 3.5000088214874268, "noise_std_eff": 0.0008590526878833771, "elapsed_sec": 1109.0154540538788}
273
+ {"step": 8460, "epoch": 1, "loss": 3.2526278495788574, "loss_en": 3.2549779415130615, "loss_zh": 3.250277519226074, "noise_std_eff": 0.009582125395536423, "elapsed_sec": 1111.8788907527924}
274
+ {"step": 8480, "epoch": 1, "loss": 3.743441104888916, "loss_en": 3.71409010887146, "loss_zh": 3.772791862487793, "noise_std_eff": 0.03294204473495484, "elapsed_sec": 1114.838086605072}
275
+ {"step": 8500, "epoch": 1, "loss": 3.630321502685547, "loss_en": 3.6685917377471924, "loss_zh": 3.5920512676239014, "noise_std_eff": 0.047718736529350284, "elapsed_sec": 1117.7304451465607}
276
+ {"step": 8520, "epoch": 1, "loss": 4.230419635772705, "loss_en": 4.250317573547363, "loss_zh": 4.210521697998047, "noise_std_eff": 0.02067673057317734, "elapsed_sec": 1120.6625652313232}
277
+ {"step": 8540, "epoch": 1, "loss": 3.7328035831451416, "loss_en": 3.7327797412872314, "loss_zh": 3.7328274250030518, "noise_std_eff": 0.049210169911384584, "elapsed_sec": 1123.5583896636963}
278
+ {"step": 8560, "epoch": 1, "loss": 4.112390518188477, "loss_en": 4.165154933929443, "loss_zh": 4.05962610244751, "noise_std_eff": 0.03326732218265534, "elapsed_sec": 1126.4792580604553}
279
+ {"step": 8580, "epoch": 1, "loss": 4.163373947143555, "loss_en": 4.1309380531311035, "loss_zh": 4.195809841156006, "noise_std_eff": 0.03610822558403015, "elapsed_sec": 1129.355785369873}
280
+ {"step": 8600, "epoch": 1, "loss": 3.5848708152770996, "loss_en": 3.573042154312134, "loss_zh": 3.5966992378234863, "noise_std_eff": 0.0787243962287903, "elapsed_sec": 1132.2987744808197}
281
+ {"step": 8620, "epoch": 1, "loss": 4.012681484222412, "loss_en": 3.961442232131958, "loss_zh": 4.063920497894287, "noise_std_eff": 0.01985893696546555, "elapsed_sec": 1135.2124366760254}
282
+ {"step": 8640, "epoch": 1, "loss": 3.6543023586273193, "loss_en": 3.6619653701782227, "loss_zh": 3.646639347076416, "noise_std_eff": 0.04020561575889588, "elapsed_sec": 1138.1341016292572}
283
+ {"step": 8660, "epoch": 1, "loss": 3.513272523880005, "loss_en": 3.484212636947632, "loss_zh": 3.542332410812378, "noise_std_eff": 0.05388525128364563, "elapsed_sec": 1140.9989612102509}
284
+ {"step": 8680, "epoch": 1, "loss": 3.5340662002563477, "loss_en": 3.4789645671844482, "loss_zh": 3.589167833328247, "noise_std_eff": 0.09993416666984559, "elapsed_sec": 1143.9150202274323}
285
+ {"step": 8700, "epoch": 1, "loss": 3.0914011001586914, "loss_en": 3.1010963916778564, "loss_zh": 3.0817058086395264, "noise_std_eff": 0.09430249333381653, "elapsed_sec": 1146.81858420372}
286
+ {"step": 8720, "epoch": 1, "loss": 3.800922393798828, "loss_en": 3.833238124847412, "loss_zh": 3.768606662750244, "noise_std_eff": 0.0017678707838058472, "elapsed_sec": 1149.7471058368683}
287
+ {"step": 8740, "epoch": 1, "loss": 3.8056163787841797, "loss_en": 3.78627610206604, "loss_zh": 3.8249568939208984, "noise_std_eff": 0.01386040449142456, "elapsed_sec": 1152.6270163059235}
288
+ {"step": 8760, "epoch": 1, "loss": 4.071047782897949, "loss_en": 4.060496807098389, "loss_zh": 4.08159875869751, "noise_std_eff": 0.00520959310233593, "elapsed_sec": 1155.5716757774353}
289
+ {"step": 8780, "epoch": 1, "loss": 3.9843227863311768, "loss_en": 3.9267940521240234, "loss_zh": 4.04185152053833, "noise_std_eff": 0.04185356795787812, "elapsed_sec": 1158.4514861106873}
290
+ {"step": 8800, "epoch": 1, "loss": 3.7265591621398926, "loss_en": 3.727391004562378, "loss_zh": 3.725727081298828, "noise_std_eff": 0.05955263376235962, "elapsed_sec": 1161.3507704734802}
291
+ {"step": 8820, "epoch": 1, "loss": 3.3118433952331543, "loss_en": 3.291285514831543, "loss_zh": 3.3324012756347656, "noise_std_eff": 0.07901257276535034, "elapsed_sec": 1188.0026712417603}
292
+ {"step": 8840, "epoch": 1, "loss": 4.100512504577637, "loss_en": 4.104438781738281, "loss_zh": 4.096585750579834, "noise_std_eff": 0.01249251812696457, "elapsed_sec": 1190.9220435619354}
293
+ {"step": 8860, "epoch": 1, "loss": 3.6254312992095947, "loss_en": 3.6421751976013184, "loss_zh": 3.608687400817871, "noise_std_eff": 0.09726861715316773, "elapsed_sec": 1193.809360742569}
294
+ {"step": 8880, "epoch": 1, "loss": 4.050227165222168, "loss_en": 4.033214092254639, "loss_zh": 4.067240238189697, "noise_std_eff": 0.02873016893863678, "elapsed_sec": 1196.7469463348389}
295
+ {"step": 8900, "epoch": 1, "loss": 3.70131778717041, "loss_en": 3.7140393257141113, "loss_zh": 3.688596487045288, "noise_std_eff": 0.039335972070693975, "elapsed_sec": 1199.6536662578583}
296
+ {"step": 8920, "epoch": 1, "loss": 4.82855224609375, "loss_en": 4.8973469734191895, "loss_zh": 4.7597575187683105, "noise_std_eff": 0.09343634247779847, "elapsed_sec": 1202.5615675449371}
297
+ {"step": 8940, "epoch": 1, "loss": 3.9617409706115723, "loss_en": 3.9332315921783447, "loss_zh": 3.990250587463379, "noise_std_eff": 0.05541018843650818, "elapsed_sec": 1205.4306774139404}
298
+ {"step": 8960, "epoch": 1, "loss": 3.97458553314209, "loss_en": 3.9363231658935547, "loss_zh": 4.012847900390625, "noise_std_eff": 0.04417921006679535, "elapsed_sec": 1208.3501861095428}
299
+ {"step": 8980, "epoch": 1, "loss": 3.9722766876220703, "loss_en": 3.9378154277801514, "loss_zh": 4.00673770904541, "noise_std_eff": 0.017683137953281403, "elapsed_sec": 1211.306168794632}
300
+ {"step": 9000, "epoch": 1, "loss": 3.8520593643188477, "loss_en": 3.8437976837158203, "loss_zh": 3.860320806503296, "noise_std_eff": 0.05922219157218933, "elapsed_sec": 1214.2102725505829}
301
+ {"step": 9020, "epoch": 1, "loss": 3.6771652698516846, "loss_en": 3.6734752655029297, "loss_zh": 3.6808552742004395, "noise_std_eff": 0.0437646746635437, "elapsed_sec": 1217.0980427265167}
302
+ {"step": 9040, "epoch": 1, "loss": 3.569725751876831, "loss_en": 3.56965970993042, "loss_zh": 3.569791793823242, "noise_std_eff": 0.0779486119747162, "elapsed_sec": 1220.0108096599579}
303
+ {"step": 9060, "epoch": 1, "loss": 3.7148613929748535, "loss_en": 3.7368361949920654, "loss_zh": 3.6928865909576416, "noise_std_eff": 0.04699406027793884, "elapsed_sec": 1222.5843591690063}
304
+ {"step": 9080, "epoch": 1, "loss": 3.306452751159668, "loss_en": 3.3142213821411133, "loss_zh": 3.2986843585968018, "noise_std_eff": 0.037092337012290956, "elapsed_sec": 1224.70258975029}
305
+ {"step": 9100, "epoch": 1, "loss": 3.8735954761505127, "loss_en": 3.9153106212615967, "loss_zh": 3.8318803310394287, "noise_std_eff": 0.07852134108543396, "elapsed_sec": 1227.4702055454254}
306
+ {"step": 9120, "epoch": 1, "loss": 3.7449896335601807, "loss_en": 3.714818000793457, "loss_zh": 3.7751612663269043, "noise_std_eff": 0.08640040159225465, "elapsed_sec": 1230.4029150009155}
307
+ {"step": 9140, "epoch": 1, "loss": 3.960367202758789, "loss_en": 3.9279375076293945, "loss_zh": 3.9927966594696045, "noise_std_eff": 0.03794514238834382, "elapsed_sec": 1233.287302017212}
308
+ {"step": 9160, "epoch": 1, "loss": 3.6764352321624756, "loss_en": 3.6277852058410645, "loss_zh": 3.7250852584838867, "noise_std_eff": 0.012035230547189713, "elapsed_sec": 1236.213936805725}
309
+ {"step": 9180, "epoch": 1, "loss": 3.920943021774292, "loss_en": 3.9564013481140137, "loss_zh": 3.8854846954345703, "noise_std_eff": 0.051544690132141115, "elapsed_sec": 1239.1175773143768}
310
+ {"step": 9200, "epoch": 1, "loss": 3.685372829437256, "loss_en": 3.741286277770996, "loss_zh": 3.6294596195220947, "noise_std_eff": 0.01835634559392929, "elapsed_sec": 1242.0500349998474}
311
+ {"step": 9220, "epoch": 1, "loss": 3.531259536743164, "loss_en": 3.516256332397461, "loss_zh": 3.546262502670288, "noise_std_eff": 0.002895603142678738, "elapsed_sec": 1268.9301600456238}
312
+ {"step": 9240, "epoch": 1, "loss": 3.578312873840332, "loss_en": 3.584909439086914, "loss_zh": 3.57171630859375, "noise_std_eff": 0.08726067543029786, "elapsed_sec": 1271.8221108913422}
313
+ {"step": 9260, "epoch": 1, "loss": 3.181194543838501, "loss_en": 3.1895358562469482, "loss_zh": 3.1728532314300537, "noise_std_eff": 0.018740224838256835, "elapsed_sec": 1274.6908557415009}
314
+ {"step": 9280, "epoch": 1, "loss": 4.0054402351379395, "loss_en": 4.023370265960693, "loss_zh": 3.9875099658966064, "noise_std_eff": 0.010142586380243302, "elapsed_sec": 1277.5939407348633}
315
+ {"step": 9300, "epoch": 1, "loss": 3.972337245941162, "loss_en": 3.9634690284729004, "loss_zh": 3.981205701828003, "noise_std_eff": 0.0010060683824121952, "elapsed_sec": 1280.482771396637}
316
+ {"step": 9320, "epoch": 1, "loss": 3.1942851543426514, "loss_en": 3.195141315460205, "loss_zh": 3.1934289932250977, "noise_std_eff": 0.05028247833251953, "elapsed_sec": 1283.3614490032196}
317
+ {"step": 9340, "epoch": 1, "loss": 3.680271625518799, "loss_en": 3.617046356201172, "loss_zh": 3.743497133255005, "noise_std_eff": 0.051123481988906865, "elapsed_sec": 1286.2265441417694}
318
+ {"step": 9360, "epoch": 1, "loss": 3.46328067779541, "loss_en": 3.4854438304901123, "loss_zh": 3.441117525100708, "noise_std_eff": 0.08195280432701112, "elapsed_sec": 1289.130009651184}
319
+ {"step": 9380, "epoch": 1, "loss": 4.133525848388672, "loss_en": 4.142343044281006, "loss_zh": 4.124708652496338, "noise_std_eff": 0.0062904804944992065, "elapsed_sec": 1291.998866558075}
320
+ {"step": 9400, "epoch": 1, "loss": 3.994016170501709, "loss_en": 4.019174098968506, "loss_zh": 3.968858480453491, "noise_std_eff": 0.01899314671754837, "elapsed_sec": 1294.9252166748047}
321
+ {"step": 9420, "epoch": 1, "loss": 3.818162202835083, "loss_en": 3.81063175201416, "loss_zh": 3.825692653656006, "noise_std_eff": 0.05749322772026062, "elapsed_sec": 1297.810040473938}
322
+ {"step": 9440, "epoch": 1, "loss": 3.2396082878112793, "loss_en": 3.2922308444976807, "loss_zh": 3.186985731124878, "noise_std_eff": 0.05166313052177429, "elapsed_sec": 1300.742474079132}
323
+ {"step": 9460, "epoch": 1, "loss": 3.4295523166656494, "loss_en": 3.4432411193847656, "loss_zh": 3.415863513946533, "noise_std_eff": 0.04854913353919983, "elapsed_sec": 1303.6387977600098}
324
+ {"step": 9480, "epoch": 1, "loss": 3.558279514312744, "loss_en": 3.532487630844116, "loss_zh": 3.584071397781372, "noise_std_eff": 0.07009373307228088, "elapsed_sec": 1306.5462114810944}
325
+ {"step": 9500, "epoch": 1, "loss": 3.247934341430664, "loss_en": 3.297992467880249, "loss_zh": 3.197876453399658, "noise_std_eff": 0.020621006190776826, "elapsed_sec": 1309.4113323688507}
326
+ {"step": 9520, "epoch": 1, "loss": 3.6795248985290527, "loss_en": 3.6071712970733643, "loss_zh": 3.7518787384033203, "noise_std_eff": 0.06719552874565125, "elapsed_sec": 1312.3379580974579}
327
+ {"step": 9540, "epoch": 1, "loss": 4.046760559082031, "loss_en": 4.031491279602051, "loss_zh": 4.062029838562012, "noise_std_eff": 0.03748695850372315, "elapsed_sec": 1315.1287655830383}
328
+ {"step": 9560, "epoch": 1, "loss": 3.720766067504883, "loss_en": 3.661451816558838, "loss_zh": 3.7800803184509277, "noise_std_eff": 0.06338794231414795, "elapsed_sec": 1318.0690033435822}
329
+ {"step": 9580, "epoch": 1, "loss": 3.45519757270813, "loss_en": 3.4156734943389893, "loss_zh": 3.4947216510772705, "noise_std_eff": 0.08261491656303406, "elapsed_sec": 1320.939175605774}
330
+ {"step": 9600, "epoch": 1, "loss": 4.2550458908081055, "loss_en": 4.2809014320373535, "loss_zh": 4.229190826416016, "noise_std_eff": 0.042413833737373355, "elapsed_sec": 1323.8953793048859}
331
+ {"step": 9620, "epoch": 1, "loss": 3.8810110092163086, "loss_en": 3.8341939449310303, "loss_zh": 3.927828073501587, "noise_std_eff": 0.03271717727184296, "elapsed_sec": 1350.835245847702}
332
+ {"step": 9640, "epoch": 1, "loss": 3.958874225616455, "loss_en": 3.9613494873046875, "loss_zh": 3.9563992023468018, "noise_std_eff": 0.004385604709386826, "elapsed_sec": 1353.7711200714111}
333
+ {"step": 9660, "epoch": 1, "loss": 3.7717251777648926, "loss_en": 3.710686445236206, "loss_zh": 3.832763671875, "noise_std_eff": 0.0005932699888944626, "elapsed_sec": 1356.6315882205963}
334
+ {"step": 9680, "epoch": 1, "loss": 3.472569465637207, "loss_en": 3.461338996887207, "loss_zh": 3.483799934387207, "noise_std_eff": 0.029250001907348635, "elapsed_sec": 1359.5827195644379}
335
+ {"step": 9700, "epoch": 1, "loss": 3.106825113296509, "loss_en": 3.109557628631592, "loss_zh": 3.104092597961426, "noise_std_eff": 0.038186419010162356, "elapsed_sec": 1362.45201587677}
336
+ {"step": 9720, "epoch": 1, "loss": 3.4786057472229004, "loss_en": 3.430051326751709, "loss_zh": 3.5271599292755127, "noise_std_eff": 0.05477593541145325, "elapsed_sec": 1365.3631036281586}
337
+ {"step": 9740, "epoch": 1, "loss": 3.6689541339874268, "loss_en": 3.641533136367798, "loss_zh": 3.6963751316070557, "noise_std_eff": 0.07007312774658203, "elapsed_sec": 1368.2089939117432}
338
+ {"step": 9760, "epoch": 1, "loss": 3.2034897804260254, "loss_en": 3.187429904937744, "loss_zh": 3.2195494174957275, "noise_std_eff": 0.015748606622219087, "elapsed_sec": 1371.1075944900513}
339
+ {"step": 9780, "epoch": 1, "loss": 3.301723003387451, "loss_en": 3.280560255050659, "loss_zh": 3.3228859901428223, "noise_std_eff": 0.03947202265262604, "elapsed_sec": 1373.9782974720001}
340
+ {"step": 9800, "epoch": 1, "loss": 3.5934789180755615, "loss_en": 3.5584757328033447, "loss_zh": 3.6284821033477783, "noise_std_eff": 0.09810816645622254, "elapsed_sec": 1376.8857583999634}
341
+ {"step": 9820, "epoch": 1, "loss": 4.076752662658691, "loss_en": 4.072906494140625, "loss_zh": 4.0805983543396, "noise_std_eff": 0.025322949886322024, "elapsed_sec": 1379.7143063545227}
342
+ {"step": 9840, "epoch": 1, "loss": 3.958932876586914, "loss_en": 3.9180586338043213, "loss_zh": 3.9998068809509277, "noise_std_eff": 0.036765563488006595, "elapsed_sec": 1382.6177036762238}
343
+ {"step": 9860, "epoch": 1, "loss": 4.192903995513916, "loss_en": 4.131876468658447, "loss_zh": 4.253931522369385, "noise_std_eff": 0.08136175870895386, "elapsed_sec": 1385.5031321048737}
344
+ {"step": 9880, "epoch": 1, "loss": 4.177336692810059, "loss_en": 4.189936637878418, "loss_zh": 4.164736747741699, "noise_std_eff": 0.011671922355890275, "elapsed_sec": 1388.387945652008}
345
+ {"step": 9900, "epoch": 1, "loss": 3.3091187477111816, "loss_en": 3.341279983520508, "loss_zh": 3.2769577503204346, "noise_std_eff": 0.027223712205886843, "elapsed_sec": 1391.2512381076813}
346
+ {"step": 9920, "epoch": 1, "loss": 3.989631414413452, "loss_en": 4.0266499519348145, "loss_zh": 3.95261287689209, "noise_std_eff": 0.06537860631942749, "elapsed_sec": 1394.1379375457764}
347
+ {"step": 9940, "epoch": 1, "loss": 3.89479923248291, "loss_en": 3.8951456546783447, "loss_zh": 3.8944525718688965, "noise_std_eff": 0.030731332302093507, "elapsed_sec": 1396.9748375415802}
348
+ {"step": 9960, "epoch": 1, "loss": 3.378223419189453, "loss_en": 3.378347635269165, "loss_zh": 3.378099203109741, "noise_std_eff": 0.005718488246202469, "elapsed_sec": 1399.9174754619598}
349
+ {"step": 9980, "epoch": 1, "loss": 3.5565905570983887, "loss_en": 3.5286920070648193, "loss_zh": 3.584489345550537, "noise_std_eff": 0.009595926105976106, "elapsed_sec": 1402.7991681098938}
350
+ {"step": 10000, "epoch": 1, "loss": 3.0504584312438965, "loss_en": 3.058213949203491, "loss_zh": 3.0427026748657227, "noise_std_eff": 0.04697500467300415, "elapsed_sec": 1405.709993839264}
351
+ {"step": 10020, "epoch": 1, "loss": 4.186410427093506, "loss_en": 4.161046028137207, "loss_zh": 4.211774826049805, "noise_std_eff": 0.07053865790367127, "elapsed_sec": 1432.2681703567505}
352
+ {"step": 10040, "epoch": 1, "loss": 3.322239875793457, "loss_en": 3.3408095836639404, "loss_zh": 3.3036704063415527, "noise_std_eff": 0.03689582049846649, "elapsed_sec": 1434.701054573059}
353
+ {"step": 10060, "epoch": 1, "loss": 3.875682830810547, "loss_en": 3.892967939376831, "loss_zh": 3.8583977222442627, "noise_std_eff": 0.08576672673225404, "elapsed_sec": 1437.3787014484406}
354
+ {"step": 10080, "epoch": 1, "loss": 3.4531478881835938, "loss_en": 3.4310247898101807, "loss_zh": 3.475271224975586, "noise_std_eff": 0.03710196316242218, "elapsed_sec": 1439.7650680541992}
355
+ {"step": 10100, "epoch": 1, "loss": 3.4193525314331055, "loss_en": 3.4343676567077637, "loss_zh": 3.4043376445770264, "noise_std_eff": 0.03091442584991455, "elapsed_sec": 1442.0598616600037}
356
+ {"step": 10120, "epoch": 1, "loss": 3.3212594985961914, "loss_en": 3.39028000831604, "loss_zh": 3.2522389888763428, "noise_std_eff": 0.062117463350296026, "elapsed_sec": 1444.8447086811066}
357
+ {"step": 10140, "epoch": 1, "loss": 3.982513427734375, "loss_en": 4.006033897399902, "loss_zh": 3.9589927196502686, "noise_std_eff": 0.023282925784587863, "elapsed_sec": 1447.2893345355988}
358
+ {"step": 10160, "epoch": 1, "loss": 3.705012559890747, "loss_en": 3.613847017288208, "loss_zh": 3.796178102493286, "noise_std_eff": 0.0214910164475441, "elapsed_sec": 1450.016276359558}
359
+ {"step": 10180, "epoch": 1, "loss": 3.5634491443634033, "loss_en": 3.548682451248169, "loss_zh": 3.5782158374786377, "noise_std_eff": 0.020824865996837617, "elapsed_sec": 1452.2967100143433}
360
+ {"step": 10200, "epoch": 1, "loss": 3.7070157527923584, "loss_en": 3.6760451793670654, "loss_zh": 3.7379863262176514, "noise_std_eff": 0.005172419175505639, "elapsed_sec": 1454.847262620926}
361
+ {"step": 10220, "epoch": 1, "loss": 3.8473000526428223, "loss_en": 3.818227529525757, "loss_zh": 3.876372814178467, "noise_std_eff": 0.004251835867762566, "elapsed_sec": 1457.1105816364288}
362
+ {"step": 10240, "epoch": 1, "loss": 3.2116713523864746, "loss_en": 3.2085657119750977, "loss_zh": 3.2147769927978516, "noise_std_eff": 0.04285654425621033, "elapsed_sec": 1459.5634124279022}
363
+ {"step": 10260, "epoch": 1, "loss": 3.678584098815918, "loss_en": 3.666261911392212, "loss_zh": 3.690906286239624, "noise_std_eff": 0.06147485971450806, "elapsed_sec": 1462.075234413147}
364
+ {"step": 10280, "epoch": 1, "loss": 4.311173439025879, "loss_en": 4.253952503204346, "loss_zh": 4.368394374847412, "noise_std_eff": 0.035890936851501465, "elapsed_sec": 1464.6743083000183}
365
+ {"step": 10300, "epoch": 1, "loss": 4.178630828857422, "loss_en": 4.185459613800049, "loss_zh": 4.171802043914795, "noise_std_eff": 0.08785259127616883, "elapsed_sec": 1467.2793581485748}
366
+ {"step": 10320, "epoch": 1, "loss": 3.453090190887451, "loss_en": 3.4476828575134277, "loss_zh": 3.4584975242614746, "noise_std_eff": 0.06023868918418884, "elapsed_sec": 1469.789680480957}
367
+ {"step": 10340, "epoch": 1, "loss": 3.599801540374756, "loss_en": 3.6012401580810547, "loss_zh": 3.598362922668457, "noise_std_eff": 0.02978309988975525, "elapsed_sec": 1472.6568386554718}
368
+ {"step": 10360, "epoch": 1, "loss": 3.533884286880493, "loss_en": 3.5513253211975098, "loss_zh": 3.5164432525634766, "noise_std_eff": 0.011494288593530655, "elapsed_sec": 1475.5914220809937}
369
+ {"step": 10380, "epoch": 1, "loss": 3.6984286308288574, "loss_en": 3.6855578422546387, "loss_zh": 3.711299180984497, "noise_std_eff": 0.09331399202346802, "elapsed_sec": 1478.4339408874512}
370
+ {"step": 10400, "epoch": 1, "loss": 3.6462841033935547, "loss_en": 3.6473500728607178, "loss_zh": 3.6452183723449707, "noise_std_eff": 0.07363440990447999, "elapsed_sec": 1481.3458604812622}
371
+ {"step": 10420, "epoch": 1, "loss": 4.465914249420166, "loss_en": 4.476414203643799, "loss_zh": 4.455414295196533, "noise_std_eff": 0.06051921844482422, "elapsed_sec": 1508.1561682224274}
372
+ {"step": 10440, "epoch": 1, "loss": 3.651386260986328, "loss_en": 3.641475200653076, "loss_zh": 3.661297082901001, "noise_std_eff": 0.09474236369132996, "elapsed_sec": 1511.0977380275726}
373
+ {"step": 10460, "epoch": 1, "loss": 3.821646213531494, "loss_en": 3.815573215484619, "loss_zh": 3.8277194499969482, "noise_std_eff": 0.037922865152359014, "elapsed_sec": 1513.9348974227905}
374
+ {"step": 10480, "epoch": 1, "loss": 3.529566764831543, "loss_en": 3.5421104431152344, "loss_zh": 3.5170228481292725, "noise_std_eff": 0.003670328110456467, "elapsed_sec": 1516.874336719513}
375
+ {"step": 10500, "epoch": 1, "loss": 4.075508117675781, "loss_en": 4.085723400115967, "loss_zh": 4.0652923583984375, "noise_std_eff": 0.08261539340019226, "elapsed_sec": 1519.7584717273712}
376
+ {"step": 10520, "epoch": 1, "loss": 4.125009059906006, "loss_en": 4.113340854644775, "loss_zh": 4.136677265167236, "noise_std_eff": 0.012150155007839204, "elapsed_sec": 1522.648530960083}
377
+ {"step": 10540, "epoch": 1, "loss": 3.5355653762817383, "loss_en": 3.5390429496765137, "loss_zh": 3.532087564468384, "noise_std_eff": 0.0746106505393982, "elapsed_sec": 1525.5116531848907}
378
+ {"step": 10560, "epoch": 1, "loss": 4.236502170562744, "loss_en": 4.230234146118164, "loss_zh": 4.242770195007324, "noise_std_eff": 0.06080980300903321, "elapsed_sec": 1528.4196817874908}
379
+ {"step": 10580, "epoch": 1, "loss": 3.4051461219787598, "loss_en": 3.41838002204895, "loss_zh": 3.3919122219085693, "noise_std_eff": 0.0675950586795807, "elapsed_sec": 1531.2705976963043}
380
+ {"step": 10600, "epoch": 1, "loss": 3.6814541816711426, "loss_en": 3.668816328048706, "loss_zh": 3.694091796875, "noise_std_eff": 0.07816383242607117, "elapsed_sec": 1534.1506762504578}
381
+ {"step": 10620, "epoch": 1, "loss": 3.2788872718811035, "loss_en": 3.2794883251190186, "loss_zh": 3.2782862186431885, "noise_std_eff": 0.012157982587814331, "elapsed_sec": 1537.030476808548}
382
+ {"step": 10640, "epoch": 1, "loss": 3.7223801612854004, "loss_en": 3.746772527694702, "loss_zh": 3.6979875564575195, "noise_std_eff": 0.06229730248451233, "elapsed_sec": 1539.9490032196045}
383
+ {"step": 10660, "epoch": 1, "loss": 3.856074094772339, "loss_en": 3.8740904331207275, "loss_zh": 3.83805775642395, "noise_std_eff": 0.07467631101608277, "elapsed_sec": 1542.7983338832855}
384
+ {"step": 10680, "epoch": 1, "loss": 4.088120460510254, "loss_en": 4.05402946472168, "loss_zh": 4.122211933135986, "noise_std_eff": 0.003943268582224846, "elapsed_sec": 1545.8101348876953}
385
+ {"step": 10700, "epoch": 1, "loss": 3.562570095062256, "loss_en": 3.573585271835327, "loss_zh": 3.5515551567077637, "noise_std_eff": 0.04008565545082093, "elapsed_sec": 1548.6789927482605}
386
+ {"step": 10720, "epoch": 1, "loss": 3.8575775623321533, "loss_en": 3.9196908473968506, "loss_zh": 3.795464277267456, "noise_std_eff": 0.02123248726129532, "elapsed_sec": 1551.6028988361359}
387
+ {"step": 10740, "epoch": 1, "loss": 3.733478307723999, "loss_en": 3.7023844718933105, "loss_zh": 3.7645721435546875, "noise_std_eff": 0.09774198532104493, "elapsed_sec": 1554.494701385498}
388
+ {"step": 10760, "epoch": 1, "loss": 3.940499782562256, "loss_en": 3.915944814682007, "loss_zh": 3.965054512023926, "noise_std_eff": 0.09388702511787415, "elapsed_sec": 1557.4075968265533}
389
+ {"step": 10780, "epoch": 1, "loss": 4.0506696701049805, "loss_en": 4.013960838317871, "loss_zh": 4.087378978729248, "noise_std_eff": 0.042225399613380434, "elapsed_sec": 1560.2369167804718}
390
+ {"step": 10800, "epoch": 1, "loss": 3.3718109130859375, "loss_en": 3.3494532108306885, "loss_zh": 3.3941686153411865, "noise_std_eff": 0.008522410690784455, "elapsed_sec": 1563.1108376979828}
391
+ {"step": 10820, "epoch": 1, "loss": 3.911465883255005, "loss_en": 3.9770829677581787, "loss_zh": 3.845848798751831, "noise_std_eff": 0.04120121002197266, "elapsed_sec": 1590.3298707008362}
392
+ {"step": 10840, "epoch": 1, "loss": 3.5005104541778564, "loss_en": 3.498476505279541, "loss_zh": 3.502544403076172, "noise_std_eff": 0.09216528534889222, "elapsed_sec": 1593.2424504756927}
393
+ {"step": 10860, "epoch": 1, "loss": 3.956484317779541, "loss_en": 4.00261926651001, "loss_zh": 3.9103493690490723, "noise_std_eff": 0.027221450209617616, "elapsed_sec": 1596.1137342453003}
394
+ {"step": 10880, "epoch": 1, "loss": 3.6560168266296387, "loss_en": 3.643190383911133, "loss_zh": 3.6688430309295654, "noise_std_eff": 0.0478312224149704, "elapsed_sec": 1599.0101673603058}
395
+ {"step": 10900, "epoch": 1, "loss": 3.617995023727417, "loss_en": 3.6102750301361084, "loss_zh": 3.6257150173187256, "noise_std_eff": 0.010822129249572755, "elapsed_sec": 1601.8702204227448}
396
+ {"step": 10920, "epoch": 1, "loss": 3.8382625579833984, "loss_en": 3.8667714595794678, "loss_zh": 3.80975341796875, "noise_std_eff": 0.05308158397674561, "elapsed_sec": 1604.7664215564728}
397
+ {"step": 10940, "epoch": 1, "loss": 3.6023526191711426, "loss_en": 3.529086112976074, "loss_zh": 3.675618886947632, "noise_std_eff": 0.012410853803157807, "elapsed_sec": 1607.626708984375}
398
+ {"step": 10960, "epoch": 1, "loss": 3.511279821395874, "loss_en": 3.451533079147339, "loss_zh": 3.571026563644409, "noise_std_eff": 0.03940794169902802, "elapsed_sec": 1610.5268294811249}
399
+ {"step": 10980, "epoch": 1, "loss": 3.876275062561035, "loss_en": 3.8408992290496826, "loss_zh": 3.9116508960723877, "noise_std_eff": 0.07882388830184937, "elapsed_sec": 1613.3747510910034}
400
+ {"step": 11000, "epoch": 1, "loss": 3.510075569152832, "loss_en": 3.514531135559082, "loss_zh": 3.505620241165161, "noise_std_eff": 0.05088722109794617, "elapsed_sec": 1616.2340242862701}
401
+ {"step": 11020, "epoch": 1, "loss": 3.8704159259796143, "loss_en": 3.8158023357391357, "loss_zh": 3.9250295162200928, "noise_std_eff": 0.09691095352172852, "elapsed_sec": 1619.096536397934}
402
+ {"step": 11040, "epoch": 1, "loss": 3.6468892097473145, "loss_en": 3.6285512447357178, "loss_zh": 3.665227174758911, "noise_std_eff": 0.09563291072845459, "elapsed_sec": 1621.9643425941467}
403
+ {"step": 11060, "epoch": 1, "loss": 3.566514015197754, "loss_en": 3.5571811199188232, "loss_zh": 3.5758471488952637, "noise_std_eff": 0.04678272008895874, "elapsed_sec": 1624.8427057266235}
404
+ {"step": 11080, "epoch": 1, "loss": 3.5482449531555176, "loss_en": 3.5395686626434326, "loss_zh": 3.5569212436676025, "noise_std_eff": 0.0021537886932492256, "elapsed_sec": 1627.7362313270569}
405
+ {"step": 11100, "epoch": 1, "loss": 4.175683975219727, "loss_en": 4.213717460632324, "loss_zh": 4.137650012969971, "noise_std_eff": 0.053604358434677125, "elapsed_sec": 1630.5796358585358}
406
+ {"step": 11120, "epoch": 1, "loss": 4.1378936767578125, "loss_en": 4.083233833312988, "loss_zh": 4.192553997039795, "noise_std_eff": 0.09018648862838746, "elapsed_sec": 1633.4752385616302}
407
+ {"step": 11140, "epoch": 1, "loss": 3.858964443206787, "loss_en": 3.8829123973846436, "loss_zh": 3.8350167274475098, "noise_std_eff": 0.031493619084358215, "elapsed_sec": 1636.3604941368103}
408
+ {"step": 11160, "epoch": 1, "loss": 3.9347922801971436, "loss_en": 3.9584388732910156, "loss_zh": 3.9111456871032715, "noise_std_eff": 0.06307595372200013, "elapsed_sec": 1639.2793066501617}
409
+ {"step": 11180, "epoch": 1, "loss": 3.851994514465332, "loss_en": 3.863992214202881, "loss_zh": 3.8399970531463623, "noise_std_eff": 0.036253973841667175, "elapsed_sec": 1642.1691496372223}
410
+ {"step": 11200, "epoch": 1, "loss": 3.5735108852386475, "loss_en": 3.5345494747161865, "loss_zh": 3.6124722957611084, "noise_std_eff": 0.007094521075487137, "elapsed_sec": 1645.064216375351}
411
+ {"step": 11220, "epoch": 1, "loss": 3.698060989379883, "loss_en": 3.6739487648010254, "loss_zh": 3.7221732139587402, "noise_std_eff": 0.02661395370960236, "elapsed_sec": 1671.9927468299866}
412
+ {"step": 11240, "epoch": 1, "loss": 3.6327571868896484, "loss_en": 3.6596102714538574, "loss_zh": 3.6059041023254395, "noise_std_eff": 0.08873094320297242, "elapsed_sec": 1674.8998641967773}
413
+ {"step": 11260, "epoch": 1, "loss": 3.467738628387451, "loss_en": 3.4705491065979004, "loss_zh": 3.464928150177002, "noise_std_eff": 0.0859546959400177, "elapsed_sec": 1677.7818429470062}
414
+ {"step": 11280, "epoch": 1, "loss": 3.695219039916992, "loss_en": 3.7197530269622803, "loss_zh": 3.670685291290283, "noise_std_eff": 0.06977440714836121, "elapsed_sec": 1680.6746730804443}
415
+ {"step": 11300, "epoch": 1, "loss": 4.147974491119385, "loss_en": 4.160583972930908, "loss_zh": 4.135365009307861, "noise_std_eff": 0.03708575665950775, "elapsed_sec": 1683.5299818515778}
416
+ {"step": 11320, "epoch": 1, "loss": 3.737264394760132, "loss_en": 3.6602587699890137, "loss_zh": 3.81427001953125, "noise_std_eff": 0.04981276988983155, "elapsed_sec": 1686.437955379486}
417
+ {"step": 11340, "epoch": 1, "loss": 3.8108506202697754, "loss_en": 3.7865521907806396, "loss_zh": 3.835149049758911, "noise_std_eff": 0.06676542162895203, "elapsed_sec": 1689.2933859825134}
418
+ {"step": 11360, "epoch": 1, "loss": 3.897794723510742, "loss_en": 3.9177114963531494, "loss_zh": 3.877877712249756, "noise_std_eff": 0.011227068305015565, "elapsed_sec": 1692.195570230484}
419
+ {"step": 11380, "epoch": 1, "loss": 3.7160732746124268, "loss_en": 3.7800204753875732, "loss_zh": 3.6521260738372803, "noise_std_eff": 0.027089133858680725, "elapsed_sec": 1695.073510169983}
420
+ {"step": 11400, "epoch": 1, "loss": 3.458230495452881, "loss_en": 3.4688756465911865, "loss_zh": 3.447585105895996, "noise_std_eff": 0.09519551396369935, "elapsed_sec": 1697.962674856186}
421
+ {"step": 11420, "epoch": 1, "loss": 3.622724771499634, "loss_en": 3.627143621444702, "loss_zh": 3.6183059215545654, "noise_std_eff": 0.026788491010665896, "elapsed_sec": 1700.8263192176819}
422
+ {"step": 11440, "epoch": 1, "loss": 3.2644500732421875, "loss_en": 3.2805657386779785, "loss_zh": 3.2483344078063965, "noise_std_eff": 0.07969592809677124, "elapsed_sec": 1703.7267677783966}
423
+ {"step": 11460, "epoch": 1, "loss": 3.462134838104248, "loss_en": 3.4742674827575684, "loss_zh": 3.450002431869507, "noise_std_eff": 0.010133448243141175, "elapsed_sec": 1706.5931596755981}
424
+ {"step": 11480, "epoch": 1, "loss": 3.508009433746338, "loss_en": 3.4877254962921143, "loss_zh": 3.5282933712005615, "noise_std_eff": 0.03932346999645234, "elapsed_sec": 1709.4898228645325}
425
+ {"step": 11500, "epoch": 1, "loss": 3.708055257797241, "loss_en": 3.701148271560669, "loss_zh": 3.7149622440338135, "noise_std_eff": 0.07697247862815858, "elapsed_sec": 1712.3342645168304}
426
+ {"step": 11520, "epoch": 1, "loss": 4.214751243591309, "loss_en": 4.229159832000732, "loss_zh": 4.200343132019043, "noise_std_eff": 0.05221159458160401, "elapsed_sec": 1715.2541272640228}
427
+ {"step": 11540, "epoch": 1, "loss": 3.6519241333007812, "loss_en": 3.6477293968200684, "loss_zh": 3.6561191082000732, "noise_std_eff": 0.04590151906013489, "elapsed_sec": 1718.1150097846985}
428
+ {"step": 11560, "epoch": 1, "loss": 3.5742130279541016, "loss_en": 3.5737879276275635, "loss_zh": 3.5746383666992188, "noise_std_eff": 0.025521644949913026, "elapsed_sec": 1721.0146045684814}
429
+ {"step": 11580, "epoch": 1, "loss": 3.727097511291504, "loss_en": 3.7252678871154785, "loss_zh": 3.7289273738861084, "noise_std_eff": 0.0992542326450348, "elapsed_sec": 1723.874629020691}
430
+ {"step": 11600, "epoch": 1, "loss": 3.592559337615967, "loss_en": 3.6851003170013428, "loss_zh": 3.50001859664917, "noise_std_eff": 0.046543252468109135, "elapsed_sec": 1726.7623755931854}
431
+ {"step": 11620, "epoch": 1, "loss": 3.548936367034912, "loss_en": 3.5720255374908447, "loss_zh": 3.5258474349975586, "noise_std_eff": 0.06815584301948548, "elapsed_sec": 1753.7002654075623}
432
+ {"step": 11640, "epoch": 1, "loss": 4.906235694885254, "loss_en": 4.9103922843933105, "loss_zh": 4.9020795822143555, "noise_std_eff": 0.030637085437774658, "elapsed_sec": 1756.618592262268}
433
+ {"step": 11660, "epoch": 1, "loss": 3.7427895069122314, "loss_en": 3.7455673217773438, "loss_zh": 3.740011692047119, "noise_std_eff": 0.0037908781319856647, "elapsed_sec": 1759.461948633194}
434
+ {"step": 11680, "epoch": 1, "loss": 3.5504767894744873, "loss_en": 3.6045877933502197, "loss_zh": 3.496365785598755, "noise_std_eff": 0.0188378781080246, "elapsed_sec": 1762.3468811511993}
435
+ {"step": 11700, "epoch": 1, "loss": 3.089099884033203, "loss_en": 3.0355634689331055, "loss_zh": 3.14263653755188, "noise_std_eff": 0.027173948287963868, "elapsed_sec": 1765.1943097114563}
436
+ {"step": 11720, "epoch": 1, "loss": 3.826079845428467, "loss_en": 3.8068153858184814, "loss_zh": 3.8453445434570312, "noise_std_eff": 0.00781858041882515, "elapsed_sec": 1768.0784730911255}
437
+ {"step": 11740, "epoch": 1, "loss": 3.987473487854004, "loss_en": 3.994858741760254, "loss_zh": 3.980088233947754, "noise_std_eff": 0.09829875826835632, "elapsed_sec": 1770.9535121917725}
438
+ {"step": 11760, "epoch": 1, "loss": 3.173861265182495, "loss_en": 3.0742571353912354, "loss_zh": 3.273465394973755, "noise_std_eff": 0.08957258462905884, "elapsed_sec": 1773.8248059749603}
439
+ {"step": 11780, "epoch": 1, "loss": 4.059998989105225, "loss_en": 4.0758442878723145, "loss_zh": 4.044153690338135, "noise_std_eff": 0.09418490529060364, "elapsed_sec": 1776.713751077652}
440
+ {"step": 11800, "epoch": 1, "loss": 2.979100227355957, "loss_en": 2.96765398979187, "loss_zh": 2.990546226501465, "noise_std_eff": 0.03524700701236725, "elapsed_sec": 1779.6103131771088}
441
+ {"step": 11820, "epoch": 1, "loss": 3.653884172439575, "loss_en": 3.641037940979004, "loss_zh": 3.6667304039001465, "noise_std_eff": 0.016752034425735474, "elapsed_sec": 1782.4550454616547}
442
+ {"step": 11840, "epoch": 1, "loss": 3.82308292388916, "loss_en": 3.855865240097046, "loss_zh": 3.7903006076812744, "noise_std_eff": 0.04375113248825074, "elapsed_sec": 1785.346336364746}
443
+ {"step": 11860, "epoch": 1, "loss": 3.6194210052490234, "loss_en": 3.6537036895751953, "loss_zh": 3.5851380825042725, "noise_std_eff": 0.0002006014343351126, "elapsed_sec": 1788.1825757026672}
444
+ {"step": 11880, "epoch": 1, "loss": 3.7634963989257812, "loss_en": 3.7472646236419678, "loss_zh": 3.7797281742095947, "noise_std_eff": 0.01716257929801941, "elapsed_sec": 1791.1086161136627}
445
+ {"step": 11900, "epoch": 1, "loss": 3.736603260040283, "loss_en": 3.7474944591522217, "loss_zh": 3.7257120609283447, "noise_std_eff": 0.05121246576309205, "elapsed_sec": 1793.9654297828674}
446
+ {"step": 11920, "epoch": 1, "loss": 3.5314106941223145, "loss_en": 3.5253682136535645, "loss_zh": 3.5374534130096436, "noise_std_eff": 0.04475103318691254, "elapsed_sec": 1796.902948141098}
447
+ {"step": 11940, "epoch": 1, "loss": 3.3263025283813477, "loss_en": 3.2672080993652344, "loss_zh": 3.385396957397461, "noise_std_eff": 0.022478100657463074, "elapsed_sec": 1799.7505147457123}
448
+ {"step": 11960, "epoch": 1, "loss": 4.054657936096191, "loss_en": 4.093346118927002, "loss_zh": 4.015969753265381, "noise_std_eff": 0.07551268339157105, "elapsed_sec": 1802.6378865242004}
planB_polish_robust_dec_noise10/probe_leakage.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_polish_robust_dec_noise10/ckpt.pt",
3
+ "max_train_groups": 8000,
4
+ "max_valid_groups": 946,
5
+ "probe_hidden": 256,
6
+ "epochs": 20,
7
+ "lr": 0.01,
8
+ "baseline_majority_lang_acc": 0.5,
9
+ "baseline_majority_len_acc": 0.9271249771118164,
10
+ "lang_probe": {
11
+ "train_acc": 0.589875,
12
+ "valid_acc": 0.5856236786469344,
13
+ "best_valid_acc": 0.5983086680761099
14
+ },
15
+ "len_probe": {
16
+ "train_acc": 0.92875,
17
+ "valid_acc": 0.9011627906976745,
18
+ "best_valid_acc": 0.9011627906976745
19
+ },
20
+ "lang_probe_length_matched": {
21
+ "train_acc": 0.581898766700925,
22
+ "valid_acc": 0.5843439911797134,
23
+ "best_valid_acc": 0.6008820286659317
24
+ },
25
+ "lang_bucketwise_valid_acc": {
26
+ "bucket_0_acc": 0.40625,
27
+ "bucket_1_acc": 0.574999988079071,
28
+ "bucket_2_acc": 0.5565217137336731,
29
+ "bucket_3_acc": 0.5912023782730103
30
+ },
31
+ "length_matched_sizes": {
32
+ "train": 15568,
33
+ "valid": 1814
34
+ }
35
+ }
planB_robust/ckpt.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7261050d9314ae32e9557ceb5a11f8f0cce6d6b7dd76a594ee9f4c7b97cd7e2
3
+ size 3646627249
planB_robust/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_jsonl": "data/groups_train.jsonl",
3
+ "valid_jsonl": "data/groups_valid.jsonl",
4
+ "run_dir": "runs/planB_robust",
5
+ "resume": "runs/planB_polish/ckpt.pt",
6
+ "backbone": "google/mt5-small",
7
+ "num_latents": 16,
8
+ "latent_dropout": 0.05,
9
+ "latent_noise_std": 0.01,
10
+ "batch_size": 4,
11
+ "grad_accum": 8,
12
+ "epochs": 1,
13
+ "max_doc_len": 256,
14
+ "max_sum_len": 64,
15
+ "eval_every": 400,
16
+ "max_train_examples": 2000,
17
+ "max_valid_examples": 200,
18
+ "lambda_align": 0.5,
19
+ "tau": 0.07,
20
+ "lambda_varcov": 10.0,
21
+ "var_target_std": 0.05,
22
+ "lambda_mean": 0.1,
23
+ "lambda_mean_diff": 0.1,
24
+ "lambda_pair": 0.2,
25
+ "lambda_lang": 0.5,
26
+ "lambda_len": 1.0,
27
+ "adv_start_step": 0,
28
+ "grl_alpha": 1.0,
29
+ "grl_warmup": 200,
30
+ "lr_model": 0.0001,
31
+ "lr_lang": 0.001,
32
+ "lr_len": 0.001,
33
+ "adv_clf_steps": 5,
34
+ "adv_clf_weight_decay": 0.0,
35
+ "adv_queue_size": 4096,
36
+ "adv_clf_batch": 256,
37
+ "adv_mix_current": 0.5
38
+ }
planB_robust/diag.jsonl ADDED
@@ -0,0 +1 @@
 
 
1
+ {"step": 3200, "nll_en": 4.761491537094116, "nll_zh": 4.767379446029663, "nll": 4.76443549156189, "lang_acc_valid_evalmode": 0.505, "len_acc_valid_evalmode": 0.895}
planB_robust/diag_full.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_robust/ckpt.pt",
3
+ "nll_en": 4.721819901819209,
4
+ "nll_zh": 4.722286191097526,
5
+ "swap_delta_en": 0.4393228224417624,
6
+ "swap_delta_zh": 0.44674703989170067,
7
+ "ablate_zero_delta_en": 0.6053643881896585,
8
+ "ablate_mean_delta_en": 0.1964060761207758,
9
+ "ablate_noise_delta_en": 2.6808454793805048,
10
+ "ablate_zero_delta_zh": 0.6048980989113418,
11
+ "ablate_mean_delta_zh": 0.20642961155284534,
12
+ "ablate_noise_delta_zh": 2.7120590088987453,
13
+ "inv_top1_full": 0.02431289665400982,
14
+ "inv_top5_full": 0.09936574846506119,
15
+ "diag_sim_mean": 0.9522932171821594,
16
+ "offdiag_sim_mean": 0.46395495533943176,
17
+ "sim_margin": 0.48833826184272766,
18
+ "n_valid": 946
19
+ }
planB_robust/logs.jsonl ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 3020, "epoch": 0, "loss": 6.876278400421143, "loss_render": 5.574617385864258, "loss_align": 0.3832191526889801, "loss_varcov": 0.0555385947227478, "loss_mean": 0.7199406623840332, "loss_mean_diff": 0.010353880003094673, "mean_diff_l2": 0.10175401717424393, "loss_pair": 0.00015211131540127099, "loss_lang": 0.8553144931793213, "loss_len": 0.05394825339317322, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6785714626312256, "len_acc_clf_detached_mix": 0.8571428656578064, "adv_buf_fill": 160, "elapsed_sec": 8.78326153755188}
2
+ {"step": 3040, "epoch": 0, "loss": 9.037178039550781, "loss_render": 5.321714401245117, "loss_align": 1.336050271987915, "loss_varcov": 0.055316247045993805, "loss_mean": 0.6666261553764343, "loss_mean_diff": 0.14929142594337463, "mean_diff_l2": 0.38638249039649963, "loss_pair": 0.0005262988852337003, "loss_lang": 2.7093873023986816, "loss_len": 1.0578848123550415, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.75390625, "len_acc_clf_detached_mix": 0.91015625, "adv_buf_fill": 320, "elapsed_sec": 15.842936515808105}
3
+ {"step": 3060, "epoch": 0, "loss": 7.815122127532959, "loss_render": 5.638380527496338, "loss_align": 0.0609850250184536, "loss_varcov": 0.042207784950733185, "loss_mean": 0.48525118827819824, "loss_mean_diff": 0.004880893509835005, "mean_diff_l2": 0.06986339390277863, "loss_pair": 4.3406202166806906e-05, "loss_lang": 2.4034109115600586, "loss_len": 0.47344404458999634, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.875, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.875, "lang_acc_clf_detached_mix": 0.734375, "len_acc_clf_detached_mix": 0.91015625, "adv_buf_fill": 480, "elapsed_sec": 22.870468378067017}
4
+ {"step": 3080, "epoch": 0, "loss": 7.281988620758057, "loss_render": 5.997339248657227, "loss_align": 0.3001667559146881, "loss_varcov": 0.04566030949354172, "loss_mean": 0.5067746639251709, "loss_mean_diff": 0.004238659515976906, "mean_diff_l2": 0.06510499119758606, "loss_pair": 5.6281940487679094e-05, "loss_lang": 1.0878978967666626, "loss_len": 0.08290104568004608, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.69921875, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 640, "elapsed_sec": 31.33722496032715}
5
+ {"step": 3100, "epoch": 0, "loss": 7.5562944412231445, "loss_render": 5.1810760498046875, "loss_align": 0.42862170934677124, "loss_varcov": 0.05040165036916733, "loss_mean": 0.6074227094650269, "loss_mean_diff": 0.013583110645413399, "mean_diff_l2": 0.11654660105705261, "loss_pair": 0.0001526661217212677, "loss_lang": 1.676601767539978, "loss_len": 0.7564594149589539, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.625, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.625, "lang_acc_clf_detached_mix": 0.625, "len_acc_clf_detached_mix": 0.9296875, "adv_buf_fill": 800, "elapsed_sec": 39.41167426109314}
6
+ {"step": 3120, "epoch": 0, "loss": 6.675807952880859, "loss_render": 4.7243428230285645, "loss_align": 0.6009824275970459, "loss_varcov": 0.04853154718875885, "loss_mean": 0.507441520690918, "loss_mean_diff": 0.00899539329111576, "mean_diff_l2": 0.09484405070543289, "loss_pair": 8.519425318809226e-05, "loss_lang": 1.939558506011963, "loss_len": 0.14421868324279785, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6015625, "len_acc_clf_detached_mix": 0.91796875, "adv_buf_fill": 960, "elapsed_sec": 46.780131578445435}
7
+ {"step": 3140, "epoch": 0, "loss": 6.587226390838623, "loss_render": 4.884444236755371, "loss_align": 0.34145528078079224, "loss_varcov": 0.0480743944644928, "loss_mean": 0.6097285747528076, "loss_mean_diff": 0.008472876623272896, "mean_diff_l2": 0.09204822778701782, "loss_pair": 5.829449582961388e-05, "loss_lang": 1.9136477708816528, "loss_len": 0.03265539929270744, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.64453125, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 1120, "elapsed_sec": 54.58237099647522}
8
+ {"step": 3160, "epoch": 0, "loss": 7.046058654785156, "loss_render": 5.024228572845459, "loss_align": 0.7382848262786865, "loss_varcov": 0.05277714878320694, "loss_mean": 0.6311275959014893, "loss_mean_diff": 0.0076501755975186825, "mean_diff_l2": 0.08746527880430222, "loss_pair": 4.0146631363313645e-05, "loss_lang": 1.208702802658081, "loss_len": 0.4566788971424103, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.71875, "len_acc_clf_detached_mix": 0.92578125, "adv_buf_fill": 1280, "elapsed_sec": 62.27759289741516}
9
+ {"step": 3180, "epoch": 0, "loss": 7.21535587310791, "loss_render": 5.199711799621582, "loss_align": 0.5909280776977539, "loss_varcov": 0.05281803011894226, "loss_mean": 0.6203291416168213, "loss_mean_diff": 0.007108135148882866, "mean_diff_l2": 0.08430975675582886, "loss_pair": 5.844556289957836e-05, "loss_lang": 2.22522234916687, "loss_len": 0.01663285121321678, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.5625, "len_acc_clf_detached_mix": 0.9375, "adv_buf_fill": 1440, "elapsed_sec": 69.89759159088135}
10
+ {"step": 3200, "epoch": 0, "loss": 7.056690216064453, "loss_render": 4.64270544052124, "loss_align": 0.5576940774917603, "loss_varcov": 0.05199909955263138, "loss_mean": 0.6144702434539795, "loss_mean_diff": 0.019337860867381096, "mean_diff_l2": 0.1390606313943863, "loss_pair": 8.246280049206689e-05, "loss_lang": 1.8289759159088135, "loss_len": 0.6372612714767456, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.625, "len_acc_clf_detached_mix": 0.9296875, "adv_buf_fill": 1600, "elapsed_sec": 77.87204480171204}
11
+ {"step": 3220, "epoch": 0, "loss": 6.683023929595947, "loss_render": 5.158309459686279, "loss_align": 0.1592254340648651, "loss_varcov": 0.045349638909101486, "loss_mean": 0.4908464848995209, "loss_mean_diff": 0.026845186948776245, "mean_diff_l2": 0.1638450026512146, "loss_pair": 0.000139553943881765, "loss_lang": 1.777904748916626, "loss_len": 0.05085552856326103, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6484375, "len_acc_clf_detached_mix": 0.92578125, "adv_buf_fill": 1760, "elapsed_sec": 96.39360666275024}
12
+ {"step": 3240, "epoch": 0, "loss": 7.937066078186035, "loss_render": 4.9136528968811035, "loss_align": 0.3106898069381714, "loss_varcov": 0.04477532207965851, "loss_mean": 0.4960465133190155, "loss_mean_diff": 0.01573428139090538, "mean_diff_l2": 0.12543636560440063, "loss_pair": 8.262126357294619e-05, "loss_lang": 2.070652961730957, "loss_len": 1.3337937593460083, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.5, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.5, "lang_acc_clf_detached_mix": 0.6171875, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 1920, "elapsed_sec": 105.07620596885681}
13
+ {"step": 3260, "epoch": 0, "loss": 6.735896110534668, "loss_render": 5.343853950500488, "loss_align": 0.44823068380355835, "loss_varcov": 0.05234198272228241, "loss_mean": 0.6424336433410645, "loss_mean_diff": 0.007120629772543907, "mean_diff_l2": 0.08438382297754288, "loss_pair": 7.646014273632318e-05, "loss_lang": 1.001953125, "loss_len": 0.07855981588363647, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.703125, "len_acc_clf_detached_mix": 0.92578125, "adv_buf_fill": 2080, "elapsed_sec": 113.16577768325806}
14
+ {"step": 3280, "epoch": 0, "loss": 5.892277240753174, "loss_render": 4.204550743103027, "loss_align": 0.7301828861236572, "loss_varcov": 0.06106958165764809, "loss_mean": 0.7883446216583252, "loss_mean_diff": 0.006495799403637648, "mean_diff_l2": 0.08059652149677277, "loss_pair": 4.376809738459997e-05, "loss_lang": 1.1500236988067627, "loss_len": 0.05743511766195297, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.609375, "len_acc_clf_detached_mix": 0.9375, "adv_buf_fill": 2240, "elapsed_sec": 121.94113636016846}
15
+ {"step": 3300, "epoch": 0, "loss": 6.675613880157471, "loss_render": 4.888925552368164, "loss_align": 0.3387090563774109, "loss_varcov": 0.054840683937072754, "loss_mean": 0.6974101066589355, "loss_mean_diff": 0.021088238805532455, "mean_diff_l2": 0.1452178955078125, "loss_pair": 0.00015820734552107751, "loss_lang": 1.7381736040115356, "loss_len": 0.12795868515968323, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.55859375, "len_acc_clf_detached_mix": 0.94140625, "adv_buf_fill": 2400, "elapsed_sec": 130.49251198768616}
16
+ {"step": 3320, "epoch": 0, "loss": 6.463747024536133, "loss_render": 4.755879878997803, "loss_align": 0.7946306467056274, "loss_varcov": 0.06582553684711456, "loss_mean": 0.8660086393356323, "loss_mean_diff": 0.006592818535864353, "mean_diff_l2": 0.08119617402553558, "loss_pair": 6.574299914063886e-05, "loss_lang": 1.0204485654830933, "loss_len": 0.05479826778173447, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.7109375, "len_acc_clf_detached_mix": 0.94921875, "adv_buf_fill": 2560, "elapsed_sec": 138.41628503799438}
17
+ {"step": 3340, "epoch": 0, "loss": 6.213773727416992, "loss_render": 4.764244079589844, "loss_align": 0.07418331503868103, "loss_varcov": 0.045371875166893005, "loss_mean": 0.5181745290756226, "loss_mean_diff": 0.017265059053897858, "mean_diff_l2": 0.1313965767621994, "loss_pair": 0.00020561416749842465, "loss_lang": 1.683219075202942, "loss_len": 0.06352486461400986, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6328125, "len_acc_clf_detached_mix": 0.8984375, "adv_buf_fill": 2720, "elapsed_sec": 146.91128516197205}
18
+ {"step": 3360, "epoch": 0, "loss": 6.712790012359619, "loss_render": 5.113739013671875, "loss_align": 0.2539443373680115, "loss_varcov": 0.04548342153429985, "loss_mean": 0.5077573657035828, "loss_mean_diff": 0.009698632173240185, "mean_diff_l2": 0.09848163276910782, "loss_pair": 8.462463301839307e-05, "loss_lang": 1.7478320598602295, "loss_len": 0.0915665552020073, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6484375, "len_acc_clf_detached_mix": 0.91796875, "adv_buf_fill": 2880, "elapsed_sec": 155.44744443893433}
19
+ {"step": 3380, "epoch": 0, "loss": 6.560072898864746, "loss_render": 4.628041744232178, "loss_align": 0.4736449420452118, "loss_varcov": 0.04999326169490814, "loss_mean": 0.5792374014854431, "loss_mean_diff": 0.004861920140683651, "mean_diff_l2": 0.06972747296094894, "loss_pair": 5.205247725825757e-05, "loss_lang": 1.543569564819336, "loss_len": 0.3650703430175781, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.625, "len_acc_batch": 0.875, "lang_acc_cur_detached": 0.625, "len_acc_cur_detached": 0.875, "lang_acc_clf_detached_mix": 0.66015625, "len_acc_clf_detached_mix": 0.9296875, "adv_buf_fill": 3040, "elapsed_sec": 163.11043548583984}
20
+ {"step": 3400, "epoch": 0, "loss": 6.325645923614502, "loss_render": 5.069282531738281, "loss_align": 0.1756565272808075, "loss_varcov": 0.046286843717098236, "loss_mean": 0.5082923173904419, "loss_mean_diff": 0.009171318262815475, "mean_diff_l2": 0.09576699882745743, "loss_pair": 5.370082362787798e-05, "loss_lang": 1.2323110103607178, "loss_len": 0.03775423392653465, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.57421875, "len_acc_clf_detached_mix": 0.87890625, "adv_buf_fill": 3200, "elapsed_sec": 171.33779048919678}
21
+ {"step": 3420, "epoch": 0, "loss": 6.82662296295166, "loss_render": 4.924870014190674, "loss_align": 0.7773740887641907, "loss_varcov": 0.05529933422803879, "loss_mean": 0.6768600940704346, "loss_mean_diff": 0.11194519698619843, "mean_diff_l2": 0.33458212018013, "loss_pair": 0.00036621722392737865, "loss_lang": 1.7035551071166992, "loss_len": 0.029340676963329315, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.375, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.375, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.61328125, "len_acc_clf_detached_mix": 0.91796875, "adv_buf_fill": 3360, "elapsed_sec": 179.39320826530457}
22
+ {"step": 3440, "epoch": 0, "loss": 6.1194047927856445, "loss_render": 4.756137847900391, "loss_align": 0.19583454728126526, "loss_varcov": 0.04695400223135948, "loss_mean": 0.5665853023529053, "loss_mean_diff": 0.0060009402222931385, "mean_diff_l2": 0.0774657353758812, "loss_pair": 4.0852726669982076e-05, "loss_lang": 1.2927026748657227, "loss_len": 0.09219157695770264, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.6171875, "len_acc_clf_detached_mix": 0.9140625, "adv_buf_fill": 3520, "elapsed_sec": 187.6694211959839}
23
+ {"step": 3460, "epoch": 0, "loss": 6.550313949584961, "loss_render": 4.864479064941406, "loss_align": 0.30802199244499207, "loss_varcov": 0.042097095400094986, "loss_mean": 0.4444228410720825, "loss_mean_diff": 0.0038946648128330708, "mean_diff_l2": 0.062407251447439194, "loss_pair": 2.7830676117446274e-05, "loss_lang": 1.9070429801940918, "loss_len": 0.11249455064535141, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 1.0, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 1.0, "lang_acc_clf_detached_mix": 0.63671875, "len_acc_clf_detached_mix": 0.9375, "adv_buf_fill": 3680, "elapsed_sec": 196.0950767993927}
24
+ {"step": 3480, "epoch": 0, "loss": 7.749448299407959, "loss_render": 5.214535713195801, "loss_align": 0.4330580532550812, "loss_varcov": 0.04625701904296875, "loss_mean": 0.516429603099823, "loss_mean_diff": 0.07339179515838623, "mean_diff_l2": 0.2709091901779175, "loss_pair": 0.0002953222137875855, "loss_lang": 1.407311201095581, "loss_len": 1.0931166410446167, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.5703125, "len_acc_clf_detached_mix": 0.8984375, "adv_buf_fill": 3840, "elapsed_sec": 203.9322428703308}
25
+ {"step": 3500, "epoch": 0, "loss": 7.0917510986328125, "loss_render": 4.889008045196533, "loss_align": 0.03125523030757904, "loss_varcov": 0.04502371698617935, "loss_mean": 0.5110816955566406, "loss_mean_diff": 0.034921273589134216, "mean_diff_l2": 0.18687234818935394, "loss_pair": 0.00014367523544933647, "loss_lang": 1.2161797285079956, "loss_len": 1.074159026145935, "grl_alpha_eff": 1.0, "lang_acc_batch": 0.5, "len_acc_batch": 0.75, "lang_acc_cur_detached": 0.5, "len_acc_cur_detached": 0.75, "lang_acc_clf_detached_mix": 0.625, "len_acc_clf_detached_mix": 0.89453125, "adv_buf_fill": 4000, "elapsed_sec": 212.18113350868225}
planB_robust/probe_leakage.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ckpt": "runs/planB_robust/ckpt.pt",
3
+ "max_train_groups": 8000,
4
+ "max_valid_groups": 946,
5
+ "probe_hidden": 256,
6
+ "epochs": 20,
7
+ "lr": 0.01,
8
+ "baseline_majority_lang_acc": 0.5,
9
+ "baseline_majority_len_acc": 0.9271249771118164,
10
+ "lang_probe": {
11
+ "train_acc": 0.662875,
12
+ "valid_acc": 0.6617336152219874,
13
+ "best_valid_acc": 0.6633192389006343
14
+ },
15
+ "len_probe": {
16
+ "train_acc": 0.9288125,
17
+ "valid_acc": 0.9011627906976745,
18
+ "best_valid_acc": 0.9011627906976745
19
+ },
20
+ "lang_probe_length_matched": {
21
+ "train_acc": 0.6607785200411099,
22
+ "valid_acc": 0.6471885336273429,
23
+ "best_valid_acc": 0.6670341786108048
24
+ },
25
+ "lang_bucketwise_valid_acc": {
26
+ "bucket_0_acc": 0.59375,
27
+ "bucket_1_acc": 0.7250000238418579,
28
+ "bucket_2_acc": 0.6521738767623901,
29
+ "bucket_3_acc": 0.6621701121330261
30
+ },
31
+ "length_matched_sizes": {
32
+ "train": 15568,
33
+ "valid": 1814
34
+ }
35
+ }