| {"event": "hub", "repo": "AdaptiveChunking/ali_pkm_s1", "keep_steps": [0, 1, 2, 4, 6, 8, 16, 32, 64, 128, 256, 512, 1000, 1024, 2048, 3000, 4096, 6000], "t": 5.7} |
| {"event": "start", "run": "ali_pkm_s1", "arch": "pkm", "objective": "ratio", "condition": "A", "params": 39425792, "langs": ["en", "tr", "fi"], "premium": {"en": 1.0, "tr": 1.12, "fi": 1.104}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "global", "p3_objective": "global", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {}, "weights": {"en": 0.74, "tr": 0.1536, "fi": 0.1064}, "device": "cuda", "pkm_params": 16908800, "pkm_config": {"n_keys": 256, "topk": 8, "heads": 1, "d_key": 256}, "t": 5.7} |
| {"event": "probe_schedule", "n": 45, "steps": [0, 1, 2, 4, 6, 8, 16, 32, 64, 128, 200, 256, 400, 512, 600, 800, 1000, 1024, 1200, 1400, 1600, 1800, 2000, 2048, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4096, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 5.7} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2674.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.153652, "cps": 20.890625, "mean_p": 0.478126, "boundary_entropy": 0.685438, "bpb": 8.041593}, "tr": {"chunks": 5645.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.311827, "cps": 44.101562, "mean_p": 0.486063, "boundary_entropy": 0.686375, "bpb": 8.017619}, "fi": {"chunks": 4732.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.251555, "cps": 36.96875, "mean_p": 0.48451, "boundary_entropy": 0.686595, "bpb": 8.075705}}, "bpb": {"en": 8.04159, "tr": 8.01762, "fi": 8.0757}, "gini_cpc": 0.147064, "gini_cps": 0.151764, "bpic": 4.346947, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 2.029439514461233, "fi": 1.6371767549398106}, "t": 6.9} |
| {"event": "train", "step": 0, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 5.5856, "bpb": 8.05832, "chunk_loss": 0.030141, "F": 0.16724, "G": 0.47713, "chunks": 171.25, "bytes_per_chunk": 5.98, "boundary_entropy": 0.69022, "router_noise": 0.0, "grad_norm": 11.404, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.014, "alpha": 0.03, "t": 7.7} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2661.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.152905, "cps": 20.789062, "mean_p": 0.478106, "boundary_entropy": 0.685439, "bpb": 8.00835}, "tr": {"chunks": 5624.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.310667, "cps": 43.9375, "mean_p": 0.486, "boundary_entropy": 0.686373, "bpb": 7.994075}, "fi": {"chunks": 4708.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.250279, "cps": 36.78125, "mean_p": 0.484402, "boundary_entropy": 0.686587, "bpb": 8.047965}}, "bpb": {"en": 8.00835, "tr": 7.99408, "fi": 8.04797}, "gini_cpc": 0.147334, "gini_cps": 0.152031, "bpic": 4.366351, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 2.03176748657516, "fi": 1.6368309050864178}, "t": 8.3} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2658.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.152732, "cps": 20.765625, "mean_p": 0.478081, "boundary_entropy": 0.685441, "bpb": 7.943327}, "tr": {"chunks": 5584.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.308457, "cps": 43.625, "mean_p": 0.485899, "boundary_entropy": 0.686371, "bpb": 7.946404}, "fi": {"chunks": 4664.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.24794, "cps": 36.4375, "mean_p": 0.484227, "boundary_entropy": 0.686575, "bpb": 7.994118}}, "bpb": {"en": 7.94333, "tr": 7.9464, "fi": 7.99412}, "gini_cpc": 0.1464, "gini_cps": 0.151144, "bpic": 4.395785, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 2.019593674470477, "fi": 1.6233635931524144}, "t": 9.3} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2617.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.150376, "cps": 20.445312, "mean_p": 0.478047, "boundary_entropy": 0.685455, "bpb": 7.727459}, "tr": {"chunks": 5457.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.301442, "cps": 42.632812, "mean_p": 0.485577, "boundary_entropy": 0.686365, "bpb": 7.79109}, "fi": {"chunks": 4538.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.241242, "cps": 35.453125, "mean_p": 0.48368, "boundary_entropy": 0.686539, "bpb": 7.814348}}, "bpb": {"en": 7.72746, "tr": 7.79109, "fi": 7.81435}, "gini_cpc": 0.145312, "gini_cps": 0.150122, "bpic": 4.498256, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 2.0045818781083935, "fi": 1.6042535376997016}, "t": 10.8} |
| {"event": "probe", "step": 6, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2582.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.148365, "cps": 20.171875, "mean_p": 0.478078, "boundary_entropy": 0.685489, "bpb": 7.44398}, "tr": {"chunks": 5246.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.289786, "cps": 40.984375, "mean_p": 0.485087, "boundary_entropy": 0.686359, "bpb": 7.589876}, "fi": {"chunks": 4288.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.227952, "cps": 33.5, "mean_p": 0.482836, "boundary_entropy": 0.686485, "bpb": 7.576384}}, "bpb": {"en": 7.44398, "tr": 7.58988, "fi": 7.57638}, "gini_cpc": 0.141541, "gini_cps": 0.146583, "bpic": 4.682403, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.953195059529614, "fi": 1.536422913880053}, "t": 12.0} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2629.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.151066, "cps": 20.539062, "mean_p": 0.478861, "boundary_entropy": 0.685614, "bpb": 7.169338}, "tr": {"chunks": 5159.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.28498, "cps": 40.304688, "mean_p": 0.484963, "boundary_entropy": 0.686391, "bpb": 7.398299}, "fi": {"chunks": 4109.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.218436, "cps": 32.101562, "mean_p": 0.482342, "boundary_entropy": 0.686467, "bpb": 7.34044}}, "bpb": {"en": 7.16934, "tr": 7.3983, "fi": 7.34044}, "gini_cpc": 0.136408, "gini_cps": 0.141772, "bpic": 4.768597, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.8864639509344137, "fi": 1.4459650368051244}, "t": 13.1} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2598.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.149285, "cps": 20.296875, "mean_p": 0.481102, "boundary_entropy": 0.686032, "bpb": 6.514911}, "tr": {"chunks": 4390.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.242501, "cps": 34.296875, "mean_p": 0.482955, "boundary_entropy": 0.686437, "bpb": 6.89823}, "fi": {"chunks": 2920.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155228, "cps": 22.8125, "mean_p": 0.478263, "boundary_entropy": 0.686272, "bpb": 6.693927}}, "bpb": {"en": 6.51491, "tr": 6.89823, "fi": 6.69393}, "gini_cpc": 0.113607, "gini_cps": 0.120576, "bpic": 5.725878, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.624422297913186, "fi": 1.0398146728309796}, "t": 14.7} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3690.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.212032, "cps": 28.828125, "mean_p": 0.490503, "boundary_entropy": 0.687173, "bpb": 6.161246}, "tr": {"chunks": 3946.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.217975, "cps": 30.828125, "mean_p": 0.486225, "boundary_entropy": 0.687284, "bpb": 6.577238}, "fi": {"chunks": 3149.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.167402, "cps": 24.601562, "mean_p": 0.485037, "boundary_entropy": 0.687297, "bpb": 6.341533}}, "bpb": {"en": 6.16125, "tr": 6.57724, "fi": 6.34153}, "gini_cpc": 0.056436, "gini_cps": 0.049266, "bpic": 5.260269, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.0280264376968467, "fi": 0.7895116289422424}, "t": 17.2} |
| {"event": "train", "step": 50, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 4.09846, "bpb": 5.91283, "chunk_loss": 0.030558, "F": 0.17438, "G": 0.48631, "chunks": 178.56, "bytes_per_chunk": 5.735, "boundary_entropy": 0.69164, "router_noise": 0.0, "grad_norm": 2.362, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 19.7} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3768.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.216514, "cps": 29.4375, "mean_p": 0.484628, "boundary_entropy": 0.686198, "bpb": 5.428619}, "tr": {"chunks": 4469.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.246865, "cps": 34.914062, "mean_p": 0.481569, "boundary_entropy": 0.686143, "bpb": 5.921806}, "fi": {"chunks": 3960.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.210515, "cps": 30.9375, "mean_p": 0.482643, "boundary_entropy": 0.686565, "bpb": 5.654159}}, "bpb": {"en": 5.42862, "tr": 5.92181, "fi": 5.65416}, "gini_cpc": 0.03596, "gini_cps": 0.038315, "bpic": 4.651308, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1401789776195732, "fi": 0.9722915884356862}, "t": 21.4} |
| {"event": "train", "step": 100, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 3.20853, "bpb": 4.62893, "chunk_loss": 0.032975, "F": 0.21317, "G": 0.4616, "chunks": 218.28, "bytes_per_chunk": 4.691, "boundary_entropy": 0.68336, "router_noise": 0.0, "grad_norm": 1.6458, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.108, "alpha": 0.03, "t": 25.1} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3646.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.209504, "cps": 28.484375, "mean_p": 0.42395, "boundary_entropy": 0.619307, "bpb": 4.122926}, "tr": {"chunks": 4228.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.233552, "cps": 33.03125, "mean_p": 0.409815, "boundary_entropy": 0.640318, "bpb": 4.373508}, "fi": {"chunks": 4338.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.23061, "cps": 33.890625, "mean_p": 0.415607, "boundary_entropy": 0.64285, "bpb": 4.328399}}, "bpb": {"en": 4.12293, "tr": 4.37351, "fi": 4.3284}, "gini_cpc": 0.023798, "gini_cps": 0.037777, "bpic": 4.645594, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1147869679349613, "fi": 1.1007409414543778}, "t": 28.1} |
| {"event": "train", "step": 150, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.74912, "bpb": 3.96614, "chunk_loss": 0.03305, "F": 0.22708, "G": 0.39748, "chunks": 232.53, "bytes_per_chunk": 4.404, "boundary_entropy": 0.58136, "router_noise": 0.0, "grad_norm": 2.1881, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.126, "alpha": 0.03, "t": 31.4} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3218.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.184911, "cps": 25.140625, "mean_p": 0.324442, "boundary_entropy": 0.463783, "bpb": 3.542093}, "tr": {"chunks": 4391.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.242556, "cps": 34.304688, "mean_p": 0.311023, "boundary_entropy": 0.48939, "bpb": 3.775809}, "fi": {"chunks": 3756.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.19967, "cps": 29.34375, "mean_p": 0.332285, "boundary_entropy": 0.526586, "bpb": 3.881188}}, "bpb": {"en": 3.54209, "tr": 3.77581, "fi": 3.88119}, "gini_cpc": 0.061279, "gini_cps": 0.068808, "bpic": 4.991817, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.3117496775495046, "fi": 1.0798210282460716}, "t": 36.3} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.55178, "tr": 3.78443, "fi": 3.87929}, "bytes_per_chunk": {"en": 5.184, "tr": 4.463, "fi": 5.282}, "bpb_mean": 3.7385, "frozen_inventory": false, "t": 37.0} |
| {"event": "train", "step": 200, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.48406, "bpb": 3.58374, "chunk_loss": 0.031015, "F": 0.19598, "G": 0.32601, "chunks": 200.69, "bytes_per_chunk": 5.102, "boundary_entropy": 0.4748, "router_noise": 0.0, "grad_norm": 1.5547, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.122, "alpha": 0.03, "t": 37.5} |
| {"event": "train", "step": 250, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.36959, "bpb": 3.4186, "chunk_loss": 0.030585, "F": 0.18817, "G": 0.28822, "chunks": 192.69, "bytes_per_chunk": 5.314, "boundary_entropy": 0.39138, "router_noise": 0.0, "grad_norm": 2.1927, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.098, "alpha": 0.03, "t": 42.4} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3134.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.180084, "cps": 24.484375, "mean_p": 0.273736, "boundary_entropy": 0.362873, "bpb": 3.290841}, "tr": {"chunks": 3701.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.204441, "cps": 28.914062, "mean_p": 0.27242, "boundary_entropy": 0.397611, "bpb": 3.511639}, "fi": {"chunks": 3322.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.176599, "cps": 25.953125, "mean_p": 0.283323, "boundary_entropy": 0.438278, "bpb": 3.639907}}, "bpb": {"en": 3.29084, "tr": 3.51164, "fi": 3.63991}, "gini_cpc": 0.033079, "gini_cps": 0.037216, "bpic": 5.585508, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1352556231711182, "fi": 0.9806473808570438}, "t": 43.2} |
| {"event": "train", "step": 300, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.20632, "bpb": 3.18304, "chunk_loss": 0.030556, "F": 0.1879, "G": 0.28709, "chunks": 192.41, "bytes_per_chunk": 5.322, "boundary_entropy": 0.36852, "router_noise": 0.0, "grad_norm": 1.7774, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.113, "alpha": 0.03, "t": 48.1} |
| {"event": "train", "step": 350, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.19169, "bpb": 3.16194, "chunk_loss": 0.030528, "F": 0.18872, "G": 0.27981, "chunks": 193.25, "bytes_per_chunk": 5.299, "boundary_entropy": 0.35925, "router_noise": 0.0, "grad_norm": 1.5387, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.094, "alpha": 0.03, "t": 52.8} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3453.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198414, "cps": 26.976562, "mean_p": 0.257127, "boundary_entropy": 0.282475, "bpb": 2.880122}, "tr": {"chunks": 3515.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.194167, "cps": 27.460938, "mean_p": 0.301596, "boundary_entropy": 0.412571, "bpb": 3.123512}, "fi": {"chunks": 3417.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181649, "cps": 26.695312, "mean_p": 0.283167, "boundary_entropy": 0.408974, "bpb": 3.345087}}, "bpb": {"en": 2.88012, "tr": 3.12351, "fi": 3.34509}, "gini_cpc": 0.019464, "gini_cps": 0.006291, "bpic": 5.462879, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9785934842485704, "fi": 0.9155048243626608}, "t": 57.4} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 2.8922, "tr": 3.13102, "fi": 3.33894}, "bytes_per_chunk": {"en": 4.794, "tr": 5.715, "fi": 5.683}, "bpb_mean": 3.12072, "frozen_inventory": false, "t": 58.0} |
| {"event": "train", "step": 400, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.0322, "bpb": 2.93185, "chunk_loss": 0.030692, "F": 0.20041, "G": 0.26498, "chunks": 205.22, "bytes_per_chunk": 4.99, "boundary_entropy": 0.29087, "router_noise": 0.0, "grad_norm": 1.4496, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.118, "alpha": 0.03, "t": 58.7} |
| {"event": "train", "step": 450, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.06152, "bpb": 2.97414, "chunk_loss": 0.030618, "F": 0.19772, "G": 0.2547, "chunks": 202.47, "bytes_per_chunk": 5.058, "boundary_entropy": 0.27471, "router_noise": 0.0, "grad_norm": 1.157, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 63.0} |
| {"event": "train", "step": 500, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.02184, "bpb": 2.9169, "chunk_loss": 0.030492, "F": 0.19455, "G": 0.25286, "chunks": 199.22, "bytes_per_chunk": 5.14, "boundary_entropy": 0.27007, "router_noise": 0.0, "grad_norm": 1.0297, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 67.5} |
| {"event": "probe", "step": 512, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3521.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.202321, "cps": 27.507812, "mean_p": 0.245108, "boundary_entropy": 0.243641, "bpb": 2.687305}, "tr": {"chunks": 3231.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178479, "cps": 25.242188, "mean_p": 0.284617, "boundary_entropy": 0.383649, "bpb": 2.977423}, "fi": {"chunks": 3329.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.176971, "cps": 26.007812, "mean_p": 0.252256, "boundary_entropy": 0.351108, "bpb": 3.242355}}, "bpb": {"en": 2.6873, "tr": 2.97742, "fi": 3.24236}, "gini_cpc": 0.0303, "gini_cps": 0.019178, "bpic": 5.627616, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8821541909597018, "fi": 0.8747017730339637}, "t": 68.8} |
| {"event": "train", "step": 550, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.95106, "bpb": 2.81479, "chunk_loss": 0.030815, "F": 0.20764, "G": 0.25369, "chunks": 212.62, "bytes_per_chunk": 4.816, "boundary_entropy": 0.24665, "router_noise": 0.0, "grad_norm": 1.3349, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.096, "alpha": 0.03, "t": 72.3} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3448.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198127, "cps": 26.9375, "mean_p": 0.231246, "boundary_entropy": 0.206714, "bpb": 2.585266}, "tr": {"chunks": 3271.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.180688, "cps": 25.554688, "mean_p": 0.289934, "boundary_entropy": 0.380642, "bpb": 2.88308}, "fi": {"chunks": 3246.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.172559, "cps": 25.359375, "mean_p": 0.242136, "boundary_entropy": 0.326811, "bpb": 3.130869}}, "bpb": {"en": 2.58527, "tr": 2.88308, "fi": 3.13087}, "gini_cpc": 0.030914, "gini_cps": 0.013514, "bpic": 5.693126, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9119832370658593, "fi": 0.8709505446842638}, "t": 76.8} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.59916, "tr": 2.89427, "fi": 3.12833}, "bytes_per_chunk": {"en": 4.804, "tr": 5.95, "fi": 5.912}, "bpb_mean": 2.87392, "frozen_inventory": false, "t": 77.3} |
| {"event": "train", "step": 600, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.92466, "bpb": 2.7767, "chunk_loss": 0.030515, "F": 0.19589, "G": 0.24888, "chunks": 200.59, "bytes_per_chunk": 5.105, "boundary_entropy": 0.24306, "router_noise": 0.0, "grad_norm": 1.1756, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.116, "alpha": 0.03, "t": 78.1} |
| {"event": "train", "step": 650, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.8584, "bpb": 2.68111, "chunk_loss": 0.030523, "F": 0.19748, "G": 0.25406, "chunks": 202.22, "bytes_per_chunk": 5.064, "boundary_entropy": 0.25019, "router_noise": 0.0, "grad_norm": 1.2143, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.107, "alpha": 0.03, "t": 83.4} |
| {"event": "train", "step": 700, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.87513, "bpb": 2.70524, "chunk_loss": 0.030377, "F": 0.18939, "G": 0.26005, "chunks": 193.94, "bytes_per_chunk": 5.28, "boundary_entropy": 0.27422, "router_noise": 0.0, "grad_norm": 1.0961, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.094, "alpha": 0.03, "t": 88.1} |
| {"event": "train", "step": 750, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.77816, "bpb": 2.56534, "chunk_loss": 0.030427, "F": 0.19672, "G": 0.24134, "chunks": 201.44, "bytes_per_chunk": 5.083, "boundary_entropy": 0.22057, "router_noise": 0.0, "grad_norm": 1.0479, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 92.7} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3532.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.202954, "cps": 27.59375, "mean_p": 0.225301, "boundary_entropy": 0.184236, "bpb": 2.416091}, "tr": {"chunks": 3357.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185439, "cps": 26.226562, "mean_p": 0.289889, "boundary_entropy": 0.372075, "bpb": 2.713281}, "fi": {"chunks": 3609.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.191856, "cps": 28.195312, "mean_p": 0.261768, "boundary_entropy": 0.319729, "bpb": 2.982123}}, "bpb": {"en": 2.41609, "tr": 2.71328, "fi": 2.98212}, "gini_cpc": 0.020123, "gini_cps": 0.016003, "bpic": 5.404077, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9137012417118128, "fi": 0.9453190806107099}, "t": 98.2} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.434, "tr": 2.72476, "fi": 2.98117}, "bytes_per_chunk": {"en": 4.717, "tr": 5.76, "fi": 5.503}, "bpb_mean": 2.71331, "frozen_inventory": false, "t": 98.7} |
| {"event": "train", "step": 800, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.79157, "bpb": 2.58469, "chunk_loss": 0.030548, "F": 0.19928, "G": 0.2508, "chunks": 204.06, "bytes_per_chunk": 5.018, "boundary_entropy": 0.2434, "router_noise": 0.0, "grad_norm": 0.9211, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.135, "alpha": 0.03, "t": 99.4} |
| {"event": "train", "step": 850, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.72959, "bpb": 2.49526, "chunk_loss": 0.030494, "F": 0.20236, "G": 0.22793, "chunks": 207.22, "bytes_per_chunk": 4.942, "boundary_entropy": 0.18381, "router_noise": 0.0, "grad_norm": 1.227, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 103.5} |
| {"event": "train", "step": 900, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.70428, "bpb": 2.45875, "chunk_loss": 0.030441, "F": 0.19894, "G": 0.23752, "chunks": 203.72, "bytes_per_chunk": 5.027, "boundary_entropy": 0.20503, "router_noise": 0.0, "grad_norm": 1.0716, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.095, "alpha": 0.03, "t": 108.3} |
| {"event": "train", "step": 950, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.70732, "bpb": 2.46314, "chunk_loss": 0.03051, "F": 0.20239, "G": 0.23556, "chunks": 207.25, "bytes_per_chunk": 4.941, "boundary_entropy": 0.19154, "router_noise": 0.0, "grad_norm": 0.9453, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.105, "alpha": 0.03, "t": 113.5} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3499.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201057, "cps": 27.335938, "mean_p": 0.21839, "boundary_entropy": 0.169853, "bpb": 2.328647}, "tr": {"chunks": 3233.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178589, "cps": 25.257812, "mean_p": 0.271784, "boundary_entropy": 0.341013, "bpb": 2.620238}, "fi": {"chunks": 3376.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179469, "cps": 26.375, "mean_p": 0.240087, "boundary_entropy": 0.284904, "bpb": 2.861889}}, "bpb": {"en": 2.32865, "tr": 2.62024, "fi": 2.86189}, "gini_cpc": 0.02679, "gini_cps": 0.017544, "bpic": 5.612584, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8882502346729948, "fi": 0.8926284656252326}, "t": 118.8} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.34331, "tr": 2.62823, "fi": 2.86275}, "bytes_per_chunk": {"en": 4.745, "tr": 5.952, "fi": 5.841}, "bpb_mean": 2.61143, "frozen_inventory": false, "t": 119.4} |
| {"event": "train", "step": 1000, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.71973, "bpb": 2.48104, "chunk_loss": 0.030473, "F": 0.20126, "G": 0.23471, "chunks": 206.09, "bytes_per_chunk": 4.969, "boundary_entropy": 0.2001, "router_noise": 0.0, "grad_norm": 0.9761, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.13, "alpha": 0.03, "t": 120.0} |
| {"event": "probe", "step": 1024, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3512.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201804, "cps": 27.4375, "mean_p": 0.220141, "boundary_entropy": 0.164203, "bpb": 2.319219}, "tr": {"chunks": 3159.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174501, "cps": 24.679688, "mean_p": 0.263874, "boundary_entropy": 0.331865, "bpb": 2.59409}, "fi": {"chunks": 3422.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181915, "cps": 26.734375, "mean_p": 0.240891, "boundary_entropy": 0.280177, "bpb": 2.830634}}, "bpb": {"en": 2.31922, "tr": 2.59409, "fi": 2.83063}, "gini_cpc": 0.032607, "gini_cps": 0.023316, "bpic": 5.620925, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8647064280489636, "fi": 0.901441887646556}, "t": 122.7} |
| {"event": "train", "step": 1050, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.73394, "bpb": 2.50154, "chunk_loss": 0.030343, "F": 0.19461, "G": 0.23059, "chunks": 199.28, "bytes_per_chunk": 5.138, "boundary_entropy": 0.19834, "router_noise": 0.0, "grad_norm": 1.0071, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.112, "alpha": 0.03, "t": 125.6} |
| {"event": "train", "step": 1100, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.71891, "bpb": 2.47986, "chunk_loss": 0.030502, "F": 0.20438, "G": 0.23018, "chunks": 209.28, "bytes_per_chunk": 4.893, "boundary_entropy": 0.17882, "router_noise": 0.0, "grad_norm": 0.9473, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 130.0} |
| {"event": "train", "step": 1150, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.72874, "bpb": 2.49405, "chunk_loss": 0.030466, "F": 0.1972, "G": 0.24035, "chunks": 201.94, "bytes_per_chunk": 5.071, "boundary_entropy": 0.21192, "router_noise": 0.0, "grad_norm": 0.928, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.098, "alpha": 0.03, "t": 134.9} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3474.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199621, "cps": 27.140625, "mean_p": 0.213844, "boundary_entropy": 0.150804, "bpb": 2.262501}, "tr": {"chunks": 3152.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174115, "cps": 24.625, "mean_p": 0.254575, "boundary_entropy": 0.312425, "bpb": 2.53211}, "fi": {"chunks": 3275.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.1741, "cps": 25.585938, "mean_p": 0.228668, "boundary_entropy": 0.252983, "bpb": 2.745095}}, "bpb": {"en": 2.2625, "tr": 2.53211, "fi": 2.74509}, "gini_cpc": 0.031056, "gini_cps": 0.021681, "bpic": 5.729926, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.87222787814537, "fi": 0.8721551045583975}, "t": 139.5} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.27784, "tr": 2.54065, "fi": 2.75148}, "bytes_per_chunk": {"en": 4.755, "tr": 6.106, "fi": 5.915}, "bpb_mean": 2.52332, "frozen_inventory": false, "t": 139.9} |
| {"event": "train", "step": 1200, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.62605, "bpb": 2.3459, "chunk_loss": 0.030386, "F": 0.19702, "G": 0.23158, "chunks": 201.75, "bytes_per_chunk": 5.076, "boundary_entropy": 0.18695, "router_noise": 0.0, "grad_norm": 0.9089, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.109, "alpha": 0.03, "t": 140.4} |
| {"event": "train", "step": 1250, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.66704, "bpb": 2.40503, "chunk_loss": 0.030381, "F": 0.19516, "G": 0.23031, "chunks": 199.84, "bytes_per_chunk": 5.124, "boundary_entropy": 0.18891, "router_noise": 0.0, "grad_norm": 0.9558, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 145.0} |
| {"event": "train", "step": 1300, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.58847, "bpb": 2.29168, "chunk_loss": 0.03043, "F": 0.20129, "G": 0.22649, "chunks": 206.12, "bytes_per_chunk": 4.968, "boundary_entropy": 0.16283, "router_noise": 0.0, "grad_norm": 0.7792, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 149.6} |
| {"event": "train", "step": 1350, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.61666, "bpb": 2.33235, "chunk_loss": 0.030441, "F": 0.20294, "G": 0.21994, "chunks": 207.81, "bytes_per_chunk": 4.928, "boundary_entropy": 0.14771, "router_noise": 0.0, "grad_norm": 0.9511, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 153.9} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3500.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201115, "cps": 27.34375, "mean_p": 0.21103, "boundary_entropy": 0.134734, "bpb": 2.205757}, "tr": {"chunks": 3230.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178423, "cps": 25.234375, "mean_p": 0.254094, "boundary_entropy": 0.301856, "bpb": 2.469125}, "fi": {"chunks": 3352.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.178194, "cps": 26.1875, "mean_p": 0.221042, "boundary_entropy": 0.232965, "bpb": 2.67739}}, "bpb": {"en": 2.20576, "tr": 2.46912, "fi": 2.67739}, "gini_cpc": 0.027398, "gini_cps": 0.017854, "bpic": 5.627058, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8871724497123602, "fi": 0.8860295419853126}, "t": 159.8} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.22125, "tr": 2.47882, "fi": 2.68337}, "bytes_per_chunk": {"en": 4.727, "tr": 5.917, "fi": 5.802}, "bpb_mean": 2.46115, "frozen_inventory": false, "t": 160.3} |
| {"event": "train", "step": 1400, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.58688, "bpb": 2.28938, "chunk_loss": 0.03038, "F": 0.20023, "G": 0.22423, "chunks": 205.03, "bytes_per_chunk": 4.994, "boundary_entropy": 0.16356, "router_noise": 0.0, "grad_norm": 1.0011, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.138, "alpha": 0.03, "t": 160.8} |
| {"event": "train", "step": 1450, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.55688, "bpb": 2.24611, "chunk_loss": 0.030478, "F": 0.20239, "G": 0.23066, "chunks": 207.25, "bytes_per_chunk": 4.941, "boundary_entropy": 0.16964, "router_noise": 0.0, "grad_norm": 0.797, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 165.0} |
| {"event": "train", "step": 1500, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.55566, "bpb": 2.24435, "chunk_loss": 0.030348, "F": 0.19858, "G": 0.21711, "chunks": 203.34, "bytes_per_chunk": 5.036, "boundary_entropy": 0.15128, "router_noise": 0.0, "grad_norm": 1.2143, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 169.4} |
| {"event": "train", "step": 1550, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.54557, "bpb": 2.22978, "chunk_loss": 0.03049, "F": 0.20395, "G": 0.22502, "chunks": 208.84, "bytes_per_chunk": 4.903, "boundary_entropy": 0.14759, "router_noise": 0.0, "grad_norm": 0.8174, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 173.6} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3457.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198644, "cps": 27.007812, "mean_p": 0.206239, "boundary_entropy": 0.129393, "bpb": 2.167974}, "tr": {"chunks": 3213.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.177484, "cps": 25.101562, "mean_p": 0.244226, "boundary_entropy": 0.285093, "bpb": 2.428662}, "fi": {"chunks": 3476.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184785, "cps": 27.15625, "mean_p": 0.229845, "boundary_entropy": 0.234379, "bpb": 2.60859}}, "bpb": {"en": 2.16797, "tr": 2.42866, "fi": 2.60859}, "gini_cpc": 0.025149, "gini_cps": 0.017281, "bpic": 5.591563, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.893480163032636, "fi": 0.9302348912442631}, "t": 178.0} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.18394, "tr": 2.43702, "fi": 2.61702}, "bytes_per_chunk": {"en": 4.796, "tr": 5.98, "fi": 5.677}, "bpb_mean": 2.41266, "frozen_inventory": false, "t": 178.5} |
| {"event": "train", "step": 1600, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.5237, "bpb": 2.19824, "chunk_loss": 0.030323, "F": 0.19547, "G": 0.22331, "chunks": 200.16, "bytes_per_chunk": 5.116, "boundary_entropy": 0.16896, "router_noise": 0.0, "grad_norm": 0.816, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.107, "alpha": 0.03, "t": 179.0} |
| {"event": "train", "step": 1650, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.57501, "bpb": 2.27226, "chunk_loss": 0.0304, "F": 0.202, "G": 0.22086, "chunks": 206.84, "bytes_per_chunk": 4.951, "boundary_entropy": 0.14576, "router_noise": 0.0, "grad_norm": 0.8192, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.078, "alpha": 0.03, "t": 182.9} |
| {"event": "train", "step": 1700, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.57326, "bpb": 2.26973, "chunk_loss": 0.030339, "F": 0.19662, "G": 0.22347, "chunks": 201.34, "bytes_per_chunk": 5.086, "boundary_entropy": 0.16183, "router_noise": 0.0, "grad_norm": 0.8788, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 187.5} |
| {"event": "train", "step": 1750, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.54065, "bpb": 2.22269, "chunk_loss": 0.030456, "F": 0.20367, "G": 0.22167, "chunks": 208.56, "bytes_per_chunk": 4.91, "boundary_entropy": 0.13675, "router_noise": 0.0, "grad_norm": 0.8072, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 191.8} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3463.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198989, "cps": 27.054688, "mean_p": 0.207319, "boundary_entropy": 0.123627, "bpb": 2.135887}, "tr": {"chunks": 3456.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190908, "cps": 27.0, "mean_p": 0.246439, "boundary_entropy": 0.2695, "bpb": 2.366684}, "fi": {"chunks": 3409.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181224, "cps": 26.632812, "mean_p": 0.21801, "boundary_entropy": 0.212888, "bpb": 2.581253}}, "bpb": {"en": 2.13589, "tr": 2.36668, "fi": 2.58125}, "gini_cpc": 0.020737, "gini_cps": 0.003486, "bpic": 5.493029, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9593891686211966, "fi": 0.9107239263931762}, "t": 196.5} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.15295, "tr": 2.37573, "fi": 2.58907}, "bytes_per_chunk": {"en": 4.778, "tr": 5.67, "fi": 5.764}, "bpb_mean": 2.37258, "frozen_inventory": false, "t": 197.0} |
| {"event": "train", "step": 1800, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.59533, "bpb": 2.30158, "chunk_loss": 0.030407, "F": 0.19958, "G": 0.22527, "chunks": 204.38, "bytes_per_chunk": 5.01, "boundary_entropy": 0.15986, "router_noise": 0.0, "grad_norm": 0.8847, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.116, "alpha": 0.03, "t": 197.6} |
| {"event": "train", "step": 1850, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.53878, "bpb": 2.21998, "chunk_loss": 0.030423, "F": 0.20114, "G": 0.22266, "chunks": 205.97, "bytes_per_chunk": 4.972, "boundary_entropy": 0.15062, "router_noise": 0.0, "grad_norm": 0.7699, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 202.1} |
| {"event": "train", "step": 1900, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.58091, "bpb": 2.28077, "chunk_loss": 0.03042, "F": 0.20233, "G": 0.21982, "chunks": 207.19, "bytes_per_chunk": 4.942, "boundary_entropy": 0.13562, "router_noise": 0.0, "grad_norm": 0.806, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 206.7} |
| {"event": "train", "step": 1950, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.54146, "bpb": 2.22385, "chunk_loss": 0.03043, "F": 0.20023, "G": 0.22049, "chunks": 205.03, "bytes_per_chunk": 4.994, "boundary_entropy": 0.13526, "router_noise": 0.0, "grad_norm": 0.7753, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 211.1} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3439.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19761, "cps": 26.867188, "mean_p": 0.204373, "boundary_entropy": 0.108876, "bpb": 2.106523}, "tr": {"chunks": 3356.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185384, "cps": 26.21875, "mean_p": 0.242406, "boundary_entropy": 0.264709, "bpb": 2.338119}, "fi": {"chunks": 3372.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179257, "cps": 26.34375, "mean_p": 0.2151, "boundary_entropy": 0.206983, "bpb": 2.514036}}, "bpb": {"en": 2.10652, "tr": 2.33812, "fi": 2.51404}, "gini_cpc": 0.021761, "gini_cps": 0.005442, "bpic": 5.580014, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9381306930358457, "fi": 0.9071260251556673}, "t": 215.5} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.12345, "tr": 2.34591, "fi": 2.51994}, "bytes_per_chunk": {"en": 4.808, "tr": 5.712, "fi": 5.761}, "bpb_mean": 2.32977, "frozen_inventory": false, "t": 216.0} |
| {"event": "train", "step": 2000, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.56029, "bpb": 2.25103, "chunk_loss": 0.030342, "F": 0.19659, "G": 0.21789, "chunks": 201.31, "bytes_per_chunk": 5.087, "boundary_entropy": 0.14213, "router_noise": 0.0, "grad_norm": 0.7892, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.113, "alpha": 0.03, "t": 216.8} |
| {"event": "probe", "step": 2048, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3442.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197782, "cps": 26.890625, "mean_p": 0.205358, "boundary_entropy": 0.106357, "bpb": 2.097771}, "tr": {"chunks": 3255.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179804, "cps": 25.429688, "mean_p": 0.229955, "boundary_entropy": 0.242221, "bpb": 2.334735}, "fi": {"chunks": 3415.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181543, "cps": 26.679688, "mean_p": 0.216838, "boundary_entropy": 0.196545, "bpb": 2.504589}}, "bpb": {"en": 2.09777, "tr": 2.33473, "fi": 2.50459}, "gini_cpc": 0.021435, "gini_cps": 0.012329, "bpic": 5.610364, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9091042659469767, "fi": 0.9178930442092078}, "t": 221.6} |
| {"event": "train", "step": 2050, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.61932, "bpb": 2.33619, "chunk_loss": 0.03035, "F": 0.19595, "G": 0.22035, "chunks": 200.66, "bytes_per_chunk": 5.103, "boundary_entropy": 0.14472, "router_noise": 0.0, "grad_norm": 0.9084, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.117, "alpha": 0.03, "t": 222.6} |
| {"event": "train", "step": 2100, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.50427, "bpb": 2.1702, "chunk_loss": 0.03041, "F": 0.20148, "G": 0.22149, "chunks": 206.31, "bytes_per_chunk": 4.963, "boundary_entropy": 0.13854, "router_noise": 0.0, "grad_norm": 0.7817, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.101, "alpha": 0.03, "t": 227.7} |
| {"event": "train", "step": 2150, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.51784, "bpb": 2.18978, "chunk_loss": 0.030306, "F": 0.19537, "G": 0.21386, "chunks": 200.06, "bytes_per_chunk": 5.118, "boundary_entropy": 0.12798, "router_noise": 0.0, "grad_norm": 0.7233, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.079, "alpha": 0.03, "t": 231.6} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3456.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198586, "cps": 27.0, "mean_p": 0.205204, "boundary_entropy": 0.100809, "bpb": 2.071765}, "tr": {"chunks": 3314.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.183064, "cps": 25.890625, "mean_p": 0.240352, "boundary_entropy": 0.25636, "bpb": 2.293795}, "fi": {"chunks": 3431.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182393, "cps": 26.804688, "mean_p": 0.212443, "boundary_entropy": 0.188115, "bpb": 2.474633}}, "bpb": {"en": 2.07177, "tr": 2.2938, "fi": 2.47463}, "gini_cpc": 0.019139, "gini_cps": 0.00928, "bpic": 5.561416, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9218331867953132, "fi": 0.9184578301555236}, "t": 236.7} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.0898, "tr": 2.3028, "fi": 2.48327}, "bytes_per_chunk": {"en": 4.793, "tr": 5.702, "fi": 5.691}, "bpb_mean": 2.29195, "frozen_inventory": false, "t": 237.2} |
| {"event": "train", "step": 2200, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.48789, "bpb": 2.14657, "chunk_loss": 0.030411, "F": 0.20239, "G": 0.21846, "chunks": 207.25, "bytes_per_chunk": 4.941, "boundary_entropy": 0.12025, "router_noise": 0.0, "grad_norm": 0.6997, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.122, "alpha": 0.03, "t": 237.7} |
| {"event": "train", "step": 2250, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.51709, "bpb": 2.1887, "chunk_loss": 0.030319, "F": 0.19635, "G": 0.21523, "chunks": 201.06, "bytes_per_chunk": 5.093, "boundary_entropy": 0.12857, "router_noise": 0.0, "grad_norm": 0.7543, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 242.2} |
| {"event": "train", "step": 2300, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.50383, "bpb": 2.16956, "chunk_loss": 0.030364, "F": 0.20071, "G": 0.21467, "chunks": 205.53, "bytes_per_chunk": 4.982, "boundary_entropy": 0.1187, "router_noise": 0.0, "grad_norm": 0.763, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 246.6} |
| {"event": "train", "step": 2350, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.50118, "bpb": 2.16574, "chunk_loss": 0.030257, "F": 0.19, "G": 0.22234, "chunks": 194.56, "bytes_per_chunk": 5.263, "boundary_entropy": 0.15644, "router_noise": 0.0, "grad_norm": 0.7983, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 251.2} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3433.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197265, "cps": 26.820312, "mean_p": 0.20221, "boundary_entropy": 0.099117, "bpb": 2.051839}, "tr": {"chunks": 3259.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.180025, "cps": 25.460938, "mean_p": 0.232573, "boundary_entropy": 0.247175, "bpb": 2.265035}, "fi": {"chunks": 3386.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180001, "cps": 26.453125, "mean_p": 0.213677, "boundary_entropy": 0.19372, "bpb": 2.441311}}, "bpb": {"en": 2.05184, "tr": 2.26504, "fi": 2.44131}, "gini_cpc": 0.020652, "gini_cps": 0.01151, "bpic": 5.629292, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9126076938225723, "fi": 0.9124842711924154}, "t": 255.9} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.06863, "tr": 2.27215, "fi": 2.45309}, "bytes_per_chunk": {"en": 4.819, "tr": 5.892, "fi": 5.76}, "bpb_mean": 2.26463, "frozen_inventory": false, "t": 256.4} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 256.9} |
| {"event": "train", "step": 2400, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.45422, "bpb": 2.098, "chunk_loss": 0.030475, "F": 0.20569, "G": 0.21948, "chunks": 210.62, "bytes_per_chunk": 4.862, "boundary_entropy": 0.11922, "router_noise": 0.0, "grad_norm": 0.755, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.115, "alpha": 0.03, "t": 257.0} |
| {"event": "train", "step": 2450, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.64055, "bpb": 2.36681, "chunk_loss": 0.030425, "F": 0.20282, "G": 0.2204, "chunks": 207.69, "bytes_per_chunk": 4.93, "boundary_entropy": 0.12082, "router_noise": 0.0, "grad_norm": 0.7138, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 261.4} |
| {"event": "train", "step": 2500, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.49355, "bpb": 2.15474, "chunk_loss": 0.030356, "F": 0.19962, "G": 0.21468, "chunks": 204.41, "bytes_per_chunk": 5.01, "boundary_entropy": 0.11462, "router_noise": 0.0, "grad_norm": 0.7696, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.1, "alpha": 0.03, "t": 266.4} |
| {"event": "train", "step": 2550, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.48694, "bpb": 2.1452, "chunk_loss": 0.030411, "F": 0.20029, "G": 0.22202, "chunks": 205.09, "bytes_per_chunk": 4.993, "boundary_entropy": 0.13915, "router_noise": 0.0, "grad_norm": 0.7226, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 270.6} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3415.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196231, "cps": 26.679688, "mean_p": 0.201197, "boundary_entropy": 0.093544, "bpb": 2.03052}, "tr": {"chunks": 3260.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.180081, "cps": 25.46875, "mean_p": 0.233414, "boundary_entropy": 0.243367, "bpb": 2.243497}, "fi": {"chunks": 3387.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180054, "cps": 26.460938, "mean_p": 0.209415, "boundary_entropy": 0.179906, "bpb": 2.414173}}, "bpb": {"en": 2.03052, "tr": 2.2435, "fi": 2.41417}, "gini_cpc": 0.019383, "gini_cps": 0.01027, "bpic": 5.638243, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9176994276043162, "fi": 0.9175647593504097}, "t": 275.3} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.04539, "tr": 2.2504, "fi": 2.42185}, "bytes_per_chunk": {"en": 4.83, "tr": 5.904, "fi": 5.711}, "bpb_mean": 2.23921, "frozen_inventory": false, "t": 275.8} |
| {"event": "train", "step": 2600, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.50962, "bpb": 2.17792, "chunk_loss": 0.030311, "F": 0.19434, "G": 0.21384, "chunks": 199.0, "bytes_per_chunk": 5.146, "boundary_entropy": 0.12651, "router_noise": 0.0, "grad_norm": 0.7425, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.118, "alpha": 0.03, "t": 276.5} |
| {"event": "train", "step": 2650, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.46619, "bpb": 2.11526, "chunk_loss": 0.030308, "F": 0.19333, "G": 0.21661, "chunks": 197.97, "bytes_per_chunk": 5.173, "boundary_entropy": 0.13406, "router_noise": 0.0, "grad_norm": 0.7497, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 280.8} |
| {"event": "train", "step": 2700, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.546, "bpb": 2.2304, "chunk_loss": 0.030364, "F": 0.19833, "G": 0.21719, "chunks": 203.09, "bytes_per_chunk": 5.042, "boundary_entropy": 0.12519, "router_noise": 0.0, "grad_norm": 0.831, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 285.1} |
| {"event": "train", "step": 2750, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.48079, "bpb": 2.13633, "chunk_loss": 0.030303, "F": 0.19516, "G": 0.21593, "chunks": 199.84, "bytes_per_chunk": 5.124, "boundary_entropy": 0.12219, "router_noise": 0.0, "grad_norm": 0.7356, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 289.6} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3441.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197725, "cps": 26.882812, "mean_p": 0.202344, "boundary_entropy": 0.092955, "bpb": 2.010708}, "tr": {"chunks": 3189.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176159, "cps": 24.914062, "mean_p": 0.223866, "boundary_entropy": 0.22384, "bpb": 2.225846}, "fi": {"chunks": 3295.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175163, "cps": 25.742188, "mean_p": 0.20113, "boundary_entropy": 0.165398, "bpb": 2.377736}}, "bpb": {"en": 2.01071, "tr": 2.22585, "fi": 2.37774}, "gini_cpc": 0.027394, "gini_cps": 0.016927, "bpic": 5.716071, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8909296560835336, "fi": 0.8858964942742279}, "t": 294.2} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.02775, "tr": 2.23204, "fi": 2.38654}, "bytes_per_chunk": {"en": 4.808, "tr": 6.01, "fi": 5.864}, "bpb_mean": 2.21544, "frozen_inventory": false, "t": 294.7} |
| {"event": "train", "step": 2800, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.46201, "bpb": 2.10923, "chunk_loss": 0.030386, "F": 0.20184, "G": 0.21319, "chunks": 206.69, "bytes_per_chunk": 4.954, "boundary_entropy": 0.10656, "router_noise": 0.0, "grad_norm": 0.6956, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.122, "alpha": 0.03, "t": 295.7} |
| {"event": "train", "step": 2850, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.54804, "bpb": 2.23334, "chunk_loss": 0.030397, "F": 0.20001, "G": 0.21555, "chunks": 204.81, "bytes_per_chunk": 5.0, "boundary_entropy": 0.11089, "router_noise": 0.0, "grad_norm": 0.7631, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.097, "alpha": 0.03, "t": 300.5} |
| {"event": "train", "step": 2900, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.42387, "bpb": 2.05422, "chunk_loss": 0.030294, "F": 0.19415, "G": 0.21508, "chunks": 198.81, "bytes_per_chunk": 5.151, "boundary_entropy": 0.12356, "router_noise": 0.0, "grad_norm": 0.7248, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 304.8} |
| {"event": "train", "step": 2950, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.36254, "bpb": 1.96573, "chunk_loss": 0.03052, "F": 0.20975, "G": 0.22047, "chunks": 214.78, "bytes_per_chunk": 4.768, "boundary_entropy": 0.09942, "router_noise": 0.0, "grad_norm": 0.6641, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 308.9} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3417.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196345, "cps": 26.695312, "mean_p": 0.201712, "boundary_entropy": 0.085913, "bpb": 1.997979}, "tr": {"chunks": 3225.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178147, "cps": 25.195312, "mean_p": 0.225904, "boundary_entropy": 0.228678, "bpb": 2.200097}, "fi": {"chunks": 3324.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.176705, "cps": 25.96875, "mean_p": 0.205964, "boundary_entropy": 0.167644, "bpb": 2.355058}}, "bpb": {"en": 1.99798, "tr": 2.2001, "fi": 2.35506}, "gini_cpc": 0.023755, "gini_cps": 0.012844, "bpic": 5.692555, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.907315455696229, "fi": 0.8999704980866695}, "t": 313.7} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.01218, "tr": 2.20828, "fi": 2.36316}, "bytes_per_chunk": {"en": 4.814, "tr": 5.924, "fi": 5.745}, "bpb_mean": 2.19454, "frozen_inventory": false, "t": 314.2} |
| {"event": "train", "step": 3000, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.45436, "bpb": 2.0982, "chunk_loss": 0.030342, "F": 0.19717, "G": 0.21531, "chunks": 201.91, "bytes_per_chunk": 5.072, "boundary_entropy": 0.11791, "router_noise": 0.0, "grad_norm": 0.8023, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.121, "alpha": 0.03, "t": 315.0} |
| {"event": "train", "step": 3050, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.41723, "bpb": 2.04464, "chunk_loss": 0.030355, "F": 0.20026, "G": 0.21457, "chunks": 205.06, "bytes_per_chunk": 4.994, "boundary_entropy": 0.10752, "router_noise": 0.0, "grad_norm": 0.6985, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.102, "alpha": 0.03, "t": 320.1} |
| {"event": "train", "step": 3100, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.51262, "bpb": 2.18224, "chunk_loss": 0.030276, "F": 0.19379, "G": 0.21099, "chunks": 198.44, "bytes_per_chunk": 5.16, "boundary_entropy": 0.11152, "router_noise": 0.0, "grad_norm": 0.8406, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.106, "alpha": 0.03, "t": 325.4} |
| {"event": "train", "step": 3150, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.41494, "bpb": 2.04133, "chunk_loss": 0.030275, "F": 0.19498, "G": 0.20738, "chunks": 199.66, "bytes_per_chunk": 5.129, "boundary_entropy": 0.10194, "router_noise": 0.0, "grad_norm": 0.7018, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 329.9} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3417.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196345, "cps": 26.695312, "mean_p": 0.200884, "boundary_entropy": 0.0823, "bpb": 1.978666}, "tr": {"chunks": 3253.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179694, "cps": 25.414062, "mean_p": 0.227598, "boundary_entropy": 0.224015, "bpb": 2.168638}, "fi": {"chunks": 3398.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180639, "cps": 26.546875, "mean_p": 0.204867, "boundary_entropy": 0.165716, "bpb": 2.330256}}, "bpb": {"en": 1.97867, "tr": 2.16864, "fi": 2.33026}, "gini_cpc": 0.019941, "gini_cps": 0.010859, "bpic": 5.634883, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9151929232185527, "fi": 0.920005942388238}, "t": 334.7} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 1.99451, "tr": 2.17706, "fi": 2.34029}, "bytes_per_chunk": {"en": 4.825, "tr": 5.935, "fi": 5.703}, "bpb_mean": 2.17062, "frozen_inventory": false, "t": 335.2} |
| {"event": "train", "step": 3200, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.34695, "bpb": 1.94324, "chunk_loss": 0.030371, "F": 0.20312, "G": 0.21125, "chunks": 208.0, "bytes_per_chunk": 4.923, "boundary_entropy": 0.09005, "router_noise": 0.0, "grad_norm": 0.738, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.119, "alpha": 0.03, "t": 335.8} |
| {"event": "train", "step": 3250, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.34876, "bpb": 1.94585, "chunk_loss": 0.030373, "F": 0.19965, "G": 0.21259, "chunks": 204.44, "bytes_per_chunk": 5.009, "boundary_entropy": 0.10336, "router_noise": 0.0, "grad_norm": 0.6606, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 340.0} |
| {"event": "train", "step": 3300, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.47685, "bpb": 2.13064, "chunk_loss": 0.030367, "F": 0.1994, "G": 0.21384, "chunks": 204.19, "bytes_per_chunk": 5.015, "boundary_entropy": 0.10241, "router_noise": 0.0, "grad_norm": 0.737, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.099, "alpha": 0.03, "t": 344.9} |
| {"event": "train", "step": 3350, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.42453, "bpb": 2.05516, "chunk_loss": 0.030283, "F": 0.19522, "G": 0.20933, "chunks": 199.91, "bytes_per_chunk": 5.122, "boundary_entropy": 0.1, "router_noise": 0.0, "grad_norm": 0.7324, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 349.2} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3405.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195656, "cps": 26.601562, "mean_p": 0.199318, "boundary_entropy": 0.078832, "bpb": 1.961684}, "tr": {"chunks": 3182.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175772, "cps": 24.859375, "mean_p": 0.221524, "boundary_entropy": 0.214709, "bpb": 2.149737}, "fi": {"chunks": 3337.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177396, "cps": 26.070312, "mean_p": 0.201789, "boundary_entropy": 0.153606, "bpb": 2.30881}}, "bpb": {"en": 1.96168, "tr": 2.14974, "fi": 2.30881}, "gini_cpc": 0.024153, "gini_cps": 0.014981, "bpic": 5.716647, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8983728692958867, "fi": 0.9066743448685124}, "t": 353.7} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 1.97963, "tr": 2.15797, "fi": 2.31645}, "bytes_per_chunk": {"en": 4.838, "tr": 5.984, "fi": 5.799}, "bpb_mean": 2.15135, "frozen_inventory": false, "t": 354.4} |
| {"event": "train", "step": 3400, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.38333, "bpb": 1.99572, "chunk_loss": 0.030369, "F": 0.20145, "G": 0.21147, "chunks": 206.28, "bytes_per_chunk": 4.964, "boundary_entropy": 0.09346, "router_noise": 0.0, "grad_norm": 0.7363, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.121, "alpha": 0.03, "t": 355.2} |
| {"event": "train", "step": 3450, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.53978, "bpb": 2.22144, "chunk_loss": 0.030414, "F": 0.20035, "G": 0.21536, "chunks": 205.16, "bytes_per_chunk": 4.991, "boundary_entropy": 0.10613, "router_noise": 0.0, "grad_norm": 0.7496, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.094, "alpha": 0.03, "t": 359.9} |
| {"event": "train", "step": 3500, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.41566, "bpb": 2.04237, "chunk_loss": 0.030301, "F": 0.19598, "G": 0.20926, "chunks": 200.69, "bytes_per_chunk": 5.102, "boundary_entropy": 0.0999, "router_noise": 0.0, "grad_norm": 0.7546, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 364.4} |
| {"event": "train", "step": 3550, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.40683, "bpb": 2.02962, "chunk_loss": 0.030288, "F": 0.19592, "G": 0.20859, "chunks": 200.62, "bytes_per_chunk": 5.104, "boundary_entropy": 0.09614, "router_noise": 0.0, "grad_norm": 0.6856, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.099, "alpha": 0.03, "t": 369.4} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3426.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196863, "cps": 26.765625, "mean_p": 0.2008, "boundary_entropy": 0.078625, "bpb": 1.944696}, "tr": {"chunks": 3250.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179528, "cps": 25.390625, "mean_p": 0.222803, "boundary_entropy": 0.212239, "bpb": 2.138896}, "fi": {"chunks": 3297.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.17527, "cps": 25.757812, "mean_p": 0.199193, "boundary_entropy": 0.153865, "bpb": 2.290514}}, "bpb": {"en": 1.9447, "tr": 2.1389, "fi": 2.29051}, "gini_cpc": 0.026094, "gini_cps": 0.011765, "bpic": 5.688559, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9119469414799319, "fi": 0.890315276460535}, "t": 374.1} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 1.96047, "tr": 2.14466, "fi": 2.30073}, "bytes_per_chunk": {"en": 4.818, "tr": 5.946, "fi": 5.857}, "bpb_mean": 2.13529, "frozen_inventory": false, "t": 374.7} |
| {"event": "train", "step": 3600, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.43468, "bpb": 2.06981, "chunk_loss": 0.030315, "F": 0.19598, "G": 0.21683, "chunks": 200.69, "bytes_per_chunk": 5.102, "boundary_entropy": 0.11678, "router_noise": 0.0, "grad_norm": 0.7079, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.123, "alpha": 0.03, "t": 375.5} |
| {"event": "train", "step": 3650, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.35789, "bpb": 1.95903, "chunk_loss": 0.030328, "F": 0.20029, "G": 0.20872, "chunks": 205.09, "bytes_per_chunk": 4.993, "boundary_entropy": 0.08902, "router_noise": 0.0, "grad_norm": 0.6787, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 380.1} |
| {"event": "train", "step": 3700, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.46657, "bpb": 2.11582, "chunk_loss": 0.030255, "F": 0.19046, "G": 0.2087, "chunks": 195.03, "bytes_per_chunk": 5.25, "boundary_entropy": 0.10946, "router_noise": 0.0, "grad_norm": 0.7864, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 384.7} |
| {"event": "train", "step": 3750, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.45032, "bpb": 2.09237, "chunk_loss": 0.030484, "F": 0.20108, "G": 0.21662, "chunks": 205.91, "bytes_per_chunk": 4.973, "boundary_entropy": 0.10524, "router_noise": 0.0, "grad_norm": 0.7922, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 389.3} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3391.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194851, "cps": 26.492188, "mean_p": 0.198342, "boundary_entropy": 0.075543, "bpb": 1.931849}, "tr": {"chunks": 3256.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.17986, "cps": 25.4375, "mean_p": 0.219374, "boundary_entropy": 0.205952, "bpb": 2.126911}, "fi": {"chunks": 3386.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180001, "cps": 26.453125, "mean_p": 0.201504, "boundary_entropy": 0.150418, "bpb": 2.273202}}, "bpb": {"en": 1.93185, "tr": 2.12691, "fi": 2.2732}, "gini_cpc": 0.018017, "gini_cps": 0.00897, "bpic": 5.65454, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9230605177721446, "fi": 0.9237860521980424}, "t": 394.2} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 1.9491, "tr": 2.13428, "fi": 2.28197}, "bytes_per_chunk": {"en": 4.843, "tr": 5.858, "fi": 5.716}, "bpb_mean": 2.12178, "frozen_inventory": false, "t": 394.6} |
| {"event": "train", "step": 3800, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.34277, "bpb": 1.93721, "chunk_loss": 0.030361, "F": 0.20193, "G": 0.20918, "chunks": 206.78, "bytes_per_chunk": 4.952, "boundary_entropy": 0.08514, "router_noise": 0.0, "grad_norm": 0.7159, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.123, "alpha": 0.03, "t": 395.5} |
| {"event": "train", "step": 3850, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.35465, "bpb": 1.95434, "chunk_loss": 0.030389, "F": 0.20261, "G": 0.21159, "chunks": 207.47, "bytes_per_chunk": 4.936, "boundary_entropy": 0.08805, "router_noise": 0.0, "grad_norm": 0.718, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 400.2} |
| {"event": "train", "step": 3900, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.42794, "bpb": 2.06008, "chunk_loss": 0.030313, "F": 0.19754, "G": 0.2135, "chunks": 202.28, "bytes_per_chunk": 5.062, "boundary_entropy": 0.1029, "router_noise": 0.0, "grad_norm": 0.7496, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 404.7} |
| {"event": "train", "step": 3950, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.42333, "bpb": 2.05343, "chunk_loss": 0.030329, "F": 0.19864, "G": 0.2073, "chunks": 203.41, "bytes_per_chunk": 5.034, "boundary_entropy": 0.0826, "router_noise": 0.0, "grad_norm": 0.7174, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 409.3} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3413.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196116, "cps": 26.664062, "mean_p": 0.199952, "boundary_entropy": 0.075117, "bpb": 1.923393}, "tr": {"chunks": 3183.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175827, "cps": 24.867188, "mean_p": 0.215927, "boundary_entropy": 0.200945, "bpb": 2.1076}, "fi": {"chunks": 3345.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177821, "cps": 26.132812, "mean_p": 0.200664, "boundary_entropy": 0.147632, "bpb": 2.26688}}, "bpb": {"en": 1.92339, "tr": 2.1076, "fi": 2.26688}, "gini_cpc": 0.024603, "gini_cps": 0.015424, "bpic": 5.706871, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8965487700932738, "fi": 0.9067176518318146}, "t": 413.8} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 1.93959, "tr": 2.11501, "fi": 2.27599}, "bytes_per_chunk": {"en": 4.828, "tr": 5.986, "fi": 5.79}, "bpb_mean": 2.1102, "frozen_inventory": false, "t": 414.4} |
| {"event": "train", "step": 4000, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.3651, "bpb": 1.96943, "chunk_loss": 0.030287, "F": 0.19342, "G": 0.21038, "chunks": 198.06, "bytes_per_chunk": 5.17, "boundary_entropy": 0.10623, "router_noise": 0.0, "grad_norm": 0.707, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.117, "alpha": 0.03, "t": 415.2} |
| {"event": "train", "step": 4050, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.4995, "bpb": 2.16333, "chunk_loss": 0.030329, "F": 0.19409, "G": 0.21106, "chunks": 198.75, "bytes_per_chunk": 5.152, "boundary_entropy": 0.10429, "router_noise": 0.0, "grad_norm": 0.8185, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 419.5} |
| {"event": "probe", "step": 4096, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3386.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194564, "cps": 26.453125, "mean_p": 0.197826, "boundary_entropy": 0.072276, "bpb": 1.91954}, "tr": {"chunks": 3207.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.177153, "cps": 25.054688, "mean_p": 0.215542, "boundary_entropy": 0.198167, "bpb": 2.100894}, "fi": {"chunks": 3310.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175961, "cps": 25.859375, "mean_p": 0.197191, "boundary_entropy": 0.143579, "bpb": 2.25184}}, "bpb": {"en": 1.91954, "tr": 2.10089, "fi": 2.25184}, "gini_cpc": 0.022645, "gini_cps": 0.01205, "bpic": 5.728769, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9105117924833805, "fi": 0.9043848462696183}, "t": 424.0} |
| {"event": "train", "step": 4100, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.43677, "bpb": 2.07282, "chunk_loss": 0.030323, "F": 0.19659, "G": 0.21352, "chunks": 201.31, "bytes_per_chunk": 5.087, "boundary_entropy": 0.10566, "router_noise": 0.0, "grad_norm": 0.719, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.111, "alpha": 0.03, "t": 425.0} |
| {"event": "train", "step": 4150, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.37507, "bpb": 1.9838, "chunk_loss": 0.030417, "F": 0.20206, "G": 0.21273, "chunks": 206.91, "bytes_per_chunk": 4.949, "boundary_entropy": 0.08693, "router_noise": 0.0, "grad_norm": 0.6748, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.098, "alpha": 0.03, "t": 429.9} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3391.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194851, "cps": 26.492188, "mean_p": 0.198501, "boundary_entropy": 0.072691, "bpb": 1.911432}, "tr": {"chunks": 3239.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178921, "cps": 25.304688, "mean_p": 0.220356, "boundary_entropy": 0.206992, "bpb": 2.09142}, "fi": {"chunks": 3317.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.176333, "cps": 25.914062, "mean_p": 0.197014, "boundary_entropy": 0.146863, "bpb": 2.250054}}, "bpb": {"en": 1.91143, "tr": 2.09142, "fi": 2.25005}, "gini_cpc": 0.022442, "gini_cps": 0.010187, "bpic": 5.703428, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9182410986068724, "fi": 0.9049611149264343}, "t": 434.6} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 1.9274, "tr": 2.10043, "fi": 2.25576}, "bytes_per_chunk": {"en": 4.847, "tr": 5.905, "fi": 5.848}, "bpb_mean": 2.09453, "frozen_inventory": false, "t": 435.2} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 435.8} |
| {"event": "train", "step": 4200, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.4532, "bpb": 2.09652, "chunk_loss": 0.030382, "F": 0.20117, "G": 0.2141, "chunks": 206.0, "bytes_per_chunk": 4.971, "boundary_entropy": 0.09208, "router_noise": 0.0, "grad_norm": 0.7079, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.119, "alpha": 0.03, "t": 435.9} |
| {"event": "train", "step": 4250, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.34376, "bpb": 1.93864, "chunk_loss": 0.030277, "F": 0.19565, "G": 0.2054, "chunks": 200.34, "bytes_per_chunk": 5.111, "boundary_entropy": 0.08293, "router_noise": 0.0, "grad_norm": 0.6845, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 440.2} |
| {"event": "train", "step": 4300, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.35757, "bpb": 1.95855, "chunk_loss": 0.030237, "F": 0.19055, "G": 0.21107, "chunks": 195.12, "bytes_per_chunk": 5.248, "boundary_entropy": 0.11206, "router_noise": 0.0, "grad_norm": 0.7142, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.079, "alpha": 0.03, "t": 444.2} |
| {"event": "train", "step": 4350, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.33258, "bpb": 1.92251, "chunk_loss": 0.030271, "F": 0.19412, "G": 0.20946, "chunks": 198.78, "bytes_per_chunk": 5.151, "boundary_entropy": 0.09738, "router_noise": 0.0, "grad_norm": 0.704, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 448.3} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3387.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194622, "cps": 26.460938, "mean_p": 0.197997, "boundary_entropy": 0.068504, "bpb": 1.901901}, "tr": {"chunks": 3235.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.1787, "cps": 25.273438, "mean_p": 0.2198, "boundary_entropy": 0.202875, "bpb": 2.078699}, "fi": {"chunks": 3299.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175376, "cps": 25.773438, "mean_p": 0.196589, "boundary_entropy": 0.139236, "bpb": 2.230777}}, "bpb": {"en": 1.9019, "tr": 2.0787, "fi": 2.23078}, "gini_cpc": 0.023383, "gini_cps": 0.010214, "bpic": 5.718375, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.918190208406409, "fi": 0.9011132085946169}, "t": 453.4} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 1.91806, "tr": 2.08612, "fi": 2.23776}, "bytes_per_chunk": {"en": 4.846, "tr": 5.904, "fi": 5.855}, "bpb_mean": 2.08064, "frozen_inventory": false, "t": 454.0} |
| {"event": "train", "step": 4400, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.36957, "bpb": 1.97588, "chunk_loss": 0.030405, "F": 0.2045, "G": 0.2113, "chunks": 209.41, "bytes_per_chunk": 4.89, "boundary_entropy": 0.07547, "router_noise": 0.0, "grad_norm": 0.7681, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.127, "alpha": 0.03, "t": 454.7} |
| {"event": "train", "step": 4450, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.39727, "bpb": 2.01584, "chunk_loss": 0.030249, "F": 0.19308, "G": 0.20746, "chunks": 197.72, "bytes_per_chunk": 5.179, "boundary_entropy": 0.09238, "router_noise": 0.0, "grad_norm": 0.6991, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 458.7} |
| {"event": "train", "step": 4500, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.4111, "bpb": 2.03579, "chunk_loss": 0.030291, "F": 0.19556, "G": 0.20688, "chunks": 200.25, "bytes_per_chunk": 5.114, "boundary_entropy": 0.08851, "router_noise": 0.0, "grad_norm": 0.6968, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.095, "alpha": 0.03, "t": 463.5} |
| {"event": "train", "step": 4550, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.32738, "bpb": 1.91501, "chunk_loss": 0.030268, "F": 0.19193, "G": 0.20806, "chunks": 196.53, "bytes_per_chunk": 5.21, "boundary_entropy": 0.0991, "router_noise": 0.0, "grad_norm": 0.6796, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 468.0} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3367.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193472, "cps": 26.304688, "mean_p": 0.196231, "boundary_entropy": 0.068381, "bpb": 1.890876}, "tr": {"chunks": 3186.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175993, "cps": 24.890625, "mean_p": 0.215474, "boundary_entropy": 0.195879, "bpb": 2.070177}, "fi": {"chunks": 3305.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175695, "cps": 25.820312, "mean_p": 0.196547, "boundary_entropy": 0.135699, "bpb": 2.21934}}, "bpb": {"en": 1.89088, "tr": 2.07018, "fi": 2.21934}, "gini_cpc": 0.02174, "gini_cps": 0.01224, "bpic": 5.75492, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9096539796423794, "fi": 0.9081144456722575}, "t": 472.4} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 1.90867, "tr": 2.07828, "fi": 2.22691}, "bytes_per_chunk": {"en": 4.867, "tr": 5.965, "fi": 5.807}, "bpb_mean": 2.07129, "frozen_inventory": false, "t": 473.1} |
| {"event": "train", "step": 4600, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.39258, "bpb": 2.00906, "chunk_loss": 0.030256, "F": 0.19327, "G": 0.20347, "chunks": 197.91, "bytes_per_chunk": 5.174, "boundary_entropy": 0.08438, "router_noise": 0.0, "grad_norm": 0.7196, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.114, "alpha": 0.03, "t": 473.8} |
| {"event": "train", "step": 4650, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.39016, "bpb": 2.00558, "chunk_loss": 0.030352, "F": 0.2002, "G": 0.21238, "chunks": 205.0, "bytes_per_chunk": 4.995, "boundary_entropy": 0.09139, "router_noise": 0.0, "grad_norm": 0.7877, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.104, "alpha": 0.03, "t": 478.9} |
| {"event": "train", "step": 4700, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.36702, "bpb": 1.97219, "chunk_loss": 0.030292, "F": 0.19492, "G": 0.21087, "chunks": 199.59, "bytes_per_chunk": 5.13, "boundary_entropy": 0.10168, "router_noise": 0.0, "grad_norm": 0.746, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 483.1} |
| {"event": "train", "step": 4750, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.46317, "bpb": 2.11091, "chunk_loss": 0.030302, "F": 0.19382, "G": 0.20866, "chunks": 198.47, "bytes_per_chunk": 5.16, "boundary_entropy": 0.09864, "router_noise": 0.0, "grad_norm": 0.8404, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 487.5} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3378.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194104, "cps": 26.390625, "mean_p": 0.196865, "boundary_entropy": 0.06862, "bpb": 1.88165}, "tr": {"chunks": 3186.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175993, "cps": 24.890625, "mean_p": 0.211649, "boundary_entropy": 0.190845, "bpb": 2.060974}, "fi": {"chunks": 3346.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177875, "cps": 26.140625, "mean_p": 0.195894, "boundary_entropy": 0.134319, "bpb": 2.206823}}, "bpb": {"en": 1.88165, "tr": 2.06097, "fi": 2.20682}, "gini_cpc": 0.022035, "gini_cps": 0.012916, "bpic": 5.724723, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9066918145221704, "fi": 0.9163861740319924}, "t": 491.9} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 1.89993, "tr": 2.06822, "fi": 2.21352}, "bytes_per_chunk": {"en": 4.85, "tr": 5.983, "fi": 5.784}, "bpb_mean": 2.06056, "frozen_inventory": false, "t": 492.5} |
| {"event": "train", "step": 4800, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.3447, "bpb": 1.94, "chunk_loss": 0.030279, "F": 0.19708, "G": 0.20542, "chunks": 201.81, "bytes_per_chunk": 5.074, "boundary_entropy": 0.07951, "router_noise": 0.0, "grad_norm": 0.7063, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.113, "alpha": 0.03, "t": 493.2} |
| {"event": "train", "step": 4850, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.33773, "bpb": 1.92993, "chunk_loss": 0.03041, "F": 0.20428, "G": 0.21334, "chunks": 209.19, "bytes_per_chunk": 4.895, "boundary_entropy": 0.08575, "router_noise": 0.0, "grad_norm": 0.6806, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 497.4} |
| {"event": "train", "step": 4900, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.39435, "bpb": 2.01162, "chunk_loss": 0.030304, "F": 0.19662, "G": 0.20926, "chunks": 201.34, "bytes_per_chunk": 5.086, "boundary_entropy": 0.08981, "router_noise": 0.0, "grad_norm": 0.8182, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.074, "alpha": 0.03, "t": 501.1} |
| {"event": "train", "step": 4950, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.33109, "bpb": 1.92035, "chunk_loss": 0.030459, "F": 0.19772, "G": 0.21446, "chunks": 202.47, "bytes_per_chunk": 5.058, "boundary_entropy": 0.10071, "router_noise": 0.0, "grad_norm": 0.7174, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 505.5} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3383.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194392, "cps": 26.429688, "mean_p": 0.197024, "boundary_entropy": 0.065896, "bpb": 1.875585}, "tr": {"chunks": 3253.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179694, "cps": 25.414062, "mean_p": 0.214434, "boundary_entropy": 0.191447, "bpb": 2.047277}, "fi": {"chunks": 3368.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179044, "cps": 26.3125, "mean_p": 0.195957, "boundary_entropy": 0.131682, "bpb": 2.193685}}, "bpb": {"en": 1.87559, "tr": 2.04728, "fi": 2.19369}, "gini_cpc": 0.018498, "gini_cps": 0.008663, "bpic": 5.670932, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9243908420448699, "fi": 0.9210481229080669}, "t": 510.0} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 1.89278, "tr": 2.05588, "fi": 2.20241}, "bytes_per_chunk": {"en": 4.851, "tr": 5.885, "fi": 5.776}, "bpb_mean": 2.05036, "frozen_inventory": false, "t": 510.5} |
| {"event": "train", "step": 5000, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.44019, "bpb": 2.07776, "chunk_loss": 0.0303, "F": 0.19373, "G": 0.20877, "chunks": 198.38, "bytes_per_chunk": 5.162, "boundary_entropy": 0.09331, "router_noise": 0.0, "grad_norm": 0.8199, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.113, "alpha": 0.03, "t": 511.1} |
| {"event": "train", "step": 5050, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.28632, "bpb": 1.85577, "chunk_loss": 0.030307, "F": 0.19855, "G": 0.20903, "chunks": 203.31, "bytes_per_chunk": 5.037, "boundary_entropy": 0.08364, "router_noise": 0.0, "grad_norm": 0.7432, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 515.3} |
| {"event": "train", "step": 5100, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.338, "bpb": 1.93033, "chunk_loss": 0.030224, "F": 0.19254, "G": 0.20247, "chunks": 197.16, "bytes_per_chunk": 5.194, "boundary_entropy": 0.08505, "router_noise": 0.0, "grad_norm": 0.7153, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 519.5} |
| {"event": "train", "step": 5150, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.37381, "bpb": 1.98199, "chunk_loss": 0.030373, "F": 0.19791, "G": 0.21044, "chunks": 202.66, "bytes_per_chunk": 5.053, "boundary_entropy": 0.09108, "router_noise": 0.0, "grad_norm": 0.7421, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 523.8} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3380.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194219, "cps": 26.40625, "mean_p": 0.196301, "boundary_entropy": 0.066111, "bpb": 1.868718}, "tr": {"chunks": 3225.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178147, "cps": 25.195312, "mean_p": 0.213916, "boundary_entropy": 0.189768, "bpb": 2.041202}, "fi": {"chunks": 3293.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175057, "cps": 25.726562, "mean_p": 0.193272, "boundary_entropy": 0.128295, "bpb": 2.188595}}, "bpb": {"en": 1.86872, "tr": 2.0412, "fi": 2.18859}, "gini_cpc": 0.023336, "gini_cps": 0.01044, "bpic": 5.731663, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9172476071343239, "fi": 0.9013371409590071}, "t": 528.4} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 1.88686, "tr": 2.04852, "fi": 2.19747}, "bytes_per_chunk": {"en": 4.856, "tr": 5.949, "fi": 5.862}, "bpb_mean": 2.04428, "frozen_inventory": false, "t": 528.9} |
| {"event": "train", "step": 5200, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.32251, "bpb": 1.90797, "chunk_loss": 0.030339, "F": 0.19925, "G": 0.20825, "chunks": 204.03, "bytes_per_chunk": 5.019, "boundary_entropy": 0.08137, "router_noise": 0.0, "grad_norm": 0.7246, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.115, "alpha": 0.03, "t": 529.5} |
| {"event": "train", "step": 5250, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.3129, "bpb": 1.89411, "chunk_loss": 0.030292, "F": 0.19516, "G": 0.20963, "chunks": 199.84, "bytes_per_chunk": 5.124, "boundary_entropy": 0.09331, "router_noise": 0.0, "grad_norm": 0.7096, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 533.9} |
| {"event": "train", "step": 5300, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.31296, "bpb": 1.89419, "chunk_loss": 0.0303, "F": 0.19864, "G": 0.20627, "chunks": 203.41, "bytes_per_chunk": 5.034, "boundary_entropy": 0.07915, "router_noise": 0.0, "grad_norm": 0.6985, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 538.1} |
| {"event": "train", "step": 5350, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.31516, "bpb": 1.89737, "chunk_loss": 0.030298, "F": 0.19702, "G": 0.20652, "chunks": 201.75, "bytes_per_chunk": 5.076, "boundary_entropy": 0.08194, "router_noise": 0.0, "grad_norm": 0.7596, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 542.7} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3393.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194966, "cps": 26.507812, "mean_p": 0.196819, "boundary_entropy": 0.067437, "bpb": 1.863193}, "tr": {"chunks": 3229.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178368, "cps": 25.226562, "mean_p": 0.215145, "boundary_entropy": 0.192718, "bpb": 2.034134}, "fi": {"chunks": 3373.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.17931, "cps": 26.351562, "mean_p": 0.196016, "boundary_entropy": 0.13238, "bpb": 2.178528}}, "bpb": {"en": 1.86319, "tr": 2.03413, "fi": 2.17853}, "gini_cpc": 0.020023, "gini_cps": 0.010939, "bpic": 5.676038, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9148665610425616, "fi": 0.9196968908601316}, "t": 547.9} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 1.88135, "tr": 2.04181, "fi": 2.18751}, "bytes_per_chunk": {"en": 4.844, "tr": 5.899, "fi": 5.735}, "bpb_mean": 2.03689, "frozen_inventory": false, "t": 548.3} |
| {"event": "train", "step": 5400, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.37599, "bpb": 1.98513, "chunk_loss": 0.030304, "F": 0.19498, "G": 0.21424, "chunks": 199.66, "bytes_per_chunk": 5.129, "boundary_entropy": 0.10901, "router_noise": 0.0, "grad_norm": 0.8857, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.137, "alpha": 0.03, "t": 549.5} |
| {"event": "train", "step": 5450, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.31841, "bpb": 1.90206, "chunk_loss": 0.030302, "F": 0.19751, "G": 0.20795, "chunks": 202.25, "bytes_per_chunk": 5.063, "boundary_entropy": 0.08499, "router_noise": 0.0, "grad_norm": 0.7353, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 553.9} |
| {"event": "train", "step": 5500, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.3176, "bpb": 1.90089, "chunk_loss": 0.030371, "F": 0.20209, "G": 0.21291, "chunks": 206.94, "bytes_per_chunk": 4.948, "boundary_entropy": 0.08429, "router_noise": 0.0, "grad_norm": 0.7102, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.081, "alpha": 0.03, "t": 558.0} |
| {"event": "train", "step": 5550, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.31457, "bpb": 1.89653, "chunk_loss": 0.030331, "F": 0.19662, "G": 0.20917, "chunks": 201.34, "bytes_per_chunk": 5.086, "boundary_entropy": 0.08829, "router_noise": 0.0, "grad_norm": 0.7158, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.076, "alpha": 0.03, "t": 561.8} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3386.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194564, "cps": 26.453125, "mean_p": 0.196938, "boundary_entropy": 0.063634, "bpb": 1.858274}, "tr": {"chunks": 3204.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176987, "cps": 25.03125, "mean_p": 0.214467, "boundary_entropy": 0.191255, "bpb": 2.030872}, "fi": {"chunks": 3353.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.178247, "cps": 26.195312, "mean_p": 0.196808, "boundary_entropy": 0.131951, "bpb": 2.173756}}, "bpb": {"en": 1.85827, "tr": 2.03087, "fi": 2.17376}, "gini_cpc": 0.021313, "gini_cps": 0.012203, "bpic": 5.705723, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9096600508627226, "fi": 0.9161336524296164}, "t": 566.5} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 1.87656, "tr": 2.0388, "fi": 2.18218}, "bytes_per_chunk": {"en": 4.849, "tr": 5.944, "fi": 5.767}, "bpb_mean": 2.03251, "frozen_inventory": false, "t": 567.2} |
| {"event": "train", "step": 5600, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.36755, "bpb": 1.97296, "chunk_loss": 0.030249, "F": 0.19211, "G": 0.20706, "chunks": 196.72, "bytes_per_chunk": 5.205, "boundary_entropy": 0.0934, "router_noise": 0.0, "grad_norm": 0.7647, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.122, "alpha": 0.03, "t": 567.9} |
| {"event": "train", "step": 5650, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.32065, "bpb": 1.90529, "chunk_loss": 0.030283, "F": 0.19391, "G": 0.20517, "chunks": 198.56, "bytes_per_chunk": 5.157, "boundary_entropy": 0.0882, "router_noise": 0.0, "grad_norm": 0.7385, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.102, "alpha": 0.03, "t": 573.0} |
| {"event": "train", "step": 5700, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.44709, "bpb": 2.0877, "chunk_loss": 0.030262, "F": 0.19226, "G": 0.20755, "chunks": 196.88, "bytes_per_chunk": 5.201, "boundary_entropy": 0.09633, "router_noise": 0.0, "grad_norm": 0.8335, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 577.5} |
| {"event": "train", "step": 5750, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.27429, "bpb": 1.83841, "chunk_loss": 0.030416, "F": 0.20288, "G": 0.21372, "chunks": 207.75, "bytes_per_chunk": 4.929, "boundary_entropy": 0.08661, "router_noise": 0.0, "grad_norm": 0.7512, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 581.8} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3384.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194449, "cps": 26.4375, "mean_p": 0.196653, "boundary_entropy": 0.064285, "bpb": 1.853399}, "tr": {"chunks": 3238.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178865, "cps": 25.296875, "mean_p": 0.214764, "boundary_entropy": 0.189648, "bpb": 2.02369}, "fi": {"chunks": 3327.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.176865, "cps": 25.992188, "mean_p": 0.195435, "boundary_entropy": 0.129241, "bpb": 2.168346}}, "bpb": {"en": 1.8534, "tr": 2.02369, "fi": 2.16835}, "gini_cpc": 0.021308, "gini_cps": 0.009783, "bpic": 5.702282, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9198564518321676, "fi": 0.9095669747329821}, "t": 586.6} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 1.87201, "tr": 2.03221, "fi": 2.17705}, "bytes_per_chunk": {"en": 4.851, "tr": 5.916, "fi": 5.827}, "bpb_mean": 2.02709, "frozen_inventory": false, "t": 587.2} |
| {"event": "train", "step": 5800, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.26129, "bpb": 1.81966, "chunk_loss": 0.030307, "F": 0.1972, "G": 0.20973, "chunks": 201.94, "bytes_per_chunk": 5.071, "boundary_entropy": 0.08735, "router_noise": 0.0, "grad_norm": 0.7624, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.122, "alpha": 0.03, "t": 587.9} |
| {"event": "train", "step": 5850, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.3814, "bpb": 1.99293, "chunk_loss": 0.030334, "F": 0.19464, "G": 0.21136, "chunks": 199.31, "bytes_per_chunk": 5.138, "boundary_entropy": 0.10089, "router_noise": 0.0, "grad_norm": 0.7937, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 592.2} |
| {"event": "train", "step": 5900, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.35752, "bpb": 1.95849, "chunk_loss": 0.03036, "F": 0.19992, "G": 0.21162, "chunks": 204.72, "bytes_per_chunk": 5.002, "boundary_entropy": 0.0898, "router_noise": 0.0, "grad_norm": 0.7494, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 596.6} |
| {"event": "train", "step": 5950, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.30889, "bpb": 1.88832, "chunk_loss": 0.030261, "F": 0.19226, "G": 0.20508, "chunks": 196.88, "bytes_per_chunk": 5.201, "boundary_entropy": 0.09206, "router_noise": 0.0, "grad_norm": 0.7396, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 600.8} |
| {"event": "train", "step": 5999, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.35642, "bpb": 1.9569, "chunk_loss": 0.030347, "F": 0.19934, "G": 0.2092, "chunks": 204.12, "bytes_per_chunk": 5.017, "boundary_entropy": 0.08507, "router_noise": 0.0, "grad_norm": 0.7347, "lr": 3e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 605.1} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3379.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194162, "cps": 26.398438, "mean_p": 0.196211, "boundary_entropy": 0.064795, "bpb": 1.852409}, "tr": {"chunks": 3199.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176711, "cps": 24.992188, "mean_p": 0.213154, "boundary_entropy": 0.188803, "bpb": 2.01886}, "fi": {"chunks": 3358.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.178513, "cps": 26.234375, "mean_p": 0.196278, "boundary_entropy": 0.129924, "bpb": 2.163747}}, "bpb": {"en": 1.85241, "tr": 2.01886, "fi": 2.16375}, "gini_cpc": 0.021176, "gini_cps": 0.012077, "bpic": 5.709742, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9101220095714507, "fi": 0.9194005025206029}, "t": 605.4} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 1.86952, "tr": 2.02676, "fi": 2.17251}, "bytes_per_chunk": {"en": 4.86, "tr": 5.948, "fi": 5.785}, "bpb_mean": 2.02293, "frozen_inventory": false, "t": 605.9} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s1/probe/reps/th0.30", "/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s1/probe/reps/th0.40", "/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s1/probe/reps/th0.50", "/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s1/probe/reps/th0.60", "/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s1/probe/reps/th0.70"], "t": 607.3} |
| {"event": "final", "run": "ali_pkm_s1", "arch": "pkm", "objective": "ratio", "condition": "A", "step": 6000, "params": 39425792, "seed": 1, "per_lang": {"en": {"cpc": 0.1941619261046946, "cps": 26.3984375, "bpb": 1.8760144233204743, "bytes_per_chunk": 4.847247647047946}, "tr": {"cpc": 0.17671104236866816, "cps": 24.9921875, "bpb": 2.017146923011753, "bytes_per_chunk": 5.970233553867702}, "fi": {"cpc": 0.1785125724310244, "cps": 26.234375, "bpb": 2.189288604082032, "bytes_per_chunk": 5.75805300209768}}, "bpb_mean": 2.0274833168047532, "gini_cpc": 0.021176244412134837, "gini_cps": 0.012077294685990392, "bpic": 5.709742351046699, "mean_boundary_entropy": 0.08506667613983154, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "probe_overhead_sec_total": 0.0, "probe_overhead_seq_frac": 0.0, "wall_sec": 608.1963963508606, "t": 608.2} |
|
|
