{"event": "start", "run": "ali_pkm_s0", "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": 6.8}
{"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 6.8}
{"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6384.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.366833, "cps": 49.875, "mean_p": 0.496725, "boundary_entropy": 0.686995, "bpb": 8.163205}, "tr": {"chunks": 5290.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.292217, "cps": 41.328125, "mean_p": 0.489567, "boundary_entropy": 0.687288, "bpb": 8.170913}, "fi": {"chunks": 8895.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.472862, "cps": 69.492188, "mean_p": 0.500688, "boundary_entropy": 0.687512, "bpb": 8.214113}}, "bpb": {"en": 8.1632, "tr": 8.17091, "fi": 8.21411}, "gini_cpc": 0.106395, "gini_cps": 0.116842, "bpic": 2.758131, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.7965927738475616, "fi": 1.2890367847879758}, "t": 8.2}
{"event": "train", "step": 0, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 5.66465, "bpb": 8.17237, "chunk_loss": 0.038537, "F": 0.28732, "G": 0.4892, "chunks": 294.22, "bytes_per_chunk": 3.48, "boundary_entropy": 0.69137, "router_noise": 0.0, "grad_norm": 11.014, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.018, "alpha": 0.03, "t": 9.3}
{"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6300.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.362007, "cps": 49.21875, "mean_p": 0.496449, "boundary_entropy": 0.686986, "bpb": 8.13129}, "tr": {"chunks": 5228.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.288792, "cps": 40.84375, "mean_p": 0.48937, "boundary_entropy": 0.687277, "bpb": 8.148735}, "fi": {"chunks": 8819.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.468821, "cps": 68.898438, "mean_p": 0.500414, "boundary_entropy": 0.687508, "bpb": 8.186495}}, "bpb": {"en": 8.13129, "tr": 8.14873, "fi": 8.1865}, "gini_cpc": 0.107197, "gini_cps": 0.117659, "bpic": 2.788224, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.7977532795143136, "fi": 1.295063400087588}, "t": 10.2}
{"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6166.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.354307, "cps": 48.171875, "mean_p": 0.49591, "boundary_entropy": 0.686966, "bpb": 8.067266}, "tr": {"chunks": 5127.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.283213, "cps": 40.054688, "mean_p": 0.488963, "boundary_entropy": 0.687253, "bpb": 8.104625}, "fi": {"chunks": 8673.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.46106, "cps": 67.757812, "mean_p": 0.499879, "boundary_entropy": 0.687499, "bpb": 8.130403}}, "bpb": {"en": 8.06727, "tr": 8.10462, "fi": 8.1304}, "gini_cpc": 0.107926, "gini_cps": 0.118401, "bpic": 2.84143, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.7993433491695421, "fi": 1.3013018966187253}, "t": 12.2}
{"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5655.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.324944, "cps": 44.179688, "mean_p": 0.493975, "boundary_entropy": 0.686887, "bpb": 7.852263}, "tr": {"chunks": 4730.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.261283, "cps": 36.953125, "mean_p": 0.487511, "boundary_entropy": 0.687164, "bpb": 7.95159}, "fi": {"chunks": 8065.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.428739, "cps": 63.007812, "mean_p": 0.497911, "boundary_entropy": 0.687457, "bpb": 7.941342}}, "bpb": {"en": 7.85226, "tr": 7.95159, "fi": 7.94134}, "gini_cpc": 0.109991, "gini_cps": 0.120506, "bpic": 3.074905, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8040852586679436, "fi": 1.319422844610785}, "t": 14.1}
{"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4318.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.248118, "cps": 33.734375, "mean_p": 0.488601, "boundary_entropy": 0.686589, "bpb": 7.290592}, "tr": {"chunks": 3740.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.206596, "cps": 29.21875, "mean_p": 0.483543, "boundary_entropy": 0.686875, "bpb": 7.525508}, "fi": {"chunks": 6331.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.336558, "cps": 49.460938, "mean_p": 0.49207, "boundary_entropy": 0.687247, "bpb": 7.435641}}, "bpb": {"en": 7.29059, "tr": 7.52551, "fi": 7.43564}, "gini_cpc": 0.109497, "gini_cps": 0.120045, "bpic": 3.942734, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8326500893182971, "fi": 1.3564441355883807}, "t": 16.1}
{"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4393.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.252428, "cps": 34.320312, "mean_p": 0.490856, "boundary_entropy": 0.686925, "bpb": 6.660971}, "tr": {"chunks": 4518.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.249572, "cps": 35.296875, "mean_p": 0.488316, "boundary_entropy": 0.687348, "bpb": 6.978847}, "fi": {"chunks": 5592.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.297273, "cps": 43.6875, "mean_p": 0.491195, "boundary_entropy": 0.68741, "bpb": 6.839705}}, "bpb": {"en": 6.66097, "tr": 6.97885, "fi": 6.8397}, "gini_cpc": 0.039787, "gini_cps": 0.055115, "bpic": 3.911742, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9886864725898402, "fi": 1.1776553143328299}, "t": 18.9}
{"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4013.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.230592, "cps": 31.351562, "mean_p": 0.494584, "boundary_entropy": 0.68755, "bpb": 6.307318}, "tr": {"chunks": 4068.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.224714, "cps": 31.78125, "mean_p": 0.491881, "boundary_entropy": 0.687956, "bpb": 6.636482}, "fi": {"chunks": 4313.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.229281, "cps": 33.695312, "mean_p": 0.492881, "boundary_entropy": 0.687997, "bpb": 6.431598}}, "bpb": {"en": 6.30732, "tr": 6.63648, "fi": 6.4316}, "gini_cpc": 0.005724, "gini_cps": 0.016137, "bpic": 4.577376, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9745078756429054, "fi": 0.9943116666061498}, "t": 23.2}
{"event": "train", "step": 50, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 4.22575, "bpb": 6.09647, "chunk_loss": 0.030278, "F": 0.17026, "G": 0.49113, "chunks": 174.34, "bytes_per_chunk": 5.873, "boundary_entropy": 0.69209, "router_noise": 0.0, "grad_norm": 2.4391, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.359, "alpha": 0.03, "t": 27.3}
{"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4096.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.235362, "cps": 32.0, "mean_p": 0.488477, "boundary_entropy": 0.686632, "bpb": 5.623363}, "tr": {"chunks": 3944.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.217864, "cps": 30.8125, "mean_p": 0.488364, "boundary_entropy": 0.68753, "bpb": 6.010324}, "fi": {"chunks": 5387.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.286375, "cps": 42.085938, "mean_p": 0.493073, "boundary_entropy": 0.687705, "bpb": 5.817171}}, "bpb": {"en": 5.62336, "tr": 6.01032, "fi": 5.81717}, "gini_cpc": 0.061755, "gini_cps": 0.071647, "bpic": 4.225218, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.925657932214274, "fi": 1.2167441429092352}, "t": 30.3}
{"event": "train", "step": 100, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 3.20122, "bpb": 4.61838, "chunk_loss": 0.031711, "F": 0.19785, "G": 0.42062, "chunks": 202.59, "bytes_per_chunk": 5.054, "boundary_entropy": 0.66572, "router_noise": 0.0, "grad_norm": 1.7175, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 36.9}
{"event": "probe", "step": 128, "phase": 1, "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": 0.37897, "boundary_entropy": 0.611724, "bpb": 4.196391}, "tr": {"chunks": 2771.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.153069, "cps": 21.648438, "mean_p": 0.410483, "boundary_entropy": 0.651016, "bpb": 4.522156}, "fi": {"chunks": 3139.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.16687, "cps": 24.523438, "mean_p": 0.390751, "boundary_entropy": 0.635004, "bpb": 4.43297}}, "bpb": {"en": 4.19639, "tr": 4.52216, "fi": 4.43297}, "gini_cpc": 0.055883, "gini_cps": 0.046174, "bpic": 6.082556, "worst_lang": "en", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.7795879465842637, "fi": 0.8498818873327484}, "t": 41.9}
{"event": "train", "step": 150, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.81465, "bpb": 4.06068, "chunk_loss": 0.031351, "F": 0.19815, "G": 0.36288, "chunks": 202.91, "bytes_per_chunk": 5.047, "boundary_entropy": 0.59433, "router_noise": 0.0, "grad_norm": 2.0106, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 45.7}
{"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2937.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168764, "cps": 22.945312, "mean_p": 0.370662, "boundary_entropy": 0.576076, "bpb": 3.648882}, "tr": {"chunks": 3213.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.177484, "cps": 25.101562, "mean_p": 0.374576, "boundary_entropy": 0.596908, "bpb": 3.794721}, "fi": {"chunks": 2993.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159109, "cps": 23.382812, "mean_p": 0.365921, "boundary_entropy": 0.602781, "bpb": 3.889555}}, "bpb": {"en": 3.64888, "tr": 3.79472, "fi": 3.88955}, "gini_cpc": 0.024241, "gini_cps": 0.020125, "bpic": 6.204966, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.0516720883908148, "fi": 0.942790086147517}, "t": 54.5}
{"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.6557, "tr": 3.80416, "fi": 3.88969}, "bytes_per_chunk": {"en": 5.785, "tr": 5.84, "fi": 6.613}, "bpb_mean": 3.78319, "frozen_inventory": false, "t": 55.6}
{"event": "train", "step": 200, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.5596, "bpb": 3.69272, "chunk_loss": 0.030186, "F": 0.16968, "G": 0.37401, "chunks": 173.75, "bytes_per_chunk": 5.894, "boundary_entropy": 0.58468, "router_noise": 0.0, "grad_norm": 1.1477, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.21, "alpha": 0.03, "t": 56.3}
{"event": "train", "step": 250, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.39313, "bpb": 3.45256, "chunk_loss": 0.032099, "F": 0.22275, "G": 0.34034, "chunks": 228.09, "bytes_per_chunk": 4.489, "boundary_entropy": 0.51893, "router_noise": 0.0, "grad_norm": 2.1217, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 64.6}
{"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3399.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195311, "cps": 26.554688, "mean_p": 0.319303, "boundary_entropy": 0.496789, "bpb": 3.349338}, "tr": {"chunks": 3350.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185052, "cps": 26.171875, "mean_p": 0.33313, "boundary_entropy": 0.5141, "bpb": 3.522568}, "fi": {"chunks": 3870.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.205731, "cps": 30.234375, "mean_p": 0.348346, "boundary_entropy": 0.553044, "bpb": 3.639689}}, "bpb": {"en": 3.34934, "tr": 3.52257, "fi": 3.63969}, "gini_cpc": 0.023521, "gini_cps": 0.032646, "bpic": 5.342499, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9474738037938151, "fi": 1.053348393409182}, "t": 66.4}
{"event": "train", "step": 300, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.27735, "bpb": 3.28553, "chunk_loss": 0.03119, "F": 0.20743, "G": 0.29559, "chunks": 212.41, "bytes_per_chunk": 4.821, "boundary_entropy": 0.42026, "router_noise": 0.0, "grad_norm": 1.2803, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.19, "alpha": 0.03, "t": 74.1}
{"event": "train", "step": 350, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.19285, "bpb": 3.16361, "chunk_loss": 0.030862, "F": 0.1973, "G": 0.29232, "chunks": 202.03, "bytes_per_chunk": 5.069, "boundary_entropy": 0.40988, "router_noise": 0.0, "grad_norm": 1.1287, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.157, "alpha": 0.03, "t": 81.9}
{"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3679.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.2114, "cps": 28.742188, "mean_p": 0.271664, "boundary_entropy": 0.352732, "bpb": 2.902508}, "tr": {"chunks": 3887.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.214716, "cps": 30.367188, "mean_p": 0.301217, "boundary_entropy": 0.435308, "bpb": 3.179606}, "fi": {"chunks": 4439.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.235979, "cps": 34.679688, "mean_p": 0.302442, "boundary_entropy": 0.42284, "bpb": 3.387244}}, "bpb": {"en": 2.90251, "tr": 3.17961, "fi": 3.38724}, "gini_cpc": 0.024748, "gini_cps": 0.042205, "bpic": 4.725698, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.0156833228937905, "fi": 1.1162657353980638}, "t": 90.3}
{"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 2.91623, "tr": 3.18637, "fi": 3.37975}, "bytes_per_chunk": {"en": 4.526, "tr": 5.256, "fi": 4.479}, "bpb_mean": 3.16078, "frozen_inventory": false, "t": 91.6}
{"event": "train", "step": 400, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.04585, "bpb": 2.95154, "chunk_loss": 0.031369, "F": 0.21857, "G": 0.28395, "chunks": 223.81, "bytes_per_chunk": 4.575, "boundary_entropy": 0.35969, "router_noise": 0.0, "grad_norm": 1.5073, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.209, "alpha": 0.03, "t": 92.4}
{"event": "train", "step": 450, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 2.01565, "bpb": 2.90797, "chunk_loss": 0.03112, "F": 0.2128, "G": 0.27318, "chunks": 217.91, "bytes_per_chunk": 4.699, "boundary_entropy": 0.34439, "router_noise": 0.0, "grad_norm": 1.3646, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.164, "alpha": 0.03, "t": 100.6}
{"event": "train", "step": 500, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.99743, "bpb": 2.88168, "chunk_loss": 0.03138, "F": 0.22235, "G": 0.27498, "chunks": 227.69, "bytes_per_chunk": 4.497, "boundary_entropy": 0.3314, "router_noise": 0.0, "grad_norm": 1.1611, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.161, "alpha": 0.03, "t": 108.7}
{"event": "train", "step": 550, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.97294, "bpb": 2.84636, "chunk_loss": 0.031312, "F": 0.22214, "G": 0.27182, "chunks": 227.47, "bytes_per_chunk": 4.502, "boundary_entropy": 0.31703, "router_noise": 0.0, "grad_norm": 1.2267, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.159, "alpha": 0.03, "t": 116.6}
{"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3702.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.212722, "cps": 28.921875, "mean_p": 0.256487, "boundary_entropy": 0.293744, "bpb": 2.598719}, "tr": {"chunks": 3744.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.206817, "cps": 29.25, "mean_p": 0.281484, "boundary_entropy": 0.386166, "bpb": 2.940939}, "fi": {"chunks": 4629.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.246079, "cps": 36.164062, "mean_p": 0.285712, "boundary_entropy": 0.337783, "bpb": 3.153334}}, "bpb": {"en": 2.59872, "tr": 2.94094, "fi": 3.15333}, "gini_cpc": 0.039325, "gini_cps": 0.05118, "bpic": 4.698302, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9722389019934642, "fi": 1.1568125808660354}, "t": 125.1}
{"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.61017, "tr": 2.9503, "fi": 3.15298}, "bytes_per_chunk": {"en": 4.461, "tr": 5.523, "fi": 4.245}, "bpb_mean": 2.90449, "frozen_inventory": false, "t": 126.4}
{"event": "train", "step": 600, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.87667, "bpb": 2.70746, "chunk_loss": 0.031212, "F": 0.22134, "G": 0.26831, "chunks": 226.66, "bytes_per_chunk": 4.518, "boundary_entropy": 0.30501, "router_noise": 0.0, "grad_norm": 1.2956, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.207, "alpha": 0.03, "t": 127.0}
{"event": "train", "step": 650, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.90716, "bpb": 2.75145, "chunk_loss": 0.031269, "F": 0.22122, "G": 0.26759, "chunks": 226.53, "bytes_per_chunk": 4.52, "boundary_entropy": 0.30042, "router_noise": 0.0, "grad_norm": 1.1682, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.156, "alpha": 0.03, "t": 134.8}
{"event": "train", "step": 700, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.85511, "bpb": 2.67636, "chunk_loss": 0.031154, "F": 0.21957, "G": 0.26328, "chunks": 224.84, "bytes_per_chunk": 4.554, "boundary_entropy": 0.29087, "router_noise": 0.0, "grad_norm": 1.2477, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.161, "alpha": 0.03, "t": 142.8}
{"event": "train", "step": 750, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.81561, "bpb": 2.61937, "chunk_loss": 0.030735, "F": 0.20187, "G": 0.25803, "chunks": 206.72, "bytes_per_chunk": 4.954, "boundary_entropy": 0.29273, "router_noise": 0.0, "grad_norm": 1.1556, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 150.9}
{"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3276.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188243, "cps": 25.59375, "mean_p": 0.237492, "boundary_entropy": 0.267425, "bpb": 2.439517}, "tr": {"chunks": 3727.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.205877, "cps": 29.117188, "mean_p": 0.25799, "boundary_entropy": 0.31951, "bpb": 2.786131}, "fi": {"chunks": 4626.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.24592, "cps": 36.140625, "mean_p": 0.280144, "boundary_entropy": 0.292116, "bpb": 2.965003}}, "bpb": {"en": 2.43952, "tr": 2.78613, "fi": 2.965}, "gini_cpc": 0.060076, "gini_cps": 0.077393, "bpic": 4.878493, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.0936769733792329, "fi": 1.3063933833430406}, "t": 159.6}
{"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.45259, "tr": 2.79767, "fi": 2.96968}, "bytes_per_chunk": {"en": 5.0, "tr": 5.461, "fi": 4.339}, "bpb_mean": 2.73998, "frozen_inventory": false, "t": 160.9}
{"event": "train", "step": 800, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.80828, "bpb": 2.6088, "chunk_loss": 0.030814, "F": 0.20486, "G": 0.2534, "chunks": 209.78, "bytes_per_chunk": 4.881, "boundary_entropy": 0.27673, "router_noise": 0.0, "grad_norm": 0.9179, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.213, "alpha": 0.03, "t": 161.6}
{"event": "train", "step": 850, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.77431, "bpb": 2.55979, "chunk_loss": 0.030755, "F": 0.20547, "G": 0.25147, "chunks": 210.41, "bytes_per_chunk": 4.867, "boundary_entropy": 0.27184, "router_noise": 0.0, "grad_norm": 1.0933, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.155, "alpha": 0.03, "t": 169.4}
{"event": "train", "step": 900, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.62278, "bpb": 2.34118, "chunk_loss": 0.030697, "F": 0.20322, "G": 0.24682, "chunks": 208.09, "bytes_per_chunk": 4.921, "boundary_entropy": 0.25636, "router_noise": 0.0, "grad_norm": 0.9316, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.149, "alpha": 0.03, "t": 176.8}
{"event": "train", "step": 950, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.78535, "bpb": 2.57572, "chunk_loss": 0.030807, "F": 0.20743, "G": 0.249, "chunks": 212.41, "bytes_per_chunk": 4.821, "boundary_entropy": 0.25407, "router_noise": 0.0, "grad_norm": 0.9788, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.161, "alpha": 0.03, "t": 184.9}
{"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3259.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187267, "cps": 25.460938, "mean_p": 0.230541, "boundary_entropy": 0.24684, "bpb": 2.360245}, "tr": {"chunks": 3788.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.209247, "cps": 29.59375, "mean_p": 0.25003, "boundary_entropy": 0.292115, "bpb": 2.658556}, "fi": {"chunks": 4636.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.246452, "cps": 36.21875, "mean_p": 0.275987, "boundary_entropy": 0.260568, "bpb": 2.833999}}, "bpb": {"en": 2.36025, "tr": 2.65856, "fi": 2.834}, "gini_cpc": 0.061367, "gini_cps": 0.078576, "bpic": 4.855945, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1173755875167763, "fi": 1.3160467093012191}, "t": 193.6}
{"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.36829, "tr": 2.66886, "fi": 2.84033}, "bytes_per_chunk": {"en": 5.042, "tr": 5.285, "fi": 4.319}, "bpb_mean": 2.62583, "frozen_inventory": false, "t": 194.9}
{"event": "train", "step": 1000, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.72884, "bpb": 2.49419, "chunk_loss": 0.03106, "F": 0.21335, "G": 0.2531, "chunks": 218.47, "bytes_per_chunk": 4.687, "boundary_entropy": 0.25214, "router_noise": 0.0, "grad_norm": 1.1161, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.225, "alpha": 0.03, "t": 196.1}
{"event": "train", "step": 1050, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.74437, "bpb": 2.51659, "chunk_loss": 0.030699, "F": 0.20441, "G": 0.2441, "chunks": 209.31, "bytes_per_chunk": 4.892, "boundary_entropy": 0.24414, "router_noise": 0.0, "grad_norm": 0.9367, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.159, "alpha": 0.03, "t": 204.1}
{"event": "train", "step": 1100, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.73801, "bpb": 2.50741, "chunk_loss": 0.030802, "F": 0.20859, "G": 0.24552, "chunks": 213.59, "bytes_per_chunk": 4.794, "boundary_entropy": 0.23132, "router_noise": 0.0, "grad_norm": 1.0426, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.149, "alpha": 0.03, "t": 211.5}
{"event": "train", "step": 1150, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.59521, "bpb": 2.3014, "chunk_loss": 0.030642, "F": 0.20389, "G": 0.23606, "chunks": 208.78, "bytes_per_chunk": 4.905, "boundary_entropy": 0.22581, "router_noise": 0.0, "grad_norm": 0.9428, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.155, "alpha": 0.03, "t": 219.3}
{"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3257.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187152, "cps": 25.445312, "mean_p": 0.221264, "boundary_entropy": 0.221296, "bpb": 2.284899}, "tr": {"chunks": 3787.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.209192, "cps": 29.585938, "mean_p": 0.23779, "boundary_entropy": 0.260286, "bpb": 2.560992}, "fi": {"chunks": 4718.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.250811, "cps": 36.859375, "mean_p": 0.274635, "boundary_entropy": 0.232972, "bpb": 2.736806}}, "bpb": {"en": 2.2849, "tr": 2.56099, "fi": 2.73681}, "gini_cpc": 0.065578, "gini_cps": 0.082809, "bpic": 4.823329, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.117766566577011, "fi": 1.3401469266858554}, "t": 227.9}
{"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.29427, "tr": 2.57045, "fi": 2.74667}, "bytes_per_chunk": {"en": 5.057, "tr": 5.29, "fi": 4.25}, "bpb_mean": 2.53713, "frozen_inventory": false, "t": 229.2}
{"event": "train", "step": 1200, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.636, "bpb": 2.36025, "chunk_loss": 0.030605, "F": 0.20328, "G": 0.23548, "chunks": 208.16, "bytes_per_chunk": 4.919, "boundary_entropy": 0.22356, "router_noise": 0.0, "grad_norm": 0.8114, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.213, "alpha": 0.03, "t": 229.9}
{"event": "train", "step": 1250, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.69086, "bpb": 2.43939, "chunk_loss": 0.030732, "F": 0.2081, "G": 0.23826, "chunks": 213.09, "bytes_per_chunk": 4.805, "boundary_entropy": 0.21983, "router_noise": 0.0, "grad_norm": 0.8871, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.159, "alpha": 0.03, "t": 237.9}
{"event": "train", "step": 1300, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.63211, "bpb": 2.35464, "chunk_loss": 0.030561, "F": 0.20129, "G": 0.23227, "chunks": 206.12, "bytes_per_chunk": 4.968, "boundary_entropy": 0.21572, "router_noise": 0.0, "grad_norm": 0.9316, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.161, "alpha": 0.03, "t": 245.9}
{"event": "train", "step": 1350, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.61711, "bpb": 2.333, "chunk_loss": 0.030812, "F": 0.21225, "G": 0.2364, "chunks": 217.34, "bytes_per_chunk": 4.711, "boundary_entropy": 0.20271, "router_noise": 0.0, "grad_norm": 1.0191, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.153, "alpha": 0.03, "t": 253.6}
{"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3268.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187784, "cps": 25.53125, "mean_p": 0.214618, "boundary_entropy": 0.196119, "bpb": 2.231815}, "tr": {"chunks": 3938.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.217533, "cps": 30.765625, "mean_p": 0.229666, "boundary_entropy": 0.228542, "bpb": 2.499804}, "fi": {"chunks": 4714.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.250598, "cps": 36.828125, "mean_p": 0.267047, "boundary_entropy": 0.205213, "bpb": 2.641856}}, "bpb": {"en": 2.23181, "tr": 2.4998, "fi": 2.64186}, "gini_cpc": 0.063844, "gini_cps": 0.080872, "bpic": 4.759396, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.158423162819636, "fi": 1.3345036534588377}, "t": 262.1}
{"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.24078, "tr": 2.50925, "fi": 2.65363}, "bytes_per_chunk": {"en": 5.095, "tr": 4.995, "fi": 4.216}, "bpb_mean": 2.46788, "frozen_inventory": false, "t": 263.4}
{"event": "train", "step": 1400, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.69736, "bpb": 2.44878, "chunk_loss": 0.030617, "F": 0.20319, "G": 0.23187, "chunks": 208.06, "bytes_per_chunk": 4.922, "boundary_entropy": 0.20673, "router_noise": 0.0, "grad_norm": 0.9021, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.211, "alpha": 0.03, "t": 264.1}
{"event": "train", "step": 1450, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.65055, "bpb": 2.38123, "chunk_loss": 0.030647, "F": 0.20612, "G": 0.23281, "chunks": 211.06, "bytes_per_chunk": 4.852, "boundary_entropy": 0.20232, "router_noise": 0.0, "grad_norm": 0.8298, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.157, "alpha": 0.03, "t": 272.0}
{"event": "train", "step": 1500, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.61162, "bpb": 2.32507, "chunk_loss": 0.030537, "F": 0.20367, "G": 0.2246, "chunks": 208.56, "bytes_per_chunk": 4.91, "boundary_entropy": 0.18845, "router_noise": 0.0, "grad_norm": 0.8951, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.152, "alpha": 0.03, "t": 279.6}
{"event": "train", "step": 1550, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.57514, "bpb": 2.27244, "chunk_loss": 0.030537, "F": 0.20227, "G": 0.22188, "chunks": 207.12, "bytes_per_chunk": 4.944, "boundary_entropy": 0.17869, "router_noise": 0.0, "grad_norm": 0.7638, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 287.7}
{"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3289.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.18899, "cps": 25.695312, "mean_p": 0.211031, "boundary_entropy": 0.180053, "bpb": 2.188763}, "tr": {"chunks": 3826.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.211346, "cps": 29.890625, "mean_p": 0.219879, "boundary_entropy": 0.203438, "bpb": 2.437759}, "fi": {"chunks": 4749.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.252459, "cps": 37.101562, "mean_p": 0.264072, "boundary_entropy": 0.183284, "bpb": 2.589903}}, "bpb": {"en": 2.18876, "tr": 2.43776, "fi": 2.5899}, "gini_cpc": 0.064817, "gini_cps": 0.082041, "bpic": 4.781861, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1182905655212672, "fi": 1.3358279707316927}, "t": 296.1}
{"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.20004, "tr": 2.44885, "fi": 2.59851}, "bytes_per_chunk": {"en": 5.006, "tr": 5.085, "fi": 4.197}, "bpb_mean": 2.4158, "frozen_inventory": false, "t": 297.2}
{"event": "train", "step": 1600, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.58077, "bpb": 2.28057, "chunk_loss": 0.030638, "F": 0.2074, "G": 0.22606, "chunks": 212.38, "bytes_per_chunk": 4.822, "boundary_entropy": 0.17873, "router_noise": 0.0, "grad_norm": 0.8866, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.204, "alpha": 0.03, "t": 297.9}
{"event": "train", "step": 1650, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.59909, "bpb": 2.307, "chunk_loss": 0.030513, "F": 0.20273, "G": 0.22174, "chunks": 207.59, "bytes_per_chunk": 4.933, "boundary_entropy": 0.1795, "router_noise": 0.0, "grad_norm": 0.7805, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.156, "alpha": 0.03, "t": 305.7}
{"event": "train", "step": 1700, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.61476, "bpb": 2.3296, "chunk_loss": 0.030567, "F": 0.20685, "G": 0.22325, "chunks": 211.81, "bytes_per_chunk": 4.834, "boundary_entropy": 0.17309, "router_noise": 0.0, "grad_norm": 0.8314, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.16, "alpha": 0.03, "t": 313.7}
{"event": "train", "step": 1750, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.56538, "bpb": 2.25836, "chunk_loss": 0.030436, "F": 0.19919, "G": 0.21778, "chunks": 203.97, "bytes_per_chunk": 5.02, "boundary_entropy": 0.17449, "router_noise": 0.0, "grad_norm": 0.7495, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 321.8}
{"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3267.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187726, "cps": 25.523438, "mean_p": 0.204976, "boundary_entropy": 0.16757, "bpb": 2.155856}, "tr": {"chunks": 3901.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215489, "cps": 30.476562, "mean_p": 0.216326, "boundary_entropy": 0.196381, "bpb": 2.389394}, "fi": {"chunks": 4760.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.253043, "cps": 37.1875, "mean_p": 0.261902, "boundary_entropy": 0.171081, "bpb": 2.525087}}, "bpb": {"en": 2.15586, "tr": 2.38939, "fi": 2.52509}, "gini_cpc": 0.066353, "gini_cps": 0.083445, "bpic": 4.756204, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.147890296353886, "fi": 1.3479384290466783}, "t": 330.5}
{"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.16662, "tr": 2.39861, "fi": 2.53531}, "bytes_per_chunk": {"en": 5.087, "tr": 5.095, "fi": 4.21}, "bpb_mean": 2.36685, "frozen_inventory": false, "t": 331.8}
{"event": "train", "step": 1800, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.53756, "bpb": 2.21823, "chunk_loss": 0.030404, "F": 0.19833, "G": 0.21412, "chunks": 203.09, "bytes_per_chunk": 5.042, "boundary_entropy": 0.16691, "router_noise": 0.0, "grad_norm": 0.7743, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.211, "alpha": 0.03, "t": 332.3}
{"event": "train", "step": 1850, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.5643, "bpb": 2.25681, "chunk_loss": 0.030541, "F": 0.20605, "G": 0.22014, "chunks": 211.0, "bytes_per_chunk": 4.853, "boundary_entropy": 0.16401, "router_noise": 0.0, "grad_norm": 0.8669, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.151, "alpha": 0.03, "t": 339.9}
{"event": "train", "step": 1900, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.50339, "bpb": 2.16894, "chunk_loss": 0.030507, "F": 0.20331, "G": 0.21813, "chunks": 208.19, "bytes_per_chunk": 4.919, "boundary_entropy": 0.15876, "router_noise": 0.0, "grad_norm": 0.7164, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 348.0}
{"event": "train", "step": 1950, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.57782, "bpb": 2.27631, "chunk_loss": 0.030666, "F": 0.20929, "G": 0.22486, "chunks": 214.31, "bytes_per_chunk": 4.778, "boundary_entropy": 0.16077, "router_noise": 0.0, "grad_norm": 0.7787, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.158, "alpha": 0.03, "t": 355.9}
{"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3277.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188301, "cps": 25.601562, "mean_p": 0.203862, "boundary_entropy": 0.157481, "bpb": 2.126752}, "tr": {"chunks": 3860.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213224, "cps": 30.15625, "mean_p": 0.215897, "boundary_entropy": 0.181391, "bpb": 2.344722}, "fi": {"chunks": 4732.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.251555, "cps": 36.96875, "mean_p": 0.261743, "boundary_entropy": 0.163317, "bpb": 2.480568}}, "bpb": {"en": 2.12675, "tr": 2.34472, "fi": 2.48057}, "gini_cpc": 0.06457, "gini_cps": 0.081726, "bpic": 4.779847, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1323597713696443, "fi": 1.3359202449524117}, "t": 364.6}
{"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.13716, "tr": 2.35685, "fi": 2.49282}, "bytes_per_chunk": {"en": 5.015, "tr": 5.107, "fi": 4.21}, "bpb_mean": 2.32894, "frozen_inventory": false, "t": 365.9}
{"event": "train", "step": 2000, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.5423, "bpb": 2.22507, "chunk_loss": 0.0306, "F": 0.21036, "G": 0.22299, "chunks": 215.41, "bytes_per_chunk": 4.754, "boundary_entropy": 0.15753, "router_noise": 0.0, "grad_norm": 0.8546, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.215, "alpha": 0.03, "t": 366.6}
{"event": "train", "step": 2050, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.55897, "bpb": 2.24912, "chunk_loss": 0.030416, "F": 0.19968, "G": 0.21536, "chunks": 204.47, "bytes_per_chunk": 5.008, "boundary_entropy": 0.15988, "router_noise": 0.0, "grad_norm": 0.8166, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.153, "alpha": 0.03, "t": 374.3}
{"event": "train", "step": 2100, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.48236, "bpb": 2.13859, "chunk_loss": 0.030388, "F": 0.19846, "G": 0.21212, "chunks": 203.22, "bytes_per_chunk": 5.039, "boundary_entropy": 0.14802, "router_noise": 0.0, "grad_norm": 0.7462, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.158, "alpha": 0.03, "t": 382.2}
{"event": "train", "step": 2150, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.48848, "bpb": 2.14743, "chunk_loss": 0.030574, "F": 0.20834, "G": 0.21969, "chunks": 213.34, "bytes_per_chunk": 4.8, "boundary_entropy": 0.14425, "router_noise": 0.0, "grad_norm": 0.7591, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.157, "alpha": 0.03, "t": 390.0}
{"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3251.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186807, "cps": 25.398438, "mean_p": 0.200417, "boundary_entropy": 0.146358, "bpb": 2.098515}, "tr": {"chunks": 3858.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213114, "cps": 30.140625, "mean_p": 0.211671, "boundary_entropy": 0.170861, "bpb": 2.309827}, "fi": {"chunks": 4810.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.255701, "cps": 37.578125, "mean_p": 0.26009, "boundary_entropy": 0.146385, "bpb": 2.458448}}, "bpb": {"en": 2.09851, "tr": 2.30983, "fi": 2.45845}, "gini_cpc": 0.070055, "gini_cps": 0.0872, "bpic": 4.759795, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1408244558679252, "fi": 1.368801094001803}, "t": 398.7}
{"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.11126, "tr": 2.31801, "fi": 2.47092}, "bytes_per_chunk": {"en": 5.01, "tr": 5.151, "fi": 4.164}, "bpb_mean": 2.30006, "frozen_inventory": false, "t": 400.0}
{"event": "train", "step": 2200, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.53244, "bpb": 2.21084, "chunk_loss": 0.030652, "F": 0.20828, "G": 0.22021, "chunks": 213.28, "bytes_per_chunk": 4.801, "boundary_entropy": 0.14483, "router_noise": 0.0, "grad_norm": 0.7698, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.212, "alpha": 0.03, "t": 400.7}
{"event": "train", "step": 2250, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.50416, "bpb": 2.17004, "chunk_loss": 0.030423, "F": 0.1998, "G": 0.21229, "chunks": 204.59, "bytes_per_chunk": 5.005, "boundary_entropy": 0.13929, "router_noise": 0.0, "grad_norm": 0.7967, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 408.9}
{"event": "train", "step": 2300, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.54434, "bpb": 2.22802, "chunk_loss": 0.030443, "F": 0.20038, "G": 0.21526, "chunks": 205.19, "bytes_per_chunk": 4.991, "boundary_entropy": 0.14439, "router_noise": 0.0, "grad_norm": 0.7177, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.156, "alpha": 0.03, "t": 416.7}
{"event": "train", "step": 2350, "arch": "pkm", "phase": 1, "objective": "global", "lm_loss": 1.46244, "bpb": 2.10985, "chunk_loss": 0.030456, "F": 0.20212, "G": 0.21438, "chunks": 206.97, "bytes_per_chunk": 4.948, "boundary_entropy": 0.13423, "router_noise": 0.0, "grad_norm": 0.7254, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.151, "alpha": 0.03, "t": 424.3}
{"event": "probe", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3260.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187324, "cps": 25.46875, "mean_p": 0.201979, "boundary_entropy": 0.134996, "bpb": 2.073266}, "tr": {"chunks": 3912.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216097, "cps": 30.5625, "mean_p": 0.21315, "boundary_entropy": 0.159102, "bpb": 2.273365}, "fi": {"chunks": 4825.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.256499, "cps": 37.695312, "mean_p": 0.263476, "boundary_entropy": 0.143599, "bpb": 2.398507}}, "bpb": {"en": 2.07327, "tr": 2.27337, "fi": 2.39851}, "gini_cpc": 0.069882, "gini_cps": 0.086966, "bpic": 4.728849, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1535988510191681, "fi": 1.3692790212488255}, "t": 432.3}
{"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.08454, "tr": 2.28064, "fi": 2.41212}, "bytes_per_chunk": {"en": 5.009, "tr": 4.998, "fi": 4.135}, "bpb_mean": 2.2591, "frozen_inventory": false, "t": 433.7}
{"event": "phase_transition", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 434.3}
{"event": "train", "step": 2400, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.46431, "bpb": 2.11255, "chunk_loss": 0.030454, "F": 0.20331, "G": 0.21496, "chunks": 208.19, "bytes_per_chunk": 4.919, "boundary_entropy": 0.13821, "router_noise": 0.0, "grad_norm": 0.6832, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.204, "alpha": 0.03, "t": 434.5}
{"event": "train", "step": 2450, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.47764, "bpb": 2.13179, "chunk_loss": 0.030667, "F": 0.21198, "G": 0.22333, "chunks": 217.06, "bytes_per_chunk": 4.718, "boundary_entropy": 0.13718, "router_noise": 0.0, "grad_norm": 0.8263, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.155, "alpha": 0.03, "t": 442.2}
{"event": "train", "step": 2500, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.57619, "bpb": 2.27396, "chunk_loss": 0.030548, "F": 0.2074, "G": 0.21716, "chunks": 212.38, "bytes_per_chunk": 4.822, "boundary_entropy": 0.13798, "router_noise": 0.0, "grad_norm": 0.7624, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.159, "alpha": 0.03, "t": 450.2}
{"event": "train", "step": 2550, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.50138, "bpb": 2.16604, "chunk_loss": 0.030414, "F": 0.20197, "G": 0.21432, "chunks": 206.81, "bytes_per_chunk": 4.951, "boundary_entropy": 0.13174, "router_noise": 0.0, "grad_norm": 0.7633, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.156, "alpha": 0.03, "t": 457.9}
{"event": "probe", "step": 2600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3282.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188588, "cps": 25.640625, "mean_p": 0.198531, "boundary_entropy": 0.129611, "bpb": 2.054865}, "tr": {"chunks": 3808.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210352, "cps": 29.75, "mean_p": 0.208829, "boundary_entropy": 0.154197, "bpb": 2.256325}, "fi": {"chunks": 4802.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.255276, "cps": 37.515625, "mean_p": 0.257778, "boundary_entropy": 0.131635, "bpb": 2.360229}}, "bpb": {"en": 2.05487, "tr": 2.25633, "fi": 2.36023}, "gini_cpc": 0.067957, "gini_cps": 0.085211, "bpic": 4.770602, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1154033172559903, "fi": 1.3536170491088249}, "t": 466.3}
{"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.06471, "tr": 2.26372, "fi": 2.37399}, "bytes_per_chunk": {"en": 5.023, "tr": 5.153, "fi": 4.159}, "bpb_mean": 2.23414, "frozen_inventory": false, "t": 467.6}
{"event": "train", "step": 2600, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.50627, "bpb": 2.17309, "chunk_loss": 0.030605, "F": 0.20929, "G": 0.2196, "chunks": 214.31, "bytes_per_chunk": 4.778, "boundary_entropy": 0.12871, "router_noise": 0.0, "grad_norm": 0.6996, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.208, "alpha": 0.03, "t": 468.3}
{"event": "train", "step": 2650, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.44228, "bpb": 2.08077, "chunk_loss": 0.030309, "F": 0.19556, "G": 0.20544, "chunks": 200.25, "bytes_per_chunk": 5.114, "boundary_entropy": 0.13137, "router_noise": 0.0, "grad_norm": 0.6525, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.16, "alpha": 0.03, "t": 476.3}
{"event": "train", "step": 2700, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.52201, "bpb": 2.1958, "chunk_loss": 0.030439, "F": 0.20126, "G": 0.21146, "chunks": 206.09, "bytes_per_chunk": 4.969, "boundary_entropy": 0.13411, "router_noise": 0.0, "grad_norm": 0.7353, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.163, "alpha": 0.03, "t": 484.5}
{"event": "train", "step": 2750, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.45897, "bpb": 2.10485, "chunk_loss": 0.030421, "F": 0.20285, "G": 0.213, "chunks": 207.72, "bytes_per_chunk": 4.93, "boundary_entropy": 0.13078, "router_noise": 0.0, "grad_norm": 0.7229, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.163, "alpha": 0.03, "t": 492.6}
{"event": "probe", "step": 2800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3263.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187496, "cps": 25.492188, "mean_p": 0.197974, "boundary_entropy": 0.131044, "bpb": 2.033299}, "tr": {"chunks": 3841.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212175, "cps": 30.007812, "mean_p": 0.206614, "boundary_entropy": 0.152099, "bpb": 2.237158}, "fi": {"chunks": 4738.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.251874, "cps": 37.015625, "mean_p": 0.255094, "boundary_entropy": 0.132154, "bpb": 2.335468}}, "bpb": {"en": 2.0333, "tr": 2.23716, "fi": 2.33547}, "gini_cpc": 0.065872, "gini_cps": 0.083038, "bpic": 4.790745, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1316204890092514, "fi": 1.3433532159906765}, "t": 500.6}
{"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.04524, "tr": 2.24406, "fi": 2.34723}, "bytes_per_chunk": {"en": 5.085, "tr": 5.19, "fi": 4.218}, "bpb_mean": 2.21217, "frozen_inventory": false, "t": 501.9}
{"event": "train", "step": 2800, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.45124, "bpb": 2.09369, "chunk_loss": 0.030426, "F": 0.20377, "G": 0.21191, "chunks": 208.66, "bytes_per_chunk": 4.908, "boundary_entropy": 0.12849, "router_noise": 0.0, "grad_norm": 0.7283, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.2, "alpha": 0.03, "t": 502.6}
{"event": "train", "step": 2850, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.55536, "bpb": 2.24391, "chunk_loss": 0.030637, "F": 0.20963, "G": 0.2198, "chunks": 214.66, "bytes_per_chunk": 4.77, "boundary_entropy": 0.12767, "router_noise": 0.0, "grad_norm": 0.7291, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.163, "alpha": 0.03, "t": 510.7}
{"event": "train", "step": 2900, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.4957, "bpb": 2.15783, "chunk_loss": 0.030648, "F": 0.21219, "G": 0.22024, "chunks": 217.28, "bytes_per_chunk": 4.713, "boundary_entropy": 0.12408, "router_noise": 0.0, "grad_norm": 0.6846, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.159, "alpha": 0.03, "t": 518.7}
{"event": "train", "step": 2950, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.43563, "bpb": 2.07118, "chunk_loss": 0.03038, "F": 0.19897, "G": 0.20889, "chunks": 203.75, "bytes_per_chunk": 5.026, "boundary_entropy": 0.12359, "router_noise": 0.0, "grad_norm": 0.679, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.158, "alpha": 0.03, "t": 526.6}
{"event": "probe", "step": 3000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3245.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186462, "cps": 25.351562, "mean_p": 0.196979, "boundary_entropy": 0.116556, "bpb": 2.015844}, "tr": {"chunks": 3856.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213003, "cps": 30.125, "mean_p": 0.203752, "boundary_entropy": 0.139874, "bpb": 2.202227}, "fi": {"chunks": 4763.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.253203, "cps": 37.210938, "mean_p": 0.255589, "boundary_entropy": 0.119925, "bpb": 2.312866}}, "bpb": {"en": 2.01584, "tr": 2.20223, "fi": 2.31287}, "gini_cpc": 0.068172, "gini_cps": 0.0853, "bpic": 4.781861, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1423413378351084, "fi": 1.3579322952942248}, "t": 534.9}
{"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.02673, "tr": 2.20873, "fi": 2.32397}, "bytes_per_chunk": {"en": 5.032, "tr": 5.163, "fi": 4.202}, "bpb_mean": 2.18647, "frozen_inventory": false, "t": 536.1}
{"event": "train", "step": 3000, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.45757, "bpb": 2.10282, "chunk_loss": 0.030402, "F": 0.20248, "G": 0.21041, "chunks": 207.34, "bytes_per_chunk": 4.939, "boundary_entropy": 0.11786, "router_noise": 0.0, "grad_norm": 0.6939, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.213, "alpha": 0.03, "t": 537.2}
{"event": "train", "step": 3050, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.44734, "bpb": 2.08807, "chunk_loss": 0.03053, "F": 0.20944, "G": 0.2171, "chunks": 214.47, "bytes_per_chunk": 4.775, "boundary_entropy": 0.11985, "router_noise": 0.0, "grad_norm": 0.669, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.154, "alpha": 0.03, "t": 544.9}
{"event": "train", "step": 3100, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.53981, "bpb": 2.22147, "chunk_loss": 0.030481, "F": 0.20309, "G": 0.21033, "chunks": 207.97, "bytes_per_chunk": 4.924, "boundary_entropy": 0.12349, "router_noise": 0.0, "grad_norm": 0.7435, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.161, "alpha": 0.03, "t": 553.0}
{"event": "train", "step": 3150, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.46298, "bpb": 2.11064, "chunk_loss": 0.030457, "F": 0.20395, "G": 0.21254, "chunks": 208.84, "bytes_per_chunk": 4.903, "boundary_entropy": 0.11694, "router_noise": 0.0, "grad_norm": 0.6953, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.157, "alpha": 0.03, "t": 560.8}
{"event": "probe", "step": 3200, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3220.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.185026, "cps": 25.15625, "mean_p": 0.194287, "boundary_entropy": 0.116551, "bpb": 1.998082}, "tr": {"chunks": 3807.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210297, "cps": 29.742188, "mean_p": 0.202685, "boundary_entropy": 0.138729, "bpb": 2.176903}, "fi": {"chunks": 4736.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.251768, "cps": 37.0, "mean_p": 0.252805, "boundary_entropy": 0.120514, "bpb": 2.288719}}, "bpb": {"en": 1.99808, "tr": 2.1769, "fi": 2.28872}, "gini_cpc": 0.068761, "gini_cps": 0.085919, "bpic": 4.822919, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1365814766640718, "fi": 1.3607177774600627}, "t": 569.6}
{"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.01147, "tr": 2.18602, "fi": 2.30097}, "bytes_per_chunk": {"en": 5.057, "tr": 5.16, "fi": 4.226}, "bpb_mean": 2.16615, "frozen_inventory": false, "t": 571.0}
{"event": "train", "step": 3200, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.42956, "bpb": 2.06242, "chunk_loss": 0.03057, "F": 0.20914, "G": 0.21549, "chunks": 214.16, "bytes_per_chunk": 4.782, "boundary_entropy": 0.1188, "router_noise": 0.0, "grad_norm": 0.6969, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.217, "alpha": 0.03, "t": 571.7}
{"event": "train", "step": 3250, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.40745, "bpb": 2.03052, "chunk_loss": 0.030296, "F": 0.19623, "G": 0.20416, "chunks": 200.94, "bytes_per_chunk": 5.096, "boundary_entropy": 0.11416, "router_noise": 0.0, "grad_norm": 0.6895, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.154, "alpha": 0.03, "t": 579.4}
{"event": "train", "step": 3300, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.45177, "bpb": 2.09446, "chunk_loss": 0.030474, "F": 0.20535, "G": 0.21141, "chunks": 210.28, "bytes_per_chunk": 4.87, "boundary_entropy": 0.11537, "router_noise": 0.0, "grad_norm": 0.704, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 587.5}
{"event": "train", "step": 3350, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.48179, "bpb": 2.13777, "chunk_loss": 0.030653, "F": 0.21368, "G": 0.21802, "chunks": 218.81, "bytes_per_chunk": 4.68, "boundary_entropy": 0.11786, "router_noise": 0.0, "grad_norm": 0.76, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.157, "alpha": 0.03, "t": 595.3}
{"event": "probe", "step": 3400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3233.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.185773, "cps": 25.257812, "mean_p": 0.195086, "boundary_entropy": 0.115629, "bpb": 1.983137}, "tr": {"chunks": 3803.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210076, "cps": 29.710938, "mean_p": 0.203191, "boundary_entropy": 0.143494, "bpb": 2.15444}, "fi": {"chunks": 4743.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.25214, "cps": 37.054688, "mean_p": 0.251783, "boundary_entropy": 0.119969, "bpb": 2.262152}}, "bpb": {"en": 1.98314, "tr": 2.15444, "fi": 2.26215}, "gini_cpc": 0.06828, "gini_cps": 0.085463, "bpic": 4.816368, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1308218451453491, "fi": 1.3572493951957973}, "t": 604.0}
{"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 1.99644, "tr": 2.16463, "fi": 2.27289}, "bytes_per_chunk": {"en": 5.089, "tr": 5.165, "fi": 4.216}, "bpb_mean": 2.14465, "frozen_inventory": false, "t": 605.4}
{"event": "train", "step": 3400, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.39866, "bpb": 2.01783, "chunk_loss": 0.030356, "F": 0.19852, "G": 0.20429, "chunks": 203.28, "bytes_per_chunk": 5.037, "boundary_entropy": 0.11434, "router_noise": 0.0, "grad_norm": 0.7073, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.212, "alpha": 0.03, "t": 605.9}
{"event": "train", "step": 3450, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.39599, "bpb": 2.01399, "chunk_loss": 0.030298, "F": 0.19635, "G": 0.204, "chunks": 201.06, "bytes_per_chunk": 5.093, "boundary_entropy": 0.10979, "router_noise": 0.0, "grad_norm": 0.6579, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 614.0}
{"event": "train", "step": 3500, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.48017, "bpb": 2.13543, "chunk_loss": 0.030548, "F": 0.20618, "G": 0.21244, "chunks": 211.12, "bytes_per_chunk": 4.85, "boundary_entropy": 0.11101, "router_noise": 0.0, "grad_norm": 0.6952, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.153, "alpha": 0.03, "t": 621.7}
{"event": "train", "step": 3550, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.48241, "bpb": 2.13866, "chunk_loss": 0.030691, "F": 0.21365, "G": 0.22151, "chunks": 218.78, "bytes_per_chunk": 4.68, "boundary_entropy": 0.11772, "router_noise": 0.0, "grad_norm": 0.7436, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.159, "alpha": 0.03, "t": 629.7}
{"event": "probe", "step": 3600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3226.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.18537, "cps": 25.203125, "mean_p": 0.195028, "boundary_entropy": 0.106404, "bpb": 1.967314}, "tr": {"chunks": 3810.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210462, "cps": 29.765625, "mean_p": 0.202281, "boundary_entropy": 0.132178, "bpb": 2.143699}, "fi": {"chunks": 4760.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.253043, "cps": 37.1875, "mean_p": 0.252119, "boundary_entropy": 0.108969, "bpb": 2.243616}}, "bpb": {"en": 1.96731, "tr": 2.1437, "fi": 2.24362}, "gini_cpc": 0.069529, "gini_cps": 0.086696, "bpic": 4.809427, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.135361547422771, "fi": 1.365069698603688}, "t": 638.5}
{"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 1.98099, "tr": 2.15027, "fi": 2.25546}, "bytes_per_chunk": {"en": 5.048, "tr": 5.202, "fi": 4.197}, "bpb_mean": 2.12891, "frozen_inventory": false, "t": 639.8}
{"event": "train", "step": 3600, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.46121, "bpb": 2.10808, "chunk_loss": 0.030358, "F": 0.19934, "G": 0.20717, "chunks": 204.12, "bytes_per_chunk": 5.017, "boundary_entropy": 0.11105, "router_noise": 0.0, "grad_norm": 0.6988, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.214, "alpha": 0.03, "t": 640.3}
{"event": "train", "step": 3650, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.47724, "bpb": 2.13121, "chunk_loss": 0.030603, "F": 0.21133, "G": 0.2169, "chunks": 216.41, "bytes_per_chunk": 4.732, "boundary_entropy": 0.11476, "router_noise": 0.0, "grad_norm": 0.6981, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.161, "alpha": 0.03, "t": 648.4}
{"event": "train", "step": 3700, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.37853, "bpb": 1.98879, "chunk_loss": 0.030368, "F": 0.20081, "G": 0.20723, "chunks": 205.62, "bytes_per_chunk": 4.98, "boundary_entropy": 0.11004, "router_noise": 0.0, "grad_norm": 0.6961, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.153, "alpha": 0.03, "t": 656.0}
{"event": "train", "step": 3750, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.39931, "bpb": 2.01878, "chunk_loss": 0.03057, "F": 0.21082, "G": 0.21429, "chunks": 215.88, "bytes_per_chunk": 4.743, "boundary_entropy": 0.1048, "router_noise": 0.0, "grad_norm": 0.6563, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.154, "alpha": 0.03, "t": 663.7}
{"event": "probe", "step": 3800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3260.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187324, "cps": 25.46875, "mean_p": 0.195674, "boundary_entropy": 0.111331, "bpb": 1.952829}, "tr": {"chunks": 3853.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212838, "cps": 30.101562, "mean_p": 0.203245, "boundary_entropy": 0.134186, "bpb": 2.122362}, "fi": {"chunks": 4784.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254319, "cps": 37.375, "mean_p": 0.2521, "boundary_entropy": 0.111478, "bpb": 2.221751}}, "bpb": {"en": 1.95283, "tr": 2.12236, "fi": 2.22175}, "gini_cpc": 0.068243, "gini_cps": 0.0854, "bpic": 4.768597, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1362005043396868, "fi": 1.357643696923188}, "t": 672.5}
{"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 1.96657, "tr": 2.12981, "fi": 2.23188}, "bytes_per_chunk": {"en": 5.047, "tr": 5.167, "fi": 4.185}, "bpb_mean": 2.10942, "frozen_inventory": false, "t": 673.9}
{"event": "train", "step": 3800, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.38519, "bpb": 1.9984, "chunk_loss": 0.030462, "F": 0.20004, "G": 0.20531, "chunks": 204.84, "bytes_per_chunk": 4.999, "boundary_entropy": 0.11409, "router_noise": 0.0, "grad_norm": 0.6724, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.215, "alpha": 0.03, "t": 674.5}
{"event": "train", "step": 3850, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.372, "bpb": 1.97938, "chunk_loss": 0.030266, "F": 0.19577, "G": 0.20042, "chunks": 200.47, "bytes_per_chunk": 5.108, "boundary_entropy": 0.11025, "router_noise": 0.0, "grad_norm": 0.7392, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 682.7}
{"event": "train", "step": 3900, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.40791, "bpb": 2.03118, "chunk_loss": 0.030393, "F": 0.20224, "G": 0.20684, "chunks": 207.09, "bytes_per_chunk": 4.945, "boundary_entropy": 0.10969, "router_noise": 0.0, "grad_norm": 0.6805, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.155, "alpha": 0.03, "t": 690.4}
{"event": "train", "step": 3950, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.4317, "bpb": 2.06551, "chunk_loss": 0.030429, "F": 0.20404, "G": 0.20874, "chunks": 208.94, "bytes_per_chunk": 4.901, "boundary_entropy": 0.10763, "router_noise": 0.0, "grad_norm": 0.6873, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.154, "alpha": 0.03, "t": 698.1}
{"event": "probe", "step": 4000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3254.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186979, "cps": 25.421875, "mean_p": 0.193368, "boundary_entropy": 0.10061, "bpb": 1.941202}, "tr": {"chunks": 3848.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212561, "cps": 30.0625, "mean_p": 0.200517, "boundary_entropy": 0.128242, "bpb": 2.106363}, "fi": {"chunks": 4784.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254319, "cps": 37.375, "mean_p": 0.252468, "boundary_entropy": 0.104146, "bpb": 2.205625}}, "bpb": {"en": 1.9412, "tr": 2.10636, "fi": 2.20563}, "gini_cpc": 0.068659, "gini_cps": 0.085815, "bpic": 4.77301, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1368183719324316, "fi": 1.36014703502446}, "t": 706.8}
{"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 1.95627, "tr": 2.11489, "fi": 2.21622}, "bytes_per_chunk": {"en": 5.022, "tr": 5.154, "fi": 4.197}, "bpb_mean": 2.0958, "frozen_inventory": false, "t": 708.2}
{"event": "train", "step": 4000, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.45388, "bpb": 2.0975, "chunk_loss": 0.030479, "F": 0.2074, "G": 0.21168, "chunks": 212.38, "bytes_per_chunk": 4.822, "boundary_entropy": 0.10468, "router_noise": 0.0, "grad_norm": 0.776, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.212, "alpha": 0.03, "t": 708.8}
{"event": "train", "step": 4050, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.37354, "bpb": 1.9816, "chunk_loss": 0.030387, "F": 0.2012, "G": 0.206, "chunks": 206.03, "bytes_per_chunk": 4.97, "boundary_entropy": 0.09941, "router_noise": 0.0, "grad_norm": 0.6721, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.16, "alpha": 0.03, "t": 716.8}
{"event": "train", "step": 4100, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.44193, "bpb": 2.08027, "chunk_loss": 0.030591, "F": 0.21115, "G": 0.21566, "chunks": 216.22, "bytes_per_chunk": 4.736, "boundary_entropy": 0.10538, "router_noise": 0.0, "grad_norm": 0.7283, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.157, "alpha": 0.03, "t": 724.6}
{"event": "train", "step": 4150, "arch": "pkm", "phase": 2, "objective": "global", "lm_loss": 1.39326, "bpb": 2.01005, "chunk_loss": 0.030679, "F": 0.21503, "G": 0.21862, "chunks": 220.19, "bytes_per_chunk": 4.651, "boundary_entropy": 0.1003, "router_noise": 0.0, "grad_norm": 0.7056, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.161, "alpha": 0.03, "t": 732.6}
{"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3257.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187152, "cps": 25.445312, "mean_p": 0.193587, "boundary_entropy": 0.099907, "bpb": 1.929583}, "tr": {"chunks": 3827.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.211401, "cps": 29.898438, "mean_p": 0.199824, "boundary_entropy": 0.128047, "bpb": 2.081066}, "fi": {"chunks": 4808.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.255595, "cps": 37.5625, "mean_p": 0.252069, "boundary_entropy": 0.10199, "bpb": 2.182711}}, "bpb": {"en": 1.92958, "tr": 2.08107, "fi": 2.18271}, "gini_cpc": 0.069753, "gini_cps": 0.086949, "bpic": 4.770602, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1295729205941962, "fi": 1.365711408119032}, "t": 740.8}
{"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 1.94255, "tr": 2.09103, "fi": 2.19188}, "bytes_per_chunk": {"en": 5.02, "tr": 5.179, "fi": 4.162}, "bpb_mean": 2.07515, "frozen_inventory": false, "t": 742.2}
{"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 742.6}
{"event": "train", "step": 4200, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.38058, "bpb": 1.99176, "chunk_loss": 0.030322, "F": 0.19989, "G": 0.20394, "chunks": 204.69, "bytes_per_chunk": 5.003, "boundary_entropy": 0.09971, "router_noise": 0.0, "grad_norm": 0.6643, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.202, "alpha": 0.03, "t": 742.8}
{"event": "train", "step": 4250, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.3872, "bpb": 2.00131, "chunk_loss": 0.030348, "F": 0.19806, "G": 0.2027, "chunks": 202.81, "bytes_per_chunk": 5.049, "boundary_entropy": 0.10454, "router_noise": 0.0, "grad_norm": 0.7179, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.16, "alpha": 0.03, "t": 750.8}
{"event": "train", "step": 4300, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.47225, "bpb": 2.12401, "chunk_loss": 0.030454, "F": 0.20535, "G": 0.21039, "chunks": 210.28, "bytes_per_chunk": 4.87, "boundary_entropy": 0.10975, "router_noise": 0.0, "grad_norm": 0.7982, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.156, "alpha": 0.03, "t": 758.6}
{"event": "train", "step": 4350, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.3239, "bpb": 1.90998, "chunk_loss": 0.030452, "F": 0.20285, "G": 0.20608, "chunks": 207.72, "bytes_per_chunk": 4.93, "boundary_entropy": 0.10054, "router_noise": 0.0, "grad_norm": 0.711, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.155, "alpha": 0.03, "t": 766.3}
{"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3270.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187899, "cps": 25.546875, "mean_p": 0.193497, "boundary_entropy": 0.100355, "bpb": 1.916452}, "tr": {"chunks": 3860.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213224, "cps": 30.15625, "mean_p": 0.201351, "boundary_entropy": 0.129434, "bpb": 2.068318}, "fi": {"chunks": 4786.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254426, "cps": 37.390625, "mean_p": 0.251559, "boundary_entropy": 0.104065, "bpb": 2.16602}}, "bpb": {"en": 1.91645, "tr": 2.06832, "fi": 2.16602}, "gini_cpc": 0.067655, "gini_cps": 0.084816, "bpic": 4.760994, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1347837831126375, "fi": 1.354057722423782}, "t": 774.9}
{"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 1.93163, "tr": 2.07659, "fi": 2.17524}, "bytes_per_chunk": {"en": 5.008, "tr": 5.139, "fi": 4.183}, "bpb_mean": 2.06115, "frozen_inventory": false, "t": 776.1}
{"event": "train", "step": 4400, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.44181, "bpb": 2.08009, "chunk_loss": 0.030664, "F": 0.21188, "G": 0.21548, "chunks": 216.97, "bytes_per_chunk": 4.72, "boundary_entropy": 0.10431, "router_noise": 0.0, "grad_norm": 0.7478, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.207, "alpha": 0.03, "t": 776.7}
{"event": "train", "step": 4450, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.37992, "bpb": 1.99081, "chunk_loss": 0.030463, "F": 0.20554, "G": 0.20943, "chunks": 210.47, "bytes_per_chunk": 4.865, "boundary_entropy": 0.10385, "router_noise": 0.0, "grad_norm": 0.682, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.156, "alpha": 0.03, "t": 784.5}
{"event": "train", "step": 4500, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.38678, "bpb": 2.0007, "chunk_loss": 0.030555, "F": 0.21033, "G": 0.21406, "chunks": 215.38, "bytes_per_chunk": 4.754, "boundary_entropy": 0.10175, "router_noise": 0.0, "grad_norm": 0.7346, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 792.6}
{"event": "train", "step": 4550, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.36799, "bpb": 1.9736, "chunk_loss": 0.030465, "F": 0.20752, "G": 0.21132, "chunks": 212.5, "bytes_per_chunk": 4.819, "boundary_entropy": 0.10207, "router_noise": 0.0, "grad_norm": 0.6832, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.158, "alpha": 0.03, "t": 800.5}
{"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3247.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186577, "cps": 25.367188, "mean_p": 0.192726, "boundary_entropy": 0.09743, "bpb": 1.906995}, "tr": {"chunks": 3866.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213556, "cps": 30.203125, "mean_p": 0.199634, "boundary_entropy": 0.128494, "bpb": 2.054104}, "fi": {"chunks": 4795.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254904, "cps": 37.460938, "mean_p": 0.251149, "boundary_entropy": 0.100258, "bpb": 2.156157}}, "bpb": {"en": 1.907, "tr": 2.0541, "fi": 2.15616}, "gini_cpc": 0.06954, "gini_cps": 0.086664, "bpic": 4.764192, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.144598387752824, "fi": 1.36621345978705}, "t": 809.5}
{"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 1.92132, "tr": 2.06335, "fi": 2.16672}, "bytes_per_chunk": {"en": 5.014, "tr": 5.138, "fi": 4.175}, "bpb_mean": 2.05046, "frozen_inventory": false, "t": 810.8}
{"event": "train", "step": 4600, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.42585, "bpb": 2.05707, "chunk_loss": 0.030417, "F": 0.20242, "G": 0.20637, "chunks": 207.28, "bytes_per_chunk": 4.94, "boundary_entropy": 0.10265, "router_noise": 0.0, "grad_norm": 0.7529, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.22, "alpha": 0.03, "t": 811.5}
{"event": "train", "step": 4650, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.34202, "bpb": 1.93613, "chunk_loss": 0.030426, "F": 0.20435, "G": 0.20709, "chunks": 209.25, "bytes_per_chunk": 4.894, "boundary_entropy": 0.10301, "router_noise": 0.0, "grad_norm": 0.6543, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.15, "alpha": 0.03, "t": 819.0}
{"event": "train", "step": 4700, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.35889, "bpb": 1.96047, "chunk_loss": 0.030426, "F": 0.20349, "G": 0.20648, "chunks": 208.38, "bytes_per_chunk": 4.914, "boundary_entropy": 0.10023, "router_noise": 0.0, "grad_norm": 0.6929, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.157, "alpha": 0.03, "t": 826.9}
{"event": "train", "step": 4750, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.32692, "bpb": 1.91434, "chunk_loss": 0.030436, "F": 0.20566, "G": 0.20894, "chunks": 210.59, "bytes_per_chunk": 4.862, "boundary_entropy": 0.09994, "router_noise": 0.0, "grad_norm": 0.6764, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.164, "alpha": 0.03, "t": 835.0}
{"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3267.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187726, "cps": 25.523438, "mean_p": 0.192717, "boundary_entropy": 0.095911, "bpb": 1.897449}, "tr": {"chunks": 3844.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.21234, "cps": 30.03125, "mean_p": 0.197705, "boundary_entropy": 0.129747, "bpb": 2.044327}, "fi": {"chunks": 4798.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.255064, "cps": 37.484375, "mean_p": 0.250472, "boundary_entropy": 0.10215, "bpb": 2.14106}}, "bpb": {"en": 1.89745, "tr": 2.04433, "fi": 2.14106}, "gini_cpc": 0.068523, "gini_cps": 0.085705, "bpic": 4.763792, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.131117738832181, "fi": 1.3586992820516728}, "t": 843.6}
{"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 1.91095, "tr": 2.05265, "fi": 2.15108}, "bytes_per_chunk": {"en": 5.014, "tr": 5.183, "fi": 4.188}, "bpb_mean": 2.03823, "frozen_inventory": false, "t": 844.9}
{"event": "train", "step": 4800, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.39828, "bpb": 2.0173, "chunk_loss": 0.030648, "F": 0.21359, "G": 0.2178, "chunks": 218.72, "bytes_per_chunk": 4.682, "boundary_entropy": 0.10306, "router_noise": 0.0, "grad_norm": 0.7313, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.209, "alpha": 0.03, "t": 845.5}
{"event": "train", "step": 4850, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.34933, "bpb": 1.94668, "chunk_loss": 0.030538, "F": 0.20822, "G": 0.21212, "chunks": 213.22, "bytes_per_chunk": 4.803, "boundary_entropy": 0.10099, "router_noise": 0.0, "grad_norm": 0.6996, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.155, "alpha": 0.03, "t": 853.2}
{"event": "train", "step": 4900, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.35992, "bpb": 1.96195, "chunk_loss": 0.030314, "F": 0.19751, "G": 0.20155, "chunks": 202.25, "bytes_per_chunk": 5.063, "boundary_entropy": 0.09683, "router_noise": 0.0, "grad_norm": 0.7222, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.151, "alpha": 0.03, "t": 860.8}
{"event": "train", "step": 4950, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.34036, "bpb": 1.93373, "chunk_loss": 0.030481, "F": 0.20612, "G": 0.20941, "chunks": 211.06, "bytes_per_chunk": 4.852, "boundary_entropy": 0.0937, "router_noise": 0.0, "grad_norm": 0.6978, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 868.9}
{"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3266.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187669, "cps": 25.515625, "mean_p": 0.192216, "boundary_entropy": 0.093825, "bpb": 1.887198}, "tr": {"chunks": 3854.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212893, "cps": 30.109375, "mean_p": 0.198749, "boundary_entropy": 0.128123, "bpb": 2.033389}, "fi": {"chunks": 4779.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254053, "cps": 37.335938, "mean_p": 0.249448, "boundary_entropy": 0.100295, "bpb": 2.128475}}, "bpb": {"en": 1.8872, "tr": 2.03339, "fi": 2.12847}, "gini_cpc": 0.067607, "gini_cps": 0.084769, "bpic": 4.767796, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1344075249611845, "fi": 1.3537332213959448}, "t": 877.6}
{"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 1.90172, "tr": 2.04114, "fi": 2.1382}, "bytes_per_chunk": {"en": 5.017, "tr": 5.167, "fi": 4.189}, "bpb_mean": 2.02702, "frozen_inventory": false, "t": 878.9}
{"event": "train", "step": 5000, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.34205, "bpb": 1.93618, "chunk_loss": 0.030486, "F": 0.20731, "G": 0.2095, "chunks": 212.28, "bytes_per_chunk": 4.824, "boundary_entropy": 0.09744, "router_noise": 0.0, "grad_norm": 0.7063, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.216, "alpha": 0.03, "t": 879.7}
{"event": "train", "step": 5050, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.35701, "bpb": 1.95775, "chunk_loss": 0.030403, "F": 0.2049, "G": 0.20777, "chunks": 209.81, "bytes_per_chunk": 4.881, "boundary_entropy": 0.09752, "router_noise": 0.0, "grad_norm": 0.682, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.16, "alpha": 0.03, "t": 887.7}
{"event": "train", "step": 5100, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.32262, "bpb": 1.90814, "chunk_loss": 0.030244, "F": 0.19562, "G": 0.19833, "chunks": 200.31, "bytes_per_chunk": 5.112, "boundary_entropy": 0.09387, "router_noise": 0.0, "grad_norm": 0.6954, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.155, "alpha": 0.03, "t": 895.5}
{"event": "train", "step": 5150, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.31429, "bpb": 1.89613, "chunk_loss": 0.030473, "F": 0.20737, "G": 0.20971, "chunks": 212.34, "bytes_per_chunk": 4.822, "boundary_entropy": 0.09329, "router_noise": 0.0, "grad_norm": 0.6982, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.154, "alpha": 0.03, "t": 903.2}
{"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3240.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186175, "cps": 25.3125, "mean_p": 0.192126, "boundary_entropy": 0.091335, "bpb": 1.880751}, "tr": {"chunks": 3851.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212727, "cps": 30.085938, "mean_p": 0.198907, "boundary_entropy": 0.12729, "bpb": 2.021778}, "fi": {"chunks": 4787.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254479, "cps": 37.398438, "mean_p": 0.249094, "boundary_entropy": 0.097247, "bpb": 2.119847}}, "bpb": {"en": 1.88075, "tr": 2.02178, "fi": 2.11985}, "gini_cpc": 0.069693, "gini_cps": 0.086827, "bpic": 4.776225, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1426206726529877, "fi": 1.3668808341061278}, "t": 912.0}
{"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 1.89602, "tr": 2.03011, "fi": 2.12986}, "bytes_per_chunk": {"en": 5.023, "tr": 5.157, "fi": 4.178}, "bpb_mean": 2.01866, "frozen_inventory": false, "t": 913.2}
{"event": "train", "step": 5200, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.35889, "bpb": 1.96047, "chunk_loss": 0.030601, "F": 0.21112, "G": 0.21473, "chunks": 216.19, "bytes_per_chunk": 4.737, "boundary_entropy": 0.09746, "router_noise": 0.0, "grad_norm": 0.7019, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.214, "alpha": 0.03, "t": 913.8}
{"event": "train", "step": 5250, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.3843, "bpb": 1.99713, "chunk_loss": 0.030501, "F": 0.20984, "G": 0.21248, "chunks": 214.88, "bytes_per_chunk": 4.766, "boundary_entropy": 0.09673, "router_noise": 0.0, "grad_norm": 0.7394, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.163, "alpha": 0.03, "t": 922.0}
{"event": "train", "step": 5300, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.3753, "bpb": 1.98415, "chunk_loss": 0.030388, "F": 0.20358, "G": 0.20664, "chunks": 208.47, "bytes_per_chunk": 4.912, "boundary_entropy": 0.0964, "router_noise": 0.0, "grad_norm": 0.6487, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.164, "alpha": 0.03, "t": 930.2}
{"event": "train", "step": 5350, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.33202, "bpb": 1.9217, "chunk_loss": 0.030626, "F": 0.20847, "G": 0.21052, "chunks": 213.47, "bytes_per_chunk": 4.797, "boundary_entropy": 0.09901, "router_noise": 0.0, "grad_norm": 0.7563, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.152, "alpha": 0.03, "t": 937.8}
{"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3256.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187094, "cps": 25.4375, "mean_p": 0.191377, "boundary_entropy": 0.092459, "bpb": 1.875561}, "tr": {"chunks": 3851.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212727, "cps": 30.085938, "mean_p": 0.197848, "boundary_entropy": 0.127996, "bpb": 2.015354}, "fi": {"chunks": 4785.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254372, "cps": 37.382812, "mean_p": 0.249324, "boundary_entropy": 0.101082, "bpb": 2.111785}}, "bpb": {"en": 1.87556, "tr": 2.01535, "fi": 2.11179}, "gini_cpc": 0.068561, "gini_cps": 0.085716, "bpic": 4.770602, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1370058290527272, "fi": 1.359595700905307}, "t": 946.2}
{"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 1.89054, "tr": 2.02349, "fi": 2.12117}, "bytes_per_chunk": {"en": 5.014, "tr": 5.171, "fi": 4.193}, "bpb_mean": 2.01173, "frozen_inventory": false, "t": 947.6}
{"event": "train", "step": 5400, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.2607, "bpb": 1.81881, "chunk_loss": 0.030376, "F": 0.19974, "G": 0.20107, "chunks": 204.53, "bytes_per_chunk": 5.007, "boundary_entropy": 0.0952, "router_noise": 0.0, "grad_norm": 0.6584, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.206, "alpha": 0.03, "t": 948.1}
{"event": "train", "step": 5450, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.30885, "bpb": 1.88827, "chunk_loss": 0.030457, "F": 0.20685, "G": 0.20953, "chunks": 211.81, "bytes_per_chunk": 4.834, "boundary_entropy": 0.09269, "router_noise": 0.0, "grad_norm": 0.6675, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.161, "alpha": 0.03, "t": 956.2}
{"event": "train", "step": 5500, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.32224, "bpb": 1.90759, "chunk_loss": 0.030407, "F": 0.20456, "G": 0.20677, "chunks": 209.47, "bytes_per_chunk": 4.889, "boundary_entropy": 0.09404, "router_noise": 0.0, "grad_norm": 0.6957, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.164, "alpha": 0.03, "t": 964.4}
{"event": "train", "step": 5550, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.30828, "bpb": 1.88745, "chunk_loss": 0.030461, "F": 0.20813, "G": 0.21027, "chunks": 213.12, "bytes_per_chunk": 4.805, "boundary_entropy": 0.09078, "router_noise": 0.0, "grad_norm": 0.6837, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.162, "alpha": 0.03, "t": 972.5}
{"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3244.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186405, "cps": 25.34375, "mean_p": 0.19092, "boundary_entropy": 0.091594, "bpb": 1.870454}, "tr": {"chunks": 3855.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212948, "cps": 30.117188, "mean_p": 0.198051, "boundary_entropy": 0.124363, "bpb": 2.007663}, "fi": {"chunks": 4773.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.253735, "cps": 37.289062, "mean_p": 0.248567, "boundary_entropy": 0.098814, "bpb": 2.102823}}, "bpb": {"en": 1.87045, "tr": 2.00766, "fi": 2.10282}, "gini_cpc": 0.06873, "gini_cps": 0.08586, "bpic": 4.778639, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1423971359121692, "fi": 1.3612027743559283}, "t": 980.3}
{"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 1.88582, "tr": 2.01587, "fi": 2.11349}, "bytes_per_chunk": {"en": 5.025, "tr": 5.154, "fi": 4.199}, "bpb_mean": 2.00506, "frozen_inventory": false, "t": 981.5}
{"event": "train", "step": 5600, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.33302, "bpb": 1.92315, "chunk_loss": 0.030474, "F": 0.20825, "G": 0.21064, "chunks": 213.25, "bytes_per_chunk": 4.802, "boundary_entropy": 0.09927, "router_noise": 0.0, "grad_norm": 0.7036, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.192, "alpha": 0.03, "t": 982.1}
{"event": "train", "step": 5650, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.32279, "bpb": 1.90839, "chunk_loss": 0.030391, "F": 0.20233, "G": 0.20606, "chunks": 207.19, "bytes_per_chunk": 4.942, "boundary_entropy": 0.09219, "router_noise": 0.0, "grad_norm": 0.6747, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.11, "alpha": 0.03, "t": 987.6}
{"event": "train", "step": 5700, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.39584, "bpb": 2.01378, "chunk_loss": 0.030507, "F": 0.21021, "G": 0.21263, "chunks": 215.25, "bytes_per_chunk": 4.757, "boundary_entropy": 0.09609, "router_noise": 0.0, "grad_norm": 0.7069, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.124, "alpha": 0.03, "t": 993.8}
{"event": "train", "step": 5750, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.38133, "bpb": 1.99283, "chunk_loss": 0.030532, "F": 0.20795, "G": 0.21037, "chunks": 212.94, "bytes_per_chunk": 4.809, "boundary_entropy": 0.09332, "router_noise": 0.0, "grad_norm": 0.7251, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.125, "alpha": 0.03, "t": 1000.0}
{"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3257.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187152, "cps": 25.445312, "mean_p": 0.191703, "boundary_entropy": 0.089386, "bpb": 1.866831}, "tr": {"chunks": 3857.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213059, "cps": 30.132812, "mean_p": 0.197984, "boundary_entropy": 0.125211, "bpb": 2.002969}, "fi": {"chunks": 4774.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.253788, "cps": 37.296875, "mean_p": 0.248756, "boundary_entropy": 0.097505, "bpb": 2.096553}}, "bpb": {"en": 1.86683, "tr": 2.00297, "fi": 2.09655}, "gini_cpc": 0.067927, "gini_cps": 0.085072, "bpic": 4.772207, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1384276861070852, "fi": 1.356053715133165}, "t": 1006.8}
{"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 1.88208, "tr": 2.01097, "fi": 2.10643}, "bytes_per_chunk": {"en": 5.009, "tr": 5.155, "fi": 4.19}, "bpb_mean": 1.99983, "frozen_inventory": false, "t": 1007.7}
{"event": "train", "step": 5800, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.29889, "bpb": 1.8739, "chunk_loss": 0.030377, "F": 0.20312, "G": 0.20492, "chunks": 208.0, "bytes_per_chunk": 4.923, "boundary_entropy": 0.09108, "router_noise": 0.0, "grad_norm": 0.7237, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.164, "alpha": 0.03, "t": 1008.2}
{"event": "train", "step": 5850, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.35062, "bpb": 1.94853, "chunk_loss": 0.030315, "F": 0.19757, "G": 0.20085, "chunks": 202.31, "bytes_per_chunk": 5.061, "boundary_entropy": 0.09337, "router_noise": 0.0, "grad_norm": 0.7402, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.12, "alpha": 0.03, "t": 1014.3}
{"event": "train", "step": 5900, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.38107, "bpb": 1.99247, "chunk_loss": 0.030506, "F": 0.20996, "G": 0.21228, "chunks": 215.0, "bytes_per_chunk": 4.763, "boundary_entropy": 0.09391, "router_noise": 0.0, "grad_norm": 0.7474, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.125, "alpha": 0.03, "t": 1020.5}
{"event": "train", "step": 5950, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.31817, "bpb": 1.90172, "chunk_loss": 0.030375, "F": 0.20129, "G": 0.203, "chunks": 206.12, "bytes_per_chunk": 4.968, "boundary_entropy": 0.0953, "router_noise": 0.0, "grad_norm": 0.7925, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.121, "alpha": 0.03, "t": 1026.6}
{"event": "train", "step": 5999, "arch": "pkm", "phase": 3, "objective": "global", "lm_loss": 1.33626, "bpb": 1.92782, "chunk_loss": 0.030576, "F": 0.2121, "G": 0.21447, "chunks": 217.19, "bytes_per_chunk": 4.715, "boundary_entropy": 0.09505, "router_noise": 0.0, "grad_norm": 0.7639, "lr": 3e-05, "router_trainable": true, "sec_per_step": 0.122, "alpha": 0.03, "t": 1032.7}
{"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3249.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186692, "cps": 25.382812, "mean_p": 0.191187, "boundary_entropy": 0.088328, "bpb": 1.863309}, "tr": {"chunks": 3860.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213224, "cps": 30.15625, "mean_p": 0.197614, "boundary_entropy": 0.123394, "bpb": 1.997068}, "fi": {"chunks": 4780.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254107, "cps": 37.34375, "mean_p": 0.248387, "boundary_entropy": 0.09704, "bpb": 2.091726}}, "bpb": {"en": 1.86331, "tr": 1.99707, "fi": 2.09173}, "gini_cpc": 0.068718, "gini_cps": 0.08585, "bpic": 4.771806, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.1421184890053322, "fi": 1.3611012161456582}, "t": 1033.3}
{"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 1.87819, "tr": 2.00545, "fi": 2.10171}, "bytes_per_chunk": {"en": 5.019, "tr": 5.159, "fi": 4.193}, "bpb_mean": 1.99512, "frozen_inventory": false, "t": 1034.4}
{"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s0/probe/reps/th0.30", "/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s0/probe/reps/th0.40", "/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s0/probe/reps/th0.50", "/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s0/probe/reps/th0.60", "/local/scratch/sas245/beetle-hnet/runs/ali_pkm_s0/probe/reps/th0.70"], "t": 1035.9}
{"event": "final", "run": "ali_pkm_s0", "arch": "pkm", "objective": "ratio", "condition": "A", "step": 6000, "params": 39425792, "seed": 0, "per_lang": {"en": {"cpc": 0.18669194966385105, "cps": 25.3828125, "bpb": 1.8851913960726823, "bytes_per_chunk": 4.955744182089721}, "tr": {"cpc": 0.2132243274595371, "cps": 30.15625, "bpb": 1.9970845663471601, "bytes_per_chunk": 5.164025333162607}, "fi": {"cpc": 0.25410663973207165, "cps": 37.34375, "bpb": 2.120231539618918, "bytes_per_chunk": 4.178326081066004}}, "bpb_mean": 2.00083583401292, "gini_cpc": 0.06871796928498064, "gini_cps": 0.08584966495640245, "bpic": 4.771805870973169, "mean_boundary_entropy": 0.09505325555801392, "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": 1037.8729794025421, "t": 1037.9}