{"event": "start", "run": "ali_hnet-ffn_s2", "arch": "hnet_ffn", "objective": "ratio", "condition": "A", "params": 22785033, "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", "ffn_params": 268041, "matched_to_ali_params": 268096, "t": 1.8}
{"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": 1.8}
{"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10933.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.628225, "cps": 85.414062, "mean_p": 0.513249, "boundary_entropy": 0.686295, "bpb": 8.12598}, "tr": {"chunks": 8187.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.452245, "cps": 63.960938, "mean_p": 0.498002, "boundary_entropy": 0.686826, "bpb": 8.097536}, "fi": {"chunks": 13540.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.719792, "cps": 105.78125, "mean_p": 0.517169, "boundary_entropy": 0.686661, "bpb": 8.065117}}, "bpb": {"en": 8.12598, "tr": 8.09754, "fi": 8.06512}, "gini_cpc": 0.099077, "gini_cps": 0.109267, "bpic": 1.737048, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.7198781817505053, "fi": 1.1457544508645294}, "t": 3.5}
{"event": "train", "step": 0, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 5.62735, "bpb": 8.11854, "chunk_loss": 0.062086, "F": 0.59833, "G": 0.50836, "chunks": 612.69, "bytes_per_chunk": 1.671, "boundary_entropy": 0.69105, "router_noise": 0.0, "grad_norm": 11.5524, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.009, "alpha": 0.03, "t": 4.0}
{"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10907.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.626731, "cps": 85.210938, "mean_p": 0.513143, "boundary_entropy": 0.686304, "bpb": 8.093969}, "tr": {"chunks": 8206.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.453295, "cps": 64.109375, "mean_p": 0.498082, "boundary_entropy": 0.686828, "bpb": 8.075965}, "fi": {"chunks": 13504.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.717878, "cps": 105.5, "mean_p": 0.517038, "boundary_entropy": 0.686672, "bpb": 8.035786}}, "bpb": {"en": 8.09397, "tr": 8.07596, "fi": 8.03579}, "gini_cpc": 0.098108, "gini_cps": 0.108287, "bpic": 1.739338, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.7232688618519676, "fi": 1.1454321082690706}, "t": 4.6}
{"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10880.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.62518, "cps": 85.0, "mean_p": 0.512978, "boundary_entropy": 0.686318, "bpb": 8.030086}, "tr": {"chunks": 8244.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.455394, "cps": 64.40625, "mean_p": 0.498245, "boundary_entropy": 0.686832, "bpb": 8.031196}, "fi": {"chunks": 13466.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.715858, "cps": 105.203125, "mean_p": 0.516799, "boundary_entropy": 0.686692, "bpb": 7.975283}}, "bpb": {"en": 8.03009, "tr": 8.0312, "fi": 7.97528}, "gini_cpc": 0.09666, "gini_cps": 0.106822, "bpic": 1.740779, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.7284213333181695, "fi": 1.1450434097539925}, "t": 5.8}
{"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10715.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.615698, "cps": 83.710938, "mean_p": 0.512099, "boundary_entropy": 0.68638, "bpb": 7.816331}, "tr": {"chunks": 8386.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.463238, "cps": 65.515625, "mean_p": 0.498669, "boundary_entropy": 0.686846, "bpb": 7.876218}, "fi": {"chunks": 13215.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.702514, "cps": 103.242188, "mean_p": 0.515683, "boundary_entropy": 0.686774, "bpb": 7.770535}}, "bpb": {"en": 7.81633, "tr": 7.87622, "fi": 7.77054}, "gini_cpc": 0.089544, "gini_cps": 0.09962, "bpic": 1.755539, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.7523782831425373, "fi": 1.1410041627086822}, "t": 6.9}
{"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 9629.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.553295, "cps": 75.226562, "mean_p": 0.507272, "boundary_entropy": 0.686615, "bpb": 7.288799}, "tr": {"chunks": 8636.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.477048, "cps": 67.46875, "mean_p": 0.49959, "boundary_entropy": 0.686903, "bpb": 7.474701}, "fi": {"chunks": 11869.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.630961, "cps": 92.726562, "mean_p": 0.510276, "boundary_entropy": 0.687082, "bpb": 7.292564}}, "bpb": {"en": 7.2888, "tr": 7.4747, "fi": 7.29256}, "gini_cpc": 0.061764, "gini_cps": 0.071525, "bpic": 1.882657, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8621940386161193, "fi": 1.1403684145516706}, "t": 8.4}
{"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3750.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.21548, "cps": 29.296875, "mean_p": 0.484373, "boundary_entropy": 0.686159, "bpb": 6.721976}, "tr": {"chunks": 7257.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.400873, "cps": 56.695312, "mean_p": 0.49433, "boundary_entropy": 0.686995, "bpb": 7.01154}, "fi": {"chunks": 4578.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.243368, "cps": 35.765625, "mean_p": 0.48651, "boundary_entropy": 0.686901, "bpb": 6.78458}}, "bpb": {"en": 6.72198, "tr": 7.01154, "fi": 6.78458}, "gini_cpc": 0.143762, "gini_cps": 0.150016, "bpic": 3.640167, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.8603704137435784, "fi": 1.1294233374089628}, "t": 10.0}
{"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2608.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.149859, "cps": 20.375, "mean_p": 0.48676, "boundary_entropy": 0.686947, "bpb": 6.304009}, "tr": {"chunks": 4640.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.256311, "cps": 36.25, "mean_p": 0.490153, "boundary_entropy": 0.687556, "bpb": 6.652033}, "fi": {"chunks": 1387.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.073733, "cps": 10.835938, "mean_p": 0.479298, "boundary_entropy": 0.686961, "bpb": 6.420794}}, "bpb": {"en": 6.30401, "tr": 6.65203, "fi": 6.42079}, "gini_cpc": 0.253631, "gini_cps": 0.251148, "bpic": 6.570006, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.7103459447625702, "fi": 0.4920181353554717}, "t": 12.8}
{"event": "train", "step": 50, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 4.21163, "bpb": 6.0761, "chunk_loss": 0.033993, "F": 0.22281, "G": 0.48818, "chunks": 228.16, "bytes_per_chunk": 4.488, "boundary_entropy": 0.69158, "router_noise": 0.0, "grad_norm": 2.4782, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.22, "alpha": 0.03, "t": 15.0}
{"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4876.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.280182, "cps": 38.09375, "mean_p": 0.485334, "boundary_entropy": 0.68548, "bpb": 5.547737}, "tr": {"chunks": 5407.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.29868, "cps": 42.242188, "mean_p": 0.486468, "boundary_entropy": 0.686459, "bpb": 5.954005}, "fi": {"chunks": 4779.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.254053, "cps": 37.335938, "mean_p": 0.48406, "boundary_entropy": 0.686307, "bpb": 5.716127}}, "bpb": {"en": 5.54774, "tr": 5.95401, "fi": 5.71613}, "gini_cpc": 0.035719, "gini_cps": 0.027796, "bpic": 3.766565, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.0660221813406896, "fi": 0.9067458369727556}, "t": 16.7}
{"event": "train", "step": 100, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 3.25148, "bpb": 4.6909, "chunk_loss": 0.029919, "F": 0.16522, "G": 0.45347, "chunks": 169.19, "bytes_per_chunk": 6.052, "boundary_entropy": 0.68219, "router_noise": 0.0, "grad_norm": 1.9765, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.113, "alpha": 0.03, "t": 20.6}
{"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4280.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.245935, "cps": 33.4375, "mean_p": 0.424539, "boundary_entropy": 0.648271, "bpb": 4.16585}, "tr": {"chunks": 4788.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.264487, "cps": 37.40625, "mean_p": 0.442392, "boundary_entropy": 0.662168, "bpb": 4.408264}, "fi": {"chunks": 3939.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.209399, "cps": 30.773438, "mean_p": 0.435589, "boundary_entropy": 0.663163, "bpb": 4.357254}}, "bpb": {"en": 4.16585, "tr": 4.40826, "fi": 4.35725}, "gini_cpc": 0.05102, "gini_cps": 0.043515, "bpic": 4.361651, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.0754344428893647, "fi": 0.8514407830574375}, "t": 23.7}
{"event": "train", "step": 150, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 2.79107, "bpb": 4.02666, "chunk_loss": 0.031521, "F": 0.19849, "G": 0.3822, "chunks": 203.25, "bytes_per_chunk": 5.038, "boundary_entropy": 0.61806, "router_noise": 0.0, "grad_norm": 1.3854, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.108, "alpha": 0.03, "t": 26.0}
{"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3610.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.207435, "cps": 28.203125, "mean_p": 0.375933, "boundary_entropy": 0.580348, "bpb": 3.631521}, "tr": {"chunks": 4022.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.222173, "cps": 31.421875, "mean_p": 0.358701, "boundary_entropy": 0.579311, "bpb": 3.812162}, "fi": {"chunks": 3226.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171495, "cps": 25.203125, "mean_p": 0.374806, "boundary_entropy": 0.614648, "bpb": 3.855995}}, "bpb": {"en": 3.63152, "tr": 3.81216, "fi": 3.856}, "gini_cpc": 0.056205, "gini_cps": 0.048873, "bpic": 5.224903, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.0710467633423577, "fi": 0.8267408516647079}, "t": 30.6}
{"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.63854, "tr": 3.82212, "fi": 3.854}, "bytes_per_chunk": {"en": 5.269, "tr": 5.588, "fi": 7.149}, "bpb_mean": 3.77155, "frozen_inventory": false, "t": 30.9}
{"event": "train", "step": 200, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 2.52819, "bpb": 3.64741, "chunk_loss": 0.030754, "F": 0.18307, "G": 0.36237, "chunks": 187.47, "bytes_per_chunk": 5.462, "boundary_entropy": 0.57903, "router_noise": 0.0, "grad_norm": 1.2142, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.104, "alpha": 0.03, "t": 31.2}
{"event": "train", "step": 250, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 2.38266, "bpb": 3.43745, "chunk_loss": 0.031106, "F": 0.19662, "G": 0.33631, "chunks": 201.34, "bytes_per_chunk": 5.086, "boundary_entropy": 0.53017, "router_noise": 0.0, "grad_norm": 1.3711, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 35.6}
{"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3467.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199219, "cps": 27.085938, "mean_p": 0.332033, "boundary_entropy": 0.516916, "bpb": 3.355315}, "tr": {"chunks": 4785.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.264321, "cps": 37.382812, "mean_p": 0.321431, "boundary_entropy": 0.507522, "bpb": 3.516065}, "fi": {"chunks": 4628.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.246026, "cps": 36.15625, "mean_p": 0.361115, "boundary_entropy": 0.573658, "bpb": 3.646463}}, "bpb": {"en": 3.35531, "tr": 3.51607, "fi": 3.64646}, "gini_cpc": 0.061166, "gini_cps": 0.068219, "bpic": 4.404658, "worst_lang": "tr", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.3267884102794731, "fi": 1.2349567419059633}, "t": 36.4}
{"event": "train", "step": 300, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 2.27393, "bpb": 3.28058, "chunk_loss": 0.03087, "F": 0.19626, "G": 0.29936, "chunks": 200.97, "bytes_per_chunk": 5.095, "boundary_entropy": 0.46184, "router_noise": 0.0, "grad_norm": 1.0752, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.102, "alpha": 0.03, "t": 40.7}
{"event": "train", "step": 350, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 2.17133, "bpb": 3.13256, "chunk_loss": 0.031112, "F": 0.20627, "G": 0.28824, "chunks": 211.22, "bytes_per_chunk": 4.848, "boundary_entropy": 0.42092, "router_noise": 0.0, "grad_norm": 1.3701, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.111, "alpha": 0.03, "t": 46.3}
{"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3044.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.174912, "cps": 23.78125, "mean_p": 0.259474, "boundary_entropy": 0.376096, "bpb": 2.907867}, "tr": {"chunks": 3440.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190024, "cps": 26.875, "mean_p": 0.269641, "boundary_entropy": 0.424887, "bpb": 3.184426}, "fi": {"chunks": 3789.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.201425, "cps": 29.601562, "mean_p": 0.279551, "boundary_entropy": 0.444234, "bpb": 3.371965}}, "bpb": {"en": 2.90787, "tr": 3.18443, "fi": 3.37197}, "gini_cpc": 0.031208, "gini_cps": 0.048347, "bpic": 5.522437, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 1.0863940121293085, "fi": 1.1515749095836858}, "t": 51.9}
{"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 2.92006, "tr": 3.19252, "fi": 3.36498}, "bytes_per_chunk": {"en": 5.308, "tr": 5.519, "fi": 5.346}, "bpb_mean": 3.15919, "frozen_inventory": false, "t": 52.7}
{"event": "train", "step": 400, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 2.08572, "bpb": 3.00906, "chunk_loss": 0.030436, "F": 0.18481, "G": 0.26727, "chunks": 189.25, "bytes_per_chunk": 5.411, "boundary_entropy": 0.38254, "router_noise": 0.0, "grad_norm": 1.304, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.136, "alpha": 0.03, "t": 53.1}
{"event": "train", "step": 450, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.99881, "bpb": 2.88367, "chunk_loss": 0.030711, "F": 0.19696, "G": 0.26856, "chunks": 201.69, "bytes_per_chunk": 5.077, "boundary_entropy": 0.35663, "router_noise": 0.0, "grad_norm": 1.4123, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.117, "alpha": 0.03, "t": 58.9}
{"event": "train", "step": 500, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.90919, "bpb": 2.75437, "chunk_loss": 0.031209, "F": 0.2168, "G": 0.27002, "chunks": 222.0, "bytes_per_chunk": 4.613, "boundary_entropy": 0.32324, "router_noise": 0.0, "grad_norm": 1.6604, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.118, "alpha": 0.03, "t": 64.8}
{"event": "train", "step": 550, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.90424, "bpb": 2.74724, "chunk_loss": 0.030727, "F": 0.19833, "G": 0.26516, "chunks": 203.09, "bytes_per_chunk": 5.042, "boundary_entropy": 0.32826, "router_noise": 0.0, "grad_norm": 1.586, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.105, "alpha": 0.03, "t": 70.0}
{"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3397.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195196, "cps": 26.539062, "mean_p": 0.252593, "boundary_entropy": 0.303275, "bpb": 2.589129}, "tr": {"chunks": 3079.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170082, "cps": 24.054688, "mean_p": 0.260232, "boundary_entropy": 0.36212, "bpb": 2.926289}, "fi": {"chunks": 4071.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.216416, "cps": 31.804688, "mean_p": 0.273037, "boundary_entropy": 0.358101, "bpb": 3.142158}}, "bpb": {"en": 2.58913, "tr": 2.92629, "fi": 3.14216}, "gini_cpc": 0.053102, "gini_cps": 0.062703, "bpic": 5.37897, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8713401192934989, "fi": 1.108709559903843}, "t": 76.0}
{"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.60268, "tr": 2.93561, "fi": 3.14246}, "bytes_per_chunk": {"en": 4.915, "tr": 6.386, "fi": 4.843}, "bpb_mean": 2.89358, "frozen_inventory": false, "t": 76.9}
{"event": "train", "step": 600, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.89153, "bpb": 2.7289, "chunk_loss": 0.030755, "F": 0.20166, "G": 0.2597, "chunks": 206.5, "bytes_per_chunk": 4.959, "boundary_entropy": 0.30736, "router_noise": 0.0, "grad_norm": 1.14, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.144, "alpha": 0.03, "t": 77.2}
{"event": "train", "step": 650, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.8441, "bpb": 2.66047, "chunk_loss": 0.030758, "F": 0.20038, "G": 0.26332, "chunks": 205.19, "bytes_per_chunk": 4.991, "boundary_entropy": 0.30679, "router_noise": 0.0, "grad_norm": 1.1516, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.112, "alpha": 0.03, "t": 82.8}
{"event": "train", "step": 700, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.98755, "bpb": 2.86743, "chunk_loss": 0.031047, "F": 0.20511, "G": 0.27009, "chunks": 210.03, "bytes_per_chunk": 4.875, "boundary_entropy": 0.30858, "router_noise": 0.0, "grad_norm": 1.068, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.116, "alpha": 0.03, "t": 88.6}
{"event": "train", "step": 750, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.78155, "bpb": 2.57024, "chunk_loss": 0.03103, "F": 0.21405, "G": 0.26331, "chunks": 219.19, "bytes_per_chunk": 4.672, "boundary_entropy": 0.2815, "router_noise": 0.0, "grad_norm": 1.2143, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.114, "alpha": 0.03, "t": 94.3}
{"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3522.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.202379, "cps": 27.515625, "mean_p": 0.247382, "boundary_entropy": 0.263907, "bpb": 2.435469}, "tr": {"chunks": 3406.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.188146, "cps": 26.609375, "mean_p": 0.258574, "boundary_entropy": 0.334289, "bpb": 2.773219}, "fi": {"chunks": 4031.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.21429, "cps": 31.492188, "mean_p": 0.266026, "boundary_entropy": 0.315534, "bpb": 2.98871}}, "bpb": {"en": 2.43547, "tr": 2.77322, "fi": 2.98871}, "gini_cpc": 0.028818, "gini_cps": 0.038021, "bpic": 5.17675, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9296700943051501, "fi": 1.058853028923868}, "t": 100.0}
{"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.44732, "tr": 2.7819, "fi": 2.99008}, "bytes_per_chunk": {"en": 4.701, "tr": 5.729, "fi": 4.882}, "bpb_mean": 2.73976, "frozen_inventory": false, "t": 100.6}
{"event": "train", "step": 800, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.86196, "bpb": 2.68625, "chunk_loss": 0.031044, "F": 0.21274, "G": 0.26254, "chunks": 217.84, "bytes_per_chunk": 4.701, "boundary_entropy": 0.27791, "router_noise": 0.0, "grad_norm": 1.1702, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.134, "alpha": 0.03, "t": 101.0}
{"event": "train", "step": 850, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.80701, "bpb": 2.60697, "chunk_loss": 0.030842, "F": 0.20737, "G": 0.25904, "chunks": 212.34, "bytes_per_chunk": 4.822, "boundary_entropy": 0.27421, "router_noise": 0.0, "grad_norm": 1.0643, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 105.3}
{"event": "train", "step": 900, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.60479, "bpb": 2.31523, "chunk_loss": 0.030765, "F": 0.20486, "G": 0.25257, "chunks": 209.78, "bytes_per_chunk": 4.881, "boundary_entropy": 0.25277, "router_noise": 0.0, "grad_norm": 1.1386, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.094, "alpha": 0.03, "t": 110.0}
{"event": "train", "step": 950, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.68808, "bpb": 2.43538, "chunk_loss": 0.030821, "F": 0.2085, "G": 0.24901, "chunks": 213.5, "bytes_per_chunk": 4.796, "boundary_entropy": 0.2469, "router_noise": 0.0, "grad_norm": 1.1258, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.076, "alpha": 0.03, "t": 113.8}
{"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3495.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200827, "cps": 27.304688, "mean_p": 0.239266, "boundary_entropy": 0.243448, "bpb": 2.336907}, "tr": {"chunks": 3472.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.191791, "cps": 27.125, "mean_p": 0.246548, "boundary_entropy": 0.307312, "bpb": 2.653829}, "fi": {"chunks": 3979.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.211525, "cps": 31.085938, "mean_p": 0.260092, "boundary_entropy": 0.291464, "bpb": 2.860235}}, "bpb": {"en": 2.33691, "tr": 2.65383, "fi": 2.86024}, "gini_cpc": 0.021776, "gini_cps": 0.030879, "bpic": 5.182898, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9550060111441467, "fi": 1.0532682571128253}, "t": 118.5}
{"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.3462, "tr": 2.66069, "fi": 2.86596}, "bytes_per_chunk": {"en": 4.815, "tr": 5.68, "fi": 4.945}, "bpb_mean": 2.62428, "frozen_inventory": false, "t": 119.2}
{"event": "train", "step": 1000, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.73743, "bpb": 2.50658, "chunk_loss": 0.030729, "F": 0.2056, "G": 0.24949, "chunks": 210.53, "bytes_per_chunk": 4.864, "boundary_entropy": 0.25605, "router_noise": 0.0, "grad_norm": 0.9806, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.121, "alpha": 0.03, "t": 119.9}
{"event": "train", "step": 1050, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.63366, "bpb": 2.35687, "chunk_loss": 0.031001, "F": 0.21396, "G": 0.24893, "chunks": 219.09, "bytes_per_chunk": 4.674, "boundary_entropy": 0.24135, "router_noise": 0.0, "grad_norm": 1.0058, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 124.4}
{"event": "train", "step": 1100, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.74199, "bpb": 2.51315, "chunk_loss": 0.030778, "F": 0.20663, "G": 0.24844, "chunks": 211.59, "bytes_per_chunk": 4.839, "boundary_entropy": 0.24746, "router_noise": 0.0, "grad_norm": 0.988, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 128.7}
{"event": "train", "step": 1150, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.64488, "bpb": 2.37306, "chunk_loss": 0.030697, "F": 0.20505, "G": 0.24547, "chunks": 209.97, "bytes_per_chunk": 4.877, "boundary_entropy": 0.23633, "router_noise": 0.0, "grad_norm": 0.8383, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 133.0}
{"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3531.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.202896, "cps": 27.585938, "mean_p": 0.235576, "boundary_entropy": 0.221401, "bpb": 2.266667}, "tr": {"chunks": 3357.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185439, "cps": 26.226562, "mean_p": 0.237658, "boundary_entropy": 0.286545, "bpb": 2.564464}, "fi": {"chunks": 4201.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.223327, "cps": 32.820312, "mean_p": 0.259745, "boundary_entropy": 0.2624, "bpb": 2.747601}}, "bpb": {"en": 2.26667, "tr": 2.56446, "fi": 2.7476}, "gini_cpc": 0.041295, "gini_cps": 0.050741, "bpic": 5.116061, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9139600072857895, "fi": 1.100695524821496}, "t": 137.6}
{"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.27649, "tr": 2.57285, "fi": 2.75415}, "bytes_per_chunk": {"en": 4.709, "tr": 5.825, "fi": 4.671}, "bpb_mean": 2.5345, "frozen_inventory": false, "t": 138.3}
{"event": "train", "step": 1200, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.62102, "bpb": 2.33864, "chunk_loss": 0.030705, "F": 0.20627, "G": 0.24542, "chunks": 211.22, "bytes_per_chunk": 4.848, "boundary_entropy": 0.23401, "router_noise": 0.0, "grad_norm": 0.9235, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.112, "alpha": 0.03, "t": 138.6}
{"event": "train", "step": 1250, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.63894, "bpb": 2.36448, "chunk_loss": 0.030543, "F": 0.19672, "G": 0.23917, "chunks": 201.44, "bytes_per_chunk": 5.083, "boundary_entropy": 0.23345, "router_noise": 0.0, "grad_norm": 0.9154, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 143.0}
{"event": "train", "step": 1300, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.67425, "bpb": 2.41543, "chunk_loss": 0.030849, "F": 0.21207, "G": 0.24521, "chunks": 217.16, "bytes_per_chunk": 4.715, "boundary_entropy": 0.22383, "router_noise": 0.0, "grad_norm": 0.9232, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 147.2}
{"event": "train", "step": 1350, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.59753, "bpb": 2.30475, "chunk_loss": 0.030659, "F": 0.20465, "G": 0.23992, "chunks": 209.56, "bytes_per_chunk": 4.886, "boundary_entropy": 0.21954, "router_noise": 0.0, "grad_norm": 0.8901, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.078, "alpha": 0.03, "t": 151.1}
{"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3492.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200655, "cps": 27.28125, "mean_p": 0.229628, "boundary_entropy": 0.211123, "bpb": 2.214484}, "tr": {"chunks": 3334.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.184168, "cps": 26.046875, "mean_p": 0.232413, "boundary_entropy": 0.281886, "bpb": 2.50977}, "fi": {"chunks": 4030.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.214236, "cps": 31.484375, "mean_p": 0.247436, "boundary_entropy": 0.249422, "bpb": 2.652041}}, "bpb": {"en": 2.21448, "tr": 2.50977, "fi": 2.65204}, "gini_cpc": 0.033461, "gini_cps": 0.042741, "bpic": 5.225866, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9178356646854492, "fi": 1.067684770243922}, "t": 154.9}
{"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.22548, "tr": 2.51845, "fi": 2.65912}, "bytes_per_chunk": {"en": 4.795, "tr": 5.925, "fi": 4.869}, "bpb_mean": 2.46768, "frozen_inventory": false, "t": 155.5}
{"event": "train", "step": 1400, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.6031, "bpb": 2.31279, "chunk_loss": 0.030569, "F": 0.19913, "G": 0.23574, "chunks": 203.91, "bytes_per_chunk": 5.022, "boundary_entropy": 0.22531, "router_noise": 0.0, "grad_norm": 0.8692, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 155.7}
{"event": "train", "step": 1450, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.65093, "bpb": 2.3818, "chunk_loss": 0.030541, "F": 0.19949, "G": 0.23687, "chunks": 204.28, "bytes_per_chunk": 5.013, "boundary_entropy": 0.22124, "router_noise": 0.0, "grad_norm": 0.8477, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 160.0}
{"event": "train", "step": 1500, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.66757, "bpb": 2.40579, "chunk_loss": 0.030755, "F": 0.20633, "G": 0.24267, "chunks": 211.28, "bytes_per_chunk": 4.847, "boundary_entropy": 0.22032, "router_noise": 0.0, "grad_norm": 0.8593, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 164.6}
{"event": "train", "step": 1550, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.59384, "bpb": 2.29942, "chunk_loss": 0.030647, "F": 0.20401, "G": 0.23888, "chunks": 208.91, "bytes_per_chunk": 4.902, "boundary_entropy": 0.22086, "router_noise": 0.0, "grad_norm": 0.8961, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.116, "alpha": 0.03, "t": 170.4}
{"event": "probe", "step": 1600, "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": 0.223889, "boundary_entropy": 0.19361, "bpb": 2.175288}, "tr": {"chunks": 3386.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.187041, "cps": 26.453125, "mean_p": 0.230063, "boundary_entropy": 0.267978, "bpb": 2.436076}, "fi": {"chunks": 4133.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.219712, "cps": 32.289062, "mean_p": 0.244608, "boundary_entropy": 0.223873, "bpb": 2.595208}}, "bpb": {"en": 2.17529, "tr": 2.43608, "fi": 2.59521}, "gini_cpc": 0.036039, "gini_cps": 0.045446, "bpic": 5.177222, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9465168434503497, "fi": 1.111848120393942}, "t": 176.2}
{"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.18557, "tr": 2.44528, "fi": 2.59958}, "bytes_per_chunk": {"en": 4.854, "tr": 5.596, "fi": 4.76}, "bpb_mean": 2.41014, "frozen_inventory": false, "t": 177.1}
{"event": "train", "step": 1600, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.51595, "bpb": 2.18706, "chunk_loss": 0.030683, "F": 0.20807, "G": 0.23491, "chunks": 213.06, "bytes_per_chunk": 4.806, "boundary_entropy": 0.20032, "router_noise": 0.0, "grad_norm": 0.8517, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.141, "alpha": 0.03, "t": 177.5}
{"event": "train", "step": 1650, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.58318, "bpb": 2.28405, "chunk_loss": 0.030576, "F": 0.20441, "G": 0.23214, "chunks": 209.31, "bytes_per_chunk": 4.892, "boundary_entropy": 0.20244, "router_noise": 0.0, "grad_norm": 0.8631, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.114, "alpha": 0.03, "t": 183.2}
{"event": "train", "step": 1700, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.54574, "bpb": 2.23003, "chunk_loss": 0.03045, "F": 0.19693, "G": 0.22819, "chunks": 201.66, "bytes_per_chunk": 5.078, "boundary_entropy": 0.20951, "router_noise": 0.0, "grad_norm": 0.7856, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.121, "alpha": 0.03, "t": 189.2}
{"event": "train", "step": 1750, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.57374, "bpb": 2.27043, "chunk_loss": 0.030507, "F": 0.19867, "G": 0.23231, "chunks": 203.44, "bytes_per_chunk": 5.033, "boundary_entropy": 0.2079, "router_noise": 0.0, "grad_norm": 0.8367, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.109, "alpha": 0.03, "t": 194.6}
{"event": "probe", "step": 1800, "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": 0.224847, "boundary_entropy": 0.179356, "bpb": 2.138063}, "tr": {"chunks": 3225.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178147, "cps": 25.195312, "mean_p": 0.222361, "boundary_entropy": 0.249968, "bpb": 2.378426}, "fi": {"chunks": 4251.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.225985, "cps": 33.210938, "mean_p": 0.245091, "boundary_entropy": 0.205064, "bpb": 2.543451}}, "bpb": {"en": 2.13806, "tr": 2.37843, "fi": 2.54345}, "gini_cpc": 0.052632, "gini_cps": 0.06225, "bpic": 5.163087, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8827724692807558, "fi": 1.1198215851506454}, "t": 200.8}
{"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.15136, "tr": 2.3877, "fi": 2.54899}, "bytes_per_chunk": {"en": 4.758, "tr": 6.4, "fi": 4.709}, "bpb_mean": 2.36268, "frozen_inventory": false, "t": 201.6}
{"event": "train", "step": 1800, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.54535, "bpb": 2.22947, "chunk_loss": 0.030527, "F": 0.19861, "G": 0.22988, "chunks": 203.38, "bytes_per_chunk": 5.035, "boundary_entropy": 0.19539, "router_noise": 0.0, "grad_norm": 0.7328, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.145, "alpha": 0.03, "t": 201.9}
{"event": "train", "step": 1850, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.55577, "bpb": 2.24451, "chunk_loss": 0.030581, "F": 0.20322, "G": 0.22967, "chunks": 208.09, "bytes_per_chunk": 4.921, "boundary_entropy": 0.18689, "router_noise": 0.0, "grad_norm": 0.7551, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.118, "alpha": 0.03, "t": 207.8}
{"event": "train", "step": 1900, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.53962, "bpb": 2.2212, "chunk_loss": 0.030603, "F": 0.20172, "G": 0.23025, "chunks": 206.56, "bytes_per_chunk": 4.957, "boundary_entropy": 0.19709, "router_noise": 0.0, "grad_norm": 0.8895, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.11, "alpha": 0.03, "t": 213.2}
{"event": "train", "step": 1950, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.49243, "bpb": 2.15313, "chunk_loss": 0.030647, "F": 0.20584, "G": 0.23245, "chunks": 210.78, "bytes_per_chunk": 4.858, "boundary_entropy": 0.19528, "router_noise": 0.0, "grad_norm": 0.8216, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.112, "alpha": 0.03, "t": 218.8}
{"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3492.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200655, "cps": 27.28125, "mean_p": 0.218948, "boundary_entropy": 0.170937, "bpb": 2.110065}, "tr": {"chunks": 3433.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.189637, "cps": 26.820312, "mean_p": 0.228127, "boundary_entropy": 0.255642, "bpb": 2.330108}, "fi": {"chunks": 4191.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.222795, "cps": 32.742188, "mean_p": 0.239208, "boundary_entropy": 0.199774, "bpb": 2.4983}}, "bpb": {"en": 2.11006, "tr": 2.33011, "fi": 2.4983}, "gini_cpc": 0.036056, "gini_cps": 0.04546, "bpic": 5.103634, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9450899330729295, "fi": 1.110339174216446}, "t": 223.7}
{"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.12207, "tr": 2.34052, "fi": 2.50756}, "bytes_per_chunk": {"en": 4.788, "tr": 5.697, "fi": 4.751}, "bpb_mean": 2.32339, "frozen_inventory": false, "t": 224.3}
{"event": "train", "step": 2000, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.65344, "bpb": 2.38541, "chunk_loss": 0.030771, "F": 0.21286, "G": 0.23587, "chunks": 217.97, "bytes_per_chunk": 4.698, "boundary_entropy": 0.18979, "router_noise": 0.0, "grad_norm": 0.8898, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.118, "alpha": 0.03, "t": 224.7}
{"event": "train", "step": 2050, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.48443, "bpb": 2.14158, "chunk_loss": 0.030466, "F": 0.19653, "G": 0.22141, "chunks": 201.25, "bytes_per_chunk": 5.088, "boundary_entropy": 0.18568, "router_noise": 0.0, "grad_norm": 0.7708, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 228.9}
{"event": "train", "step": 2100, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.5441, "bpb": 2.22767, "chunk_loss": 0.030523, "F": 0.20291, "G": 0.22676, "chunks": 207.78, "bytes_per_chunk": 4.928, "boundary_entropy": 0.18147, "router_noise": 0.0, "grad_norm": 0.764, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 233.3}
{"event": "train", "step": 2150, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.50475, "bpb": 2.17089, "chunk_loss": 0.030582, "F": 0.20654, "G": 0.22785, "chunks": 211.5, "bytes_per_chunk": 4.842, "boundary_entropy": 0.17462, "router_noise": 0.0, "grad_norm": 0.7082, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 237.6}
{"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3447.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198069, "cps": 26.929688, "mean_p": 0.217472, "boundary_entropy": 0.168257, "bpb": 2.087226}, "tr": {"chunks": 3310.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.182843, "cps": 25.859375, "mean_p": 0.222964, "boundary_entropy": 0.244994, "bpb": 2.299592}, "fi": {"chunks": 4107.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.21833, "cps": 32.085938, "mean_p": 0.23911, "boundary_entropy": 0.19358, "bpb": 2.452383}}, "bpb": {"en": 2.08723, "tr": 2.29959, "fi": 2.45238}, "gini_cpc": 0.03948, "gini_cps": 0.048908, "bpic": 5.222018, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9231245036440978, "fi": 1.1022894637088765}, "t": 242.2}
{"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.09711, "tr": 2.30825, "fi": 2.46007}, "bytes_per_chunk": {"en": 4.864, "tr": 5.767, "fi": 4.808}, "bpb_mean": 2.28848, "frozen_inventory": false, "t": 242.9}
{"event": "train", "step": 2200, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.48095, "bpb": 2.13655, "chunk_loss": 0.030452, "F": 0.20013, "G": 0.22388, "chunks": 204.94, "bytes_per_chunk": 4.997, "boundary_entropy": 0.17647, "router_noise": 0.0, "grad_norm": 0.6986, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.112, "alpha": 0.03, "t": 243.2}
{"event": "train", "step": 2250, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.50532, "bpb": 2.17171, "chunk_loss": 0.030554, "F": 0.20319, "G": 0.2258, "chunks": 208.06, "bytes_per_chunk": 4.922, "boundary_entropy": 0.17574, "router_noise": 0.0, "grad_norm": 0.8173, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 247.5}
{"event": "train", "step": 2300, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.51549, "bpb": 2.18639, "chunk_loss": 0.030496, "F": 0.20367, "G": 0.22438, "chunks": 208.56, "bytes_per_chunk": 4.91, "boundary_entropy": 0.17655, "router_noise": 0.0, "grad_norm": 0.7801, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 251.8}
{"event": "train", "step": 2350, "arch": "hnet_ffn", "phase": 1, "objective": "global", "lm_loss": 1.56293, "bpb": 2.25484, "chunk_loss": 0.030506, "F": 0.20383, "G": 0.22697, "chunks": 208.72, "bytes_per_chunk": 4.906, "boundary_entropy": 0.17332, "router_noise": 0.0, "grad_norm": 0.7109, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 256.0}
{"event": "probe", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3477.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199793, "cps": 27.164062, "mean_p": 0.215523, "boundary_entropy": 0.153721, "bpb": 2.064218}, "tr": {"chunks": 3286.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181517, "cps": 25.671875, "mean_p": 0.216838, "boundary_entropy": 0.229502, "bpb": 2.269917}, "fi": {"chunks": 4220.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.224337, "cps": 32.96875, "mean_p": 0.237476, "boundary_entropy": 0.17401, "bpb": 2.398859}}, "bpb": {"en": 2.06422, "tr": 2.26992, "fi": 2.39886}, "gini_cpc": 0.047134, "gini_cps": 0.056694, "bpic": 5.165437, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.908524068749158, "fi": 1.1228454850527354}, "t": 260.5}
{"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.07726, "tr": 2.27823, "fi": 2.40784}, "bytes_per_chunk": {"en": 4.787, "tr": 6.274, "fi": 4.713}, "bpb_mean": 2.25444, "frozen_inventory": false, "t": 260.9}
{"event": "phase_transition", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 261.2}
{"event": "train", "step": 2400, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.50332, "bpb": 2.16884, "chunk_loss": 0.030616, "F": 0.20807, "G": 0.22883, "chunks": 213.06, "bytes_per_chunk": 4.806, "boundary_entropy": 0.16855, "router_noise": 0.0, "grad_norm": 0.7812, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.108, "alpha": 0.03, "t": 261.3}
{"event": "train", "step": 2450, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.46581, "bpb": 2.11472, "chunk_loss": 0.030457, "F": 0.19852, "G": 0.22409, "chunks": 203.28, "bytes_per_chunk": 5.037, "boundary_entropy": 0.17604, "router_noise": 0.0, "grad_norm": 0.6994, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 265.7}
{"event": "train", "step": 2500, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.44045, "bpb": 2.07813, "chunk_loss": 0.030635, "F": 0.21201, "G": 0.22841, "chunks": 217.09, "bytes_per_chunk": 4.717, "boundary_entropy": 0.15818, "router_noise": 0.0, "grad_norm": 0.7771, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 270.0}
{"event": "train", "step": 2550, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.46923, "bpb": 2.11965, "chunk_loss": 0.030411, "F": 0.19769, "G": 0.22091, "chunks": 202.44, "bytes_per_chunk": 5.058, "boundary_entropy": 0.168, "router_noise": 0.0, "grad_norm": 0.7457, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 274.3}
{"event": "probe", "step": 2600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3473.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199563, "cps": 27.132812, "mean_p": 0.213304, "boundary_entropy": 0.150006, "bpb": 2.044844}, "tr": {"chunks": 3375.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.186433, "cps": 26.367188, "mean_p": 0.221173, "boundary_entropy": 0.230595, "bpb": 2.245755}, "fi": {"chunks": 4204.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.223486, "cps": 32.84375, "mean_p": 0.236253, "boundary_entropy": 0.16721, "bpb": 2.384459}}, "bpb": {"en": 2.04484, "tr": 2.24576, "fi": 2.38446}, "gini_cpc": 0.04053, "gini_cps": 0.050006, "bpic": 5.133189, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9342058072247078, "fi": 1.1198765760665028}, "t": 278.7}
{"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.0566, "tr": 2.25754, "fi": 2.39062}, "bytes_per_chunk": {"en": 4.765, "tr": 6.018, "fi": 4.74}, "bpb_mean": 2.23492, "frozen_inventory": false, "t": 279.2}
{"event": "train", "step": 2600, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.42399, "bpb": 2.05438, "chunk_loss": 0.030493, "F": 0.20505, "G": 0.22187, "chunks": 209.97, "bytes_per_chunk": 4.877, "boundary_entropy": 0.16112, "router_noise": 0.0, "grad_norm": 0.7364, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.106, "alpha": 0.03, "t": 279.6}
{"event": "train", "step": 2650, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.46783, "bpb": 2.11763, "chunk_loss": 0.03038, "F": 0.19635, "G": 0.21922, "chunks": 201.06, "bytes_per_chunk": 5.093, "boundary_entropy": 0.16978, "router_noise": 0.0, "grad_norm": 0.6965, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 283.7}
{"event": "train", "step": 2700, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.45055, "bpb": 2.0927, "chunk_loss": 0.030551, "F": 0.20566, "G": 0.22438, "chunks": 210.59, "bytes_per_chunk": 4.862, "boundary_entropy": 0.16259, "router_noise": 0.0, "grad_norm": 0.6602, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 288.0}
{"event": "train", "step": 2750, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.45681, "bpb": 2.10173, "chunk_loss": 0.030598, "F": 0.20743, "G": 0.22601, "chunks": 212.41, "bytes_per_chunk": 4.821, "boundary_entropy": 0.16349, "router_noise": 0.0, "grad_norm": 0.6854, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 292.4}
{"event": "probe", "step": 2800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3446.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198012, "cps": 26.921875, "mean_p": 0.210935, "boundary_entropy": 0.142563, "bpb": 2.026357}, "tr": {"chunks": 3331.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.184003, "cps": 26.023438, "mean_p": 0.217245, "boundary_entropy": 0.222502, "bpb": 2.212027}, "fi": {"chunks": 4188.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.222636, "cps": 32.71875, "mean_p": 0.233256, "boundary_entropy": 0.162019, "bpb": 2.337631}}, "bpb": {"en": 2.02636, "tr": 2.21203, "fi": 2.33763}, "gini_cpc": 0.042595, "gini_cps": 0.052105, "bpic": 5.173917, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9292507672530589, "fi": 1.1243554689000486}, "t": 297.2}
{"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.03676, "tr": 2.22034, "fi": 2.34418}, "bytes_per_chunk": {"en": 4.825, "tr": 6.075, "fi": 4.754}, "bpb_mean": 2.20043, "frozen_inventory": false, "t": 297.7}
{"event": "train", "step": 2800, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.41499, "bpb": 2.0414, "chunk_loss": 0.030444, "F": 0.20203, "G": 0.21943, "chunks": 206.88, "bytes_per_chunk": 4.95, "boundary_entropy": 0.15365, "router_noise": 0.0, "grad_norm": 0.7044, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.115, "alpha": 0.03, "t": 298.1}
{"event": "train", "step": 2850, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.46384, "bpb": 2.11187, "chunk_loss": 0.030675, "F": 0.21429, "G": 0.22779, "chunks": 219.44, "bytes_per_chunk": 4.666, "boundary_entropy": 0.14538, "router_noise": 0.0, "grad_norm": 0.7393, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 302.4}
{"event": "train", "step": 2900, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.39414, "bpb": 2.01132, "chunk_loss": 0.030384, "F": 0.19879, "G": 0.2165, "chunks": 203.56, "bytes_per_chunk": 5.03, "boundary_entropy": 0.1591, "router_noise": 0.0, "grad_norm": 0.6755, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.079, "alpha": 0.03, "t": 306.4}
{"event": "train", "step": 2950, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.5457, "bpb": 2.22997, "chunk_loss": 0.030625, "F": 0.20584, "G": 0.22568, "chunks": 210.78, "bytes_per_chunk": 4.858, "boundary_entropy": 0.15948, "router_noise": 0.0, "grad_norm": 0.6955, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 310.7}
{"event": "probe", "step": 3000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3443.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197839, "cps": 26.898438, "mean_p": 0.210184, "boundary_entropy": 0.136656, "bpb": 2.007236}, "tr": {"chunks": 3282.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181296, "cps": 25.640625, "mean_p": 0.213621, "boundary_entropy": 0.218353, "bpb": 2.186495}, "fi": {"chunks": 4186.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.222529, "cps": 32.703125, "mean_p": 0.231797, "boundary_entropy": 0.152957, "bpb": 2.318779}}, "bpb": {"en": 2.00724, "tr": 2.18649, "fi": 2.31878}, "gini_cpc": 0.045688, "gini_cps": 0.055235, "bpic": 5.199523, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9163789885383168, "fi": 1.1247977477413282}, "t": 315.1}
{"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.02109, "tr": 2.19549, "fi": 2.32231}, "bytes_per_chunk": {"en": 4.799, "tr": 6.137, "fi": 4.76}, "bpb_mean": 2.17963, "frozen_inventory": false, "t": 315.7}
{"event": "train", "step": 3000, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.49272, "bpb": 2.15354, "chunk_loss": 0.030462, "F": 0.20102, "G": 0.21615, "chunks": 205.84, "bytes_per_chunk": 4.975, "boundary_entropy": 0.14649, "router_noise": 0.0, "grad_norm": 0.7095, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.114, "alpha": 0.03, "t": 316.4}
{"event": "train", "step": 3050, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.44185, "bpb": 2.08016, "chunk_loss": 0.030558, "F": 0.20746, "G": 0.2193, "chunks": 212.44, "bytes_per_chunk": 4.82, "boundary_entropy": 0.1387, "router_noise": 0.0, "grad_norm": 0.721, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 320.9}
{"event": "train", "step": 3100, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.49037, "bpb": 2.15014, "chunk_loss": 0.03037, "F": 0.19592, "G": 0.21487, "chunks": 200.62, "bytes_per_chunk": 5.104, "boundary_entropy": 0.1547, "router_noise": 0.0, "grad_norm": 0.7086, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 325.3}
{"event": "train", "step": 3150, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.45213, "bpb": 2.09499, "chunk_loss": 0.030455, "F": 0.20203, "G": 0.2158, "chunks": 206.88, "bytes_per_chunk": 4.95, "boundary_entropy": 0.14367, "router_noise": 0.0, "grad_norm": 0.7173, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 329.6}
{"event": "probe", "step": 3200, "phase": 2, "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": 0.210019, "boundary_entropy": 0.128529, "bpb": 1.987305}, "tr": {"chunks": 3257.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179915, "cps": 25.445312, "mean_p": 0.212749, "boundary_entropy": 0.211644, "bpb": 2.163112}, "fi": {"chunks": 4164.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.22136, "cps": 32.53125, "mean_p": 0.230847, "boundary_entropy": 0.149888, "bpb": 2.297331}}, "bpb": {"en": 1.98731, "tr": 2.16311, "fi": 2.29733}, "gini_cpc": 0.045981, "gini_cps": 0.055499, "bpic": 5.207159, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.901283692614045, "fi": 1.1089019405744023}, "t": 334.4}
{"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.00083, "tr": 2.17209, "fi": 2.30441}, "bytes_per_chunk": {"en": 4.765, "tr": 6.172, "fi": 4.791}, "bpb_mean": 2.15911, "frozen_inventory": false, "t": 335.0}
{"event": "train", "step": 3200, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.43331, "bpb": 2.06782, "chunk_loss": 0.030492, "F": 0.20587, "G": 0.2191, "chunks": 210.81, "bytes_per_chunk": 4.857, "boundary_entropy": 0.13609, "router_noise": 0.0, "grad_norm": 0.647, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.115, "alpha": 0.03, "t": 335.4}
{"event": "train", "step": 3250, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.43747, "bpb": 2.07384, "chunk_loss": 0.030505, "F": 0.20248, "G": 0.21974, "chunks": 207.34, "bytes_per_chunk": 4.939, "boundary_entropy": 0.14888, "router_noise": 0.0, "grad_norm": 0.7127, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.078, "alpha": 0.03, "t": 339.3}
{"event": "train", "step": 3300, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.41316, "bpb": 2.03876, "chunk_loss": 0.030547, "F": 0.2077, "G": 0.21931, "chunks": 212.69, "bytes_per_chunk": 4.815, "boundary_entropy": 0.13842, "router_noise": 0.0, "grad_norm": 0.6828, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 343.6}
{"event": "train", "step": 3350, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.35877, "bpb": 1.96029, "chunk_loss": 0.030477, "F": 0.20657, "G": 0.21617, "chunks": 211.53, "bytes_per_chunk": 4.841, "boundary_entropy": 0.13432, "router_noise": 0.0, "grad_norm": 0.7163, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 347.8}
{"event": "probe", "step": 3400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3501.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201172, "cps": 27.351562, "mean_p": 0.210309, "boundary_entropy": 0.125891, "bpb": 1.976739}, "tr": {"chunks": 3260.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.180081, "cps": 25.46875, "mean_p": 0.213626, "boundary_entropy": 0.212202, "bpb": 2.166897}, "fi": {"chunks": 4205.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.223539, "cps": 32.851562, "mean_p": 0.230035, "boundary_entropy": 0.142946, "bpb": 2.267044}}, "bpb": {"en": 1.97674, "tr": 2.1669, "fi": 2.26704}, "gini_cpc": 0.047905, "gini_cps": 0.05745, "bpic": 5.173445, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.8951566824532247, "fi": 1.1111843755661388}, "t": 352.4}
{"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 1.98914, "tr": 2.17519, "fi": 2.27314}, "bytes_per_chunk": {"en": 4.753, "tr": 6.238, "fi": 4.735}, "bpb_mean": 2.14582, "frozen_inventory": false, "t": 353.0}
{"event": "train", "step": 3400, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.42745, "bpb": 2.05937, "chunk_loss": 0.030434, "F": 0.19962, "G": 0.21721, "chunks": 204.41, "bytes_per_chunk": 5.01, "boundary_entropy": 0.14403, "router_noise": 0.0, "grad_norm": 0.761, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.11, "alpha": 0.03, "t": 353.3}
{"event": "train", "step": 3450, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.39054, "bpb": 2.00613, "chunk_loss": 0.030464, "F": 0.20276, "G": 0.21903, "chunks": 207.62, "bytes_per_chunk": 4.932, "boundary_entropy": 0.14556, "router_noise": 0.0, "grad_norm": 0.8092, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 357.5}
{"event": "train", "step": 3500, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.36297, "bpb": 1.96636, "chunk_loss": 0.030409, "F": 0.20132, "G": 0.21123, "chunks": 206.16, "bytes_per_chunk": 4.967, "boundary_entropy": 0.12948, "router_noise": 0.0, "grad_norm": 0.6579, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.08, "alpha": 0.03, "t": 361.5}
{"event": "train", "step": 3550, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.40946, "bpb": 2.03342, "chunk_loss": 0.030465, "F": 0.20172, "G": 0.2173, "chunks": 206.56, "bytes_per_chunk": 4.957, "boundary_entropy": 0.14308, "router_noise": 0.0, "grad_norm": 0.6877, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 365.7}
{"event": "probe", "step": 3600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3515.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201977, "cps": 27.460938, "mean_p": 0.209649, "boundary_entropy": 0.12275, "bpb": 1.960584}, "tr": {"chunks": 3298.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.18218, "cps": 25.765625, "mean_p": 0.211656, "boundary_entropy": 0.200581, "bpb": 2.124539}, "fi": {"chunks": 4229.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.224815, "cps": 33.039062, "mean_p": 0.229696, "boundary_entropy": 0.137749, "bpb": 2.244386}}, "bpb": {"en": 1.96058, "tr": 2.12454, "fi": 2.24439}, "gini_cpc": 0.046675, "gini_cps": 0.05621, "bpic": 5.137837, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9019841182221945, "fi": 1.113075420520952}, "t": 370.1}
{"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 1.9729, "tr": 2.13389, "fi": 2.24978}, "bytes_per_chunk": {"en": 4.731, "tr": 6.139, "fi": 4.732}, "bpb_mean": 2.11886, "frozen_inventory": false, "t": 370.7}
{"event": "train", "step": 3600, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.40466, "bpb": 2.02649, "chunk_loss": 0.030545, "F": 0.20798, "G": 0.2201, "chunks": 212.97, "bytes_per_chunk": 4.808, "boundary_entropy": 0.13287, "router_noise": 0.0, "grad_norm": 0.6216, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.107, "alpha": 0.03, "t": 371.0}
{"event": "train", "step": 3650, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.44375, "bpb": 2.0829, "chunk_loss": 0.030517, "F": 0.2059, "G": 0.21845, "chunks": 210.84, "bytes_per_chunk": 4.857, "boundary_entropy": 0.13037, "router_noise": 0.0, "grad_norm": 0.7064, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 375.2}
{"event": "train", "step": 3700, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.39443, "bpb": 2.01173, "chunk_loss": 0.030463, "F": 0.2048, "G": 0.21723, "chunks": 209.72, "bytes_per_chunk": 4.883, "boundary_entropy": 0.13464, "router_noise": 0.0, "grad_norm": 0.6849, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 379.7}
{"event": "train", "step": 3750, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.37693, "bpb": 1.9865, "chunk_loss": 0.030417, "F": 0.20273, "G": 0.21548, "chunks": 207.59, "bytes_per_chunk": 4.933, "boundary_entropy": 0.13721, "router_noise": 0.0, "grad_norm": 0.6755, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.081, "alpha": 0.03, "t": 383.8}
{"event": "probe", "step": 3800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3445.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197954, "cps": 26.914062, "mean_p": 0.206389, "boundary_entropy": 0.121541, "bpb": 1.951136}, "tr": {"chunks": 3328.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.183837, "cps": 26.0, "mean_p": 0.209619, "boundary_entropy": 0.196107, "bpb": 2.115601}, "fi": {"chunks": 4237.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.225241, "cps": 33.101562, "mean_p": 0.229777, "boundary_entropy": 0.136322, "bpb": 2.227625}}, "bpb": {"en": 1.95114, "tr": 2.1156, "fi": 2.22763}, "gini_cpc": 0.045471, "gini_cps": 0.055041, "bpic": 5.15277, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9286833517638587, "fi": 1.1378407270118562}, "t": 388.2}
{"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 1.96459, "tr": 2.12368, "fi": 2.23373}, "bytes_per_chunk": {"en": 4.797, "tr": 5.86, "fi": 4.717}, "bpb_mean": 2.10733, "frozen_inventory": false, "t": 388.9}
{"event": "train", "step": 3800, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.45127, "bpb": 2.09374, "chunk_loss": 0.030488, "F": 0.20624, "G": 0.21761, "chunks": 211.19, "bytes_per_chunk": 4.849, "boundary_entropy": 0.13201, "router_noise": 0.0, "grad_norm": 0.7183, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.11, "alpha": 0.03, "t": 389.2}
{"event": "train", "step": 3850, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.54898, "bpb": 2.2347, "chunk_loss": 0.030707, "F": 0.20309, "G": 0.2196, "chunks": 207.97, "bytes_per_chunk": 4.924, "boundary_entropy": 0.1394, "router_noise": 0.0, "grad_norm": 0.7034, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 393.4}
{"event": "train", "step": 3900, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.4362, "bpb": 2.072, "chunk_loss": 0.03048, "F": 0.20389, "G": 0.21469, "chunks": 208.78, "bytes_per_chunk": 4.905, "boundary_entropy": 0.12744, "router_noise": 0.0, "grad_norm": 0.7418, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 397.5}
{"event": "train", "step": 3950, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.492, "bpb": 2.1525, "chunk_loss": 0.030412, "F": 0.19913, "G": 0.21309, "chunks": 203.91, "bytes_per_chunk": 5.022, "boundary_entropy": 0.13353, "router_noise": 0.0, "grad_norm": 0.7808, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.081, "alpha": 0.03, "t": 401.5}
{"event": "probe", "step": 4000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3459.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198759, "cps": 27.023438, "mean_p": 0.204989, "boundary_entropy": 0.11867, "bpb": 1.936472}, "tr": {"chunks": 3367.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185991, "cps": 26.304688, "mean_p": 0.212531, "boundary_entropy": 0.197096, "bpb": 2.101044}, "fi": {"chunks": 4230.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.224868, "cps": 33.046875, "mean_p": 0.228528, "boundary_entropy": 0.131749, "bpb": 2.204217}}, "bpb": {"en": 1.93647, "tr": 2.10104, "fi": 2.20422}, "gini_cpc": 0.042515, "gini_cps": 0.052038, "bpic": 5.131331, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9357635471189985, "fi": 1.1313631839575697}, "t": 405.7}
{"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 1.95015, "tr": 2.10841, "fi": 2.21059}, "bytes_per_chunk": {"en": 4.786, "tr": 5.91, "fi": 4.722}, "bpb_mean": 2.08971, "frozen_inventory": false, "t": 406.3}
{"event": "train", "step": 4000, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.4143, "bpb": 2.04041, "chunk_loss": 0.03052, "F": 0.2049, "G": 0.2163, "chunks": 209.81, "bytes_per_chunk": 4.881, "boundary_entropy": 0.13017, "router_noise": 0.0, "grad_norm": 0.6586, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.102, "alpha": 0.03, "t": 406.6}
{"event": "train", "step": 4050, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.42457, "bpb": 2.05522, "chunk_loss": 0.030449, "F": 0.20114, "G": 0.21651, "chunks": 205.97, "bytes_per_chunk": 4.972, "boundary_entropy": 0.13526, "router_noise": 0.0, "grad_norm": 0.6846, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.077, "alpha": 0.03, "t": 410.4}
{"event": "train", "step": 4100, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.36196, "bpb": 1.96489, "chunk_loss": 0.030402, "F": 0.20184, "G": 0.213, "chunks": 206.69, "bytes_per_chunk": 4.954, "boundary_entropy": 0.12682, "router_noise": 0.0, "grad_norm": 0.7254, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 414.8}
{"event": "train", "step": 4150, "arch": "hnet_ffn", "phase": 2, "objective": "global", "lm_loss": 1.42521, "bpb": 2.05615, "chunk_loss": 0.030625, "F": 0.21097, "G": 0.22117, "chunks": 216.03, "bytes_per_chunk": 4.74, "boundary_entropy": 0.1272, "router_noise": 0.0, "grad_norm": 0.7363, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 419.0}
{"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3473.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199563, "cps": 27.132812, "mean_p": 0.205159, "boundary_entropy": 0.113849, "bpb": 1.928657}, "tr": {"chunks": 3286.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181517, "cps": 25.671875, "mean_p": 0.211968, "boundary_entropy": 0.199305, "bpb": 2.082253}, "fi": {"chunks": 4301.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.228643, "cps": 33.601562, "mean_p": 0.231298, "boundary_entropy": 0.129026, "bpb": 2.191808}}, "bpb": {"en": 1.92866, "tr": 2.08225, "fi": 2.19181}, "gini_cpc": 0.051527, "gini_cps": 0.061181, "bpic": 5.129476, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9095704540860415, "fi": 1.1457157834590934}, "t": 423.3}
{"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 1.94136, "tr": 2.09231, "fi": 2.19898}, "bytes_per_chunk": {"en": 4.78, "tr": 6.181, "fi": 4.661}, "bpb_mean": 2.07755, "frozen_inventory": false, "t": 423.9}
{"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 424.2}
{"event": "train", "step": 4200, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.35489, "bpb": 1.95469, "chunk_loss": 0.030325, "F": 0.19604, "G": 0.20975, "chunks": 200.75, "bytes_per_chunk": 5.101, "boundary_entropy": 0.12942, "router_noise": 0.0, "grad_norm": 0.6923, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.105, "alpha": 0.03, "t": 424.2}
{"event": "train", "step": 4250, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.38858, "bpb": 2.00329, "chunk_loss": 0.030493, "F": 0.20514, "G": 0.21794, "chunks": 210.06, "bytes_per_chunk": 4.875, "boundary_entropy": 0.1254, "router_noise": 0.0, "grad_norm": 0.6848, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.1, "alpha": 0.03, "t": 429.3}
{"event": "train", "step": 4300, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.38643, "bpb": 2.00019, "chunk_loss": 0.030323, "F": 0.19641, "G": 0.20935, "chunks": 201.12, "bytes_per_chunk": 5.091, "boundary_entropy": 0.12989, "router_noise": 0.0, "grad_norm": 0.7506, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.101, "alpha": 0.03, "t": 434.3}
{"event": "train", "step": 4350, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.40415, "bpb": 2.02576, "chunk_loss": 0.030436, "F": 0.2045, "G": 0.21452, "chunks": 209.41, "bytes_per_chunk": 4.89, "boundary_entropy": 0.12157, "router_noise": 0.0, "grad_norm": 0.6817, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 438.9}
{"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3484.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200195, "cps": 27.21875, "mean_p": 0.204735, "boundary_entropy": 0.111409, "bpb": 1.917818}, "tr": {"chunks": 3270.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.180633, "cps": 25.546875, "mean_p": 0.210849, "boundary_entropy": 0.195868, "bpb": 2.073244}, "fi": {"chunks": 4210.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.223805, "cps": 32.890625, "mean_p": 0.227536, "boundary_entropy": 0.126435, "bpb": 2.172148}}, "bpb": {"en": 1.91782, "tr": 2.07324, "fi": 2.17215}, "gini_cpc": 0.047601, "gini_cps": 0.057157, "bpic": 5.174389, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.902283831523316, "fi": 1.1179340556106452}, "t": 446.2}
{"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 1.93085, "tr": 2.0824, "fi": 2.18007}, "bytes_per_chunk": {"en": 4.777, "tr": 6.06, "fi": 4.703}, "bpb_mean": 2.06444, "frozen_inventory": false, "t": 446.8}
{"event": "train", "step": 4400, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.36015, "bpb": 1.96229, "chunk_loss": 0.030397, "F": 0.20444, "G": 0.21238, "chunks": 209.34, "bytes_per_chunk": 4.891, "boundary_entropy": 0.11841, "router_noise": 0.0, "grad_norm": 0.6494, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 447.1}
{"event": "train", "step": 4450, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.35726, "bpb": 1.95811, "chunk_loss": 0.030328, "F": 0.1973, "G": 0.20701, "chunks": 202.03, "bytes_per_chunk": 5.069, "boundary_entropy": 0.12216, "router_noise": 0.0, "grad_norm": 0.6687, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.08, "alpha": 0.03, "t": 451.2}
{"event": "train", "step": 4500, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.33037, "bpb": 1.91932, "chunk_loss": 0.030457, "F": 0.20642, "G": 0.21563, "chunks": 211.38, "bytes_per_chunk": 4.844, "boundary_entropy": 0.11642, "router_noise": 0.0, "grad_norm": 0.6548, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 455.3}
{"event": "train", "step": 4550, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.36707, "bpb": 1.97227, "chunk_loss": 0.030338, "F": 0.19879, "G": 0.21076, "chunks": 203.56, "bytes_per_chunk": 5.03, "boundary_entropy": 0.12413, "router_noise": 0.0, "grad_norm": 0.6625, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 459.4}
{"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3487.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200368, "cps": 27.242188, "mean_p": 0.204531, "boundary_entropy": 0.109976, "bpb": 1.908935}, "tr": {"chunks": 3370.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.186157, "cps": 26.328125, "mean_p": 0.209639, "boundary_entropy": 0.187533, "bpb": 2.063422}, "fi": {"chunks": 4277.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.227367, "cps": 33.414062, "mean_p": 0.229059, "boundary_entropy": 0.120862, "bpb": 2.165914}}, "bpb": {"en": 1.90894, "tr": 2.06342, "fi": 2.16591}, "gini_cpc": 0.044753, "gini_cps": 0.054308, "bpic": 5.095384, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.929076600691759, "fi": 1.1347482970115677}, "t": 463.7}
{"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 1.92156, "tr": 2.07135, "fi": 2.17218}, "bytes_per_chunk": {"en": 4.771, "tr": 6.05, "fi": 4.663}, "bpb_mean": 2.05503, "frozen_inventory": false, "t": 464.4}
{"event": "train", "step": 4600, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.36156, "bpb": 1.96431, "chunk_loss": 0.030614, "F": 0.21356, "G": 0.22101, "chunks": 218.69, "bytes_per_chunk": 4.682, "boundary_entropy": 0.12076, "router_noise": 0.0, "grad_norm": 0.6989, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.107, "alpha": 0.03, "t": 464.7}
{"event": "train", "step": 4650, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.39592, "bpb": 2.01389, "chunk_loss": 0.030483, "F": 0.20435, "G": 0.21824, "chunks": 209.25, "bytes_per_chunk": 4.894, "boundary_entropy": 0.12996, "router_noise": 0.0, "grad_norm": 0.6734, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.08, "alpha": 0.03, "t": 468.7}
{"event": "train", "step": 4700, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.42463, "bpb": 2.05531, "chunk_loss": 0.030549, "F": 0.21066, "G": 0.21716, "chunks": 215.72, "bytes_per_chunk": 4.747, "boundary_entropy": 0.11496, "router_noise": 0.0, "grad_norm": 0.6956, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.081, "alpha": 0.03, "t": 472.8}
{"event": "train", "step": 4750, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.41072, "bpb": 2.03524, "chunk_loss": 0.030483, "F": 0.20599, "G": 0.21616, "chunks": 210.94, "bytes_per_chunk": 4.855, "boundary_entropy": 0.11974, "router_noise": 0.0, "grad_norm": 0.7025, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.081, "alpha": 0.03, "t": 476.8}
{"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3478.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199851, "cps": 27.171875, "mean_p": 0.20416, "boundary_entropy": 0.103925, "bpb": 1.901122}, "tr": {"chunks": 3307.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.182677, "cps": 25.835938, "mean_p": 0.206504, "boundary_entropy": 0.181809, "bpb": 2.047108}, "fi": {"chunks": 4243.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.22556, "cps": 33.148438, "mean_p": 0.225936, "boundary_entropy": 0.11548, "bpb": 2.15187}}, "bpb": {"en": 1.90112, "tr": 2.04711, "fi": 2.15187}, "gini_cpc": 0.047014, "gini_cps": 0.056583, "bpic": 5.14436, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9140673280430297, "fi": 1.1286406571662935}, "t": 480.9}
{"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 1.91517, "tr": 2.05827, "fi": 2.15575}, "bytes_per_chunk": {"en": 4.775, "tr": 6.009, "fi": 4.695}, "bpb_mean": 2.04306, "frozen_inventory": false, "t": 481.4}
{"event": "train", "step": 4800, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.37624, "bpb": 1.9855, "chunk_loss": 0.030548, "F": 0.21274, "G": 0.21693, "chunks": 217.84, "bytes_per_chunk": 4.701, "boundary_entropy": 0.10919, "router_noise": 0.0, "grad_norm": 0.6554, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.097, "alpha": 0.03, "t": 481.7}
{"event": "train", "step": 4850, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.35336, "bpb": 1.95249, "chunk_loss": 0.030411, "F": 0.20096, "G": 0.2137, "chunks": 205.78, "bytes_per_chunk": 4.976, "boundary_entropy": 0.12638, "router_noise": 0.0, "grad_norm": 0.6768, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 485.8}
{"event": "train", "step": 4900, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.38606, "bpb": 1.99966, "chunk_loss": 0.030431, "F": 0.19931, "G": 0.20986, "chunks": 204.09, "bytes_per_chunk": 5.017, "boundary_entropy": 0.11786, "router_noise": 0.0, "grad_norm": 0.6828, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 489.9}
{"event": "train", "step": 4950, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.34206, "bpb": 1.93618, "chunk_loss": 0.030428, "F": 0.20093, "G": 0.21308, "chunks": 205.75, "bytes_per_chunk": 4.977, "boundary_entropy": 0.12336, "router_noise": 0.0, "grad_norm": 0.6581, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 494.0}
{"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3478.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199851, "cps": 27.171875, "mean_p": 0.203939, "boundary_entropy": 0.105287, "bpb": 1.893718}, "tr": {"chunks": 3281.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181241, "cps": 25.632812, "mean_p": 0.206198, "boundary_entropy": 0.184517, "bpb": 2.041701}, "fi": {"chunks": 4254.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.226144, "cps": 33.234375, "mean_p": 0.226532, "boundary_entropy": 0.115422, "bpb": 2.139806}}, "bpb": {"en": 1.89372, "tr": 2.0417, "fi": 2.13981}, "gini_cpc": 0.049298, "gini_cps": 0.0589, "bpic": 5.151367, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9068808295461689, "fi": 1.131566664055011}, "t": 498.3}
{"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 1.90701, "tr": 2.05149, "fi": 2.14557}, "bytes_per_chunk": {"en": 4.776, "tr": 6.176, "fi": 4.686}, "bpb_mean": 2.03469, "frozen_inventory": false, "t": 498.9}
{"event": "train", "step": 5000, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.33099, "bpb": 1.92021, "chunk_loss": 0.030492, "F": 0.20996, "G": 0.21382, "chunks": 215.0, "bytes_per_chunk": 4.763, "boundary_entropy": 0.10793, "router_noise": 0.0, "grad_norm": 0.6274, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.104, "alpha": 0.03, "t": 499.2}
{"event": "train", "step": 5050, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.34727, "bpb": 1.94371, "chunk_loss": 0.03057, "F": 0.21136, "G": 0.21716, "chunks": 216.44, "bytes_per_chunk": 4.731, "boundary_entropy": 0.11183, "router_noise": 0.0, "grad_norm": 0.6915, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.077, "alpha": 0.03, "t": 503.1}
{"event": "train", "step": 5100, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.3789, "bpb": 1.98933, "chunk_loss": 0.030482, "F": 0.20425, "G": 0.21603, "chunks": 209.16, "bytes_per_chunk": 4.896, "boundary_entropy": 0.1202, "router_noise": 0.0, "grad_norm": 0.7135, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.074, "alpha": 0.03, "t": 506.8}
{"event": "train", "step": 5150, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.37351, "bpb": 1.98156, "chunk_loss": 0.03053, "F": 0.20886, "G": 0.21764, "chunks": 213.88, "bytes_per_chunk": 4.788, "boundary_entropy": 0.1199, "router_noise": 0.0, "grad_norm": 0.6967, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.076, "alpha": 0.03, "t": 510.6}
{"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3477.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199793, "cps": 27.164062, "mean_p": 0.203263, "boundary_entropy": 0.103305, "bpb": 1.886856}, "tr": {"chunks": 3291.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181793, "cps": 25.710938, "mean_p": 0.207023, "boundary_entropy": 0.184176, "bpb": 2.033519}, "fi": {"chunks": 4265.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.226729, "cps": 33.320312, "mean_p": 0.226053, "boundary_entropy": 0.110602, "bpb": 2.130936}}, "bpb": {"en": 1.88686, "tr": 2.03352, "fi": 2.13094}, "gini_cpc": 0.049246, "gini_cps": 0.058854, "bpic": 5.142028, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9099064851653922, "fi": 1.134818955864909}, "t": 514.8}
{"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 1.90079, "tr": 2.04377, "fi": 2.13612}, "bytes_per_chunk": {"en": 4.771, "tr": 6.09, "fi": 4.694}, "bpb_mean": 2.02689, "frozen_inventory": false, "t": 515.3}
{"event": "train", "step": 5200, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.32545, "bpb": 1.91222, "chunk_loss": 0.030371, "F": 0.20251, "G": 0.20896, "chunks": 207.38, "bytes_per_chunk": 4.938, "boundary_entropy": 0.11075, "router_noise": 0.0, "grad_norm": 0.6898, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.101, "alpha": 0.03, "t": 515.7}
{"event": "train", "step": 5250, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.32266, "bpb": 1.9082, "chunk_loss": 0.030442, "F": 0.20514, "G": 0.21247, "chunks": 210.06, "bytes_per_chunk": 4.875, "boundary_entropy": 0.1116, "router_noise": 0.0, "grad_norm": 0.6713, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.076, "alpha": 0.03, "t": 519.5}
{"event": "train", "step": 5300, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.38213, "bpb": 1.99399, "chunk_loss": 0.030465, "F": 0.20544, "G": 0.21507, "chunks": 210.38, "bytes_per_chunk": 4.867, "boundary_entropy": 0.11445, "router_noise": 0.0, "grad_norm": 0.6988, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.076, "alpha": 0.03, "t": 523.3}
{"event": "train", "step": 5350, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.3121, "bpb": 1.89296, "chunk_loss": 0.030331, "F": 0.20074, "G": 0.20703, "chunks": 205.56, "bytes_per_chunk": 4.981, "boundary_entropy": 0.10937, "router_noise": 0.0, "grad_norm": 0.712, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.08, "alpha": 0.03, "t": 527.3}
{"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3496.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200885, "cps": 27.3125, "mean_p": 0.204531, "boundary_entropy": 0.101833, "bpb": 1.880691}, "tr": {"chunks": 3327.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.183782, "cps": 25.992188, "mean_p": 0.208402, "boundary_entropy": 0.183087, "bpb": 2.026455}, "fi": {"chunks": 4258.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.226357, "cps": 33.265625, "mean_p": 0.226811, "boundary_entropy": 0.111778, "bpb": 2.124528}}, "bpb": {"en": 1.88069, "tr": 2.02646, "fi": 2.12453}, "gini_cpc": 0.046452, "gini_cps": 0.056012, "bpic": 5.119755, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9148606448657447, "fi": 1.1267990441349194}, "t": 531.2}
{"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 1.89474, "tr": 2.03693, "fi": 2.13119}, "bytes_per_chunk": {"en": 4.749, "tr": 6.131, "fi": 4.678}, "bpb_mean": 2.02095, "frozen_inventory": false, "t": 531.6}
{"event": "train", "step": 5400, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.38111, "bpb": 1.99252, "chunk_loss": 0.030599, "F": 0.20935, "G": 0.21701, "chunks": 214.38, "bytes_per_chunk": 4.777, "boundary_entropy": 0.11085, "router_noise": 0.0, "grad_norm": 0.7214, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.096, "alpha": 0.03, "t": 532.1}
{"event": "train", "step": 5450, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.31944, "bpb": 1.90355, "chunk_loss": 0.030573, "F": 0.21072, "G": 0.21827, "chunks": 215.78, "bytes_per_chunk": 4.746, "boundary_entropy": 0.1142, "router_noise": 0.0, "grad_norm": 0.7033, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.079, "alpha": 0.03, "t": 536.1}
{"event": "train", "step": 5500, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.3058, "bpb": 1.88387, "chunk_loss": 0.03042, "F": 0.20459, "G": 0.21242, "chunks": 209.5, "bytes_per_chunk": 4.888, "boundary_entropy": 0.10983, "router_noise": 0.0, "grad_norm": 0.6752, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 540.1}
{"event": "train", "step": 5550, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.36312, "bpb": 1.96657, "chunk_loss": 0.030371, "F": 0.2001, "G": 0.21023, "chunks": 204.91, "bytes_per_chunk": 4.997, "boundary_entropy": 0.11776, "router_noise": 0.0, "grad_norm": 0.7618, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.079, "alpha": 0.03, "t": 544.1}
{"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3479.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199908, "cps": 27.179688, "mean_p": 0.203028, "boundary_entropy": 0.103509, "bpb": 1.877123}, "tr": {"chunks": 3330.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.183947, "cps": 26.015625, "mean_p": 0.20916, "boundary_entropy": 0.185308, "bpb": 2.019522}, "fi": {"chunks": 4232.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.224975, "cps": 33.0625, "mean_p": 0.226128, "boundary_entropy": 0.111384, "bpb": 2.115882}}, "bpb": {"en": 1.87712, "tr": 2.01952, "fi": 2.11588}, "gini_cpc": 0.044925, "gini_cps": 0.054464, "bpic": 5.138303, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9201600493182499, "fi": 1.1253910760751389}, "t": 548.8}
{"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 1.89167, "tr": 2.02934, "fi": 2.12244}, "bytes_per_chunk": {"en": 4.778, "tr": 5.978, "fi": 4.698}, "bpb_mean": 2.01448, "frozen_inventory": false, "t": 549.4}
{"event": "train", "step": 5600, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.35598, "bpb": 1.95627, "chunk_loss": 0.030486, "F": 0.20621, "G": 0.21408, "chunks": 211.16, "bytes_per_chunk": 4.849, "boundary_entropy": 0.11663, "router_noise": 0.0, "grad_norm": 0.7807, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.114, "alpha": 0.03, "t": 549.8}
{"event": "train", "step": 5650, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.33498, "bpb": 1.92596, "chunk_loss": 0.03053, "F": 0.21066, "G": 0.21503, "chunks": 215.72, "bytes_per_chunk": 4.747, "boundary_entropy": 0.10135, "router_noise": 0.0, "grad_norm": 0.6969, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 554.1}
{"event": "train", "step": 5700, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.34513, "bpb": 1.94062, "chunk_loss": 0.030373, "F": 0.2016, "G": 0.20962, "chunks": 206.44, "bytes_per_chunk": 4.96, "boundary_entropy": 0.11173, "router_noise": 0.0, "grad_norm": 0.7446, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 558.1}
{"event": "train", "step": 5750, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.3989, "bpb": 2.01818, "chunk_loss": 0.030626, "F": 0.21216, "G": 0.21859, "chunks": 217.25, "bytes_per_chunk": 4.713, "boundary_entropy": 0.10662, "router_noise": 0.0, "grad_norm": 0.7591, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.076, "alpha": 0.03, "t": 561.9}
{"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3493.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200713, "cps": 27.289062, "mean_p": 0.203357, "boundary_entropy": 0.098629, "bpb": 1.872843}, "tr": {"chunks": 3325.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.183671, "cps": 25.976562, "mean_p": 0.207437, "boundary_entropy": 0.182037, "bpb": 2.016203}, "fi": {"chunks": 4260.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.226463, "cps": 33.28125, "mean_p": 0.226536, "boundary_entropy": 0.106365, "bpb": 2.112238}}, "bpb": {"en": 1.87284, "tr": 2.0162, "fi": 2.11224}, "gini_cpc": 0.046702, "gini_cps": 0.056268, "bpic": 5.121141, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9150959489547509, "fi": 1.1282965240915697}, "t": 566.2}
{"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 1.88649, "tr": 2.02559, "fi": 2.11862}, "bytes_per_chunk": {"en": 4.755, "tr": 6.042, "fi": 4.67}, "bpb_mean": 2.01023, "frozen_inventory": false, "t": 566.7}
{"event": "train", "step": 5800, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.32096, "bpb": 1.90575, "chunk_loss": 0.030525, "F": 0.20697, "G": 0.21489, "chunks": 211.94, "bytes_per_chunk": 4.832, "boundary_entropy": 0.1142, "router_noise": 0.0, "grad_norm": 0.7032, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.102, "alpha": 0.03, "t": 567.1}
{"event": "train", "step": 5850, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.34479, "bpb": 1.94012, "chunk_loss": 0.030488, "F": 0.20911, "G": 0.21398, "chunks": 214.12, "bytes_per_chunk": 4.782, "boundary_entropy": 0.10094, "router_noise": 0.0, "grad_norm": 0.7427, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.078, "alpha": 0.03, "t": 571.0}
{"event": "train", "step": 5900, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.27617, "bpb": 1.84113, "chunk_loss": 0.030364, "F": 0.20166, "G": 0.20925, "chunks": 206.5, "bytes_per_chunk": 4.959, "boundary_entropy": 0.10898, "router_noise": 0.0, "grad_norm": 0.696, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.075, "alpha": 0.03, "t": 574.7}
{"event": "train", "step": 5950, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.38167, "bpb": 1.99332, "chunk_loss": 0.030504, "F": 0.20755, "G": 0.21463, "chunks": 212.53, "bytes_per_chunk": 4.818, "boundary_entropy": 0.11068, "router_noise": 0.0, "grad_norm": 0.7331, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 578.9}
{"event": "train", "step": 5999, "arch": "hnet_ffn", "phase": 3, "objective": "global", "lm_loss": 1.38224, "bpb": 1.99415, "chunk_loss": 0.030561, "F": 0.20959, "G": 0.2155, "chunks": 214.62, "bytes_per_chunk": 4.771, "boundary_entropy": 0.10757, "router_noise": 0.0, "grad_norm": 0.7062, "lr": 3e-05, "router_trainable": true, "sec_per_step": 0.078, "alpha": 0.03, "t": 582.8}
{"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3466.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199161, "cps": 27.078125, "mean_p": 0.202963, "boundary_entropy": 0.100151, "bpb": 1.871516}, "tr": {"chunks": 3309.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.182787, "cps": 25.851562, "mean_p": 0.208028, "boundary_entropy": 0.182198, "bpb": 2.011767}, "fi": {"chunks": 4242.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.225506, "cps": 33.140625, "mean_p": 0.22588, "boundary_entropy": 0.107201, "bpb": 2.110077}}, "bpb": {"en": 1.87152, "tr": 2.01177, "fi": 2.11008}, "gini_cpc": 0.046883, "gini_cps": 0.056458, "bpic": 5.149496, "worst_lang": "fi", "cpc_parity_vs_en": {"en": 1.0, "tr": 0.9177867373587294, "fi": 1.1322813200941377}, "t": 583.2}
{"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 1.88541, "tr": 2.02176, "fi": 2.11664}, "bytes_per_chunk": {"en": 4.79, "tr": 6.076, "fi": 4.687}, "bpb_mean": 2.00793, "frozen_inventory": false, "t": 583.7}
{"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/runs/ali_hnet-ffn_s2/probe/reps/th0.30", "/local/scratch/sas245/beetle-hnet/runs/ali_hnet-ffn_s2/probe/reps/th0.40", "/local/scratch/sas245/beetle-hnet/runs/ali_hnet-ffn_s2/probe/reps/th0.50", "/local/scratch/sas245/beetle-hnet/runs/ali_hnet-ffn_s2/probe/reps/th0.60", "/local/scratch/sas245/beetle-hnet/runs/ali_hnet-ffn_s2/probe/reps/th0.70"], "t": 584.8}
{"event": "final", "run": "ali_hnet-ffn_s2", "arch": "hnet_ffn", "objective": "ratio", "condition": "A", "step": 6000, "params": 22785033, "seed": 2, "per_lang": {"en": {"cpc": 0.19916106418433605, "cps": 27.078125, "bpb": 1.8920453991596864, "bytes_per_chunk": 4.746277033938966}, "tr": {"cpc": 0.18278738330663427, "cps": 25.8515625, "bpb": 2.0128654149185237, "bytes_per_chunk": 6.104606213031531}, "fi": {"cpc": 0.2255063526659933, "cps": 33.140625, "bpb": 2.1365384401236085, "bytes_per_chunk": 4.661124989998311}}, "bpb_mean": 2.013816418067273, "gini_cpc": 0.04688301565627673, "gini_cps": 0.056458200962149485, "bpic": 5.149496233094308, "mean_boundary_entropy": 0.1075679287314415, "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": 586.1055707931519, "ffn_params": 268041, "matched_to_ali_params": 268096, "t": 586.1}