{ "screen": "SP-2-SPINE-LM", "seed": 0, "steps": 3000, "batch": 32, "block": 512, "lrs": [ 0.001 ], "text_mb": 24.0, "d_spine": 384, "layers_spine": 8, "n_spine": 12018432, "tf_mult": 2.0, "d_transformer": 624, "layers_transformer": 5, "n_transformer": 24045216, "device": "cuda", "thresholds": { "pass": 1.05, "concern": 1.15 }, "corpora": { "text": { "spine_lr0.001": { "best_bpb": 1.6573128700256348, "history": [ { "step": 0, "train_bits": 8.097306429979653, "eval_bpb": 5.986931800842285, "best": 5.986931800842285, "kBps": 11.0 }, { "step": 250, "train_bits": 2.1230272788424283, "eval_bpb": 2.182271718978882, "best": 2.182271718978882, "kBps": 21.7 }, { "step": 500, "train_bits": 1.9552195068293143, "eval_bpb": 2.0323281288146973, "best": 2.0323281288146973, "kBps": 21.7 }, { "step": 750, "train_bits": 1.7949351158781939, "eval_bpb": 1.9211971759796143, "best": 1.9211971759796143, "kBps": 21.7 }, { "step": 1000, "train_bits": 1.7489086029063365, "eval_bpb": 1.850945234298706, "best": 1.850945234298706, "kBps": 21.7 }, { "step": 1250, "train_bits": 1.6664138567496238, "eval_bpb": 1.812852382659912, "best": 1.812852382659912, "kBps": 21.7 }, { "step": 1500, "train_bits": 1.6738565779483627, "eval_bpb": 1.7703351974487305, "best": 1.7703351974487305, "kBps": 21.7 }, { "step": 1750, "train_bits": 1.6136710772962721, "eval_bpb": 1.7218865156173706, "best": 1.7218865156173706, "kBps": 21.7 }, { "step": 2000, "train_bits": 1.584476162399056, "eval_bpb": 1.6893564462661743, "best": 1.6893564462661743, "kBps": 21.7 }, { "step": 2250, "train_bits": 1.5259576895817784, "eval_bpb": 1.6723499298095703, "best": 1.6723499298095703, "kBps": 21.7 }, { "step": 2500, "train_bits": 1.4685067130784064, "eval_bpb": 1.662462830543518, "best": 1.662462830543518, "kBps": 21.7 }, { "step": 2750, "train_bits": 1.5642197018151327, "eval_bpb": 1.6594675779342651, "best": 1.6594675779342651, "kBps": 21.7 }, { "step": 2999, "train_bits": 1.4700798383858484, "eval_bpb": 1.6573128700256348, "best": 1.6573128700256348, "kBps": 21.7 } ], "wall_s": 2268.9 }, "tf_lr0.001": { "best_bpb": 1.870753288269043, "history": [ { "step": 0, "train_bits": 8.26989102012048, "eval_bpb": 6.739991664886475, "best": 6.739991664886475, "kBps": 11.4 }, { "step": 250, "train_bits": 3.5156865472026206, "eval_bpb": 3.5838325023651123, "best": 3.5838325023651123, "kBps": 18.5 }, { "step": 500, "train_bits": 3.118333703065567, "eval_bpb": 3.109508991241455, "best": 3.109508991241455, "kBps": 18.5 }, { "step": 750, "train_bits": 2.671179498752868, "eval_bpb": 2.7187116146087646, "best": 2.7187116146087646, "kBps": 18.6 }, { "step": 1000, "train_bits": 2.3104986872419415, "eval_bpb": 2.379302978515625, "best": 2.379302978515625, "kBps": 18.6 }, { "step": 1250, "train_bits": 2.0811331650055025, "eval_bpb": 2.1693010330200195, "best": 2.1693010330200195, "kBps": 18.6 }, { "step": 1500, "train_bits": 1.9972204218924259, "eval_bpb": 2.0676300525665283, "best": 2.0676300525665283, "kBps": 18.6 }, { "step": 1750, "train_bits": 1.9581717610140252, "eval_bpb": 1.989850640296936, "best": 1.989850640296936, "kBps": 18.6 }, { "step": 2000, "train_bits": 1.8832767562162893, "eval_bpb": 1.9484683275222778, "best": 1.9484683275222778, "kBps": 18.6 }, { "step": 2250, "train_bits": 1.836628181996261, "eval_bpb": 1.9090824127197266, "best": 1.9090824127197266, "kBps": 18.6 }, { "step": 2500, "train_bits": 1.8263085350138892, "eval_bpb": 1.8851382732391357, "best": 1.8851382732391357, "kBps": 18.6 }, { "step": 2750, "train_bits": 1.8128861490273014, "eval_bpb": 1.8729664087295532, "best": 1.8729664087295532, "kBps": 18.6 }, { "step": 2999, "train_bits": 1.7961690913981345, "eval_bpb": 1.870753288269043, "best": 1.870753288269043, "kBps": 18.6 } ], "wall_s": 2642.3 }, "spine_best": 1.6573128700256348, "tf_best": 1.870753288269043, "ratio": 0.8859066988780232, "verdict": "PASS" } }, "partial": false, "verdict": "PASS" }