| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.6148445799270241, |
| "eval_steps": 1024, |
| "global_step": 13312, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.011823934229365849, |
| "grad_norm": 2.166717529296875, |
| "learning_rate": 0.000498046875, |
| "loss": 6.863234996795654, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.023647868458731697, |
| "grad_norm": 0.9976925849914551, |
| "learning_rate": 0.000998046875, |
| "loss": 1.0293713808059692, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.03547180268809755, |
| "grad_norm": 0.9951309561729431, |
| "learning_rate": 0.000999640996023194, |
| "loss": 0.5387133359909058, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.047295736917463395, |
| "grad_norm": 1.1936287879943848, |
| "learning_rate": 0.0009985588674043958, |
| "loss": 0.4395764172077179, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.047295736917463395, |
| "eval_bleu": 0.630805532752935, |
| "eval_ce_loss": 1.5585694417833738, |
| "eval_cos_loss": 0.12051772464015713, |
| "eval_loss": 0.41065708449169924, |
| "eval_mse_loss": 0.22680107500727317, |
| "eval_rec_loss": 0.015947288232876763, |
| "flow/cos_sim": 0.8794823028997744, |
| "flow/improvement_ratio": 0.9763019112147153, |
| "flow/mag_ratio_mean": 0.8738831932141901, |
| "flow/mag_ratio_std": 0.17382950611310463, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.047295736917463395, |
| "eval_bleu": 0.630805532752935, |
| "eval_ce_loss": 1.5585694417833738, |
| "eval_cos_loss": 0.12051772464015713, |
| "eval_loss": 0.41065708449169924, |
| "eval_mse_loss": 0.22680107500727317, |
| "eval_rec_loss": 0.015947288232876763, |
| "eval_runtime": 149.5036, |
| "eval_samples_per_second": 187.24, |
| "eval_steps_per_second": 2.93, |
| "flow/cos_sim": 0.8794823028997744, |
| "flow/improvement_ratio": 0.9763019112147153, |
| "flow/mag_ratio_mean": 0.8738831932141901, |
| "flow/mag_ratio_std": 0.17382950611310463, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.05911967114682925, |
| "grad_norm": 1.413312315940857, |
| "learning_rate": 0.0009967551747861387, |
| "loss": 0.40079358220100403, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.0709436053761951, |
| "grad_norm": 0.4399195909500122, |
| "learning_rate": 0.000994232528651847, |
| "loss": 0.38004180788993835, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.08276753960556095, |
| "grad_norm": 1.061546802520752, |
| "learning_rate": 0.0009909945800260092, |
| "loss": 0.3571281135082245, |
| "step": 1792 |
| }, |
| { |
| "epoch": 0.09459147383492679, |
| "grad_norm": 0.4095996916294098, |
| "learning_rate": 0.0009870460151900522, |
| "loss": 0.34413549304008484, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.09459147383492679, |
| "eval_bleu": 0.6370628963183373, |
| "eval_ce_loss": 1.4389138337411838, |
| "eval_cos_loss": 0.09142084970866164, |
| "eval_loss": 0.33612148960431415, |
| "eval_mse_loss": 0.17856517835566987, |
| "eval_rec_loss": 0.004522840211489004, |
| "flow/cos_sim": 0.9085791546460156, |
| "flow/improvement_ratio": 0.9776561622205935, |
| "flow/mag_ratio_mean": 0.9092166116248527, |
| "flow/mag_ratio_std": 0.14505240416417928, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.09459147383492679, |
| "eval_bleu": 0.6370628963183373, |
| "eval_ce_loss": 1.4389138337411838, |
| "eval_cos_loss": 0.09142084970866164, |
| "eval_loss": 0.33612148960431415, |
| "eval_mse_loss": 0.17856517835566987, |
| "eval_rec_loss": 0.004522840211489004, |
| "eval_runtime": 144.0023, |
| "eval_samples_per_second": 194.393, |
| "eval_steps_per_second": 3.042, |
| "flow/cos_sim": 0.9085791546460156, |
| "flow/improvement_ratio": 0.9776561622205935, |
| "flow/mag_ratio_mean": 0.9092166116248527, |
| "flow/mag_ratio_std": 0.14505240416417928, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.10641540806429264, |
| "grad_norm": 1.3680005073547363, |
| "learning_rate": 0.0009823925488998885, |
| "loss": 0.3346008360385895, |
| "step": 2304 |
| }, |
| { |
| "epoch": 0.1182393422936585, |
| "grad_norm": 0.9834803938865662, |
| "learning_rate": 0.0009770409161149525, |
| "loss": 0.32358497381210327, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.13006327652302435, |
| "grad_norm": 0.7012990117073059, |
| "learning_rate": 0.0009709988622506973, |
| "loss": 0.31395941972732544, |
| "step": 2816 |
| }, |
| { |
| "epoch": 0.1418872107523902, |
| "grad_norm": 0.4390225112438202, |
| "learning_rate": 0.000964275131968659, |
| "loss": 0.30854108929634094, |
| "step": 3072 |
| }, |
| { |
| "epoch": 0.1418872107523902, |
| "eval_bleu": 0.6707740050213097, |
| "eval_ce_loss": 1.2450400457534616, |
| "eval_cos_loss": 0.08531694868622033, |
| "eval_loss": 0.3036313052523082, |
| "eval_mse_loss": 0.16824311817481638, |
| "eval_rec_loss": 0.0023524861730428916, |
| "flow/cos_sim": 0.9146830740312463, |
| "flow/improvement_ratio": 0.9760218352487643, |
| "flow/mag_ratio_mean": 0.9162033348595171, |
| "flow/mag_ratio_std": 0.12317379108166586, |
| "step": 3072 |
| }, |
| { |
| "epoch": 0.1418872107523902, |
| "eval_bleu": 0.6707740050213097, |
| "eval_ce_loss": 1.2450400457534616, |
| "eval_cos_loss": 0.08531694868622033, |
| "eval_loss": 0.3036313052523082, |
| "eval_mse_loss": 0.16824311817481638, |
| "eval_rec_loss": 0.0023524861730428916, |
| "eval_runtime": 145.6816, |
| "eval_samples_per_second": 192.152, |
| "eval_steps_per_second": 3.007, |
| "flow/cos_sim": 0.9146830740312463, |
| "flow/improvement_ratio": 0.9760218352487643, |
| "flow/mag_ratio_mean": 0.9162033348595171, |
| "flow/mag_ratio_std": 0.12317379108166586, |
| "step": 3072 |
| }, |
| { |
| "epoch": 0.15371114498175603, |
| "grad_norm": 0.6193675398826599, |
| "learning_rate": 0.0009568794565203123, |
| "loss": 0.30133724212646484, |
| "step": 3328 |
| }, |
| { |
| "epoch": 0.1655350792111219, |
| "grad_norm": 0.5207021236419678, |
| "learning_rate": 0.0009488225396630347, |
| "loss": 0.29369795322418213, |
| "step": 3584 |
| }, |
| { |
| "epoch": 0.17735901344048774, |
| "grad_norm": 0.5521863102912903, |
| "learning_rate": 0.0009401160421685646, |
| "loss": 0.2897779941558838, |
| "step": 3840 |
| }, |
| { |
| "epoch": 0.18918294766985358, |
| "grad_norm": 0.4839329421520233, |
| "learning_rate": 0.0009307725649463714, |
| "loss": 0.2864342927932739, |
| "step": 4096 |
| }, |
| { |
| "epoch": 0.18918294766985358, |
| "eval_bleu": 0.6704121655421459, |
| "eval_ce_loss": 1.2046921730858007, |
| "eval_cos_loss": 0.07965970783575213, |
| "eval_loss": 0.2842630836857508, |
| "eval_mse_loss": 0.1541883000534147, |
| "eval_rec_loss": 0.0016395924527162992, |
| "flow/cos_sim": 0.9203403159363629, |
| "flow/improvement_ratio": 0.9744544083669305, |
| "flow/mag_ratio_mean": 0.9212693351588838, |
| "flow/mag_ratio_std": 0.10148854444832563, |
| "step": 4096 |
| }, |
| { |
| "epoch": 0.18918294766985358, |
| "eval_bleu": 0.6704121655421459, |
| "eval_ce_loss": 1.2046921730858007, |
| "eval_cos_loss": 0.07965970783575213, |
| "eval_loss": 0.2842630836857508, |
| "eval_mse_loss": 0.1541883000534147, |
| "eval_rec_loss": 0.0016395924527162992, |
| "eval_runtime": 144.2359, |
| "eval_samples_per_second": 194.078, |
| "eval_steps_per_second": 3.037, |
| "flow/cos_sim": 0.9203403159363629, |
| "flow/improvement_ratio": 0.9744544083669305, |
| "flow/mag_ratio_mean": 0.9212693351588838, |
| "flow/mag_ratio_std": 0.10148854444832563, |
| "step": 4096 |
| }, |
| { |
| "epoch": 0.20100688189921945, |
| "grad_norm": 0.9023747444152832, |
| "learning_rate": 0.0009208056308063659, |
| "loss": 0.30037108063697815, |
| "step": 4352 |
| }, |
| { |
| "epoch": 0.2128308161285853, |
| "grad_norm": 0.8910952806472778, |
| "learning_rate": 0.0009102296648873445, |
| "loss": 0.28084370493888855, |
| "step": 4608 |
| }, |
| { |
| "epoch": 0.22465475035795113, |
| "grad_norm": 0.43350881338119507, |
| "learning_rate": 0.0008990599737794927, |
| "loss": 0.276326060295105, |
| "step": 4864 |
| }, |
| { |
| "epoch": 0.236478684587317, |
| "grad_norm": 0.4241744577884674, |
| "learning_rate": 0.0008873127233711644, |
| "loss": 0.2720130980014801, |
| "step": 5120 |
| }, |
| { |
| "epoch": 0.236478684587317, |
| "eval_bleu": 0.6884395396484843, |
| "eval_ce_loss": 1.1380460211947627, |
| "eval_cos_loss": 0.0743916420410621, |
| "eval_loss": 0.26738106996909666, |
| "eval_mse_loss": 0.14492133019925796, |
| "eval_rec_loss": 0.0012159729793058606, |
| "flow/cos_sim": 0.9256083738041795, |
| "flow/improvement_ratio": 0.9722109245110865, |
| "flow/mag_ratio_mean": 0.9361066815515632, |
| "flow/mag_ratio_std": 0.08903367672812994, |
| "step": 5120 |
| }, |
| { |
| "epoch": 0.236478684587317, |
| "eval_bleu": 0.6884395396484843, |
| "eval_ce_loss": 1.1380460211947627, |
| "eval_cos_loss": 0.0743916420410621, |
| "eval_loss": 0.26738106996909666, |
| "eval_mse_loss": 0.14492133019925796, |
| "eval_rec_loss": 0.0012159729793058606, |
| "eval_runtime": 145.1812, |
| "eval_samples_per_second": 192.814, |
| "eval_steps_per_second": 3.017, |
| "flow/cos_sim": 0.9256083738041795, |
| "flow/improvement_ratio": 0.9722109245110865, |
| "flow/mag_ratio_mean": 0.9361066815515632, |
| "flow/mag_ratio_std": 0.08903367672812994, |
| "step": 5120 |
| }, |
| { |
| "epoch": 0.24830261881668284, |
| "grad_norm": 0.7429539561271667, |
| "learning_rate": 0.0008750049154520011, |
| "loss": 0.2687661945819855, |
| "step": 5376 |
| }, |
| { |
| "epoch": 0.2601265530460487, |
| "grad_norm": 0.4944668114185333, |
| "learning_rate": 0.0008621543631062487, |
| "loss": 0.26752030849456787, |
| "step": 5632 |
| }, |
| { |
| "epoch": 0.27195048727541454, |
| "grad_norm": 0.4112600088119507, |
| "learning_rate": 0.0008487796649318904, |
| "loss": 0.26332396268844604, |
| "step": 5888 |
| }, |
| { |
| "epoch": 0.2837744215047804, |
| "grad_norm": 0.5803443789482117, |
| "learning_rate": 0.0008349001781229053, |
| "loss": 0.2650764286518097, |
| "step": 6144 |
| }, |
| { |
| "epoch": 0.2837744215047804, |
| "eval_bleu": 0.6901447099765771, |
| "eval_ce_loss": 1.1182073426845411, |
| "eval_cos_loss": 0.07293294577701995, |
| "eval_loss": 0.2625204250216484, |
| "eval_mse_loss": 0.14227469792785166, |
| "eval_rec_loss": 0.0011316965778382733, |
| "flow/cos_sim": 0.9270670652117359, |
| "flow/improvement_ratio": 0.972748759402532, |
| "flow/mag_ratio_mean": 0.9329337213681713, |
| "flow/mag_ratio_std": 0.08477971859310315, |
| "step": 6144 |
| }, |
| { |
| "epoch": 0.2837744215047804, |
| "eval_bleu": 0.6901447099765771, |
| "eval_ce_loss": 1.1182073426845411, |
| "eval_cos_loss": 0.07293294577701995, |
| "eval_loss": 0.2625204250216484, |
| "eval_mse_loss": 0.14227469792785166, |
| "eval_rec_loss": 0.0011316965778382733, |
| "eval_runtime": 143.6154, |
| "eval_samples_per_second": 194.916, |
| "eval_steps_per_second": 3.05, |
| "flow/cos_sim": 0.9270670652117359, |
| "flow/improvement_ratio": 0.972748759402532, |
| "flow/mag_ratio_mean": 0.9329337213681713, |
| "flow/mag_ratio_std": 0.08477971859310315, |
| "step": 6144 |
| }, |
| { |
| "epoch": 0.2955983557341462, |
| "grad_norm": 0.5521038770675659, |
| "learning_rate": 0.0008205359904536107, |
| "loss": 0.2589625120162964, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.30742228996351206, |
| "grad_norm": 0.5204539895057678, |
| "learning_rate": 0.0008057078912056363, |
| "loss": 0.2580261528491974, |
| "step": 6656 |
| }, |
| { |
| "epoch": 0.3192462241928779, |
| "grad_norm": 0.5325313806533813, |
| "learning_rate": 0.0007904373410796086, |
| "loss": 0.25551143288612366, |
| "step": 6912 |
| }, |
| { |
| "epoch": 0.3310701584222438, |
| "grad_norm": 0.3639472424983978, |
| "learning_rate": 0.0007747464411350876, |
| "loss": 0.25460946559906006, |
| "step": 7168 |
| }, |
| { |
| "epoch": 0.3310701584222438, |
| "eval_bleu": 0.7071091482680488, |
| "eval_ce_loss": 1.0378137462215336, |
| "eval_cos_loss": 0.07100786366521224, |
| "eval_loss": 0.2512795734283042, |
| "eval_mse_loss": 0.13953843853993503, |
| "eval_rec_loss": 0.0008589719193517103, |
| "flow/cos_sim": 0.9289921618759904, |
| "flow/improvement_ratio": 0.9735576084487514, |
| "flow/mag_ratio_mean": 0.9378764381691745, |
| "flow/mag_ratio_std": 0.08363016963549401, |
| "step": 7168 |
| }, |
| { |
| "epoch": 0.3310701584222438, |
| "eval_bleu": 0.7071091482680488, |
| "eval_ce_loss": 1.0378137462215336, |
| "eval_cos_loss": 0.07100786366521224, |
| "eval_loss": 0.2512795734283042, |
| "eval_mse_loss": 0.13953843853993503, |
| "eval_rec_loss": 0.0008589719193517103, |
| "eval_runtime": 145.8938, |
| "eval_samples_per_second": 191.872, |
| "eval_steps_per_second": 3.002, |
| "flow/cos_sim": 0.9289921618759904, |
| "flow/improvement_ratio": 0.9735576084487514, |
| "flow/mag_ratio_mean": 0.9378764381691745, |
| "flow/mag_ratio_std": 0.08363016963549401, |
| "step": 7168 |
| }, |
| { |
| "epoch": 0.34289409265160964, |
| "grad_norm": 0.26728010177612305, |
| "learning_rate": 0.000758657900803716, |
| "loss": 0.25558242201805115, |
| "step": 7424 |
| }, |
| { |
| "epoch": 0.3547180268809755, |
| "grad_norm": 0.3367791175842285, |
| "learning_rate": 0.000742195005021869, |
| "loss": 0.2547113597393036, |
| "step": 7680 |
| }, |
| { |
| "epoch": 0.3665419611103413, |
| "grad_norm": 0.6501409411430359, |
| "learning_rate": 0.0007253815805303786, |
| "loss": 0.2516462504863739, |
| "step": 7936 |
| }, |
| { |
| "epoch": 0.37836589533970716, |
| "grad_norm": 0.41470277309417725, |
| "learning_rate": 0.0007082419613901028, |
| "loss": 0.25756070017814636, |
| "step": 8192 |
| }, |
| { |
| "epoch": 0.37836589533970716, |
| "eval_bleu": 0.7183146382916741, |
| "eval_ce_loss": 1.0048845314408001, |
| "eval_cos_loss": 0.07200061048478841, |
| "eval_loss": 0.2508213480245577, |
| "eval_mse_loss": 0.14229333698409333, |
| "eval_rec_loss": 0.0008394962100133504, |
| "flow/cos_sim": 0.9279994152180137, |
| "flow/improvement_ratio": 0.9741122624645494, |
| "flow/mag_ratio_mean": 0.9306073022759669, |
| "flow/mag_ratio_std": 0.08248136017352478, |
| "step": 8192 |
| }, |
| { |
| "epoch": 0.37836589533970716, |
| "eval_bleu": 0.7183146382916741, |
| "eval_ce_loss": 1.0048845314408001, |
| "eval_cos_loss": 0.07200061048478841, |
| "eval_loss": 0.2508213480245577, |
| "eval_mse_loss": 0.14229333698409333, |
| "eval_rec_loss": 0.0008394962100133504, |
| "eval_runtime": 144.5073, |
| "eval_samples_per_second": 193.713, |
| "eval_steps_per_second": 3.031, |
| "flow/cos_sim": 0.9279994152180137, |
| "flow/improvement_ratio": 0.9741122624645494, |
| "flow/mag_ratio_mean": 0.9306073022759669, |
| "flow/mag_ratio_std": 0.08248136017352478, |
| "step": 8192 |
| }, |
| { |
| "epoch": 0.390189829569073, |
| "grad_norm": 0.5503519773483276, |
| "learning_rate": 0.0006908009537632514, |
| "loss": 0.24980439245700836, |
| "step": 8448 |
| }, |
| { |
| "epoch": 0.4020137637984389, |
| "grad_norm": 0.5455402135848999, |
| "learning_rate": 0.0006730838000114403, |
| "loss": 0.25091099739074707, |
| "step": 8704 |
| }, |
| { |
| "epoch": 0.41383769802780473, |
| "grad_norm": 0.29903262853622437, |
| "learning_rate": 0.0006551161421624341, |
| "loss": 0.24810341000556946, |
| "step": 8960 |
| }, |
| { |
| "epoch": 0.4256616322571706, |
| "grad_norm": 0.5439685583114624, |
| "learning_rate": 0.0006369239847984517, |
| "loss": 0.24486252665519714, |
| "step": 9216 |
| }, |
| { |
| "epoch": 0.4256616322571706, |
| "eval_bleu": 0.7059557795131848, |
| "eval_ce_loss": 1.0441748459045201, |
| "eval_cos_loss": 0.06742798249539175, |
| "eval_loss": 0.24551006137780404, |
| "eval_mse_loss": 0.13372647651730607, |
| "eval_rec_loss": 0.0006233008009167367, |
| "flow/cos_sim": 0.9325720338516583, |
| "flow/improvement_ratio": 0.9731077969618583, |
| "flow/mag_ratio_mean": 0.934824386688128, |
| "flow/mag_ratio_std": 0.079195149838244, |
| "step": 9216 |
| }, |
| { |
| "epoch": 0.4256616322571706, |
| "eval_bleu": 0.7059557795131848, |
| "eval_ce_loss": 1.0441748459045201, |
| "eval_cos_loss": 0.06742798249539175, |
| "eval_loss": 0.24551006137780404, |
| "eval_mse_loss": 0.13372647651730607, |
| "eval_rec_loss": 0.0006233008009167367, |
| "eval_runtime": 143.7536, |
| "eval_samples_per_second": 194.729, |
| "eval_steps_per_second": 3.047, |
| "flow/cos_sim": 0.9325720338516583, |
| "flow/improvement_ratio": 0.9731077969618583, |
| "flow/mag_ratio_mean": 0.934824386688128, |
| "flow/mag_ratio_std": 0.079195149838244, |
| "step": 9216 |
| }, |
| { |
| "epoch": 0.4374855664865364, |
| "grad_norm": 0.6247985363006592, |
| "learning_rate": 0.0006185336574197479, |
| "loss": 0.24821333587169647, |
| "step": 9472 |
| }, |
| { |
| "epoch": 0.44930950071590225, |
| "grad_norm": 0.5992924571037292, |
| "learning_rate": 0.0005999717763379407, |
| "loss": 0.2463277131319046, |
| "step": 9728 |
| }, |
| { |
| "epoch": 0.4611334349452681, |
| "grad_norm": 0.5051174759864807, |
| "learning_rate": 0.0005812652061542363, |
| "loss": 0.24300755560398102, |
| "step": 9984 |
| }, |
| { |
| "epoch": 0.472957369174634, |
| "grad_norm": 0.40211889147758484, |
| "learning_rate": 0.0005624410208783071, |
| "loss": 0.24105887115001678, |
| "step": 10240 |
| }, |
| { |
| "epoch": 0.472957369174634, |
| "eval_bleu": 0.7120039792514781, |
| "eval_ce_loss": 1.005759668472695, |
| "eval_cos_loss": 0.06631913622118295, |
| "eval_loss": 0.2399323729619588, |
| "eval_mse_loss": 0.1321649712820848, |
| "eval_rec_loss": 0.0005595201312656808, |
| "flow/cos_sim": 0.9336808790627136, |
| "flow/improvement_ratio": 0.9762771404497156, |
| "flow/mag_ratio_mean": 0.926664409572131, |
| "flow/mag_ratio_std": 0.0785302072763443, |
| "step": 10240 |
| }, |
| { |
| "epoch": 0.472957369174634, |
| "eval_bleu": 0.7120039792514781, |
| "eval_ce_loss": 1.005759668472695, |
| "eval_cos_loss": 0.06631913622118295, |
| "eval_loss": 0.2399323729619588, |
| "eval_mse_loss": 0.1321649712820848, |
| "eval_rec_loss": 0.0005595201312656808, |
| "eval_runtime": 145.1926, |
| "eval_samples_per_second": 192.799, |
| "eval_steps_per_second": 3.017, |
| "flow/cos_sim": 0.9336808790627136, |
| "flow/improvement_ratio": 0.9762771404497156, |
| "flow/mag_ratio_mean": 0.926664409572131, |
| "flow/mag_ratio_std": 0.0785302072763443, |
| "step": 10240 |
| }, |
| { |
| "epoch": 0.48478130340399983, |
| "grad_norm": 0.5163204073905945, |
| "learning_rate": 0.0005435264647440881, |
| "loss": 0.2431734949350357, |
| "step": 10496 |
| }, |
| { |
| "epoch": 0.49660523763336567, |
| "grad_norm": 0.3941446840763092, |
| "learning_rate": 0.000524548912779213, |
| "loss": 0.23893995583057404, |
| "step": 10752 |
| }, |
| { |
| "epoch": 0.5084291718627315, |
| "grad_norm": 0.600309431552887, |
| "learning_rate": 0.0005055358311851499, |
| "loss": 0.24141335487365723, |
| "step": 11008 |
| }, |
| { |
| "epoch": 0.5202531060920974, |
| "grad_norm": 0.2635146975517273, |
| "learning_rate": 0.0004865147375853812, |
| "loss": 0.23994745314121246, |
| "step": 11264 |
| }, |
| { |
| "epoch": 0.5202531060920974, |
| "eval_bleu": 0.721584420569106, |
| "eval_ce_loss": 0.9708276380142665, |
| "eval_cos_loss": 0.06598166635761794, |
| "eval_loss": 0.236370724664159, |
| "eval_mse_loss": 0.13219797124737473, |
| "eval_rec_loss": 0.0004918215776982103, |
| "flow/cos_sim": 0.9340183556624199, |
| "flow/improvement_ratio": 0.9756744954411842, |
| "flow/mag_ratio_mean": 0.9309794351662675, |
| "flow/mag_ratio_std": 0.07776250285372886, |
| "step": 11264 |
| }, |
| { |
| "epoch": 0.5202531060920974, |
| "eval_bleu": 0.721584420569106, |
| "eval_ce_loss": 0.9708276380142665, |
| "eval_cos_loss": 0.06598166635761794, |
| "eval_loss": 0.236370724664159, |
| "eval_mse_loss": 0.13219797124737473, |
| "eval_rec_loss": 0.0004918215776982103, |
| "eval_runtime": 145.3935, |
| "eval_samples_per_second": 192.533, |
| "eval_steps_per_second": 3.013, |
| "flow/cos_sim": 0.9340183556624199, |
| "flow/improvement_ratio": 0.9756744954411842, |
| "flow/mag_ratio_mean": 0.9309794351662675, |
| "flow/mag_ratio_std": 0.07776250285372886, |
| "step": 11264 |
| }, |
| { |
| "epoch": 0.5320770403214632, |
| "grad_norm": 0.7631673216819763, |
| "learning_rate": 0.0004675131611991607, |
| "loss": 0.2390274852514267, |
| "step": 11520 |
| }, |
| { |
| "epoch": 0.5439009745508291, |
| "grad_norm": 0.3735812306404114, |
| "learning_rate": 0.0004485586029984899, |
| "loss": 0.23720039427280426, |
| "step": 11776 |
| }, |
| { |
| "epoch": 0.5557249087801949, |
| "grad_norm": 0.31787776947021484, |
| "learning_rate": 0.00042967849590597266, |
| "loss": 0.2358924150466919, |
| "step": 12032 |
| }, |
| { |
| "epoch": 0.5675488430095608, |
| "grad_norm": 0.4151494801044464, |
| "learning_rate": 0.0004109001650911621, |
| "loss": 0.23592068254947662, |
| "step": 12288 |
| }, |
| { |
| "epoch": 0.5675488430095608, |
| "eval_bleu": 0.7193234778931977, |
| "eval_ce_loss": 0.9832807419370843, |
| "eval_cos_loss": 0.06677046464968762, |
| "eval_loss": 0.2394936037118032, |
| "eval_mse_loss": 0.1340124544147487, |
| "eval_rec_loss": 0.00047602696994042604, |
| "flow/cos_sim": 0.933229555278064, |
| "flow/improvement_ratio": 0.9741054156871691, |
| "flow/mag_ratio_mean": 0.9458617875565133, |
| "flow/mag_ratio_std": 0.07636498908201854, |
| "step": 12288 |
| }, |
| { |
| "epoch": 0.5675488430095608, |
| "eval_bleu": 0.7193234778931977, |
| "eval_ce_loss": 0.9832807419370843, |
| "eval_cos_loss": 0.06677046464968762, |
| "eval_loss": 0.2394936037118032, |
| "eval_mse_loss": 0.1340124544147487, |
| "eval_rec_loss": 0.00047602696994042604, |
| "eval_runtime": 146.0267, |
| "eval_samples_per_second": 191.698, |
| "eval_steps_per_second": 2.999, |
| "flow/cos_sim": 0.933229555278064, |
| "flow/improvement_ratio": 0.9741054156871691, |
| "flow/mag_ratio_mean": 0.9458617875565133, |
| "flow/mag_ratio_std": 0.07636498908201854, |
| "step": 12288 |
| }, |
| { |
| "epoch": 0.5793727772389267, |
| "grad_norm": 0.5829018950462341, |
| "learning_rate": 0.0003922507884228551, |
| "loss": 0.23718097805976868, |
| "step": 12544 |
| }, |
| { |
| "epoch": 0.5911967114682924, |
| "grad_norm": 0.24561040103435516, |
| "learning_rate": 0.00037375735713457723, |
| "loss": 0.23286746442317963, |
| "step": 12800 |
| }, |
| { |
| "epoch": 0.6030206456976583, |
| "grad_norm": 0.2852168381214142, |
| "learning_rate": 0.00035544663676018276, |
| "loss": 0.23593547940254211, |
| "step": 13056 |
| }, |
| { |
| "epoch": 0.6148445799270241, |
| "grad_norm": 0.387464314699173, |
| "learning_rate": 0.00033734512839611255, |
| "loss": 0.23200258612632751, |
| "step": 13312 |
| }, |
| { |
| "epoch": 0.6148445799270241, |
| "eval_bleu": 0.7243775501478836, |
| "eval_ce_loss": 0.9625263297802782, |
| "eval_cos_loss": 0.06494614632468518, |
| "eval_loss": 0.2341489744581044, |
| "eval_mse_loss": 0.13099264763465754, |
| "eval_rec_loss": 0.00040907774940659357, |
| "flow/cos_sim": 0.9350538734159514, |
| "flow/improvement_ratio": 0.9740996707792151, |
| "flow/mag_ratio_mean": 0.9335339481699957, |
| "flow/mag_ratio_std": 0.07727514897987722, |
| "step": 13312 |
| }, |
| { |
| "epoch": 0.6148445799270241, |
| "eval_bleu": 0.7243775501478836, |
| "eval_ce_loss": 0.9625263297802782, |
| "eval_cos_loss": 0.06494614632468518, |
| "eval_loss": 0.2341489744581044, |
| "eval_mse_loss": 0.13099264763465754, |
| "eval_rec_loss": 0.00040907774940659357, |
| "eval_runtime": 144.1869, |
| "eval_samples_per_second": 194.144, |
| "eval_steps_per_second": 3.038, |
| "flow/cos_sim": 0.9350538734159514, |
| "flow/improvement_ratio": 0.9740996707792151, |
| "flow/mag_ratio_mean": 0.9335339481699957, |
| "flow/mag_ratio_std": 0.07727514897987722, |
| "step": 13312 |
| } |
| ], |
| "logging_steps": 256, |
| "max_steps": 21651, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 1024, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 64, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
