| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "eval_steps": 500, | |
| "global_step": 113400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.912750482559204, | |
| "learning_rate": 9.900352733686068e-06, | |
| "loss": 2.4424, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.366978645324707, | |
| "learning_rate": 9.800440917107585e-06, | |
| "loss": 2.3536, | |
| "step": 2268 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.4429948329925537, | |
| "learning_rate": 9.700440917107584e-06, | |
| "loss": 2.3138, | |
| "step": 3402 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.427384853363037, | |
| "learning_rate": 9.600440917107585e-06, | |
| "loss": 2.2871, | |
| "step": 4536 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.3265113830566406, | |
| "learning_rate": 9.500529100529102e-06, | |
| "loss": 2.2576, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 4.512721538543701, | |
| "learning_rate": 9.400529100529101e-06, | |
| "loss": 2.2087, | |
| "step": 6804 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 3.427581548690796, | |
| "learning_rate": 9.300617283950618e-06, | |
| "loss": 2.1972, | |
| "step": 7938 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.985769510269165, | |
| "learning_rate": 9.200617283950617e-06, | |
| "loss": 2.1687, | |
| "step": 9072 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 5.713807106018066, | |
| "learning_rate": 9.100705467372136e-06, | |
| "loss": 2.1537, | |
| "step": 10206 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.9860029220581055, | |
| "learning_rate": 9.000705467372135e-06, | |
| "loss": 2.1393, | |
| "step": 11340 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.900705467372134e-06, | |
| "loss": 2.1331, | |
| "step": 12474 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 5.642363548278809, | |
| "learning_rate": 8.800793650793651e-06, | |
| "loss": 2.1241, | |
| "step": 13608 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 3.955806016921997, | |
| "learning_rate": 8.700881834215168e-06, | |
| "loss": 2.1189, | |
| "step": 14742 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 4.731072425842285, | |
| "learning_rate": 8.600881834215168e-06, | |
| "loss": 2.1023, | |
| "step": 15876 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 5.199740886688232, | |
| "learning_rate": 8.500881834215169e-06, | |
| "loss": 2.0996, | |
| "step": 17010 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 8.10754680633545, | |
| "learning_rate": 8.400970017636686e-06, | |
| "loss": 2.0948, | |
| "step": 18144 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 7.051990509033203, | |
| "learning_rate": 8.300970017636685e-06, | |
| "loss": 2.0864, | |
| "step": 19278 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 3.324284553527832, | |
| "learning_rate": 8.200970017636684e-06, | |
| "loss": 2.0647, | |
| "step": 20412 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 4.6707682609558105, | |
| "learning_rate": 8.101058201058201e-06, | |
| "loss": 2.0577, | |
| "step": 21546 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 3.083761692047119, | |
| "learning_rate": 8.001058201058202e-06, | |
| "loss": 2.0609, | |
| "step": 22680 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 4.806012153625488, | |
| "learning_rate": 7.901058201058201e-06, | |
| "loss": 2.0601, | |
| "step": 23814 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 4.715734958648682, | |
| "learning_rate": 7.801146384479718e-06, | |
| "loss": 2.0389, | |
| "step": 24948 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 3.61401104927063, | |
| "learning_rate": 7.70114638447972e-06, | |
| "loss": 2.0428, | |
| "step": 26082 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 2.6958115100860596, | |
| "learning_rate": 7.6011463844797185e-06, | |
| "loss": 2.0315, | |
| "step": 27216 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 2.96954345703125, | |
| "learning_rate": 7.5011463844797186e-06, | |
| "loss": 2.0378, | |
| "step": 28350 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 5.391402721405029, | |
| "learning_rate": 7.4013227513227516e-06, | |
| "loss": 2.0096, | |
| "step": 29484 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 2.2738614082336426, | |
| "learning_rate": 7.301322751322752e-06, | |
| "loss": 2.0087, | |
| "step": 30618 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 4.0258378982543945, | |
| "learning_rate": 7.201322751322753e-06, | |
| "loss": 2.0135, | |
| "step": 31752 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 3.196429491043091, | |
| "learning_rate": 7.101410934744269e-06, | |
| "loss": 2.013, | |
| "step": 32886 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 3.6472020149230957, | |
| "learning_rate": 7.001410934744268e-06, | |
| "loss": 2.0002, | |
| "step": 34020 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 3.440370559692383, | |
| "learning_rate": 6.901499118165785e-06, | |
| "loss": 2.005, | |
| "step": 35154 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 5.474171161651611, | |
| "learning_rate": 6.801587301587303e-06, | |
| "loss": 1.9974, | |
| "step": 36288 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 2.209702253341675, | |
| "learning_rate": 6.701587301587302e-06, | |
| "loss": 2.0003, | |
| "step": 37422 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 5.641165256500244, | |
| "learning_rate": 6.601587301587302e-06, | |
| "loss": 1.9966, | |
| "step": 38556 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 3.8642516136169434, | |
| "learning_rate": 6.501587301587302e-06, | |
| "loss": 1.9866, | |
| "step": 39690 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "grad_norm": 3.8914637565612793, | |
| "learning_rate": 6.401675485008819e-06, | |
| "loss": 1.9769, | |
| "step": 40824 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "grad_norm": 2.3981971740722656, | |
| "learning_rate": 6.301675485008819e-06, | |
| "loss": 1.9695, | |
| "step": 41958 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "grad_norm": 3.656935453414917, | |
| "learning_rate": 6.201763668430335e-06, | |
| "loss": 1.9774, | |
| "step": 43092 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "grad_norm": 5.288485050201416, | |
| "learning_rate": 6.1018518518518525e-06, | |
| "loss": 1.984, | |
| "step": 44226 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 5.823598861694336, | |
| "learning_rate": 6.001851851851852e-06, | |
| "loss": 1.9786, | |
| "step": 45360 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "grad_norm": 9.096850395202637, | |
| "learning_rate": 5.901940035273369e-06, | |
| "loss": 1.9815, | |
| "step": 46494 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "grad_norm": 2.720540761947632, | |
| "learning_rate": 5.801940035273369e-06, | |
| "loss": 1.9756, | |
| "step": 47628 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "grad_norm": 6.2244696617126465, | |
| "learning_rate": 5.702028218694886e-06, | |
| "loss": 1.9715, | |
| "step": 48762 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "grad_norm": 1.7180719375610352, | |
| "learning_rate": 5.602028218694886e-06, | |
| "loss": 1.9783, | |
| "step": 49896 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "grad_norm": 3.5073482990264893, | |
| "learning_rate": 5.502116402116403e-06, | |
| "loss": 1.9619, | |
| "step": 51030 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "grad_norm": 3.795732021331787, | |
| "learning_rate": 5.402204585537919e-06, | |
| "loss": 1.9664, | |
| "step": 52164 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "grad_norm": 2.003103733062744, | |
| "learning_rate": 5.302204585537919e-06, | |
| "loss": 1.9772, | |
| "step": 53298 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "grad_norm": 8.892159461975098, | |
| "learning_rate": 5.202292768959436e-06, | |
| "loss": 1.9586, | |
| "step": 54432 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "grad_norm": 3.5988032817840576, | |
| "learning_rate": 5.102292768959436e-06, | |
| "loss": 1.9623, | |
| "step": 55566 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 6.822943687438965, | |
| "learning_rate": 5.002380952380953e-06, | |
| "loss": 1.972, | |
| "step": 56700 | |
| }, | |
| { | |
| "epoch": 51.0, | |
| "grad_norm": 3.6781959533691406, | |
| "learning_rate": 4.902380952380953e-06, | |
| "loss": 1.9518, | |
| "step": 57834 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "grad_norm": 6.689939975738525, | |
| "learning_rate": 4.802469135802469e-06, | |
| "loss": 1.9515, | |
| "step": 58968 | |
| }, | |
| { | |
| "epoch": 53.0, | |
| "grad_norm": 3.9311599731445312, | |
| "learning_rate": 4.7024691358024695e-06, | |
| "loss": 1.9482, | |
| "step": 60102 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "grad_norm": 2.7226245403289795, | |
| "learning_rate": 4.6025573192239865e-06, | |
| "loss": 1.9569, | |
| "step": 61236 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "grad_norm": 4.933896541595459, | |
| "learning_rate": 4.502645502645503e-06, | |
| "loss": 1.9498, | |
| "step": 62370 | |
| }, | |
| { | |
| "epoch": 56.0, | |
| "grad_norm": 5.327332019805908, | |
| "learning_rate": 4.402645502645503e-06, | |
| "loss": 1.9405, | |
| "step": 63504 | |
| }, | |
| { | |
| "epoch": 57.0, | |
| "grad_norm": 2.307694911956787, | |
| "learning_rate": 4.30273368606702e-06, | |
| "loss": 1.9582, | |
| "step": 64638 | |
| }, | |
| { | |
| "epoch": 58.0, | |
| "grad_norm": 2.4888343811035156, | |
| "learning_rate": 4.20273368606702e-06, | |
| "loss": 1.9462, | |
| "step": 65772 | |
| }, | |
| { | |
| "epoch": 59.0, | |
| "grad_norm": 3.256455898284912, | |
| "learning_rate": 4.102821869488536e-06, | |
| "loss": 1.9481, | |
| "step": 66906 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 2.2037665843963623, | |
| "learning_rate": 4.002821869488536e-06, | |
| "loss": 1.9526, | |
| "step": 68040 | |
| }, | |
| { | |
| "epoch": 61.0, | |
| "grad_norm": 5.307345390319824, | |
| "learning_rate": 3.902910052910053e-06, | |
| "loss": 1.9317, | |
| "step": 69174 | |
| }, | |
| { | |
| "epoch": 62.0, | |
| "grad_norm": 5.948610305786133, | |
| "learning_rate": 3.802910052910053e-06, | |
| "loss": 1.9486, | |
| "step": 70308 | |
| }, | |
| { | |
| "epoch": 63.0, | |
| "grad_norm": 4.689316272735596, | |
| "learning_rate": 3.7029982363315697e-06, | |
| "loss": 1.9503, | |
| "step": 71442 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "grad_norm": 4.029778957366943, | |
| "learning_rate": 3.6030864197530867e-06, | |
| "loss": 1.9316, | |
| "step": 72576 | |
| }, | |
| { | |
| "epoch": 65.0, | |
| "grad_norm": 4.771517276763916, | |
| "learning_rate": 3.5030864197530868e-06, | |
| "loss": 1.9331, | |
| "step": 73710 | |
| }, | |
| { | |
| "epoch": 66.0, | |
| "grad_norm": 2.0614895820617676, | |
| "learning_rate": 3.4031746031746033e-06, | |
| "loss": 1.9369, | |
| "step": 74844 | |
| }, | |
| { | |
| "epoch": 67.0, | |
| "grad_norm": 4.460591793060303, | |
| "learning_rate": 3.303174603174604e-06, | |
| "loss": 1.9512, | |
| "step": 75978 | |
| }, | |
| { | |
| "epoch": 68.0, | |
| "grad_norm": 2.2756810188293457, | |
| "learning_rate": 3.2032627865961204e-06, | |
| "loss": 1.9298, | |
| "step": 77112 | |
| }, | |
| { | |
| "epoch": 69.0, | |
| "grad_norm": 2.6889700889587402, | |
| "learning_rate": 3.10326278659612e-06, | |
| "loss": 1.9309, | |
| "step": 78246 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "grad_norm": 6.319788455963135, | |
| "learning_rate": 3.003350970017637e-06, | |
| "loss": 1.9377, | |
| "step": 79380 | |
| }, | |
| { | |
| "epoch": 71.0, | |
| "grad_norm": 3.2527384757995605, | |
| "learning_rate": 2.9034391534391536e-06, | |
| "loss": 1.9509, | |
| "step": 80514 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "grad_norm": 3.7406935691833496, | |
| "learning_rate": 2.8034391534391537e-06, | |
| "loss": 1.9285, | |
| "step": 81648 | |
| }, | |
| { | |
| "epoch": 73.0, | |
| "grad_norm": 1.6912527084350586, | |
| "learning_rate": 2.7035273368606703e-06, | |
| "loss": 1.9321, | |
| "step": 82782 | |
| }, | |
| { | |
| "epoch": 74.0, | |
| "grad_norm": 4.06993293762207, | |
| "learning_rate": 2.603527336860671e-06, | |
| "loss": 1.9334, | |
| "step": 83916 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "grad_norm": 1.849129319190979, | |
| "learning_rate": 2.503615520282187e-06, | |
| "loss": 1.9308, | |
| "step": 85050 | |
| }, | |
| { | |
| "epoch": 76.0, | |
| "grad_norm": 3.185353994369507, | |
| "learning_rate": 2.403615520282187e-06, | |
| "loss": 1.9368, | |
| "step": 86184 | |
| }, | |
| { | |
| "epoch": 77.0, | |
| "grad_norm": 2.2649121284484863, | |
| "learning_rate": 2.303703703703704e-06, | |
| "loss": 1.9328, | |
| "step": 87318 | |
| }, | |
| { | |
| "epoch": 78.0, | |
| "grad_norm": 3.588413953781128, | |
| "learning_rate": 2.203703703703704e-06, | |
| "loss": 1.9286, | |
| "step": 88452 | |
| }, | |
| { | |
| "epoch": 79.0, | |
| "grad_norm": 2.376042127609253, | |
| "learning_rate": 2.1037918871252207e-06, | |
| "loss": 1.9131, | |
| "step": 89586 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 5.023166179656982, | |
| "learning_rate": 2.0038800705467372e-06, | |
| "loss": 1.9246, | |
| "step": 90720 | |
| }, | |
| { | |
| "epoch": 81.0, | |
| "grad_norm": 3.989335775375366, | |
| "learning_rate": 1.9038800705467373e-06, | |
| "loss": 1.9278, | |
| "step": 91854 | |
| }, | |
| { | |
| "epoch": 82.0, | |
| "grad_norm": 1.964154601097107, | |
| "learning_rate": 1.803968253968254e-06, | |
| "loss": 1.9324, | |
| "step": 92988 | |
| }, | |
| { | |
| "epoch": 83.0, | |
| "grad_norm": 3.225398302078247, | |
| "learning_rate": 1.7039682539682542e-06, | |
| "loss": 1.9182, | |
| "step": 94122 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "grad_norm": 4.1602935791015625, | |
| "learning_rate": 1.6040564373897708e-06, | |
| "loss": 1.9336, | |
| "step": 95256 | |
| }, | |
| { | |
| "epoch": 85.0, | |
| "grad_norm": 4.833250522613525, | |
| "learning_rate": 1.5040564373897709e-06, | |
| "loss": 1.9248, | |
| "step": 96390 | |
| }, | |
| { | |
| "epoch": 86.0, | |
| "grad_norm": 2.619542121887207, | |
| "learning_rate": 1.4041446208112876e-06, | |
| "loss": 1.9369, | |
| "step": 97524 | |
| }, | |
| { | |
| "epoch": 87.0, | |
| "grad_norm": 2.0560250282287598, | |
| "learning_rate": 1.3041446208112875e-06, | |
| "loss": 1.9184, | |
| "step": 98658 | |
| }, | |
| { | |
| "epoch": 88.0, | |
| "grad_norm": 7.1246209144592285, | |
| "learning_rate": 1.2042328042328043e-06, | |
| "loss": 1.9279, | |
| "step": 99792 | |
| }, | |
| { | |
| "epoch": 89.0, | |
| "grad_norm": 3.3801794052124023, | |
| "learning_rate": 1.104320987654321e-06, | |
| "loss": 1.9329, | |
| "step": 100926 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "grad_norm": 1.7479180097579956, | |
| "learning_rate": 1.0043209876543212e-06, | |
| "loss": 1.9211, | |
| "step": 102060 | |
| }, | |
| { | |
| "epoch": 91.0, | |
| "grad_norm": 3.091583013534546, | |
| "learning_rate": 9.044091710758378e-07, | |
| "loss": 1.9236, | |
| "step": 103194 | |
| }, | |
| { | |
| "epoch": 92.0, | |
| "grad_norm": 2.1435961723327637, | |
| "learning_rate": 8.044091710758378e-07, | |
| "loss": 1.9214, | |
| "step": 104328 | |
| }, | |
| { | |
| "epoch": 93.0, | |
| "grad_norm": 3.5798821449279785, | |
| "learning_rate": 7.044973544973546e-07, | |
| "loss": 1.9142, | |
| "step": 105462 | |
| }, | |
| { | |
| "epoch": 94.0, | |
| "grad_norm": 1.9765084981918335, | |
| "learning_rate": 6.044973544973545e-07, | |
| "loss": 1.9336, | |
| "step": 106596 | |
| }, | |
| { | |
| "epoch": 95.0, | |
| "grad_norm": 1.3698736429214478, | |
| "learning_rate": 5.045855379188714e-07, | |
| "loss": 1.927, | |
| "step": 107730 | |
| }, | |
| { | |
| "epoch": 96.0, | |
| "grad_norm": 1.0919588804244995, | |
| "learning_rate": 4.045855379188713e-07, | |
| "loss": 1.9213, | |
| "step": 108864 | |
| }, | |
| { | |
| "epoch": 97.0, | |
| "grad_norm": 1.2900702953338623, | |
| "learning_rate": 3.04673721340388e-07, | |
| "loss": 1.9118, | |
| "step": 109998 | |
| }, | |
| { | |
| "epoch": 98.0, | |
| "grad_norm": 1.611956238746643, | |
| "learning_rate": 2.047619047619048e-07, | |
| "loss": 1.9227, | |
| "step": 111132 | |
| }, | |
| { | |
| "epoch": 99.0, | |
| "grad_norm": 4.691267490386963, | |
| "learning_rate": 1.0476190476190476e-07, | |
| "loss": 1.9204, | |
| "step": 112266 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 1.494165301322937, | |
| "learning_rate": 4.8500881834215175e-09, | |
| "loss": 1.9094, | |
| "step": 113400 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 113400, | |
| "total_flos": 3.571154905966387e+18, | |
| "train_loss": 1.224746272576549, | |
| "train_runtime": 52658.2042, | |
| "train_samples_per_second": 258.353, | |
| "train_steps_per_second": 2.154 | |
| } | |
| ], | |
| "logging_steps": 1134, | |
| "max_steps": 113400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 530, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.571154905966387e+18, | |
| "train_batch_size": 60, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |