| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.6621403168444876, |
| "eval_steps": 1024, |
| "global_step": 14336, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.011823934229365849, |
| "grad_norm": 0.17313924431800842, |
| "learning_rate": 0.000498046875, |
| "loss": 1.7064213752746582, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.023647868458731697, |
| "grad_norm": 0.15720121562480927, |
| "learning_rate": 0.000998046875, |
| "loss": 1.1101791858673096, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.03547180268809755, |
| "grad_norm": 0.145066499710083, |
| "learning_rate": 0.000999640996023194, |
| "loss": 1.0319945812225342, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.047295736917463395, |
| "grad_norm": 0.1564043015241623, |
| "learning_rate": 0.0009985588674043958, |
| "loss": 0.9714428186416626, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.047295736917463395, |
| "eval_loss": 0.9119398776254698, |
| "eval_mse_loss": 0.9119398776254698, |
| "flow/cos_sim": 0.705863389794685, |
| "flow/improvement_ratio": 0.4745521226974383, |
| "flow/mag_ratio_mean": 0.6946878124317623, |
| "flow/mag_ratio_std": 0.17089739207127322, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.047295736917463395, |
| "eval_loss": 0.9119398776254698, |
| "eval_mse_loss": 0.9119398776254698, |
| "eval_runtime": 100.3557, |
| "eval_samples_per_second": 278.938, |
| "eval_steps_per_second": 4.364, |
| "flow/cos_sim": 0.705863389794685, |
| "flow/improvement_ratio": 0.4745521226974383, |
| "flow/mag_ratio_mean": 0.6946878124317623, |
| "flow/mag_ratio_std": 0.17089739207127322, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.05911967114682925, |
| "grad_norm": 0.16647565364837646, |
| "learning_rate": 0.0009967551747861387, |
| "loss": 0.929740846157074, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.0709436053761951, |
| "grad_norm": 0.18282100558280945, |
| "learning_rate": 0.000994232528651847, |
| "loss": 0.895330548286438, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.08276753960556095, |
| "grad_norm": 0.17141221463680267, |
| "learning_rate": 0.0009909945800260092, |
| "loss": 0.8683519959449768, |
| "step": 1792 |
| }, |
| { |
| "epoch": 0.09459147383492679, |
| "grad_norm": 0.19230426847934723, |
| "learning_rate": 0.0009870460151900522, |
| "loss": 0.8509060144424438, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.09459147383492679, |
| "eval_loss": 0.8054492728078746, |
| "eval_mse_loss": 0.8054492728078746, |
| "flow/cos_sim": 0.7493123941497716, |
| "flow/improvement_ratio": 0.476775190669652, |
| "flow/mag_ratio_mean": 0.7334817982699773, |
| "flow/mag_ratio_std": 0.1754700492846367, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.09459147383492679, |
| "eval_loss": 0.8054492728078746, |
| "eval_mse_loss": 0.8054492728078746, |
| "eval_runtime": 101.0321, |
| "eval_samples_per_second": 277.07, |
| "eval_steps_per_second": 4.335, |
| "flow/cos_sim": 0.7493123941497716, |
| "flow/improvement_ratio": 0.476775190669652, |
| "flow/mag_ratio_mean": 0.7334817982699773, |
| "flow/mag_ratio_std": 0.1754700492846367, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.10641540806429264, |
| "grad_norm": 0.1906282752752304, |
| "learning_rate": 0.0009823925488998885, |
| "loss": 0.8359999060630798, |
| "step": 2304 |
| }, |
| { |
| "epoch": 0.1182393422936585, |
| "grad_norm": 0.24338403344154358, |
| "learning_rate": 0.0009770409161149525, |
| "loss": 0.8268019556999207, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.13006327652302435, |
| "grad_norm": 0.2402328997850418, |
| "learning_rate": 0.0009709988622506973, |
| "loss": 0.8164141178131104, |
| "step": 2816 |
| }, |
| { |
| "epoch": 0.1418872107523902, |
| "grad_norm": 0.19165532290935516, |
| "learning_rate": 0.000964275131968659, |
| "loss": 0.8093737363815308, |
| "step": 3072 |
| }, |
| { |
| "epoch": 0.1418872107523902, |
| "eval_loss": 0.7729098539646357, |
| "eval_mse_loss": 0.7729098539646357, |
| "flow/cos_sim": 0.7615711595097633, |
| "flow/improvement_ratio": 0.48455443985113816, |
| "flow/mag_ratio_mean": 0.7450809510059009, |
| "flow/mag_ratio_std": 0.17553651608424645, |
| "step": 3072 |
| }, |
| { |
| "epoch": 0.1418872107523902, |
| "eval_loss": 0.7729098539646357, |
| "eval_mse_loss": 0.7729098539646357, |
| "eval_runtime": 100.8602, |
| "eval_samples_per_second": 277.543, |
| "eval_steps_per_second": 4.343, |
| "flow/cos_sim": 0.7615711595097633, |
| "flow/improvement_ratio": 0.48455443985113816, |
| "flow/mag_ratio_mean": 0.7450809510059009, |
| "flow/mag_ratio_std": 0.17553651608424645, |
| "step": 3072 |
| }, |
| { |
| "epoch": 0.15371114498175603, |
| "grad_norm": 0.23683102428913116, |
| "learning_rate": 0.0009568794565203123, |
| "loss": 0.8039254546165466, |
| "step": 3328 |
| }, |
| { |
| "epoch": 0.1655350792111219, |
| "grad_norm": 0.24094223976135254, |
| "learning_rate": 0.0009488225396630347, |
| "loss": 0.8009832501411438, |
| "step": 3584 |
| }, |
| { |
| "epoch": 0.17735901344048774, |
| "grad_norm": 0.22388198971748352, |
| "learning_rate": 0.0009401160421685646, |
| "loss": 0.7924661636352539, |
| "step": 3840 |
| }, |
| { |
| "epoch": 0.18918294766985358, |
| "grad_norm": 0.22257934510707855, |
| "learning_rate": 0.0009307725649463714, |
| "loss": 0.7945993542671204, |
| "step": 4096 |
| }, |
| { |
| "epoch": 0.18918294766985358, |
| "eval_loss": 0.7515348514465436, |
| "eval_mse_loss": 0.7515348514465436, |
| "flow/cos_sim": 0.7693686923479925, |
| "flow/improvement_ratio": 0.4754656074934354, |
| "flow/mag_ratio_mean": 0.7561575877884207, |
| "flow/mag_ratio_std": 0.1765889780570383, |
| "step": 4096 |
| }, |
| { |
| "epoch": 0.18918294766985358, |
| "eval_loss": 0.7515348514465436, |
| "eval_mse_loss": 0.7515348514465436, |
| "eval_runtime": 100.3871, |
| "eval_samples_per_second": 278.851, |
| "eval_steps_per_second": 4.363, |
| "flow/cos_sim": 0.7693686923479925, |
| "flow/improvement_ratio": 0.4754656074934354, |
| "flow/mag_ratio_mean": 0.7561575877884207, |
| "flow/mag_ratio_std": 0.1765889780570383, |
| "step": 4096 |
| }, |
| { |
| "epoch": 0.20100688189921945, |
| "grad_norm": 0.22517219185829163, |
| "learning_rate": 0.0009208056308063659, |
| "loss": 0.7901711463928223, |
| "step": 4352 |
| }, |
| { |
| "epoch": 0.2128308161285853, |
| "grad_norm": 0.2154047191143036, |
| "learning_rate": 0.0009102296648873445, |
| "loss": 0.7860417366027832, |
| "step": 4608 |
| }, |
| { |
| "epoch": 0.22465475035795113, |
| "grad_norm": 0.2263442873954773, |
| "learning_rate": 0.0008990599737794927, |
| "loss": 0.785137414932251, |
| "step": 4864 |
| }, |
| { |
| "epoch": 0.236478684587317, |
| "grad_norm": 0.2566796839237213, |
| "learning_rate": 0.0008873127233711644, |
| "loss": 0.7781056761741638, |
| "step": 5120 |
| }, |
| { |
| "epoch": 0.236478684587317, |
| "eval_loss": 0.7397306067486332, |
| "eval_mse_loss": 0.7397306067486332, |
| "flow/cos_sim": 0.7743443068304018, |
| "flow/improvement_ratio": 0.47565921052405824, |
| "flow/mag_ratio_mean": 0.7624385751545701, |
| "flow/mag_ratio_std": 0.17572799042615717, |
| "step": 5120 |
| }, |
| { |
| "epoch": 0.236478684587317, |
| "eval_loss": 0.7397306067486332, |
| "eval_mse_loss": 0.7397306067486332, |
| "eval_runtime": 100.3693, |
| "eval_samples_per_second": 278.9, |
| "eval_steps_per_second": 4.364, |
| "flow/cos_sim": 0.7743443068304018, |
| "flow/improvement_ratio": 0.47565921052405824, |
| "flow/mag_ratio_mean": 0.7624385751545701, |
| "flow/mag_ratio_std": 0.17572799042615717, |
| "step": 5120 |
| }, |
| { |
| "epoch": 0.24830261881668284, |
| "grad_norm": 0.2051643282175064, |
| "learning_rate": 0.0008750049154520011, |
| "loss": 0.7773663997650146, |
| "step": 5376 |
| }, |
| { |
| "epoch": 0.2601265530460487, |
| "grad_norm": 0.23271359503269196, |
| "learning_rate": 0.0008621543631062487, |
| "loss": 0.7789486050605774, |
| "step": 5632 |
| }, |
| { |
| "epoch": 0.27195048727541454, |
| "grad_norm": 0.22768262028694153, |
| "learning_rate": 0.0008487796649318904, |
| "loss": 0.7753859758377075, |
| "step": 5888 |
| }, |
| { |
| "epoch": 0.2837744215047804, |
| "grad_norm": 0.23808346688747406, |
| "learning_rate": 0.0008349001781229053, |
| "loss": 0.7760079503059387, |
| "step": 6144 |
| }, |
| { |
| "epoch": 0.2837744215047804, |
| "eval_loss": 0.7382390670580407, |
| "eval_mse_loss": 0.7382390670580407, |
| "flow/cos_sim": 0.7747776206225565, |
| "flow/improvement_ratio": 0.48073648889315185, |
| "flow/mag_ratio_mean": 0.7609562668354, |
| "flow/mag_ratio_std": 0.17754725645802336, |
| "step": 6144 |
| }, |
| { |
| "epoch": 0.2837744215047804, |
| "eval_loss": 0.7382390670580407, |
| "eval_mse_loss": 0.7382390670580407, |
| "eval_runtime": 101.1596, |
| "eval_samples_per_second": 276.721, |
| "eval_steps_per_second": 4.33, |
| "flow/cos_sim": 0.7747776206225565, |
| "flow/improvement_ratio": 0.48073648889315185, |
| "flow/mag_ratio_mean": 0.7609562668354, |
| "flow/mag_ratio_std": 0.17754725645802336, |
| "step": 6144 |
| }, |
| { |
| "epoch": 0.2955983557341462, |
| "grad_norm": 0.22078828513622284, |
| "learning_rate": 0.0008205359904536107, |
| "loss": 0.7707664966583252, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.30742228996351206, |
| "grad_norm": 0.21839560568332672, |
| "learning_rate": 0.0008057078912056363, |
| "loss": 0.7725305557250977, |
| "step": 6656 |
| }, |
| { |
| "epoch": 0.3192462241928779, |
| "grad_norm": 0.23544862866401672, |
| "learning_rate": 0.0007904373410796086, |
| "loss": 0.7722088694572449, |
| "step": 6912 |
| }, |
| { |
| "epoch": 0.3310701584222438, |
| "grad_norm": 0.23949392139911652, |
| "learning_rate": 0.0007747464411350876, |
| "loss": 0.7707281708717346, |
| "step": 7168 |
| }, |
| { |
| "epoch": 0.3310701584222438, |
| "eval_loss": 0.7320967807889529, |
| "eval_mse_loss": 0.7320967807889529, |
| "flow/cos_sim": 0.7774923427464211, |
| "flow/improvement_ratio": 0.48102973714538905, |
| "flow/mag_ratio_mean": 0.7606384374507486, |
| "flow/mag_ratio_std": 0.1749520302839475, |
| "step": 7168 |
| }, |
| { |
| "epoch": 0.3310701584222438, |
| "eval_loss": 0.7320967807889529, |
| "eval_mse_loss": 0.7320967807889529, |
| "eval_runtime": 101.0362, |
| "eval_samples_per_second": 277.059, |
| "eval_steps_per_second": 4.335, |
| "flow/cos_sim": 0.7774923427464211, |
| "flow/improvement_ratio": 0.48102973714538905, |
| "flow/mag_ratio_mean": 0.7606384374507486, |
| "flow/mag_ratio_std": 0.1749520302839475, |
| "step": 7168 |
| }, |
| { |
| "epoch": 0.34289409265160964, |
| "grad_norm": 0.21281319856643677, |
| "learning_rate": 0.000758657900803716, |
| "loss": 0.7705408930778503, |
| "step": 7424 |
| }, |
| { |
| "epoch": 0.3547180268809755, |
| "grad_norm": 0.2182183712720871, |
| "learning_rate": 0.000742195005021869, |
| "loss": 0.7691267728805542, |
| "step": 7680 |
| }, |
| { |
| "epoch": 0.3665419611103413, |
| "grad_norm": 0.20540344715118408, |
| "learning_rate": 0.0007253815805303786, |
| "loss": 0.7687782049179077, |
| "step": 7936 |
| }, |
| { |
| "epoch": 0.37836589533970716, |
| "grad_norm": 0.25242024660110474, |
| "learning_rate": 0.000708309515673374, |
| "loss": 0.7675922513008118, |
| "step": 8192 |
| }, |
| { |
| "epoch": 0.37836589533970716, |
| "eval_loss": 0.72715279745729, |
| "eval_mse_loss": 0.72715279745729, |
| "flow/cos_sim": 0.7793520910826992, |
| "flow/improvement_ratio": 0.47947503408612724, |
| "flow/mag_ratio_mean": 0.7625162698090349, |
| "flow/mag_ratio_std": 0.17530882001331408, |
| "step": 8192 |
| }, |
| { |
| "epoch": 0.37836589533970716, |
| "eval_loss": 0.72715279745729, |
| "eval_mse_loss": 0.72715279745729, |
| "eval_runtime": 100.656, |
| "eval_samples_per_second": 278.106, |
| "eval_steps_per_second": 4.351, |
| "flow/cos_sim": 0.7793520910826992, |
| "flow/improvement_ratio": 0.47947503408612724, |
| "flow/mag_ratio_mean": 0.7625162698090349, |
| "flow/mag_ratio_std": 0.17530882001331408, |
| "step": 8192 |
| }, |
| { |
| "epoch": 0.390189829569073, |
| "grad_norm": 0.18503619730472565, |
| "learning_rate": 0.0006908696365085842, |
| "loss": 0.7652720808982849, |
| "step": 8448 |
| }, |
| { |
| "epoch": 0.4020137637984389, |
| "grad_norm": 0.20311063528060913, |
| "learning_rate": 0.0006731535118143318, |
| "loss": 0.7671474814414978, |
| "step": 8704 |
| }, |
| { |
| "epoch": 0.41383769802780473, |
| "grad_norm": 0.17226819694042206, |
| "learning_rate": 0.0006551867821290267, |
| "loss": 0.7648245096206665, |
| "step": 8960 |
| }, |
| { |
| "epoch": 0.4256616322571706, |
| "grad_norm": 0.20406724512577057, |
| "learning_rate": 0.0006369954506915572, |
| "loss": 0.7627850770950317, |
| "step": 9216 |
| }, |
| { |
| "epoch": 0.4256616322571706, |
| "eval_loss": 0.7245442253541728, |
| "eval_mse_loss": 0.7245442253541728, |
| "flow/cos_sim": 0.7804642662610093, |
| "flow/improvement_ratio": 0.478297841671395, |
| "flow/mag_ratio_mean": 0.7637837749637969, |
| "flow/mag_ratio_std": 0.17519798711554646, |
| "step": 9216 |
| }, |
| { |
| "epoch": 0.4256616322571706, |
| "eval_loss": 0.7245442253541728, |
| "eval_mse_loss": 0.7245442253541728, |
| "eval_runtime": 100.5176, |
| "eval_samples_per_second": 278.488, |
| "eval_steps_per_second": 4.357, |
| "flow/cos_sim": 0.7804642662610093, |
| "flow/improvement_ratio": 0.478297841671395, |
| "flow/mag_ratio_mean": 0.7637837749637969, |
| "flow/mag_ratio_std": 0.17519798711554646, |
| "step": 9216 |
| }, |
| { |
| "epoch": 0.4374855664865364, |
| "grad_norm": 0.19549715518951416, |
| "learning_rate": 0.0006186058458068149, |
| "loss": 0.7646156549453735, |
| "step": 9472 |
| }, |
| { |
| "epoch": 0.44930950071590225, |
| "grad_norm": 0.22935132682323456, |
| "learning_rate": 0.0006000445827407526, |
| "loss": 0.7637522220611572, |
| "step": 9728 |
| }, |
| { |
| "epoch": 0.4611334349452681, |
| "grad_norm": 0.1892293095588684, |
| "learning_rate": 0.0005813385252001232, |
| "loss": 0.7614347338676453, |
| "step": 9984 |
| }, |
| { |
| "epoch": 0.472957369174634, |
| "grad_norm": 0.19596756994724274, |
| "learning_rate": 0.00056251474645265, |
| "loss": 0.7602492570877075, |
| "step": 10240 |
| }, |
| { |
| "epoch": 0.472957369174634, |
| "eval_loss": 0.7225635307564583, |
| "eval_mse_loss": 0.7225635307564583, |
| "flow/cos_sim": 0.781412697137763, |
| "flow/improvement_ratio": 0.48148406279958, |
| "flow/mag_ratio_mean": 0.7642885947336345, |
| "flow/mag_ratio_std": 0.17388889067657462, |
| "step": 10240 |
| }, |
| { |
| "epoch": 0.472957369174634, |
| "eval_loss": 0.7225635307564583, |
| "eval_mse_loss": 0.7225635307564583, |
| "eval_runtime": 100.4132, |
| "eval_samples_per_second": 278.778, |
| "eval_steps_per_second": 4.362, |
| "flow/cos_sim": 0.781412697137763, |
| "flow/improvement_ratio": 0.48148406279958, |
| "flow/mag_ratio_mean": 0.7642885947336345, |
| "flow/mag_ratio_std": 0.17388889067657462, |
| "step": 10240 |
| }, |
| { |
| "epoch": 0.48478130340399983, |
| "grad_norm": 0.2039279341697693, |
| "learning_rate": 0.0005436004901439003, |
| "loss": 0.7591882944107056, |
| "step": 10496 |
| }, |
| { |
| "epoch": 0.49660523763336567, |
| "grad_norm": 0.18449348211288452, |
| "learning_rate": 0.0005246973484120841, |
| "loss": 0.7581324577331543, |
| "step": 10752 |
| }, |
| { |
| "epoch": 0.5084291718627315, |
| "grad_norm": 0.18734613060951233, |
| "learning_rate": 0.0005056844377834413, |
| "loss": 0.7585782408714294, |
| "step": 11008 |
| }, |
| { |
| "epoch": 0.5202531060920974, |
| "grad_norm": 0.19812439382076263, |
| "learning_rate": 0.0004866633000708374, |
| "loss": 0.758400559425354, |
| "step": 11264 |
| }, |
| { |
| "epoch": 0.5202531060920974, |
| "eval_loss": 0.7193547954025878, |
| "eval_mse_loss": 0.7193547954025878, |
| "flow/cos_sim": 0.7829214582160183, |
| "flow/improvement_ratio": 0.4840309496747849, |
| "flow/mag_ratio_mean": 0.7646983984398515, |
| "flow/mag_ratio_std": 0.17315611838614015, |
| "step": 11264 |
| }, |
| { |
| "epoch": 0.5202531060920974, |
| "eval_loss": 0.7193547954025878, |
| "eval_mse_loss": 0.7193547954025878, |
| "eval_runtime": 100.2674, |
| "eval_samples_per_second": 279.183, |
| "eval_steps_per_second": 4.368, |
| "flow/cos_sim": 0.7829214582160183, |
| "flow/improvement_ratio": 0.4840309496747849, |
| "flow/mag_ratio_mean": 0.7646983984398515, |
| "flow/mag_ratio_std": 0.17315611838614015, |
| "step": 11264 |
| }, |
| { |
| "epoch": 0.5320770403214632, |
| "grad_norm": 0.20949678122997284, |
| "learning_rate": 0.00046766146455737116, |
| "loss": 0.756182074546814, |
| "step": 11520 |
| }, |
| { |
| "epoch": 0.5439009745508291, |
| "grad_norm": 0.20575636625289917, |
| "learning_rate": 0.00044870643259007823, |
| "loss": 0.7595986723899841, |
| "step": 11776 |
| }, |
| { |
| "epoch": 0.5557249087801949, |
| "grad_norm": 0.18551619350910187, |
| "learning_rate": 0.000429825637777245, |
| "loss": 0.7588385939598083, |
| "step": 12032 |
| }, |
| { |
| "epoch": 0.5675488430095608, |
| "grad_norm": 0.21281953155994415, |
| "learning_rate": 0.00041104640628376166, |
| "loss": 0.755641758441925, |
| "step": 12288 |
| }, |
| { |
| "epoch": 0.5675488430095608, |
| "eval_loss": 0.7185944062934074, |
| "eval_mse_loss": 0.7185944062934074, |
| "flow/cos_sim": 0.7829709806943048, |
| "flow/improvement_ratio": 0.4833299268598426, |
| "flow/mag_ratio_mean": 0.7666111965429837, |
| "flow/mag_ratio_std": 0.17369252120114897, |
| "step": 12288 |
| }, |
| { |
| "epoch": 0.5675488430095608, |
| "eval_loss": 0.7185944062934074, |
| "eval_mse_loss": 0.7185944062934074, |
| "eval_runtime": 100.2742, |
| "eval_samples_per_second": 279.165, |
| "eval_steps_per_second": 4.368, |
| "flow/cos_sim": 0.7829709806943048, |
| "flow/improvement_ratio": 0.4833299268598426, |
| "flow/mag_ratio_mean": 0.7666111965429837, |
| "flow/mag_ratio_std": 0.17369252120114897, |
| "step": 12288 |
| }, |
| { |
| "epoch": 0.5793727772389267, |
| "grad_norm": 0.1611846536397934, |
| "learning_rate": 0.00039239591728197724, |
| "loss": 0.7571735382080078, |
| "step": 12544 |
| }, |
| { |
| "epoch": 0.5911967114682924, |
| "grad_norm": 0.17743456363677979, |
| "learning_rate": 0.0003739011636152962, |
| "loss": 0.7581431269645691, |
| "step": 12800 |
| }, |
| { |
| "epoch": 0.6030206456976583, |
| "grad_norm": 0.17940299212932587, |
| "learning_rate": 0.0003556600555022182, |
| "loss": 0.7566176652908325, |
| "step": 13056 |
| }, |
| { |
| "epoch": 0.6148445799270241, |
| "grad_norm": 0.1930851936340332, |
| "learning_rate": 0.00033755594309981104, |
| "loss": 0.7547513246536255, |
| "step": 13312 |
| }, |
| { |
| "epoch": 0.6148445799270241, |
| "eval_loss": 0.7150395362888842, |
| "eval_mse_loss": 0.7150395362888842, |
| "flow/cos_sim": 0.7844073471685523, |
| "flow/improvement_ratio": 0.47775867540542394, |
| "flow/mag_ratio_mean": 0.7674100999146292, |
| "flow/mag_ratio_std": 0.1724472795039007, |
| "step": 13312 |
| }, |
| { |
| "epoch": 0.6148445799270241, |
| "eval_loss": 0.7150395362888842, |
| "eval_mse_loss": 0.7150395362888842, |
| "eval_runtime": 100.7945, |
| "eval_samples_per_second": 277.724, |
| "eval_steps_per_second": 4.345, |
| "flow/cos_sim": 0.7844073471685523, |
| "flow/improvement_ratio": 0.47775867540542394, |
| "flow/mag_ratio_mean": 0.7674100999146292, |
| "flow/mag_ratio_std": 0.1724472795039007, |
| "step": 13312 |
| }, |
| { |
| "epoch": 0.62666851415639, |
| "grad_norm": 0.17439058423042297, |
| "learning_rate": 0.000319686935899715, |
| "loss": 0.7525946497917175, |
| "step": 13568 |
| }, |
| { |
| "epoch": 0.6384924483857558, |
| "grad_norm": 0.16240930557250977, |
| "learning_rate": 0.0003020788957071146, |
| "loss": 0.7547707557678223, |
| "step": 13824 |
| }, |
| { |
| "epoch": 0.6503163826151217, |
| "grad_norm": 0.16282112896442413, |
| "learning_rate": 0.0002847573066297746, |
| "loss": 0.7539600729942322, |
| "step": 14080 |
| }, |
| { |
| "epoch": 0.6621403168444876, |
| "grad_norm": 0.19493776559829712, |
| "learning_rate": 0.0002677472381949038, |
| "loss": 0.753170371055603, |
| "step": 14336 |
| }, |
| { |
| "epoch": 0.6621403168444876, |
| "eval_loss": 0.7136020788319035, |
| "eval_mse_loss": 0.7136020788319035, |
| "flow/cos_sim": 0.7848518109757062, |
| "flow/improvement_ratio": 0.47987128467592477, |
| "flow/mag_ratio_mean": 0.7672100954404161, |
| "flow/mag_ratio_std": 0.17320666945280005, |
| "step": 14336 |
| }, |
| { |
| "epoch": 0.6621403168444876, |
| "eval_loss": 0.7136020788319035, |
| "eval_mse_loss": 0.7136020788319035, |
| "eval_runtime": 100.4201, |
| "eval_samples_per_second": 278.759, |
| "eval_steps_per_second": 4.362, |
| "flow/cos_sim": 0.7848518109757062, |
| "flow/improvement_ratio": 0.47987128467592477, |
| "flow/mag_ratio_mean": 0.7672100954404161, |
| "flow/mag_ratio_std": 0.17320666945280005, |
| "step": 14336 |
| } |
| ], |
| "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 |
| } |
|
|
