| { | |
| "best_metric": 0.4830440580844879, | |
| "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l57-l/checkpoint-20000", | |
| "epoch": 2.34641638225256, | |
| "eval_steps": 500, | |
| "global_step": 22000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05332764505119454, | |
| "grad_norm": 414.62841796875, | |
| "learning_rate": 4.911120591581342e-07, | |
| "loss": 0.3812, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.05332764505119454, | |
| "eval_loss": 1.1163015365600586, | |
| "eval_runtime": 61.8962, | |
| "eval_samples_per_second": 15.946, | |
| "eval_steps_per_second": 2.003, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.10665529010238908, | |
| "grad_norm": 34.08418655395508, | |
| "learning_rate": 4.822241183162685e-07, | |
| "loss": 0.2683, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.10665529010238908, | |
| "eval_loss": 0.9684447646141052, | |
| "eval_runtime": 61.821, | |
| "eval_samples_per_second": 15.965, | |
| "eval_steps_per_second": 2.006, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1599829351535836, | |
| "grad_norm": 350.4574890136719, | |
| "learning_rate": 4.733361774744027e-07, | |
| "loss": 0.2119, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1599829351535836, | |
| "eval_loss": 0.9099885821342468, | |
| "eval_runtime": 61.9606, | |
| "eval_samples_per_second": 15.929, | |
| "eval_steps_per_second": 2.001, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21331058020477817, | |
| "grad_norm": 0.01787523180246353, | |
| "learning_rate": 4.6444823663253695e-07, | |
| "loss": 0.1889, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.21331058020477817, | |
| "eval_loss": 0.8620074987411499, | |
| "eval_runtime": 63.2824, | |
| "eval_samples_per_second": 15.597, | |
| "eval_steps_per_second": 1.959, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2666382252559727, | |
| "grad_norm": 118.66907501220703, | |
| "learning_rate": 4.5556029579067116e-07, | |
| "loss": 0.2071, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.2666382252559727, | |
| "eval_loss": 0.791786253452301, | |
| "eval_runtime": 63.7932, | |
| "eval_samples_per_second": 15.472, | |
| "eval_steps_per_second": 1.944, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3199658703071672, | |
| "grad_norm": 7.21087323185543e-09, | |
| "learning_rate": 4.4667235494880547e-07, | |
| "loss": 0.1588, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3199658703071672, | |
| "eval_loss": 0.7657254338264465, | |
| "eval_runtime": 62.4738, | |
| "eval_samples_per_second": 15.799, | |
| "eval_steps_per_second": 1.985, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.37329351535836175, | |
| "grad_norm": 0.0014836661284789443, | |
| "learning_rate": 4.377844141069397e-07, | |
| "loss": 0.1718, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.37329351535836175, | |
| "eval_loss": 0.7610095739364624, | |
| "eval_runtime": 63.9118, | |
| "eval_samples_per_second": 15.443, | |
| "eval_steps_per_second": 1.94, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.42662116040955633, | |
| "grad_norm": 3.2739710807800293, | |
| "learning_rate": 4.2889647326507393e-07, | |
| "loss": 0.1113, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.42662116040955633, | |
| "eval_loss": 0.7458393573760986, | |
| "eval_runtime": 62.1803, | |
| "eval_samples_per_second": 15.873, | |
| "eval_steps_per_second": 1.994, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.47994880546075086, | |
| "grad_norm": 1.788787841796875, | |
| "learning_rate": 4.2000853242320814e-07, | |
| "loss": 0.1313, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.47994880546075086, | |
| "eval_loss": 0.7168156504631042, | |
| "eval_runtime": 62.1477, | |
| "eval_samples_per_second": 15.882, | |
| "eval_steps_per_second": 1.995, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5332764505119454, | |
| "grad_norm": 4.031916250823997e-05, | |
| "learning_rate": 4.1112059158134245e-07, | |
| "loss": 0.1649, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5332764505119454, | |
| "eval_loss": 0.7019046545028687, | |
| "eval_runtime": 62.2092, | |
| "eval_samples_per_second": 15.866, | |
| "eval_steps_per_second": 1.993, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.58660409556314, | |
| "grad_norm": 3.883213139488362e-06, | |
| "learning_rate": 4.0223265073947665e-07, | |
| "loss": 0.1245, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.58660409556314, | |
| "eval_loss": 0.6812178492546082, | |
| "eval_runtime": 62.492, | |
| "eval_samples_per_second": 15.794, | |
| "eval_steps_per_second": 1.984, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6399317406143344, | |
| "grad_norm": 80.17021942138672, | |
| "learning_rate": 3.933447098976109e-07, | |
| "loss": 0.1286, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6399317406143344, | |
| "eval_loss": 0.6501584649085999, | |
| "eval_runtime": 62.5236, | |
| "eval_samples_per_second": 15.786, | |
| "eval_steps_per_second": 1.983, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.693259385665529, | |
| "grad_norm": 0.0001418297761119902, | |
| "learning_rate": 3.8445676905574517e-07, | |
| "loss": 0.1076, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.693259385665529, | |
| "eval_loss": 0.6153982281684875, | |
| "eval_runtime": 62.2927, | |
| "eval_samples_per_second": 15.845, | |
| "eval_steps_per_second": 1.991, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7465870307167235, | |
| "grad_norm": 5.804526495012396e-07, | |
| "learning_rate": 3.755688282138794e-07, | |
| "loss": 0.1477, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7465870307167235, | |
| "eval_loss": 0.6118016839027405, | |
| "eval_runtime": 63.1588, | |
| "eval_samples_per_second": 15.627, | |
| "eval_steps_per_second": 1.963, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7999146757679181, | |
| "grad_norm": 2.66489315032959, | |
| "learning_rate": 3.6668088737201363e-07, | |
| "loss": 0.1315, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7999146757679181, | |
| "eval_loss": 0.601563036441803, | |
| "eval_runtime": 62.9163, | |
| "eval_samples_per_second": 15.688, | |
| "eval_steps_per_second": 1.971, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8532423208191127, | |
| "grad_norm": 851.2547607421875, | |
| "learning_rate": 3.5779294653014783e-07, | |
| "loss": 0.1413, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8532423208191127, | |
| "eval_loss": 0.584888756275177, | |
| "eval_runtime": 62.6735, | |
| "eval_samples_per_second": 15.748, | |
| "eval_steps_per_second": 1.979, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9065699658703071, | |
| "grad_norm": 6.041275628376752e-05, | |
| "learning_rate": 3.4890500568828214e-07, | |
| "loss": 0.124, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9065699658703071, | |
| "eval_loss": 0.5766209363937378, | |
| "eval_runtime": 63.3493, | |
| "eval_samples_per_second": 15.58, | |
| "eval_steps_per_second": 1.957, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9598976109215017, | |
| "grad_norm": 3.0901598930358887, | |
| "learning_rate": 3.4001706484641635e-07, | |
| "loss": 0.1215, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.9598976109215017, | |
| "eval_loss": 0.5558739304542542, | |
| "eval_runtime": 63.0862, | |
| "eval_samples_per_second": 15.645, | |
| "eval_steps_per_second": 1.966, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0132252559726962, | |
| "grad_norm": 2509.93212890625, | |
| "learning_rate": 3.311291240045506e-07, | |
| "loss": 0.131, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0132252559726962, | |
| "eval_loss": 0.563326358795166, | |
| "eval_runtime": 62.3607, | |
| "eval_samples_per_second": 15.827, | |
| "eval_steps_per_second": 1.988, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0665529010238908, | |
| "grad_norm": 32.05735778808594, | |
| "learning_rate": 3.2224118316268486e-07, | |
| "loss": 0.0348, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0665529010238908, | |
| "eval_loss": 0.553093671798706, | |
| "eval_runtime": 62.9895, | |
| "eval_samples_per_second": 15.669, | |
| "eval_steps_per_second": 1.969, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1198805460750854, | |
| "grad_norm": 0.004324838053435087, | |
| "learning_rate": 3.133532423208191e-07, | |
| "loss": 0.0687, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.1198805460750854, | |
| "eval_loss": 0.5494681000709534, | |
| "eval_runtime": 62.3211, | |
| "eval_samples_per_second": 15.837, | |
| "eval_steps_per_second": 1.99, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.17320819112628, | |
| "grad_norm": 0.09391242265701294, | |
| "learning_rate": 3.044653014789533e-07, | |
| "loss": 0.0749, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.17320819112628, | |
| "eval_loss": 0.5473881363868713, | |
| "eval_runtime": 62.3454, | |
| "eval_samples_per_second": 15.831, | |
| "eval_steps_per_second": 1.989, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.2265358361774745, | |
| "grad_norm": 1.9322777752850406e-11, | |
| "learning_rate": 2.955773606370876e-07, | |
| "loss": 0.1083, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.2265358361774745, | |
| "eval_loss": 0.541618287563324, | |
| "eval_runtime": 63.1868, | |
| "eval_samples_per_second": 15.62, | |
| "eval_steps_per_second": 1.962, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.2798634812286689, | |
| "grad_norm": 0.0002804399700835347, | |
| "learning_rate": 2.8668941979522184e-07, | |
| "loss": 0.0485, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.2798634812286689, | |
| "eval_loss": 0.5412191152572632, | |
| "eval_runtime": 62.9217, | |
| "eval_samples_per_second": 15.686, | |
| "eval_steps_per_second": 1.971, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3331911262798635, | |
| "grad_norm": 8.90180089641035e-08, | |
| "learning_rate": 2.778014789533561e-07, | |
| "loss": 0.0449, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.3331911262798635, | |
| "eval_loss": 0.5510581731796265, | |
| "eval_runtime": 62.525, | |
| "eval_samples_per_second": 15.786, | |
| "eval_steps_per_second": 1.983, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.386518771331058, | |
| "grad_norm": 1.0178096429905303e-12, | |
| "learning_rate": 2.689135381114903e-07, | |
| "loss": 0.0847, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.386518771331058, | |
| "eval_loss": 0.5492247939109802, | |
| "eval_runtime": 62.8314, | |
| "eval_samples_per_second": 15.709, | |
| "eval_steps_per_second": 1.974, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.4398464163822526, | |
| "grad_norm": 0.06238893046975136, | |
| "learning_rate": 2.600255972696246e-07, | |
| "loss": 0.0702, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.4398464163822526, | |
| "eval_loss": 0.5508909821510315, | |
| "eval_runtime": 63.08, | |
| "eval_samples_per_second": 15.647, | |
| "eval_steps_per_second": 1.966, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.493174061433447, | |
| "grad_norm": 3.558750449883519e-08, | |
| "learning_rate": 2.511376564277588e-07, | |
| "loss": 0.0487, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.493174061433447, | |
| "eval_loss": 0.544665515422821, | |
| "eval_runtime": 62.2772, | |
| "eval_samples_per_second": 15.849, | |
| "eval_steps_per_second": 1.991, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.5465017064846416, | |
| "grad_norm": 0.29742875695228577, | |
| "learning_rate": 2.422497155858931e-07, | |
| "loss": 0.072, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.5465017064846416, | |
| "eval_loss": 0.5490172505378723, | |
| "eval_runtime": 62.363, | |
| "eval_samples_per_second": 15.827, | |
| "eval_steps_per_second": 1.988, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.5998293515358362, | |
| "grad_norm": 664.765380859375, | |
| "learning_rate": 2.333617747440273e-07, | |
| "loss": 0.0325, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.5998293515358362, | |
| "eval_loss": 0.5443011522293091, | |
| "eval_runtime": 62.1023, | |
| "eval_samples_per_second": 15.893, | |
| "eval_steps_per_second": 1.997, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.6531569965870307, | |
| "grad_norm": 0.0024932425003498793, | |
| "learning_rate": 2.2447383390216154e-07, | |
| "loss": 0.0894, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.6531569965870307, | |
| "eval_loss": 0.5264189839363098, | |
| "eval_runtime": 62.2333, | |
| "eval_samples_per_second": 15.86, | |
| "eval_steps_per_second": 1.993, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.7064846416382253, | |
| "grad_norm": 1.1819661855697632, | |
| "learning_rate": 2.155858930602958e-07, | |
| "loss": 0.0503, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.7064846416382253, | |
| "eval_loss": 0.52089524269104, | |
| "eval_runtime": 62.2763, | |
| "eval_samples_per_second": 15.849, | |
| "eval_steps_per_second": 1.991, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.75981228668942, | |
| "grad_norm": 9.822515023572009e-11, | |
| "learning_rate": 2.0669795221843002e-07, | |
| "loss": 0.0568, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.75981228668942, | |
| "eval_loss": 0.5083423852920532, | |
| "eval_runtime": 63.6327, | |
| "eval_samples_per_second": 15.511, | |
| "eval_steps_per_second": 1.949, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.8131399317406145, | |
| "grad_norm": 9.34051013617987e-13, | |
| "learning_rate": 1.9781001137656428e-07, | |
| "loss": 0.0589, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.8131399317406145, | |
| "eval_loss": 0.5093010067939758, | |
| "eval_runtime": 62.0049, | |
| "eval_samples_per_second": 15.918, | |
| "eval_steps_per_second": 2.0, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.8664675767918089, | |
| "grad_norm": 0.00014494755305349827, | |
| "learning_rate": 1.889220705346985e-07, | |
| "loss": 0.0892, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.8664675767918089, | |
| "eval_loss": 0.4983443319797516, | |
| "eval_runtime": 62.0795, | |
| "eval_samples_per_second": 15.899, | |
| "eval_steps_per_second": 1.997, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.9197952218430034, | |
| "grad_norm": 1.963637918933614e-09, | |
| "learning_rate": 1.8003412969283277e-07, | |
| "loss": 0.0584, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.9197952218430034, | |
| "eval_loss": 0.48863324522972107, | |
| "eval_runtime": 63.6145, | |
| "eval_samples_per_second": 15.515, | |
| "eval_steps_per_second": 1.949, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.9731228668941978, | |
| "grad_norm": 386.5431823730469, | |
| "learning_rate": 1.7114618885096697e-07, | |
| "loss": 0.063, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.9731228668941978, | |
| "eval_loss": 0.4945463240146637, | |
| "eval_runtime": 62.2077, | |
| "eval_samples_per_second": 15.866, | |
| "eval_steps_per_second": 1.993, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.0264505119453924, | |
| "grad_norm": 0.10051131248474121, | |
| "learning_rate": 1.6225824800910123e-07, | |
| "loss": 0.0493, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.0264505119453924, | |
| "eval_loss": 0.4956463575363159, | |
| "eval_runtime": 62.1433, | |
| "eval_samples_per_second": 15.883, | |
| "eval_steps_per_second": 1.995, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.079778156996587, | |
| "grad_norm": 1.1283660569461063e-05, | |
| "learning_rate": 1.533703071672355e-07, | |
| "loss": 0.0246, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.079778156996587, | |
| "eval_loss": 0.48714280128479004, | |
| "eval_runtime": 63.7786, | |
| "eval_samples_per_second": 15.475, | |
| "eval_steps_per_second": 1.944, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.1331058020477816, | |
| "grad_norm": 2.3514085114584304e-06, | |
| "learning_rate": 1.4448236632536972e-07, | |
| "loss": 0.0385, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.1331058020477816, | |
| "eval_loss": 0.4830440580844879, | |
| "eval_runtime": 62.2754, | |
| "eval_samples_per_second": 15.849, | |
| "eval_steps_per_second": 1.991, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.186433447098976, | |
| "grad_norm": 0.045034199953079224, | |
| "learning_rate": 1.3559442548350398e-07, | |
| "loss": 0.0574, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.186433447098976, | |
| "eval_loss": 0.4848896861076355, | |
| "eval_runtime": 62.2583, | |
| "eval_samples_per_second": 15.853, | |
| "eval_steps_per_second": 1.992, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.2397610921501707, | |
| "grad_norm": 0.001391001744195819, | |
| "learning_rate": 1.267064846416382e-07, | |
| "loss": 0.039, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.2397610921501707, | |
| "eval_loss": 0.4871656596660614, | |
| "eval_runtime": 62.0998, | |
| "eval_samples_per_second": 15.894, | |
| "eval_steps_per_second": 1.997, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.2930887372013653, | |
| "grad_norm": 8.033894971504196e-08, | |
| "learning_rate": 1.1781854379977247e-07, | |
| "loss": 0.0653, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.2930887372013653, | |
| "eval_loss": 0.4837980270385742, | |
| "eval_runtime": 63.2915, | |
| "eval_samples_per_second": 15.595, | |
| "eval_steps_per_second": 1.959, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.34641638225256, | |
| "grad_norm": 2.780315089268015e-09, | |
| "learning_rate": 1.0893060295790671e-07, | |
| "loss": 0.0325, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.34641638225256, | |
| "eval_loss": 0.4875977635383606, | |
| "eval_runtime": 61.9485, | |
| "eval_samples_per_second": 15.933, | |
| "eval_steps_per_second": 2.002, | |
| "step": 22000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 28128, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7915403244422520.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |