PIXELZX/XION0.2-27B

1{2  "best_global_step": null,3  "best_metric": null,4  "best_model_checkpoint": null,5  "epoch": 3.0,6  "eval_steps": 22,7  "global_step": 66,8  "is_hyper_param_search": false,9  "is_local_process_zero": true,10  "is_world_process_zero": true,11  "log_history": [12    {13      "epoch": 0.045454545454545456,14      "grad_norm": 13.541594505310059,15      "learning_rate": 1e-05,16      "loss": 1.687255859375,17      "step": 118    },19    {20      "epoch": 0.09090909090909091,21      "grad_norm": 270.9620056152344,22      "learning_rate": 9.994336695915041e-06,23      "loss": 4.320972442626953,24      "step": 225    },26    {27      "epoch": 0.13636363636363635,28      "grad_norm": 8.045546531677246,29      "learning_rate": 9.977359612865424e-06,30      "loss": 0.990966796875,31      "step": 332    },33    {34      "epoch": 0.18181818181818182,35      "grad_norm": 4.051703453063965,36      "learning_rate": 9.949107209404664e-06,37      "loss": 1.09613037109375,38      "step": 439    },40    {41      "epoch": 0.22727272727272727,42      "grad_norm": 2.503697633743286,43      "learning_rate": 9.909643486313533e-06,44      "loss": 1.016815185546875,45      "step": 546    },47    {48      "epoch": 0.2727272727272727,49      "grad_norm": 3.1819581985473633,50      "learning_rate": 9.859057841617709e-06,51      "loss": 0.91766357421875,52      "step": 653    },54    {55      "epoch": 0.3181818181818182,56      "grad_norm": 1.810927391052246,57      "learning_rate": 9.797464868072489e-06,58      "loss": 0.9587249755859375,59      "step": 760    },61    {62      "epoch": 0.36363636363636365,63      "grad_norm": 1.6619356870651245,64      "learning_rate": 9.725004093573343e-06,65      "loss": 0.80902099609375,66      "step": 867    },68    {69      "epoch": 0.4090909090909091,70      "grad_norm": 2.2837491035461426,71      "learning_rate": 9.641839665080363e-06,72      "loss": 0.8079757690429688,73      "step": 974    },75    {76      "epoch": 0.45454545454545453,77      "grad_norm": 3.019440174102783,78      "learning_rate": 9.548159976772593e-06,79      "loss": 0.8288116455078125,80      "step": 1081    },82    {83      "epoch": 0.5,84      "grad_norm": 1.1921716928482056,85      "learning_rate": 9.444177243274619e-06,86      "loss": 0.6759376525878906,87      "step": 1188    },89    {90      "epoch": 0.5454545454545454,91      "grad_norm": 2.224130153656006,92      "learning_rate": 9.330127018922195e-06,93      "loss": 0.54034423828125,94      "step": 1295    },96    {97      "epoch": 0.5909090909090909,98      "grad_norm": 1.386547327041626,99      "learning_rate": 9.206267664155906e-06,100      "loss": 0.49675750732421875,101      "step": 13102    },103    {104      "epoch": 0.6363636363636364,105      "grad_norm": 1.4452524185180664,106      "learning_rate": 9.07287976025168e-06,107      "loss": 0.4326019287109375,108      "step": 14109    },110    {111      "epoch": 0.6818181818181818,112      "grad_norm": 2.8639578819274902,113      "learning_rate": 8.930265473713939e-06,114      "loss": 0.587005615234375,115      "step": 15116    },117    {118      "epoch": 0.7272727272727273,119      "grad_norm": 1.1509618759155273,120      "learning_rate": 8.778747871771293e-06,121      "loss": 0.49512481689453125,122      "step": 16123    },124    {125      "epoch": 0.7727272727272727,126      "grad_norm": 1.06486177444458,127      "learning_rate": 8.61867019052535e-06,128      "loss": 0.6640739440917969,129      "step": 17130    },131    {132      "epoch": 0.8181818181818182,133      "grad_norm": 0.9991989731788635,134      "learning_rate": 8.450395057410561e-06,135      "loss": 0.437957763671875,136      "step": 18137    },138    {139      "epoch": 0.8636363636363636,140      "grad_norm": 0.7819029092788696,141      "learning_rate": 8.274303669726427e-06,142      "loss": 0.5671710968017578,143      "step": 19144    },145    {146      "epoch": 0.9090909090909091,147      "grad_norm": 0.8972281813621521,148      "learning_rate": 8.090794931103026e-06,149      "loss": 0.3094291687011719,150      "step": 20151    },152    {153      "epoch": 0.9545454545454546,154      "grad_norm": 0.8230551481246948,155      "learning_rate": 7.900284547855992e-06,156      "loss": 0.42014312744140625,157      "step": 21158    },159    {160      "epoch": 1.0,161      "grad_norm": 0.9112458229064941,162      "learning_rate": 7.703204087277989e-06,163      "loss": 0.3252525329589844,164      "step": 22165    },166    {167      "epoch": 1.0,168      "eval_loss": 0.59442138671875,169      "eval_runtime": 13.685,170      "eval_samples_per_second": 0.365,171      "eval_steps_per_second": 0.073,172      "step": 22173    },174    {175      "checkpoint_runtime": 130.4115176    },177    {178      "epoch": 1.0454545454545454,179      "grad_norm": 1.6123707294464111,180      "learning_rate": 7.500000000000001e-06,181      "loss": 0.5166397094726562,182      "step": 23183    },184    {185      "epoch": 1.0909090909090908,186      "grad_norm": 1.3553637266159058,187      "learning_rate": 7.291132608637053e-06,188      "loss": 0.3297233581542969,189      "step": 24190    },191    {192      "epoch": 1.1363636363636362,193      "grad_norm": 1.3093994855880737,194      "learning_rate": 7.0770750650094335e-06,195      "loss": 0.49069786071777344,196      "step": 25197    },198    {199      "epoch": 1.1818181818181819,200      "grad_norm": 0.996208667755127,201      "learning_rate": 6.858312278301638e-06,202      "loss": 0.2602996826171875,203      "step": 26204    },205    {206      "epoch": 1.2272727272727273,207      "grad_norm": 0.8691992163658142,208      "learning_rate": 6.635339816587109e-06,209      "loss": 0.09621810913085938,210      "step": 27211    },212    {213      "epoch": 1.2727272727272727,214      "grad_norm": 0.73832106590271,215      "learning_rate": 6.408662784207149e-06,216      "loss": 0.2693767547607422,217      "step": 28218    },219    {220      "epoch": 1.3181818181818181,221      "grad_norm": 0.9352995753288269,222      "learning_rate": 6.178794677547138e-06,223      "loss": 0.2981605529785156,224      "step": 29225    },226    {227      "epoch": 1.3636363636363638,228      "grad_norm": 0.6079849004745483,229      "learning_rate": 5.946256221802052e-06,230      "loss": 0.26114463806152344,231      "step": 30232    },233    {234      "epoch": 1.4090909090909092,235      "grad_norm": 1.1613916158676147,236      "learning_rate": 5.711574191366427e-06,237      "loss": 0.3663311004638672,238      "step": 31239    },240    {241      "epoch": 1.4545454545454546,242      "grad_norm": 1.0265836715698242,243      "learning_rate": 5.475280216520913e-06,244      "loss": 0.13217544555664062,245      "step": 32246    },247    {248      "epoch": 1.5,249      "grad_norm": 0.9267441630363464,250      "learning_rate": 5.237909579118713e-06,251      "loss": 0.3060760498046875,252      "step": 33253    },254    {255      "epoch": 1.5454545454545454,256      "grad_norm": 0.6758596301078796,257      "learning_rate": 5e-06,258      "loss": 0.053920745849609375,259      "step": 34260    },261    {262      "epoch": 1.5909090909090908,263      "grad_norm": 0.6371091604232788,264      "learning_rate": 4.762090420881289e-06,265      "loss": 0.11317062377929688,266      "step": 35267    },268    {269      "epoch": 1.6363636363636362,270      "grad_norm": 0.6892620325088501,271      "learning_rate": 4.524719783479088e-06,272      "loss": 0.1634387969970703,273      "step": 36274    },275    {276      "epoch": 1.6818181818181817,277      "grad_norm": 1.267598032951355,278      "learning_rate": 4.2884258086335755e-06,279      "loss": 0.1157989501953125,280      "step": 37281    },282    {283      "epoch": 1.7272727272727273,284      "grad_norm": 0.618965208530426,285      "learning_rate": 4.053743778197951e-06,286      "loss": 0.2590446472167969,287      "step": 38288    },289    {290      "epoch": 1.7727272727272727,291      "grad_norm": 0.8626583814620972,292      "learning_rate": 3.821205322452863e-06,293      "loss": 0.4122934341430664,294      "step": 39295    },296    {297      "epoch": 1.8181818181818183,298      "grad_norm": 0.5675498843193054,299      "learning_rate": 3.5913372157928515e-06,300      "loss": 0.18222999572753906,301      "step": 40302    },303    {304      "epoch": 1.8636363636363638,305      "grad_norm": 1.346577525138855,306      "learning_rate": 3.3646601834128924e-06,307      "loss": 0.36645030975341797,308      "step": 41309    },310    {311      "epoch": 1.9090909090909092,312      "grad_norm": 0.5783951282501221,313      "learning_rate": 3.141687721698363e-06,314      "loss": 0.19093704223632812,315      "step": 42316    },317    {318      "epoch": 1.9545454545454546,319      "grad_norm": 0.7677918672561646,320      "learning_rate": 2.9229249349905686e-06,321      "loss": 0.24652671813964844,322      "step": 43323    },324    {325      "epoch": 2.0,326      "grad_norm": 0.6522639393806458,327      "learning_rate": 2.708867391362948e-06,328      "loss": 0.13533401489257812,329      "step": 44330    },331    {332      "epoch": 2.0,333      "eval_loss": 0.59344482421875,334      "eval_runtime": 4.3126,335      "eval_samples_per_second": 1.159,336      "eval_steps_per_second": 0.232,337      "step": 44338    },339    {340      "checkpoint_runtime": 134.2782341    },342    {343      "epoch": 2.0454545454545454,344      "grad_norm": 1.315116286277771,345      "learning_rate": 2.5000000000000015e-06,346      "loss": 0.2724342346191406,347      "step": 45348    },349    {350      "epoch": 2.090909090909091,351      "grad_norm": 0.5187699198722839,352      "learning_rate": 2.296795912722014e-06,353      "loss": 0.18209326267242432,354      "step": 46355    },356    {357      "epoch": 2.1363636363636362,358      "grad_norm": 0.9916004538536072,359      "learning_rate": 2.09971545214401e-06,360      "loss": 0.2614755630493164,361      "step": 47362    },363    {364      "epoch": 2.1818181818181817,365      "grad_norm": 0.5112076997756958,366      "learning_rate": 1.9092050688969736e-06,367      "loss": 0.1507863998413086,368      "step": 48369    },370    {371      "epoch": 2.227272727272727,372      "grad_norm": 0.3336804211139679,373      "learning_rate": 1.7256963302735752e-06,374      "loss": 0.024288177490234375,375      "step": 49376    },377    {378      "epoch": 2.2727272727272725,379      "grad_norm": 0.40631377696990967,380      "learning_rate": 1.549604942589441e-06,381      "loss": 0.19109153747558594,382      "step": 50383    },384    {385      "epoch": 2.3181818181818183,386      "grad_norm": 0.8899633288383484,387      "learning_rate": 1.3813298094746491e-06,388      "loss": 0.17161941528320312,389      "step": 51390    },391    {392      "epoch": 2.3636363636363638,393      "grad_norm": 0.5578643679618835,394      "learning_rate": 1.2212521282287093e-06,395      "loss": 0.17924213409423828,396      "step": 52397    },398    {399      "epoch": 2.409090909090909,400      "grad_norm": 1.057443618774414,401      "learning_rate": 1.0697345262860638e-06,402      "loss": 0.21059799194335938,403      "step": 53404    },405    {406      "epoch": 2.4545454545454546,407      "grad_norm": 0.6124624609947205,408      "learning_rate": 9.271202397483214e-07,409      "loss": 0.050838470458984375,410      "step": 54411    },412    {413      "epoch": 2.5,414      "grad_norm": 0.5138548016548157,415      "learning_rate": 7.937323358440935e-07,416      "loss": 0.22317028045654297,417      "step": 55418    },419    {420      "epoch": 2.5454545454545454,421      "grad_norm": 0.5870931148529053,422      "learning_rate": 6.698729810778065e-07,423      "loss": 0.02097320556640625,424      "step": 56425    },426    {427      "epoch": 2.590909090909091,428      "grad_norm": 0.47857600450515747,429      "learning_rate": 5.558227567253832e-07,430      "loss": 0.056339263916015625,431      "step": 57432    },433    {434      "epoch": 2.6363636363636362,435      "grad_norm": 0.4128285050392151,436      "learning_rate": 4.5184002322740784e-07,437      "loss": 0.10932445526123047,438      "step": 58439    },440    {441      "epoch": 2.6818181818181817,442      "grad_norm": 0.9088109135627747,443      "learning_rate": 3.581603349196372e-07,444      "loss": 0.04517364501953125,445      "step": 59446    },447    {448      "epoch": 2.7272727272727275,449      "grad_norm": 0.5057564973831177,450      "learning_rate": 2.7499590642665773e-07,451      "loss": 0.19766902923583984,452      "step": 60453    },454    {455      "epoch": 2.7727272727272725,456      "grad_norm": 0.792408287525177,457      "learning_rate": 2.0253513192751374e-07,458      "loss": 0.33954524993896484,459      "step": 61460    },461    {462      "epoch": 2.8181818181818183,463      "grad_norm": 0.47474050521850586,464      "learning_rate": 1.4094215838229176e-07,465      "loss": 0.12943077087402344,466      "step": 62467    },468    {469      "epoch": 2.8636363636363638,470      "grad_norm": 0.9088992476463318,471      "learning_rate": 9.035651368646647e-08,472      "loss": 0.3276538848876953,473      "step": 63474    },475    {476      "epoch": 2.909090909090909,477      "grad_norm": 0.4451668858528137,478      "learning_rate": 5.089279059533658e-08,479      "loss": 0.1589193344116211,480      "step": 64481    },482    {483      "epoch": 2.9545454545454546,484      "grad_norm": 0.6257097721099854,485      "learning_rate": 2.264038713457706e-08,486      "loss": 0.2027444839477539,487      "step": 65488    },489    {490      "epoch": 3.0,491      "grad_norm": 0.4987045228481293,492      "learning_rate": 5.6633040849601865e-09,493      "loss": 0.09051704406738281,494      "step": 66495    },496    {497      "epoch": 3.0,498      "eval_loss": 0.61431884765625,499      "eval_runtime": 4.3337,500      "eval_samples_per_second": 1.154,501      "eval_steps_per_second": 0.231,502      "step": 66503    }504  ],505  "logging_steps": 1.0,506  "max_steps": 66,507  "num_input_tokens_seen": 0,508  "num_train_epochs": 3,509  "save_steps": 22,510  "stateful_callbacks": {511    "TrainerControl": {512      "args": {513        "should_epoch_stop": false,514        "should_evaluate": false,515        "should_log": false,516        "should_save": true,517        "should_training_stop": true518      },519      "attributes": {}520    }521  },522  "total_flos": 8.409622391826678e+17,523  "train_batch_size": 1,524  "trial_name": null,525  "trial_params": null526}527