{
  "best_metric": 0.7917771801779077,
  "best_model_checkpoint": "/mnt/bn/chenhaobo-va-data/wangmingze/work_dir/RS2/v1.0_0.5b_2expertlayer_add_stage2/checkpoint-2500",
  "epoch": 0.9998716420647291,
  "eval_steps": 250,
  "global_step": 3408,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014669478316677363,
      "grad_norm": 0.0,
      "learning_rate": 4.854368932038836e-06,
      "loss": 0.7701,
      "step": 50
    },
    {
      "epoch": 0.029338956633354726,
      "grad_norm": 0.0,
      "learning_rate": 9.708737864077671e-06,
      "loss": 0.7258,
      "step": 100
    },
    {
      "epoch": 0.04400843495003209,
      "grad_norm": 0.0,
      "learning_rate": 9.995010921874192e-06,
      "loss": 0.6919,
      "step": 150
    },
    {
      "epoch": 0.05867791326670945,
      "grad_norm": 0.0,
      "learning_rate": 9.978761072117746e-06,
      "loss": 0.6531,
      "step": 200
    },
    {
      "epoch": 0.07334739158338681,
      "grad_norm": 0.0,
      "learning_rate": 9.951266820514078e-06,
      "loss": 0.6419,
      "step": 250
    },
    {
      "epoch": 0.07334739158338681,
      "eval_Bleu_1": 0.8169531364770959,
      "eval_Bleu_2": 0.7062374873836518,
      "eval_Bleu_3": 0.596860567007575,
      "eval_Bleu_4": 0.491661197404511,
      "eval_CIDEr": 1.3575382686490522,
      "eval_METEOR": 0.3647716618103635,
      "eval_ROUGE_L": 0.7326745249907421,
      "eval_Sm": 0.7366614132136673,
      "eval_runtime": 57.0866,
      "eval_samples_per_second": 33.791,
      "eval_steps_per_second": 4.239,
      "step": 250
    },
    {
      "epoch": 0.08801686990006417,
      "grad_norm": 0.0,
      "learning_rate": 9.91259026211235e-06,
      "loss": 0.6598,
      "step": 300
    },
    {
      "epoch": 0.10268634821674154,
      "grad_norm": 0.0,
      "learning_rate": 9.862818746909562e-06,
      "loss": 0.6676,
      "step": 350
    },
    {
      "epoch": 0.1173558265334189,
      "grad_norm": 0.0,
      "learning_rate": 9.802064682572885e-06,
      "loss": 0.656,
      "step": 400
    },
    {
      "epoch": 0.13202530485009628,
      "grad_norm": 0.0,
      "learning_rate": 9.730465280569857e-06,
      "loss": 0.6482,
      "step": 450
    },
    {
      "epoch": 0.14669478316677362,
      "grad_norm": 0.0,
      "learning_rate": 9.648182246279889e-06,
      "loss": 0.6443,
      "step": 500
    },
    {
      "epoch": 0.14669478316677362,
      "eval_Bleu_1": 0.8405638946544479,
      "eval_Bleu_2": 0.753658526473079,
      "eval_Bleu_3": 0.6722570876632427,
      "eval_Bleu_4": 0.6007924360982545,
      "eval_CIDEr": 1.3490450172101873,
      "eval_METEOR": 0.39011245456693056,
      "eval_ROUGE_L": 0.7481350696503057,
      "eval_Sm": 0.7720212443814196,
      "eval_runtime": 57.3785,
      "eval_samples_per_second": 33.619,
      "eval_steps_per_second": 4.218,
      "step": 500
    },
    {
      "epoch": 0.16136426148345098,
      "grad_norm": 0.0,
      "learning_rate": 9.555401413786872e-06,
      "loss": 0.6441,
      "step": 550
    },
    {
      "epoch": 0.17603373980012835,
      "grad_norm": 0.0,
      "learning_rate": 9.452332326177762e-06,
      "loss": 0.628,
      "step": 600
    },
    {
      "epoch": 0.1907032181168057,
      "grad_norm": 0.0,
      "learning_rate": 9.339207762294993e-06,
      "loss": 0.6225,
      "step": 650
    },
    {
      "epoch": 0.20537269643348308,
      "grad_norm": 0.0,
      "learning_rate": 9.21628321101153e-06,
      "loss": 0.6275,
      "step": 700
    },
    {
      "epoch": 0.22004217475016044,
      "grad_norm": 0.0,
      "learning_rate": 9.083836294215928e-06,
      "loss": 0.6193,
      "step": 750
    },
    {
      "epoch": 0.22004217475016044,
      "eval_Bleu_1": 0.8357740951145508,
      "eval_Bleu_2": 0.7348878179410974,
      "eval_Bleu_3": 0.6449140850517517,
      "eval_Bleu_4": 0.5652896325584327,
      "eval_CIDEr": 1.338564458509915,
      "eval_METEOR": 0.38648091177368565,
      "eval_ROUGE_L": 0.7400486871573863,
      "eval_Sm": 0.757595922499855,
      "eval_runtime": 55.9851,
      "eval_samples_per_second": 34.456,
      "eval_steps_per_second": 4.323,
      "step": 750
    },
    {
      "epoch": 0.2347116530668378,
      "grad_norm": 0.0,
      "learning_rate": 8.942166139810558e-06,
      "loss": 0.6161,
      "step": 800
    },
    {
      "epoch": 0.24938113138351517,
      "grad_norm": 0.0,
      "learning_rate": 8.791592706139029e-06,
      "loss": 0.608,
      "step": 850
    },
    {
      "epoch": 0.26405060970019256,
      "grad_norm": 0.0,
      "learning_rate": 8.63245605936865e-06,
      "loss": 0.6089,
      "step": 900
    },
    {
      "epoch": 0.2787200880168699,
      "grad_norm": 0.0,
      "learning_rate": 8.465115605459853e-06,
      "loss": 0.6008,
      "step": 950
    },
    {
      "epoch": 0.29338956633354724,
      "grad_norm": 0.0,
      "learning_rate": 8.289949278457234e-06,
      "loss": 0.6067,
      "step": 1000
    },
    {
      "epoch": 0.29338956633354724,
      "eval_Bleu_1": 0.8439012874997293,
      "eval_Bleu_2": 0.7577885636766362,
      "eval_Bleu_3": 0.6802456566151902,
      "eval_Bleu_4": 0.6162934805152959,
      "eval_CIDEr": 1.3433378767916668,
      "eval_METEOR": 0.39601512054991583,
      "eval_ROUGE_L": 0.7434272419402116,
      "eval_Sm": 0.7747684299492725,
      "eval_runtime": 58.4063,
      "eval_samples_per_second": 33.027,
      "eval_steps_per_second": 4.143,
      "step": 1000
    },
    {
      "epoch": 0.3080590446502246,
      "grad_norm": 0.0,
      "learning_rate": 8.10735268693537e-06,
      "loss": 0.5935,
      "step": 1050
    },
    {
      "epoch": 0.32272852296690197,
      "grad_norm": 0.0,
      "learning_rate": 7.9177382205272e-06,
      "loss": 0.5982,
      "step": 1100
    },
    {
      "epoch": 0.33739800128357933,
      "grad_norm": 0.0,
      "learning_rate": 7.7215341185528e-06,
      "loss": 0.5896,
      "step": 1150
    },
    {
      "epoch": 0.3520674796002567,
      "grad_norm": 0.0,
      "learning_rate": 7.519183502852059e-06,
      "loss": 0.5894,
      "step": 1200
    },
    {
      "epoch": 0.36673695791693406,
      "grad_norm": 0.0,
      "learning_rate": 7.311143377005595e-06,
      "loss": 0.5895,
      "step": 1250
    },
    {
      "epoch": 0.36673695791693406,
      "eval_Bleu_1": 0.8266465939028358,
      "eval_Bleu_2": 0.710305533232487,
      "eval_Bleu_3": 0.6009466460819967,
      "eval_Bleu_4": 0.4936679318924604,
      "eval_CIDEr": 1.3599632203211036,
      "eval_METEOR": 0.3750138198781453,
      "eval_ROUGE_L": 0.7407217686677534,
      "eval_Sm": 0.7423416851898657,
      "eval_runtime": 54.3677,
      "eval_samples_per_second": 35.481,
      "eval_steps_per_second": 4.451,
      "step": 1250
    },
    {
      "epoch": 0.3814064362336114,
      "grad_norm": 0.0,
      "learning_rate": 7.097883594204085e-06,
      "loss": 0.5828,
      "step": 1300
    },
    {
      "epoch": 0.3960759145502888,
      "grad_norm": 0.0,
      "learning_rate": 6.8798857960971146e-06,
      "loss": 0.5847,
      "step": 1350
    },
    {
      "epoch": 0.41074539286696615,
      "grad_norm": 0.0,
      "learning_rate": 6.657642325018093e-06,
      "loss": 0.5858,
      "step": 1400
    },
    {
      "epoch": 0.4254148711836435,
      "grad_norm": 0.0,
      "learning_rate": 6.4316551120419455e-06,
      "loss": 0.5813,
      "step": 1450
    },
    {
      "epoch": 0.4400843495003209,
      "grad_norm": 0.0,
      "learning_rate": 6.202434543386897e-06,
      "loss": 0.576,
      "step": 1500
    },
    {
      "epoch": 0.4400843495003209,
      "eval_Bleu_1": 0.8414514317798398,
      "eval_Bleu_2": 0.7384309037975365,
      "eval_Bleu_3": 0.6464269430105737,
      "eval_Bleu_4": 0.5645144727483805,
      "eval_CIDEr": 1.3688482781282445,
      "eval_METEOR": 0.39291224897822313,
      "eval_ROUGE_L": 0.7474174911662436,
      "eval_Sm": 0.768423122755273,
      "eval_runtime": 56.4566,
      "eval_samples_per_second": 34.168,
      "eval_steps_per_second": 4.286,
      "step": 1500
    },
    {
      "epoch": 0.45475382781699825,
      "grad_norm": 0.0,
      "learning_rate": 5.970498307720523e-06,
      "loss": 0.5751,
      "step": 1550
    },
    {
      "epoch": 0.4694233061336756,
      "grad_norm": 0.0,
      "learning_rate": 5.736370226973418e-06,
      "loss": 0.5739,
      "step": 1600
    },
    {
      "epoch": 0.484092784450353,
      "grad_norm": 0.0,
      "learning_rate": 5.500579073301046e-06,
      "loss": 0.5701,
      "step": 1650
    },
    {
      "epoch": 0.49876226276703034,
      "grad_norm": 0.0,
      "learning_rate": 5.263657374865634e-06,
      "loss": 0.5696,
      "step": 1700
    },
    {
      "epoch": 0.5134317410837077,
      "grad_norm": 0.0,
      "learning_rate": 5.026140213135231e-06,
      "loss": 0.565,
      "step": 1750
    },
    {
      "epoch": 0.5134317410837077,
      "eval_Bleu_1": 0.8646246048814135,
      "eval_Bleu_2": 0.7805335599133837,
      "eval_Bleu_3": 0.7063089691499992,
      "eval_Bleu_4": 0.6442486681073776,
      "eval_CIDEr": 1.3635428927748194,
      "eval_METEOR": 0.40681446738731897,
      "eval_ROUGE_L": 0.7515735652449509,
      "eval_Sm": 0.7915448983786166,
      "eval_runtime": 58.531,
      "eval_samples_per_second": 32.957,
      "eval_steps_per_second": 4.135,
      "step": 1750
    },
    {
      "epoch": 0.5281012194003851,
      "grad_norm": 0.0,
      "learning_rate": 4.788564014416187e-06,
      "loss": 0.5715,
      "step": 1800
    },
    {
      "epoch": 0.5427706977170624,
      "grad_norm": 0.0,
      "learning_rate": 4.551465338348348e-06,
      "loss": 0.5629,
      "step": 1850
    },
    {
      "epoch": 0.5574401760337399,
      "grad_norm": 0.0,
      "learning_rate": 4.31537966609912e-06,
      "loss": 0.5561,
      "step": 1900
    },
    {
      "epoch": 0.5721096543504172,
      "grad_norm": 0.0,
      "learning_rate": 4.080840190993248e-06,
      "loss": 0.5635,
      "step": 1950
    },
    {
      "epoch": 0.5867791326670945,
      "grad_norm": 0.0,
      "learning_rate": 3.848376614309611e-06,
      "loss": 0.5562,
      "step": 2000
    },
    {
      "epoch": 0.5867791326670945,
      "eval_Bleu_1": 0.8621401806286906,
      "eval_Bleu_2": 0.7698355783797655,
      "eval_Bleu_3": 0.6875894468539042,
      "eval_Bleu_4": 0.6167406606295135,
      "eval_CIDEr": 1.3869771604149415,
      "eval_METEOR": 0.4032516282402591,
      "eval_ROUGE_L": 0.7573257699480513,
      "eval_Sm": 0.7910738048081913,
      "eval_runtime": 64.025,
      "eval_samples_per_second": 30.129,
      "eval_steps_per_second": 3.78,
      "step": 2000
    },
    {
      "epoch": 0.6014486109837719,
      "grad_norm": 0.0,
      "learning_rate": 3.618513948964714e-06,
      "loss": 0.5573,
      "step": 2050
    },
    {
      "epoch": 0.6161180893004492,
      "grad_norm": 0.0,
      "learning_rate": 3.3917713337847404e-06,
      "loss": 0.5583,
      "step": 2100
    },
    {
      "epoch": 0.6307875676171266,
      "grad_norm": 0.0,
      "learning_rate": 3.168660861044061e-06,
      "loss": 0.555,
      "step": 2150
    },
    {
      "epoch": 0.6454570459338039,
      "grad_norm": 0.0,
      "learning_rate": 2.9496864199182094e-06,
      "loss": 0.5581,
      "step": 2200
    },
    {
      "epoch": 0.6601265242504814,
      "grad_norm": 0.0,
      "learning_rate": 2.7353425584633464e-06,
      "loss": 0.5513,
      "step": 2250
    },
    {
      "epoch": 0.6601265242504814,
      "eval_Bleu_1": 0.8503795445464875,
      "eval_Bleu_2": 0.7509102158542702,
      "eval_Bleu_3": 0.6589892932186134,
      "eval_Bleu_4": 0.578193091773488,
      "eval_CIDEr": 1.3746830539156165,
      "eval_METEOR": 0.40021882792006797,
      "eval_ROUGE_L": 0.7539686207703685,
      "eval_Sm": 0.7767658985948852,
      "eval_runtime": 59.1294,
      "eval_samples_per_second": 32.623,
      "eval_steps_per_second": 4.093,
      "step": 2250
    },
    {
      "epoch": 0.6747960025671587,
      "grad_norm": 0.0,
      "learning_rate": 2.526113366692381e-06,
      "loss": 0.5501,
      "step": 2300
    },
    {
      "epoch": 0.6894654808838361,
      "grad_norm": 0.0,
      "learning_rate": 2.3224713832703283e-06,
      "loss": 0.5463,
      "step": 2350
    },
    {
      "epoch": 0.7041349592005134,
      "grad_norm": 0.0,
      "learning_rate": 2.1248765282980715e-06,
      "loss": 0.5428,
      "step": 2400
    },
    {
      "epoch": 0.7188044375171908,
      "grad_norm": 0.0,
      "learning_rate": 1.9337750645948257e-06,
      "loss": 0.5474,
      "step": 2450
    },
    {
      "epoch": 0.7334739158338681,
      "grad_norm": 0.0,
      "learning_rate": 1.7495985898252244e-06,
      "loss": 0.5405,
      "step": 2500
    },
    {
      "epoch": 0.7334739158338681,
      "eval_Bleu_1": 0.8642067420940046,
      "eval_Bleu_2": 0.7708274701376591,
      "eval_Bleu_3": 0.6848356904871902,
      "eval_Bleu_4": 0.610421078936172,
      "eval_CIDEr": 1.3919047346615983,
      "eval_METEOR": 0.40507692559924574,
      "eval_ROUGE_L": 0.7597059815146143,
      "eval_Sm": 0.7917771801779077,
      "eval_runtime": 58.1719,
      "eval_samples_per_second": 33.16,
      "eval_steps_per_second": 4.16,
      "step": 2500
    },
    {
      "epoch": 0.7481433941505455,
      "grad_norm": 0.0,
      "learning_rate": 1.5727630617472505e-06,
      "loss": 0.5413,
      "step": 2550
    },
    {
      "epoch": 0.7628128724672228,
      "grad_norm": 0.0,
      "learning_rate": 1.4036678587825181e-06,
      "loss": 0.5416,
      "step": 2600
    },
    {
      "epoch": 0.7774823507839003,
      "grad_norm": 0.0,
      "learning_rate": 1.242694878030537e-06,
      "loss": 0.539,
      "step": 2650
    },
    {
      "epoch": 0.7921518291005776,
      "grad_norm": 0.0,
      "learning_rate": 1.0902076727640748e-06,
      "loss": 0.5424,
      "step": 2700
    },
    {
      "epoch": 0.806821307417255,
      "grad_norm": 0.0,
      "learning_rate": 9.465506313536e-07,
      "loss": 0.5398,
      "step": 2750
    },
    {
      "epoch": 0.806821307417255,
      "eval_Bleu_1": 0.8440182071199624,
      "eval_Bleu_2": 0.7453206956121321,
      "eval_Bleu_3": 0.6567459444563346,
      "eval_Bleu_4": 0.5802106326806715,
      "eval_CIDEr": 1.3639134120747893,
      "eval_METEOR": 0.3938719364603724,
      "eval_ROUGE_L": 0.7522720619431441,
      "eval_Sm": 0.7725670107897443,
      "eval_runtime": 57.2843,
      "eval_samples_per_second": 33.674,
      "eval_steps_per_second": 4.225,
      "step": 2750
    },
    {
      "epoch": 0.8214907857339323,
      "grad_norm": 0.0,
      "learning_rate": 8.120481994751395e-07,
      "loss": 0.5365,
      "step": 2800
    },
    {
      "epoch": 0.8361602640506097,
      "grad_norm": 0.0,
      "learning_rate": 6.870041473581829e-07,
      "loss": 0.5339,
      "step": 2850
    },
    {
      "epoch": 0.850829742367287,
      "grad_norm": 0.0,
      "learning_rate": 5.717008837285643e-07,
      "loss": 0.5403,
      "step": 2900
    },
    {
      "epoch": 0.8654992206839645,
      "grad_norm": 0.0,
      "learning_rate": 4.663988179957296e-07,
      "loss": 0.5384,
      "step": 2950
    },
    {
      "epoch": 0.8801686990006418,
      "grad_norm": 0.0,
      "learning_rate": 3.713357721248778e-07,
      "loss": 0.5405,
      "step": 3000
    },
    {
      "epoch": 0.8801686990006418,
      "eval_Bleu_1": 0.8414596679328293,
      "eval_Bleu_2": 0.7401714139842944,
      "eval_Bleu_3": 0.6473427959433954,
      "eval_Bleu_4": 0.5629518243939922,
      "eval_CIDEr": 1.3640353499862266,
      "eval_METEOR": 0.39264881149023945,
      "eval_ROUGE_L": 0.7533214693096227,
      "eval_Sm": 0.7682393637950202,
      "eval_runtime": 56.0109,
      "eval_samples_per_second": 34.44,
      "eval_steps_per_second": 4.321,
      "step": 3000
    },
    {
      "epoch": 0.8948381773173192,
      "grad_norm": 0.0,
      "learning_rate": 2.867264435222733e-07,
      "loss": 0.5424,
      "step": 3050
    },
    {
      "epoch": 0.9095076556339965,
      "grad_norm": 0.0,
      "learning_rate": 2.1276192014675956e-07,
      "loss": 0.5353,
      "step": 3100
    },
    {
      "epoch": 0.9241771339506739,
      "grad_norm": 0.0,
      "learning_rate": 1.4960924894258988e-07,
      "loss": 0.5386,
      "step": 3150
    },
    {
      "epoch": 0.9388466122673512,
      "grad_norm": 0.0,
      "learning_rate": 9.741105856827193e-08,
      "loss": 0.5349,
      "step": 3200
    },
    {
      "epoch": 0.9535160905840286,
      "grad_norm": 0.0,
      "learning_rate": 5.6285237273466245e-08,
      "loss": 0.5394,
      "step": 3250
    },
    {
      "epoch": 0.9535160905840286,
      "eval_Bleu_1": 0.8448620121122421,
      "eval_Bleu_2": 0.7411775242513393,
      "eval_Bleu_3": 0.6479512210204962,
      "eval_Bleu_4": 0.5639978439699083,
      "eval_CIDEr": 1.3674205674001598,
      "eval_METEOR": 0.393529859811792,
      "eval_ROUGE_L": 0.7542710559497389,
      "eval_Sm": 0.7698048317828997,
      "eval_runtime": 60.7238,
      "eval_samples_per_second": 31.767,
      "eval_steps_per_second": 3.985,
      "step": 3250
    },
    {
      "epoch": 0.968185568900706,
      "grad_norm": 0.0,
      "learning_rate": 2.632466665145106e-08,
      "loss": 0.5362,
      "step": 3300
    },
    {
      "epoch": 0.9828550472173834,
      "grad_norm": 0.0,
      "learning_rate": 7.597011868473725e-09,
      "loss": 0.5397,
      "step": 3350
    },
    {
      "epoch": 0.9975245255340607,
      "grad_norm": 0.0,
      "learning_rate": 1.4456884373625913e-10,
      "loss": 0.5335,
      "step": 3400
    },
    {
      "epoch": 0.9998716420647291,
      "step": 3408,
      "total_flos": 3.085365241802916e+17,
      "train_loss": 0.5832489245514355,
      "train_runtime": 15945.5719,
      "train_samples_per_second": 54.72,
      "train_steps_per_second": 0.214
    }
  ],
  "logging_steps": 50,
  "max_steps": 3408,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 250,
  "total_flos": 3.085365241802916e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}