| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.009741248097412482, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 4.8706240487062404e-05, | |
| "grad_norm": 5.851158142089844, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 2.1245, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 9.741248097412481e-05, | |
| "grad_norm": 18.33625602722168, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 3.869, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00014611872146118722, | |
| "grad_norm": 15.931517601013184, | |
| "learning_rate": 3e-06, | |
| "loss": 3.1648, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00019482496194824962, | |
| "grad_norm": 26.00968360900879, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.3093, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.000243531202435312, | |
| "grad_norm": 27.92300033569336, | |
| "learning_rate": 5e-06, | |
| "loss": 2.221, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00029223744292237444, | |
| "grad_norm": 9.341499328613281, | |
| "learning_rate": 6e-06, | |
| "loss": 2.2322, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0003409436834094368, | |
| "grad_norm": 13.460984230041504, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 2.5544, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.00038964992389649923, | |
| "grad_norm": 12.86645793914795, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.537, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00043835616438356166, | |
| "grad_norm": 15.244658470153809, | |
| "learning_rate": 9e-06, | |
| "loss": 2.7317, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.000487062404870624, | |
| "grad_norm": 11.827373504638672, | |
| "learning_rate": 1e-05, | |
| "loss": 2.2533, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0005357686453576865, | |
| "grad_norm": 10.726211547851562, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 2.2206, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0005844748858447489, | |
| "grad_norm": 13.430566787719727, | |
| "learning_rate": 1.2e-05, | |
| "loss": 2.4726, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0006331811263318112, | |
| "grad_norm": 15.749054908752441, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 2.3328, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0006818873668188736, | |
| "grad_norm": 13.098236083984375, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 2.3669, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0007305936073059361, | |
| "grad_norm": 14.811712265014648, | |
| "learning_rate": 1.5e-05, | |
| "loss": 2.3254, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0007792998477929985, | |
| "grad_norm": 13.960674285888672, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.3478, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0008280060882800608, | |
| "grad_norm": 13.510299682617188, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 2.3439, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0008767123287671233, | |
| "grad_norm": 21.18514060974121, | |
| "learning_rate": 1.8e-05, | |
| "loss": 2.6721, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0009254185692541857, | |
| "grad_norm": 17.1453800201416, | |
| "learning_rate": 1.9e-05, | |
| "loss": 2.1923, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.000974124809741248, | |
| "grad_norm": 15.410392761230469, | |
| "learning_rate": 2e-05, | |
| "loss": 2.4277, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0010228310502283105, | |
| "grad_norm": 17.85428810119629, | |
| "learning_rate": 2.1e-05, | |
| "loss": 2.1419, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.001071537290715373, | |
| "grad_norm": 16.322954177856445, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 2.4326, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0011202435312024353, | |
| "grad_norm": 18.91599464416504, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 2.1728, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0011689497716894977, | |
| "grad_norm": 14.737173080444336, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.2486, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0012176560121765602, | |
| "grad_norm": 14.996482849121094, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.9959, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0012663622526636225, | |
| "grad_norm": 12.736573219299316, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 1.8652, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.001315068493150685, | |
| "grad_norm": 13.891550064086914, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 2.3333, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0013637747336377472, | |
| "grad_norm": 11.90274429321289, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.8707, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0014124809741248097, | |
| "grad_norm": 16.16830062866211, | |
| "learning_rate": 2.9e-05, | |
| "loss": 1.9199, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0014611872146118722, | |
| "grad_norm": 12.124021530151367, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8475, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0015098934550989344, | |
| "grad_norm": 13.702016830444336, | |
| "learning_rate": 3.1e-05, | |
| "loss": 1.6096, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.001558599695585997, | |
| "grad_norm": 16.256940841674805, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.0673, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0016073059360730594, | |
| "grad_norm": 11.912320137023926, | |
| "learning_rate": 3.3e-05, | |
| "loss": 1.5261, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0016560121765601217, | |
| "grad_norm": 12.158217430114746, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.566, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0017047184170471841, | |
| "grad_norm": 11.484949111938477, | |
| "learning_rate": 3.5e-05, | |
| "loss": 2.0858, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0017534246575342466, | |
| "grad_norm": 12.884050369262695, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.0508, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0018021308980213089, | |
| "grad_norm": 14.054976463317871, | |
| "learning_rate": 3.7e-05, | |
| "loss": 1.2539, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0018508371385083714, | |
| "grad_norm": 13.9093599319458, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.5044, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0018995433789954338, | |
| "grad_norm": 42.75831604003906, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 1.2133, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.001948249619482496, | |
| "grad_norm": 13.14990234375, | |
| "learning_rate": 4e-05, | |
| "loss": 1.4692, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0019969558599695586, | |
| "grad_norm": 11.442179679870605, | |
| "learning_rate": 4.1e-05, | |
| "loss": 1.1273, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.002045662100456621, | |
| "grad_norm": 9.139272689819336, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.6377, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0020943683409436835, | |
| "grad_norm": 9.682751655578613, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.8092, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.002143074581430746, | |
| "grad_norm": 10.700410842895508, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.9342, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.002191780821917808, | |
| "grad_norm": 8.799379348754883, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.6458, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0022404870624048705, | |
| "grad_norm": 7.537528038024902, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.3726, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.002289193302891933, | |
| "grad_norm": 9.551314353942871, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.6441, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.0023378995433789955, | |
| "grad_norm": 13.743586540222168, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.0114, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.002386605783866058, | |
| "grad_norm": 9.87704849243164, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.5438, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.0024353120243531205, | |
| "grad_norm": 20.906862258911133, | |
| "learning_rate": 5e-05, | |
| "loss": 1.8715, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0024840182648401825, | |
| "grad_norm": 8.490747451782227, | |
| "learning_rate": 5.1000000000000006e-05, | |
| "loss": 2.5756, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.002532724505327245, | |
| "grad_norm": 12.285213470458984, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 2.3924, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0025814307458143075, | |
| "grad_norm": 15.3045015335083, | |
| "learning_rate": 5.300000000000001e-05, | |
| "loss": 1.8956, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.00263013698630137, | |
| "grad_norm": 15.969414710998535, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.8218, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0026788432267884324, | |
| "grad_norm": 13.07938003540039, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.8747, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0027275494672754945, | |
| "grad_norm": 9.398449897766113, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.8304, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.002776255707762557, | |
| "grad_norm": 7.938607215881348, | |
| "learning_rate": 5.6999999999999996e-05, | |
| "loss": 1.9411, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0028249619482496194, | |
| "grad_norm": 9.889793395996094, | |
| "learning_rate": 5.8e-05, | |
| "loss": 2.0788, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.002873668188736682, | |
| "grad_norm": 7.4011101722717285, | |
| "learning_rate": 5.9e-05, | |
| "loss": 1.954, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0029223744292237444, | |
| "grad_norm": 7.340896129608154, | |
| "learning_rate": 6e-05, | |
| "loss": 1.9234, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.002971080669710807, | |
| "grad_norm": 13.956856727600098, | |
| "learning_rate": 6.1e-05, | |
| "loss": 2.3137, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.003019786910197869, | |
| "grad_norm": 12.846822738647461, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.5154, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.0030684931506849314, | |
| "grad_norm": 10.943364143371582, | |
| "learning_rate": 6.3e-05, | |
| "loss": 1.9186, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.003117199391171994, | |
| "grad_norm": 9.687166213989258, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.5523, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.0031659056316590563, | |
| "grad_norm": 9.057082176208496, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.9528, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.003214611872146119, | |
| "grad_norm": 12.989787101745605, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.9981, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.0032633181126331813, | |
| "grad_norm": 12.194509506225586, | |
| "learning_rate": 6.7e-05, | |
| "loss": 2.0458, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.0033120243531202433, | |
| "grad_norm": 14.83133316040039, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 2.1763, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.003360730593607306, | |
| "grad_norm": 12.523411750793457, | |
| "learning_rate": 6.9e-05, | |
| "loss": 2.1116, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0034094368340943683, | |
| "grad_norm": 10.275344848632812, | |
| "learning_rate": 7e-05, | |
| "loss": 2.2137, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0034581430745814308, | |
| "grad_norm": 11.111023902893066, | |
| "learning_rate": 7.1e-05, | |
| "loss": 2.3323, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.0035068493150684932, | |
| "grad_norm": 11.215889930725098, | |
| "learning_rate": 7.2e-05, | |
| "loss": 2.2838, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.0035555555555555557, | |
| "grad_norm": 11.466020584106445, | |
| "learning_rate": 7.3e-05, | |
| "loss": 2.3188, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.0036042617960426178, | |
| "grad_norm": 12.254678726196289, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.899, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.0036529680365296802, | |
| "grad_norm": 12.059733390808105, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.8779, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.0037016742770167427, | |
| "grad_norm": 14.260489463806152, | |
| "learning_rate": 7.6e-05, | |
| "loss": 1.941, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.003750380517503805, | |
| "grad_norm": 11.856407165527344, | |
| "learning_rate": 7.7e-05, | |
| "loss": 1.8031, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.0037990867579908677, | |
| "grad_norm": 13.23192024230957, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 1.8466, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.00384779299847793, | |
| "grad_norm": 14.03378677368164, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 2.0481, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.003896499238964992, | |
| "grad_norm": 12.832358360290527, | |
| "learning_rate": 8e-05, | |
| "loss": 2.223, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.003945205479452055, | |
| "grad_norm": 11.655765533447266, | |
| "learning_rate": 8.1e-05, | |
| "loss": 1.6975, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.003993911719939117, | |
| "grad_norm": 13.00943660736084, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.6346, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.00404261796042618, | |
| "grad_norm": 13.812478065490723, | |
| "learning_rate": 8.3e-05, | |
| "loss": 1.818, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.004091324200913242, | |
| "grad_norm": 10.820585250854492, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.4317, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.004140030441400305, | |
| "grad_norm": 11.540773391723633, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.6072, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.004188736681887367, | |
| "grad_norm": 12.939353942871094, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.5664, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.0042374429223744296, | |
| "grad_norm": 13.942463874816895, | |
| "learning_rate": 8.7e-05, | |
| "loss": 1.6459, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.004286149162861492, | |
| "grad_norm": 10.26823902130127, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.2702, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.004334855403348554, | |
| "grad_norm": 11.739928245544434, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 1.4725, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.004383561643835616, | |
| "grad_norm": 9.966146469116211, | |
| "learning_rate": 9e-05, | |
| "loss": 0.7986, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.004432267884322679, | |
| "grad_norm": 12.615833282470703, | |
| "learning_rate": 9.1e-05, | |
| "loss": 1.2814, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.004480974124809741, | |
| "grad_norm": 10.074495315551758, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 0.6983, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.0045296803652968036, | |
| "grad_norm": 14.656659126281738, | |
| "learning_rate": 9.300000000000001e-05, | |
| "loss": 0.8942, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.004578386605783866, | |
| "grad_norm": 8.69896411895752, | |
| "learning_rate": 9.4e-05, | |
| "loss": 0.6231, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0046270928462709285, | |
| "grad_norm": 9.49130630493164, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.5908, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.004675799086757991, | |
| "grad_norm": 10.93470287322998, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.9713, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.0047245053272450535, | |
| "grad_norm": 14.361600875854492, | |
| "learning_rate": 9.7e-05, | |
| "loss": 1.1417, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.004773211567732116, | |
| "grad_norm": 10.079813003540039, | |
| "learning_rate": 9.8e-05, | |
| "loss": 0.8555, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.004821917808219178, | |
| "grad_norm": 10.335041999816895, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 0.7722, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.004870624048706241, | |
| "grad_norm": 16.433164596557617, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3996, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0049193302891933025, | |
| "grad_norm": 10.94658374786377, | |
| "learning_rate": 9.997532801828658e-05, | |
| "loss": 2.4347, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.004968036529680365, | |
| "grad_norm": 17.193456649780273, | |
| "learning_rate": 9.990133642141359e-05, | |
| "loss": 3.2929, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.0050167427701674275, | |
| "grad_norm": 12.511935234069824, | |
| "learning_rate": 9.977809823015401e-05, | |
| "loss": 1.2897, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.00506544901065449, | |
| "grad_norm": 10.709287643432617, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 1.7389, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0051141552511415524, | |
| "grad_norm": 12.061141014099121, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 1.0578, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.005162861491628615, | |
| "grad_norm": 8.664139747619629, | |
| "learning_rate": 9.911436253643445e-05, | |
| "loss": 1.5287, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.005211567732115677, | |
| "grad_norm": 8.517009735107422, | |
| "learning_rate": 9.879583809693738e-05, | |
| "loss": 1.5803, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.00526027397260274, | |
| "grad_norm": 7.50566291809082, | |
| "learning_rate": 9.842915805643155e-05, | |
| "loss": 1.6515, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.005308980213089802, | |
| "grad_norm": 7.030799388885498, | |
| "learning_rate": 9.801468428384716e-05, | |
| "loss": 1.6422, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.005357686453576865, | |
| "grad_norm": 8.954855918884277, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.0945, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.005406392694063927, | |
| "grad_norm": 11.065245628356934, | |
| "learning_rate": 9.704403844771128e-05, | |
| "loss": 1.7066, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.005455098934550989, | |
| "grad_norm": 8.86803150177002, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 1.2714, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.005503805175038051, | |
| "grad_norm": 8.038043022155762, | |
| "learning_rate": 9.588773128419906e-05, | |
| "loss": 1.3595, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.005552511415525114, | |
| "grad_norm": 7.1317877769470215, | |
| "learning_rate": 9.524135262330098e-05, | |
| "loss": 1.012, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.005601217656012176, | |
| "grad_norm": 9.007568359375, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 1.7679, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.005649923896499239, | |
| "grad_norm": 9.552806854248047, | |
| "learning_rate": 9.381533400219318e-05, | |
| "loss": 1.6309, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.005698630136986301, | |
| "grad_norm": 13.193597793579102, | |
| "learning_rate": 9.30371013501972e-05, | |
| "loss": 1.8146, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.005747336377473364, | |
| "grad_norm": 10.52649974822998, | |
| "learning_rate": 9.221639627510076e-05, | |
| "loss": 2.1017, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.005796042617960426, | |
| "grad_norm": 13.295526504516602, | |
| "learning_rate": 9.135402871372808e-05, | |
| "loss": 1.8865, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.005844748858447489, | |
| "grad_norm": 11.041900634765625, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.8419, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.005893455098934551, | |
| "grad_norm": 12.388705253601074, | |
| "learning_rate": 8.950775061878453e-05, | |
| "loss": 2.1317, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.005942161339421614, | |
| "grad_norm": 10.880697250366211, | |
| "learning_rate": 8.852566213878947e-05, | |
| "loss": 2.2668, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.005990867579908676, | |
| "grad_norm": 9.947246551513672, | |
| "learning_rate": 8.750555348152298e-05, | |
| "loss": 2.0057, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.006039573820395738, | |
| "grad_norm": 11.419881820678711, | |
| "learning_rate": 8.644843137107059e-05, | |
| "loss": 2.2503, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.0060882800608828, | |
| "grad_norm": 17.00235939025879, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 1.7933, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.006136986301369863, | |
| "grad_norm": 12.406261444091797, | |
| "learning_rate": 8.422735529643444e-05, | |
| "loss": 2.0982, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.006185692541856925, | |
| "grad_norm": 11.130414009094238, | |
| "learning_rate": 8.306559326618259e-05, | |
| "loss": 2.1638, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.006234398782343988, | |
| "grad_norm": 12.033727645874023, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 2.3324, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.00628310502283105, | |
| "grad_norm": 10.567495346069336, | |
| "learning_rate": 8.064535268264883e-05, | |
| "loss": 2.0464, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.006331811263318113, | |
| "grad_norm": 11.884528160095215, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.7445, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.006380517503805175, | |
| "grad_norm": 11.059581756591797, | |
| "learning_rate": 7.810416889260653e-05, | |
| "loss": 1.9192, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.006429223744292238, | |
| "grad_norm": 11.430746078491211, | |
| "learning_rate": 7.679133974894983e-05, | |
| "loss": 1.9024, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.0064779299847793, | |
| "grad_norm": 10.843358039855957, | |
| "learning_rate": 7.545207078751857e-05, | |
| "loss": 1.882, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.006526636225266363, | |
| "grad_norm": 12.441071510314941, | |
| "learning_rate": 7.408768370508576e-05, | |
| "loss": 1.2063, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.006575342465753425, | |
| "grad_norm": 12.600017547607422, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 1.7018, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.006624048706240487, | |
| "grad_norm": 10.350911140441895, | |
| "learning_rate": 7.128896457825364e-05, | |
| "loss": 1.0953, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.006672754946727549, | |
| "grad_norm": 11.342569351196289, | |
| "learning_rate": 6.985739453173903e-05, | |
| "loss": 1.2846, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.006721461187214612, | |
| "grad_norm": 9.398454666137695, | |
| "learning_rate": 6.840622763423391e-05, | |
| "loss": 1.0346, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.006770167427701674, | |
| "grad_norm": 8.416460990905762, | |
| "learning_rate": 6.693689601226458e-05, | |
| "loss": 0.8632, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.006818873668188737, | |
| "grad_norm": 9.184191703796387, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.0037, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.006867579908675799, | |
| "grad_norm": 11.087357521057129, | |
| "learning_rate": 6.394955530196147e-05, | |
| "loss": 1.1403, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.0069162861491628615, | |
| "grad_norm": 11.673155784606934, | |
| "learning_rate": 6.243449435824276e-05, | |
| "loss": 1.3567, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.006964992389649924, | |
| "grad_norm": 9.134658813476562, | |
| "learning_rate": 6.090716206982714e-05, | |
| "loss": 0.9909, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.0070136986301369865, | |
| "grad_norm": 7.556820869445801, | |
| "learning_rate": 5.9369065729286245e-05, | |
| "loss": 0.6842, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.007062404870624049, | |
| "grad_norm": 8.291131973266602, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 0.682, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.0071111111111111115, | |
| "grad_norm": 7.839075088500977, | |
| "learning_rate": 5.6266661678215216e-05, | |
| "loss": 0.6249, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.007159817351598174, | |
| "grad_norm": 8.085784912109375, | |
| "learning_rate": 5.470541566592573e-05, | |
| "loss": 0.7631, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.0072085235920852355, | |
| "grad_norm": 10.273898124694824, | |
| "learning_rate": 5.313952597646568e-05, | |
| "loss": 1.0338, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.007257229832572298, | |
| "grad_norm": 7.386903285980225, | |
| "learning_rate": 5.157053795390642e-05, | |
| "loss": 0.4861, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.0073059360730593605, | |
| "grad_norm": 14.1038236618042, | |
| "learning_rate": 5e-05, | |
| "loss": 1.4195, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.007354642313546423, | |
| "grad_norm": 6.475687026977539, | |
| "learning_rate": 4.8429462046093585e-05, | |
| "loss": 2.2988, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.0074033485540334855, | |
| "grad_norm": 12.853471755981445, | |
| "learning_rate": 4.6860474023534335e-05, | |
| "loss": 2.8033, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.007452054794520548, | |
| "grad_norm": 9.609049797058105, | |
| "learning_rate": 4.529458433407429e-05, | |
| "loss": 1.9798, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.00750076103500761, | |
| "grad_norm": 6.795541286468506, | |
| "learning_rate": 4.373333832178478e-05, | |
| "loss": 1.5672, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.007549467275494673, | |
| "grad_norm": 6.115726470947266, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 1.0559, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.007598173515981735, | |
| "grad_norm": 6.5415239334106445, | |
| "learning_rate": 4.063093427071376e-05, | |
| "loss": 1.0537, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.007646879756468798, | |
| "grad_norm": 10.223687171936035, | |
| "learning_rate": 3.9092837930172884e-05, | |
| "loss": 1.9506, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.00769558599695586, | |
| "grad_norm": 11.674057006835938, | |
| "learning_rate": 3.756550564175727e-05, | |
| "loss": 1.8997, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.007744292237442923, | |
| "grad_norm": 7.180853843688965, | |
| "learning_rate": 3.605044469803854e-05, | |
| "loss": 1.6581, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.007792998477929984, | |
| "grad_norm": 9.071992874145508, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.4177, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.007841704718417048, | |
| "grad_norm": 7.320943355560303, | |
| "learning_rate": 3.3063103987735433e-05, | |
| "loss": 1.5026, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.00789041095890411, | |
| "grad_norm": 5.533285617828369, | |
| "learning_rate": 3.1593772365766105e-05, | |
| "loss": 0.8148, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.007939117199391173, | |
| "grad_norm": 7.4456610679626465, | |
| "learning_rate": 3.0142605468260978e-05, | |
| "loss": 1.2012, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.007987823439878234, | |
| "grad_norm": 7.7538862228393555, | |
| "learning_rate": 2.8711035421746367e-05, | |
| "loss": 2.0698, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.008036529680365296, | |
| "grad_norm": 8.773526191711426, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 2.6465, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.00808523592085236, | |
| "grad_norm": 9.068692207336426, | |
| "learning_rate": 2.591231629491423e-05, | |
| "loss": 2.0785, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.008133942161339421, | |
| "grad_norm": 11.246406555175781, | |
| "learning_rate": 2.4547929212481435e-05, | |
| "loss": 2.3037, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.008182648401826484, | |
| "grad_norm": 10.597480773925781, | |
| "learning_rate": 2.3208660251050158e-05, | |
| "loss": 2.0094, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.008231354642313546, | |
| "grad_norm": 9.89315128326416, | |
| "learning_rate": 2.1895831107393484e-05, | |
| "loss": 1.8189, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.00828006088280061, | |
| "grad_norm": 9.914687156677246, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 1.7016, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.00832876712328767, | |
| "grad_norm": 9.6238431930542, | |
| "learning_rate": 1.9354647317351188e-05, | |
| "loss": 2.3516, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.008377473363774734, | |
| "grad_norm": 8.912046432495117, | |
| "learning_rate": 1.8128800512565513e-05, | |
| "loss": 2.2398, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.008426179604261796, | |
| "grad_norm": 9.197094917297363, | |
| "learning_rate": 1.6934406733817414e-05, | |
| "loss": 2.3229, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.008474885844748859, | |
| "grad_norm": 9.847735404968262, | |
| "learning_rate": 1.5772644703565565e-05, | |
| "loss": 2.0305, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.00852359208523592, | |
| "grad_norm": 10.200972557067871, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 1.656, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.008572298325722984, | |
| "grad_norm": 10.505123138427734, | |
| "learning_rate": 1.3551568628929434e-05, | |
| "loss": 2.147, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.008621004566210046, | |
| "grad_norm": 8.955830574035645, | |
| "learning_rate": 1.2494446518477022e-05, | |
| "loss": 1.4854, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.008669710806697107, | |
| "grad_norm": 12.1397066116333, | |
| "learning_rate": 1.1474337861210543e-05, | |
| "loss": 2.39, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.00871841704718417, | |
| "grad_norm": 12.608519554138184, | |
| "learning_rate": 1.049224938121548e-05, | |
| "loss": 2.3129, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.008767123287671232, | |
| "grad_norm": 9.86025619506836, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.6276, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.008815829528158296, | |
| "grad_norm": 12.893600463867188, | |
| "learning_rate": 8.645971286271904e-06, | |
| "loss": 2.4241, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.008864535768645357, | |
| "grad_norm": 11.86868667602539, | |
| "learning_rate": 7.783603724899257e-06, | |
| "loss": 2.2967, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.00891324200913242, | |
| "grad_norm": 11.338330268859863, | |
| "learning_rate": 6.962898649802823e-06, | |
| "loss": 1.6438, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.008961948249619482, | |
| "grad_norm": 11.070121765136719, | |
| "learning_rate": 6.184665997806832e-06, | |
| "loss": 1.3556, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.009010654490106546, | |
| "grad_norm": 10.048547744750977, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 1.6426, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.009059360730593607, | |
| "grad_norm": 14.009288787841797, | |
| "learning_rate": 4.758647376699032e-06, | |
| "loss": 1.7888, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.00910806697108067, | |
| "grad_norm": 10.863405227661133, | |
| "learning_rate": 4.112268715800943e-06, | |
| "loss": 1.5613, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.009156773211567732, | |
| "grad_norm": 11.333900451660156, | |
| "learning_rate": 3.511175705587433e-06, | |
| "loss": 1.4206, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.009205479452054794, | |
| "grad_norm": 10.498023986816406, | |
| "learning_rate": 2.9559615522887273e-06, | |
| "loss": 1.8235, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.009254185692541857, | |
| "grad_norm": 9.978139877319336, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.4594, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.009302891933028919, | |
| "grad_norm": 10.376583099365234, | |
| "learning_rate": 1.985315716152847e-06, | |
| "loss": 1.3848, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.009351598173515982, | |
| "grad_norm": 9.732892036437988, | |
| "learning_rate": 1.5708419435684462e-06, | |
| "loss": 1.0441, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.009400304414003044, | |
| "grad_norm": 10.896247863769531, | |
| "learning_rate": 1.2041619030626284e-06, | |
| "loss": 1.2706, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.009449010654490107, | |
| "grad_norm": 9.157602310180664, | |
| "learning_rate": 8.856374635655695e-07, | |
| "loss": 0.7031, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.009497716894977169, | |
| "grad_norm": 7.2205047607421875, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 0.4888, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.009546423135464232, | |
| "grad_norm": 6.679050922393799, | |
| "learning_rate": 3.9426493427611177e-07, | |
| "loss": 0.4518, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.009595129375951294, | |
| "grad_norm": 8.90599250793457, | |
| "learning_rate": 2.219017698460002e-07, | |
| "loss": 0.8044, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.009643835616438357, | |
| "grad_norm": 11.641607284545898, | |
| "learning_rate": 9.866357858642205e-08, | |
| "loss": 1.0654, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.009692541856925418, | |
| "grad_norm": 10.517057418823242, | |
| "learning_rate": 2.467198171342e-08, | |
| "loss": 0.7411, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.009741248097412482, | |
| "grad_norm": 12.261541366577148, | |
| "learning_rate": 0.0, | |
| "loss": 1.2423, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 239, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5236987211022336.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |