pbe-lora
/
gcp_models
/lora-logo_fix_full_deepseek33b_gpt35i_lr_0.0002_alpha_1024_r_1024
/checkpoint-902
/trainer_state.json
| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9789590254706533, | |
| "eval_steps": 113, | |
| "global_step": 902, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.428594393312656, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.1469, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "eval_loss": 2.179504632949829, | |
| "eval_runtime": 172.7557, | |
| "eval_samples_per_second": 13.447, | |
| "eval_steps_per_second": 0.423, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 3.7159828982580754, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.1946, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.643082734610657, | |
| "learning_rate": 1.2e-05, | |
| "loss": 2.232, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.75889245136616, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.1482, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.121791718587376, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9094, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.357880867767518, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.5871, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.946129682154882, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.326, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.108395610487625, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.0602, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.130447473758986, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.962, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.250625735269271, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7239, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.086877732413652, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.6204, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.1321791358110234, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.5483, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8099243715089002, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 0.5086, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8079844588753853, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.5112, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9169759048137539, | |
| "learning_rate": 6e-05, | |
| "loss": 0.4889, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.6917931287891316, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.4691, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.6294590670553853, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 0.4549, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.6001819771732335, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.4412, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.6389927422614288, | |
| "learning_rate": 7.6e-05, | |
| "loss": 0.4077, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.5185334799357709, | |
| "learning_rate": 8e-05, | |
| "loss": 0.4219, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.5274718517549238, | |
| "learning_rate": 8.4e-05, | |
| "loss": 0.4124, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.5333597700699511, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.4333, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.4817408646893783, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 0.4022, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.5041871581629092, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.3971, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.5892498993207921, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3769, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.4448471314677223, | |
| "learning_rate": 0.00010400000000000001, | |
| "loss": 0.3905, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.41460492934076704, | |
| "learning_rate": 0.00010800000000000001, | |
| "loss": 0.3814, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.39759977947743247, | |
| "learning_rate": 0.00011200000000000001, | |
| "loss": 0.3675, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.42656614420804834, | |
| "learning_rate": 0.000116, | |
| "loss": 0.3949, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.3248894659390531, | |
| "learning_rate": 0.00012, | |
| "loss": 0.3734, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.35364315198613105, | |
| "learning_rate": 0.000124, | |
| "loss": 0.3452, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.3377660731261332, | |
| "learning_rate": 0.00012800000000000002, | |
| "loss": 0.365, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.3029616210555195, | |
| "learning_rate": 0.000132, | |
| "loss": 0.3502, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.43234275224940705, | |
| "learning_rate": 0.00013600000000000003, | |
| "loss": 0.3728, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.3831666942704003, | |
| "learning_rate": 0.00014, | |
| "loss": 0.3922, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.3421076533380305, | |
| "learning_rate": 0.000144, | |
| "loss": 0.3544, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.30146273904800347, | |
| "learning_rate": 0.000148, | |
| "loss": 0.3237, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.3234891150012619, | |
| "learning_rate": 0.000152, | |
| "loss": 0.3771, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.32536194802044366, | |
| "learning_rate": 0.00015600000000000002, | |
| "loss": 0.3719, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.32489341033312524, | |
| "learning_rate": 0.00016, | |
| "loss": 0.3716, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.3135978131489568, | |
| "learning_rate": 0.000164, | |
| "loss": 0.3818, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.28403149178513853, | |
| "learning_rate": 0.000168, | |
| "loss": 0.3576, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.30747408490186307, | |
| "learning_rate": 0.000172, | |
| "loss": 0.3801, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.3491185114838396, | |
| "learning_rate": 0.00017600000000000002, | |
| "loss": 0.3612, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.4062709802932099, | |
| "learning_rate": 0.00018, | |
| "loss": 0.3752, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.4174448098463489, | |
| "learning_rate": 0.00018400000000000003, | |
| "loss": 0.3705, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.30034892483536746, | |
| "learning_rate": 0.000188, | |
| "loss": 0.3504, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.24699047743823555, | |
| "learning_rate": 0.000192, | |
| "loss": 0.3622, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.2944161708858025, | |
| "learning_rate": 0.000196, | |
| "loss": 0.3462, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.24968532785704994, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3074, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.4989866478305154, | |
| "learning_rate": 0.00019999932018605637, | |
| "loss": 0.3426, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.2729667152979656, | |
| "learning_rate": 0.00019999728075346845, | |
| "loss": 0.3452, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.26553719135078474, | |
| "learning_rate": 0.00019999388172996495, | |
| "loss": 0.3771, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.2563892875653707, | |
| "learning_rate": 0.0001999891231617599, | |
| "loss": 0.3709, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.40493567306146466, | |
| "learning_rate": 0.0001999830051135521, | |
| "loss": 0.3659, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.282074263507646, | |
| "learning_rate": 0.00019997552766852432, | |
| "loss": 0.3579, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.24830919197340875, | |
| "learning_rate": 0.00019996669092834193, | |
| "loss": 0.3164, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.31153665113975915, | |
| "learning_rate": 0.0001999564950131517, | |
| "loss": 0.3554, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.2883397422934534, | |
| "learning_rate": 0.00019994494006158017, | |
| "loss": 0.3318, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.29607763624002476, | |
| "learning_rate": 0.00019993202623073172, | |
| "loss": 0.3515, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.26631042457457627, | |
| "learning_rate": 0.0001999177536961863, | |
| "loss": 0.3404, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.30709845993919116, | |
| "learning_rate": 0.00019990212265199738, | |
| "loss": 0.3454, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.2689959836388791, | |
| "learning_rate": 0.0001998851333106889, | |
| "loss": 0.3448, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.28079893446027976, | |
| "learning_rate": 0.00019986678590325273, | |
| "loss": 0.3537, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.2777161046281277, | |
| "learning_rate": 0.00019984708067914532, | |
| "loss": 0.3512, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.2612936496089595, | |
| "learning_rate": 0.0001998260179062844, | |
| "loss": 0.3355, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.270747618988818, | |
| "learning_rate": 0.0001998035978710453, | |
| "loss": 0.3557, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.2524425572972277, | |
| "learning_rate": 0.00019977982087825713, | |
| "loss": 0.324, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.2655824178335376, | |
| "learning_rate": 0.00019975468725119843, | |
| "loss": 0.3274, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.29931436509996834, | |
| "learning_rate": 0.000199728197331593, | |
| "loss": 0.3733, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.2945779980864737, | |
| "learning_rate": 0.00019970035147960524, | |
| "loss": 0.3161, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.42061603746317366, | |
| "learning_rate": 0.00019967115007383507, | |
| "loss": 0.3486, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.24899746628510128, | |
| "learning_rate": 0.000199640593511313, | |
| "loss": 0.3428, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.2748778960772334, | |
| "learning_rate": 0.00019960868220749448, | |
| "loss": 0.3215, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.2826057447583215, | |
| "learning_rate": 0.00019957541659625458, | |
| "loss": 0.3663, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.26758814883425874, | |
| "learning_rate": 0.00019954079712988183, | |
| "loss": 0.3473, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.26191530390935236, | |
| "learning_rate": 0.00019950482427907211, | |
| "loss": 0.3464, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.27009995213133564, | |
| "learning_rate": 0.00019946749853292232, | |
| "loss": 0.3427, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.2828252836889427, | |
| "learning_rate": 0.00019942882039892377, | |
| "loss": 0.3369, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.28359570095347486, | |
| "learning_rate": 0.00019938879040295508, | |
| "loss": 0.3474, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.23726568176198512, | |
| "learning_rate": 0.0001993474090892753, | |
| "loss": 0.3348, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.28359002617947426, | |
| "learning_rate": 0.00019930467702051628, | |
| "loss": 0.3434, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.24507300294017476, | |
| "learning_rate": 0.0001992605947776752, | |
| "loss": 0.3211, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.2459171433488338, | |
| "learning_rate": 0.00019921516296010644, | |
| "loss": 0.3539, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.23982609581354497, | |
| "learning_rate": 0.0001991683821855137, | |
| "loss": 0.3367, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.24948097532302946, | |
| "learning_rate": 0.00019912025308994148, | |
| "loss": 0.3313, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.25673241215001325, | |
| "learning_rate": 0.00019907077632776632, | |
| "loss": 0.3384, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.2248520328170477, | |
| "learning_rate": 0.00019901995257168807, | |
| "loss": 0.3075, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.2785744560779341, | |
| "learning_rate": 0.00019896778251272078, | |
| "loss": 0.3505, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.2612264615045669, | |
| "learning_rate": 0.00019891426686018305, | |
| "loss": 0.3319, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.2588107693763323, | |
| "learning_rate": 0.00019885940634168864, | |
| "loss": 0.3036, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.2539991795599821, | |
| "learning_rate": 0.0001988032017031364, | |
| "loss": 0.3275, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.2228639788098274, | |
| "learning_rate": 0.00019874565370870038, | |
| "loss": 0.3128, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.2295330254763288, | |
| "learning_rate": 0.00019868676314081904, | |
| "loss": 0.3226, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.26774082666954935, | |
| "learning_rate": 0.00019862653080018506, | |
| "loss": 0.3438, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.2096704558396889, | |
| "learning_rate": 0.0001985649575057341, | |
| "loss": 0.3158, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.222343383957648, | |
| "learning_rate": 0.00019850204409463385, | |
| "loss": 0.3127, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.24355910663526567, | |
| "learning_rate": 0.00019843779142227256, | |
| "loss": 0.3278, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.2589007730313338, | |
| "learning_rate": 0.00019837220036224756, | |
| "loss": 0.3433, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.2048208439355147, | |
| "learning_rate": 0.00019830527180635308, | |
| "loss": 0.3038, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.22100911829011585, | |
| "learning_rate": 0.00019823700666456853, | |
| "loss": 0.3295, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.2865876628348663, | |
| "learning_rate": 0.0001981674058650458, | |
| "loss": 0.3374, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.23267706317431597, | |
| "learning_rate": 0.00019809647035409672, | |
| "loss": 0.3146, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.22896912398059824, | |
| "learning_rate": 0.0001980242010961803, | |
| "loss": 0.3302, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.21199810980682918, | |
| "learning_rate": 0.00019795059907388952, | |
| "loss": 0.3089, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.2314314536279127, | |
| "learning_rate": 0.00019787566528793807, | |
| "loss": 0.3219, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.22578014749699987, | |
| "learning_rate": 0.00019779940075714648, | |
| "loss": 0.3089, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.26739038717214403, | |
| "learning_rate": 0.0001977218065184287, | |
| "loss": 0.3314, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.27640406493967007, | |
| "learning_rate": 0.00019764288362677753, | |
| "loss": 0.315, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.275170691575301, | |
| "learning_rate": 0.0001975626331552507, | |
| "loss": 0.3192, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.28590415176453904, | |
| "learning_rate": 0.00019748105619495594, | |
| "loss": 0.3235, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.26618886451702856, | |
| "learning_rate": 0.0001973981538550364, | |
| "loss": 0.3235, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.22783361042781863, | |
| "learning_rate": 0.00019731392726265537, | |
| "loss": 0.3319, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.3324408233165741, | |
| "eval_runtime": 173.7614, | |
| "eval_samples_per_second": 13.369, | |
| "eval_steps_per_second": 0.42, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.25003343596979083, | |
| "learning_rate": 0.00019722837756298113, | |
| "loss": 0.3269, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.2051643008869916, | |
| "learning_rate": 0.0001971415059191712, | |
| "loss": 0.3331, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.26679577582874325, | |
| "learning_rate": 0.00019705331351235674, | |
| "loss": 0.319, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.22959410656566043, | |
| "learning_rate": 0.0001969638015416263, | |
| "loss": 0.326, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.24732040292675736, | |
| "learning_rate": 0.0001968729712240095, | |
| "loss": 0.3456, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.21472772586501115, | |
| "learning_rate": 0.00019678082379446078, | |
| "loss": 0.3154, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.22004349633491344, | |
| "learning_rate": 0.00019668736050584224, | |
| "loss": 0.3296, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.20481829759367912, | |
| "learning_rate": 0.00019659258262890683, | |
| "loss": 0.3168, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.22596082871640896, | |
| "learning_rate": 0.00019649649145228102, | |
| "loss": 0.3198, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.22416198789548378, | |
| "learning_rate": 0.00019639908828244718, | |
| "loss": 0.3122, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.25019709447245486, | |
| "learning_rate": 0.000196300374443726, | |
| "loss": 0.3256, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.23470807250735432, | |
| "learning_rate": 0.0001962003512782584, | |
| "loss": 0.3245, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.24693935678122386, | |
| "learning_rate": 0.00019609902014598718, | |
| "loss": 0.3292, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.2260534058582815, | |
| "learning_rate": 0.00019599638242463868, | |
| "loss": 0.325, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.23667285933893983, | |
| "learning_rate": 0.00019589243950970402, | |
| "loss": 0.3337, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.21298574370036155, | |
| "learning_rate": 0.00019578719281442003, | |
| "loss": 0.3092, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.23372716379048458, | |
| "learning_rate": 0.00019568064376975012, | |
| "loss": 0.317, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.22334413345013637, | |
| "learning_rate": 0.0001955727938243648, | |
| "loss": 0.2834, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.2248381817587477, | |
| "learning_rate": 0.00019546364444462207, | |
| "loss": 0.302, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.2360215129073668, | |
| "learning_rate": 0.00019535319711454728, | |
| "loss": 0.3293, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.22623309491159815, | |
| "learning_rate": 0.00019524145333581317, | |
| "loss": 0.3273, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.2070555672669519, | |
| "learning_rate": 0.00019512841462771924, | |
| "loss": 0.2972, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.22460201914808312, | |
| "learning_rate": 0.00019501408252717138, | |
| "loss": 0.3019, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.2225333175600629, | |
| "learning_rate": 0.00019489845858866066, | |
| "loss": 0.2983, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.21017634852282913, | |
| "learning_rate": 0.0001947815443842424, | |
| "loss": 0.3002, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.24382065213677345, | |
| "learning_rate": 0.00019466334150351476, | |
| "loss": 0.3061, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.22222229848717998, | |
| "learning_rate": 0.00019454385155359702, | |
| "loss": 0.3189, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.21144991906498956, | |
| "learning_rate": 0.00019442307615910793, | |
| "loss": 0.3093, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.20966684015092005, | |
| "learning_rate": 0.00019430101696214336, | |
| "loss": 0.2975, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.2285163015371888, | |
| "learning_rate": 0.0001941776756222542, | |
| "loss": 0.3291, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.22173438888731478, | |
| "learning_rate": 0.00019405305381642375, | |
| "loss": 0.3052, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.23472615797287633, | |
| "learning_rate": 0.00019392715323904481, | |
| "loss": 0.3158, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.19206597044565454, | |
| "learning_rate": 0.00019379997560189675, | |
| "loss": 0.3047, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.21928888599234125, | |
| "learning_rate": 0.00019367152263412217, | |
| "loss": 0.3196, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.21469742386937682, | |
| "learning_rate": 0.00019354179608220348, | |
| "loss": 0.2981, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.39037984784670976, | |
| "learning_rate": 0.000193410797709939, | |
| "loss": 0.2962, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.3224683875107671, | |
| "learning_rate": 0.00019327852929841916, | |
| "loss": 0.3268, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.22377973835731155, | |
| "learning_rate": 0.0001931449926460022, | |
| "loss": 0.3093, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.21525051451055552, | |
| "learning_rate": 0.00019301018956828964, | |
| "loss": 0.3173, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.19693490433987484, | |
| "learning_rate": 0.00019287412189810172, | |
| "loss": 0.3069, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.1985719221741534, | |
| "learning_rate": 0.00019273679148545245, | |
| "loss": 0.3239, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.19043619193681682, | |
| "learning_rate": 0.00019259820019752443, | |
| "loss": 0.3041, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.2029259146228578, | |
| "learning_rate": 0.0001924583499186434, | |
| "loss": 0.3049, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.1913024173068621, | |
| "learning_rate": 0.00019231724255025284, | |
| "loss": 0.3263, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.20418535912509125, | |
| "learning_rate": 0.00019217488001088784, | |
| "loss": 0.3093, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.21506493193668452, | |
| "learning_rate": 0.00019203126423614916, | |
| "loss": 0.3386, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.2078781708947467, | |
| "learning_rate": 0.00019188639717867696, | |
| "loss": 0.3098, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.21203565844547, | |
| "learning_rate": 0.00019174028080812415, | |
| "loss": 0.3245, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.22944231505967116, | |
| "learning_rate": 0.0001915929171111296, | |
| "loss": 0.3174, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.19389795184150505, | |
| "learning_rate": 0.00019144430809129128, | |
| "loss": 0.2985, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.2294568217196969, | |
| "learning_rate": 0.00019129445576913888, | |
| "loss": 0.2916, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.21471312626755512, | |
| "learning_rate": 0.00019114336218210634, | |
| "loss": 0.3203, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.200256096520887, | |
| "learning_rate": 0.00019099102938450416, | |
| "loss": 0.314, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.20286230376549838, | |
| "learning_rate": 0.00019083745944749162, | |
| "loss": 0.2953, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.22752630009694177, | |
| "learning_rate": 0.00019068265445904836, | |
| "loss": 0.3098, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.1969764402722988, | |
| "learning_rate": 0.00019052661652394618, | |
| "loss": 0.2798, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.21419927863312113, | |
| "learning_rate": 0.0001903693477637204, | |
| "loss": 0.315, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.1978129640154948, | |
| "learning_rate": 0.00019021085031664087, | |
| "loss": 0.2769, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.1832605646042805, | |
| "learning_rate": 0.00019005112633768313, | |
| "loss": 0.2787, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.25130865190581947, | |
| "learning_rate": 0.00018989017799849896, | |
| "loss": 0.3042, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.22333188194958617, | |
| "learning_rate": 0.0001897280074873868, | |
| "loss": 0.3366, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.23303955610127655, | |
| "learning_rate": 0.00018956461700926215, | |
| "loss": 0.3069, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.21946538873928723, | |
| "learning_rate": 0.00018940000878562758, | |
| "loss": 0.3026, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.22034016868322132, | |
| "learning_rate": 0.00018923418505454237, | |
| "loss": 0.3031, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.1918273691770089, | |
| "learning_rate": 0.00018906714807059218, | |
| "loss": 0.287, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.1984010069624954, | |
| "learning_rate": 0.00018889890010485847, | |
| "loss": 0.3039, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.19139523377149348, | |
| "learning_rate": 0.00018872944344488747, | |
| "loss": 0.3152, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.18841860181390324, | |
| "learning_rate": 0.0001885587803946592, | |
| "loss": 0.3171, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.19736334769102976, | |
| "learning_rate": 0.0001883869132745561, | |
| "loss": 0.2851, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.19610383893082234, | |
| "learning_rate": 0.00018821384442133145, | |
| "loss": 0.307, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.19693236953217128, | |
| "learning_rate": 0.00018803957618807764, | |
| "loss": 0.3219, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.1978461815572098, | |
| "learning_rate": 0.0001878641109441942, | |
| "loss": 0.2936, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.22453262012783057, | |
| "learning_rate": 0.00018768745107535542, | |
| "loss": 0.3225, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.1951859636493649, | |
| "learning_rate": 0.00018750959898347825, | |
| "loss": 0.2892, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.19051492201816908, | |
| "learning_rate": 0.00018733055708668926, | |
| "loss": 0.2922, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.19631810708829814, | |
| "learning_rate": 0.00018715032781929208, | |
| "loss": 0.2928, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.20533375929006709, | |
| "learning_rate": 0.00018696891363173405, | |
| "loss": 0.3212, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.21698315667726276, | |
| "learning_rate": 0.00018678631699057302, | |
| "loss": 0.3419, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.23742145444149265, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.3081, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.22429285552735956, | |
| "learning_rate": 0.00018641758629402467, | |
| "loss": 0.3132, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.20957107721564286, | |
| "learning_rate": 0.00018623145725200278, | |
| "loss": 0.3176, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.23119517784848204, | |
| "learning_rate": 0.0001860441557830405, | |
| "loss": 0.3174, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.22081382473360098, | |
| "learning_rate": 0.00018585568443374087, | |
| "loss": 0.3029, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.20857278454177905, | |
| "learning_rate": 0.00018566604576661288, | |
| "loss": 0.2803, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.20277792365487685, | |
| "learning_rate": 0.00018547524236003674, | |
| "loss": 0.3032, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.22068595819466755, | |
| "learning_rate": 0.0001852832768082288, | |
| "loss": 0.3196, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.22826508701764245, | |
| "learning_rate": 0.00018509015172120621, | |
| "loss": 0.307, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.23318100222090005, | |
| "learning_rate": 0.00018489586972475155, | |
| "loss": 0.3243, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.20790305208366794, | |
| "learning_rate": 0.00018470043346037698, | |
| "loss": 0.3026, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.23045901908374763, | |
| "learning_rate": 0.00018450384558528845, | |
| "loss": 0.3215, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.1895969081217627, | |
| "learning_rate": 0.0001843061087723496, | |
| "loss": 0.2827, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.21471234551245583, | |
| "learning_rate": 0.00018410722571004522, | |
| "loss": 0.2758, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.19174460263968052, | |
| "learning_rate": 0.00018390719910244487, | |
| "loss": 0.2935, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.2042082548764658, | |
| "learning_rate": 0.00018370603166916616, | |
| "loss": 0.3219, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.20862412820843024, | |
| "learning_rate": 0.00018350372614533753, | |
| "loss": 0.3079, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.19365588486473745, | |
| "learning_rate": 0.00018330028528156138, | |
| "loss": 0.2878, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.20923759592087007, | |
| "learning_rate": 0.0001830957118438764, | |
| "loss": 0.3052, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.19346679246935292, | |
| "learning_rate": 0.00018289000861372007, | |
| "loss": 0.2873, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.20707125547144337, | |
| "learning_rate": 0.00018268317838789088, | |
| "loss": 0.2888, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.210272509030541, | |
| "learning_rate": 0.00018247522397851028, | |
| "loss": 0.3063, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.22635533463893145, | |
| "learning_rate": 0.0001822661482129844, | |
| "loss": 0.2844, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.20691107915045812, | |
| "learning_rate": 0.00018205595393396568, | |
| "loss": 0.2986, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.21549907787627992, | |
| "learning_rate": 0.00018184464399931412, | |
| "loss": 0.3098, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.20061134281496176, | |
| "learning_rate": 0.00018163222128205853, | |
| "loss": 0.2871, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.20502352214692726, | |
| "learning_rate": 0.00018141868867035745, | |
| "loss": 0.294, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.2291358128642933, | |
| "learning_rate": 0.00018120404906745973, | |
| "loss": 0.2757, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.1972835362678214, | |
| "learning_rate": 0.00018098830539166536, | |
| "loss": 0.3084, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.21611715417590088, | |
| "learning_rate": 0.00018077146057628545, | |
| "loss": 0.2816, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.21599858891116713, | |
| "learning_rate": 0.00018055351756960262, | |
| "loss": 0.3085, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.19318954639097616, | |
| "learning_rate": 0.00018033447933483076, | |
| "loss": 0.2557, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.18823294633329682, | |
| "learning_rate": 0.00018011434885007482, | |
| "loss": 0.2902, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.21452443929587184, | |
| "learning_rate": 0.00017989312910829023, | |
| "loss": 0.311, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.20666447125090429, | |
| "learning_rate": 0.00017967082311724227, | |
| "loss": 0.2883, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.297645628452301, | |
| "eval_runtime": 173.6984, | |
| "eval_samples_per_second": 13.374, | |
| "eval_steps_per_second": 0.42, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.21336924271339083, | |
| "learning_rate": 0.00017944743389946524, | |
| "loss": 0.3026, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.20792583408257218, | |
| "learning_rate": 0.0001792229644922212, | |
| "loss": 0.2843, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.2129371399765845, | |
| "learning_rate": 0.0001789974179474588, | |
| "loss": 0.3091, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.19527091537284477, | |
| "learning_rate": 0.00017877079733177184, | |
| "loss": 0.297, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.20516345655708346, | |
| "learning_rate": 0.00017854310572635733, | |
| "loss": 0.2935, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.1969309791573733, | |
| "learning_rate": 0.00017831434622697385, | |
| "loss": 0.2898, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.2696175229862278, | |
| "learning_rate": 0.0001780845219438994, | |
| "loss": 0.2924, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.234973041162424, | |
| "learning_rate": 0.00017785363600188894, | |
| "loss": 0.3179, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.23939104068785497, | |
| "learning_rate": 0.00017762169154013216, | |
| "loss": 0.2796, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.20378572036784126, | |
| "learning_rate": 0.00017738869171221068, | |
| "loss": 0.2784, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.19159802735904477, | |
| "learning_rate": 0.0001771546396860551, | |
| "loss": 0.2834, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.23302725481391803, | |
| "learning_rate": 0.00017691953864390207, | |
| "loss": 0.2997, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.23471887316578038, | |
| "learning_rate": 0.0001766833917822509, | |
| "loss": 0.3, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.1796225243725005, | |
| "learning_rate": 0.00017644620231182015, | |
| "loss": 0.2901, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.20037902218428075, | |
| "learning_rate": 0.00017620797345750403, | |
| "loss": 0.294, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.20415485179217374, | |
| "learning_rate": 0.0001759687084583285, | |
| "loss": 0.3162, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.18182779738483645, | |
| "learning_rate": 0.00017572841056740722, | |
| "loss": 0.275, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.17275776985467062, | |
| "learning_rate": 0.00017548708305189722, | |
| "loss": 0.2592, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.18388792575433974, | |
| "learning_rate": 0.00017524472919295487, | |
| "loss": 0.2998, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.1821375376929017, | |
| "learning_rate": 0.00017500135228569068, | |
| "loss": 0.2586, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.18937126829543396, | |
| "learning_rate": 0.00017475695563912505, | |
| "loss": 0.2858, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.1833938565560613, | |
| "learning_rate": 0.00017451154257614287, | |
| "loss": 0.271, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.20382927797400074, | |
| "learning_rate": 0.0001742651164334486, | |
| "loss": 0.2931, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.23510651870506527, | |
| "learning_rate": 0.00017401768056152085, | |
| "loss": 0.3078, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.2191619740250893, | |
| "learning_rate": 0.00017376923832456665, | |
| "loss": 0.3111, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.20648235757385622, | |
| "learning_rate": 0.00017351979310047602, | |
| "loss": 0.2816, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.19445561313531604, | |
| "learning_rate": 0.00017326934828077573, | |
| "loss": 0.2894, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.19169798824869538, | |
| "learning_rate": 0.00017301790727058345, | |
| "loss": 0.2802, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.2006703950203184, | |
| "learning_rate": 0.0001727654734885612, | |
| "loss": 0.2749, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.21061482297381626, | |
| "learning_rate": 0.0001725120503668691, | |
| "loss": 0.3042, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.20740853426573325, | |
| "learning_rate": 0.00017225764135111868, | |
| "loss": 0.3025, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.1945016623980029, | |
| "learning_rate": 0.00017200224990032576, | |
| "loss": 0.2964, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.21095668740779128, | |
| "learning_rate": 0.00017174587948686374, | |
| "loss": 0.3047, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.18724414137242903, | |
| "learning_rate": 0.00017148853359641626, | |
| "loss": 0.2678, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.18247589724576602, | |
| "learning_rate": 0.00017123021572792982, | |
| "loss": 0.2796, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.20187877784359254, | |
| "learning_rate": 0.00017097092939356623, | |
| "loss": 0.2819, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.21960455156021205, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.2849, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.21920804195394356, | |
| "learning_rate": 0.00017044946544164433, | |
| "loss": 0.286, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.2153094965416609, | |
| "learning_rate": 0.00017018729491405536, | |
| "loss": 0.2728, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.18899974451413623, | |
| "learning_rate": 0.00016992417010043142, | |
| "loss": 0.2643, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.20452371771386804, | |
| "learning_rate": 0.00016966009457829086, | |
| "loss": 0.2805, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.19546419928152936, | |
| "learning_rate": 0.0001693950719380782, | |
| "loss": 0.2749, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.19627215184487343, | |
| "learning_rate": 0.00016912910578311503, | |
| "loss": 0.273, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.18828419766156645, | |
| "learning_rate": 0.00016886219972955146, | |
| "loss": 0.273, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.19027059707604657, | |
| "learning_rate": 0.00016859435740631658, | |
| "loss": 0.3046, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.18426233664144975, | |
| "learning_rate": 0.00016832558245506935, | |
| "loss": 0.2643, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.18658729778193736, | |
| "learning_rate": 0.00016805587853014895, | |
| "loss": 0.285, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.1796032355554721, | |
| "learning_rate": 0.00016778524929852512, | |
| "loss": 0.261, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.19790847765561645, | |
| "learning_rate": 0.0001675136984397484, | |
| "loss": 0.3036, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.2013507251688789, | |
| "learning_rate": 0.0001672412296459, | |
| "loss": 0.2929, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.20727255991158675, | |
| "learning_rate": 0.00016696784662154163, | |
| "loss": 0.28, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.22175173491924988, | |
| "learning_rate": 0.0001666935530836651, | |
| "loss": 0.2953, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.2110368794764076, | |
| "learning_rate": 0.00016641835276164183, | |
| "loss": 0.3012, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.1839097327175649, | |
| "learning_rate": 0.00016614224939717217, | |
| "loss": 0.2985, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.18075701410157072, | |
| "learning_rate": 0.00016586524674423446, | |
| "loss": 0.2614, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.19006328055997038, | |
| "learning_rate": 0.00016558734856903404, | |
| "loss": 0.2741, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.1936501426398552, | |
| "learning_rate": 0.00016530855864995195, | |
| "loss": 0.2486, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.2146360520459362, | |
| "learning_rate": 0.0001650288807774937, | |
| "loss": 0.2949, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.1934454081094374, | |
| "learning_rate": 0.00016474831875423767, | |
| "loss": 0.25, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.20138529225313043, | |
| "learning_rate": 0.0001644668763947833, | |
| "loss": 0.2826, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.20336180418412392, | |
| "learning_rate": 0.00016418455752569943, | |
| "loss": 0.281, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.2366660122305481, | |
| "learning_rate": 0.00016390136598547217, | |
| "loss": 0.2665, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.18478597928089585, | |
| "learning_rate": 0.00016361730562445263, | |
| "loss": 0.3022, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.18679948920602824, | |
| "learning_rate": 0.0001633323803048047, | |
| "loss": 0.2844, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.19949957744833288, | |
| "learning_rate": 0.00016304659390045252, | |
| "loss": 0.2912, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.20044341182168152, | |
| "learning_rate": 0.0001627599502970277, | |
| "loss": 0.2729, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.18558206726419454, | |
| "learning_rate": 0.00016247245339181662, | |
| "loss": 0.2693, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.2083655759577377, | |
| "learning_rate": 0.00016218410709370736, | |
| "loss": 0.3022, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.21456219690969847, | |
| "learning_rate": 0.00016189491532313664, | |
| "loss": 0.2933, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.2316084982608295, | |
| "learning_rate": 0.00016160488201203644, | |
| "loss": 0.2631, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.20943766303868608, | |
| "learning_rate": 0.00016131401110378043, | |
| "loss": 0.2847, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.2346281512353159, | |
| "learning_rate": 0.00016102230655313076, | |
| "loss": 0.2898, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.18690479698056445, | |
| "learning_rate": 0.0001607297723261837, | |
| "loss": 0.2761, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.22244239877246086, | |
| "learning_rate": 0.00016043641240031623, | |
| "loss": 0.2794, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.19513520777123738, | |
| "learning_rate": 0.00016014223076413173, | |
| "loss": 0.2757, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.1954004471181246, | |
| "learning_rate": 0.00015984723141740576, | |
| "loss": 0.2713, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.21399036326669596, | |
| "learning_rate": 0.00015955141837103168, | |
| "loss": 0.2767, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.2045546981741005, | |
| "learning_rate": 0.0001592547956469662, | |
| "loss": 0.2807, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.21321779886240727, | |
| "learning_rate": 0.00015895736727817455, | |
| "loss": 0.2899, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.20121770170139855, | |
| "learning_rate": 0.00015865913730857582, | |
| "loss": 0.2706, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.21901656836549904, | |
| "learning_rate": 0.00015836010979298782, | |
| "loss": 0.3043, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.19476889300368513, | |
| "learning_rate": 0.0001580602887970721, | |
| "loss": 0.294, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.1864916489760236, | |
| "learning_rate": 0.00015775967839727842, | |
| "loss": 0.2789, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.2005263777787887, | |
| "learning_rate": 0.0001574582826807897, | |
| "loss": 0.2786, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.20330910409013883, | |
| "learning_rate": 0.0001571561057454661, | |
| "loss": 0.2856, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.2075041911121307, | |
| "learning_rate": 0.00015685315169978954, | |
| "loss": 0.301, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.198545224318664, | |
| "learning_rate": 0.0001565494246628077, | |
| "loss": 0.2667, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.18262873322406858, | |
| "learning_rate": 0.0001562449287640781, | |
| "loss": 0.2761, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.20682904427400386, | |
| "learning_rate": 0.0001559396681436118, | |
| "loss": 0.2798, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.17100330949740156, | |
| "learning_rate": 0.00015563364695181741, | |
| "loss": 0.2638, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.19147233406100644, | |
| "learning_rate": 0.00015532686934944438, | |
| "loss": 0.2772, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.20859821686266017, | |
| "learning_rate": 0.00015501933950752656, | |
| "loss": 0.2899, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.22660943071408993, | |
| "learning_rate": 0.00015471106160732542, | |
| "loss": 0.2809, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.1970595992988207, | |
| "learning_rate": 0.00015440203984027324, | |
| "loss": 0.2664, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.21441465355834557, | |
| "learning_rate": 0.00015409227840791617, | |
| "loss": 0.2872, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.1928160770028818, | |
| "learning_rate": 0.000153781781521857, | |
| "loss": 0.2932, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.18232874606722474, | |
| "learning_rate": 0.00015347055340369804, | |
| "loss": 0.2865, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.18944518082470313, | |
| "learning_rate": 0.00015315859828498354, | |
| "loss": 0.2895, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.18688411692002266, | |
| "learning_rate": 0.00015284592040714227, | |
| "loss": 0.3068, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.15360887670295487, | |
| "learning_rate": 0.00015253252402142988, | |
| "loss": 0.2541, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.17886321037354558, | |
| "learning_rate": 0.00015221841338887104, | |
| "loss": 0.2735, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.18210509607368616, | |
| "learning_rate": 0.0001519035927802015, | |
| "loss": 0.2674, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.2113800439853146, | |
| "learning_rate": 0.00015158806647581002, | |
| "loss": 0.2611, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.21664373470568593, | |
| "learning_rate": 0.00015127183876568022, | |
| "loss": 0.2734, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.21152226314764647, | |
| "learning_rate": 0.0001509549139493323, | |
| "loss": 0.2605, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.20787608314066813, | |
| "learning_rate": 0.0001506372963357644, | |
| "loss": 0.2829, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.20890612340451087, | |
| "learning_rate": 0.00015031899024339415, | |
| "loss": 0.2761, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.18123961862890825, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.2625, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.20005061805451332, | |
| "learning_rate": 0.00014968032994266224, | |
| "loss": 0.2739, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.22011995491515385, | |
| "learning_rate": 0.00014935998441770407, | |
| "loss": 0.2769, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.1759458248293069, | |
| "learning_rate": 0.00014903896778063267, | |
| "loss": 0.2751, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.17963701036394614, | |
| "learning_rate": 0.00014871728439607966, | |
| "loss": 0.2861, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.22395035633400334, | |
| "learning_rate": 0.00014839493863774212, | |
| "loss": 0.2748, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.27847620844841003, | |
| "eval_runtime": 174.097, | |
| "eval_samples_per_second": 13.343, | |
| "eval_steps_per_second": 0.419, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.19931473943900235, | |
| "learning_rate": 0.00014807193488832282, | |
| "loss": 0.261, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.1895068181039295, | |
| "learning_rate": 0.00014774827753947088, | |
| "loss": 0.2666, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.19682173311096884, | |
| "learning_rate": 0.00014742397099172183, | |
| "loss": 0.2564, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.2183356533233642, | |
| "learning_rate": 0.00014709901965443794, | |
| "loss": 0.2904, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.2113562507034691, | |
| "learning_rate": 0.00014677342794574817, | |
| "loss": 0.2915, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.2034784414842117, | |
| "learning_rate": 0.00014644720029248829, | |
| "loss": 0.2717, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.1967572374697097, | |
| "learning_rate": 0.00014612034113014035, | |
| "loss": 0.2887, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.1899953535101574, | |
| "learning_rate": 0.00014579285490277274, | |
| "loss": 0.2922, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.19441384522232566, | |
| "learning_rate": 0.0001454647460629795, | |
| "loss": 0.2785, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.20710674744690838, | |
| "learning_rate": 0.00014513601907181992, | |
| "loss": 0.2929, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.18640646337351455, | |
| "learning_rate": 0.00014480667839875786, | |
| "loss": 0.261, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.1754057901024035, | |
| "learning_rate": 0.00014447672852160095, | |
| "loss": 0.267, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.1845116395890587, | |
| "learning_rate": 0.0001441461739264397, | |
| "loss": 0.2608, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.1832842348380918, | |
| "learning_rate": 0.00014381501910758662, | |
| "loss": 0.264, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.21078762275275081, | |
| "learning_rate": 0.00014348326856751496, | |
| "loss": 0.2903, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.2040423439668467, | |
| "learning_rate": 0.00014315092681679755, | |
| "loss": 0.2866, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.20916646529152538, | |
| "learning_rate": 0.00014281799837404552, | |
| "loss": 0.2669, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.20829199241699958, | |
| "learning_rate": 0.00014248448776584688, | |
| "loss": 0.2773, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.22863270132432637, | |
| "learning_rate": 0.0001421503995267048, | |
| "loss": 0.2691, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.19107899619014693, | |
| "learning_rate": 0.00014181573819897617, | |
| "loss": 0.2854, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.18310920202899897, | |
| "learning_rate": 0.00014148050833280977, | |
| "loss": 0.2523, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.19142180009852441, | |
| "learning_rate": 0.00014114471448608426, | |
| "loss": 0.2668, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.17916118207750328, | |
| "learning_rate": 0.0001408083612243465, | |
| "loss": 0.2826, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.1795532305403206, | |
| "learning_rate": 0.0001404714531207492, | |
| "loss": 0.2876, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.19812580507096825, | |
| "learning_rate": 0.0001401339947559889, | |
| "loss": 0.2813, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.16985158453017796, | |
| "learning_rate": 0.00013979599071824362, | |
| "loss": 0.2824, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.20447909651191687, | |
| "learning_rate": 0.00013945744560311057, | |
| "loss": 0.2567, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.20178806044954373, | |
| "learning_rate": 0.0001391183640135435, | |
| "loss": 0.2755, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.2265042443851448, | |
| "learning_rate": 0.00013877875055979023, | |
| "loss": 0.2905, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.22127748544952333, | |
| "learning_rate": 0.00013843860985933003, | |
| "loss": 0.2624, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.20805077265871227, | |
| "learning_rate": 0.00013809794653681074, | |
| "loss": 0.2396, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.21236429860308073, | |
| "learning_rate": 0.00013775676522398588, | |
| "loss": 0.2916, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.19893678645018886, | |
| "learning_rate": 0.00013741507055965168, | |
| "loss": 0.2551, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.18256634499151977, | |
| "learning_rate": 0.00013707286718958413, | |
| "loss": 0.2633, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.18221655160895703, | |
| "learning_rate": 0.00013673015976647568, | |
| "loss": 0.2672, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.19429764298431815, | |
| "learning_rate": 0.00013638695294987204, | |
| "loss": 0.2417, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.18793583918739298, | |
| "learning_rate": 0.0001360432514061087, | |
| "loss": 0.2613, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.2011435035031983, | |
| "learning_rate": 0.00013569905980824788, | |
| "loss": 0.2685, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.2052637622415086, | |
| "learning_rate": 0.00013535438283601435, | |
| "loss": 0.2959, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.1945089020971851, | |
| "learning_rate": 0.00013500922517573245, | |
| "loss": 0.2482, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.20812822716018067, | |
| "learning_rate": 0.00013466359152026195, | |
| "loss": 0.2741, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.22384533508071883, | |
| "learning_rate": 0.0001343174865689344, | |
| "loss": 0.2914, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.19177959970339475, | |
| "learning_rate": 0.0001339709150274893, | |
| "loss": 0.2694, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.18196106404878304, | |
| "learning_rate": 0.0001336238816080099, | |
| "loss": 0.2649, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.2526481416447456, | |
| "learning_rate": 0.00013327639102885937, | |
| "loss": 0.2813, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.26250954621985834, | |
| "learning_rate": 0.0001329284480146166, | |
| "loss": 0.2793, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.19318862080071506, | |
| "learning_rate": 0.00013258005729601177, | |
| "loss": 0.2441, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.21447364833995924, | |
| "learning_rate": 0.00013223122360986225, | |
| "loss": 0.2887, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.18021814111984105, | |
| "learning_rate": 0.00013188195169900813, | |
| "loss": 0.2722, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.1986609859985159, | |
| "learning_rate": 0.0001315322463122477, | |
| "loss": 0.2638, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.24674716367898283, | |
| "learning_rate": 0.00013118211220427298, | |
| "loss": 0.2753, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.22439353357611744, | |
| "learning_rate": 0.0001308315541356049, | |
| "loss": 0.2853, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.17776179888083857, | |
| "learning_rate": 0.00013048057687252865, | |
| "loss": 0.2441, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.21002490821212472, | |
| "learning_rate": 0.00013012918518702914, | |
| "loss": 0.2882, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.18514107951857695, | |
| "learning_rate": 0.00012977738385672557, | |
| "loss": 0.2643, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.18577497415553973, | |
| "learning_rate": 0.000129425177664807, | |
| "loss": 0.2588, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.21465706083350347, | |
| "learning_rate": 0.00012907257139996704, | |
| "loss": 0.287, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.1742663966229961, | |
| "learning_rate": 0.0001287195698563388, | |
| "loss": 0.2668, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.20553702826350395, | |
| "learning_rate": 0.0001283661778334297, | |
| "loss": 0.2739, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.16765771945192084, | |
| "learning_rate": 0.0001280124001360562, | |
| "loss": 0.2594, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.1943563901020515, | |
| "learning_rate": 0.0001276582415742786, | |
| "loss": 0.2708, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.1896178424172107, | |
| "learning_rate": 0.0001273037069633354, | |
| "loss": 0.2639, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.0712870595212587, | |
| "learning_rate": 0.00012694880112357808, | |
| "loss": 0.2765, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.2092081458414992, | |
| "learning_rate": 0.00012659352888040547, | |
| "loss": 0.2589, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.21211010351459872, | |
| "learning_rate": 0.0001262378950641979, | |
| "loss": 0.285, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.23377789093835202, | |
| "learning_rate": 0.00012588190451025207, | |
| "loss": 0.3038, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.24727367430206143, | |
| "learning_rate": 0.00012552556205871478, | |
| "loss": 0.2577, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.1800512798597796, | |
| "learning_rate": 0.00012516887255451735, | |
| "loss": 0.2392, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.2114523591213948, | |
| "learning_rate": 0.00012481184084730976, | |
| "loss": 0.27, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.22674225384993477, | |
| "learning_rate": 0.0001244544717913947, | |
| "loss": 0.2372, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.17864125878915538, | |
| "learning_rate": 0.00012409677024566144, | |
| "loss": 0.2601, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.1904684467485375, | |
| "learning_rate": 0.00012373874107352004, | |
| "loss": 0.2647, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.17298512767214175, | |
| "learning_rate": 0.0001233803891428349, | |
| "loss": 0.2438, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.17544690229663243, | |
| "learning_rate": 0.00012302171932585885, | |
| "loss": 0.2585, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.20074139062085292, | |
| "learning_rate": 0.0001226627364991667, | |
| "loss": 0.2575, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.18591883669880138, | |
| "learning_rate": 0.0001223034455435891, | |
| "loss": 0.2593, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.20298533039339473, | |
| "learning_rate": 0.00012194385134414608, | |
| "loss": 0.2779, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.19947012767946487, | |
| "learning_rate": 0.00012158395878998063, | |
| "loss": 0.2776, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.18367461189103365, | |
| "learning_rate": 0.00012122377277429231, | |
| "loss": 0.2934, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.18639410238342366, | |
| "learning_rate": 0.00012086329819427065, | |
| "loss": 0.2848, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.18355116451736245, | |
| "learning_rate": 0.00012050253995102854, | |
| "loss": 0.2864, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.16618761241876043, | |
| "learning_rate": 0.00012014150294953563, | |
| "loss": 0.2722, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.1679981782550472, | |
| "learning_rate": 0.00011978019209855174, | |
| "loss": 0.2417, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.17373986628880816, | |
| "learning_rate": 0.00011941861231055994, | |
| "loss": 0.2464, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.16872090115264807, | |
| "learning_rate": 0.0001190567685016998, | |
| "loss": 0.2541, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.17551340253440462, | |
| "learning_rate": 0.00011869466559170073, | |
| "loss": 0.2521, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.18233468138593362, | |
| "learning_rate": 0.00011833230850381487, | |
| "loss": 0.2712, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.18504780357931186, | |
| "learning_rate": 0.00011796970216475018, | |
| "loss": 0.2754, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.18478336853897961, | |
| "learning_rate": 0.00011760685150460362, | |
| "loss": 0.2592, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.19917097640828008, | |
| "learning_rate": 0.00011724376145679394, | |
| "loss": 0.2855, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.19144241808727208, | |
| "learning_rate": 0.00011688043695799468, | |
| "loss": 0.2502, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.18405239160044018, | |
| "learning_rate": 0.00011651688294806706, | |
| "loss": 0.2477, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.1796721265270086, | |
| "learning_rate": 0.00011615310436999279, | |
| "loss": 0.249, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.198923200273133, | |
| "learning_rate": 0.00011578910616980683, | |
| "loss": 0.2559, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.2056661440897021, | |
| "learning_rate": 0.00011542489329653024, | |
| "loss": 0.2645, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.21059350822537876, | |
| "learning_rate": 0.00011506047070210282, | |
| "loss": 0.2747, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.19445339084850757, | |
| "learning_rate": 0.00011469584334131578, | |
| "loss": 0.2732, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.1978300160069901, | |
| "learning_rate": 0.0001143310161717444, | |
| "loss": 0.2653, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.1992099703929807, | |
| "learning_rate": 0.00011396599415368061, | |
| "loss": 0.2775, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.20064532246856512, | |
| "learning_rate": 0.00011360078225006562, | |
| "loss": 0.2642, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.19513907362313387, | |
| "learning_rate": 0.00011323538542642227, | |
| "loss": 0.2554, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.191943659814926, | |
| "learning_rate": 0.00011286980865078763, | |
| "loss": 0.2778, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.1921885618320068, | |
| "learning_rate": 0.0001125040568936456, | |
| "loss": 0.2734, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.19140570504247198, | |
| "learning_rate": 0.00011213813512785898, | |
| "loss": 0.2809, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.1772706721475436, | |
| "learning_rate": 0.00011177204832860213, | |
| "loss": 0.2558, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.17875694766580855, | |
| "learning_rate": 0.00011140580147329338, | |
| "loss": 0.2569, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.1891065652583873, | |
| "learning_rate": 0.000111039399541527, | |
| "loss": 0.271, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.17377167224369017, | |
| "learning_rate": 0.00011067284751500583, | |
| "loss": 0.2711, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.16281001225472738, | |
| "learning_rate": 0.00011030615037747353, | |
| "loss": 0.2464, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.17949964198592142, | |
| "learning_rate": 0.0001099393131146466, | |
| "loss": 0.2514, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.18463419257880492, | |
| "learning_rate": 0.00010957234071414674, | |
| "loss": 0.2799, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.18137088353673406, | |
| "learning_rate": 0.00010920523816543309, | |
| "loss": 0.2615, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.18259357360917502, | |
| "learning_rate": 0.00010883801045973425, | |
| "loss": 0.2812, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.2611839771270752, | |
| "eval_runtime": 174.0726, | |
| "eval_samples_per_second": 13.345, | |
| "eval_steps_per_second": 0.419, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.20671514454621873, | |
| "learning_rate": 0.00010847066258998053, | |
| "loss": 0.2973, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.18947687688376835, | |
| "learning_rate": 0.00010810319955073601, | |
| "loss": 0.2752, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.20255312746628473, | |
| "learning_rate": 0.00010773562633813061, | |
| "loss": 0.2548, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.18423910547834596, | |
| "learning_rate": 0.00010736794794979227, | |
| "loss": 0.2553, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.19708052534061443, | |
| "learning_rate": 0.00010700016938477891, | |
| "loss": 0.2494, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.17998530753425868, | |
| "learning_rate": 0.00010663229564351041, | |
| "loss": 0.2546, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.2085635770627596, | |
| "learning_rate": 0.00010626433172770078, | |
| "loss": 0.2596, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.18909283426176846, | |
| "learning_rate": 0.00010589628264029006, | |
| "loss": 0.2384, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.20378232017976675, | |
| "learning_rate": 0.0001055281533853762, | |
| "loss": 0.2283, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.208268642315818, | |
| "learning_rate": 0.00010515994896814731, | |
| "loss": 0.2436, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.1809796643912464, | |
| "learning_rate": 0.00010479167439481328, | |
| "loss": 0.2476, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.1954805890621439, | |
| "learning_rate": 0.00010442333467253789, | |
| "loss": 0.2412, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.1812710951170547, | |
| "learning_rate": 0.00010405493480937076, | |
| "loss": 0.2325, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.20086044909183742, | |
| "learning_rate": 0.00010368647981417916, | |
| "loss": 0.2552, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.18278992994480658, | |
| "learning_rate": 0.00010331797469657992, | |
| "loss": 0.2381, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.17652404986801007, | |
| "learning_rate": 0.00010294942446687146, | |
| "loss": 0.2528, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.19696889409419022, | |
| "learning_rate": 0.00010258083413596537, | |
| "loss": 0.2251, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.20649006123097716, | |
| "learning_rate": 0.00010221220871531869, | |
| "loss": 0.241, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.2048554346722386, | |
| "learning_rate": 0.00010184355321686538, | |
| "loss": 0.2507, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.19936391739850573, | |
| "learning_rate": 0.00010147487265294844, | |
| "loss": 0.2483, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.20014649703171655, | |
| "learning_rate": 0.0001011061720362516, | |
| "loss": 0.243, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.19698954610511737, | |
| "learning_rate": 0.00010073745637973124, | |
| "loss": 0.2655, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.18526644345070792, | |
| "learning_rate": 0.00010036873069654832, | |
| "loss": 0.2372, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.19481973475599917, | |
| "learning_rate": 0.0001, | |
| "loss": 0.2446, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.20394580318594288, | |
| "learning_rate": 9.96312693034517e-05, | |
| "loss": 0.2511, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.220457159395997, | |
| "learning_rate": 9.926254362026875e-05, | |
| "loss": 0.2445, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.18633719328240075, | |
| "learning_rate": 9.889382796374844e-05, | |
| "loss": 0.2377, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.1927353006687828, | |
| "learning_rate": 9.852512734705159e-05, | |
| "loss": 0.2528, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.22794683697537815, | |
| "learning_rate": 9.815644678313462e-05, | |
| "loss": 0.2592, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.20639017103117685, | |
| "learning_rate": 9.778779128468132e-05, | |
| "loss": 0.2433, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.20823587084943018, | |
| "learning_rate": 9.741916586403465e-05, | |
| "loss": 0.2411, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.20402677890788284, | |
| "learning_rate": 9.705057553312856e-05, | |
| "loss": 0.2577, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.19085117547328076, | |
| "learning_rate": 9.668202530342009e-05, | |
| "loss": 0.2209, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.194053406187504, | |
| "learning_rate": 9.631352018582089e-05, | |
| "loss": 0.241, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.20333052287999634, | |
| "learning_rate": 9.594506519062925e-05, | |
| "loss": 0.2526, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.1889581802336701, | |
| "learning_rate": 9.557666532746213e-05, | |
| "loss": 0.2118, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.1947056880617223, | |
| "learning_rate": 9.520832560518673e-05, | |
| "loss": 0.2245, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.19441196737143498, | |
| "learning_rate": 9.48400510318527e-05, | |
| "loss": 0.2498, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.19134129181023776, | |
| "learning_rate": 9.447184661462382e-05, | |
| "loss": 0.2186, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.1932180061270486, | |
| "learning_rate": 9.410371735970997e-05, | |
| "loss": 0.2376, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.21806461386450074, | |
| "learning_rate": 9.373566827229923e-05, | |
| "loss": 0.2519, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.19951555170904356, | |
| "learning_rate": 9.336770435648964e-05, | |
| "loss": 0.2467, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.22251564739691396, | |
| "learning_rate": 9.29998306152211e-05, | |
| "loss": 0.2661, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.21386539336545002, | |
| "learning_rate": 9.263205205020775e-05, | |
| "loss": 0.243, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.2097758786130228, | |
| "learning_rate": 9.226437366186941e-05, | |
| "loss": 0.2278, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.227474113330377, | |
| "learning_rate": 9.1896800449264e-05, | |
| "loss": 0.2381, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.2127655256693725, | |
| "learning_rate": 9.15293374100195e-05, | |
| "loss": 0.2156, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.21595500178149832, | |
| "learning_rate": 9.116198954026577e-05, | |
| "loss": 0.2534, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.20638365496977454, | |
| "learning_rate": 9.07947618345669e-05, | |
| "loss": 0.2252, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.2111994035780588, | |
| "learning_rate": 9.042765928585327e-05, | |
| "loss": 0.2427, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.2015346478285064, | |
| "learning_rate": 9.006068688535342e-05, | |
| "loss": 0.2434, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.20144655863058586, | |
| "learning_rate": 8.969384962252646e-05, | |
| "loss": 0.2324, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.1966831938081588, | |
| "learning_rate": 8.932715248499418e-05, | |
| "loss": 0.2342, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.2077508601935168, | |
| "learning_rate": 8.896060045847304e-05, | |
| "loss": 0.2462, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.21104976527581248, | |
| "learning_rate": 8.859419852670664e-05, | |
| "loss": 0.2381, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.22776314339414724, | |
| "learning_rate": 8.822795167139788e-05, | |
| "loss": 0.2531, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.2094206489789516, | |
| "learning_rate": 8.786186487214107e-05, | |
| "loss": 0.2416, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.21100736273500442, | |
| "learning_rate": 8.749594310635443e-05, | |
| "loss": 0.2582, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.19053851221493054, | |
| "learning_rate": 8.713019134921238e-05, | |
| "loss": 0.2238, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.19389337918079758, | |
| "learning_rate": 8.676461457357776e-05, | |
| "loss": 0.2338, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.19404522798986354, | |
| "learning_rate": 8.639921774993439e-05, | |
| "loss": 0.2391, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.21274678156633864, | |
| "learning_rate": 8.60340058463194e-05, | |
| "loss": 0.2604, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.1969735372029067, | |
| "learning_rate": 8.566898382825559e-05, | |
| "loss": 0.2219, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.1909564045414605, | |
| "learning_rate": 8.530415665868424e-05, | |
| "loss": 0.2379, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.2076961618333992, | |
| "learning_rate": 8.49395292978972e-05, | |
| "loss": 0.2343, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.2275056200459885, | |
| "learning_rate": 8.457510670346976e-05, | |
| "loss": 0.2558, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.20774675383309715, | |
| "learning_rate": 8.421089383019319e-05, | |
| "loss": 0.2549, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.2221943514742673, | |
| "learning_rate": 8.384689563000723e-05, | |
| "loss": 0.2464, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.22707800753207244, | |
| "learning_rate": 8.348311705193293e-05, | |
| "loss": 0.243, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.2447932865549153, | |
| "learning_rate": 8.311956304200533e-05, | |
| "loss": 0.264, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.2072307507148345, | |
| "learning_rate": 8.275623854320608e-05, | |
| "loss": 0.2449, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.19366898457969708, | |
| "learning_rate": 8.239314849539638e-05, | |
| "loss": 0.2402, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.20056495036693472, | |
| "learning_rate": 8.203029783524984e-05, | |
| "loss": 0.2257, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.2084543244379053, | |
| "learning_rate": 8.166769149618517e-05, | |
| "loss": 0.2493, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.207683866170019, | |
| "learning_rate": 8.130533440829928e-05, | |
| "loss": 0.2435, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.20125088482715547, | |
| "learning_rate": 8.094323149830023e-05, | |
| "loss": 0.2413, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.18512657144962646, | |
| "learning_rate": 8.058138768944013e-05, | |
| "loss": 0.2272, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.19840383281608243, | |
| "learning_rate": 8.021980790144827e-05, | |
| "loss": 0.2301, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.19751451861458907, | |
| "learning_rate": 7.985849705046439e-05, | |
| "loss": 0.2392, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.21690288735429328, | |
| "learning_rate": 7.949746004897151e-05, | |
| "loss": 0.2441, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.21086255260898568, | |
| "learning_rate": 7.913670180572936e-05, | |
| "loss": 0.2533, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.21659430069441296, | |
| "learning_rate": 7.877622722570771e-05, | |
| "loss": 0.2451, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.21073430482336133, | |
| "learning_rate": 7.84160412100194e-05, | |
| "loss": 0.2354, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.19999698025688695, | |
| "learning_rate": 7.805614865585396e-05, | |
| "loss": 0.2274, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.21741693336883022, | |
| "learning_rate": 7.769655445641093e-05, | |
| "loss": 0.243, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.19692922935201004, | |
| "learning_rate": 7.733726350083332e-05, | |
| "loss": 0.237, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.20151776142559946, | |
| "learning_rate": 7.697828067414119e-05, | |
| "loss": 0.2363, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.22663937283958677, | |
| "learning_rate": 7.661961085716512e-05, | |
| "loss": 0.2426, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.21759095693447697, | |
| "learning_rate": 7.626125892647997e-05, | |
| "loss": 0.2675, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.20761753236557173, | |
| "learning_rate": 7.590322975433857e-05, | |
| "loss": 0.237, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.18838661893932568, | |
| "learning_rate": 7.554552820860534e-05, | |
| "loss": 0.2258, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.2110298449100788, | |
| "learning_rate": 7.518815915269024e-05, | |
| "loss": 0.2496, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.20854600045531033, | |
| "learning_rate": 7.483112744548268e-05, | |
| "loss": 0.2626, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.19208651929110873, | |
| "learning_rate": 7.447443794128525e-05, | |
| "loss": 0.2335, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.18632455778406, | |
| "learning_rate": 7.411809548974792e-05, | |
| "loss": 0.2247, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.20794441259868549, | |
| "learning_rate": 7.376210493580212e-05, | |
| "loss": 0.2479, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.20925285349145628, | |
| "learning_rate": 7.34064711195946e-05, | |
| "loss": 0.2438, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.19745634320778846, | |
| "learning_rate": 7.305119887642191e-05, | |
| "loss": 0.2622, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.1856897506744385, | |
| "learning_rate": 7.269629303666463e-05, | |
| "loss": 0.2251, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.19811525650313866, | |
| "learning_rate": 7.234175842572145e-05, | |
| "loss": 0.2464, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.20082814748751002, | |
| "learning_rate": 7.198759986394381e-05, | |
| "loss": 0.2406, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.20448100943924474, | |
| "learning_rate": 7.163382216657034e-05, | |
| "loss": 0.2536, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.2087548947902547, | |
| "learning_rate": 7.128043014366123e-05, | |
| "loss": 0.2432, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.18979837479724626, | |
| "learning_rate": 7.092742860003297e-05, | |
| "loss": 0.2286, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.1938685062447375, | |
| "learning_rate": 7.057482233519302e-05, | |
| "loss": 0.2332, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.19389222639217057, | |
| "learning_rate": 7.022261614327448e-05, | |
| "loss": 0.2259, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.2028721556190883, | |
| "learning_rate": 6.98708148129709e-05, | |
| "loss": 0.2522, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.2030667349807296, | |
| "learning_rate": 6.951942312747134e-05, | |
| "loss": 0.2414, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.2118463451987899, | |
| "learning_rate": 6.916844586439513e-05, | |
| "loss": 0.2521, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.21226696216160676, | |
| "learning_rate": 6.881788779572705e-05, | |
| "loss": 0.2443, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.20461446026763713, | |
| "learning_rate": 6.84677536877523e-05, | |
| "loss": 0.2323, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.19401289577735006, | |
| "learning_rate": 6.811804830099187e-05, | |
| "loss": 0.2452, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.19494475075646883, | |
| "learning_rate": 6.776877639013777e-05, | |
| "loss": 0.2276, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_loss": 0.25228264927864075, | |
| "eval_runtime": 174.1389, | |
| "eval_samples_per_second": 13.34, | |
| "eval_steps_per_second": 0.419, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.18369247947269843, | |
| "learning_rate": 6.741994270398826e-05, | |
| "loss": 0.2327, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.19844990451218875, | |
| "learning_rate": 6.70715519853834e-05, | |
| "loss": 0.2121, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.2000478247742355, | |
| "learning_rate": 6.672360897114062e-05, | |
| "loss": 0.216, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.2066389631842727, | |
| "learning_rate": 6.637611839199013e-05, | |
| "loss": 0.2245, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.19561029619648587, | |
| "learning_rate": 6.602908497251072e-05, | |
| "loss": 0.2269, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.17858369393459886, | |
| "learning_rate": 6.568251343106562e-05, | |
| "loss": 0.2371, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.2042913512728603, | |
| "learning_rate": 6.533640847973808e-05, | |
| "loss": 0.2347, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.19680273871868867, | |
| "learning_rate": 6.499077482426756e-05, | |
| "loss": 0.2246, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.2019623623327301, | |
| "learning_rate": 6.464561716398565e-05, | |
| "loss": 0.225, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.22138397726642448, | |
| "learning_rate": 6.430094019175217e-05, | |
| "loss": 0.249, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.20441012808266243, | |
| "learning_rate": 6.395674859389127e-05, | |
| "loss": 0.2226, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.22123583080158427, | |
| "learning_rate": 6.361304705012798e-05, | |
| "loss": 0.242, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.20417002392136502, | |
| "learning_rate": 6.326984023352435e-05, | |
| "loss": 0.2214, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.20372731516937811, | |
| "learning_rate": 6.292713281041588e-05, | |
| "loss": 0.2308, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.19749044128114374, | |
| "learning_rate": 6.258492944034833e-05, | |
| "loss": 0.2331, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.19648293746816736, | |
| "learning_rate": 6.224323477601417e-05, | |
| "loss": 0.2325, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.19271942162617067, | |
| "learning_rate": 6.190205346318927e-05, | |
| "loss": 0.2312, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.21072725131725892, | |
| "learning_rate": 6.156139014066999e-05, | |
| "loss": 0.2423, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.19198580990324404, | |
| "learning_rate": 6.122124944020977e-05, | |
| "loss": 0.2347, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.21149357156024817, | |
| "learning_rate": 6.0881635986456555e-05, | |
| "loss": 0.2246, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.19079790950210723, | |
| "learning_rate": 6.054255439688947e-05, | |
| "loss": 0.2266, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.19482559547798342, | |
| "learning_rate": 6.020400928175637e-05, | |
| "loss": 0.2265, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.19971007179240052, | |
| "learning_rate": 5.986600524401112e-05, | |
| "loss": 0.2355, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.19343926484407545, | |
| "learning_rate": 5.9528546879250825e-05, | |
| "loss": 0.2218, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.19930073976975177, | |
| "learning_rate": 5.91916387756535e-05, | |
| "loss": 0.2403, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.20394273446146363, | |
| "learning_rate": 5.885528551391576e-05, | |
| "loss": 0.2105, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.22138051923326063, | |
| "learning_rate": 5.8519491667190264e-05, | |
| "loss": 0.2329, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.20190957372704954, | |
| "learning_rate": 5.818426180102382e-05, | |
| "loss": 0.2486, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.20629696816954632, | |
| "learning_rate": 5.784960047329519e-05, | |
| "loss": 0.2228, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.20235775048560617, | |
| "learning_rate": 5.751551223415313e-05, | |
| "loss": 0.24, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.1922696841143549, | |
| "learning_rate": 5.718200162595449e-05, | |
| "loss": 0.2286, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.19801052068324668, | |
| "learning_rate": 5.684907318320247e-05, | |
| "loss": 0.2328, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.22189721177634955, | |
| "learning_rate": 5.651673143248508e-05, | |
| "loss": 0.2366, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.20510345484208217, | |
| "learning_rate": 5.61849808924134e-05, | |
| "loss": 0.2445, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.20437222653878181, | |
| "learning_rate": 5.5853826073560314e-05, | |
| "loss": 0.2283, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.18788556351853072, | |
| "learning_rate": 5.5523271478399084e-05, | |
| "loss": 0.2264, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.1947377617771126, | |
| "learning_rate": 5.5193321601242156e-05, | |
| "loss": 0.2369, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.19219173101094308, | |
| "learning_rate": 5.48639809281801e-05, | |
| "loss": 0.2267, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.21019734225509798, | |
| "learning_rate": 5.4535253937020526e-05, | |
| "loss": 0.2342, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.19871629669119603, | |
| "learning_rate": 5.420714509722729e-05, | |
| "loss": 0.2412, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.18878853628731276, | |
| "learning_rate": 5.387965886985966e-05, | |
| "loss": 0.2125, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.19185339023163583, | |
| "learning_rate": 5.355279970751174e-05, | |
| "loss": 0.2335, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.19314964556917028, | |
| "learning_rate": 5.322657205425183e-05, | |
| "loss": 0.2172, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.19994293680338385, | |
| "learning_rate": 5.29009803455621e-05, | |
| "loss": 0.2315, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.19241810973460463, | |
| "learning_rate": 5.257602900827821e-05, | |
| "loss": 0.2172, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.22168479611060188, | |
| "learning_rate": 5.225172246052914e-05, | |
| "loss": 0.2308, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.21457620781731124, | |
| "learning_rate": 5.192806511167718e-05, | |
| "loss": 0.2531, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.2202429827455574, | |
| "learning_rate": 5.160506136225787e-05, | |
| "loss": 0.2587, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.1981490531845343, | |
| "learning_rate": 5.1282715603920374e-05, | |
| "loss": 0.2374, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.20755864391862947, | |
| "learning_rate": 5.0961032219367374e-05, | |
| "loss": 0.2328, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.2050356684741965, | |
| "learning_rate": 5.064001558229591e-05, | |
| "loss": 0.234, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.19524780155524296, | |
| "learning_rate": 5.03196700573378e-05, | |
| "loss": 0.226, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.2207091761071787, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.2284, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.207367533948898, | |
| "learning_rate": 4.968100975660582e-05, | |
| "loss": 0.2275, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.20040401651948256, | |
| "learning_rate": 4.936270366423563e-05, | |
| "loss": 0.2349, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.19722408386051055, | |
| "learning_rate": 4.904508605066773e-05, | |
| "loss": 0.2266, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.21069454400535126, | |
| "learning_rate": 4.872816123431977e-05, | |
| "loss": 0.2399, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.27442431176210424, | |
| "learning_rate": 4.841193352419003e-05, | |
| "loss": 0.255, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.21268499575308422, | |
| "learning_rate": 4.809640721979855e-05, | |
| "loss": 0.2536, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.1958296129113393, | |
| "learning_rate": 4.778158661112896e-05, | |
| "loss": 0.2227, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.2037287597321259, | |
| "learning_rate": 4.746747597857014e-05, | |
| "loss": 0.264, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.21857028629384495, | |
| "learning_rate": 4.715407959285776e-05, | |
| "loss": 0.2539, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.17653319803507914, | |
| "learning_rate": 4.6841401715016463e-05, | |
| "loss": 0.2173, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.2088118045650298, | |
| "learning_rate": 4.6529446596301994e-05, | |
| "loss": 0.2181, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.20018277943155008, | |
| "learning_rate": 4.6218218478142984e-05, | |
| "loss": 0.2451, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.19840613289975464, | |
| "learning_rate": 4.590772159208384e-05, | |
| "loss": 0.2222, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.1995822171363863, | |
| "learning_rate": 4.559796015972677e-05, | |
| "loss": 0.241, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.19895719674336662, | |
| "learning_rate": 4.52889383926746e-05, | |
| "loss": 0.2371, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.18187801681389232, | |
| "learning_rate": 4.498066049247344e-05, | |
| "loss": 0.2218, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.2177749086338868, | |
| "learning_rate": 4.4673130650555605e-05, | |
| "loss": 0.2486, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.21347066621586283, | |
| "learning_rate": 4.436635304818258e-05, | |
| "loss": 0.2262, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.20310412813205128, | |
| "learning_rate": 4.40603318563882e-05, | |
| "loss": 0.227, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.20257831329266338, | |
| "learning_rate": 4.375507123592194e-05, | |
| "loss": 0.2071, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.21375954762824598, | |
| "learning_rate": 4.34505753371923e-05, | |
| "loss": 0.2316, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.21010933001283658, | |
| "learning_rate": 4.314684830021046e-05, | |
| "loss": 0.2269, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.23866662146097914, | |
| "learning_rate": 4.2843894254533904e-05, | |
| "loss": 0.2621, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.202396176516729, | |
| "learning_rate": 4.2541717319210315e-05, | |
| "loss": 0.2229, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.2114385200169175, | |
| "learning_rate": 4.224032160272159e-05, | |
| "loss": 0.2244, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.19545378343350944, | |
| "learning_rate": 4.1939711202927936e-05, | |
| "loss": 0.2084, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.2037569083756898, | |
| "learning_rate": 4.1639890207012176e-05, | |
| "loss": 0.2336, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.21516998837226808, | |
| "learning_rate": 4.1340862691424184e-05, | |
| "loss": 0.2236, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.20361533038162538, | |
| "learning_rate": 4.104263272182546e-05, | |
| "loss": 0.2298, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.2299271783012964, | |
| "learning_rate": 4.074520435303383e-05, | |
| "loss": 0.2436, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.20022559518426747, | |
| "learning_rate": 4.044858162896834e-05, | |
| "loss": 0.2399, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.212614510340799, | |
| "learning_rate": 4.015276858259427e-05, | |
| "loss": 0.2287, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.20401412514620756, | |
| "learning_rate": 3.985776923586829e-05, | |
| "loss": 0.2314, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.1987739253038705, | |
| "learning_rate": 3.9563587599683795e-05, | |
| "loss": 0.2133, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.18748268405111385, | |
| "learning_rate": 3.927022767381634e-05, | |
| "loss": 0.2099, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.19769569764762104, | |
| "learning_rate": 3.897769344686929e-05, | |
| "loss": 0.2294, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.23235953320355546, | |
| "learning_rate": 3.868598889621957e-05, | |
| "loss": 0.2433, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.19438645171196242, | |
| "learning_rate": 3.839511798796357e-05, | |
| "loss": 0.2317, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.23669503631409483, | |
| "learning_rate": 3.810508467686337e-05, | |
| "loss": 0.2455, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.21634566116480686, | |
| "learning_rate": 3.781589290629266e-05, | |
| "loss": 0.2343, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.20952846432796216, | |
| "learning_rate": 3.752754660818339e-05, | |
| "loss": 0.2135, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.20360318165711974, | |
| "learning_rate": 3.7240049702972345e-05, | |
| "loss": 0.2068, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.21078292163104015, | |
| "learning_rate": 3.6953406099547514e-05, | |
| "loss": 0.2365, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.1948997332089136, | |
| "learning_rate": 3.6667619695195285e-05, | |
| "loss": 0.2232, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.20947930741143364, | |
| "learning_rate": 3.638269437554741e-05, | |
| "loss": 0.2324, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.2012585295919145, | |
| "learning_rate": 3.609863401452787e-05, | |
| "loss": 0.2135, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.2118635337112509, | |
| "learning_rate": 3.581544247430056e-05, | |
| "loss": 0.2296, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.20679430628191067, | |
| "learning_rate": 3.553312360521675e-05, | |
| "loss": 0.2228, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.20244338222777755, | |
| "learning_rate": 3.525168124576239e-05, | |
| "loss": 0.2214, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.21511489923748392, | |
| "learning_rate": 3.49711192225063e-05, | |
| "loss": 0.2461, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.20512373280874766, | |
| "learning_rate": 3.469144135004809e-05, | |
| "loss": 0.224, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.2298619558228232, | |
| "learning_rate": 3.4412651430966e-05, | |
| "loss": 0.2572, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.19862016808278388, | |
| "learning_rate": 3.413475325576554e-05, | |
| "loss": 0.2403, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.22555024826650924, | |
| "learning_rate": 3.385775060282783e-05, | |
| "loss": 0.2282, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.19215563610758657, | |
| "learning_rate": 3.3581647238358215e-05, | |
| "loss": 0.2144, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.20476097833745566, | |
| "learning_rate": 3.330644691633492e-05, | |
| "loss": 0.225, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.19750951099732988, | |
| "learning_rate": 3.3032153378458376e-05, | |
| "loss": 0.2287, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.21089379422961432, | |
| "learning_rate": 3.275877035410002e-05, | |
| "loss": 0.2321, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.2078264905766209, | |
| "learning_rate": 3.248630156025158e-05, | |
| "loss": 0.2381, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.2179758901349511, | |
| "learning_rate": 3.2214750701474874e-05, | |
| "loss": 0.2483, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_loss": 0.2440394163131714, | |
| "eval_runtime": 174.4083, | |
| "eval_samples_per_second": 13.319, | |
| "eval_steps_per_second": 0.419, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.2131438455471886, | |
| "learning_rate": 3.194412146985106e-05, | |
| "loss": 0.2348, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.20229002928013867, | |
| "learning_rate": 3.167441754493066e-05, | |
| "loss": 0.2186, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.20234425592543342, | |
| "learning_rate": 3.1405642593683425e-05, | |
| "loss": 0.2273, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.22859532849176858, | |
| "learning_rate": 3.113780027044856e-05, | |
| "loss": 0.2254, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.20559865195742957, | |
| "learning_rate": 3.0870894216884995e-05, | |
| "loss": 0.2291, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.21685988567299588, | |
| "learning_rate": 3.060492806192185e-05, | |
| "loss": 0.2164, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.19802233527020463, | |
| "learning_rate": 3.0339905421709137e-05, | |
| "loss": 0.234, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.19515973792765512, | |
| "learning_rate": 3.0075829899568597e-05, | |
| "loss": 0.2232, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.19173680197953472, | |
| "learning_rate": 2.981270508594467e-05, | |
| "loss": 0.2308, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.19991596550087715, | |
| "learning_rate": 2.955053455835568e-05, | |
| "loss": 0.218, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.18665703388000335, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.2366, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.18942165113318024, | |
| "learning_rate": 2.9029070606433785e-05, | |
| "loss": 0.2132, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.20485337976012102, | |
| "learning_rate": 2.8769784272070188e-05, | |
| "loss": 0.2335, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.20246216943076425, | |
| "learning_rate": 2.8511466403583766e-05, | |
| "loss": 0.2352, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.19311011267813682, | |
| "learning_rate": 2.8254120513136296e-05, | |
| "loss": 0.2067, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.210224208337888, | |
| "learning_rate": 2.799775009967428e-05, | |
| "loss": 0.2333, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.21970393594166632, | |
| "learning_rate": 2.7742358648881352e-05, | |
| "loss": 0.229, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.19295096373345785, | |
| "learning_rate": 2.7487949633130882e-05, | |
| "loss": 0.2204, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.19297425613578706, | |
| "learning_rate": 2.7234526511438818e-05, | |
| "loss": 0.2307, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.21270027750348675, | |
| "learning_rate": 2.6982092729416587e-05, | |
| "loss": 0.2304, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.1821212153306606, | |
| "learning_rate": 2.6730651719224285e-05, | |
| "loss": 0.2217, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.20641341779566383, | |
| "learning_rate": 2.6480206899524007e-05, | |
| "loss": 0.2297, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.19895365786093475, | |
| "learning_rate": 2.623076167543337e-05, | |
| "loss": 0.2294, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.211723910335701, | |
| "learning_rate": 2.5982319438479164e-05, | |
| "loss": 0.2279, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.20694547351763354, | |
| "learning_rate": 2.5734883566551414e-05, | |
| "loss": 0.2176, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.20794623990405123, | |
| "learning_rate": 2.548845742385717e-05, | |
| "loss": 0.241, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.18729243114097596, | |
| "learning_rate": 2.5243044360874968e-05, | |
| "loss": 0.2268, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.2096570854756017, | |
| "learning_rate": 2.4998647714309342e-05, | |
| "loss": 0.2233, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.20227926508122493, | |
| "learning_rate": 2.4755270807045174e-05, | |
| "loss": 0.2287, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.20682124517055636, | |
| "learning_rate": 2.451291694810275e-05, | |
| "loss": 0.2315, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.19809452439226158, | |
| "learning_rate": 2.42715894325928e-05, | |
| "loss": 0.2219, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.2095130134600125, | |
| "learning_rate": 2.403129154167153e-05, | |
| "loss": 0.2291, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.20788084999909145, | |
| "learning_rate": 2.379202654249596e-05, | |
| "loss": 0.2428, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.2172405068460128, | |
| "learning_rate": 2.3553797688179856e-05, | |
| "loss": 0.2372, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.19461394318035657, | |
| "learning_rate": 2.3316608217749147e-05, | |
| "loss": 0.2145, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.18940028210697493, | |
| "learning_rate": 2.3080461356097937e-05, | |
| "loss": 0.221, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.1994412806184569, | |
| "learning_rate": 2.2845360313944885e-05, | |
| "loss": 0.2192, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.22225917578862417, | |
| "learning_rate": 2.2611308287789344e-05, | |
| "loss": 0.2326, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.22182577419896116, | |
| "learning_rate": 2.2378308459867825e-05, | |
| "loss": 0.2233, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.20974195726874859, | |
| "learning_rate": 2.2146363998111076e-05, | |
| "loss": 0.2075, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.20126278684107707, | |
| "learning_rate": 2.191547805610066e-05, | |
| "loss": 0.2272, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.2115414591515168, | |
| "learning_rate": 2.168565377302615e-05, | |
| "loss": 0.2011, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.22823267316076018, | |
| "learning_rate": 2.145689427364268e-05, | |
| "loss": 0.2346, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.20563465172572046, | |
| "learning_rate": 2.1229202668228197e-05, | |
| "loss": 0.223, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.2125772917653009, | |
| "learning_rate": 2.1002582052541188e-05, | |
| "loss": 0.2351, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.20342137649326203, | |
| "learning_rate": 2.077703550777882e-05, | |
| "loss": 0.2214, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.2182233891977982, | |
| "learning_rate": 2.055256610053479e-05, | |
| "loss": 0.2355, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.20932056445296426, | |
| "learning_rate": 2.032917688275774e-05, | |
| "loss": 0.233, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.19761955501105657, | |
| "learning_rate": 2.0106870891709795e-05, | |
| "loss": 0.2166, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.20268753517625793, | |
| "learning_rate": 1.988565114992519e-05, | |
| "loss": 0.2322, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.21415195057982833, | |
| "learning_rate": 1.9665520665169235e-05, | |
| "loss": 0.2333, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.18762518756686802, | |
| "learning_rate": 1.9446482430397385e-05, | |
| "loss": 0.2183, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.17327672319544696, | |
| "learning_rate": 1.9228539423714565e-05, | |
| "loss": 0.1868, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.20722565430523956, | |
| "learning_rate": 1.9011694608334675e-05, | |
| "loss": 0.2233, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.21924150970052983, | |
| "learning_rate": 1.8795950932540284e-05, | |
| "loss": 0.2208, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.20907553351335467, | |
| "learning_rate": 1.858131132964259e-05, | |
| "loss": 0.2223, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.19056857008632616, | |
| "learning_rate": 1.8367778717941474e-05, | |
| "loss": 0.2101, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.21084097137078958, | |
| "learning_rate": 1.8155356000685895e-05, | |
| "loss": 0.2453, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.20113307615771472, | |
| "learning_rate": 1.794404606603434e-05, | |
| "loss": 0.2285, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.1977232826916787, | |
| "learning_rate": 1.773385178701561e-05, | |
| "loss": 0.2066, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.22423101983076504, | |
| "learning_rate": 1.752477602148973e-05, | |
| "loss": 0.2358, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.1879155996937992, | |
| "learning_rate": 1.7316821612109136e-05, | |
| "loss": 0.2272, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.19578435167746402, | |
| "learning_rate": 1.7109991386279966e-05, | |
| "loss": 0.2179, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.19857068398335803, | |
| "learning_rate": 1.6904288156123636e-05, | |
| "loss": 0.2261, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.20885572306052286, | |
| "learning_rate": 1.6699714718438642e-05, | |
| "loss": 0.2323, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.20555096566027478, | |
| "learning_rate": 1.649627385466248e-05, | |
| "loss": 0.2159, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.22223342099052376, | |
| "learning_rate": 1.6293968330833874e-05, | |
| "loss": 0.2445, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.20268806818030535, | |
| "learning_rate": 1.609280089755515e-05, | |
| "loss": 0.2336, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.21879721029918314, | |
| "learning_rate": 1.5892774289954814e-05, | |
| "loss": 0.2326, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.1979735202861903, | |
| "learning_rate": 1.5693891227650426e-05, | |
| "loss": 0.2444, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.19636639452399096, | |
| "learning_rate": 1.549615441471153e-05, | |
| "loss": 0.2229, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.19526467691493388, | |
| "learning_rate": 1.5299566539623046e-05, | |
| "loss": 0.2212, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.21208951867669978, | |
| "learning_rate": 1.5104130275248452e-05, | |
| "loss": 0.2186, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.20045769407509645, | |
| "learning_rate": 1.4909848278793782e-05, | |
| "loss": 0.2129, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.2147971015918436, | |
| "learning_rate": 1.4716723191771231e-05, | |
| "loss": 0.2262, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.206242119031671, | |
| "learning_rate": 1.452475763996326e-05, | |
| "loss": 0.2236, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.20912939904357944, | |
| "learning_rate": 1.4333954233387136e-05, | |
| "loss": 0.2123, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.20917641203494783, | |
| "learning_rate": 1.414431556625917e-05, | |
| "loss": 0.2104, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.2084574875835098, | |
| "learning_rate": 1.3955844216959502e-05, | |
| "loss": 0.2313, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.23628135092170213, | |
| "learning_rate": 1.3768542747997215e-05, | |
| "loss": 0.2307, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.21532128754886334, | |
| "learning_rate": 1.3582413705975327e-05, | |
| "loss": 0.229, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.19053700675294885, | |
| "learning_rate": 1.339745962155613e-05, | |
| "loss": 0.2185, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.20258235891664356, | |
| "learning_rate": 1.3213683009427003e-05, | |
| "loss": 0.2267, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.20530653449904004, | |
| "learning_rate": 1.3031086368265998e-05, | |
| "loss": 0.24, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.2065814367954602, | |
| "learning_rate": 1.2849672180707928e-05, | |
| "loss": 0.2405, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.20567628728139853, | |
| "learning_rate": 1.2669442913310725e-05, | |
| "loss": 0.2144, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.21176419914000316, | |
| "learning_rate": 1.249040101652178e-05, | |
| "loss": 0.2173, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.21884981500734577, | |
| "learning_rate": 1.2312548924644585e-05, | |
| "loss": 0.2405, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.19687803356921588, | |
| "learning_rate": 1.2135889055805837e-05, | |
| "loss": 0.2204, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.20109138915685001, | |
| "learning_rate": 1.196042381192236e-05, | |
| "loss": 0.2297, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.19234466174457304, | |
| "learning_rate": 1.1786155578668545e-05, | |
| "loss": 0.2258, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.1978258826111612, | |
| "learning_rate": 1.161308672544389e-05, | |
| "loss": 0.2038, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.21048922943923923, | |
| "learning_rate": 1.1441219605340792e-05, | |
| "loss": 0.2283, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.22257127311445948, | |
| "learning_rate": 1.1270556555112532e-05, | |
| "loss": 0.2283, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.19149094284354706, | |
| "learning_rate": 1.1101099895141542e-05, | |
| "loss": 0.215, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.20549949771292994, | |
| "learning_rate": 1.0932851929407827e-05, | |
| "loss": 0.2385, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.1758290014045967, | |
| "learning_rate": 1.0765814945457653e-05, | |
| "loss": 0.1882, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.1898211512831452, | |
| "learning_rate": 1.059999121437244e-05, | |
| "loss": 0.2049, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.20222538943850363, | |
| "learning_rate": 1.0435382990737853e-05, | |
| "loss": 0.2175, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.22196981794940074, | |
| "learning_rate": 1.0271992512613237e-05, | |
| "loss": 0.2463, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.2027360971190236, | |
| "learning_rate": 1.0109822001501079e-05, | |
| "loss": 0.2145, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.20140001638042115, | |
| "learning_rate": 9.948873662316883e-06, | |
| "loss": 0.2106, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.21591974789953125, | |
| "learning_rate": 9.789149683359144e-06, | |
| "loss": 0.2421, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.20965313194879998, | |
| "learning_rate": 9.630652236279625e-06, | |
| "loss": 0.2313, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.1989390148317295, | |
| "learning_rate": 9.473383476053821e-06, | |
| "loss": 0.2319, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.17721856904495678, | |
| "learning_rate": 9.31734554095165e-06, | |
| "loss": 0.2086, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.18710003353683868, | |
| "learning_rate": 9.162540552508392e-06, | |
| "loss": 0.2135, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.19769078883988017, | |
| "learning_rate": 9.008970615495826e-06, | |
| "loss": 0.2203, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.19064162760273085, | |
| "learning_rate": 8.856637817893688e-06, | |
| "loss": 0.2022, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.1926418728967082, | |
| "learning_rate": 8.70554423086114e-06, | |
| "loss": 0.1937, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.21141489190301602, | |
| "learning_rate": 8.555691908708707e-06, | |
| "loss": 0.2266, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.20876156881963107, | |
| "learning_rate": 8.40708288887042e-06, | |
| "loss": 0.2343, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.18416645127890868, | |
| "learning_rate": 8.259719191875892e-06, | |
| "loss": 0.1982, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "eval_loss": 0.23803040385246277, | |
| "eval_runtime": 174.4554, | |
| "eval_samples_per_second": 13.316, | |
| "eval_steps_per_second": 0.418, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.2040866955904897, | |
| "learning_rate": 8.113602821323041e-06, | |
| "loss": 0.2136, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.1982756149597501, | |
| "learning_rate": 7.968735763850854e-06, | |
| "loss": 0.2167, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.21789914937542643, | |
| "learning_rate": 7.825119989112173e-06, | |
| "loss": 0.2397, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.2147093525916075, | |
| "learning_rate": 7.682757449747135e-06, | |
| "loss": 0.2358, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.20816472900077765, | |
| "learning_rate": 7.541650081356588e-06, | |
| "loss": 0.2234, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.21148961713513456, | |
| "learning_rate": 7.401799802475573e-06, | |
| "loss": 0.2319, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.1956708576913421, | |
| "learning_rate": 7.263208514547548e-06, | |
| "loss": 0.2137, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.20721619820091447, | |
| "learning_rate": 7.125878101898298e-06, | |
| "loss": 0.2188, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.22116900534807907, | |
| "learning_rate": 6.989810431710375e-06, | |
| "loss": 0.2267, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.19978803749852295, | |
| "learning_rate": 6.855007353997822e-06, | |
| "loss": 0.2101, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.21381752099366957, | |
| "learning_rate": 6.7214707015808565e-06, | |
| "loss": 0.2261, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.19219567545117702, | |
| "learning_rate": 6.589202290061014e-06, | |
| "loss": 0.2034, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.18779844136231152, | |
| "learning_rate": 6.4582039177965455e-06, | |
| "loss": 0.2061, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.20174766286042234, | |
| "learning_rate": 6.328477365877849e-06, | |
| "loss": 0.2233, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.21231582457977347, | |
| "learning_rate": 6.200024398103255e-06, | |
| "loss": 0.2128, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.18391901989584938, | |
| "learning_rate": 6.072846760955186e-06, | |
| "loss": 0.1992, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.20294386783873802, | |
| "learning_rate": 5.946946183576241e-06, | |
| "loss": 0.222, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.20469202298707184, | |
| "learning_rate": 5.822324377745791e-06, | |
| "loss": 0.2235, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.22515194087302773, | |
| "learning_rate": 5.698983037856665e-06, | |
| "loss": 0.2438, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.20643509196206, | |
| "learning_rate": 5.57692384089209e-06, | |
| "loss": 0.2131, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.2108231693828425, | |
| "learning_rate": 5.456148446402976e-06, | |
| "loss": 0.2078, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.21597167356066369, | |
| "learning_rate": 5.336658496485248e-06, | |
| "loss": 0.2222, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.1933627658122827, | |
| "learning_rate": 5.2184556157576e-06, | |
| "loss": 0.2157, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.22222814897657234, | |
| "learning_rate": 5.10154141133935e-06, | |
| "loss": 0.2533, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.20557125384851288, | |
| "learning_rate": 4.985917472828639e-06, | |
| "loss": 0.2264, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.21870759144189225, | |
| "learning_rate": 4.871585372280774e-06, | |
| "loss": 0.2208, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.20383293324564078, | |
| "learning_rate": 4.758546664186869e-06, | |
| "loss": 0.2128, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.2207651135648432, | |
| "learning_rate": 4.646802885452739e-06, | |
| "loss": 0.2331, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.20935706303755758, | |
| "learning_rate": 4.536355555377947e-06, | |
| "loss": 0.2056, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.19500728601207568, | |
| "learning_rate": 4.427206175635202e-06, | |
| "loss": 0.2216, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.1831341530266785, | |
| "learning_rate": 4.319356230249905e-06, | |
| "loss": 0.2185, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.20328065520250121, | |
| "learning_rate": 4.212807185579992e-06, | |
| "loss": 0.22, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.197367783334187, | |
| "learning_rate": 4.107560490295992e-06, | |
| "loss": 0.2088, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.199986101949123, | |
| "learning_rate": 4.003617575361318e-06, | |
| "loss": 0.2252, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.20378685375125896, | |
| "learning_rate": 3.90097985401282e-06, | |
| "loss": 0.2177, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.2089863714675373, | |
| "learning_rate": 3.7996487217416223e-06, | |
| "loss": 0.2281, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.22174346947327125, | |
| "learning_rate": 3.699625556274e-06, | |
| "loss": 0.2298, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.20182626270268705, | |
| "learning_rate": 3.600911717552824e-06, | |
| "loss": 0.2239, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.20313048355330246, | |
| "learning_rate": 3.5035085477190143e-06, | |
| "loss": 0.225, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.2123857689013769, | |
| "learning_rate": 3.40741737109318e-06, | |
| "loss": 0.2285, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.20877332815435723, | |
| "learning_rate": 3.3126394941577656e-06, | |
| "loss": 0.2179, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.20261893136881134, | |
| "learning_rate": 3.2191762055392428e-06, | |
| "loss": 0.2053, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.2165479347703477, | |
| "learning_rate": 3.127028775990515e-06, | |
| "loss": 0.234, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.20854006087046636, | |
| "learning_rate": 3.03619845837374e-06, | |
| "loss": 0.2194, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.20839858622666318, | |
| "learning_rate": 2.94668648764328e-06, | |
| "loss": 0.2011, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.2001464877253074, | |
| "learning_rate": 2.8584940808288107e-06, | |
| "loss": 0.2074, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.2003840882274367, | |
| "learning_rate": 2.7716224370188836e-06, | |
| "loss": 0.2102, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.18643967510262804, | |
| "learning_rate": 2.6860727373446472e-06, | |
| "loss": 0.2039, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.21745550785786122, | |
| "learning_rate": 2.6018461449636265e-06, | |
| "loss": 0.2225, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.22552267236084578, | |
| "learning_rate": 2.5189438050440607e-06, | |
| "loss": 0.233, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.21764248387495064, | |
| "learning_rate": 2.4373668447493224e-06, | |
| "loss": 0.2214, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.1980950415461294, | |
| "learning_rate": 2.357116373222479e-06, | |
| "loss": 0.2098, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.1880075122855425, | |
| "learning_rate": 2.2781934815713225e-06, | |
| "loss": 0.216, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.21062925563667326, | |
| "learning_rate": 2.2005992428535184e-06, | |
| "loss": 0.2247, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.2119014226422388, | |
| "learning_rate": 2.1243347120619595e-06, | |
| "loss": 0.2329, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.20411445722959737, | |
| "learning_rate": 2.0494009261104785e-06, | |
| "loss": 0.1941, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.19915577884332822, | |
| "learning_rate": 1.9757989038197146e-06, | |
| "loss": 0.2151, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.21265487566748392, | |
| "learning_rate": 1.9035296459033125e-06, | |
| "loss": 0.2401, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.1996989913105502, | |
| "learning_rate": 1.8325941349542353e-06, | |
| "loss": 0.2146, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.1933371213186986, | |
| "learning_rate": 1.7629933354314733e-06, | |
| "loss": 0.2002, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.20045617018763978, | |
| "learning_rate": 1.6947281936469218e-06, | |
| "loss": 0.2293, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.19920942808294725, | |
| "learning_rate": 1.6277996377524695e-06, | |
| "loss": 0.2183, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.1936935465614516, | |
| "learning_rate": 1.562208577727442e-06, | |
| "loss": 0.1957, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.18683523311557979, | |
| "learning_rate": 1.4979559053661663e-06, | |
| "loss": 0.2009, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.18917173941955473, | |
| "learning_rate": 1.4350424942659146e-06, | |
| "loss": 0.2109, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.19682753422160834, | |
| "learning_rate": 1.3734691998149474e-06, | |
| "loss": 0.2088, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.2048854194962544, | |
| "learning_rate": 1.3132368591809552e-06, | |
| "loss": 0.2291, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.20142459735505794, | |
| "learning_rate": 1.2543462912996463e-06, | |
| "loss": 0.2278, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.18935162489689705, | |
| "learning_rate": 1.1967982968635993e-06, | |
| "loss": 0.2113, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.20159497258222603, | |
| "learning_rate": 1.1405936583113841e-06, | |
| "loss": 0.2266, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.19452768000808426, | |
| "learning_rate": 1.0857331398169579e-06, | |
| "loss": 0.2053, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.2014716516584308, | |
| "learning_rate": 1.032217487279219e-06, | |
| "loss": 0.1975, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.19846952960822706, | |
| "learning_rate": 9.800474283119144e-07, | |
| "loss": 0.2062, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.20327559720564944, | |
| "learning_rate": 9.292236722337033e-07, | |
| "loss": 0.2108, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.19290238138729213, | |
| "learning_rate": 8.797469100585431e-07, | |
| "loss": 0.2039, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.21287614440666355, | |
| "learning_rate": 8.316178144862963e-07, | |
| "loss": 0.2307, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.22567801639890173, | |
| "learning_rate": 7.848370398935711e-07, | |
| "loss": 0.2573, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.21560397460264963, | |
| "learning_rate": 7.394052223248182e-07, | |
| "loss": 0.2241, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.1957878241687257, | |
| "learning_rate": 6.953229794837146e-07, | |
| "loss": 0.2193, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.20291967284647494, | |
| "learning_rate": 6.525909107247041e-07, | |
| "loss": 0.2064, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.18847141357662414, | |
| "learning_rate": 6.11209597044926e-07, | |
| "loss": 0.2105, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.2044881782029401, | |
| "learning_rate": 5.71179601076266e-07, | |
| "loss": 0.2177, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.21142542004279102, | |
| "learning_rate": 5.325014670776951e-07, | |
| "loss": 0.233, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.19576617003967908, | |
| "learning_rate": 4.951757209279095e-07, | |
| "loss": 0.2214, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.20534907108193845, | |
| "learning_rate": 4.5920287011818054e-07, | |
| "loss": 0.2122, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.2059415081571872, | |
| "learning_rate": 4.245834037454155e-07, | |
| "loss": 0.213, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.19398152799817778, | |
| "learning_rate": 3.913177925055189e-07, | |
| "loss": 0.2219, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.2070317988118509, | |
| "learning_rate": 3.5940648868703073e-07, | |
| "loss": 0.2244, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.8623179589154261, | |
| "learning_rate": 3.288499261649314e-07, | |
| "loss": 0.2217, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.215386416256104, | |
| "learning_rate": 2.9964852039476854e-07, | |
| "loss": 0.2229, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.20755620295808855, | |
| "learning_rate": 2.718026684070063e-07, | |
| "loss": 0.2208, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.20071657749706617, | |
| "learning_rate": 2.4531274880160713e-07, | |
| "loss": 0.212, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.20451448277112147, | |
| "learning_rate": 2.201791217428917e-07, | |
| "loss": 0.2293, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.18693359575878427, | |
| "learning_rate": 1.964021289546869e-07, | |
| "loss": 0.2067, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.1878379319628547, | |
| "learning_rate": 1.739820937155967e-07, | |
| "loss": 0.2023, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.183214719604713, | |
| "learning_rate": 1.5291932085468307e-07, | |
| "loss": 0.1914, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.20983709187003516, | |
| "learning_rate": 1.3321409674728057e-07, | |
| "loss": 0.2458, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.18418752748665865, | |
| "learning_rate": 1.1486668931111056e-07, | |
| "loss": 0.2111, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.21472529700497753, | |
| "learning_rate": 9.78773480026396e-08, | |
| "loss": 0.2597, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.19925573417089712, | |
| "learning_rate": 8.224630381369335e-08, | |
| "loss": 0.2274, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.288190699780251, | |
| "learning_rate": 6.797376926829246e-08, | |
| "loss": 0.226, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.20512892684223066, | |
| "learning_rate": 5.5059938419821377e-08, | |
| "loss": 0.209, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.21184481530963395, | |
| "learning_rate": 4.350498684829729e-08, | |
| "loss": 0.223, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.20194019888677378, | |
| "learning_rate": 3.330907165809416e-08, | |
| "loss": 0.219, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.21240034431305688, | |
| "learning_rate": 2.447233147570005e-08, | |
| "loss": 0.2052, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.2164638642279727, | |
| "learning_rate": 1.6994886447896376e-08, | |
| "loss": 0.2264, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.21008301667873402, | |
| "learning_rate": 1.0876838240125863e-08, | |
| "loss": 0.2419, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.20392637628490956, | |
| "learning_rate": 6.1182700350714825e-09, | |
| "loss": 0.2225, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.21551655412638365, | |
| "learning_rate": 2.719246531535102e-09, | |
| "loss": 0.2395, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.1985114294041871, | |
| "learning_rate": 6.798139436159367e-10, | |
| "loss": 0.2104, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.2212692948861265, | |
| "learning_rate": 0.0, | |
| "loss": 0.2445, | |
| "step": 902 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 902, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 451, | |
| "total_flos": 2.635365103453392e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |