Files
qwen2.5-3B-SFT/trainer_state.json
ModelHub XC 6a3ccb98ad 初始化项目,由ModelHub XC社区提供模型
Model: ricemonster/qwen2.5-3B-SFT
Source: Original Platform
2026-08-24 20:23:59 +08:00

2191 lines
51 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 15444,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006475006475006475,
"grad_norm": 1.1328125,
"learning_rate": 4.9841362341362344e-05,
"loss": 0.4244,
"step": 50
},
{
"epoch": 0.01295001295001295,
"grad_norm": 1.0078125,
"learning_rate": 4.9679487179487185e-05,
"loss": 0.3459,
"step": 100
},
{
"epoch": 0.019425019425019424,
"grad_norm": 0.98828125,
"learning_rate": 4.951761201761202e-05,
"loss": 0.3259,
"step": 150
},
{
"epoch": 0.0259000259000259,
"grad_norm": 0.9453125,
"learning_rate": 4.935573685573686e-05,
"loss": 0.3187,
"step": 200
},
{
"epoch": 0.03237503237503238,
"grad_norm": 0.953125,
"learning_rate": 4.91938616938617e-05,
"loss": 0.3079,
"step": 250
},
{
"epoch": 0.03885003885003885,
"grad_norm": 0.875,
"learning_rate": 4.903198653198653e-05,
"loss": 0.3066,
"step": 300
},
{
"epoch": 0.045325045325045325,
"grad_norm": 0.9921875,
"learning_rate": 4.887011137011137e-05,
"loss": 0.3011,
"step": 350
},
{
"epoch": 0.0518000518000518,
"grad_norm": 0.86328125,
"learning_rate": 4.8708236208236214e-05,
"loss": 0.2987,
"step": 400
},
{
"epoch": 0.05827505827505827,
"grad_norm": 0.87890625,
"learning_rate": 4.854636104636105e-05,
"loss": 0.2966,
"step": 450
},
{
"epoch": 0.06475006475006476,
"grad_norm": 0.87109375,
"learning_rate": 4.838448588448589e-05,
"loss": 0.2951,
"step": 500
},
{
"epoch": 0.07122507122507123,
"grad_norm": 0.83984375,
"learning_rate": 4.822261072261072e-05,
"loss": 0.2915,
"step": 550
},
{
"epoch": 0.0777000777000777,
"grad_norm": 0.8515625,
"learning_rate": 4.806073556073556e-05,
"loss": 0.2889,
"step": 600
},
{
"epoch": 0.08417508417508418,
"grad_norm": 0.8125,
"learning_rate": 4.78988603988604e-05,
"loss": 0.295,
"step": 650
},
{
"epoch": 0.09065009065009065,
"grad_norm": 0.79296875,
"learning_rate": 4.7736985236985236e-05,
"loss": 0.2869,
"step": 700
},
{
"epoch": 0.09712509712509712,
"grad_norm": 0.859375,
"learning_rate": 4.757511007511008e-05,
"loss": 0.2888,
"step": 750
},
{
"epoch": 0.1036001036001036,
"grad_norm": 0.75390625,
"learning_rate": 4.7413234913234916e-05,
"loss": 0.2847,
"step": 800
},
{
"epoch": 0.11007511007511007,
"grad_norm": 0.9140625,
"learning_rate": 4.725135975135975e-05,
"loss": 0.2793,
"step": 850
},
{
"epoch": 0.11655011655011654,
"grad_norm": 1.0,
"learning_rate": 4.708948458948459e-05,
"loss": 0.2858,
"step": 900
},
{
"epoch": 0.12302512302512303,
"grad_norm": 0.8984375,
"learning_rate": 4.692760942760943e-05,
"loss": 0.2831,
"step": 950
},
{
"epoch": 0.1295001295001295,
"grad_norm": 0.77734375,
"learning_rate": 4.676573426573427e-05,
"loss": 0.2789,
"step": 1000
},
{
"epoch": 0.13597513597513597,
"grad_norm": 0.74609375,
"learning_rate": 4.6603859103859105e-05,
"loss": 0.2719,
"step": 1050
},
{
"epoch": 0.14245014245014245,
"grad_norm": 0.76953125,
"learning_rate": 4.6441983941983945e-05,
"loss": 0.2765,
"step": 1100
},
{
"epoch": 0.14892514892514894,
"grad_norm": 0.8203125,
"learning_rate": 4.6280108780108785e-05,
"loss": 0.2706,
"step": 1150
},
{
"epoch": 0.1554001554001554,
"grad_norm": 0.8046875,
"learning_rate": 4.611823361823362e-05,
"loss": 0.2743,
"step": 1200
},
{
"epoch": 0.16187516187516188,
"grad_norm": 0.80859375,
"learning_rate": 4.595635845635845e-05,
"loss": 0.2737,
"step": 1250
},
{
"epoch": 0.16835016835016836,
"grad_norm": 0.73828125,
"learning_rate": 4.57944832944833e-05,
"loss": 0.2713,
"step": 1300
},
{
"epoch": 0.17482517482517482,
"grad_norm": 0.75,
"learning_rate": 4.5632608132608133e-05,
"loss": 0.271,
"step": 1350
},
{
"epoch": 0.1813001813001813,
"grad_norm": 0.7578125,
"learning_rate": 4.5470732970732974e-05,
"loss": 0.2722,
"step": 1400
},
{
"epoch": 0.18777518777518779,
"grad_norm": 0.73046875,
"learning_rate": 4.5308857808857814e-05,
"loss": 0.2702,
"step": 1450
},
{
"epoch": 0.19425019425019424,
"grad_norm": 0.80078125,
"learning_rate": 4.514698264698265e-05,
"loss": 0.2682,
"step": 1500
},
{
"epoch": 0.20072520072520073,
"grad_norm": 0.75390625,
"learning_rate": 4.498510748510749e-05,
"loss": 0.2676,
"step": 1550
},
{
"epoch": 0.2072002072002072,
"grad_norm": 0.70703125,
"learning_rate": 4.482323232323233e-05,
"loss": 0.2723,
"step": 1600
},
{
"epoch": 0.21367521367521367,
"grad_norm": 0.66796875,
"learning_rate": 4.466135716135716e-05,
"loss": 0.2644,
"step": 1650
},
{
"epoch": 0.22015022015022015,
"grad_norm": 0.7109375,
"learning_rate": 4.4499481999482e-05,
"loss": 0.2614,
"step": 1700
},
{
"epoch": 0.22662522662522663,
"grad_norm": 0.7421875,
"learning_rate": 4.4337606837606836e-05,
"loss": 0.2678,
"step": 1750
},
{
"epoch": 0.2331002331002331,
"grad_norm": 0.71875,
"learning_rate": 4.4175731675731677e-05,
"loss": 0.2697,
"step": 1800
},
{
"epoch": 0.23957523957523957,
"grad_norm": 0.65234375,
"learning_rate": 4.401385651385652e-05,
"loss": 0.264,
"step": 1850
},
{
"epoch": 0.24605024605024606,
"grad_norm": 0.75390625,
"learning_rate": 4.385198135198135e-05,
"loss": 0.2612,
"step": 1900
},
{
"epoch": 0.25252525252525254,
"grad_norm": 0.70703125,
"learning_rate": 4.36901061901062e-05,
"loss": 0.2661,
"step": 1950
},
{
"epoch": 0.259000259000259,
"grad_norm": 0.765625,
"learning_rate": 4.352823102823103e-05,
"loss": 0.2685,
"step": 2000
},
{
"epoch": 0.26547526547526545,
"grad_norm": 0.8203125,
"learning_rate": 4.3366355866355865e-05,
"loss": 0.2684,
"step": 2050
},
{
"epoch": 0.27195027195027194,
"grad_norm": 0.76953125,
"learning_rate": 4.3204480704480705e-05,
"loss": 0.2603,
"step": 2100
},
{
"epoch": 0.2784252784252784,
"grad_norm": 0.75390625,
"learning_rate": 4.3042605542605546e-05,
"loss": 0.2554,
"step": 2150
},
{
"epoch": 0.2849002849002849,
"grad_norm": 0.796875,
"learning_rate": 4.2880730380730386e-05,
"loss": 0.2621,
"step": 2200
},
{
"epoch": 0.2913752913752914,
"grad_norm": 0.78125,
"learning_rate": 4.271885521885522e-05,
"loss": 0.2634,
"step": 2250
},
{
"epoch": 0.2978502978502979,
"grad_norm": 0.71484375,
"learning_rate": 4.255698005698006e-05,
"loss": 0.2636,
"step": 2300
},
{
"epoch": 0.3043253043253043,
"grad_norm": 0.80859375,
"learning_rate": 4.23951048951049e-05,
"loss": 0.2654,
"step": 2350
},
{
"epoch": 0.3108003108003108,
"grad_norm": 0.7578125,
"learning_rate": 4.2233229733229734e-05,
"loss": 0.258,
"step": 2400
},
{
"epoch": 0.31727531727531727,
"grad_norm": 0.73046875,
"learning_rate": 4.207135457135457e-05,
"loss": 0.2606,
"step": 2450
},
{
"epoch": 0.32375032375032375,
"grad_norm": 0.82421875,
"learning_rate": 4.1909479409479415e-05,
"loss": 0.2624,
"step": 2500
},
{
"epoch": 0.33022533022533024,
"grad_norm": 0.703125,
"learning_rate": 4.174760424760425e-05,
"loss": 0.2557,
"step": 2550
},
{
"epoch": 0.3367003367003367,
"grad_norm": 0.7421875,
"learning_rate": 4.158572908572909e-05,
"loss": 0.2563,
"step": 2600
},
{
"epoch": 0.34317534317534315,
"grad_norm": 0.69140625,
"learning_rate": 4.142385392385393e-05,
"loss": 0.252,
"step": 2650
},
{
"epoch": 0.34965034965034963,
"grad_norm": 0.7734375,
"learning_rate": 4.126197876197876e-05,
"loss": 0.257,
"step": 2700
},
{
"epoch": 0.3561253561253561,
"grad_norm": 0.796875,
"learning_rate": 4.11001036001036e-05,
"loss": 0.2571,
"step": 2750
},
{
"epoch": 0.3626003626003626,
"grad_norm": 0.82421875,
"learning_rate": 4.093822843822844e-05,
"loss": 0.2583,
"step": 2800
},
{
"epoch": 0.3690753690753691,
"grad_norm": 0.69140625,
"learning_rate": 4.077635327635328e-05,
"loss": 0.2557,
"step": 2850
},
{
"epoch": 0.37555037555037557,
"grad_norm": 0.80859375,
"learning_rate": 4.061447811447812e-05,
"loss": 0.2486,
"step": 2900
},
{
"epoch": 0.382025382025382,
"grad_norm": 0.71484375,
"learning_rate": 4.045260295260295e-05,
"loss": 0.2514,
"step": 2950
},
{
"epoch": 0.3885003885003885,
"grad_norm": 0.8046875,
"learning_rate": 4.029072779072779e-05,
"loss": 0.2548,
"step": 3000
},
{
"epoch": 0.39497539497539497,
"grad_norm": 0.796875,
"learning_rate": 4.012885262885263e-05,
"loss": 0.2548,
"step": 3050
},
{
"epoch": 0.40145040145040145,
"grad_norm": 0.71875,
"learning_rate": 3.9966977466977466e-05,
"loss": 0.2565,
"step": 3100
},
{
"epoch": 0.40792540792540793,
"grad_norm": 0.8046875,
"learning_rate": 3.980510230510231e-05,
"loss": 0.2546,
"step": 3150
},
{
"epoch": 0.4144004144004144,
"grad_norm": 0.8515625,
"learning_rate": 3.9643227143227146e-05,
"loss": 0.2565,
"step": 3200
},
{
"epoch": 0.4208754208754209,
"grad_norm": 0.7265625,
"learning_rate": 3.948135198135198e-05,
"loss": 0.2539,
"step": 3250
},
{
"epoch": 0.42735042735042733,
"grad_norm": 0.7265625,
"learning_rate": 3.931947681947682e-05,
"loss": 0.2457,
"step": 3300
},
{
"epoch": 0.4338254338254338,
"grad_norm": 0.71875,
"learning_rate": 3.915760165760166e-05,
"loss": 0.2507,
"step": 3350
},
{
"epoch": 0.4403004403004403,
"grad_norm": 0.69140625,
"learning_rate": 3.89957264957265e-05,
"loss": 0.2534,
"step": 3400
},
{
"epoch": 0.4467754467754468,
"grad_norm": 0.8046875,
"learning_rate": 3.8833851333851335e-05,
"loss": 0.2546,
"step": 3450
},
{
"epoch": 0.45325045325045327,
"grad_norm": 0.734375,
"learning_rate": 3.8671976171976175e-05,
"loss": 0.2512,
"step": 3500
},
{
"epoch": 0.45972545972545975,
"grad_norm": 0.7109375,
"learning_rate": 3.8510101010101015e-05,
"loss": 0.2514,
"step": 3550
},
{
"epoch": 0.4662004662004662,
"grad_norm": 0.8046875,
"learning_rate": 3.834822584822585e-05,
"loss": 0.2488,
"step": 3600
},
{
"epoch": 0.47267547267547266,
"grad_norm": 0.80078125,
"learning_rate": 3.818635068635068e-05,
"loss": 0.2541,
"step": 3650
},
{
"epoch": 0.47915047915047915,
"grad_norm": 0.68359375,
"learning_rate": 3.802447552447553e-05,
"loss": 0.2567,
"step": 3700
},
{
"epoch": 0.48562548562548563,
"grad_norm": 0.7265625,
"learning_rate": 3.7862600362600363e-05,
"loss": 0.2525,
"step": 3750
},
{
"epoch": 0.4921004921004921,
"grad_norm": 0.8203125,
"learning_rate": 3.7700725200725204e-05,
"loss": 0.2499,
"step": 3800
},
{
"epoch": 0.4985754985754986,
"grad_norm": 0.74609375,
"learning_rate": 3.7538850038850044e-05,
"loss": 0.2557,
"step": 3850
},
{
"epoch": 0.5050505050505051,
"grad_norm": 0.79296875,
"learning_rate": 3.737697487697488e-05,
"loss": 0.2568,
"step": 3900
},
{
"epoch": 0.5115255115255115,
"grad_norm": 0.8125,
"learning_rate": 3.721509971509972e-05,
"loss": 0.2543,
"step": 3950
},
{
"epoch": 0.518000518000518,
"grad_norm": 0.765625,
"learning_rate": 3.705322455322455e-05,
"loss": 0.2546,
"step": 4000
},
{
"epoch": 0.5244755244755245,
"grad_norm": 0.84765625,
"learning_rate": 3.689134939134939e-05,
"loss": 0.258,
"step": 4050
},
{
"epoch": 0.5309505309505309,
"grad_norm": 0.7734375,
"learning_rate": 3.672947422947423e-05,
"loss": 0.2519,
"step": 4100
},
{
"epoch": 0.5374255374255374,
"grad_norm": 0.74609375,
"learning_rate": 3.6567599067599066e-05,
"loss": 0.2507,
"step": 4150
},
{
"epoch": 0.5439005439005439,
"grad_norm": 0.6640625,
"learning_rate": 3.6405723905723907e-05,
"loss": 0.2526,
"step": 4200
},
{
"epoch": 0.5503755503755504,
"grad_norm": 0.640625,
"learning_rate": 3.624384874384875e-05,
"loss": 0.2556,
"step": 4250
},
{
"epoch": 0.5568505568505568,
"grad_norm": 0.73828125,
"learning_rate": 3.608197358197358e-05,
"loss": 0.2516,
"step": 4300
},
{
"epoch": 0.5633255633255633,
"grad_norm": 0.78515625,
"learning_rate": 3.592009842009842e-05,
"loss": 0.2476,
"step": 4350
},
{
"epoch": 0.5698005698005698,
"grad_norm": 0.80078125,
"learning_rate": 3.575822325822326e-05,
"loss": 0.2449,
"step": 4400
},
{
"epoch": 0.5762755762755762,
"grad_norm": 0.67578125,
"learning_rate": 3.5596348096348095e-05,
"loss": 0.251,
"step": 4450
},
{
"epoch": 0.5827505827505828,
"grad_norm": 0.80859375,
"learning_rate": 3.5434472934472935e-05,
"loss": 0.2463,
"step": 4500
},
{
"epoch": 0.5892255892255892,
"grad_norm": 0.8359375,
"learning_rate": 3.5272597772597776e-05,
"loss": 0.2515,
"step": 4550
},
{
"epoch": 0.5957005957005957,
"grad_norm": 0.8359375,
"learning_rate": 3.5110722610722616e-05,
"loss": 0.249,
"step": 4600
},
{
"epoch": 0.6021756021756022,
"grad_norm": 0.85546875,
"learning_rate": 3.494884744884745e-05,
"loss": 0.2486,
"step": 4650
},
{
"epoch": 0.6086506086506086,
"grad_norm": 0.859375,
"learning_rate": 3.478697228697229e-05,
"loss": 0.2497,
"step": 4700
},
{
"epoch": 0.6151256151256151,
"grad_norm": 0.8828125,
"learning_rate": 3.462509712509713e-05,
"loss": 0.2469,
"step": 4750
},
{
"epoch": 0.6216006216006216,
"grad_norm": 0.7578125,
"learning_rate": 3.4463221963221964e-05,
"loss": 0.2538,
"step": 4800
},
{
"epoch": 0.6280756280756281,
"grad_norm": 0.78515625,
"learning_rate": 3.43013468013468e-05,
"loss": 0.2455,
"step": 4850
},
{
"epoch": 0.6345506345506345,
"grad_norm": 0.74609375,
"learning_rate": 3.4139471639471645e-05,
"loss": 0.2412,
"step": 4900
},
{
"epoch": 0.6410256410256411,
"grad_norm": 0.79296875,
"learning_rate": 3.397759647759648e-05,
"loss": 0.2451,
"step": 4950
},
{
"epoch": 0.6475006475006475,
"grad_norm": 0.6640625,
"learning_rate": 3.381572131572132e-05,
"loss": 0.2487,
"step": 5000
},
{
"epoch": 0.6539756539756539,
"grad_norm": 0.8203125,
"learning_rate": 3.365384615384616e-05,
"loss": 0.2457,
"step": 5050
},
{
"epoch": 0.6604506604506605,
"grad_norm": 0.8203125,
"learning_rate": 3.349197099197099e-05,
"loss": 0.2468,
"step": 5100
},
{
"epoch": 0.6669256669256669,
"grad_norm": 0.7578125,
"learning_rate": 3.333009583009583e-05,
"loss": 0.2477,
"step": 5150
},
{
"epoch": 0.6734006734006734,
"grad_norm": 0.7578125,
"learning_rate": 3.316822066822067e-05,
"loss": 0.247,
"step": 5200
},
{
"epoch": 0.6798756798756799,
"grad_norm": 0.76171875,
"learning_rate": 3.300634550634551e-05,
"loss": 0.241,
"step": 5250
},
{
"epoch": 0.6863506863506863,
"grad_norm": 0.7265625,
"learning_rate": 3.284447034447035e-05,
"loss": 0.2451,
"step": 5300
},
{
"epoch": 0.6928256928256928,
"grad_norm": 0.80859375,
"learning_rate": 3.268259518259518e-05,
"loss": 0.2431,
"step": 5350
},
{
"epoch": 0.6993006993006993,
"grad_norm": 0.76171875,
"learning_rate": 3.252072002072002e-05,
"loss": 0.2403,
"step": 5400
},
{
"epoch": 0.7057757057757058,
"grad_norm": 0.83203125,
"learning_rate": 3.235884485884486e-05,
"loss": 0.2511,
"step": 5450
},
{
"epoch": 0.7122507122507122,
"grad_norm": 0.66796875,
"learning_rate": 3.2196969696969696e-05,
"loss": 0.2412,
"step": 5500
},
{
"epoch": 0.7187257187257188,
"grad_norm": 0.78515625,
"learning_rate": 3.2035094535094536e-05,
"loss": 0.2453,
"step": 5550
},
{
"epoch": 0.7252007252007252,
"grad_norm": 0.76953125,
"learning_rate": 3.1873219373219376e-05,
"loss": 0.2473,
"step": 5600
},
{
"epoch": 0.7316757316757316,
"grad_norm": 0.703125,
"learning_rate": 3.171134421134421e-05,
"loss": 0.2455,
"step": 5650
},
{
"epoch": 0.7381507381507382,
"grad_norm": 0.94921875,
"learning_rate": 3.154946904946905e-05,
"loss": 0.2485,
"step": 5700
},
{
"epoch": 0.7446257446257446,
"grad_norm": 0.7734375,
"learning_rate": 3.138759388759389e-05,
"loss": 0.2469,
"step": 5750
},
{
"epoch": 0.7511007511007511,
"grad_norm": 0.76171875,
"learning_rate": 3.122571872571873e-05,
"loss": 0.2452,
"step": 5800
},
{
"epoch": 0.7575757575757576,
"grad_norm": 0.77734375,
"learning_rate": 3.1063843563843565e-05,
"loss": 0.2466,
"step": 5850
},
{
"epoch": 0.764050764050764,
"grad_norm": 0.78515625,
"learning_rate": 3.09019684019684e-05,
"loss": 0.2458,
"step": 5900
},
{
"epoch": 0.7705257705257705,
"grad_norm": 0.734375,
"learning_rate": 3.0740093240093245e-05,
"loss": 0.2429,
"step": 5950
},
{
"epoch": 0.777000777000777,
"grad_norm": 0.7578125,
"learning_rate": 3.057821807821808e-05,
"loss": 0.2476,
"step": 6000
},
{
"epoch": 0.7834757834757835,
"grad_norm": 0.7734375,
"learning_rate": 3.0416342916342916e-05,
"loss": 0.2442,
"step": 6050
},
{
"epoch": 0.7899507899507899,
"grad_norm": 0.75390625,
"learning_rate": 3.0254467754467756e-05,
"loss": 0.2423,
"step": 6100
},
{
"epoch": 0.7964257964257965,
"grad_norm": 0.73828125,
"learning_rate": 3.0092592592592593e-05,
"loss": 0.2443,
"step": 6150
},
{
"epoch": 0.8029008029008029,
"grad_norm": 0.734375,
"learning_rate": 2.993071743071743e-05,
"loss": 0.2415,
"step": 6200
},
{
"epoch": 0.8093758093758093,
"grad_norm": 0.734375,
"learning_rate": 2.9768842268842274e-05,
"loss": 0.2411,
"step": 6250
},
{
"epoch": 0.8158508158508159,
"grad_norm": 0.765625,
"learning_rate": 2.9606967106967108e-05,
"loss": 0.2429,
"step": 6300
},
{
"epoch": 0.8223258223258223,
"grad_norm": 0.7890625,
"learning_rate": 2.9445091945091945e-05,
"loss": 0.2425,
"step": 6350
},
{
"epoch": 0.8288008288008288,
"grad_norm": 0.7578125,
"learning_rate": 2.9283216783216782e-05,
"loss": 0.2456,
"step": 6400
},
{
"epoch": 0.8352758352758353,
"grad_norm": 0.73046875,
"learning_rate": 2.9121341621341626e-05,
"loss": 0.2404,
"step": 6450
},
{
"epoch": 0.8417508417508418,
"grad_norm": 0.8125,
"learning_rate": 2.8959466459466463e-05,
"loss": 0.2412,
"step": 6500
},
{
"epoch": 0.8482258482258482,
"grad_norm": 0.8359375,
"learning_rate": 2.8797591297591296e-05,
"loss": 0.2402,
"step": 6550
},
{
"epoch": 0.8547008547008547,
"grad_norm": 0.703125,
"learning_rate": 2.863571613571614e-05,
"loss": 0.2366,
"step": 6600
},
{
"epoch": 0.8611758611758612,
"grad_norm": 0.85546875,
"learning_rate": 2.8473840973840977e-05,
"loss": 0.2428,
"step": 6650
},
{
"epoch": 0.8676508676508676,
"grad_norm": 0.6875,
"learning_rate": 2.8311965811965814e-05,
"loss": 0.238,
"step": 6700
},
{
"epoch": 0.8741258741258742,
"grad_norm": 0.75,
"learning_rate": 2.8150090650090648e-05,
"loss": 0.2381,
"step": 6750
},
{
"epoch": 0.8806008806008806,
"grad_norm": 0.65625,
"learning_rate": 2.798821548821549e-05,
"loss": 0.2468,
"step": 6800
},
{
"epoch": 0.887075887075887,
"grad_norm": 0.75390625,
"learning_rate": 2.782634032634033e-05,
"loss": 0.2419,
"step": 6850
},
{
"epoch": 0.8935508935508936,
"grad_norm": 0.70703125,
"learning_rate": 2.7664465164465165e-05,
"loss": 0.2445,
"step": 6900
},
{
"epoch": 0.9000259000259,
"grad_norm": 0.80859375,
"learning_rate": 2.7502590002590006e-05,
"loss": 0.2457,
"step": 6950
},
{
"epoch": 0.9065009065009065,
"grad_norm": 0.76953125,
"learning_rate": 2.7340714840714843e-05,
"loss": 0.2376,
"step": 7000
},
{
"epoch": 0.912975912975913,
"grad_norm": 0.71484375,
"learning_rate": 2.717883967883968e-05,
"loss": 0.2447,
"step": 7050
},
{
"epoch": 0.9194509194509195,
"grad_norm": 0.8046875,
"learning_rate": 2.7016964516964517e-05,
"loss": 0.2364,
"step": 7100
},
{
"epoch": 0.9259259259259259,
"grad_norm": 0.73828125,
"learning_rate": 2.6855089355089357e-05,
"loss": 0.2372,
"step": 7150
},
{
"epoch": 0.9324009324009324,
"grad_norm": 0.734375,
"learning_rate": 2.6693214193214194e-05,
"loss": 0.2367,
"step": 7200
},
{
"epoch": 0.9388759388759389,
"grad_norm": 0.79296875,
"learning_rate": 2.653133903133903e-05,
"loss": 0.2438,
"step": 7250
},
{
"epoch": 0.9453509453509453,
"grad_norm": 0.8984375,
"learning_rate": 2.636946386946387e-05,
"loss": 0.2383,
"step": 7300
},
{
"epoch": 0.9518259518259519,
"grad_norm": 0.859375,
"learning_rate": 2.620758870758871e-05,
"loss": 0.239,
"step": 7350
},
{
"epoch": 0.9583009583009583,
"grad_norm": 0.84375,
"learning_rate": 2.6045713545713545e-05,
"loss": 0.2353,
"step": 7400
},
{
"epoch": 0.9647759647759647,
"grad_norm": 0.8046875,
"learning_rate": 2.5883838383838382e-05,
"loss": 0.2405,
"step": 7450
},
{
"epoch": 0.9712509712509713,
"grad_norm": 0.82421875,
"learning_rate": 2.5721963221963223e-05,
"loss": 0.241,
"step": 7500
},
{
"epoch": 0.9777259777259777,
"grad_norm": 0.75390625,
"learning_rate": 2.556008806008806e-05,
"loss": 0.2367,
"step": 7550
},
{
"epoch": 0.9842009842009842,
"grad_norm": 0.81640625,
"learning_rate": 2.5398212898212897e-05,
"loss": 0.2418,
"step": 7600
},
{
"epoch": 0.9906759906759907,
"grad_norm": 0.7265625,
"learning_rate": 2.523633773633774e-05,
"loss": 0.2425,
"step": 7650
},
{
"epoch": 0.9971509971509972,
"grad_norm": 0.79296875,
"learning_rate": 2.5074462574462578e-05,
"loss": 0.2385,
"step": 7700
},
{
"epoch": 1.0036260036260036,
"grad_norm": 0.8203125,
"learning_rate": 2.491258741258741e-05,
"loss": 0.2282,
"step": 7750
},
{
"epoch": 1.0101010101010102,
"grad_norm": 0.734375,
"learning_rate": 2.475071225071225e-05,
"loss": 0.2236,
"step": 7800
},
{
"epoch": 1.0165760165760165,
"grad_norm": 0.83984375,
"learning_rate": 2.4588837088837092e-05,
"loss": 0.2202,
"step": 7850
},
{
"epoch": 1.023051023051023,
"grad_norm": 0.8125,
"learning_rate": 2.442696192696193e-05,
"loss": 0.2222,
"step": 7900
},
{
"epoch": 1.0295260295260296,
"grad_norm": 0.78125,
"learning_rate": 2.4265086765086766e-05,
"loss": 0.2213,
"step": 7950
},
{
"epoch": 1.036001036001036,
"grad_norm": 0.7265625,
"learning_rate": 2.4103211603211603e-05,
"loss": 0.2283,
"step": 8000
},
{
"epoch": 1.0424760424760424,
"grad_norm": 0.8203125,
"learning_rate": 2.3941336441336443e-05,
"loss": 0.2232,
"step": 8050
},
{
"epoch": 1.048951048951049,
"grad_norm": 0.81640625,
"learning_rate": 2.377946127946128e-05,
"loss": 0.2287,
"step": 8100
},
{
"epoch": 1.0554260554260555,
"grad_norm": 0.78515625,
"learning_rate": 2.3617586117586117e-05,
"loss": 0.2265,
"step": 8150
},
{
"epoch": 1.0619010619010618,
"grad_norm": 0.828125,
"learning_rate": 2.3455710955710958e-05,
"loss": 0.2258,
"step": 8200
},
{
"epoch": 1.0683760683760684,
"grad_norm": 0.8203125,
"learning_rate": 2.3293835793835795e-05,
"loss": 0.2273,
"step": 8250
},
{
"epoch": 1.074851074851075,
"grad_norm": 0.8125,
"learning_rate": 2.3131960631960635e-05,
"loss": 0.225,
"step": 8300
},
{
"epoch": 1.0813260813260812,
"grad_norm": 0.74609375,
"learning_rate": 2.297008547008547e-05,
"loss": 0.2213,
"step": 8350
},
{
"epoch": 1.0878010878010878,
"grad_norm": 0.796875,
"learning_rate": 2.280821030821031e-05,
"loss": 0.2229,
"step": 8400
},
{
"epoch": 1.0942760942760943,
"grad_norm": 0.96484375,
"learning_rate": 2.264633514633515e-05,
"loss": 0.2261,
"step": 8450
},
{
"epoch": 1.1007511007511008,
"grad_norm": 0.86328125,
"learning_rate": 2.2484459984459986e-05,
"loss": 0.2246,
"step": 8500
},
{
"epoch": 1.1072261072261071,
"grad_norm": 0.7734375,
"learning_rate": 2.2322584822584823e-05,
"loss": 0.2216,
"step": 8550
},
{
"epoch": 1.1137011137011137,
"grad_norm": 0.828125,
"learning_rate": 2.216070966070966e-05,
"loss": 0.2276,
"step": 8600
},
{
"epoch": 1.1201761201761202,
"grad_norm": 0.9765625,
"learning_rate": 2.19988344988345e-05,
"loss": 0.226,
"step": 8650
},
{
"epoch": 1.1266511266511268,
"grad_norm": 0.8515625,
"learning_rate": 2.1836959336959338e-05,
"loss": 0.2275,
"step": 8700
},
{
"epoch": 1.133126133126133,
"grad_norm": 0.82421875,
"learning_rate": 2.1675084175084175e-05,
"loss": 0.2203,
"step": 8750
},
{
"epoch": 1.1396011396011396,
"grad_norm": 0.86328125,
"learning_rate": 2.1513209013209015e-05,
"loss": 0.2193,
"step": 8800
},
{
"epoch": 1.1460761460761462,
"grad_norm": 0.78515625,
"learning_rate": 2.1351333851333852e-05,
"loss": 0.2255,
"step": 8850
},
{
"epoch": 1.1525511525511525,
"grad_norm": 0.80078125,
"learning_rate": 2.1189458689458693e-05,
"loss": 0.2191,
"step": 8900
},
{
"epoch": 1.159026159026159,
"grad_norm": 0.8125,
"learning_rate": 2.1027583527583526e-05,
"loss": 0.2238,
"step": 8950
},
{
"epoch": 1.1655011655011656,
"grad_norm": 0.90625,
"learning_rate": 2.0865708365708367e-05,
"loss": 0.2201,
"step": 9000
},
{
"epoch": 1.1719761719761719,
"grad_norm": 0.78515625,
"learning_rate": 2.0703833203833207e-05,
"loss": 0.2249,
"step": 9050
},
{
"epoch": 1.1784511784511784,
"grad_norm": 0.87890625,
"learning_rate": 2.0541958041958044e-05,
"loss": 0.2226,
"step": 9100
},
{
"epoch": 1.184926184926185,
"grad_norm": 0.90234375,
"learning_rate": 2.038008288008288e-05,
"loss": 0.2211,
"step": 9150
},
{
"epoch": 1.1914011914011915,
"grad_norm": 0.80859375,
"learning_rate": 2.0218207718207718e-05,
"loss": 0.2233,
"step": 9200
},
{
"epoch": 1.1978761978761978,
"grad_norm": 0.84765625,
"learning_rate": 2.005633255633256e-05,
"loss": 0.2246,
"step": 9250
},
{
"epoch": 1.2043512043512044,
"grad_norm": 0.890625,
"learning_rate": 1.9894457394457395e-05,
"loss": 0.2266,
"step": 9300
},
{
"epoch": 1.210826210826211,
"grad_norm": 0.79296875,
"learning_rate": 1.9732582232582232e-05,
"loss": 0.2194,
"step": 9350
},
{
"epoch": 1.2173012173012172,
"grad_norm": 1.0234375,
"learning_rate": 1.9570707070707073e-05,
"loss": 0.2216,
"step": 9400
},
{
"epoch": 1.2237762237762237,
"grad_norm": 0.76953125,
"learning_rate": 1.940883190883191e-05,
"loss": 0.2269,
"step": 9450
},
{
"epoch": 1.2302512302512303,
"grad_norm": 0.91796875,
"learning_rate": 1.924695674695675e-05,
"loss": 0.2245,
"step": 9500
},
{
"epoch": 1.2367262367262368,
"grad_norm": 0.8125,
"learning_rate": 1.9085081585081584e-05,
"loss": 0.2257,
"step": 9550
},
{
"epoch": 1.2432012432012431,
"grad_norm": 0.78125,
"learning_rate": 1.8923206423206424e-05,
"loss": 0.2204,
"step": 9600
},
{
"epoch": 1.2496762496762497,
"grad_norm": 0.890625,
"learning_rate": 1.876133126133126e-05,
"loss": 0.2257,
"step": 9650
},
{
"epoch": 1.2561512561512562,
"grad_norm": 0.859375,
"learning_rate": 1.85994560994561e-05,
"loss": 0.2238,
"step": 9700
},
{
"epoch": 1.2626262626262625,
"grad_norm": 0.76171875,
"learning_rate": 1.843758093758094e-05,
"loss": 0.2218,
"step": 9750
},
{
"epoch": 1.269101269101269,
"grad_norm": 0.8984375,
"learning_rate": 1.8275705775705775e-05,
"loss": 0.2234,
"step": 9800
},
{
"epoch": 1.2755762755762756,
"grad_norm": 0.7578125,
"learning_rate": 1.8113830613830616e-05,
"loss": 0.2253,
"step": 9850
},
{
"epoch": 1.282051282051282,
"grad_norm": 0.85546875,
"learning_rate": 1.7951955451955453e-05,
"loss": 0.2197,
"step": 9900
},
{
"epoch": 1.2885262885262885,
"grad_norm": 0.8671875,
"learning_rate": 1.779008029008029e-05,
"loss": 0.2237,
"step": 9950
},
{
"epoch": 1.295001295001295,
"grad_norm": 0.82421875,
"learning_rate": 1.762820512820513e-05,
"loss": 0.2199,
"step": 10000
},
{
"epoch": 1.3014763014763016,
"grad_norm": 0.78515625,
"learning_rate": 1.7466329966329967e-05,
"loss": 0.2221,
"step": 10050
},
{
"epoch": 1.3079513079513079,
"grad_norm": 0.81640625,
"learning_rate": 1.7304454804454808e-05,
"loss": 0.2222,
"step": 10100
},
{
"epoch": 1.3144263144263144,
"grad_norm": 0.84375,
"learning_rate": 1.714257964257964e-05,
"loss": 0.2224,
"step": 10150
},
{
"epoch": 1.320901320901321,
"grad_norm": 0.77734375,
"learning_rate": 1.698070448070448e-05,
"loss": 0.2238,
"step": 10200
},
{
"epoch": 1.3273763273763275,
"grad_norm": 0.89453125,
"learning_rate": 1.681882931882932e-05,
"loss": 0.2262,
"step": 10250
},
{
"epoch": 1.3338513338513338,
"grad_norm": 0.8359375,
"learning_rate": 1.665695415695416e-05,
"loss": 0.2232,
"step": 10300
},
{
"epoch": 1.3403263403263403,
"grad_norm": 0.7890625,
"learning_rate": 1.6495078995078996e-05,
"loss": 0.2252,
"step": 10350
},
{
"epoch": 1.3468013468013469,
"grad_norm": 0.8671875,
"learning_rate": 1.6333203833203833e-05,
"loss": 0.2177,
"step": 10400
},
{
"epoch": 1.3532763532763532,
"grad_norm": 0.79296875,
"learning_rate": 1.6171328671328673e-05,
"loss": 0.2188,
"step": 10450
},
{
"epoch": 1.3597513597513597,
"grad_norm": 0.86328125,
"learning_rate": 1.600945350945351e-05,
"loss": 0.2233,
"step": 10500
},
{
"epoch": 1.3662263662263663,
"grad_norm": 0.7890625,
"learning_rate": 1.5847578347578347e-05,
"loss": 0.2225,
"step": 10550
},
{
"epoch": 1.3727013727013726,
"grad_norm": 0.85546875,
"learning_rate": 1.5685703185703188e-05,
"loss": 0.2193,
"step": 10600
},
{
"epoch": 1.3791763791763791,
"grad_norm": 0.8359375,
"learning_rate": 1.5523828023828025e-05,
"loss": 0.2185,
"step": 10650
},
{
"epoch": 1.3856513856513857,
"grad_norm": 0.8125,
"learning_rate": 1.5361952861952865e-05,
"loss": 0.2238,
"step": 10700
},
{
"epoch": 1.392126392126392,
"grad_norm": 0.80859375,
"learning_rate": 1.52000777000777e-05,
"loss": 0.2209,
"step": 10750
},
{
"epoch": 1.3986013986013985,
"grad_norm": 0.91796875,
"learning_rate": 1.5038202538202539e-05,
"loss": 0.2238,
"step": 10800
},
{
"epoch": 1.405076405076405,
"grad_norm": 0.83203125,
"learning_rate": 1.4876327376327376e-05,
"loss": 0.2238,
"step": 10850
},
{
"epoch": 1.4115514115514116,
"grad_norm": 0.75390625,
"learning_rate": 1.4714452214452215e-05,
"loss": 0.2189,
"step": 10900
},
{
"epoch": 1.4180264180264182,
"grad_norm": 0.76171875,
"learning_rate": 1.4552577052577055e-05,
"loss": 0.2196,
"step": 10950
},
{
"epoch": 1.4245014245014245,
"grad_norm": 0.78515625,
"learning_rate": 1.439070189070189e-05,
"loss": 0.2237,
"step": 11000
},
{
"epoch": 1.430976430976431,
"grad_norm": 0.8359375,
"learning_rate": 1.4228826728826731e-05,
"loss": 0.2253,
"step": 11050
},
{
"epoch": 1.4374514374514376,
"grad_norm": 0.85546875,
"learning_rate": 1.4066951566951566e-05,
"loss": 0.2232,
"step": 11100
},
{
"epoch": 1.4439264439264439,
"grad_norm": 0.89453125,
"learning_rate": 1.3905076405076407e-05,
"loss": 0.2191,
"step": 11150
},
{
"epoch": 1.4504014504014504,
"grad_norm": 0.875,
"learning_rate": 1.3743201243201245e-05,
"loss": 0.219,
"step": 11200
},
{
"epoch": 1.456876456876457,
"grad_norm": 0.796875,
"learning_rate": 1.3581326081326082e-05,
"loss": 0.2251,
"step": 11250
},
{
"epoch": 1.4633514633514633,
"grad_norm": 0.8203125,
"learning_rate": 1.3419450919450921e-05,
"loss": 0.2182,
"step": 11300
},
{
"epoch": 1.4698264698264698,
"grad_norm": 0.890625,
"learning_rate": 1.3257575757575758e-05,
"loss": 0.226,
"step": 11350
},
{
"epoch": 1.4763014763014763,
"grad_norm": 0.83984375,
"learning_rate": 1.3095700595700597e-05,
"loss": 0.2171,
"step": 11400
},
{
"epoch": 1.4827764827764827,
"grad_norm": 0.875,
"learning_rate": 1.2933825433825434e-05,
"loss": 0.2199,
"step": 11450
},
{
"epoch": 1.4892514892514892,
"grad_norm": 0.7890625,
"learning_rate": 1.2771950271950272e-05,
"loss": 0.2198,
"step": 11500
},
{
"epoch": 1.4957264957264957,
"grad_norm": 0.79296875,
"learning_rate": 1.2610075110075113e-05,
"loss": 0.2196,
"step": 11550
},
{
"epoch": 1.502201502201502,
"grad_norm": 0.9921875,
"learning_rate": 1.2448199948199948e-05,
"loss": 0.2231,
"step": 11600
},
{
"epoch": 1.5086765086765088,
"grad_norm": 0.95703125,
"learning_rate": 1.2286324786324787e-05,
"loss": 0.2187,
"step": 11650
},
{
"epoch": 1.5151515151515151,
"grad_norm": 0.83203125,
"learning_rate": 1.2124449624449625e-05,
"loss": 0.2216,
"step": 11700
},
{
"epoch": 1.5216265216265217,
"grad_norm": 0.79296875,
"learning_rate": 1.1962574462574464e-05,
"loss": 0.2198,
"step": 11750
},
{
"epoch": 1.5281015281015282,
"grad_norm": 0.88671875,
"learning_rate": 1.1800699300699301e-05,
"loss": 0.2218,
"step": 11800
},
{
"epoch": 1.5345765345765345,
"grad_norm": 0.84375,
"learning_rate": 1.163882413882414e-05,
"loss": 0.2196,
"step": 11850
},
{
"epoch": 1.541051541051541,
"grad_norm": 0.91015625,
"learning_rate": 1.1476948976948977e-05,
"loss": 0.2204,
"step": 11900
},
{
"epoch": 1.5475265475265476,
"grad_norm": 0.92578125,
"learning_rate": 1.1315073815073815e-05,
"loss": 0.2264,
"step": 11950
},
{
"epoch": 1.554001554001554,
"grad_norm": 0.7890625,
"learning_rate": 1.1153198653198654e-05,
"loss": 0.2282,
"step": 12000
},
{
"epoch": 1.5604765604765605,
"grad_norm": 0.88671875,
"learning_rate": 1.0991323491323493e-05,
"loss": 0.2203,
"step": 12050
},
{
"epoch": 1.566951566951567,
"grad_norm": 0.81640625,
"learning_rate": 1.082944832944833e-05,
"loss": 0.2225,
"step": 12100
},
{
"epoch": 1.5734265734265733,
"grad_norm": 0.77734375,
"learning_rate": 1.0667573167573168e-05,
"loss": 0.2233,
"step": 12150
},
{
"epoch": 1.5799015799015799,
"grad_norm": 0.93359375,
"learning_rate": 1.0505698005698005e-05,
"loss": 0.2243,
"step": 12200
},
{
"epoch": 1.5863765863765864,
"grad_norm": 0.875,
"learning_rate": 1.0343822843822844e-05,
"loss": 0.2177,
"step": 12250
},
{
"epoch": 1.5928515928515927,
"grad_norm": 0.91015625,
"learning_rate": 1.0181947681947683e-05,
"loss": 0.2228,
"step": 12300
},
{
"epoch": 1.5993265993265995,
"grad_norm": 0.82421875,
"learning_rate": 1.0020072520072522e-05,
"loss": 0.2229,
"step": 12350
},
{
"epoch": 1.6058016058016058,
"grad_norm": 0.93359375,
"learning_rate": 9.858197358197359e-06,
"loss": 0.2233,
"step": 12400
},
{
"epoch": 1.6122766122766121,
"grad_norm": 0.875,
"learning_rate": 9.696322196322197e-06,
"loss": 0.2204,
"step": 12450
},
{
"epoch": 1.6187516187516189,
"grad_norm": 0.796875,
"learning_rate": 9.534447034447034e-06,
"loss": 0.2219,
"step": 12500
},
{
"epoch": 1.6252266252266252,
"grad_norm": 0.98046875,
"learning_rate": 9.372571872571873e-06,
"loss": 0.218,
"step": 12550
},
{
"epoch": 1.6317016317016317,
"grad_norm": 0.86328125,
"learning_rate": 9.210696710696712e-06,
"loss": 0.2258,
"step": 12600
},
{
"epoch": 1.6381766381766383,
"grad_norm": 0.765625,
"learning_rate": 9.04882154882155e-06,
"loss": 0.2244,
"step": 12650
},
{
"epoch": 1.6446516446516446,
"grad_norm": 0.88671875,
"learning_rate": 8.886946386946387e-06,
"loss": 0.2187,
"step": 12700
},
{
"epoch": 1.6511266511266511,
"grad_norm": 0.91796875,
"learning_rate": 8.725071225071226e-06,
"loss": 0.2189,
"step": 12750
},
{
"epoch": 1.6576016576016577,
"grad_norm": 0.890625,
"learning_rate": 8.563196063196063e-06,
"loss": 0.2191,
"step": 12800
},
{
"epoch": 1.664076664076664,
"grad_norm": 0.91796875,
"learning_rate": 8.401320901320902e-06,
"loss": 0.2209,
"step": 12850
},
{
"epoch": 1.6705516705516705,
"grad_norm": 0.81640625,
"learning_rate": 8.239445739445739e-06,
"loss": 0.217,
"step": 12900
},
{
"epoch": 1.677026677026677,
"grad_norm": 0.84375,
"learning_rate": 8.077570577570579e-06,
"loss": 0.2233,
"step": 12950
},
{
"epoch": 1.6835016835016834,
"grad_norm": 0.87890625,
"learning_rate": 7.915695415695416e-06,
"loss": 0.2176,
"step": 13000
},
{
"epoch": 1.68997668997669,
"grad_norm": 0.80859375,
"learning_rate": 7.753820253820255e-06,
"loss": 0.2248,
"step": 13050
},
{
"epoch": 1.6964516964516965,
"grad_norm": 1.0703125,
"learning_rate": 7.591945091945092e-06,
"loss": 0.231,
"step": 13100
},
{
"epoch": 1.7029267029267028,
"grad_norm": 0.8359375,
"learning_rate": 7.43006993006993e-06,
"loss": 0.2181,
"step": 13150
},
{
"epoch": 1.7094017094017095,
"grad_norm": 0.78515625,
"learning_rate": 7.268194768194768e-06,
"loss": 0.2194,
"step": 13200
},
{
"epoch": 1.7158767158767159,
"grad_norm": 0.83984375,
"learning_rate": 7.106319606319607e-06,
"loss": 0.2188,
"step": 13250
},
{
"epoch": 1.7223517223517224,
"grad_norm": 0.890625,
"learning_rate": 6.944444444444445e-06,
"loss": 0.2232,
"step": 13300
},
{
"epoch": 1.728826728826729,
"grad_norm": 0.87109375,
"learning_rate": 6.782569282569283e-06,
"loss": 0.2192,
"step": 13350
},
{
"epoch": 1.7353017353017353,
"grad_norm": 0.87890625,
"learning_rate": 6.6206941206941205e-06,
"loss": 0.2208,
"step": 13400
},
{
"epoch": 1.7417767417767418,
"grad_norm": 0.921875,
"learning_rate": 6.458818958818959e-06,
"loss": 0.2206,
"step": 13450
},
{
"epoch": 1.7482517482517483,
"grad_norm": 0.76171875,
"learning_rate": 6.296943796943797e-06,
"loss": 0.2193,
"step": 13500
},
{
"epoch": 1.7547267547267547,
"grad_norm": 0.9140625,
"learning_rate": 6.135068635068635e-06,
"loss": 0.2212,
"step": 13550
},
{
"epoch": 1.7612017612017612,
"grad_norm": 0.859375,
"learning_rate": 5.9731934731934735e-06,
"loss": 0.222,
"step": 13600
},
{
"epoch": 1.7676767676767677,
"grad_norm": 0.8515625,
"learning_rate": 5.811318311318311e-06,
"loss": 0.2204,
"step": 13650
},
{
"epoch": 1.774151774151774,
"grad_norm": 0.85546875,
"learning_rate": 5.649443149443149e-06,
"loss": 0.2215,
"step": 13700
},
{
"epoch": 1.7806267806267806,
"grad_norm": 0.91015625,
"learning_rate": 5.487567987567988e-06,
"loss": 0.2262,
"step": 13750
},
{
"epoch": 1.7871017871017871,
"grad_norm": 0.8125,
"learning_rate": 5.325692825692826e-06,
"loss": 0.2216,
"step": 13800
},
{
"epoch": 1.7935767935767934,
"grad_norm": 0.984375,
"learning_rate": 5.163817663817664e-06,
"loss": 0.2181,
"step": 13850
},
{
"epoch": 1.8000518000518002,
"grad_norm": 0.75,
"learning_rate": 5.001942501942502e-06,
"loss": 0.2239,
"step": 13900
},
{
"epoch": 1.8065268065268065,
"grad_norm": 0.8046875,
"learning_rate": 4.84006734006734e-06,
"loss": 0.2211,
"step": 13950
},
{
"epoch": 1.8130018130018128,
"grad_norm": 0.88671875,
"learning_rate": 4.678192178192178e-06,
"loss": 0.2249,
"step": 14000
},
{
"epoch": 1.8194768194768196,
"grad_norm": 0.8125,
"learning_rate": 4.516317016317017e-06,
"loss": 0.2218,
"step": 14050
},
{
"epoch": 1.825951825951826,
"grad_norm": 0.765625,
"learning_rate": 4.3544418544418545e-06,
"loss": 0.2206,
"step": 14100
},
{
"epoch": 1.8324268324268325,
"grad_norm": 0.84375,
"learning_rate": 4.192566692566692e-06,
"loss": 0.2211,
"step": 14150
},
{
"epoch": 1.838901838901839,
"grad_norm": 0.83984375,
"learning_rate": 4.030691530691531e-06,
"loss": 0.224,
"step": 14200
},
{
"epoch": 1.8453768453768453,
"grad_norm": 0.8671875,
"learning_rate": 3.868816368816369e-06,
"loss": 0.2181,
"step": 14250
},
{
"epoch": 1.8518518518518519,
"grad_norm": 0.88671875,
"learning_rate": 3.706941206941207e-06,
"loss": 0.2262,
"step": 14300
},
{
"epoch": 1.8583268583268584,
"grad_norm": 0.890625,
"learning_rate": 3.545066045066045e-06,
"loss": 0.2195,
"step": 14350
},
{
"epoch": 1.8648018648018647,
"grad_norm": 0.86328125,
"learning_rate": 3.3831908831908837e-06,
"loss": 0.224,
"step": 14400
},
{
"epoch": 1.8712768712768713,
"grad_norm": 0.8828125,
"learning_rate": 3.2213157213157215e-06,
"loss": 0.2219,
"step": 14450
},
{
"epoch": 1.8777518777518778,
"grad_norm": 0.8046875,
"learning_rate": 3.0594405594405598e-06,
"loss": 0.2206,
"step": 14500
},
{
"epoch": 1.8842268842268841,
"grad_norm": 0.78515625,
"learning_rate": 2.8975653975653976e-06,
"loss": 0.216,
"step": 14550
},
{
"epoch": 1.8907018907018907,
"grad_norm": 0.87109375,
"learning_rate": 2.735690235690236e-06,
"loss": 0.222,
"step": 14600
},
{
"epoch": 1.8971768971768972,
"grad_norm": 0.8359375,
"learning_rate": 2.573815073815074e-06,
"loss": 0.2239,
"step": 14650
},
{
"epoch": 1.9036519036519035,
"grad_norm": 0.875,
"learning_rate": 2.411939911939912e-06,
"loss": 0.222,
"step": 14700
},
{
"epoch": 1.9101269101269103,
"grad_norm": 0.87890625,
"learning_rate": 2.2500647500647503e-06,
"loss": 0.2217,
"step": 14750
},
{
"epoch": 1.9166019166019166,
"grad_norm": 0.87109375,
"learning_rate": 2.088189588189588e-06,
"loss": 0.2244,
"step": 14800
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.87890625,
"learning_rate": 1.9263144263144264e-06,
"loss": 0.2193,
"step": 14850
},
{
"epoch": 1.9295519295519297,
"grad_norm": 0.859375,
"learning_rate": 1.7644392644392646e-06,
"loss": 0.2214,
"step": 14900
},
{
"epoch": 1.936026936026936,
"grad_norm": 0.86328125,
"learning_rate": 1.6025641025641025e-06,
"loss": 0.2191,
"step": 14950
},
{
"epoch": 1.9425019425019425,
"grad_norm": 0.78515625,
"learning_rate": 1.4406889406889407e-06,
"loss": 0.2192,
"step": 15000
},
{
"epoch": 1.948976948976949,
"grad_norm": 0.87109375,
"learning_rate": 1.2788137788137788e-06,
"loss": 0.2209,
"step": 15050
},
{
"epoch": 1.9554519554519554,
"grad_norm": 0.87890625,
"learning_rate": 1.116938616938617e-06,
"loss": 0.2233,
"step": 15100
},
{
"epoch": 1.961926961926962,
"grad_norm": 0.90234375,
"learning_rate": 9.550634550634551e-07,
"loss": 0.2192,
"step": 15150
},
{
"epoch": 1.9684019684019685,
"grad_norm": 0.91796875,
"learning_rate": 7.931882931882932e-07,
"loss": 0.2223,
"step": 15200
},
{
"epoch": 1.9748769748769748,
"grad_norm": 0.89453125,
"learning_rate": 6.313131313131313e-07,
"loss": 0.2202,
"step": 15250
},
{
"epoch": 1.9813519813519813,
"grad_norm": 0.796875,
"learning_rate": 4.694379694379695e-07,
"loss": 0.2197,
"step": 15300
},
{
"epoch": 1.9878269878269879,
"grad_norm": 0.859375,
"learning_rate": 3.075628075628076e-07,
"loss": 0.2177,
"step": 15350
},
{
"epoch": 1.9943019943019942,
"grad_norm": 0.94140625,
"learning_rate": 1.456876456876457e-07,
"loss": 0.2211,
"step": 15400
}
],
"logging_steps": 50,
"max_steps": 15444,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.159340177340498e+18,
"train_batch_size": 6,
"trial_name": null,
"trial_params": null
}