llavaov_sft6 / trainer_state.json
Jinghao-Guo's picture
Upload folder using huggingface_hub
3f039a6 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 3500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002857142857142857,
"grad_norm": 8.25,
"learning_rate": 9.999991910858043e-06,
"loss": 2.0631,
"step": 10
},
{
"epoch": 0.005714285714285714,
"grad_norm": 2.84375,
"learning_rate": 9.999708793637737e-06,
"loss": 1.5602,
"step": 20
},
{
"epoch": 0.008571428571428572,
"grad_norm": 1.890625,
"learning_rate": 9.999021245492833e-06,
"loss": 1.4412,
"step": 30
},
{
"epoch": 0.011428571428571429,
"grad_norm": 1.7421875,
"learning_rate": 9.997929322039716e-06,
"loss": 1.4112,
"step": 40
},
{
"epoch": 0.014285714285714285,
"grad_norm": 1.7109375,
"learning_rate": 9.996433111605053e-06,
"loss": 1.4083,
"step": 50
},
{
"epoch": 0.017142857142857144,
"grad_norm": 1.625,
"learning_rate": 9.994532735218647e-06,
"loss": 1.3948,
"step": 60
},
{
"epoch": 0.02,
"grad_norm": 1.6875,
"learning_rate": 9.992228346603645e-06,
"loss": 1.3753,
"step": 70
},
{
"epoch": 0.022857142857142857,
"grad_norm": 1.6171875,
"learning_rate": 9.98952013216411e-06,
"loss": 1.3471,
"step": 80
},
{
"epoch": 0.025714285714285714,
"grad_norm": 1.703125,
"learning_rate": 9.986408310969932e-06,
"loss": 1.3213,
"step": 90
},
{
"epoch": 0.02857142857142857,
"grad_norm": 1.5234375,
"learning_rate": 9.982893134739118e-06,
"loss": 1.319,
"step": 100
},
{
"epoch": 0.03142857142857143,
"grad_norm": 1.6328125,
"learning_rate": 9.978974887817424e-06,
"loss": 1.3487,
"step": 110
},
{
"epoch": 0.03428571428571429,
"grad_norm": 1.5703125,
"learning_rate": 9.97465388715535e-06,
"loss": 1.3334,
"step": 120
},
{
"epoch": 0.037142857142857144,
"grad_norm": 1.6328125,
"learning_rate": 9.969930482282518e-06,
"loss": 1.286,
"step": 130
},
{
"epoch": 0.04,
"grad_norm": 1.5078125,
"learning_rate": 9.964805055279374e-06,
"loss": 1.2842,
"step": 140
},
{
"epoch": 0.04285714285714286,
"grad_norm": 1.5703125,
"learning_rate": 9.959278020746308e-06,
"loss": 1.3026,
"step": 150
},
{
"epoch": 0.045714285714285714,
"grad_norm": 1.546875,
"learning_rate": 9.953349825770093e-06,
"loss": 1.3521,
"step": 160
},
{
"epoch": 0.04857142857142857,
"grad_norm": 1.4921875,
"learning_rate": 9.947020949887732e-06,
"loss": 1.284,
"step": 170
},
{
"epoch": 0.05142857142857143,
"grad_norm": 1.5625,
"learning_rate": 9.940291905047668e-06,
"loss": 1.2834,
"step": 180
},
{
"epoch": 0.054285714285714284,
"grad_norm": 1.6484375,
"learning_rate": 9.933163235568369e-06,
"loss": 1.312,
"step": 190
},
{
"epoch": 0.05714285714285714,
"grad_norm": 1.5,
"learning_rate": 9.925635518094295e-06,
"loss": 1.2968,
"step": 200
},
{
"epoch": 0.06,
"grad_norm": 1.609375,
"learning_rate": 9.917709361549257e-06,
"loss": 1.3068,
"step": 210
},
{
"epoch": 0.06285714285714286,
"grad_norm": 1.625,
"learning_rate": 9.909385407087164e-06,
"loss": 1.2879,
"step": 220
},
{
"epoch": 0.06571428571428571,
"grad_norm": 1.53125,
"learning_rate": 9.900664328040151e-06,
"loss": 1.2828,
"step": 230
},
{
"epoch": 0.06857142857142857,
"grad_norm": 1.5078125,
"learning_rate": 9.891546829864115e-06,
"loss": 1.2977,
"step": 240
},
{
"epoch": 0.07142857142857142,
"grad_norm": 1.671875,
"learning_rate": 9.882033650081656e-06,
"loss": 1.3454,
"step": 250
},
{
"epoch": 0.07428571428571429,
"grad_norm": 1.4609375,
"learning_rate": 9.87212555822241e-06,
"loss": 1.2249,
"step": 260
},
{
"epoch": 0.07714285714285714,
"grad_norm": 1.5390625,
"learning_rate": 9.861823355760806e-06,
"loss": 1.2945,
"step": 270
},
{
"epoch": 0.08,
"grad_norm": 1.4296875,
"learning_rate": 9.851127876051236e-06,
"loss": 1.3094,
"step": 280
},
{
"epoch": 0.08285714285714285,
"grad_norm": 1.5625,
"learning_rate": 9.840039984260634e-06,
"loss": 1.3061,
"step": 290
},
{
"epoch": 0.08571428571428572,
"grad_norm": 1.484375,
"learning_rate": 9.828560577298505e-06,
"loss": 1.2716,
"step": 300
},
{
"epoch": 0.08857142857142856,
"grad_norm": 1.6015625,
"learning_rate": 9.816690583744366e-06,
"loss": 1.27,
"step": 310
},
{
"epoch": 0.09142857142857143,
"grad_norm": 1.71875,
"learning_rate": 9.804430963772632e-06,
"loss": 1.2911,
"step": 320
},
{
"epoch": 0.09428571428571429,
"grad_norm": 1.484375,
"learning_rate": 9.791782709074944e-06,
"loss": 1.2916,
"step": 330
},
{
"epoch": 0.09714285714285714,
"grad_norm": 1.4453125,
"learning_rate": 9.778746842779962e-06,
"loss": 1.2836,
"step": 340
},
{
"epoch": 0.1,
"grad_norm": 2.046875,
"learning_rate": 9.765324419370595e-06,
"loss": 1.2619,
"step": 350
},
{
"epoch": 0.10285714285714286,
"grad_norm": 1.390625,
"learning_rate": 9.751516524598697e-06,
"loss": 1.2661,
"step": 360
},
{
"epoch": 0.10571428571428572,
"grad_norm": 1.59375,
"learning_rate": 9.73732427539725e-06,
"loss": 1.2996,
"step": 370
},
{
"epoch": 0.10857142857142857,
"grad_norm": 1.3515625,
"learning_rate": 9.722748819790012e-06,
"loss": 1.2307,
"step": 380
},
{
"epoch": 0.11142857142857143,
"grad_norm": 2.65625,
"learning_rate": 9.707791336798642e-06,
"loss": 1.3046,
"step": 390
},
{
"epoch": 0.11428571428571428,
"grad_norm": 1.5390625,
"learning_rate": 9.692453036347352e-06,
"loss": 1.3182,
"step": 400
},
{
"epoch": 0.11714285714285715,
"grad_norm": 1.5546875,
"learning_rate": 9.676735159165001e-06,
"loss": 1.2765,
"step": 410
},
{
"epoch": 0.12,
"grad_norm": 1.609375,
"learning_rate": 9.660638976684766e-06,
"loss": 1.3235,
"step": 420
},
{
"epoch": 0.12285714285714286,
"grad_norm": 1.65625,
"learning_rate": 9.64416579094127e-06,
"loss": 1.2747,
"step": 430
},
{
"epoch": 0.12571428571428572,
"grad_norm": 1.6171875,
"learning_rate": 9.627316934465266e-06,
"loss": 1.275,
"step": 440
},
{
"epoch": 0.12857142857142856,
"grad_norm": 1.40625,
"learning_rate": 9.610093770175856e-06,
"loss": 1.2651,
"step": 450
},
{
"epoch": 0.13142857142857142,
"grad_norm": 1.6171875,
"learning_rate": 9.592497691270234e-06,
"loss": 1.2993,
"step": 460
},
{
"epoch": 0.13428571428571429,
"grad_norm": 1.515625,
"learning_rate": 9.57453012111099e-06,
"loss": 1.2805,
"step": 470
},
{
"epoch": 0.13714285714285715,
"grad_norm": 1.59375,
"learning_rate": 9.556192513110974e-06,
"loss": 1.2978,
"step": 480
},
{
"epoch": 0.14,
"grad_norm": 2.171875,
"learning_rate": 9.537486350615727e-06,
"loss": 1.2951,
"step": 490
},
{
"epoch": 0.14285714285714285,
"grad_norm": 1.859375,
"learning_rate": 9.518413146783502e-06,
"loss": 1.275,
"step": 500
},
{
"epoch": 0.1457142857142857,
"grad_norm": 1.4921875,
"learning_rate": 9.498974444462847e-06,
"loss": 1.2869,
"step": 510
},
{
"epoch": 0.14857142857142858,
"grad_norm": 1.34375,
"learning_rate": 9.479171816067808e-06,
"loss": 1.2968,
"step": 520
},
{
"epoch": 0.15142857142857144,
"grad_norm": 1.40625,
"learning_rate": 9.459006863450745e-06,
"loss": 1.2781,
"step": 530
},
{
"epoch": 0.15428571428571428,
"grad_norm": 1.75,
"learning_rate": 9.438481217772744e-06,
"loss": 1.238,
"step": 540
},
{
"epoch": 0.15714285714285714,
"grad_norm": 1.40625,
"learning_rate": 9.417596539371677e-06,
"loss": 1.2414,
"step": 550
},
{
"epoch": 0.16,
"grad_norm": 1.6953125,
"learning_rate": 9.396354517627893e-06,
"loss": 1.2428,
"step": 560
},
{
"epoch": 0.16285714285714287,
"grad_norm": 1.4921875,
"learning_rate": 9.37475687082757e-06,
"loss": 1.2304,
"step": 570
},
{
"epoch": 0.1657142857142857,
"grad_norm": 1.453125,
"learning_rate": 9.35280534602371e-06,
"loss": 1.2377,
"step": 580
},
{
"epoch": 0.16857142857142857,
"grad_norm": 1.6171875,
"learning_rate": 9.330501718894822e-06,
"loss": 1.2865,
"step": 590
},
{
"epoch": 0.17142857142857143,
"grad_norm": 1.484375,
"learning_rate": 9.307847793601295e-06,
"loss": 1.2753,
"step": 600
},
{
"epoch": 0.1742857142857143,
"grad_norm": 2.078125,
"learning_rate": 9.284845402639447e-06,
"loss": 1.2317,
"step": 610
},
{
"epoch": 0.17714285714285713,
"grad_norm": 1.546875,
"learning_rate": 9.261496406693291e-06,
"loss": 1.2473,
"step": 620
},
{
"epoch": 0.18,
"grad_norm": 1.375,
"learning_rate": 9.237802694484035e-06,
"loss": 1.2379,
"step": 630
},
{
"epoch": 0.18285714285714286,
"grad_norm": 1.4375,
"learning_rate": 9.213766182617291e-06,
"loss": 1.2274,
"step": 640
},
{
"epoch": 0.18571428571428572,
"grad_norm": 1.3828125,
"learning_rate": 9.189388815428047e-06,
"loss": 1.2372,
"step": 650
},
{
"epoch": 0.18857142857142858,
"grad_norm": 1.5390625,
"learning_rate": 9.164672564823374e-06,
"loss": 1.2541,
"step": 660
},
{
"epoch": 0.19142857142857142,
"grad_norm": 1.828125,
"learning_rate": 9.139619430122934e-06,
"loss": 1.2972,
"step": 670
},
{
"epoch": 0.19428571428571428,
"grad_norm": 1.4375,
"learning_rate": 9.114231437897245e-06,
"loss": 1.2356,
"step": 680
},
{
"epoch": 0.19714285714285715,
"grad_norm": 1.90625,
"learning_rate": 9.088510641803747e-06,
"loss": 1.2966,
"step": 690
},
{
"epoch": 0.2,
"grad_norm": 1.4140625,
"learning_rate": 9.062459122420684e-06,
"loss": 1.231,
"step": 700
},
{
"epoch": 0.20285714285714285,
"grad_norm": 1.453125,
"learning_rate": 9.036078987078804e-06,
"loss": 1.2295,
"step": 710
},
{
"epoch": 0.2057142857142857,
"grad_norm": 1.5078125,
"learning_rate": 9.009372369690897e-06,
"loss": 1.2599,
"step": 720
},
{
"epoch": 0.20857142857142857,
"grad_norm": 1.546875,
"learning_rate": 8.982341430579174e-06,
"loss": 1.2406,
"step": 730
},
{
"epoch": 0.21142857142857144,
"grad_norm": 1.75,
"learning_rate": 8.954988356300518e-06,
"loss": 1.2162,
"step": 740
},
{
"epoch": 0.21428571428571427,
"grad_norm": 1.6171875,
"learning_rate": 8.927315359469627e-06,
"loss": 1.2414,
"step": 750
},
{
"epoch": 0.21714285714285714,
"grad_norm": 1.2890625,
"learning_rate": 8.899324678580005e-06,
"loss": 1.2257,
"step": 760
},
{
"epoch": 0.22,
"grad_norm": 1.5703125,
"learning_rate": 8.871018577822916e-06,
"loss": 1.2735,
"step": 770
},
{
"epoch": 0.22285714285714286,
"grad_norm": 1.421875,
"learning_rate": 8.842399346904215e-06,
"loss": 1.2487,
"step": 780
},
{
"epoch": 0.2257142857142857,
"grad_norm": 1.5,
"learning_rate": 8.813469300859135e-06,
"loss": 1.2696,
"step": 790
},
{
"epoch": 0.22857142857142856,
"grad_norm": 1.2890625,
"learning_rate": 8.784230779865029e-06,
"loss": 1.2262,
"step": 800
},
{
"epoch": 0.23142857142857143,
"grad_norm": 1.59375,
"learning_rate": 8.754686149052057e-06,
"loss": 1.2514,
"step": 810
},
{
"epoch": 0.2342857142857143,
"grad_norm": 1.46875,
"learning_rate": 8.724837798311883e-06,
"loss": 1.2419,
"step": 820
},
{
"epoch": 0.23714285714285716,
"grad_norm": 1.4453125,
"learning_rate": 8.694688142104346e-06,
"loss": 1.2941,
"step": 830
},
{
"epoch": 0.24,
"grad_norm": 1.4765625,
"learning_rate": 8.664239619262151e-06,
"loss": 1.2273,
"step": 840
},
{
"epoch": 0.24285714285714285,
"grad_norm": 1.4296875,
"learning_rate": 8.6334946927936e-06,
"loss": 1.249,
"step": 850
},
{
"epoch": 0.24571428571428572,
"grad_norm": 1.4921875,
"learning_rate": 8.602455849683343e-06,
"loss": 1.2663,
"step": 860
},
{
"epoch": 0.24857142857142858,
"grad_norm": 1.640625,
"learning_rate": 8.571125600691214e-06,
"loss": 1.2835,
"step": 870
},
{
"epoch": 0.25142857142857145,
"grad_norm": 1.6796875,
"learning_rate": 8.53950648014913e-06,
"loss": 1.2804,
"step": 880
},
{
"epoch": 0.2542857142857143,
"grad_norm": 1.5,
"learning_rate": 8.507601045756085e-06,
"loss": 1.2981,
"step": 890
},
{
"epoch": 0.2571428571428571,
"grad_norm": 1.71875,
"learning_rate": 8.475411878371257e-06,
"loss": 1.1956,
"step": 900
},
{
"epoch": 0.26,
"grad_norm": 1.5625,
"learning_rate": 8.442941581805245e-06,
"loss": 1.2578,
"step": 910
},
{
"epoch": 0.26285714285714284,
"grad_norm": 1.46875,
"learning_rate": 8.410192782609428e-06,
"loss": 1.2073,
"step": 920
},
{
"epoch": 0.26571428571428574,
"grad_norm": 1.59375,
"learning_rate": 8.377168129863526e-06,
"loss": 1.2343,
"step": 930
},
{
"epoch": 0.26857142857142857,
"grad_norm": 1.5,
"learning_rate": 8.34387029496129e-06,
"loss": 1.2565,
"step": 940
},
{
"epoch": 0.2714285714285714,
"grad_norm": 2.265625,
"learning_rate": 8.310301971394423e-06,
"loss": 1.2424,
"step": 950
},
{
"epoch": 0.2742857142857143,
"grad_norm": 1.546875,
"learning_rate": 8.276465874534701e-06,
"loss": 1.2834,
"step": 960
},
{
"epoch": 0.27714285714285714,
"grad_norm": 1.5859375,
"learning_rate": 8.242364741414321e-06,
"loss": 1.2526,
"step": 970
},
{
"epoch": 0.28,
"grad_norm": 1.703125,
"learning_rate": 8.208001330504497e-06,
"loss": 1.2253,
"step": 980
},
{
"epoch": 0.28285714285714286,
"grad_norm": 1.6171875,
"learning_rate": 8.17337842149233e-06,
"loss": 1.2526,
"step": 990
},
{
"epoch": 0.2857142857142857,
"grad_norm": 1.9609375,
"learning_rate": 8.138498815055959e-06,
"loss": 1.2918,
"step": 1000
},
{
"epoch": 0.2885714285714286,
"grad_norm": 1.453125,
"learning_rate": 8.103365332638e-06,
"loss": 1.2428,
"step": 1010
},
{
"epoch": 0.2914285714285714,
"grad_norm": 1.3984375,
"learning_rate": 8.067980816217336e-06,
"loss": 1.2538,
"step": 1020
},
{
"epoch": 0.29428571428571426,
"grad_norm": 1.5859375,
"learning_rate": 8.032348128079204e-06,
"loss": 1.253,
"step": 1030
},
{
"epoch": 0.29714285714285715,
"grad_norm": 1.390625,
"learning_rate": 7.996470150583678e-06,
"loss": 1.2471,
"step": 1040
},
{
"epoch": 0.3,
"grad_norm": 1.4765625,
"learning_rate": 7.96034978593251e-06,
"loss": 1.2314,
"step": 1050
},
{
"epoch": 0.3028571428571429,
"grad_norm": 1.4375,
"learning_rate": 7.923989955934363e-06,
"loss": 1.2425,
"step": 1060
},
{
"epoch": 0.3057142857142857,
"grad_norm": 1.4296875,
"learning_rate": 7.887393601768468e-06,
"loss": 1.2446,
"step": 1070
},
{
"epoch": 0.30857142857142855,
"grad_norm": 1.9921875,
"learning_rate": 7.85056368374671e-06,
"loss": 1.2594,
"step": 1080
},
{
"epoch": 0.31142857142857144,
"grad_norm": 1.65625,
"learning_rate": 7.813503181074157e-06,
"loss": 1.2098,
"step": 1090
},
{
"epoch": 0.3142857142857143,
"grad_norm": 1.59375,
"learning_rate": 7.776215091608087e-06,
"loss": 1.2196,
"step": 1100
},
{
"epoch": 0.3171428571428571,
"grad_norm": 1.4921875,
"learning_rate": 7.738702431615464e-06,
"loss": 1.2226,
"step": 1110
},
{
"epoch": 0.32,
"grad_norm": 1.6484375,
"learning_rate": 7.700968235528973e-06,
"loss": 1.2617,
"step": 1120
},
{
"epoch": 0.32285714285714284,
"grad_norm": 1.8828125,
"learning_rate": 7.663015555701552e-06,
"loss": 1.2693,
"step": 1130
},
{
"epoch": 0.32571428571428573,
"grad_norm": 2.875,
"learning_rate": 7.624847462159479e-06,
"loss": 1.258,
"step": 1140
},
{
"epoch": 0.32857142857142857,
"grad_norm": 1.5,
"learning_rate": 7.586467042354044e-06,
"loss": 1.24,
"step": 1150
},
{
"epoch": 0.3314285714285714,
"grad_norm": 1.5625,
"learning_rate": 7.547877400911798e-06,
"loss": 1.2413,
"step": 1160
},
{
"epoch": 0.3342857142857143,
"grad_norm": 1.4609375,
"learning_rate": 7.509081659383417e-06,
"loss": 1.232,
"step": 1170
},
{
"epoch": 0.33714285714285713,
"grad_norm": 1.4140625,
"learning_rate": 7.4700829559912005e-06,
"loss": 1.2422,
"step": 1180
},
{
"epoch": 0.34,
"grad_norm": 1.4921875,
"learning_rate": 7.430884445375212e-06,
"loss": 1.2341,
"step": 1190
},
{
"epoch": 0.34285714285714286,
"grad_norm": 1.59375,
"learning_rate": 7.391489298338099e-06,
"loss": 1.2236,
"step": 1200
},
{
"epoch": 0.3457142857142857,
"grad_norm": 1.65625,
"learning_rate": 7.351900701588612e-06,
"loss": 1.2481,
"step": 1210
},
{
"epoch": 0.3485714285714286,
"grad_norm": 1.765625,
"learning_rate": 7.312121857483816e-06,
"loss": 1.2099,
"step": 1220
},
{
"epoch": 0.3514285714285714,
"grad_norm": 1.53125,
"learning_rate": 7.272155983770055e-06,
"loss": 1.2585,
"step": 1230
},
{
"epoch": 0.35428571428571426,
"grad_norm": 1.640625,
"learning_rate": 7.232006313322668e-06,
"loss": 1.2337,
"step": 1240
},
{
"epoch": 0.35714285714285715,
"grad_norm": 1.34375,
"learning_rate": 7.191676093884479e-06,
"loss": 1.2221,
"step": 1250
},
{
"epoch": 0.36,
"grad_norm": 2.15625,
"learning_rate": 7.151168587803074e-06,
"loss": 1.2662,
"step": 1260
},
{
"epoch": 0.3628571428571429,
"grad_norm": 1.5078125,
"learning_rate": 7.110487071766924e-06,
"loss": 1.2171,
"step": 1270
},
{
"epoch": 0.3657142857142857,
"grad_norm": 1.65625,
"learning_rate": 7.069634836540311e-06,
"loss": 1.221,
"step": 1280
},
{
"epoch": 0.36857142857142855,
"grad_norm": 1.4375,
"learning_rate": 7.0286151866971544e-06,
"loss": 1.2431,
"step": 1290
},
{
"epoch": 0.37142857142857144,
"grad_norm": 1.5546875,
"learning_rate": 6.987431440353687e-06,
"loss": 1.2335,
"step": 1300
},
{
"epoch": 0.3742857142857143,
"grad_norm": 1.640625,
"learning_rate": 6.946086928900054e-06,
"loss": 1.1925,
"step": 1310
},
{
"epoch": 0.37714285714285717,
"grad_norm": 1.53125,
"learning_rate": 6.904584996730839e-06,
"loss": 1.2303,
"step": 1320
},
{
"epoch": 0.38,
"grad_norm": 1.6875,
"learning_rate": 6.862929000974523e-06,
"loss": 1.2511,
"step": 1330
},
{
"epoch": 0.38285714285714284,
"grad_norm": 1.328125,
"learning_rate": 6.821122311221932e-06,
"loss": 1.2166,
"step": 1340
},
{
"epoch": 0.38571428571428573,
"grad_norm": 1.4453125,
"learning_rate": 6.779168309253663e-06,
"loss": 1.2181,
"step": 1350
},
{
"epoch": 0.38857142857142857,
"grad_norm": 1.53125,
"learning_rate": 6.73707038876653e-06,
"loss": 1.2358,
"step": 1360
},
{
"epoch": 0.3914285714285714,
"grad_norm": 1.7578125,
"learning_rate": 6.6948319550990485e-06,
"loss": 1.2176,
"step": 1370
},
{
"epoch": 0.3942857142857143,
"grad_norm": 1.3828125,
"learning_rate": 6.6524564249559645e-06,
"loss": 1.2317,
"step": 1380
},
{
"epoch": 0.39714285714285713,
"grad_norm": 1.609375,
"learning_rate": 6.6099472261318864e-06,
"loss": 1.306,
"step": 1390
},
{
"epoch": 0.4,
"grad_norm": 1.5703125,
"learning_rate": 6.567307797233997e-06,
"loss": 1.253,
"step": 1400
},
{
"epoch": 0.40285714285714286,
"grad_norm": 2.75,
"learning_rate": 6.524541587403915e-06,
"loss": 1.2285,
"step": 1410
},
{
"epoch": 0.4057142857142857,
"grad_norm": 1.40625,
"learning_rate": 6.481652056038672e-06,
"loss": 1.2117,
"step": 1420
},
{
"epoch": 0.4085714285714286,
"grad_norm": 1.65625,
"learning_rate": 6.438642672510894e-06,
"loss": 1.2791,
"step": 1430
},
{
"epoch": 0.4114285714285714,
"grad_norm": 1.9296875,
"learning_rate": 6.395516915888158e-06,
"loss": 1.2448,
"step": 1440
},
{
"epoch": 0.4142857142857143,
"grad_norm": 1.328125,
"learning_rate": 6.352278274651562e-06,
"loss": 1.2291,
"step": 1450
},
{
"epoch": 0.41714285714285715,
"grad_norm": 1.546875,
"learning_rate": 6.308930246413541e-06,
"loss": 1.2544,
"step": 1460
},
{
"epoch": 0.42,
"grad_norm": 1.7890625,
"learning_rate": 6.265476337634942e-06,
"loss": 1.2402,
"step": 1470
},
{
"epoch": 0.4228571428571429,
"grad_norm": 1.4296875,
"learning_rate": 6.22192006334139e-06,
"loss": 1.2525,
"step": 1480
},
{
"epoch": 0.4257142857142857,
"grad_norm": 1.7265625,
"learning_rate": 6.178264946838942e-06,
"loss": 1.2481,
"step": 1490
},
{
"epoch": 0.42857142857142855,
"grad_norm": 1.484375,
"learning_rate": 6.13451451942909e-06,
"loss": 1.2281,
"step": 1500
},
{
"epoch": 0.43142857142857144,
"grad_norm": 1.6015625,
"learning_rate": 6.090672320123113e-06,
"loss": 1.2064,
"step": 1510
},
{
"epoch": 0.4342857142857143,
"grad_norm": 1.5546875,
"learning_rate": 6.046741895355802e-06,
"loss": 1.2248,
"step": 1520
},
{
"epoch": 0.43714285714285717,
"grad_norm": 1.484375,
"learning_rate": 6.00272679869858e-06,
"loss": 1.2337,
"step": 1530
},
{
"epoch": 0.44,
"grad_norm": 1.53125,
"learning_rate": 5.958630590572061e-06,
"loss": 1.241,
"step": 1540
},
{
"epoch": 0.44285714285714284,
"grad_norm": 1.4140625,
"learning_rate": 5.914456837958032e-06,
"loss": 1.222,
"step": 1550
},
{
"epoch": 0.44571428571428573,
"grad_norm": 1.3515625,
"learning_rate": 5.87020911411093e-06,
"loss": 1.2441,
"step": 1560
},
{
"epoch": 0.44857142857142857,
"grad_norm": 1.6171875,
"learning_rate": 5.825890998268784e-06,
"loss": 1.2635,
"step": 1570
},
{
"epoch": 0.4514285714285714,
"grad_norm": 1.5,
"learning_rate": 5.781506075363702e-06,
"loss": 1.243,
"step": 1580
},
{
"epoch": 0.4542857142857143,
"grad_norm": 1.5,
"learning_rate": 5.737057935731868e-06,
"loss": 1.2483,
"step": 1590
},
{
"epoch": 0.45714285714285713,
"grad_norm": 1.4140625,
"learning_rate": 5.692550174823129e-06,
"loss": 1.2055,
"step": 1600
},
{
"epoch": 0.46,
"grad_norm": 1.5078125,
"learning_rate": 5.647986392910149e-06,
"loss": 1.2177,
"step": 1610
},
{
"epoch": 0.46285714285714286,
"grad_norm": 2.015625,
"learning_rate": 5.6033701947971805e-06,
"loss": 1.2275,
"step": 1620
},
{
"epoch": 0.4657142857142857,
"grad_norm": 1.6328125,
"learning_rate": 5.55870518952847e-06,
"loss": 1.2497,
"step": 1630
},
{
"epoch": 0.4685714285714286,
"grad_norm": 1.328125,
"learning_rate": 5.513994990096318e-06,
"loss": 1.2057,
"step": 1640
},
{
"epoch": 0.4714285714285714,
"grad_norm": 2.421875,
"learning_rate": 5.469243213148821e-06,
"loss": 1.2558,
"step": 1650
},
{
"epoch": 0.4742857142857143,
"grad_norm": 1.6953125,
"learning_rate": 5.424453478697321e-06,
"loss": 1.2452,
"step": 1660
},
{
"epoch": 0.47714285714285715,
"grad_norm": 1.7578125,
"learning_rate": 5.379629409823571e-06,
"loss": 1.253,
"step": 1670
},
{
"epoch": 0.48,
"grad_norm": 1.578125,
"learning_rate": 5.334774632386672e-06,
"loss": 1.2358,
"step": 1680
},
{
"epoch": 0.4828571428571429,
"grad_norm": 1.828125,
"learning_rate": 5.2898927747297665e-06,
"loss": 1.2656,
"step": 1690
},
{
"epoch": 0.4857142857142857,
"grad_norm": 1.6328125,
"learning_rate": 5.244987467386536e-06,
"loss": 1.2283,
"step": 1700
},
{
"epoch": 0.48857142857142855,
"grad_norm": 1.6015625,
"learning_rate": 5.200062342787534e-06,
"loss": 1.217,
"step": 1710
},
{
"epoch": 0.49142857142857144,
"grad_norm": 1.4921875,
"learning_rate": 5.155121034966346e-06,
"loss": 1.2441,
"step": 1720
},
{
"epoch": 0.4942857142857143,
"grad_norm": 1.6171875,
"learning_rate": 5.110167179265636e-06,
"loss": 1.2197,
"step": 1730
},
{
"epoch": 0.49714285714285716,
"grad_norm": 1.515625,
"learning_rate": 5.065204412043071e-06,
"loss": 1.2542,
"step": 1740
},
{
"epoch": 0.5,
"grad_norm": 1.5546875,
"learning_rate": 5.0202363703771866e-06,
"loss": 1.2047,
"step": 1750
},
{
"epoch": 0.5028571428571429,
"grad_norm": 1.65625,
"learning_rate": 4.975266691773168e-06,
"loss": 1.2221,
"step": 1760
},
{
"epoch": 0.5057142857142857,
"grad_norm": 1.484375,
"learning_rate": 4.930299013868616e-06,
"loss": 1.2402,
"step": 1770
},
{
"epoch": 0.5085714285714286,
"grad_norm": 1.453125,
"learning_rate": 4.885336974139292e-06,
"loss": 1.1988,
"step": 1780
},
{
"epoch": 0.5114285714285715,
"grad_norm": 1.7421875,
"learning_rate": 4.840384209604881e-06,
"loss": 1.2282,
"step": 1790
},
{
"epoch": 0.5142857142857142,
"grad_norm": 1.7265625,
"learning_rate": 4.795444356534787e-06,
"loss": 1.1957,
"step": 1800
},
{
"epoch": 0.5171428571428571,
"grad_norm": 1.4453125,
"learning_rate": 4.750521050153997e-06,
"loss": 1.2562,
"step": 1810
},
{
"epoch": 0.52,
"grad_norm": 1.4453125,
"learning_rate": 4.7056179243490205e-06,
"loss": 1.2339,
"step": 1820
},
{
"epoch": 0.5228571428571429,
"grad_norm": 1.84375,
"learning_rate": 4.660738611373942e-06,
"loss": 1.2338,
"step": 1830
},
{
"epoch": 0.5257142857142857,
"grad_norm": 1.546875,
"learning_rate": 4.615886741556604e-06,
"loss": 1.2672,
"step": 1840
},
{
"epoch": 0.5285714285714286,
"grad_norm": 1.828125,
"learning_rate": 4.5710659430049475e-06,
"loss": 1.2383,
"step": 1850
},
{
"epoch": 0.5314285714285715,
"grad_norm": 1.53125,
"learning_rate": 4.5262798413135345e-06,
"loss": 1.2147,
"step": 1860
},
{
"epoch": 0.5342857142857143,
"grad_norm": 1.375,
"learning_rate": 4.4815320592702625e-06,
"loss": 1.1947,
"step": 1870
},
{
"epoch": 0.5371428571428571,
"grad_norm": 1.5390625,
"learning_rate": 4.436826216563318e-06,
"loss": 1.2432,
"step": 1880
},
{
"epoch": 0.54,
"grad_norm": 1.4296875,
"learning_rate": 4.392165929488381e-06,
"loss": 1.2544,
"step": 1890
},
{
"epoch": 0.5428571428571428,
"grad_norm": 1.59375,
"learning_rate": 4.347554810656093e-06,
"loss": 1.2482,
"step": 1900
},
{
"epoch": 0.5457142857142857,
"grad_norm": 1.6015625,
"learning_rate": 4.3029964686998335e-06,
"loss": 1.2396,
"step": 1910
},
{
"epoch": 0.5485714285714286,
"grad_norm": 1.53125,
"learning_rate": 4.25849450798381e-06,
"loss": 1.2299,
"step": 1920
},
{
"epoch": 0.5514285714285714,
"grad_norm": 1.515625,
"learning_rate": 4.2140525283115054e-06,
"loss": 1.2296,
"step": 1930
},
{
"epoch": 0.5542857142857143,
"grad_norm": 4.0,
"learning_rate": 4.169674124634482e-06,
"loss": 1.2191,
"step": 1940
},
{
"epoch": 0.5571428571428572,
"grad_norm": 1.578125,
"learning_rate": 4.125362886761577e-06,
"loss": 1.2496,
"step": 1950
},
{
"epoch": 0.56,
"grad_norm": 1.609375,
"learning_rate": 4.08112239906853e-06,
"loss": 1.2369,
"step": 1960
},
{
"epoch": 0.5628571428571428,
"grad_norm": 1.6328125,
"learning_rate": 4.036956240208039e-06,
"loss": 1.2569,
"step": 1970
},
{
"epoch": 0.5657142857142857,
"grad_norm": 1.625,
"learning_rate": 3.992867982820268e-06,
"loss": 1.2248,
"step": 1980
},
{
"epoch": 0.5685714285714286,
"grad_norm": 1.484375,
"learning_rate": 3.948861193243867e-06,
"loss": 1.2002,
"step": 1990
},
{
"epoch": 0.5714285714285714,
"grad_norm": 1.921875,
"learning_rate": 3.904939431227477e-06,
"loss": 1.23,
"step": 2000
},
{
"epoch": 0.5742857142857143,
"grad_norm": 1.6796875,
"learning_rate": 3.8611062496417895e-06,
"loss": 1.2209,
"step": 2010
},
{
"epoch": 0.5771428571428572,
"grad_norm": 1.5859375,
"learning_rate": 3.81736519419214e-06,
"loss": 1.271,
"step": 2020
},
{
"epoch": 0.58,
"grad_norm": 1.4921875,
"learning_rate": 3.7737198031317025e-06,
"loss": 1.2471,
"step": 2030
},
{
"epoch": 0.5828571428571429,
"grad_norm": 1.2109375,
"learning_rate": 3.7301736069752683e-06,
"loss": 1.2174,
"step": 2040
},
{
"epoch": 0.5857142857142857,
"grad_norm": 1.7578125,
"learning_rate": 3.686730128213668e-06,
"loss": 1.2369,
"step": 2050
},
{
"epoch": 0.5885714285714285,
"grad_norm": 1.5625,
"learning_rate": 3.6433928810288292e-06,
"loss": 1.2319,
"step": 2060
},
{
"epoch": 0.5914285714285714,
"grad_norm": 1.9375,
"learning_rate": 3.6001653710095115e-06,
"loss": 1.3194,
"step": 2070
},
{
"epoch": 0.5942857142857143,
"grad_norm": 1.65625,
"learning_rate": 3.557051094867735e-06,
"loss": 1.1943,
"step": 2080
},
{
"epoch": 0.5971428571428572,
"grad_norm": 1.5546875,
"learning_rate": 3.514053540155934e-06,
"loss": 1.2375,
"step": 2090
},
{
"epoch": 0.6,
"grad_norm": 1.78125,
"learning_rate": 3.471176184984839e-06,
"loss": 1.2498,
"step": 2100
},
{
"epoch": 0.6028571428571429,
"grad_norm": 1.4609375,
"learning_rate": 3.4284224977421333e-06,
"loss": 1.2565,
"step": 2110
},
{
"epoch": 0.6057142857142858,
"grad_norm": 1.5625,
"learning_rate": 3.3857959368118874e-06,
"loss": 1.1942,
"step": 2120
},
{
"epoch": 0.6085714285714285,
"grad_norm": 1.40625,
"learning_rate": 3.3432999502948156e-06,
"loss": 1.2208,
"step": 2130
},
{
"epoch": 0.6114285714285714,
"grad_norm": 1.6953125,
"learning_rate": 3.300937975729347e-06,
"loss": 1.2034,
"step": 2140
},
{
"epoch": 0.6142857142857143,
"grad_norm": 1.7890625,
"learning_rate": 3.2587134398135654e-06,
"loss": 1.2471,
"step": 2150
},
{
"epoch": 0.6171428571428571,
"grad_norm": 1.546875,
"learning_rate": 3.216629758128018e-06,
"loss": 1.2465,
"step": 2160
},
{
"epoch": 0.62,
"grad_norm": 1.453125,
"learning_rate": 3.1746903348594303e-06,
"loss": 1.204,
"step": 2170
},
{
"epoch": 0.6228571428571429,
"grad_norm": 1.453125,
"learning_rate": 3.1328985625253304e-06,
"loss": 1.2282,
"step": 2180
},
{
"epoch": 0.6257142857142857,
"grad_norm": 1.5859375,
"learning_rate": 3.0912578216996316e-06,
"loss": 1.2622,
"step": 2190
},
{
"epoch": 0.6285714285714286,
"grad_norm": 1.71875,
"learning_rate": 3.0497714807391776e-06,
"loss": 1.2507,
"step": 2200
},
{
"epoch": 0.6314285714285715,
"grad_norm": 1.5,
"learning_rate": 3.008442895511263e-06,
"loss": 1.2277,
"step": 2210
},
{
"epoch": 0.6342857142857142,
"grad_norm": 1.65625,
"learning_rate": 2.9672754091221807e-06,
"loss": 1.2164,
"step": 2220
},
{
"epoch": 0.6371428571428571,
"grad_norm": 1.6171875,
"learning_rate": 2.926272351646794e-06,
"loss": 1.2358,
"step": 2230
},
{
"epoch": 0.64,
"grad_norm": 1.5546875,
"learning_rate": 2.88543703985917e-06,
"loss": 1.2415,
"step": 2240
},
{
"epoch": 0.6428571428571429,
"grad_norm": 1.53125,
"learning_rate": 2.8447727769642697e-06,
"loss": 1.2594,
"step": 2250
},
{
"epoch": 0.6457142857142857,
"grad_norm": 1.78125,
"learning_rate": 2.8042828523307562e-06,
"loss": 1.2364,
"step": 2260
},
{
"epoch": 0.6485714285714286,
"grad_norm": 1.5234375,
"learning_rate": 2.7639705412249185e-06,
"loss": 1.2227,
"step": 2270
},
{
"epoch": 0.6514285714285715,
"grad_norm": 2.1875,
"learning_rate": 2.7238391045457273e-06,
"loss": 1.2627,
"step": 2280
},
{
"epoch": 0.6542857142857142,
"grad_norm": 1.546875,
"learning_rate": 2.683891788561055e-06,
"loss": 1.2754,
"step": 2290
},
{
"epoch": 0.6571428571428571,
"grad_norm": 1.3046875,
"learning_rate": 2.644131824645092e-06,
"loss": 1.2215,
"step": 2300
},
{
"epoch": 0.66,
"grad_norm": 1.7890625,
"learning_rate": 2.604562429016944e-06,
"loss": 1.2072,
"step": 2310
},
{
"epoch": 0.6628571428571428,
"grad_norm": 1.65625,
"learning_rate": 2.5651868024804847e-06,
"loss": 1.2342,
"step": 2320
},
{
"epoch": 0.6657142857142857,
"grad_norm": 1.6171875,
"learning_rate": 2.526008130165432e-06,
"loss": 1.2508,
"step": 2330
},
{
"epoch": 0.6685714285714286,
"grad_norm": 1.515625,
"learning_rate": 2.4870295812696924e-06,
"loss": 1.2335,
"step": 2340
},
{
"epoch": 0.6714285714285714,
"grad_norm": 1.4921875,
"learning_rate": 2.4482543088030188e-06,
"loss": 1.2186,
"step": 2350
},
{
"epoch": 0.6742857142857143,
"grad_norm": 1.453125,
"learning_rate": 2.4096854493319476e-06,
"loss": 1.2424,
"step": 2360
},
{
"epoch": 0.6771428571428572,
"grad_norm": 1.6015625,
"learning_rate": 2.3713261227260815e-06,
"loss": 1.2232,
"step": 2370
},
{
"epoch": 0.68,
"grad_norm": 1.5546875,
"learning_rate": 2.333179431905726e-06,
"loss": 1.1973,
"step": 2380
},
{
"epoch": 0.6828571428571428,
"grad_norm": 1.671875,
"learning_rate": 2.295248462590883e-06,
"loss": 1.2461,
"step": 2390
},
{
"epoch": 0.6857142857142857,
"grad_norm": 1.40625,
"learning_rate": 2.257536283051651e-06,
"loss": 1.17,
"step": 2400
},
{
"epoch": 0.6885714285714286,
"grad_norm": 2.03125,
"learning_rate": 2.2200459438600296e-06,
"loss": 1.2275,
"step": 2410
},
{
"epoch": 0.6914285714285714,
"grad_norm": 1.5703125,
"learning_rate": 2.1827804776431478e-06,
"loss": 1.2649,
"step": 2420
},
{
"epoch": 0.6942857142857143,
"grad_norm": 1.5703125,
"learning_rate": 2.145742898837964e-06,
"loss": 1.2343,
"step": 2430
},
{
"epoch": 0.6971428571428572,
"grad_norm": 1.5,
"learning_rate": 2.1089362034474176e-06,
"loss": 1.2209,
"step": 2440
},
{
"epoch": 0.7,
"grad_norm": 1.5859375,
"learning_rate": 2.07236336879808e-06,
"loss": 1.2363,
"step": 2450
},
{
"epoch": 0.7028571428571428,
"grad_norm": 1.640625,
"learning_rate": 2.0360273532993195e-06,
"loss": 1.2301,
"step": 2460
},
{
"epoch": 0.7057142857142857,
"grad_norm": 1.4140625,
"learning_rate": 1.9999310962039937e-06,
"loss": 1.2704,
"step": 2470
},
{
"epoch": 0.7085714285714285,
"grad_norm": 1.453125,
"learning_rate": 1.964077517370681e-06,
"loss": 1.1998,
"step": 2480
},
{
"epoch": 0.7114285714285714,
"grad_norm": 1.671875,
"learning_rate": 1.9284695170275054e-06,
"loss": 1.2305,
"step": 2490
},
{
"epoch": 0.7142857142857143,
"grad_norm": 1.59375,
"learning_rate": 1.8931099755375203e-06,
"loss": 1.21,
"step": 2500
},
{
"epoch": 0.7171428571428572,
"grad_norm": 1.4609375,
"learning_rate": 1.858001753165724e-06,
"loss": 1.2015,
"step": 2510
},
{
"epoch": 0.72,
"grad_norm": 1.4375,
"learning_rate": 1.823147689847687e-06,
"loss": 1.2115,
"step": 2520
},
{
"epoch": 0.7228571428571429,
"grad_norm": 2.015625,
"learning_rate": 1.7885506049598201e-06,
"loss": 1.2317,
"step": 2530
},
{
"epoch": 0.7257142857142858,
"grad_norm": 2.265625,
"learning_rate": 1.7542132970913251e-06,
"loss": 1.2063,
"step": 2540
},
{
"epoch": 0.7285714285714285,
"grad_norm": 1.5,
"learning_rate": 1.720138543817807e-06,
"loss": 1.2621,
"step": 2550
},
{
"epoch": 0.7314285714285714,
"grad_norm": 1.578125,
"learning_rate": 1.6863291014765893e-06,
"loss": 1.2355,
"step": 2560
},
{
"epoch": 0.7342857142857143,
"grad_norm": 1.59375,
"learning_rate": 1.6527877049437624e-06,
"loss": 1.2525,
"step": 2570
},
{
"epoch": 0.7371428571428571,
"grad_norm": 7.28125,
"learning_rate": 1.6195170674129469e-06,
"loss": 1.2434,
"step": 2580
},
{
"epoch": 0.74,
"grad_norm": 1.6328125,
"learning_rate": 1.586519880175827e-06,
"loss": 1.2228,
"step": 2590
},
{
"epoch": 0.7428571428571429,
"grad_norm": 1.3203125,
"learning_rate": 1.5537988124044495e-06,
"loss": 1.2388,
"step": 2600
},
{
"epoch": 0.7457142857142857,
"grad_norm": 1.4609375,
"learning_rate": 1.5213565109353045e-06,
"loss": 1.2369,
"step": 2610
},
{
"epoch": 0.7485714285714286,
"grad_norm": 4.09375,
"learning_rate": 1.489195600055232e-06,
"loss": 1.2123,
"step": 2620
},
{
"epoch": 0.7514285714285714,
"grad_norm": 1.5234375,
"learning_rate": 1.4573186812891344e-06,
"loss": 1.2231,
"step": 2630
},
{
"epoch": 0.7542857142857143,
"grad_norm": 1.4765625,
"learning_rate": 1.4257283331895316e-06,
"loss": 1.2269,
"step": 2640
},
{
"epoch": 0.7571428571428571,
"grad_norm": 1.5,
"learning_rate": 1.3944271111279938e-06,
"loss": 1.2389,
"step": 2650
},
{
"epoch": 0.76,
"grad_norm": 1.359375,
"learning_rate": 1.3634175470884192e-06,
"loss": 1.2535,
"step": 2660
},
{
"epoch": 0.7628571428571429,
"grad_norm": 1.5703125,
"learning_rate": 1.3327021494622327e-06,
"loss": 1.2166,
"step": 2670
},
{
"epoch": 0.7657142857142857,
"grad_norm": 1.46875,
"learning_rate": 1.3022834028454746e-06,
"loss": 1.2672,
"step": 2680
},
{
"epoch": 0.7685714285714286,
"grad_norm": 1.6640625,
"learning_rate": 1.2721637678378135e-06,
"loss": 1.2109,
"step": 2690
},
{
"epoch": 0.7714285714285715,
"grad_norm": 1.5859375,
"learning_rate": 1.242345680843517e-06,
"loss": 1.2273,
"step": 2700
},
{
"epoch": 0.7742857142857142,
"grad_norm": 1.4453125,
"learning_rate": 1.2128315538743646e-06,
"loss": 1.2322,
"step": 2710
},
{
"epoch": 0.7771428571428571,
"grad_norm": 1.4375,
"learning_rate": 1.183623774354531e-06,
"loss": 1.2197,
"step": 2720
},
{
"epoch": 0.78,
"grad_norm": 1.890625,
"learning_rate": 1.1547247049274767e-06,
"loss": 1.2347,
"step": 2730
},
{
"epoch": 0.7828571428571428,
"grad_norm": 1.640625,
"learning_rate": 1.1261366832648258e-06,
"loss": 1.2782,
"step": 2740
},
{
"epoch": 0.7857142857142857,
"grad_norm": 1.5859375,
"learning_rate": 1.0978620218772655e-06,
"loss": 1.222,
"step": 2750
},
{
"epoch": 0.7885714285714286,
"grad_norm": 1.625,
"learning_rate": 1.0699030079274957e-06,
"loss": 1.2103,
"step": 2760
},
{
"epoch": 0.7914285714285715,
"grad_norm": 1.6875,
"learning_rate": 1.0422619030452063e-06,
"loss": 1.2241,
"step": 2770
},
{
"epoch": 0.7942857142857143,
"grad_norm": 1.546875,
"learning_rate": 1.0149409431441421e-06,
"loss": 1.24,
"step": 2780
},
{
"epoch": 0.7971428571428572,
"grad_norm": 1.765625,
"learning_rate": 9.879423382412328e-07,
"loss": 1.2566,
"step": 2790
},
{
"epoch": 0.8,
"grad_norm": 1.671875,
"learning_rate": 9.612682722778206e-07,
"loss": 1.2916,
"step": 2800
},
{
"epoch": 0.8028571428571428,
"grad_norm": 1.40625,
"learning_rate": 9.349209029430051e-07,
"loss": 1.212,
"step": 2810
},
{
"epoch": 0.8057142857142857,
"grad_norm": 1.3828125,
"learning_rate": 9.089023614991032e-07,
"loss": 1.2317,
"step": 2820
},
{
"epoch": 0.8085714285714286,
"grad_norm": 1.5703125,
"learning_rate": 8.83214752609246e-07,
"loss": 1.2229,
"step": 2830
},
{
"epoch": 0.8114285714285714,
"grad_norm": 1.609375,
"learning_rate": 8.578601541671366e-07,
"loss": 1.2755,
"step": 2840
},
{
"epoch": 0.8142857142857143,
"grad_norm": 1.4140625,
"learning_rate": 8.328406171289621e-07,
"loss": 1.2019,
"step": 2850
},
{
"epoch": 0.8171428571428572,
"grad_norm": 1.7734375,
"learning_rate": 8.081581653474945e-07,
"loss": 1.1962,
"step": 2860
},
{
"epoch": 0.82,
"grad_norm": 1.78125,
"learning_rate": 7.83814795408378e-07,
"loss": 1.2464,
"step": 2870
},
{
"epoch": 0.8228571428571428,
"grad_norm": 1.609375,
"learning_rate": 7.598124764686182e-07,
"loss": 1.1909,
"step": 2880
},
{
"epoch": 0.8257142857142857,
"grad_norm": 1.53125,
"learning_rate": 7.361531500973052e-07,
"loss": 1.2326,
"step": 2890
},
{
"epoch": 0.8285714285714286,
"grad_norm": 1.6796875,
"learning_rate": 7.128387301185502e-07,
"loss": 1.2443,
"step": 2900
},
{
"epoch": 0.8314285714285714,
"grad_norm": 1.4296875,
"learning_rate": 6.898711024566762e-07,
"loss": 1.2719,
"step": 2910
},
{
"epoch": 0.8342857142857143,
"grad_norm": 1.515625,
"learning_rate": 6.672521249836689e-07,
"loss": 1.2577,
"step": 2920
},
{
"epoch": 0.8371428571428572,
"grad_norm": 1.7109375,
"learning_rate": 6.44983627368882e-07,
"loss": 1.2606,
"step": 2930
},
{
"epoch": 0.84,
"grad_norm": 1.6875,
"learning_rate": 6.230674109310436e-07,
"loss": 1.2358,
"step": 2940
},
{
"epoch": 0.8428571428571429,
"grad_norm": 1.6875,
"learning_rate": 6.015052484925426e-07,
"loss": 1.2674,
"step": 2950
},
{
"epoch": 0.8457142857142858,
"grad_norm": 1.671875,
"learning_rate": 5.802988842360169e-07,
"loss": 1.2425,
"step": 2960
},
{
"epoch": 0.8485714285714285,
"grad_norm": 1.546875,
"learning_rate": 5.594500335632785e-07,
"loss": 1.2456,
"step": 2970
},
{
"epoch": 0.8514285714285714,
"grad_norm": 1.578125,
"learning_rate": 5.389603829565393e-07,
"loss": 1.2576,
"step": 2980
},
{
"epoch": 0.8542857142857143,
"grad_norm": 1.65625,
"learning_rate": 5.188315898419971e-07,
"loss": 1.2568,
"step": 2990
},
{
"epoch": 0.8571428571428571,
"grad_norm": 1.6484375,
"learning_rate": 4.990652824557651e-07,
"loss": 1.2363,
"step": 3000
},
{
"epoch": 0.86,
"grad_norm": 1.46875,
"learning_rate": 4.796630597121616e-07,
"loss": 1.2239,
"step": 3010
},
{
"epoch": 0.8628571428571429,
"grad_norm": 1.5859375,
"learning_rate": 4.6062649107436906e-07,
"loss": 1.265,
"step": 3020
},
{
"epoch": 0.8657142857142858,
"grad_norm": 1.5,
"learning_rate": 4.4195711642748497e-07,
"loss": 1.2642,
"step": 3030
},
{
"epoch": 0.8685714285714285,
"grad_norm": 1.5859375,
"learning_rate": 4.2365644595395163e-07,
"loss": 1.2409,
"step": 3040
},
{
"epoch": 0.8714285714285714,
"grad_norm": 1.7890625,
"learning_rate": 4.057259600114072e-07,
"loss": 1.187,
"step": 3050
},
{
"epoch": 0.8742857142857143,
"grad_norm": 1.6640625,
"learning_rate": 3.8816710901292467e-07,
"loss": 1.2412,
"step": 3060
},
{
"epoch": 0.8771428571428571,
"grad_norm": 1.625,
"learning_rate": 3.709813133096957e-07,
"loss": 1.2396,
"step": 3070
},
{
"epoch": 0.88,
"grad_norm": 1.5859375,
"learning_rate": 3.5416996307613526e-07,
"loss": 1.245,
"step": 3080
},
{
"epoch": 0.8828571428571429,
"grad_norm": 1.3984375,
"learning_rate": 3.377344181974301e-07,
"loss": 1.1911,
"step": 3090
},
{
"epoch": 0.8857142857142857,
"grad_norm": 1.6484375,
"learning_rate": 3.2167600815953106e-07,
"loss": 1.2691,
"step": 3100
},
{
"epoch": 0.8885714285714286,
"grad_norm": 1.8203125,
"learning_rate": 3.059960319416183e-07,
"loss": 1.2316,
"step": 3110
},
{
"epoch": 0.8914285714285715,
"grad_norm": 1.6171875,
"learning_rate": 2.90695757911017e-07,
"loss": 1.2343,
"step": 3120
},
{
"epoch": 0.8942857142857142,
"grad_norm": 1.4765625,
"learning_rate": 2.7577642372060676e-07,
"loss": 1.2145,
"step": 3130
},
{
"epoch": 0.8971428571428571,
"grad_norm": 1.5859375,
"learning_rate": 2.6123923620869797e-07,
"loss": 1.234,
"step": 3140
},
{
"epoch": 0.9,
"grad_norm": 1.3515625,
"learning_rate": 2.4708537130141484e-07,
"loss": 1.2177,
"step": 3150
},
{
"epoch": 0.9028571428571428,
"grad_norm": 1.7734375,
"learning_rate": 2.3331597391757444e-07,
"loss": 1.2631,
"step": 3160
},
{
"epoch": 0.9057142857142857,
"grad_norm": 1.609375,
"learning_rate": 2.1993215787606937e-07,
"loss": 1.2564,
"step": 3170
},
{
"epoch": 0.9085714285714286,
"grad_norm": 1.546875,
"learning_rate": 2.0693500580577232e-07,
"loss": 1.2415,
"step": 3180
},
{
"epoch": 0.9114285714285715,
"grad_norm": 1.5390625,
"learning_rate": 1.9432556905796153e-07,
"loss": 1.2711,
"step": 3190
},
{
"epoch": 0.9142857142857143,
"grad_norm": 1.5390625,
"learning_rate": 1.82104867621275e-07,
"loss": 1.2392,
"step": 3200
},
{
"epoch": 0.9171428571428571,
"grad_norm": 2.015625,
"learning_rate": 1.7027389003920315e-07,
"loss": 1.2394,
"step": 3210
},
{
"epoch": 0.92,
"grad_norm": 1.5625,
"learning_rate": 1.5883359333012438e-07,
"loss": 1.2219,
"step": 3220
},
{
"epoch": 0.9228571428571428,
"grad_norm": 1.4609375,
"learning_rate": 1.4778490290988934e-07,
"loss": 1.2376,
"step": 3230
},
{
"epoch": 0.9257142857142857,
"grad_norm": 1.6015625,
"learning_rate": 1.3712871251697012e-07,
"loss": 1.2682,
"step": 3240
},
{
"epoch": 0.9285714285714286,
"grad_norm": 1.6796875,
"learning_rate": 1.2686588414015544e-07,
"loss": 1.2103,
"step": 3250
},
{
"epoch": 0.9314285714285714,
"grad_norm": 1.7109375,
"learning_rate": 1.169972479488285e-07,
"loss": 1.2426,
"step": 3260
},
{
"epoch": 0.9342857142857143,
"grad_norm": 1.6796875,
"learning_rate": 1.0752360222581471e-07,
"loss": 1.2023,
"step": 3270
},
{
"epoch": 0.9371428571428572,
"grad_norm": 2.203125,
"learning_rate": 9.84457133028055e-08,
"loss": 1.2508,
"step": 3280
},
{
"epoch": 0.94,
"grad_norm": 1.4765625,
"learning_rate": 8.976431549837239e-08,
"loss": 1.2167,
"step": 3290
},
{
"epoch": 0.9428571428571428,
"grad_norm": 1.6953125,
"learning_rate": 8.148011105856168e-08,
"loss": 1.2454,
"step": 3300
},
{
"epoch": 0.9457142857142857,
"grad_norm": 1.734375,
"learning_rate": 7.359377010009383e-08,
"loss": 1.2451,
"step": 3310
},
{
"epoch": 0.9485714285714286,
"grad_norm": 1.4296875,
"learning_rate": 6.610593055615733e-08,
"loss": 1.2328,
"step": 3320
},
{
"epoch": 0.9514285714285714,
"grad_norm": 1.84375,
"learning_rate": 5.9017198124801664e-08,
"loss": 1.2635,
"step": 3330
},
{
"epoch": 0.9542857142857143,
"grad_norm": 1.5390625,
"learning_rate": 5.232814621994597e-08,
"loss": 1.2532,
"step": 3340
},
{
"epoch": 0.9571428571428572,
"grad_norm": 1.671875,
"learning_rate": 4.603931592499278e-08,
"loss": 1.2491,
"step": 3350
},
{
"epoch": 0.96,
"grad_norm": 1.5703125,
"learning_rate": 4.015121594906024e-08,
"loss": 1.2173,
"step": 3360
},
{
"epoch": 0.9628571428571429,
"grad_norm": 1.4140625,
"learning_rate": 3.4664322585831745e-08,
"loss": 1.2187,
"step": 3370
},
{
"epoch": 0.9657142857142857,
"grad_norm": 1.6875,
"learning_rate": 2.9579079675028377e-08,
"loss": 1.234,
"step": 3380
},
{
"epoch": 0.9685714285714285,
"grad_norm": 1.609375,
"learning_rate": 2.4895898566505982e-08,
"loss": 1.2441,
"step": 3390
},
{
"epoch": 0.9714285714285714,
"grad_norm": 1.4375,
"learning_rate": 2.0615158086981225e-08,
"loss": 1.2438,
"step": 3400
},
{
"epoch": 0.9742857142857143,
"grad_norm": 1.46875,
"learning_rate": 1.6737204509387206e-08,
"loss": 1.1841,
"step": 3410
},
{
"epoch": 0.9771428571428571,
"grad_norm": 1.5703125,
"learning_rate": 1.3262351524864215e-08,
"loss": 1.2387,
"step": 3420
},
{
"epoch": 0.98,
"grad_norm": 1.921875,
"learning_rate": 1.0190880217383348e-08,
"loss": 1.2228,
"step": 3430
},
{
"epoch": 0.9828571428571429,
"grad_norm": 1.4375,
"learning_rate": 7.523039041012481e-09,
"loss": 1.2125,
"step": 3440
},
{
"epoch": 0.9857142857142858,
"grad_norm": 1.59375,
"learning_rate": 5.259043799813457e-09,
"loss": 1.2386,
"step": 3450
},
{
"epoch": 0.9885714285714285,
"grad_norm": 1.546875,
"learning_rate": 3.3990776303910457e-09,
"loss": 1.2625,
"step": 3460
},
{
"epoch": 0.9914285714285714,
"grad_norm": 1.5859375,
"learning_rate": 1.9432909870747972e-09,
"loss": 1.2377,
"step": 3470
},
{
"epoch": 0.9942857142857143,
"grad_norm": 1.7421875,
"learning_rate": 8.918016297515541e-10,
"loss": 1.2364,
"step": 3480
},
{
"epoch": 0.9971428571428571,
"grad_norm": 1.625,
"learning_rate": 2.4469461433751507e-10,
"loss": 1.2174,
"step": 3490
},
{
"epoch": 1.0,
"grad_norm": 1.546875,
"learning_rate": 2.02228589873954e-12,
"loss": 1.2482,
"step": 3500
}
],
"logging_steps": 10,
"max_steps": 3500,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"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": 1.412338023206486e+19,
"train_batch_size": 28,
"trial_name": null,
"trial_params": null
}