joeyzero/Qwen3-4B-Reasoning-Backfill-v0.1
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
Files
2e64cb8d176847550dc613ff727d70ece861ff08
Qwen3-4B-Reasoning-Backfill…/checkpoint-20040/trainer_state.json
ModelHub XC 2e64cb8d17 初始化项目，由ModelHub XC社区提供模型 
Model: joeyzero/Qwen3-4B-Reasoning-Backfill-v0.1
Source: Original Platform
2026-05-25 04:07:19 +08:00

22079 lines
658 KiB
JSON
Raw Blame History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0011983223487118,
"eval_steps": 500,
"global_step": 20040,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0009986019572598363,
"grad_norm": 14.6875,
"learning_rate": 5.625e-06,
"loss": 1.2935,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10,
"tokens_per_second_per_gpu": 2433.25
},
{
"epoch": 0.0019972039145196726,
"grad_norm": 4.875,
"learning_rate": 1.1875e-05,
"loss": 0.854,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 20,
"tokens_per_second_per_gpu": 2456.09
},
{
"epoch": 0.0029958058717795086,
"grad_norm": 3.765625,
"learning_rate": 1.8125e-05,
"loss": 0.6748,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 30,
"tokens_per_second_per_gpu": 2727.46
},
{
"epoch": 0.003994407829039345,
"grad_norm": 5.1875,
"learning_rate": 2.4375e-05,
"loss": 0.6458,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 40,
"tokens_per_second_per_gpu": 2542.63
},
{
"epoch": 0.004993009786299181,
"grad_norm": 4.125,
"learning_rate": 2.4999996883431864e-05,
"loss": 0.6774,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 50,
"tokens_per_second_per_gpu": 2345.48
},
{
"epoch": 0.005991611743559017,
"grad_norm": 3.375,
"learning_rate": 2.499998611011191e-05,
"loss": 0.5765,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 60,
"tokens_per_second_per_gpu": 2522.77
},
{
"epoch": 0.006990213700818854,
"grad_norm": 3.6875,
"learning_rate": 2.4999967641570623e-05,
"loss": 0.6479,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 70,
"tokens_per_second_per_gpu": 2549.09
},
{
"epoch": 0.00798881565807869,
"grad_norm": 5.0625,
"learning_rate": 2.4999941477819366e-05,
"loss": 0.655,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 80,
"tokens_per_second_per_gpu": 2546.51
},
{
"epoch": 0.008987417615338526,
"grad_norm": 3.984375,
"learning_rate": 2.499990761887425e-05,
"loss": 0.5784,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 90,
"tokens_per_second_per_gpu": 2584.81
},
{
"epoch": 0.009986019572598362,
"grad_norm": 3.515625,
"learning_rate": 2.4999866064756117e-05,
"loss": 0.6179,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 100,
"tokens_per_second_per_gpu": 2478.64
},
{
"epoch": 0.010984621529858199,
"grad_norm": 4.03125,
"learning_rate": 2.499981681549055e-05,
"loss": 0.6117,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 110,
"tokens_per_second_per_gpu": 2465.76
},
{
"epoch": 0.011983223487118035,
"grad_norm": 4.34375,
"learning_rate": 2.4999759871107865e-05,
"loss": 0.6695,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 120,
"tokens_per_second_per_gpu": 2387.83
},
{
"epoch": 0.01298182544437787,
"grad_norm": 4.75,
"learning_rate": 2.4999695231643118e-05,
"loss": 0.6279,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 130,
"tokens_per_second_per_gpu": 2498.78
},
{
"epoch": 0.013980427401637708,
"grad_norm": 3.6875,
"learning_rate": 2.499962289713611e-05,
"loss": 0.6055,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 140,
"tokens_per_second_per_gpu": 2553.78
},
{
"epoch": 0.014979029358897543,
"grad_norm": 4.15625,
"learning_rate": 2.499954286763136e-05,
"loss": 0.6795,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 150,
"tokens_per_second_per_gpu": 2560.78
},
{
"epoch": 0.01597763131615738,
"grad_norm": 3.421875,
"learning_rate": 2.4999455143178143e-05,
"loss": 0.6555,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 160,
"tokens_per_second_per_gpu": 2449.63
},
{
"epoch": 0.016976233273417216,
"grad_norm": 4.1875,
"learning_rate": 2.499935972383046e-05,
"loss": 0.6666,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 170,
"tokens_per_second_per_gpu": 2483.81
},
{
"epoch": 0.017974835230677052,
"grad_norm": 4.28125,
"learning_rate": 2.499925660964706e-05,
"loss": 0.6499,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 180,
"tokens_per_second_per_gpu": 2676.57
},
{
"epoch": 0.018973437187936888,
"grad_norm": 3.34375,
"learning_rate": 2.4999145800691415e-05,
"loss": 0.5887,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 190,
"tokens_per_second_per_gpu": 2547.88
},
{
"epoch": 0.019972039145196723,
"grad_norm": 4.25,
"learning_rate": 2.4999027297031743e-05,
"loss": 0.6962,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 200,
"tokens_per_second_per_gpu": 2651.86
},
{
"epoch": 0.020970641102456562,
"grad_norm": 3.5,
"learning_rate": 2.4998901098740997e-05,
"loss": 0.6851,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 210,
"tokens_per_second_per_gpu": 2523.71
},
{
"epoch": 0.021969243059716398,
"grad_norm": 3.734375,
"learning_rate": 2.4998767205896865e-05,
"loss": 0.6518,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 220,
"tokens_per_second_per_gpu": 2453.45
},
{
"epoch": 0.022967845016976234,
"grad_norm": 3.53125,
"learning_rate": 2.499862561858178e-05,
"loss": 0.6525,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 230,
"tokens_per_second_per_gpu": 2649.66
},
{
"epoch": 0.02396644697423607,
"grad_norm": 5.21875,
"learning_rate": 2.49984763368829e-05,
"loss": 0.5986,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 240,
"tokens_per_second_per_gpu": 2473.23
},
{
"epoch": 0.024965048931495905,
"grad_norm": 3.359375,
"learning_rate": 2.499831936089213e-05,
"loss": 0.6158,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 250,
"tokens_per_second_per_gpu": 2570.46
},
{
"epoch": 0.02596365088875574,
"grad_norm": 4.65625,
"learning_rate": 2.49981546907061e-05,
"loss": 0.6437,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 260,
"tokens_per_second_per_gpu": 2503.01
},
{
"epoch": 0.02696225284601558,
"grad_norm": 3.5,
"learning_rate": 2.499798232642619e-05,
"loss": 0.6132,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 270,
"tokens_per_second_per_gpu": 2626.8
},
{
"epoch": 0.027960854803275415,
"grad_norm": 4.8125,
"learning_rate": 2.499780226815851e-05,
"loss": 0.6843,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 280,
"tokens_per_second_per_gpu": 2601.43
},
{
"epoch": 0.02895945676053525,
"grad_norm": 3.484375,
"learning_rate": 2.4997614516013902e-05,
"loss": 0.5671,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 290,
"tokens_per_second_per_gpu": 2545.89
},
{
"epoch": 0.029958058717795086,
"grad_norm": 3.46875,
"learning_rate": 2.499741907010796e-05,
"loss": 0.6735,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 300,
"tokens_per_second_per_gpu": 2372.26
},
{
"epoch": 0.030956660675054922,
"grad_norm": 4.0,
"learning_rate": 2.4997215930560997e-05,
"loss": 0.6259,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 310,
"tokens_per_second_per_gpu": 2587.61
},
{
"epoch": 0.03195526263231476,
"grad_norm": 3.625,
"learning_rate": 2.4997005097498068e-05,
"loss": 0.6604,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 320,
"tokens_per_second_per_gpu": 2526.99
},
{
"epoch": 0.0329538645895746,
"grad_norm": 3.984375,
"learning_rate": 2.499678657104897e-05,
"loss": 0.613,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 330,
"tokens_per_second_per_gpu": 2517.74
},
{
"epoch": 0.03395246654683443,
"grad_norm": 3.375,
"learning_rate": 2.499656035134823e-05,
"loss": 0.6758,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 340,
"tokens_per_second_per_gpu": 2381.37
},
{
"epoch": 0.03495106850409427,
"grad_norm": 3.484375,
"learning_rate": 2.499632643853511e-05,
"loss": 0.5786,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 350,
"tokens_per_second_per_gpu": 2478.29
},
{
"epoch": 0.035949670461354104,
"grad_norm": 3.671875,
"learning_rate": 2.4996084832753617e-05,
"loss": 0.7036,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 360,
"tokens_per_second_per_gpu": 2411.29
},
{
"epoch": 0.03694827241861394,
"grad_norm": 4.0625,
"learning_rate": 2.4995835534152486e-05,
"loss": 0.5763,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 370,
"tokens_per_second_per_gpu": 2577.58
},
{
"epoch": 0.037946874375873775,
"grad_norm": 3.75,
"learning_rate": 2.499557854288519e-05,
"loss": 0.6163,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 380,
"tokens_per_second_per_gpu": 2664.57
},
{
"epoch": 0.03894547633313361,
"grad_norm": 3.09375,
"learning_rate": 2.4995313859109933e-05,
"loss": 0.6304,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 390,
"tokens_per_second_per_gpu": 2496.57
},
{
"epoch": 0.039944078290393446,
"grad_norm": 3.796875,
"learning_rate": 2.4995041482989668e-05,
"loss": 0.6398,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 400,
"tokens_per_second_per_gpu": 2388.44
},
{
"epoch": 0.04094268024765328,
"grad_norm": 3.78125,
"learning_rate": 2.4994761414692064e-05,
"loss": 0.6036,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 410,
"tokens_per_second_per_gpu": 2590.92
},
{
"epoch": 0.041941282204913125,
"grad_norm": 3.234375,
"learning_rate": 2.4994473654389542e-05,
"loss": 0.5739,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 420,
"tokens_per_second_per_gpu": 2513.79
},
{
"epoch": 0.04293988416217296,
"grad_norm": 3.796875,
"learning_rate": 2.4994178202259255e-05,
"loss": 0.6383,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 430,
"tokens_per_second_per_gpu": 2555.06
},
{
"epoch": 0.043938486119432796,
"grad_norm": 2.625,
"learning_rate": 2.4993875058483085e-05,
"loss": 0.5553,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 440,
"tokens_per_second_per_gpu": 2507.61
},
{
"epoch": 0.04493708807669263,
"grad_norm": 3.28125,
"learning_rate": 2.4993564223247655e-05,
"loss": 0.5436,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 450,
"tokens_per_second_per_gpu": 2448.31
},
{
"epoch": 0.04593569003395247,
"grad_norm": 4.875,
"learning_rate": 2.4993245696744318e-05,
"loss": 0.5881,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 460,
"tokens_per_second_per_gpu": 2441.27
},
{
"epoch": 0.0469342919912123,
"grad_norm": 3.109375,
"learning_rate": 2.499291947916917e-05,
"loss": 0.6493,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 470,
"tokens_per_second_per_gpu": 2348.09
},
{
"epoch": 0.04793289394847214,
"grad_norm": 4.03125,
"learning_rate": 2.4992585570723025e-05,
"loss": 0.5154,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 480,
"tokens_per_second_per_gpu": 2521.61
},
{
"epoch": 0.048931495905731974,
"grad_norm": 3.015625,
"learning_rate": 2.499224397161146e-05,
"loss": 0.577,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 490,
"tokens_per_second_per_gpu": 2723.84
},
{
"epoch": 0.04993009786299181,
"grad_norm": 2.828125,
"learning_rate": 2.4991894682044757e-05,
"loss": 0.608,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 500,
"tokens_per_second_per_gpu": 2604.08
},
{
"epoch": 0.050928699820251645,
"grad_norm": 3.171875,
"learning_rate": 2.4991537702237948e-05,
"loss": 0.5497,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 510,
"tokens_per_second_per_gpu": 2482.06
},
{
"epoch": 0.05192730177751148,
"grad_norm": 3.640625,
"learning_rate": 2.4991173032410796e-05,
"loss": 0.5923,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 520,
"tokens_per_second_per_gpu": 2453.18
},
{
"epoch": 0.052925903734771323,
"grad_norm": 3.34375,
"learning_rate": 2.4990800672787805e-05,
"loss": 0.6137,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 530,
"tokens_per_second_per_gpu": 2533.46
},
{
"epoch": 0.05392450569203116,
"grad_norm": 2.734375,
"learning_rate": 2.4990420623598197e-05,
"loss": 0.5488,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 540,
"tokens_per_second_per_gpu": 2447.38
},
{
"epoch": 0.054923107649290995,
"grad_norm": 3.109375,
"learning_rate": 2.4990032885075945e-05,
"loss": 0.5871,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 550,
"tokens_per_second_per_gpu": 2593.13
},
{
"epoch": 0.05592170960655083,
"grad_norm": 3.34375,
"learning_rate": 2.498963745745974e-05,
"loss": 0.6611,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 560,
"tokens_per_second_per_gpu": 2345.76
},
{
"epoch": 0.056920311563810666,
"grad_norm": 4.875,
"learning_rate": 2.4989234340993026e-05,
"loss": 0.6844,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 570,
"tokens_per_second_per_gpu": 2451.35
},
{
"epoch": 0.0579189135210705,
"grad_norm": 3.65625,
"learning_rate": 2.4988823535923956e-05,
"loss": 0.6293,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 580,
"tokens_per_second_per_gpu": 2638.89
},
{
"epoch": 0.05891751547833034,
"grad_norm": 3.953125,
"learning_rate": 2.4988405042505434e-05,
"loss": 0.591,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 590,
"tokens_per_second_per_gpu": 2659.59
},
{
"epoch": 0.05991611743559017,
"grad_norm": 4.21875,
"learning_rate": 2.4987978860995098e-05,
"loss": 0.641,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 600,
"tokens_per_second_per_gpu": 2304.14
},
{
"epoch": 0.06091471939285001,
"grad_norm": 3.9375,
"learning_rate": 2.498754499165531e-05,
"loss": 0.6039,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 610,
"tokens_per_second_per_gpu": 2461.25
},
{
"epoch": 0.061913321350109844,
"grad_norm": 3.9375,
"learning_rate": 2.4987103434753163e-05,
"loss": 0.6365,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 620,
"tokens_per_second_per_gpu": 2578.47
},
{
"epoch": 0.06291192330736968,
"grad_norm": 4.03125,
"learning_rate": 2.498665419056049e-05,
"loss": 0.5842,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 630,
"tokens_per_second_per_gpu": 2607.5
},
{
"epoch": 0.06391052526462952,
"grad_norm": 2.859375,
"learning_rate": 2.4986197259353855e-05,
"loss": 0.532,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 640,
"tokens_per_second_per_gpu": 2560.27
},
{
"epoch": 0.06490912722188935,
"grad_norm": 2.921875,
"learning_rate": 2.4985732641414555e-05,
"loss": 0.5259,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 650,
"tokens_per_second_per_gpu": 2467.38
},
{
"epoch": 0.0659077291791492,
"grad_norm": 3.578125,
"learning_rate": 2.4985260337028613e-05,
"loss": 0.6415,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 660,
"tokens_per_second_per_gpu": 2394.48
},
{
"epoch": 0.06690633113640902,
"grad_norm": 3.765625,
"learning_rate": 2.4984780346486793e-05,
"loss": 0.6262,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 670,
"tokens_per_second_per_gpu": 2536.05
},
{
"epoch": 0.06790493309366886,
"grad_norm": 3.734375,
"learning_rate": 2.4984292670084582e-05,
"loss": 0.4979,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 680,
"tokens_per_second_per_gpu": 2580.76
},
{
"epoch": 0.0689035350509287,
"grad_norm": 4.1875,
"learning_rate": 2.4983797308122205e-05,
"loss": 0.6399,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 690,
"tokens_per_second_per_gpu": 2450.58
},
{
"epoch": 0.06990213700818854,
"grad_norm": 2.734375,
"learning_rate": 2.4983294260904615e-05,
"loss": 0.6073,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 700,
"tokens_per_second_per_gpu": 2511.56
},
{
"epoch": 0.07090073896544838,
"grad_norm": 3.203125,
"learning_rate": 2.49827835287415e-05,
"loss": 0.5878,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 710,
"tokens_per_second_per_gpu": 2516.19
},
{
"epoch": 0.07189934092270821,
"grad_norm": 3.734375,
"learning_rate": 2.4982265111947273e-05,
"loss": 0.6766,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 720,
"tokens_per_second_per_gpu": 2545.0
},
{
"epoch": 0.07289794287996805,
"grad_norm": 3.640625,
"learning_rate": 2.4981739010841078e-05,
"loss": 0.5988,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 730,
"tokens_per_second_per_gpu": 2567.94
},
{
"epoch": 0.07389654483722788,
"grad_norm": 3.328125,
"learning_rate": 2.49812052257468e-05,
"loss": 0.6392,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 740,
"tokens_per_second_per_gpu": 2595.01
},
{
"epoch": 0.07489514679448772,
"grad_norm": 3.96875,
"learning_rate": 2.4980663756993042e-05,
"loss": 0.6307,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 750,
"tokens_per_second_per_gpu": 2628.94
},
{
"epoch": 0.07589374875174755,
"grad_norm": 3.53125,
"learning_rate": 2.498011460491314e-05,
"loss": 0.5795,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 760,
"tokens_per_second_per_gpu": 2616.04
},
{
"epoch": 0.07689235070900739,
"grad_norm": 4.03125,
"learning_rate": 2.497955776984517e-05,
"loss": 0.5619,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 770,
"tokens_per_second_per_gpu": 2441.7
},
{
"epoch": 0.07789095266626722,
"grad_norm": 4.28125,
"learning_rate": 2.497899325213192e-05,
"loss": 0.5717,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 780,
"tokens_per_second_per_gpu": 2460.02
},
{
"epoch": 0.07888955462352706,
"grad_norm": 3.53125,
"learning_rate": 2.4978421052120928e-05,
"loss": 0.5271,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 790,
"tokens_per_second_per_gpu": 2416.47
},
{
"epoch": 0.07988815658078689,
"grad_norm": 3.03125,
"learning_rate": 2.497784117016444e-05,
"loss": 0.5435,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 800,
"tokens_per_second_per_gpu": 2378.95
},
{
"epoch": 0.08088675853804674,
"grad_norm": 3.03125,
"learning_rate": 2.497725360661945e-05,
"loss": 0.6304,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 810,
"tokens_per_second_per_gpu": 2548.34
},
{
"epoch": 0.08188536049530656,
"grad_norm": 3.671875,
"learning_rate": 2.497665836184767e-05,
"loss": 0.6798,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 820,
"tokens_per_second_per_gpu": 2410.81
},
{
"epoch": 0.0828839624525664,
"grad_norm": 3.3125,
"learning_rate": 2.4976055436215544e-05,
"loss": 0.6166,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 830,
"tokens_per_second_per_gpu": 2564.2
},
{
"epoch": 0.08388256440982625,
"grad_norm": 3.84375,
"learning_rate": 2.4975444830094245e-05,
"loss": 0.6335,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 840,
"tokens_per_second_per_gpu": 2407.54
},
{
"epoch": 0.08488116636708608,
"grad_norm": 3.28125,
"learning_rate": 2.4974826543859674e-05,
"loss": 0.5788,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 850,
"tokens_per_second_per_gpu": 2543.54
},
{
"epoch": 0.08587976832434592,
"grad_norm": 4.3125,
"learning_rate": 2.4974200577892452e-05,
"loss": 0.7068,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 860,
"tokens_per_second_per_gpu": 2471.1
},
{
"epoch": 0.08687837028160575,
"grad_norm": 3.40625,
"learning_rate": 2.4973566932577947e-05,
"loss": 0.6471,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 870,
"tokens_per_second_per_gpu": 2520.86
},
{
"epoch": 0.08787697223886559,
"grad_norm": 3.734375,
"learning_rate": 2.4972925608306233e-05,
"loss": 0.646,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 880,
"tokens_per_second_per_gpu": 2564.43
},
{
"epoch": 0.08887557419612542,
"grad_norm": 3.203125,
"learning_rate": 2.4972276605472126e-05,
"loss": 0.5778,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 890,
"tokens_per_second_per_gpu": 2537.1
},
{
"epoch": 0.08987417615338526,
"grad_norm": 3.078125,
"learning_rate": 2.4971619924475162e-05,
"loss": 0.5407,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 900,
"tokens_per_second_per_gpu": 2533.92
},
{
"epoch": 0.09087277811064509,
"grad_norm": 3.828125,
"learning_rate": 2.4970955565719606e-05,
"loss": 0.6092,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 910,
"tokens_per_second_per_gpu": 2553.55
},
{
"epoch": 0.09187138006790493,
"grad_norm": 3.375,
"learning_rate": 2.4970283529614452e-05,
"loss": 0.6348,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 920,
"tokens_per_second_per_gpu": 2478.03
},
{
"epoch": 0.09286998202516476,
"grad_norm": 2.546875,
"learning_rate": 2.4969603816573416e-05,
"loss": 0.5222,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 930,
"tokens_per_second_per_gpu": 2625.45
},
{
"epoch": 0.0938685839824246,
"grad_norm": 3.625,
"learning_rate": 2.496891642701494e-05,
"loss": 0.5492,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 940,
"tokens_per_second_per_gpu": 2548.76
},
{
"epoch": 0.09486718593968445,
"grad_norm": 4.40625,
"learning_rate": 2.49682213613622e-05,
"loss": 0.5474,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 950,
"tokens_per_second_per_gpu": 2533.71
},
{
"epoch": 0.09586578789694428,
"grad_norm": 3.203125,
"learning_rate": 2.496751862004308e-05,
"loss": 0.6221,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 960,
"tokens_per_second_per_gpu": 2528.55
},
{
"epoch": 0.09686438985420412,
"grad_norm": 3.265625,
"learning_rate": 2.496680820349021e-05,
"loss": 0.5566,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 970,
"tokens_per_second_per_gpu": 2739.19
},
{
"epoch": 0.09786299181146395,
"grad_norm": 4.25,
"learning_rate": 2.4966090112140932e-05,
"loss": 0.5967,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 980,
"tokens_per_second_per_gpu": 2651.78
},
{
"epoch": 0.09886159376872379,
"grad_norm": 2.828125,
"learning_rate": 2.4965364346437317e-05,
"loss": 0.5078,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 990,
"tokens_per_second_per_gpu": 2628.16
},
{
"epoch": 0.09986019572598362,
"grad_norm": 3.6875,
"learning_rate": 2.496463090682616e-05,
"loss": 0.5847,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1000,
"tokens_per_second_per_gpu": 2624.44
},
{
"epoch": 0.10085879768324346,
"grad_norm": 4.03125,
"learning_rate": 2.496388979375898e-05,
"loss": 0.6145,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1010,
"tokens_per_second_per_gpu": 2494.73
},
{
"epoch": 0.10185739964050329,
"grad_norm": 3.734375,
"learning_rate": 2.4963141007692022e-05,
"loss": 0.6686,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1020,
"tokens_per_second_per_gpu": 2569.75
},
{
"epoch": 0.10285600159776313,
"grad_norm": 2.828125,
"learning_rate": 2.496238454908624e-05,
"loss": 0.534,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1030,
"tokens_per_second_per_gpu": 2503.62
},
{
"epoch": 0.10385460355502296,
"grad_norm": 4.09375,
"learning_rate": 2.4961620418407343e-05,
"loss": 0.6204,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1040,
"tokens_per_second_per_gpu": 2576.86
},
{
"epoch": 0.1048532055122828,
"grad_norm": 3.984375,
"learning_rate": 2.4960848616125733e-05,
"loss": 0.622,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1050,
"tokens_per_second_per_gpu": 2616.67
},
{
"epoch": 0.10585180746954265,
"grad_norm": 3.0,
"learning_rate": 2.4960069142716544e-05,
"loss": 0.5688,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1060,
"tokens_per_second_per_gpu": 2562.4
},
{
"epoch": 0.10685040942680248,
"grad_norm": 2.984375,
"learning_rate": 2.495928199865964e-05,
"loss": 0.5633,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1070,
"tokens_per_second_per_gpu": 2546.41
},
{
"epoch": 0.10784901138406232,
"grad_norm": 3.9375,
"learning_rate": 2.4958487184439597e-05,
"loss": 0.4996,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1080,
"tokens_per_second_per_gpu": 2592.55
},
{
"epoch": 0.10884761334132215,
"grad_norm": 4.28125,
"learning_rate": 2.495768470054572e-05,
"loss": 0.6219,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1090,
"tokens_per_second_per_gpu": 2584.93
},
{
"epoch": 0.10984621529858199,
"grad_norm": 2.921875,
"learning_rate": 2.495687454747203e-05,
"loss": 0.5217,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1100,
"tokens_per_second_per_gpu": 2649.48
},
{
"epoch": 0.11084481725584182,
"grad_norm": 2.8125,
"learning_rate": 2.4956056725717276e-05,
"loss": 0.5152,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1110,
"tokens_per_second_per_gpu": 2571.54
},
{
"epoch": 0.11184341921310166,
"grad_norm": 3.453125,
"learning_rate": 2.4955231235784922e-05,
"loss": 0.5394,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1120,
"tokens_per_second_per_gpu": 2618.3
},
{
"epoch": 0.11284202117036149,
"grad_norm": 2.5,
"learning_rate": 2.4954398078183154e-05,
"loss": 0.5768,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1130,
"tokens_per_second_per_gpu": 2456.04
},
{
"epoch": 0.11384062312762133,
"grad_norm": 3.71875,
"learning_rate": 2.4953557253424882e-05,
"loss": 0.5728,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1140,
"tokens_per_second_per_gpu": 2504.75
},
{
"epoch": 0.11483922508488116,
"grad_norm": 3.546875,
"learning_rate": 2.4952708762027727e-05,
"loss": 0.5992,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1150,
"tokens_per_second_per_gpu": 2646.76
},
{
"epoch": 0.115837827042141,
"grad_norm": 3.453125,
"learning_rate": 2.495185260451404e-05,
"loss": 0.6201,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1160,
"tokens_per_second_per_gpu": 2543.13
},
{
"epoch": 0.11683642899940083,
"grad_norm": 2.828125,
"learning_rate": 2.4950988781410892e-05,
"loss": 0.6482,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1170,
"tokens_per_second_per_gpu": 2634.88
},
{
"epoch": 0.11783503095666067,
"grad_norm": 2.796875,
"learning_rate": 2.495011729325006e-05,
"loss": 0.5553,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1180,
"tokens_per_second_per_gpu": 2636.98
},
{
"epoch": 0.11883363291392052,
"grad_norm": 3.125,
"learning_rate": 2.4949238140568054e-05,
"loss": 0.5984,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1190,
"tokens_per_second_per_gpu": 2433.71
},
{
"epoch": 0.11983223487118035,
"grad_norm": 3.03125,
"learning_rate": 2.4948351323906093e-05,
"loss": 0.5565,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1200,
"tokens_per_second_per_gpu": 2425.42
},
{
"epoch": 0.12083083682844019,
"grad_norm": 3.875,
"learning_rate": 2.494745684381012e-05,
"loss": 0.6125,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1210,
"tokens_per_second_per_gpu": 2579.97
},
{
"epoch": 0.12182943878570002,
"grad_norm": 3.8125,
"learning_rate": 2.494655470083079e-05,
"loss": 0.5549,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1220,
"tokens_per_second_per_gpu": 2587.4
},
{
"epoch": 0.12282804074295986,
"grad_norm": 4.53125,
"learning_rate": 2.4945644895523487e-05,
"loss": 0.5994,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1230,
"tokens_per_second_per_gpu": 2518.03
},
{
"epoch": 0.12382664270021969,
"grad_norm": 3.390625,
"learning_rate": 2.4944727428448295e-05,
"loss": 0.5928,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1240,
"tokens_per_second_per_gpu": 2612.22
},
{
"epoch": 0.12482524465747953,
"grad_norm": 2.65625,
"learning_rate": 2.4943802300170027e-05,
"loss": 0.587,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1250,
"tokens_per_second_per_gpu": 2553.1
},
{
"epoch": 0.12582384661473936,
"grad_norm": 3.671875,
"learning_rate": 2.494286951125821e-05,
"loss": 0.6287,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1260,
"tokens_per_second_per_gpu": 2586.72
},
{
"epoch": 0.1268224485719992,
"grad_norm": 2.5,
"learning_rate": 2.494192906228708e-05,
"loss": 0.5679,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1270,
"tokens_per_second_per_gpu": 2514.0
},
{
"epoch": 0.12782105052925904,
"grad_norm": 4.53125,
"learning_rate": 2.4940980953835602e-05,
"loss": 0.6084,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1280,
"tokens_per_second_per_gpu": 2554.42
},
{
"epoch": 0.1288196524865189,
"grad_norm": 3.5625,
"learning_rate": 2.494002518648745e-05,
"loss": 0.5308,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1290,
"tokens_per_second_per_gpu": 2339.2
},
{
"epoch": 0.1298182544437787,
"grad_norm": 3.375,
"learning_rate": 2.4939061760831007e-05,
"loss": 0.6598,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1300,
"tokens_per_second_per_gpu": 2681.19
},
{
"epoch": 0.13081685640103854,
"grad_norm": 2.65625,
"learning_rate": 2.4938090677459374e-05,
"loss": 0.5446,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1310,
"tokens_per_second_per_gpu": 2655.93
},
{
"epoch": 0.1318154583582984,
"grad_norm": 3.109375,
"learning_rate": 2.493711193697037e-05,
"loss": 0.6038,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1320,
"tokens_per_second_per_gpu": 2547.55
},
{
"epoch": 0.13281406031555823,
"grad_norm": 3.359375,
"learning_rate": 2.493612553996653e-05,
"loss": 0.5614,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1330,
"tokens_per_second_per_gpu": 2681.78
},
{
"epoch": 0.13381266227281804,
"grad_norm": 3.234375,
"learning_rate": 2.4935131487055094e-05,
"loss": 0.6167,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1340,
"tokens_per_second_per_gpu": 2502.16
},
{
"epoch": 0.1348112642300779,
"grad_norm": 3.578125,
"learning_rate": 2.493412977884802e-05,
"loss": 0.5644,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1350,
"tokens_per_second_per_gpu": 2587.13
},
{
"epoch": 0.13580986618733773,
"grad_norm": 3.75,
"learning_rate": 2.4933120415961975e-05,
"loss": 0.6031,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1360,
"tokens_per_second_per_gpu": 2491.03
},
{
"epoch": 0.13680846814459757,
"grad_norm": 3.75,
"learning_rate": 2.4932103399018346e-05,
"loss": 0.646,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1370,
"tokens_per_second_per_gpu": 2662.49
},
{
"epoch": 0.1378070701018574,
"grad_norm": 3.5,
"learning_rate": 2.4931078728643226e-05,
"loss": 0.534,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1380,
"tokens_per_second_per_gpu": 2355.97
},
{
"epoch": 0.13880567205911723,
"grad_norm": 3.03125,
"learning_rate": 2.493004640546742e-05,
"loss": 0.5726,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1390,
"tokens_per_second_per_gpu": 2610.61
},
{
"epoch": 0.13980427401637707,
"grad_norm": 2.828125,
"learning_rate": 2.4929006430126445e-05,
"loss": 0.5112,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1400,
"tokens_per_second_per_gpu": 2532.1
},
{
"epoch": 0.14080287597363692,
"grad_norm": 3.9375,
"learning_rate": 2.492795880326053e-05,
"loss": 0.5487,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1410,
"tokens_per_second_per_gpu": 2326.36
},
{
"epoch": 0.14180147793089676,
"grad_norm": 3.5625,
"learning_rate": 2.492690352551461e-05,
"loss": 0.5554,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1420,
"tokens_per_second_per_gpu": 2442.66
},
{
"epoch": 0.14280007988815657,
"grad_norm": 2.984375,
"learning_rate": 2.4925840597538343e-05,
"loss": 0.5526,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1430,
"tokens_per_second_per_gpu": 2556.92
},
{
"epoch": 0.14379868184541642,
"grad_norm": 3.96875,
"learning_rate": 2.4924770019986075e-05,
"loss": 0.6488,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1440,
"tokens_per_second_per_gpu": 2374.16
},
{
"epoch": 0.14479728380267626,
"grad_norm": 3.0,
"learning_rate": 2.492369179351688e-05,
"loss": 0.5764,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1450,
"tokens_per_second_per_gpu": 2619.86
},
{
"epoch": 0.1457958857599361,
"grad_norm": 3.5625,
"learning_rate": 2.4922605918794532e-05,
"loss": 0.5998,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1460,
"tokens_per_second_per_gpu": 2452.72
},
{
"epoch": 0.14679448771719592,
"grad_norm": 4.4375,
"learning_rate": 2.4921512396487515e-05,
"loss": 0.5821,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1470,
"tokens_per_second_per_gpu": 2671.47
},
{
"epoch": 0.14779308967445576,
"grad_norm": 3.890625,
"learning_rate": 2.4920411227269026e-05,
"loss": 0.5923,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1480,
"tokens_per_second_per_gpu": 2448.97
},
{
"epoch": 0.1487916916317156,
"grad_norm": 3.75,
"learning_rate": 2.4919302411816956e-05,
"loss": 0.6361,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1490,
"tokens_per_second_per_gpu": 2632.79
},
{
"epoch": 0.14979029358897544,
"grad_norm": 3.640625,
"learning_rate": 2.491818595081392e-05,
"loss": 0.5488,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1500,
"tokens_per_second_per_gpu": 2393.86
},
{
"epoch": 0.15078889554623526,
"grad_norm": 3.671875,
"learning_rate": 2.4917061844947233e-05,
"loss": 0.6485,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1510,
"tokens_per_second_per_gpu": 2451.56
},
{
"epoch": 0.1517874975034951,
"grad_norm": 2.9375,
"learning_rate": 2.491593009490891e-05,
"loss": 0.6288,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1520,
"tokens_per_second_per_gpu": 2417.47
},
{
"epoch": 0.15278609946075494,
"grad_norm": 3.203125,
"learning_rate": 2.4914790701395674e-05,
"loss": 0.5701,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1530,
"tokens_per_second_per_gpu": 2478.94
},
{
"epoch": 0.15378470141801479,
"grad_norm": 3.53125,
"learning_rate": 2.4913643665108965e-05,
"loss": 0.5149,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1540,
"tokens_per_second_per_gpu": 2434.93
},
{
"epoch": 0.15478330337527463,
"grad_norm": 3.59375,
"learning_rate": 2.4912488986754918e-05,
"loss": 0.5279,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1550,
"tokens_per_second_per_gpu": 2531.95
},
{
"epoch": 0.15578190533253444,
"grad_norm": 3.375,
"learning_rate": 2.4911326667044373e-05,
"loss": 0.5147,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1560,
"tokens_per_second_per_gpu": 2601.2
},
{
"epoch": 0.15678050728979429,
"grad_norm": 3.609375,
"learning_rate": 2.491015670669287e-05,
"loss": 0.5212,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1570,
"tokens_per_second_per_gpu": 2487.32
},
{
"epoch": 0.15777910924705413,
"grad_norm": 3.609375,
"learning_rate": 2.4908979106420665e-05,
"loss": 0.5706,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1580,
"tokens_per_second_per_gpu": 2673.55
},
{
"epoch": 0.15877771120431397,
"grad_norm": 3.28125,
"learning_rate": 2.4907793866952712e-05,
"loss": 0.5486,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1590,
"tokens_per_second_per_gpu": 2404.47
},
{
"epoch": 0.15977631316157379,
"grad_norm": 3.390625,
"learning_rate": 2.4906600989018657e-05,
"loss": 0.5771,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1600,
"tokens_per_second_per_gpu": 2514.08
},
{
"epoch": 0.16077491511883363,
"grad_norm": 3.21875,
"learning_rate": 2.4905400473352864e-05,
"loss": 0.5454,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1610,
"tokens_per_second_per_gpu": 2519.39
},
{
"epoch": 0.16177351707609347,
"grad_norm": 2.703125,
"learning_rate": 2.4904192320694395e-05,
"loss": 0.5374,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1620,
"tokens_per_second_per_gpu": 2497.08
},
{
"epoch": 0.1627721190333533,
"grad_norm": 3.421875,
"learning_rate": 2.4902976531787003e-05,
"loss": 0.6485,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1630,
"tokens_per_second_per_gpu": 2578.48
},
{
"epoch": 0.16377072099061313,
"grad_norm": 3.703125,
"learning_rate": 2.4901753107379157e-05,
"loss": 0.6222,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1640,
"tokens_per_second_per_gpu": 2447.14
},
{
"epoch": 0.16476932294787297,
"grad_norm": 3.65625,
"learning_rate": 2.490052204822402e-05,
"loss": 0.5996,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1650,
"tokens_per_second_per_gpu": 2496.6
},
{
"epoch": 0.1657679249051328,
"grad_norm": 2.4375,
"learning_rate": 2.489928335507945e-05,
"loss": 0.5879,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1660,
"tokens_per_second_per_gpu": 2662.82
},
{
"epoch": 0.16676652686239266,
"grad_norm": 3.1875,
"learning_rate": 2.489803702870801e-05,
"loss": 0.6386,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1670,
"tokens_per_second_per_gpu": 2614.24
},
{
"epoch": 0.1677651288196525,
"grad_norm": 2.9375,
"learning_rate": 2.489678306987696e-05,
"loss": 0.5964,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1680,
"tokens_per_second_per_gpu": 2560.69
},
{
"epoch": 0.1687637307769123,
"grad_norm": 3.015625,
"learning_rate": 2.489552147935827e-05,
"loss": 0.5447,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1690,
"tokens_per_second_per_gpu": 2481.65
},
{
"epoch": 0.16976233273417216,
"grad_norm": 3.0625,
"learning_rate": 2.4894252257928585e-05,
"loss": 0.6053,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1700,
"tokens_per_second_per_gpu": 2515.55
},
{
"epoch": 0.170760934691432,
"grad_norm": 2.1875,
"learning_rate": 2.489297540636927e-05,
"loss": 0.4786,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1710,
"tokens_per_second_per_gpu": 2383.9
},
{
"epoch": 0.17175953664869184,
"grad_norm": 3.625,
"learning_rate": 2.4891690925466372e-05,
"loss": 0.6433,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1720,
"tokens_per_second_per_gpu": 2585.35
},
{
"epoch": 0.17275813860595166,
"grad_norm": 3.359375,
"learning_rate": 2.4890398816010646e-05,
"loss": 0.5149,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1730,
"tokens_per_second_per_gpu": 2446.56
},
{
"epoch": 0.1737567405632115,
"grad_norm": 3.984375,
"learning_rate": 2.4889099078797536e-05,
"loss": 0.6383,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1740,
"tokens_per_second_per_gpu": 2516.58
},
{
"epoch": 0.17475534252047134,
"grad_norm": 3.5,
"learning_rate": 2.4887791714627187e-05,
"loss": 0.5973,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1750,
"tokens_per_second_per_gpu": 2361.41
},
{
"epoch": 0.17575394447773118,
"grad_norm": 3.015625,
"learning_rate": 2.4886476724304433e-05,
"loss": 0.5347,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1760,
"tokens_per_second_per_gpu": 2347.78
},
{
"epoch": 0.17675254643499103,
"grad_norm": 3.625,
"learning_rate": 2.4885154108638807e-05,
"loss": 0.5402,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1770,
"tokens_per_second_per_gpu": 2549.88
},
{
"epoch": 0.17775114839225084,
"grad_norm": 3.40625,
"learning_rate": 2.4883823868444538e-05,
"loss": 0.6241,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1780,
"tokens_per_second_per_gpu": 2497.43
},
{
"epoch": 0.17874975034951068,
"grad_norm": 2.984375,
"learning_rate": 2.4882486004540547e-05,
"loss": 0.5148,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1790,
"tokens_per_second_per_gpu": 2737.7
},
{
"epoch": 0.17974835230677053,
"grad_norm": 3.75,
"learning_rate": 2.4881140517750438e-05,
"loss": 0.5942,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1800,
"tokens_per_second_per_gpu": 2394.88
},
{
"epoch": 0.18074695426403037,
"grad_norm": 2.71875,
"learning_rate": 2.4879787408902526e-05,
"loss": 0.5466,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1810,
"tokens_per_second_per_gpu": 2383.25
},
{
"epoch": 0.18174555622129018,
"grad_norm": 3.921875,
"learning_rate": 2.4878426678829812e-05,
"loss": 0.6014,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1820,
"tokens_per_second_per_gpu": 2401.83
},
{
"epoch": 0.18274415817855003,
"grad_norm": 3.859375,
"learning_rate": 2.4877058328369982e-05,
"loss": 0.4986,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1830,
"tokens_per_second_per_gpu": 2600.69
},
{
"epoch": 0.18374276013580987,
"grad_norm": 4.15625,
"learning_rate": 2.4875682358365414e-05,
"loss": 0.6434,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1840,
"tokens_per_second_per_gpu": 2398.51
},
{
"epoch": 0.1847413620930697,
"grad_norm": 3.515625,
"learning_rate": 2.487429876966319e-05,
"loss": 0.5426,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1850,
"tokens_per_second_per_gpu": 2544.1
},
{
"epoch": 0.18573996405032953,
"grad_norm": 4.1875,
"learning_rate": 2.4872907563115065e-05,
"loss": 0.6276,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1860,
"tokens_per_second_per_gpu": 2450.1
},
{
"epoch": 0.18673856600758937,
"grad_norm": 3.625,
"learning_rate": 2.4871508739577493e-05,
"loss": 0.5667,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1870,
"tokens_per_second_per_gpu": 2557.67
},
{
"epoch": 0.1877371679648492,
"grad_norm": 3.84375,
"learning_rate": 2.487010229991162e-05,
"loss": 0.5737,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1880,
"tokens_per_second_per_gpu": 2396.57
},
{
"epoch": 0.18873576992210905,
"grad_norm": 3.21875,
"learning_rate": 2.486868824498327e-05,
"loss": 0.5918,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1890,
"tokens_per_second_per_gpu": 2489.78
},
{
"epoch": 0.1897343718793689,
"grad_norm": 3.15625,
"learning_rate": 2.486726657566296e-05,
"loss": 0.5501,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1900,
"tokens_per_second_per_gpu": 2461.71
},
{
"epoch": 0.1907329738366287,
"grad_norm": 2.859375,
"learning_rate": 2.48658372928259e-05,
"loss": 0.5414,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1910,
"tokens_per_second_per_gpu": 2426.67
},
{
"epoch": 0.19173157579388855,
"grad_norm": 2.828125,
"learning_rate": 2.4864400397351987e-05,
"loss": 0.6109,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1920,
"tokens_per_second_per_gpu": 2425.32
},
{
"epoch": 0.1927301777511484,
"grad_norm": 2.640625,
"learning_rate": 2.486295589012579e-05,
"loss": 0.5343,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1930,
"tokens_per_second_per_gpu": 2601.42
},
{
"epoch": 0.19372877970840824,
"grad_norm": 4.4375,
"learning_rate": 2.4861503772036583e-05,
"loss": 0.6402,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1940,
"tokens_per_second_per_gpu": 2407.81
},
{
"epoch": 0.19472738166566805,
"grad_norm": 2.75,
"learning_rate": 2.4860044043978316e-05,
"loss": 0.5227,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1950,
"tokens_per_second_per_gpu": 2604.35
},
{
"epoch": 0.1957259836229279,
"grad_norm": 3.640625,
"learning_rate": 2.485857670684962e-05,
"loss": 0.5519,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1960,
"tokens_per_second_per_gpu": 2780.31
},
{
"epoch": 0.19672458558018774,
"grad_norm": 4.15625,
"learning_rate": 2.485710176155381e-05,
"loss": 0.5742,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1970,
"tokens_per_second_per_gpu": 2571.81
},
{
"epoch": 0.19772318753744758,
"grad_norm": 2.375,
"learning_rate": 2.4855619208998903e-05,
"loss": 0.6111,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1980,
"tokens_per_second_per_gpu": 2562.79
},
{
"epoch": 0.1987217894947074,
"grad_norm": 3.0625,
"learning_rate": 2.4854129050097573e-05,
"loss": 0.5201,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 1990,
"tokens_per_second_per_gpu": 2591.12
},
{
"epoch": 0.19972039145196724,
"grad_norm": 2.953125,
"learning_rate": 2.48526312857672e-05,
"loss": 0.5301,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2000,
"tokens_per_second_per_gpu": 2482.81
},
{
"epoch": 0.20071899340922708,
"grad_norm": 2.90625,
"learning_rate": 2.485112591692983e-05,
"loss": 0.6424,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2010,
"tokens_per_second_per_gpu": 2546.7
},
{
"epoch": 0.20171759536648692,
"grad_norm": 3.46875,
"learning_rate": 2.4849612944512192e-05,
"loss": 0.6543,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2020,
"tokens_per_second_per_gpu": 2591.4
},
{
"epoch": 0.20271619732374677,
"grad_norm": 3.671875,
"learning_rate": 2.4848092369445704e-05,
"loss": 0.5653,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2030,
"tokens_per_second_per_gpu": 2582.03
},
{
"epoch": 0.20371479928100658,
"grad_norm": 3.6875,
"learning_rate": 2.484656419266646e-05,
"loss": 0.5981,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2040,
"tokens_per_second_per_gpu": 2614.81
},
{
"epoch": 0.20471340123826642,
"grad_norm": 3.078125,
"learning_rate": 2.4845028415115235e-05,
"loss": 0.5306,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2050,
"tokens_per_second_per_gpu": 2352.58
},
{
"epoch": 0.20571200319552627,
"grad_norm": 4.25,
"learning_rate": 2.4843485037737475e-05,
"loss": 0.5991,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2060,
"tokens_per_second_per_gpu": 2420.53
},
{
"epoch": 0.2067106051527861,
"grad_norm": 3.25,
"learning_rate": 2.4841934061483323e-05,
"loss": 0.5581,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2070,
"tokens_per_second_per_gpu": 2418.01
},
{
"epoch": 0.20770920711004592,
"grad_norm": 4.0,
"learning_rate": 2.4840375487307577e-05,
"loss": 0.5648,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2080,
"tokens_per_second_per_gpu": 2463.28
},
{
"epoch": 0.20870780906730577,
"grad_norm": 3.3125,
"learning_rate": 2.483880931616973e-05,
"loss": 0.5762,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2090,
"tokens_per_second_per_gpu": 2549.31
},
{
"epoch": 0.2097064110245656,
"grad_norm": 3.234375,
"learning_rate": 2.4837235549033945e-05,
"loss": 0.4791,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2100,
"tokens_per_second_per_gpu": 2445.48
},
{
"epoch": 0.21070501298182545,
"grad_norm": 2.921875,
"learning_rate": 2.4835654186869062e-05,
"loss": 0.5677,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2110,
"tokens_per_second_per_gpu": 2541.8
},
{
"epoch": 0.2117036149390853,
"grad_norm": 3.09375,
"learning_rate": 2.4834065230648597e-05,
"loss": 0.6072,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2120,
"tokens_per_second_per_gpu": 2517.96
},
{
"epoch": 0.2127022168963451,
"grad_norm": 3.171875,
"learning_rate": 2.483246868135074e-05,
"loss": 0.5429,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2130,
"tokens_per_second_per_gpu": 2587.14
},
{
"epoch": 0.21370081885360495,
"grad_norm": 3.28125,
"learning_rate": 2.4830864539958355e-05,
"loss": 0.5612,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2140,
"tokens_per_second_per_gpu": 2380.99
},
{
"epoch": 0.2146994208108648,
"grad_norm": 2.8125,
"learning_rate": 2.482925280745898e-05,
"loss": 0.6179,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2150,
"tokens_per_second_per_gpu": 2521.96
},
{
"epoch": 0.21569802276812464,
"grad_norm": 3.46875,
"learning_rate": 2.482763348484483e-05,
"loss": 0.5553,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2160,
"tokens_per_second_per_gpu": 2334.48
},
{
"epoch": 0.21669662472538445,
"grad_norm": 3.765625,
"learning_rate": 2.482600657311279e-05,
"loss": 0.5549,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2170,
"tokens_per_second_per_gpu": 2617.39
},
{
"epoch": 0.2176952266826443,
"grad_norm": 3.546875,
"learning_rate": 2.482437207326442e-05,
"loss": 0.6127,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2180,
"tokens_per_second_per_gpu": 2305.99
},
{
"epoch": 0.21869382863990414,
"grad_norm": 3.546875,
"learning_rate": 2.4822729986305938e-05,
"loss": 0.6552,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2190,
"tokens_per_second_per_gpu": 2493.51
},
{
"epoch": 0.21969243059716398,
"grad_norm": 4.0,
"learning_rate": 2.482108031324825e-05,
"loss": 0.5893,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2200,
"tokens_per_second_per_gpu": 2504.39
},
{
"epoch": 0.2206910325544238,
"grad_norm": 3.8125,
"learning_rate": 2.4819423055106925e-05,
"loss": 0.6117,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2210,
"tokens_per_second_per_gpu": 2556.14
},
{
"epoch": 0.22168963451168364,
"grad_norm": 3.5625,
"learning_rate": 2.48177582129022e-05,
"loss": 0.5668,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2220,
"tokens_per_second_per_gpu": 2562.64
},
{
"epoch": 0.22268823646894348,
"grad_norm": 3.515625,
"learning_rate": 2.4816085787658984e-05,
"loss": 0.5064,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2230,
"tokens_per_second_per_gpu": 2531.2
},
{
"epoch": 0.22368683842620332,
"grad_norm": 2.375,
"learning_rate": 2.4814405780406848e-05,
"loss": 0.5623,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2240,
"tokens_per_second_per_gpu": 2504.99
},
{
"epoch": 0.22468544038346316,
"grad_norm": 3.34375,
"learning_rate": 2.4812718192180042e-05,
"loss": 0.6047,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2250,
"tokens_per_second_per_gpu": 2426.11
},
{
"epoch": 0.22568404234072298,
"grad_norm": 3.546875,
"learning_rate": 2.4811023024017473e-05,
"loss": 0.5582,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2260,
"tokens_per_second_per_gpu": 2468.31
},
{
"epoch": 0.22668264429798282,
"grad_norm": 2.984375,
"learning_rate": 2.4809320276962722e-05,
"loss": 0.5309,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2270,
"tokens_per_second_per_gpu": 2504.65
},
{
"epoch": 0.22768124625524266,
"grad_norm": 2.5,
"learning_rate": 2.480760995206402e-05,
"loss": 0.5235,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2280,
"tokens_per_second_per_gpu": 2360.38
},
{
"epoch": 0.2286798482125025,
"grad_norm": 2.84375,
"learning_rate": 2.4805892050374287e-05,
"loss": 0.5618,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2290,
"tokens_per_second_per_gpu": 2366.07
},
{
"epoch": 0.22967845016976232,
"grad_norm": 3.34375,
"learning_rate": 2.480416657295109e-05,
"loss": 0.5765,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2300,
"tokens_per_second_per_gpu": 2477.47
},
{
"epoch": 0.23067705212702216,
"grad_norm": 3.234375,
"learning_rate": 2.480243352085666e-05,
"loss": 0.5606,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2310,
"tokens_per_second_per_gpu": 2415.0
},
{
"epoch": 0.231675654084282,
"grad_norm": 3.03125,
"learning_rate": 2.4800692895157906e-05,
"loss": 0.5314,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2320,
"tokens_per_second_per_gpu": 2506.95
},
{
"epoch": 0.23267425604154185,
"grad_norm": 3.171875,
"learning_rate": 2.479894469692638e-05,
"loss": 0.5777,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2330,
"tokens_per_second_per_gpu": 2500.61
},
{
"epoch": 0.23367285799880166,
"grad_norm": 4.78125,
"learning_rate": 2.479718892723831e-05,
"loss": 0.6,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2340,
"tokens_per_second_per_gpu": 2536.16
},
{
"epoch": 0.2346714599560615,
"grad_norm": 4.28125,
"learning_rate": 2.4795425587174574e-05,
"loss": 0.5407,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2350,
"tokens_per_second_per_gpu": 2492.73
},
{
"epoch": 0.23567006191332135,
"grad_norm": 3.21875,
"learning_rate": 2.479365467782073e-05,
"loss": 0.5298,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2360,
"tokens_per_second_per_gpu": 2617.39
},
{
"epoch": 0.2366686638705812,
"grad_norm": 2.859375,
"learning_rate": 2.4791876200266968e-05,
"loss": 0.5561,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2370,
"tokens_per_second_per_gpu": 2614.95
},
{
"epoch": 0.23766726582784103,
"grad_norm": 4.15625,
"learning_rate": 2.4790090155608156e-05,
"loss": 0.5084,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2380,
"tokens_per_second_per_gpu": 2559.15
},
{
"epoch": 0.23866586778510085,
"grad_norm": 3.4375,
"learning_rate": 2.4788296544943817e-05,
"loss": 0.5153,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2390,
"tokens_per_second_per_gpu": 2507.4
},
{
"epoch": 0.2396644697423607,
"grad_norm": 3.453125,
"learning_rate": 2.4786495369378133e-05,
"loss": 0.6032,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2400,
"tokens_per_second_per_gpu": 2505.49
},
{
"epoch": 0.24066307169962053,
"grad_norm": 2.625,
"learning_rate": 2.4784686630019937e-05,
"loss": 0.5533,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2410,
"tokens_per_second_per_gpu": 2484.77
},
{
"epoch": 0.24166167365688038,
"grad_norm": 2.875,
"learning_rate": 2.4782870327982725e-05,
"loss": 0.5331,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2420,
"tokens_per_second_per_gpu": 2469.52
},
{
"epoch": 0.2426602756141402,
"grad_norm": 2.828125,
"learning_rate": 2.4781046464384642e-05,
"loss": 0.5613,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2430,
"tokens_per_second_per_gpu": 2516.75
},
{
"epoch": 0.24365887757140003,
"grad_norm": 3.21875,
"learning_rate": 2.47792150403485e-05,
"loss": 0.5527,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2440,
"tokens_per_second_per_gpu": 2431.34
},
{
"epoch": 0.24465747952865988,
"grad_norm": 2.921875,
"learning_rate": 2.4777376057001745e-05,
"loss": 0.581,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2450,
"tokens_per_second_per_gpu": 2526.71
},
{
"epoch": 0.24565608148591972,
"grad_norm": 2.609375,
"learning_rate": 2.47755295154765e-05,
"loss": 0.6167,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2460,
"tokens_per_second_per_gpu": 2585.06
},
{
"epoch": 0.24665468344317956,
"grad_norm": 2.515625,
"learning_rate": 2.477367541690952e-05,
"loss": 0.5701,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2470,
"tokens_per_second_per_gpu": 2515.99
},
{
"epoch": 0.24765328540043938,
"grad_norm": 3.171875,
"learning_rate": 2.4771813762442235e-05,
"loss": 0.5737,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2480,
"tokens_per_second_per_gpu": 2562.7
},
{
"epoch": 0.24865188735769922,
"grad_norm": 3.640625,
"learning_rate": 2.4769944553220703e-05,
"loss": 0.6068,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2490,
"tokens_per_second_per_gpu": 2521.39
},
{
"epoch": 0.24965048931495906,
"grad_norm": 3.375,
"learning_rate": 2.4768067790395646e-05,
"loss": 0.6001,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2500,
"tokens_per_second_per_gpu": 2517.88
},
{
"epoch": 0.2506490912722189,
"grad_norm": 3.984375,
"learning_rate": 2.476618347512244e-05,
"loss": 0.569,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2510,
"tokens_per_second_per_gpu": 2493.93
},
{
"epoch": 0.2516476932294787,
"grad_norm": 3.265625,
"learning_rate": 2.4764291608561095e-05,
"loss": 0.5648,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2520,
"tokens_per_second_per_gpu": 2446.04
},
{
"epoch": 0.2526462951867386,
"grad_norm": 3.390625,
"learning_rate": 2.4762392191876283e-05,
"loss": 0.6216,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2530,
"tokens_per_second_per_gpu": 2451.36
},
{
"epoch": 0.2536448971439984,
"grad_norm": 3.953125,
"learning_rate": 2.476048522623732e-05,
"loss": 0.5765,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2540,
"tokens_per_second_per_gpu": 2344.68
},
{
"epoch": 0.2546434991012582,
"grad_norm": 3.1875,
"learning_rate": 2.475857071281817e-05,
"loss": 0.61,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2550,
"tokens_per_second_per_gpu": 2457.85
},
{
"epoch": 0.2556421010585181,
"grad_norm": 4.4375,
"learning_rate": 2.475664865279744e-05,
"loss": 0.5853,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2560,
"tokens_per_second_per_gpu": 2481.59
},
{
"epoch": 0.2566407030157779,
"grad_norm": 4.21875,
"learning_rate": 2.475471904735839e-05,
"loss": 0.681,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2570,
"tokens_per_second_per_gpu": 2360.86
},
{
"epoch": 0.2576393049730378,
"grad_norm": 4.46875,
"learning_rate": 2.4752781897688915e-05,
"loss": 0.6177,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2580,
"tokens_per_second_per_gpu": 2472.65
},
{
"epoch": 0.2586379069302976,
"grad_norm": 2.53125,
"learning_rate": 2.4750837204981565e-05,
"loss": 0.6154,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2590,
"tokens_per_second_per_gpu": 2493.07
},
{
"epoch": 0.2596365088875574,
"grad_norm": 2.796875,
"learning_rate": 2.4748884970433524e-05,
"loss": 0.4886,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2600,
"tokens_per_second_per_gpu": 2411.8
},
{
"epoch": 0.2606351108448173,
"grad_norm": 2.59375,
"learning_rate": 2.4746925195246624e-05,
"loss": 0.5894,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2610,
"tokens_per_second_per_gpu": 2517.37
},
{
"epoch": 0.2616337128020771,
"grad_norm": 3.265625,
"learning_rate": 2.474495788062734e-05,
"loss": 0.579,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2620,
"tokens_per_second_per_gpu": 2522.44
},
{
"epoch": 0.2626323147593369,
"grad_norm": 2.75,
"learning_rate": 2.4742983027786794e-05,
"loss": 0.5023,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2630,
"tokens_per_second_per_gpu": 2416.93
},
{
"epoch": 0.2636309167165968,
"grad_norm": 3.28125,
"learning_rate": 2.4741000637940732e-05,
"loss": 0.5997,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2640,
"tokens_per_second_per_gpu": 2394.91
},
{
"epoch": 0.2646295186738566,
"grad_norm": 3.15625,
"learning_rate": 2.473901071230955e-05,
"loss": 0.5638,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2650,
"tokens_per_second_per_gpu": 2549.4
},
{
"epoch": 0.26562812063111646,
"grad_norm": 2.703125,
"learning_rate": 2.4737013252118285e-05,
"loss": 0.5516,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2660,
"tokens_per_second_per_gpu": 2653.01
},
{
"epoch": 0.2666267225883763,
"grad_norm": 3.515625,
"learning_rate": 2.4735008258596616e-05,
"loss": 0.579,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2670,
"tokens_per_second_per_gpu": 2433.46
},
{
"epoch": 0.2676253245456361,
"grad_norm": 4.03125,
"learning_rate": 2.4732995732978844e-05,
"loss": 0.5879,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2680,
"tokens_per_second_per_gpu": 2442.08
},
{
"epoch": 0.26862392650289596,
"grad_norm": 3.625,
"learning_rate": 2.4730975676503926e-05,
"loss": 0.5778,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2690,
"tokens_per_second_per_gpu": 2478.57
},
{
"epoch": 0.2696225284601558,
"grad_norm": 3.578125,
"learning_rate": 2.472894809041544e-05,
"loss": 0.5456,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2700,
"tokens_per_second_per_gpu": 2513.98
},
{
"epoch": 0.27062113041741565,
"grad_norm": 2.78125,
"learning_rate": 2.4726912975961602e-05,
"loss": 0.5063,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2710,
"tokens_per_second_per_gpu": 2464.05
},
{
"epoch": 0.27161973237467546,
"grad_norm": 2.953125,
"learning_rate": 2.4724870334395274e-05,
"loss": 0.5402,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2720,
"tokens_per_second_per_gpu": 2574.62
},
{
"epoch": 0.2726183343319353,
"grad_norm": 3.0625,
"learning_rate": 2.4722820166973943e-05,
"loss": 0.5773,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2730,
"tokens_per_second_per_gpu": 2676.57
},
{
"epoch": 0.27361693628919515,
"grad_norm": 3.265625,
"learning_rate": 2.472076247495972e-05,
"loss": 0.5996,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2740,
"tokens_per_second_per_gpu": 2632.42
},
{
"epoch": 0.27461553824645496,
"grad_norm": 3.8125,
"learning_rate": 2.4718697259619372e-05,
"loss": 0.5912,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2750,
"tokens_per_second_per_gpu": 2505.39
},
{
"epoch": 0.2756141402037148,
"grad_norm": 4.0,
"learning_rate": 2.4716624522224274e-05,
"loss": 0.5692,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2760,
"tokens_per_second_per_gpu": 2358.23
},
{
"epoch": 0.27661274216097465,
"grad_norm": 2.859375,
"learning_rate": 2.471454426405044e-05,
"loss": 0.5161,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2770,
"tokens_per_second_per_gpu": 2587.35
},
{
"epoch": 0.27761134411823446,
"grad_norm": 3.015625,
"learning_rate": 2.4712456486378526e-05,
"loss": 0.4979,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2780,
"tokens_per_second_per_gpu": 2474.05
},
{
"epoch": 0.27860994607549433,
"grad_norm": 3.5,
"learning_rate": 2.4710361190493796e-05,
"loss": 0.5293,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2790,
"tokens_per_second_per_gpu": 2613.3
},
{
"epoch": 0.27960854803275415,
"grad_norm": 3.65625,
"learning_rate": 2.4708258377686157e-05,
"loss": 0.5265,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2800,
"tokens_per_second_per_gpu": 2237.67
},
{
"epoch": 0.28060714999001396,
"grad_norm": 3.640625,
"learning_rate": 2.470614804925014e-05,
"loss": 0.5403,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2810,
"tokens_per_second_per_gpu": 2387.78
},
{
"epoch": 0.28160575194727383,
"grad_norm": 3.390625,
"learning_rate": 2.4704030206484896e-05,
"loss": 0.5103,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2820,
"tokens_per_second_per_gpu": 2400.98
},
{
"epoch": 0.28260435390453365,
"grad_norm": 2.90625,
"learning_rate": 2.4701904850694216e-05,
"loss": 0.5899,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2830,
"tokens_per_second_per_gpu": 2402.21
},
{
"epoch": 0.2836029558617935,
"grad_norm": 3.25,
"learning_rate": 2.4699771983186505e-05,
"loss": 0.5662,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2840,
"tokens_per_second_per_gpu": 2212.44
},
{
"epoch": 0.28460155781905333,
"grad_norm": 3.0625,
"learning_rate": 2.469763160527479e-05,
"loss": 0.529,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2850,
"tokens_per_second_per_gpu": 2409.89
},
{
"epoch": 0.28560015977631314,
"grad_norm": 3.40625,
"learning_rate": 2.4695483718276734e-05,
"loss": 0.4944,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2860,
"tokens_per_second_per_gpu": 2545.09
},
{
"epoch": 0.286598761733573,
"grad_norm": 2.921875,
"learning_rate": 2.4693328323514615e-05,
"loss": 0.5462,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2870,
"tokens_per_second_per_gpu": 2623.18
},
{
"epoch": 0.28759736369083283,
"grad_norm": 3.546875,
"learning_rate": 2.4691165422315333e-05,
"loss": 0.5415,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2880,
"tokens_per_second_per_gpu": 2528.76
},
{
"epoch": 0.28859596564809264,
"grad_norm": 2.671875,
"learning_rate": 2.4688995016010407e-05,
"loss": 0.5014,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2890,
"tokens_per_second_per_gpu": 2326.85
},
{
"epoch": 0.2895945676053525,
"grad_norm": 4.28125,
"learning_rate": 2.4686817105935983e-05,
"loss": 0.628,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2900,
"tokens_per_second_per_gpu": 2523.03
},
{
"epoch": 0.29059316956261233,
"grad_norm": 3.640625,
"learning_rate": 2.4684631693432818e-05,
"loss": 0.6168,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2910,
"tokens_per_second_per_gpu": 2497.34
},
{
"epoch": 0.2915917715198722,
"grad_norm": 4.125,
"learning_rate": 2.4682438779846294e-05,
"loss": 0.56,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2920,
"tokens_per_second_per_gpu": 2362.11
},
{
"epoch": 0.292590373477132,
"grad_norm": 2.875,
"learning_rate": 2.4680238366526415e-05,
"loss": 0.5157,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2930,
"tokens_per_second_per_gpu": 2361.71
},
{
"epoch": 0.29358897543439183,
"grad_norm": 3.421875,
"learning_rate": 2.4678030454827787e-05,
"loss": 0.5326,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2940,
"tokens_per_second_per_gpu": 2318.25
},
{
"epoch": 0.2945875773916517,
"grad_norm": 3.84375,
"learning_rate": 2.4675815046109644e-05,
"loss": 0.6094,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2950,
"tokens_per_second_per_gpu": 2476.4
},
{
"epoch": 0.2955861793489115,
"grad_norm": 3.671875,
"learning_rate": 2.4673592141735834e-05,
"loss": 0.6058,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2960,
"tokens_per_second_per_gpu": 2447.05
},
{
"epoch": 0.2965847813061714,
"grad_norm": 3.375,
"learning_rate": 2.4671361743074817e-05,
"loss": 0.5322,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2970,
"tokens_per_second_per_gpu": 2349.01
},
{
"epoch": 0.2975833832634312,
"grad_norm": 2.75,
"learning_rate": 2.4669123851499664e-05,
"loss": 0.5738,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2980,
"tokens_per_second_per_gpu": 2490.05
},
{
"epoch": 0.298581985220691,
"grad_norm": 2.453125,
"learning_rate": 2.466687846838807e-05,
"loss": 0.5614,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 2990,
"tokens_per_second_per_gpu": 2518.59
},
{
"epoch": 0.2995805871779509,
"grad_norm": 2.984375,
"learning_rate": 2.4664625595122326e-05,
"loss": 0.4831,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3000,
"tokens_per_second_per_gpu": 2548.96
},
{
"epoch": 0.3005791891352107,
"grad_norm": 3.203125,
"learning_rate": 2.4662365233089346e-05,
"loss": 0.5516,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3010,
"tokens_per_second_per_gpu": 2410.0
},
{
"epoch": 0.3015777910924705,
"grad_norm": 3.75,
"learning_rate": 2.4660097383680653e-05,
"loss": 0.5786,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3020,
"tokens_per_second_per_gpu": 2585.21
},
{
"epoch": 0.3025763930497304,
"grad_norm": 2.8125,
"learning_rate": 2.4657822048292374e-05,
"loss": 0.5573,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3030,
"tokens_per_second_per_gpu": 2489.88
},
{
"epoch": 0.3035749950069902,
"grad_norm": 4.28125,
"learning_rate": 2.4655539228325244e-05,
"loss": 0.5643,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3040,
"tokens_per_second_per_gpu": 2352.79
},
{
"epoch": 0.30457359696425007,
"grad_norm": 2.5625,
"learning_rate": 2.4653248925184613e-05,
"loss": 0.557,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3050,
"tokens_per_second_per_gpu": 2514.11
},
{
"epoch": 0.3055721989215099,
"grad_norm": 3.0625,
"learning_rate": 2.4650951140280438e-05,
"loss": 0.5692,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3060,
"tokens_per_second_per_gpu": 2443.75
},
{
"epoch": 0.3065708008787697,
"grad_norm": 3.578125,
"learning_rate": 2.4648645875027264e-05,
"loss": 0.5302,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3070,
"tokens_per_second_per_gpu": 2522.86
},
{
"epoch": 0.30756940283602957,
"grad_norm": 3.8125,
"learning_rate": 2.4646333130844268e-05,
"loss": 0.6062,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3080,
"tokens_per_second_per_gpu": 2471.63
},
{
"epoch": 0.3085680047932894,
"grad_norm": 3.625,
"learning_rate": 2.4644012909155214e-05,
"loss": 0.5842,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3090,
"tokens_per_second_per_gpu": 2333.61
},
{
"epoch": 0.30956660675054926,
"grad_norm": 2.59375,
"learning_rate": 2.464168521138847e-05,
"loss": 0.5587,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3100,
"tokens_per_second_per_gpu": 2481.82
},
{
"epoch": 0.31056520870780907,
"grad_norm": 2.84375,
"learning_rate": 2.4639350038977017e-05,
"loss": 0.5658,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3110,
"tokens_per_second_per_gpu": 2511.81
},
{
"epoch": 0.3115638106650689,
"grad_norm": 4.4375,
"learning_rate": 2.463700739335842e-05,
"loss": 0.6054,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3120,
"tokens_per_second_per_gpu": 2457.39
},
{
"epoch": 0.31256241262232876,
"grad_norm": 2.609375,
"learning_rate": 2.463465727597486e-05,
"loss": 0.6113,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3130,
"tokens_per_second_per_gpu": 2327.09
},
{
"epoch": 0.31356101457958857,
"grad_norm": 2.609375,
"learning_rate": 2.4632299688273113e-05,
"loss": 0.5818,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3140,
"tokens_per_second_per_gpu": 2621.31
},
{
"epoch": 0.3145596165368484,
"grad_norm": 3.9375,
"learning_rate": 2.4629934631704554e-05,
"loss": 0.5405,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3150,
"tokens_per_second_per_gpu": 2577.83
},
{
"epoch": 0.31555821849410826,
"grad_norm": 2.625,
"learning_rate": 2.462756210772515e-05,
"loss": 0.5581,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3160,
"tokens_per_second_per_gpu": 2464.7
},
{
"epoch": 0.31655682045136807,
"grad_norm": 2.96875,
"learning_rate": 2.462518211779548e-05,
"loss": 0.6155,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3170,
"tokens_per_second_per_gpu": 2631.32
},
{
"epoch": 0.31755542240862794,
"grad_norm": 3.625,
"learning_rate": 2.4622794663380698e-05,
"loss": 0.6188,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3180,
"tokens_per_second_per_gpu": 2417.53
},
{
"epoch": 0.31855402436588776,
"grad_norm": 3.578125,
"learning_rate": 2.4620399745950573e-05,
"loss": 0.6294,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3190,
"tokens_per_second_per_gpu": 2398.46
},
{
"epoch": 0.31955262632314757,
"grad_norm": 3.390625,
"learning_rate": 2.461799736697946e-05,
"loss": 0.5619,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3200,
"tokens_per_second_per_gpu": 2374.25
},
{
"epoch": 0.32055122828040744,
"grad_norm": 2.546875,
"learning_rate": 2.46155875279463e-05,
"loss": 0.5456,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3210,
"tokens_per_second_per_gpu": 2546.56
},
{
"epoch": 0.32154983023766726,
"grad_norm": 4.03125,
"learning_rate": 2.4613170230334647e-05,
"loss": 0.546,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3220,
"tokens_per_second_per_gpu": 2627.21
},
{
"epoch": 0.3225484321949271,
"grad_norm": 3.5625,
"learning_rate": 2.461074547563262e-05,
"loss": 0.5568,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3230,
"tokens_per_second_per_gpu": 2561.46
},
{
"epoch": 0.32354703415218694,
"grad_norm": 3.265625,
"learning_rate": 2.4608313265332954e-05,
"loss": 0.5806,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3240,
"tokens_per_second_per_gpu": 2709.13
},
{
"epoch": 0.32454563610944676,
"grad_norm": 2.71875,
"learning_rate": 2.4605873600932956e-05,
"loss": 0.5327,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3250,
"tokens_per_second_per_gpu": 2512.87
},
{
"epoch": 0.3255442380667066,
"grad_norm": 3.734375,
"learning_rate": 2.4603426483934527e-05,
"loss": 0.606,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3260,
"tokens_per_second_per_gpu": 2635.98
},
{
"epoch": 0.32654284002396644,
"grad_norm": 3.4375,
"learning_rate": 2.4600971915844164e-05,
"loss": 0.5971,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3270,
"tokens_per_second_per_gpu": 2532.57
},
{
"epoch": 0.32754144198122626,
"grad_norm": 2.609375,
"learning_rate": 2.459850989817294e-05,
"loss": 0.6051,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3280,
"tokens_per_second_per_gpu": 2443.84
},
{
"epoch": 0.3285400439384861,
"grad_norm": 2.3125,
"learning_rate": 2.4596040432436516e-05,
"loss": 0.5303,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3290,
"tokens_per_second_per_gpu": 2570.41
},
{
"epoch": 0.32953864589574594,
"grad_norm": 2.890625,
"learning_rate": 2.4593563520155144e-05,
"loss": 0.5508,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3300,
"tokens_per_second_per_gpu": 2198.31
},
{
"epoch": 0.3305372478530058,
"grad_norm": 3.015625,
"learning_rate": 2.459107916285366e-05,
"loss": 0.5428,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3310,
"tokens_per_second_per_gpu": 2444.5
},
{
"epoch": 0.3315358498102656,
"grad_norm": 3.265625,
"learning_rate": 2.458858736206147e-05,
"loss": 0.5136,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3320,
"tokens_per_second_per_gpu": 2268.29
},
{
"epoch": 0.33253445176752544,
"grad_norm": 3.46875,
"learning_rate": 2.4586088119312582e-05,
"loss": 0.5684,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3330,
"tokens_per_second_per_gpu": 2519.81
},
{
"epoch": 0.3335330537247853,
"grad_norm": 2.796875,
"learning_rate": 2.458358143614557e-05,
"loss": 0.5652,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3340,
"tokens_per_second_per_gpu": 2499.16
},
{
"epoch": 0.3345316556820451,
"grad_norm": 3.171875,
"learning_rate": 2.4581067314103602e-05,
"loss": 0.5297,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3350,
"tokens_per_second_per_gpu": 2536.3
},
{
"epoch": 0.335530257639305,
"grad_norm": 3.0625,
"learning_rate": 2.4578545754734405e-05,
"loss": 0.4888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3360,
"tokens_per_second_per_gpu": 2358.22
},
{
"epoch": 0.3365288595965648,
"grad_norm": 2.53125,
"learning_rate": 2.4576016759590308e-05,
"loss": 0.5815,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3370,
"tokens_per_second_per_gpu": 2465.07
},
{
"epoch": 0.3375274615538246,
"grad_norm": 3.328125,
"learning_rate": 2.4573480330228205e-05,
"loss": 0.5625,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3380,
"tokens_per_second_per_gpu": 2604.7
},
{
"epoch": 0.3385260635110845,
"grad_norm": 3.171875,
"learning_rate": 2.4570936468209565e-05,
"loss": 0.5208,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3390,
"tokens_per_second_per_gpu": 2298.91
},
{
"epoch": 0.3395246654683443,
"grad_norm": 3.796875,
"learning_rate": 2.456838517510044e-05,
"loss": 0.5497,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3400,
"tokens_per_second_per_gpu": 2480.21
},
{
"epoch": 0.3405232674256042,
"grad_norm": 3.140625,
"learning_rate": 2.4565826452471447e-05,
"loss": 0.5188,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3410,
"tokens_per_second_per_gpu": 2538.9
},
{
"epoch": 0.341521869382864,
"grad_norm": 3.578125,
"learning_rate": 2.456326030189779e-05,
"loss": 0.5889,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3420,
"tokens_per_second_per_gpu": 2385.96
},
{
"epoch": 0.3425204713401238,
"grad_norm": 2.5625,
"learning_rate": 2.4560686724959235e-05,
"loss": 0.5191,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3430,
"tokens_per_second_per_gpu": 2397.06
},
{
"epoch": 0.3435190732973837,
"grad_norm": 2.578125,
"learning_rate": 2.4558105723240127e-05,
"loss": 0.5515,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3440,
"tokens_per_second_per_gpu": 2525.31
},
{
"epoch": 0.3445176752546435,
"grad_norm": 2.96875,
"learning_rate": 2.4555517298329373e-05,
"loss": 0.6177,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3450,
"tokens_per_second_per_gpu": 2382.04
},
{
"epoch": 0.3455162772119033,
"grad_norm": 3.015625,
"learning_rate": 2.4552921451820453e-05,
"loss": 0.5418,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3460,
"tokens_per_second_per_gpu": 2343.09
},
{
"epoch": 0.3465148791691632,
"grad_norm": 3.140625,
"learning_rate": 2.4550318185311426e-05,
"loss": 0.5588,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3470,
"tokens_per_second_per_gpu": 2374.68
},
{
"epoch": 0.347513481126423,
"grad_norm": 3.09375,
"learning_rate": 2.4547707500404905e-05,
"loss": 0.6188,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3480,
"tokens_per_second_per_gpu": 2373.24
},
{
"epoch": 0.34851208308368287,
"grad_norm": 2.640625,
"learning_rate": 2.4545089398708078e-05,
"loss": 0.5147,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3490,
"tokens_per_second_per_gpu": 2367.93
},
{
"epoch": 0.3495106850409427,
"grad_norm": 3.609375,
"learning_rate": 2.4542463881832696e-05,
"loss": 0.5253,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3500,
"tokens_per_second_per_gpu": 2692.22
},
{
"epoch": 0.3505092869982025,
"grad_norm": 2.640625,
"learning_rate": 2.453983095139507e-05,
"loss": 0.5915,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3510,
"tokens_per_second_per_gpu": 2574.64
},
{
"epoch": 0.35150788895546237,
"grad_norm": 2.953125,
"learning_rate": 2.453719060901609e-05,
"loss": 0.563,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3520,
"tokens_per_second_per_gpu": 2438.66
},
{
"epoch": 0.3525064909127222,
"grad_norm": 3.140625,
"learning_rate": 2.4534542856321197e-05,
"loss": 0.5872,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3530,
"tokens_per_second_per_gpu": 2399.74
},
{
"epoch": 0.35350509286998205,
"grad_norm": 3.609375,
"learning_rate": 2.453188769494039e-05,
"loss": 0.5894,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3540,
"tokens_per_second_per_gpu": 2545.87
},
{
"epoch": 0.35450369482724187,
"grad_norm": 3.53125,
"learning_rate": 2.4529225126508244e-05,
"loss": 0.6227,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3550,
"tokens_per_second_per_gpu": 2504.27
},
{
"epoch": 0.3555022967845017,
"grad_norm": 2.828125,
"learning_rate": 2.4526555152663876e-05,
"loss": 0.5584,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3560,
"tokens_per_second_per_gpu": 2573.94
},
{
"epoch": 0.35650089874176155,
"grad_norm": 3.515625,
"learning_rate": 2.4523877775050974e-05,
"loss": 0.5686,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3570,
"tokens_per_second_per_gpu": 2386.54
},
{
"epoch": 0.35749950069902137,
"grad_norm": 3.5,
"learning_rate": 2.4521192995317784e-05,
"loss": 0.5596,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3580,
"tokens_per_second_per_gpu": 2471.82
},
{
"epoch": 0.3584981026562812,
"grad_norm": 2.96875,
"learning_rate": 2.4518500815117107e-05,
"loss": 0.5635,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3590,
"tokens_per_second_per_gpu": 2419.2
},
{
"epoch": 0.35949670461354105,
"grad_norm": 2.90625,
"learning_rate": 2.4515801236106297e-05,
"loss": 0.5983,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3600,
"tokens_per_second_per_gpu": 2501.18
},
{
"epoch": 0.36049530657080087,
"grad_norm": 3.203125,
"learning_rate": 2.4513094259947257e-05,
"loss": 0.5698,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3610,
"tokens_per_second_per_gpu": 2637.18
},
{
"epoch": 0.36149390852806074,
"grad_norm": 3.578125,
"learning_rate": 2.4510379888306458e-05,
"loss": 0.5449,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3620,
"tokens_per_second_per_gpu": 2418.55
},
{
"epoch": 0.36249251048532055,
"grad_norm": 3.90625,
"learning_rate": 2.450765812285492e-05,
"loss": 0.5834,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3630,
"tokens_per_second_per_gpu": 2663.24
},
{
"epoch": 0.36349111244258037,
"grad_norm": 3.6875,
"learning_rate": 2.4504928965268206e-05,
"loss": 0.5551,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3640,
"tokens_per_second_per_gpu": 2536.7
},
{
"epoch": 0.36448971439984024,
"grad_norm": 3.265625,
"learning_rate": 2.450219241722644e-05,
"loss": 0.5185,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3650,
"tokens_per_second_per_gpu": 2353.01
},
{
"epoch": 0.36548831635710005,
"grad_norm": 3.1875,
"learning_rate": 2.4499448480414288e-05,
"loss": 0.5041,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3660,
"tokens_per_second_per_gpu": 2539.95
},
{
"epoch": 0.3664869183143599,
"grad_norm": 3.25,
"learning_rate": 2.449669715652097e-05,
"loss": 0.5492,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3670,
"tokens_per_second_per_gpu": 2663.58
},
{
"epoch": 0.36748552027161974,
"grad_norm": 2.359375,
"learning_rate": 2.449393844724025e-05,
"loss": 0.4435,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3680,
"tokens_per_second_per_gpu": 2529.78
},
{
"epoch": 0.36848412222887955,
"grad_norm": 3.46875,
"learning_rate": 2.4491172354270443e-05,
"loss": 0.4711,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3690,
"tokens_per_second_per_gpu": 2588.91
},
{
"epoch": 0.3694827241861394,
"grad_norm": 2.96875,
"learning_rate": 2.4488398879314405e-05,
"loss": 0.5515,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3700,
"tokens_per_second_per_gpu": 2383.03
},
{
"epoch": 0.37048132614339924,
"grad_norm": 2.765625,
"learning_rate": 2.4485618024079538e-05,
"loss": 0.5084,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3710,
"tokens_per_second_per_gpu": 2404.34
},
{
"epoch": 0.37147992810065905,
"grad_norm": 2.96875,
"learning_rate": 2.4482829790277784e-05,
"loss": 0.6272,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3720,
"tokens_per_second_per_gpu": 2450.42
},
{
"epoch": 0.3724785300579189,
"grad_norm": 3.203125,
"learning_rate": 2.4480034179625638e-05,
"loss": 0.5742,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3730,
"tokens_per_second_per_gpu": 2284.42
},
{
"epoch": 0.37347713201517874,
"grad_norm": 3.265625,
"learning_rate": 2.4477231193844125e-05,
"loss": 0.5531,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3740,
"tokens_per_second_per_gpu": 2592.06
},
{
"epoch": 0.3744757339724386,
"grad_norm": 2.453125,
"learning_rate": 2.4474420834658814e-05,
"loss": 0.6117,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3750,
"tokens_per_second_per_gpu": 2480.02
},
{
"epoch": 0.3754743359296984,
"grad_norm": 2.78125,
"learning_rate": 2.4471603103799818e-05,
"loss": 0.5261,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3760,
"tokens_per_second_per_gpu": 2634.12
},
{
"epoch": 0.37647293788695824,
"grad_norm": 3.328125,
"learning_rate": 2.446877800300178e-05,
"loss": 0.623,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3770,
"tokens_per_second_per_gpu": 2470.14
},
{
"epoch": 0.3774715398442181,
"grad_norm": 2.96875,
"learning_rate": 2.4465945534003887e-05,
"loss": 0.5036,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3780,
"tokens_per_second_per_gpu": 2491.46
},
{
"epoch": 0.3784701418014779,
"grad_norm": 2.890625,
"learning_rate": 2.4463105698549854e-05,
"loss": 0.5372,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3790,
"tokens_per_second_per_gpu": 2375.22
},
{
"epoch": 0.3794687437587378,
"grad_norm": 2.5625,
"learning_rate": 2.446025849838794e-05,
"loss": 0.5727,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3800,
"tokens_per_second_per_gpu": 2521.78
},
{
"epoch": 0.3804673457159976,
"grad_norm": 2.84375,
"learning_rate": 2.4457403935270933e-05,
"loss": 0.5809,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3810,
"tokens_per_second_per_gpu": 2429.02
},
{
"epoch": 0.3814659476732574,
"grad_norm": 3.90625,
"learning_rate": 2.4454542010956157e-05,
"loss": 0.5229,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3820,
"tokens_per_second_per_gpu": 2310.42
},
{
"epoch": 0.3824645496305173,
"grad_norm": 3.71875,
"learning_rate": 2.4451672727205465e-05,
"loss": 0.5871,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3830,
"tokens_per_second_per_gpu": 2553.33
},
{
"epoch": 0.3834631515877771,
"grad_norm": 3.890625,
"learning_rate": 2.4448796085785235e-05,
"loss": 0.513,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3840,
"tokens_per_second_per_gpu": 2377.86
},
{
"epoch": 0.3844617535450369,
"grad_norm": 3.21875,
"learning_rate": 2.4445912088466383e-05,
"loss": 0.55,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3850,
"tokens_per_second_per_gpu": 2443.88
},
{
"epoch": 0.3854603555022968,
"grad_norm": 2.46875,
"learning_rate": 2.444302073702435e-05,
"loss": 0.5702,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3860,
"tokens_per_second_per_gpu": 2648.06
},
{
"epoch": 0.3864589574595566,
"grad_norm": 3.46875,
"learning_rate": 2.444012203323911e-05,
"loss": 0.65,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3870,
"tokens_per_second_per_gpu": 2429.26
},
{
"epoch": 0.3874575594168165,
"grad_norm": 3.5,
"learning_rate": 2.4437215978895155e-05,
"loss": 0.5737,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3880,
"tokens_per_second_per_gpu": 2426.38
},
{
"epoch": 0.3884561613740763,
"grad_norm": 3.453125,
"learning_rate": 2.4434302575781508e-05,
"loss": 0.487,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3890,
"tokens_per_second_per_gpu": 2520.43
},
{
"epoch": 0.3894547633313361,
"grad_norm": 3.46875,
"learning_rate": 2.4431381825691707e-05,
"loss": 0.5012,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3900,
"tokens_per_second_per_gpu": 2466.26
},
{
"epoch": 0.390453365288596,
"grad_norm": 2.328125,
"learning_rate": 2.442845373042383e-05,
"loss": 0.5715,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3910,
"tokens_per_second_per_gpu": 2606.85
},
{
"epoch": 0.3914519672458558,
"grad_norm": 3.40625,
"learning_rate": 2.4425518291780452e-05,
"loss": 0.5278,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3920,
"tokens_per_second_per_gpu": 2637.98
},
{
"epoch": 0.39245056920311566,
"grad_norm": 3.65625,
"learning_rate": 2.4422575511568697e-05,
"loss": 0.5401,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3930,
"tokens_per_second_per_gpu": 2521.92
},
{
"epoch": 0.3934491711603755,
"grad_norm": 2.75,
"learning_rate": 2.441962539160019e-05,
"loss": 0.4578,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3940,
"tokens_per_second_per_gpu": 2370.27
},
{
"epoch": 0.3944477731176353,
"grad_norm": 3.6875,
"learning_rate": 2.4416667933691075e-05,
"loss": 0.5931,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3950,
"tokens_per_second_per_gpu": 2578.85
},
{
"epoch": 0.39544637507489516,
"grad_norm": 3.3125,
"learning_rate": 2.4413703139662016e-05,
"loss": 0.5504,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3960,
"tokens_per_second_per_gpu": 2464.95
},
{
"epoch": 0.396444977032155,
"grad_norm": 3.609375,
"learning_rate": 2.4410731011338205e-05,
"loss": 0.5592,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3970,
"tokens_per_second_per_gpu": 2423.95
},
{
"epoch": 0.3974435789894148,
"grad_norm": 3.59375,
"learning_rate": 2.4407751550549328e-05,
"loss": 0.5736,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3980,
"tokens_per_second_per_gpu": 2372.9
},
{
"epoch": 0.39844218094667466,
"grad_norm": 3.5625,
"learning_rate": 2.4404764759129605e-05,
"loss": 0.6565,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 3990,
"tokens_per_second_per_gpu": 2473.87
},
{
"epoch": 0.3994407829039345,
"grad_norm": 2.53125,
"learning_rate": 2.4401770638917754e-05,
"loss": 0.5468,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4000,
"tokens_per_second_per_gpu": 2597.7
},
{
"epoch": 0.40043938486119435,
"grad_norm": 3.28125,
"learning_rate": 2.4398769191757013e-05,
"loss": 0.5964,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4010,
"tokens_per_second_per_gpu": 2362.27
},
{
"epoch": 0.40143798681845416,
"grad_norm": 4.34375,
"learning_rate": 2.4395760419495125e-05,
"loss": 0.5754,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4020,
"tokens_per_second_per_gpu": 2248.82
},
{
"epoch": 0.402436588775714,
"grad_norm": 2.78125,
"learning_rate": 2.439274432398435e-05,
"loss": 0.5514,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4030,
"tokens_per_second_per_gpu": 2507.58
},
{
"epoch": 0.40343519073297385,
"grad_norm": 2.8125,
"learning_rate": 2.4389720907081447e-05,
"loss": 0.6159,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4040,
"tokens_per_second_per_gpu": 2335.41
},
{
"epoch": 0.40443379269023366,
"grad_norm": 3.078125,
"learning_rate": 2.438669017064769e-05,
"loss": 0.5052,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4050,
"tokens_per_second_per_gpu": 2467.85
},
{
"epoch": 0.40543239464749353,
"grad_norm": 3.46875,
"learning_rate": 2.4383652116548857e-05,
"loss": 0.5493,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4060,
"tokens_per_second_per_gpu": 2605.98
},
{
"epoch": 0.40643099660475335,
"grad_norm": 2.90625,
"learning_rate": 2.4380606746655228e-05,
"loss": 0.5805,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4070,
"tokens_per_second_per_gpu": 2334.34
},
{
"epoch": 0.40742959856201316,
"grad_norm": 3.125,
"learning_rate": 2.437755406284159e-05,
"loss": 0.5583,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4080,
"tokens_per_second_per_gpu": 2321.28
},
{
"epoch": 0.40842820051927303,
"grad_norm": 3.90625,
"learning_rate": 2.437449406698723e-05,
"loss": 0.6618,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4090,
"tokens_per_second_per_gpu": 2549.02
},
{
"epoch": 0.40942680247653285,
"grad_norm": 4.28125,
"learning_rate": 2.4371426760975934e-05,
"loss": 0.5474,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4100,
"tokens_per_second_per_gpu": 2536.58
},
{
"epoch": 0.4104254044337927,
"grad_norm": 2.90625,
"learning_rate": 2.4368352146696004e-05,
"loss": 0.6462,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4110,
"tokens_per_second_per_gpu": 2479.24
},
{
"epoch": 0.41142400639105253,
"grad_norm": 2.640625,
"learning_rate": 2.4365270226040215e-05,
"loss": 0.5162,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4120,
"tokens_per_second_per_gpu": 2587.72
},
{
"epoch": 0.41242260834831235,
"grad_norm": 3.3125,
"learning_rate": 2.4362181000905864e-05,
"loss": 0.5806,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4130,
"tokens_per_second_per_gpu": 2513.64
},
{
"epoch": 0.4134212103055722,
"grad_norm": 2.546875,
"learning_rate": 2.435908447319473e-05,
"loss": 0.5458,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4140,
"tokens_per_second_per_gpu": 2371.01
},
{
"epoch": 0.41441981226283203,
"grad_norm": 2.78125,
"learning_rate": 2.435598064481309e-05,
"loss": 0.587,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4150,
"tokens_per_second_per_gpu": 2443.3
},
{
"epoch": 0.41541841422009185,
"grad_norm": 2.609375,
"learning_rate": 2.4352869517671728e-05,
"loss": 0.5834,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4160,
"tokens_per_second_per_gpu": 2405.91
},
{
"epoch": 0.4164170161773517,
"grad_norm": 3.390625,
"learning_rate": 2.43497510936859e-05,
"loss": 0.5347,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4170,
"tokens_per_second_per_gpu": 2445.15
},
{
"epoch": 0.41741561813461153,
"grad_norm": 3.296875,
"learning_rate": 2.4346625374775372e-05,
"loss": 0.5869,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4180,
"tokens_per_second_per_gpu": 2506.63
},
{
"epoch": 0.4184142200918714,
"grad_norm": 2.671875,
"learning_rate": 2.4343492362864395e-05,
"loss": 0.4689,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4190,
"tokens_per_second_per_gpu": 2541.72
},
{
"epoch": 0.4194128220491312,
"grad_norm": 3.765625,
"learning_rate": 2.43403520598817e-05,
"loss": 0.5893,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4200,
"tokens_per_second_per_gpu": 2395.07
},
{
"epoch": 0.42041142400639103,
"grad_norm": 2.875,
"learning_rate": 2.433720446776052e-05,
"loss": 0.5039,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4210,
"tokens_per_second_per_gpu": 2406.24
},
{
"epoch": 0.4214100259636509,
"grad_norm": 2.734375,
"learning_rate": 2.4334049588438578e-05,
"loss": 0.5541,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4220,
"tokens_per_second_per_gpu": 2260.91
},
{
"epoch": 0.4224086279209107,
"grad_norm": 4.0,
"learning_rate": 2.433088742385806e-05,
"loss": 0.5895,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4230,
"tokens_per_second_per_gpu": 2364.37
},
{
"epoch": 0.4234072298781706,
"grad_norm": 3.171875,
"learning_rate": 2.432771797596567e-05,
"loss": 0.4817,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4240,
"tokens_per_second_per_gpu": 2583.05
},
{
"epoch": 0.4244058318354304,
"grad_norm": 3.21875,
"learning_rate": 2.432454124671257e-05,
"loss": 0.4883,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4250,
"tokens_per_second_per_gpu": 2388.78
},
{
"epoch": 0.4254044337926902,
"grad_norm": 2.890625,
"learning_rate": 2.4321357238054408e-05,
"loss": 0.5259,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4260,
"tokens_per_second_per_gpu": 2441.14
},
{
"epoch": 0.4264030357499501,
"grad_norm": 3.328125,
"learning_rate": 2.4318165951951325e-05,
"loss": 0.5294,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4270,
"tokens_per_second_per_gpu": 2360.56
},
{
"epoch": 0.4274016377072099,
"grad_norm": 2.75,
"learning_rate": 2.431496739036793e-05,
"loss": 0.5458,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4280,
"tokens_per_second_per_gpu": 2460.33
},
{
"epoch": 0.4284002396644697,
"grad_norm": 2.6875,
"learning_rate": 2.4311761555273326e-05,
"loss": 0.5095,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4290,
"tokens_per_second_per_gpu": 2250.36
},
{
"epoch": 0.4293988416217296,
"grad_norm": 3.328125,
"learning_rate": 2.4308548448641078e-05,
"loss": 0.5322,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4300,
"tokens_per_second_per_gpu": 2285.98
},
{
"epoch": 0.4303974435789894,
"grad_norm": 2.71875,
"learning_rate": 2.430532807244923e-05,
"loss": 0.551,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4310,
"tokens_per_second_per_gpu": 2347.4
},
{
"epoch": 0.4313960455362493,
"grad_norm": 2.421875,
"learning_rate": 2.4302100428680314e-05,
"loss": 0.4959,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4320,
"tokens_per_second_per_gpu": 2469.96
},
{
"epoch": 0.4323946474935091,
"grad_norm": 3.046875,
"learning_rate": 2.4298865519321317e-05,
"loss": 0.5482,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4330,
"tokens_per_second_per_gpu": 2174.4
},
{
"epoch": 0.4333932494507689,
"grad_norm": 3.625,
"learning_rate": 2.4295623346363713e-05,
"loss": 0.5648,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4340,
"tokens_per_second_per_gpu": 2500.3
},
{
"epoch": 0.4343918514080288,
"grad_norm": 3.734375,
"learning_rate": 2.4292373911803447e-05,
"loss": 0.5281,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4350,
"tokens_per_second_per_gpu": 2461.61
},
{
"epoch": 0.4353904533652886,
"grad_norm": 3.21875,
"learning_rate": 2.4289117217640922e-05,
"loss": 0.4731,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4360,
"tokens_per_second_per_gpu": 2301.74
},
{
"epoch": 0.43638905532254846,
"grad_norm": 3.15625,
"learning_rate": 2.428585326588103e-05,
"loss": 0.6016,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4370,
"tokens_per_second_per_gpu": 2434.79
},
{
"epoch": 0.4373876572798083,
"grad_norm": 3.484375,
"learning_rate": 2.4282582058533108e-05,
"loss": 0.5505,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4380,
"tokens_per_second_per_gpu": 2549.33
},
{
"epoch": 0.4383862592370681,
"grad_norm": 2.921875,
"learning_rate": 2.4279303597610982e-05,
"loss": 0.5635,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4390,
"tokens_per_second_per_gpu": 2540.72
},
{
"epoch": 0.43938486119432796,
"grad_norm": 3.359375,
"learning_rate": 2.4276017885132922e-05,
"loss": 0.5223,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4400,
"tokens_per_second_per_gpu": 2450.28
},
{
"epoch": 0.4403834631515878,
"grad_norm": 3.125,
"learning_rate": 2.427272492312168e-05,
"loss": 0.6017,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4410,
"tokens_per_second_per_gpu": 2663.91
},
{
"epoch": 0.4413820651088476,
"grad_norm": 3.1875,
"learning_rate": 2.426942471360447e-05,
"loss": 0.5381,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4420,
"tokens_per_second_per_gpu": 2389.82
},
{
"epoch": 0.44238066706610746,
"grad_norm": 3.390625,
"learning_rate": 2.4266117258612946e-05,
"loss": 0.4874,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4430,
"tokens_per_second_per_gpu": 2590.38
},
{
"epoch": 0.4433792690233673,
"grad_norm": 2.453125,
"learning_rate": 2.426280256018325e-05,
"loss": 0.553,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4440,
"tokens_per_second_per_gpu": 2351.79
},
{
"epoch": 0.44437787098062714,
"grad_norm": 3.0625,
"learning_rate": 2.425948062035597e-05,
"loss": 0.5712,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4450,
"tokens_per_second_per_gpu": 2433.89
},
{
"epoch": 0.44537647293788696,
"grad_norm": 3.828125,
"learning_rate": 2.425615144117615e-05,
"loss": 0.5696,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4460,
"tokens_per_second_per_gpu": 2414.88
},
{
"epoch": 0.4463750748951468,
"grad_norm": 2.859375,
"learning_rate": 2.4252815024693294e-05,
"loss": 0.5249,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4470,
"tokens_per_second_per_gpu": 2376.92
},
{
"epoch": 0.44737367685240664,
"grad_norm": 2.703125,
"learning_rate": 2.4249471372961362e-05,
"loss": 0.5048,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4480,
"tokens_per_second_per_gpu": 2420.72
},
{
"epoch": 0.44837227880966646,
"grad_norm": 3.484375,
"learning_rate": 2.424612048803877e-05,
"loss": 0.5668,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4490,
"tokens_per_second_per_gpu": 2544.45
},
{
"epoch": 0.44937088076692633,
"grad_norm": 2.859375,
"learning_rate": 2.424276237198838e-05,
"loss": 0.5316,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4500,
"tokens_per_second_per_gpu": 2420.85
},
{
"epoch": 0.45036948272418614,
"grad_norm": 2.34375,
"learning_rate": 2.4239397026877516e-05,
"loss": 0.5216,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4510,
"tokens_per_second_per_gpu": 2419.56
},
{
"epoch": 0.45136808468144596,
"grad_norm": 2.640625,
"learning_rate": 2.4236024454777938e-05,
"loss": 0.4876,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4520,
"tokens_per_second_per_gpu": 2444.45
},
{
"epoch": 0.45236668663870583,
"grad_norm": 3.140625,
"learning_rate": 2.4232644657765874e-05,
"loss": 0.5669,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4530,
"tokens_per_second_per_gpu": 2443.96
},
{
"epoch": 0.45336528859596564,
"grad_norm": 2.375,
"learning_rate": 2.422925763792198e-05,
"loss": 0.6136,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4540,
"tokens_per_second_per_gpu": 2387.65
},
{
"epoch": 0.45436389055322546,
"grad_norm": 3.578125,
"learning_rate": 2.422586339733137e-05,
"loss": 0.5344,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4550,
"tokens_per_second_per_gpu": 2422.13
},
{
"epoch": 0.45536249251048533,
"grad_norm": 3.453125,
"learning_rate": 2.4222461938083606e-05,
"loss": 0.6354,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4560,
"tokens_per_second_per_gpu": 2456.26
},
{
"epoch": 0.45636109446774514,
"grad_norm": 2.59375,
"learning_rate": 2.4219053262272678e-05,
"loss": 0.5671,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4570,
"tokens_per_second_per_gpu": 2528.45
},
{
"epoch": 0.457359696425005,
"grad_norm": 3.6875,
"learning_rate": 2.421563737199704e-05,
"loss": 0.5183,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4580,
"tokens_per_second_per_gpu": 2358.28
},
{
"epoch": 0.45835829838226483,
"grad_norm": 3.421875,
"learning_rate": 2.4212214269359577e-05,
"loss": 0.6117,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4590,
"tokens_per_second_per_gpu": 2611.44
},
{
"epoch": 0.45935690033952464,
"grad_norm": 3.546875,
"learning_rate": 2.4208783956467606e-05,
"loss": 0.5579,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4600,
"tokens_per_second_per_gpu": 2472.43
},
{
"epoch": 0.4603555022967845,
"grad_norm": 3.09375,
"learning_rate": 2.4205346435432895e-05,
"loss": 0.629,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4610,
"tokens_per_second_per_gpu": 2377.02
},
{
"epoch": 0.46135410425404433,
"grad_norm": 2.890625,
"learning_rate": 2.420190170837165e-05,
"loss": 0.5716,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4620,
"tokens_per_second_per_gpu": 2566.22
},
{
"epoch": 0.4623527062113042,
"grad_norm": 3.484375,
"learning_rate": 2.4198449777404497e-05,
"loss": 0.5158,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4630,
"tokens_per_second_per_gpu": 2220.8
},
{
"epoch": 0.463351308168564,
"grad_norm": 3.3125,
"learning_rate": 2.419499064465652e-05,
"loss": 0.6126,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4640,
"tokens_per_second_per_gpu": 2309.71
},
{
"epoch": 0.46434991012582383,
"grad_norm": 3.265625,
"learning_rate": 2.4191524312257215e-05,
"loss": 0.5655,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4650,
"tokens_per_second_per_gpu": 2538.15
},
{
"epoch": 0.4653485120830837,
"grad_norm": 3.90625,
"learning_rate": 2.418805078234052e-05,
"loss": 0.5663,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4660,
"tokens_per_second_per_gpu": 2333.92
},
{
"epoch": 0.4663471140403435,
"grad_norm": 3.015625,
"learning_rate": 2.4184570057044816e-05,
"loss": 0.6321,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4670,
"tokens_per_second_per_gpu": 2502.95
},
{
"epoch": 0.46734571599760333,
"grad_norm": 4.4375,
"learning_rate": 2.418108213851289e-05,
"loss": 0.5627,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4680,
"tokens_per_second_per_gpu": 2467.4
},
{
"epoch": 0.4683443179548632,
"grad_norm": 2.875,
"learning_rate": 2.4177587028891973e-05,
"loss": 0.5177,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4690,
"tokens_per_second_per_gpu": 2407.5
},
{
"epoch": 0.469342919912123,
"grad_norm": 3.1875,
"learning_rate": 2.4174084730333724e-05,
"loss": 0.4718,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4700,
"tokens_per_second_per_gpu": 2512.93
},
{
"epoch": 0.4703415218693829,
"grad_norm": 2.390625,
"learning_rate": 2.417057524499421e-05,
"loss": 0.488,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4710,
"tokens_per_second_per_gpu": 2444.95
},
{
"epoch": 0.4713401238266427,
"grad_norm": 3.640625,
"learning_rate": 2.416705857503394e-05,
"loss": 0.5131,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4720,
"tokens_per_second_per_gpu": 2330.08
},
{
"epoch": 0.4723387257839025,
"grad_norm": 2.8125,
"learning_rate": 2.4163534722617847e-05,
"loss": 0.5353,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4730,
"tokens_per_second_per_gpu": 2498.29
},
{
"epoch": 0.4733373277411624,
"grad_norm": 3.5625,
"learning_rate": 2.416000368991527e-05,
"loss": 0.5322,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4740,
"tokens_per_second_per_gpu": 2445.76
},
{
"epoch": 0.4743359296984222,
"grad_norm": 3.4375,
"learning_rate": 2.4156465479099988e-05,
"loss": 0.5458,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4750,
"tokens_per_second_per_gpu": 2424.4
},
{
"epoch": 0.47533453165568207,
"grad_norm": 2.859375,
"learning_rate": 2.4152920092350178e-05,
"loss": 0.5389,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4760,
"tokens_per_second_per_gpu": 2373.17
},
{
"epoch": 0.4763331336129419,
"grad_norm": 3.171875,
"learning_rate": 2.4149367531848453e-05,
"loss": 0.5211,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4770,
"tokens_per_second_per_gpu": 2600.46
},
{
"epoch": 0.4773317355702017,
"grad_norm": 2.984375,
"learning_rate": 2.4145807799781832e-05,
"loss": 0.5741,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4780,
"tokens_per_second_per_gpu": 2526.23
},
{
"epoch": 0.47833033752746157,
"grad_norm": 2.421875,
"learning_rate": 2.4142240898341748e-05,
"loss": 0.5601,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4790,
"tokens_per_second_per_gpu": 2572.8
},
{
"epoch": 0.4793289394847214,
"grad_norm": 3.015625,
"learning_rate": 2.4138666829724056e-05,
"loss": 0.5544,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4800,
"tokens_per_second_per_gpu": 2528.82
},
{
"epoch": 0.48032754144198125,
"grad_norm": 2.875,
"learning_rate": 2.4135085596129016e-05,
"loss": 0.5956,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4810,
"tokens_per_second_per_gpu": 2478.72
},
{
"epoch": 0.48132614339924107,
"grad_norm": 2.5,
"learning_rate": 2.41314971997613e-05,
"loss": 0.5668,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4820,
"tokens_per_second_per_gpu": 2397.99
},
{
"epoch": 0.4823247453565009,
"grad_norm": 3.625,
"learning_rate": 2.4127901642829996e-05,
"loss": 0.5941,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4830,
"tokens_per_second_per_gpu": 2474.64
},
{
"epoch": 0.48332334731376075,
"grad_norm": 3.859375,
"learning_rate": 2.412429892754859e-05,
"loss": 0.5743,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4840,
"tokens_per_second_per_gpu": 2584.3
},
{
"epoch": 0.48432194927102057,
"grad_norm": 2.984375,
"learning_rate": 2.4120689056134982e-05,
"loss": 0.5397,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4850,
"tokens_per_second_per_gpu": 2359.69
},
{
"epoch": 0.4853205512282804,
"grad_norm": 3.046875,
"learning_rate": 2.4117072030811474e-05,
"loss": 0.564,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4860,
"tokens_per_second_per_gpu": 2459.11
},
{
"epoch": 0.48631915318554025,
"grad_norm": 3.140625,
"learning_rate": 2.4113447853804776e-05,
"loss": 0.5758,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4870,
"tokens_per_second_per_gpu": 2461.79
},
{
"epoch": 0.48731775514280007,
"grad_norm": 2.734375,
"learning_rate": 2.4109816527345994e-05,
"loss": 0.6419,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4880,
"tokens_per_second_per_gpu": 2434.32
},
{
"epoch": 0.48831635710005994,
"grad_norm": 2.90625,
"learning_rate": 2.4106178053670643e-05,
"loss": 0.5302,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4890,
"tokens_per_second_per_gpu": 2362.72
},
{
"epoch": 0.48931495905731975,
"grad_norm": 4.0625,
"learning_rate": 2.410253243501863e-05,
"loss": 0.5507,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4900,
"tokens_per_second_per_gpu": 2235.29
},
{
"epoch": 0.49031356101457957,
"grad_norm": 3.34375,
"learning_rate": 2.409887967363427e-05,
"loss": 0.5862,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4910,
"tokens_per_second_per_gpu": 2518.31
},
{
"epoch": 0.49131216297183944,
"grad_norm": 2.96875,
"learning_rate": 2.409521977176627e-05,
"loss": 0.5617,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4920,
"tokens_per_second_per_gpu": 2549.59
},
{
"epoch": 0.49231076492909925,
"grad_norm": 3.640625,
"learning_rate": 2.4091552731667724e-05,
"loss": 0.5003,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4930,
"tokens_per_second_per_gpu": 2278.58
},
{
"epoch": 0.4933093668863591,
"grad_norm": 2.90625,
"learning_rate": 2.4087878555596137e-05,
"loss": 0.4769,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4940,
"tokens_per_second_per_gpu": 2406.03
},
{
"epoch": 0.49430796884361894,
"grad_norm": 2.71875,
"learning_rate": 2.4084197245813404e-05,
"loss": 0.5552,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4950,
"tokens_per_second_per_gpu": 2631.77
},
{
"epoch": 0.49530657080087875,
"grad_norm": 3.34375,
"learning_rate": 2.4080508804585798e-05,
"loss": 0.4799,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4960,
"tokens_per_second_per_gpu": 2549.99
},
{
"epoch": 0.4963051727581386,
"grad_norm": 4.0625,
"learning_rate": 2.407681323418399e-05,
"loss": 0.5167,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4970,
"tokens_per_second_per_gpu": 2569.42
},
{
"epoch": 0.49730377471539844,
"grad_norm": 3.296875,
"learning_rate": 2.4073110536883054e-05,
"loss": 0.5434,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4980,
"tokens_per_second_per_gpu": 2485.73
},
{
"epoch": 0.49830237667265825,
"grad_norm": 4.21875,
"learning_rate": 2.4069400714962425e-05,
"loss": 0.6331,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 4990,
"tokens_per_second_per_gpu": 2396.54
},
{
"epoch": 0.4993009786299181,
"grad_norm": 3.390625,
"learning_rate": 2.4065683770705944e-05,
"loss": 0.5372,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5000,
"tokens_per_second_per_gpu": 2274.09
},
{
"epoch": 0.5002995805871779,
"grad_norm": 3.453125,
"learning_rate": 2.4061959706401828e-05,
"loss": 0.4883,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5010,
"tokens_per_second_per_gpu": 2333.26
},
{
"epoch": 0.5012981825444378,
"grad_norm": 3.4375,
"learning_rate": 2.4058228524342684e-05,
"loss": 0.5563,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5020,
"tokens_per_second_per_gpu": 2403.31
},
{
"epoch": 0.5022967845016976,
"grad_norm": 3.703125,
"learning_rate": 2.4054490226825487e-05,
"loss": 0.5188,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5030,
"tokens_per_second_per_gpu": 2215.03
},
{
"epoch": 0.5032953864589574,
"grad_norm": 3.765625,
"learning_rate": 2.405074481615161e-05,
"loss": 0.5939,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5040,
"tokens_per_second_per_gpu": 2538.45
},
{
"epoch": 0.5042939884162173,
"grad_norm": 3.328125,
"learning_rate": 2.4046992294626797e-05,
"loss": 0.6671,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5050,
"tokens_per_second_per_gpu": 2375.83
},
{
"epoch": 0.5052925903734772,
"grad_norm": 3.640625,
"learning_rate": 2.4043232664561164e-05,
"loss": 0.5268,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5060,
"tokens_per_second_per_gpu": 2412.54
},
{
"epoch": 0.5062911923307369,
"grad_norm": 2.53125,
"learning_rate": 2.403946592826921e-05,
"loss": 0.5084,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5070,
"tokens_per_second_per_gpu": 2450.83
},
{
"epoch": 0.5072897942879968,
"grad_norm": 3.265625,
"learning_rate": 2.4035692088069813e-05,
"loss": 0.5108,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5080,
"tokens_per_second_per_gpu": 2333.46
},
{
"epoch": 0.5082883962452567,
"grad_norm": 3.453125,
"learning_rate": 2.4031911146286208e-05,
"loss": 0.59,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5090,
"tokens_per_second_per_gpu": 2426.0
},
{
"epoch": 0.5092869982025164,
"grad_norm": 3.234375,
"learning_rate": 2.4028123105246016e-05,
"loss": 0.5588,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5100,
"tokens_per_second_per_gpu": 2474.85
},
{
"epoch": 0.5102856001597763,
"grad_norm": 3.125,
"learning_rate": 2.402432796728123e-05,
"loss": 0.5887,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5110,
"tokens_per_second_per_gpu": 2506.2
},
{
"epoch": 0.5112842021170362,
"grad_norm": 2.8125,
"learning_rate": 2.4020525734728206e-05,
"loss": 0.5968,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5120,
"tokens_per_second_per_gpu": 2355.71
},
{
"epoch": 0.5122828040742959,
"grad_norm": 3.625,
"learning_rate": 2.401671640992766e-05,
"loss": 0.5221,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5130,
"tokens_per_second_per_gpu": 2426.5
},
{
"epoch": 0.5132814060315558,
"grad_norm": 4.40625,
"learning_rate": 2.401289999522469e-05,
"loss": 0.5342,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5140,
"tokens_per_second_per_gpu": 2431.84
},
{
"epoch": 0.5142800079888157,
"grad_norm": 3.515625,
"learning_rate": 2.4009076492968748e-05,
"loss": 0.5455,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5150,
"tokens_per_second_per_gpu": 2411.93
},
{
"epoch": 0.5152786099460755,
"grad_norm": 2.5625,
"learning_rate": 2.4005245905513655e-05,
"loss": 0.5442,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5160,
"tokens_per_second_per_gpu": 2429.42
},
{
"epoch": 0.5162772119033353,
"grad_norm": 3.640625,
"learning_rate": 2.400140823521759e-05,
"loss": 0.6137,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5170,
"tokens_per_second_per_gpu": 2337.28
},
{
"epoch": 0.5172758138605952,
"grad_norm": 2.78125,
"learning_rate": 2.399756348444309e-05,
"loss": 0.598,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5180,
"tokens_per_second_per_gpu": 2434.73
},
{
"epoch": 0.518274415817855,
"grad_norm": 3.484375,
"learning_rate": 2.399371165555706e-05,
"loss": 0.5828,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5190,
"tokens_per_second_per_gpu": 2276.22
},
{
"epoch": 0.5192730177751148,
"grad_norm": 3.015625,
"learning_rate": 2.398985275093075e-05,
"loss": 0.5385,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5200,
"tokens_per_second_per_gpu": 2463.48
},
{
"epoch": 0.5202716197323747,
"grad_norm": 3.84375,
"learning_rate": 2.398598677293978e-05,
"loss": 0.5646,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5210,
"tokens_per_second_per_gpu": 2286.72
},
{
"epoch": 0.5212702216896345,
"grad_norm": 2.609375,
"learning_rate": 2.3982113723964113e-05,
"loss": 0.5604,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5220,
"tokens_per_second_per_gpu": 2479.19
},
{
"epoch": 0.5222688236468943,
"grad_norm": 3.6875,
"learning_rate": 2.397823360638807e-05,
"loss": 0.5133,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5230,
"tokens_per_second_per_gpu": 2443.43
},
{
"epoch": 0.5232674256041542,
"grad_norm": 3.21875,
"learning_rate": 2.397434642260032e-05,
"loss": 0.5446,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5240,
"tokens_per_second_per_gpu": 2366.96
},
{
"epoch": 0.524266027561414,
"grad_norm": 3.65625,
"learning_rate": 2.3970452174993885e-05,
"loss": 0.5931,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5250,
"tokens_per_second_per_gpu": 2459.65
},
{
"epoch": 0.5252646295186738,
"grad_norm": 3.109375,
"learning_rate": 2.396655086596614e-05,
"loss": 0.5185,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5260,
"tokens_per_second_per_gpu": 2486.66
},
{
"epoch": 0.5262632314759337,
"grad_norm": 4.0,
"learning_rate": 2.3962642497918802e-05,
"loss": 0.7091,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5270,
"tokens_per_second_per_gpu": 2393.16
},
{
"epoch": 0.5272618334331935,
"grad_norm": 2.015625,
"learning_rate": 2.3958727073257937e-05,
"loss": 0.5092,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5280,
"tokens_per_second_per_gpu": 2549.0
},
{
"epoch": 0.5282604353904534,
"grad_norm": 3.0,
"learning_rate": 2.3954804594393943e-05,
"loss": 0.5459,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5290,
"tokens_per_second_per_gpu": 2448.94
},
{
"epoch": 0.5292590373477132,
"grad_norm": 3.140625,
"learning_rate": 2.3950875063741573e-05,
"loss": 0.5546,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5300,
"tokens_per_second_per_gpu": 2529.08
},
{
"epoch": 0.530257639304973,
"grad_norm": 2.265625,
"learning_rate": 2.3946938483719923e-05,
"loss": 0.5281,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5310,
"tokens_per_second_per_gpu": 2448.53
},
{
"epoch": 0.5312562412622329,
"grad_norm": 3.421875,
"learning_rate": 2.394299485675242e-05,
"loss": 0.5782,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5320,
"tokens_per_second_per_gpu": 2392.17
},
{
"epoch": 0.5322548432194927,
"grad_norm": 4.21875,
"learning_rate": 2.3939044185266837e-05,
"loss": 0.5397,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5330,
"tokens_per_second_per_gpu": 2232.2
},
{
"epoch": 0.5332534451767525,
"grad_norm": 3.09375,
"learning_rate": 2.393508647169528e-05,
"loss": 0.5584,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5340,
"tokens_per_second_per_gpu": 2289.8
},
{
"epoch": 0.5342520471340124,
"grad_norm": 4.0,
"learning_rate": 2.3931121718474182e-05,
"loss": 0.5214,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5350,
"tokens_per_second_per_gpu": 2614.37
},
{
"epoch": 0.5352506490912722,
"grad_norm": 3.296875,
"learning_rate": 2.3927149928044328e-05,
"loss": 0.5758,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5360,
"tokens_per_second_per_gpu": 2450.7
},
{
"epoch": 0.536249251048532,
"grad_norm": 3.15625,
"learning_rate": 2.392317110285082e-05,
"loss": 0.5795,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5370,
"tokens_per_second_per_gpu": 2441.27
},
{
"epoch": 0.5372478530057919,
"grad_norm": 2.984375,
"learning_rate": 2.3919185245343095e-05,
"loss": 0.5365,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5380,
"tokens_per_second_per_gpu": 2465.43
},
{
"epoch": 0.5382464549630517,
"grad_norm": 3.359375,
"learning_rate": 2.3915192357974927e-05,
"loss": 0.6324,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5390,
"tokens_per_second_per_gpu": 2501.65
},
{
"epoch": 0.5392450569203115,
"grad_norm": 4.03125,
"learning_rate": 2.3911192443204407e-05,
"loss": 0.5911,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5400,
"tokens_per_second_per_gpu": 2271.73
},
{
"epoch": 0.5402436588775714,
"grad_norm": 3.125,
"learning_rate": 2.390718550349395e-05,
"loss": 0.5488,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5410,
"tokens_per_second_per_gpu": 2594.14
},
{
"epoch": 0.5412422608348313,
"grad_norm": 4.0,
"learning_rate": 2.390317154131031e-05,
"loss": 0.6199,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5420,
"tokens_per_second_per_gpu": 2421.96
},
{
"epoch": 0.542240862792091,
"grad_norm": 3.171875,
"learning_rate": 2.3899150559124554e-05,
"loss": 0.5541,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5430,
"tokens_per_second_per_gpu": 2476.47
},
{
"epoch": 0.5432394647493509,
"grad_norm": 3.3125,
"learning_rate": 2.389512255941207e-05,
"loss": 0.5335,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5440,
"tokens_per_second_per_gpu": 2410.69
},
{
"epoch": 0.5442380667066108,
"grad_norm": 3.046875,
"learning_rate": 2.3891087544652573e-05,
"loss": 0.5144,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5450,
"tokens_per_second_per_gpu": 2543.43
},
{
"epoch": 0.5452366686638705,
"grad_norm": 3.859375,
"learning_rate": 2.3887045517330085e-05,
"loss": 0.5859,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5460,
"tokens_per_second_per_gpu": 2364.21
},
{
"epoch": 0.5462352706211304,
"grad_norm": 3.0625,
"learning_rate": 2.388299647993296e-05,
"loss": 0.5521,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5470,
"tokens_per_second_per_gpu": 2478.25
},
{
"epoch": 0.5472338725783903,
"grad_norm": 3.015625,
"learning_rate": 2.387894043495386e-05,
"loss": 0.5718,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5480,
"tokens_per_second_per_gpu": 2283.55
},
{
"epoch": 0.54823247453565,
"grad_norm": 3.4375,
"learning_rate": 2.3874877384889756e-05,
"loss": 0.5917,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5490,
"tokens_per_second_per_gpu": 2275.5
},
{
"epoch": 0.5492310764929099,
"grad_norm": 3.640625,
"learning_rate": 2.387080733224194e-05,
"loss": 0.6264,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5500,
"tokens_per_second_per_gpu": 2620.18
},
{
"epoch": 0.5502296784501698,
"grad_norm": 3.359375,
"learning_rate": 2.3866730279516006e-05,
"loss": 0.5682,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5510,
"tokens_per_second_per_gpu": 2320.04
},
{
"epoch": 0.5512282804074295,
"grad_norm": 2.90625,
"learning_rate": 2.3862646229221867e-05,
"loss": 0.5858,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5520,
"tokens_per_second_per_gpu": 2584.93
},
{
"epoch": 0.5522268823646894,
"grad_norm": 3.9375,
"learning_rate": 2.385855518387374e-05,
"loss": 0.5627,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5530,
"tokens_per_second_per_gpu": 2356.73
},
{
"epoch": 0.5532254843219493,
"grad_norm": 3.25,
"learning_rate": 2.3854457145990146e-05,
"loss": 0.5535,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5540,
"tokens_per_second_per_gpu": 2260.76
},
{
"epoch": 0.5542240862792092,
"grad_norm": 2.875,
"learning_rate": 2.385035211809391e-05,
"loss": 0.5486,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5550,
"tokens_per_second_per_gpu": 2326.77
},
{
"epoch": 0.5552226882364689,
"grad_norm": 2.625,
"learning_rate": 2.384624010271217e-05,
"loss": 0.5493,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5560,
"tokens_per_second_per_gpu": 2385.14
},
{
"epoch": 0.5562212901937288,
"grad_norm": 3.984375,
"learning_rate": 2.384212110237635e-05,
"loss": 0.5923,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5570,
"tokens_per_second_per_gpu": 2473.04
},
{
"epoch": 0.5572198921509887,
"grad_norm": 4.28125,
"learning_rate": 2.3837995119622185e-05,
"loss": 0.546,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5580,
"tokens_per_second_per_gpu": 2410.71
},
{
"epoch": 0.5582184941082484,
"grad_norm": 3.484375,
"learning_rate": 2.383386215698971e-05,
"loss": 0.5881,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5590,
"tokens_per_second_per_gpu": 2304.25
},
{
"epoch": 0.5592170960655083,
"grad_norm": 3.578125,
"learning_rate": 2.3829722217023255e-05,
"loss": 0.5402,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5600,
"tokens_per_second_per_gpu": 2349.77
},
{
"epoch": 0.5602156980227682,
"grad_norm": 3.796875,
"learning_rate": 2.382557530227143e-05,
"loss": 0.5821,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5610,
"tokens_per_second_per_gpu": 2402.1
},
{
"epoch": 0.5612142999800279,
"grad_norm": 3.515625,
"learning_rate": 2.3821421415287162e-05,
"loss": 0.524,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5620,
"tokens_per_second_per_gpu": 2476.49
},
{
"epoch": 0.5622129019372878,
"grad_norm": 2.859375,
"learning_rate": 2.3817260558627658e-05,
"loss": 0.5359,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5630,
"tokens_per_second_per_gpu": 2464.38
},
{
"epoch": 0.5632115038945477,
"grad_norm": 2.90625,
"learning_rate": 2.381309273485442e-05,
"loss": 0.6084,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5640,
"tokens_per_second_per_gpu": 2383.39
},
{
"epoch": 0.5642101058518074,
"grad_norm": 2.78125,
"learning_rate": 2.380891794653323e-05,
"loss": 0.5132,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5650,
"tokens_per_second_per_gpu": 2358.67
},
{
"epoch": 0.5652087078090673,
"grad_norm": 3.34375,
"learning_rate": 2.3804736196234177e-05,
"loss": 0.5258,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5660,
"tokens_per_second_per_gpu": 2580.38
},
{
"epoch": 0.5662073097663272,
"grad_norm": 3.0,
"learning_rate": 2.3800547486531614e-05,
"loss": 0.5578,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5670,
"tokens_per_second_per_gpu": 2421.23
},
{
"epoch": 0.567205911723587,
"grad_norm": 3.296875,
"learning_rate": 2.3796351820004188e-05,
"loss": 0.5441,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5680,
"tokens_per_second_per_gpu": 2427.33
},
{
"epoch": 0.5682045136808468,
"grad_norm": 3.59375,
"learning_rate": 2.379214919923483e-05,
"loss": 0.6666,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5690,
"tokens_per_second_per_gpu": 2168.69
},
{
"epoch": 0.5692031156381067,
"grad_norm": 2.90625,
"learning_rate": 2.378793962681075e-05,
"loss": 0.5647,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5700,
"tokens_per_second_per_gpu": 2451.56
},
{
"epoch": 0.5702017175953665,
"grad_norm": 3.40625,
"learning_rate": 2.378372310532344e-05,
"loss": 0.5448,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5710,
"tokens_per_second_per_gpu": 2396.81
},
{
"epoch": 0.5712003195526263,
"grad_norm": 3.8125,
"learning_rate": 2.377949963736867e-05,
"loss": 0.5739,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5720,
"tokens_per_second_per_gpu": 2300.7
},
{
"epoch": 0.5721989215098862,
"grad_norm": 3.421875,
"learning_rate": 2.3775269225546477e-05,
"loss": 0.5776,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5730,
"tokens_per_second_per_gpu": 2476.96
},
{
"epoch": 0.573197523467146,
"grad_norm": 2.828125,
"learning_rate": 2.3771031872461186e-05,
"loss": 0.5555,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5740,
"tokens_per_second_per_gpu": 2466.16
},
{
"epoch": 0.5741961254244058,
"grad_norm": 2.859375,
"learning_rate": 2.376678758072139e-05,
"loss": 0.551,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5750,
"tokens_per_second_per_gpu": 2492.07
},
{
"epoch": 0.5751947273816657,
"grad_norm": 3.15625,
"learning_rate": 2.376253635293995e-05,
"loss": 0.5041,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5760,
"tokens_per_second_per_gpu": 2523.54
},
{
"epoch": 0.5761933293389255,
"grad_norm": 3.328125,
"learning_rate": 2.3758278191734004e-05,
"loss": 0.5486,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5770,
"tokens_per_second_per_gpu": 2361.79
},
{
"epoch": 0.5771919312961853,
"grad_norm": 2.84375,
"learning_rate": 2.3754013099724952e-05,
"loss": 0.5194,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5780,
"tokens_per_second_per_gpu": 2388.88
},
{
"epoch": 0.5781905332534452,
"grad_norm": 2.53125,
"learning_rate": 2.3749741079538463e-05,
"loss": 0.5631,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5790,
"tokens_per_second_per_gpu": 2678.98
},
{
"epoch": 0.579189135210705,
"grad_norm": 2.6875,
"learning_rate": 2.3745462133804468e-05,
"loss": 0.5004,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5800,
"tokens_per_second_per_gpu": 2481.67
},
{
"epoch": 0.5801877371679649,
"grad_norm": 2.53125,
"learning_rate": 2.374117626515717e-05,
"loss": 0.4748,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5810,
"tokens_per_second_per_gpu": 2458.91
},
{
"epoch": 0.5811863391252247,
"grad_norm": 3.40625,
"learning_rate": 2.3736883476235025e-05,
"loss": 0.4887,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5820,
"tokens_per_second_per_gpu": 2547.53
},
{
"epoch": 0.5821849410824845,
"grad_norm": 2.875,
"learning_rate": 2.3732583769680754e-05,
"loss": 0.5811,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5830,
"tokens_per_second_per_gpu": 2409.01
},
{
"epoch": 0.5831835430397444,
"grad_norm": 4.0625,
"learning_rate": 2.3728277148141338e-05,
"loss": 0.5326,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5840,
"tokens_per_second_per_gpu": 2469.02
},
{
"epoch": 0.5841821449970042,
"grad_norm": 3.53125,
"learning_rate": 2.372396361426801e-05,
"loss": 0.5813,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5850,
"tokens_per_second_per_gpu": 2370.35
},
{
"epoch": 0.585180746954264,
"grad_norm": 3.296875,
"learning_rate": 2.371964317071626e-05,
"loss": 0.5683,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5860,
"tokens_per_second_per_gpu": 2452.34
},
{
"epoch": 0.5861793489115239,
"grad_norm": 2.75,
"learning_rate": 2.3715315820145835e-05,
"loss": 0.503,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5870,
"tokens_per_second_per_gpu": 2460.97
},
{
"epoch": 0.5871779508687837,
"grad_norm": 3.03125,
"learning_rate": 2.3710981565220727e-05,
"loss": 0.5877,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5880,
"tokens_per_second_per_gpu": 2396.15
},
{
"epoch": 0.5881765528260435,
"grad_norm": 2.390625,
"learning_rate": 2.370664040860919e-05,
"loss": 0.5046,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5890,
"tokens_per_second_per_gpu": 2427.32
},
{
"epoch": 0.5891751547833034,
"grad_norm": 3.796875,
"learning_rate": 2.370229235298371e-05,
"loss": 0.5362,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5900,
"tokens_per_second_per_gpu": 2388.97
},
{
"epoch": 0.5901737567405632,
"grad_norm": 4.0,
"learning_rate": 2.369793740102104e-05,
"loss": 0.5232,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5910,
"tokens_per_second_per_gpu": 2517.35
},
{
"epoch": 0.591172358697823,
"grad_norm": 2.84375,
"learning_rate": 2.3693575555402164e-05,
"loss": 0.6054,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5920,
"tokens_per_second_per_gpu": 2574.84
},
{
"epoch": 0.5921709606550829,
"grad_norm": 3.4375,
"learning_rate": 2.3689206818812317e-05,
"loss": 0.5935,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5930,
"tokens_per_second_per_gpu": 2438.65
},
{
"epoch": 0.5931695626123428,
"grad_norm": 3.421875,
"learning_rate": 2.3684831193940965e-05,
"loss": 0.5366,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5940,
"tokens_per_second_per_gpu": 2452.1
},
{
"epoch": 0.5941681645696025,
"grad_norm": 2.875,
"learning_rate": 2.3680448683481835e-05,
"loss": 0.5303,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5950,
"tokens_per_second_per_gpu": 2484.88
},
{
"epoch": 0.5951667665268624,
"grad_norm": 2.75,
"learning_rate": 2.3676059290132874e-05,
"loss": 0.5421,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5960,
"tokens_per_second_per_gpu": 2506.0
},
{
"epoch": 0.5961653684841223,
"grad_norm": 3.046875,
"learning_rate": 2.3671663016596273e-05,
"loss": 0.5615,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5970,
"tokens_per_second_per_gpu": 2361.73
},
{
"epoch": 0.597163970441382,
"grad_norm": 3.0625,
"learning_rate": 2.3667259865578463e-05,
"loss": 0.5055,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5980,
"tokens_per_second_per_gpu": 2644.14
},
{
"epoch": 0.5981625723986419,
"grad_norm": 3.875,
"learning_rate": 2.36628498397901e-05,
"loss": 0.5696,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 5990,
"tokens_per_second_per_gpu": 2531.78
},
{
"epoch": 0.5991611743559018,
"grad_norm": 3.03125,
"learning_rate": 2.3658432941946084e-05,
"loss": 0.5633,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6000,
"tokens_per_second_per_gpu": 2540.37
},
{
"epoch": 0.6001597763131615,
"grad_norm": 2.171875,
"learning_rate": 2.3654009174765532e-05,
"loss": 0.4821,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6010,
"tokens_per_second_per_gpu": 2310.62
},
{
"epoch": 0.6011583782704214,
"grad_norm": 3.28125,
"learning_rate": 2.36495785409718e-05,
"loss": 0.5541,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6020,
"tokens_per_second_per_gpu": 2425.51
},
{
"epoch": 0.6021569802276813,
"grad_norm": 2.875,
"learning_rate": 2.364514104329246e-05,
"loss": 0.4777,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6030,
"tokens_per_second_per_gpu": 2557.81
},
{
"epoch": 0.603155582184941,
"grad_norm": 2.703125,
"learning_rate": 2.364069668445933e-05,
"loss": 0.5474,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6040,
"tokens_per_second_per_gpu": 2283.58
},
{
"epoch": 0.6041541841422009,
"grad_norm": 2.453125,
"learning_rate": 2.3636245467208428e-05,
"loss": 0.5488,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6050,
"tokens_per_second_per_gpu": 2460.04
},
{
"epoch": 0.6051527860994608,
"grad_norm": 3.5,
"learning_rate": 2.3631787394280007e-05,
"loss": 0.5247,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6060,
"tokens_per_second_per_gpu": 2518.49
},
{
"epoch": 0.6061513880567206,
"grad_norm": 3.5625,
"learning_rate": 2.3627322468418547e-05,
"loss": 0.52,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6070,
"tokens_per_second_per_gpu": 2503.82
},
{
"epoch": 0.6071499900139804,
"grad_norm": 3.234375,
"learning_rate": 2.362285069237273e-05,
"loss": 0.588,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6080,
"tokens_per_second_per_gpu": 2571.65
},
{
"epoch": 0.6081485919712403,
"grad_norm": 3.0625,
"learning_rate": 2.3618372068895465e-05,
"loss": 0.5022,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6090,
"tokens_per_second_per_gpu": 2359.77
},
{
"epoch": 0.6091471939285001,
"grad_norm": 2.640625,
"learning_rate": 2.361388660074388e-05,
"loss": 0.5529,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6100,
"tokens_per_second_per_gpu": 2441.49
},
{
"epoch": 0.6101457958857599,
"grad_norm": 3.25,
"learning_rate": 2.3609394290679303e-05,
"loss": 0.5274,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6110,
"tokens_per_second_per_gpu": 2389.13
},
{
"epoch": 0.6111443978430198,
"grad_norm": 2.640625,
"learning_rate": 2.3604895141467294e-05,
"loss": 0.5642,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6120,
"tokens_per_second_per_gpu": 2468.79
},
{
"epoch": 0.6121429998002796,
"grad_norm": 2.453125,
"learning_rate": 2.36003891558776e-05,
"loss": 0.5241,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6130,
"tokens_per_second_per_gpu": 2412.56
},
{
"epoch": 0.6131416017575394,
"grad_norm": 2.671875,
"learning_rate": 2.35958763366842e-05,
"loss": 0.539,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6140,
"tokens_per_second_per_gpu": 2438.1
},
{
"epoch": 0.6141402037147993,
"grad_norm": 3.59375,
"learning_rate": 2.3591356686665255e-05,
"loss": 0.5303,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6150,
"tokens_per_second_per_gpu": 2391.96
},
{
"epoch": 0.6151388056720591,
"grad_norm": 3.015625,
"learning_rate": 2.3586830208603156e-05,
"loss": 0.5435,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6160,
"tokens_per_second_per_gpu": 2486.14
},
{
"epoch": 0.6161374076293189,
"grad_norm": 3.84375,
"learning_rate": 2.358229690528448e-05,
"loss": 0.5236,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6170,
"tokens_per_second_per_gpu": 2406.55
},
{
"epoch": 0.6171360095865788,
"grad_norm": 3.078125,
"learning_rate": 2.357775677950002e-05,
"loss": 0.5432,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6180,
"tokens_per_second_per_gpu": 2499.59
},
{
"epoch": 0.6181346115438386,
"grad_norm": 3.25,
"learning_rate": 2.357320983404475e-05,
"loss": 0.535,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6190,
"tokens_per_second_per_gpu": 2267.46
},
{
"epoch": 0.6191332135010985,
"grad_norm": 2.078125,
"learning_rate": 2.356865607171786e-05,
"loss": 0.4825,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6200,
"tokens_per_second_per_gpu": 2329.38
},
{
"epoch": 0.6201318154583583,
"grad_norm": 4.03125,
"learning_rate": 2.3564095495322726e-05,
"loss": 0.5419,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6210,
"tokens_per_second_per_gpu": 2427.66
},
{
"epoch": 0.6211304174156181,
"grad_norm": 3.546875,
"learning_rate": 2.3559528107666927e-05,
"loss": 0.5854,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6220,
"tokens_per_second_per_gpu": 2739.51
},
{
"epoch": 0.622129019372878,
"grad_norm": 3.25,
"learning_rate": 2.3554953911562233e-05,
"loss": 0.5563,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6230,
"tokens_per_second_per_gpu": 2437.26
},
{
"epoch": 0.6231276213301378,
"grad_norm": 3.0,
"learning_rate": 2.3550372909824595e-05,
"loss": 0.5111,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6240,
"tokens_per_second_per_gpu": 2474.2
},
{
"epoch": 0.6241262232873976,
"grad_norm": 2.65625,
"learning_rate": 2.3545785105274164e-05,
"loss": 0.5278,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6250,
"tokens_per_second_per_gpu": 2473.32
},
{
"epoch": 0.6251248252446575,
"grad_norm": 2.796875,
"learning_rate": 2.3541190500735284e-05,
"loss": 0.5701,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6260,
"tokens_per_second_per_gpu": 2145.78
},
{
"epoch": 0.6261234272019173,
"grad_norm": 2.65625,
"learning_rate": 2.3536589099036472e-05,
"loss": 0.5337,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6270,
"tokens_per_second_per_gpu": 2503.2
},
{
"epoch": 0.6271220291591771,
"grad_norm": 3.0625,
"learning_rate": 2.3531980903010434e-05,
"loss": 0.6164,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6280,
"tokens_per_second_per_gpu": 2354.88
},
{
"epoch": 0.628120631116437,
"grad_norm": 3.421875,
"learning_rate": 2.3527365915494065e-05,
"loss": 0.5665,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6290,
"tokens_per_second_per_gpu": 2313.86
},
{
"epoch": 0.6291192330736968,
"grad_norm": 2.4375,
"learning_rate": 2.352274413932843e-05,
"loss": 0.4569,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6300,
"tokens_per_second_per_gpu": 2415.36
},
{
"epoch": 0.6301178350309566,
"grad_norm": 3.21875,
"learning_rate": 2.3518115577358787e-05,
"loss": 0.5335,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6310,
"tokens_per_second_per_gpu": 2356.65
},
{
"epoch": 0.6311164369882165,
"grad_norm": 3.5,
"learning_rate": 2.3513480232434558e-05,
"loss": 0.5182,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6320,
"tokens_per_second_per_gpu": 2441.75
},
{
"epoch": 0.6321150389454764,
"grad_norm": 3.390625,
"learning_rate": 2.350883810740935e-05,
"loss": 0.5574,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6330,
"tokens_per_second_per_gpu": 2175.82
},
{
"epoch": 0.6331136409027361,
"grad_norm": 3.078125,
"learning_rate": 2.3504189205140942e-05,
"loss": 0.5581,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6340,
"tokens_per_second_per_gpu": 2490.39
},
{
"epoch": 0.634112242859996,
"grad_norm": 3.453125,
"learning_rate": 2.3499533528491274e-05,
"loss": 0.5094,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6350,
"tokens_per_second_per_gpu": 2165.91
},
{
"epoch": 0.6351108448172559,
"grad_norm": 3.0625,
"learning_rate": 2.349487108032648e-05,
"loss": 0.5649,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6360,
"tokens_per_second_per_gpu": 2519.19
},
{
"epoch": 0.6361094467745156,
"grad_norm": 2.890625,
"learning_rate": 2.349020186351684e-05,
"loss": 0.5346,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6370,
"tokens_per_second_per_gpu": 2634.85
},
{
"epoch": 0.6371080487317755,
"grad_norm": 3.578125,
"learning_rate": 2.348552588093681e-05,
"loss": 0.522,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6380,
"tokens_per_second_per_gpu": 2210.78
},
{
"epoch": 0.6381066506890354,
"grad_norm": 3.078125,
"learning_rate": 2.3480843135465015e-05,
"loss": 0.5757,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6390,
"tokens_per_second_per_gpu": 2360.44
},
{
"epoch": 0.6391052526462951,
"grad_norm": 2.96875,
"learning_rate": 2.3476153629984236e-05,
"loss": 0.6138,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6400,
"tokens_per_second_per_gpu": 2294.92
},
{
"epoch": 0.640103854603555,
"grad_norm": 2.90625,
"learning_rate": 2.3471457367381418e-05,
"loss": 0.564,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6410,
"tokens_per_second_per_gpu": 2408.76
},
{
"epoch": 0.6411024565608149,
"grad_norm": 3.328125,
"learning_rate": 2.346675435054767e-05,
"loss": 0.5139,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6420,
"tokens_per_second_per_gpu": 2312.73
},
{
"epoch": 0.6421010585180746,
"grad_norm": 3.921875,
"learning_rate": 2.3462044582378252e-05,
"loss": 0.5856,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6430,
"tokens_per_second_per_gpu": 2381.74
},
{
"epoch": 0.6430996604753345,
"grad_norm": 3.203125,
"learning_rate": 2.345732806577259e-05,
"loss": 0.5052,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6440,
"tokens_per_second_per_gpu": 1579.19
},
{
"epoch": 0.6440982624325944,
"grad_norm": 2.53125,
"learning_rate": 2.3452604803634253e-05,
"loss": 0.4741,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6450,
"tokens_per_second_per_gpu": 2416.04
},
{
"epoch": 0.6450968643898543,
"grad_norm": 3.28125,
"learning_rate": 2.3447874798870967e-05,
"loss": 0.4842,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6460,
"tokens_per_second_per_gpu": 2497.58
},
{
"epoch": 0.646095466347114,
"grad_norm": 3.453125,
"learning_rate": 2.3443138054394616e-05,
"loss": 0.573,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6470,
"tokens_per_second_per_gpu": 2470.82
},
{
"epoch": 0.6470940683043739,
"grad_norm": 3.859375,
"learning_rate": 2.343839457312122e-05,
"loss": 0.6474,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6480,
"tokens_per_second_per_gpu": 2508.61
},
{
"epoch": 0.6480926702616338,
"grad_norm": 2.609375,
"learning_rate": 2.3433644357970956e-05,
"loss": 0.4742,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6490,
"tokens_per_second_per_gpu": 2361.45
},
{
"epoch": 0.6490912722188935,
"grad_norm": 3.234375,
"learning_rate": 2.3428887411868144e-05,
"loss": 0.5069,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6500,
"tokens_per_second_per_gpu": 2573.21
},
{
"epoch": 0.6500898741761534,
"grad_norm": 2.375,
"learning_rate": 2.342412373774125e-05,
"loss": 0.6023,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6510,
"tokens_per_second_per_gpu": 2390.31
},
{
"epoch": 0.6510884761334133,
"grad_norm": 3.34375,
"learning_rate": 2.3419353338522877e-05,
"loss": 0.5201,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6520,
"tokens_per_second_per_gpu": 2412.2
},
{
"epoch": 0.652087078090673,
"grad_norm": 2.734375,
"learning_rate": 2.341457621714977e-05,
"loss": 0.477,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6530,
"tokens_per_second_per_gpu": 2419.49
},
{
"epoch": 0.6530856800479329,
"grad_norm": 2.71875,
"learning_rate": 2.340979237656281e-05,
"loss": 0.5396,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6540,
"tokens_per_second_per_gpu": 2518.31
},
{
"epoch": 0.6540842820051928,
"grad_norm": 2.578125,
"learning_rate": 2.3405001819707027e-05,
"loss": 0.5651,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6550,
"tokens_per_second_per_gpu": 2675.52
},
{
"epoch": 0.6550828839624525,
"grad_norm": 3.171875,
"learning_rate": 2.340020454953157e-05,
"loss": 0.5521,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6560,
"tokens_per_second_per_gpu": 2316.57
},
{
"epoch": 0.6560814859197124,
"grad_norm": 3.203125,
"learning_rate": 2.3395400568989723e-05,
"loss": 0.6035,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6570,
"tokens_per_second_per_gpu": 2570.83
},
{
"epoch": 0.6570800878769723,
"grad_norm": 2.78125,
"learning_rate": 2.3390589881038904e-05,
"loss": 0.5669,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6580,
"tokens_per_second_per_gpu": 2453.61
},
{
"epoch": 0.6580786898342321,
"grad_norm": 2.625,
"learning_rate": 2.3385772488640672e-05,
"loss": 0.509,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6590,
"tokens_per_second_per_gpu": 2478.86
},
{
"epoch": 0.6590772917914919,
"grad_norm": 4.46875,
"learning_rate": 2.3380948394760694e-05,
"loss": 0.6186,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6600,
"tokens_per_second_per_gpu": 2574.84
},
{
"epoch": 0.6600758937487518,
"grad_norm": 3.9375,
"learning_rate": 2.3376117602368776e-05,
"loss": 0.6322,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6610,
"tokens_per_second_per_gpu": 2408.42
},
{
"epoch": 0.6610744957060116,
"grad_norm": 4.09375,
"learning_rate": 2.3371280114438833e-05,
"loss": 0.4688,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6620,
"tokens_per_second_per_gpu": 2354.74
},
{
"epoch": 0.6620730976632714,
"grad_norm": 3.25,
"learning_rate": 2.3366435933948915e-05,
"loss": 0.6297,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6630,
"tokens_per_second_per_gpu": 2481.68
},
{
"epoch": 0.6630716996205313,
"grad_norm": 2.875,
"learning_rate": 2.3361585063881198e-05,
"loss": 0.5635,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6640,
"tokens_per_second_per_gpu": 2369.43
},
{
"epoch": 0.6640703015777911,
"grad_norm": 2.6875,
"learning_rate": 2.335672750722196e-05,
"loss": 0.4554,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6650,
"tokens_per_second_per_gpu": 2360.28
},
{
"epoch": 0.6650689035350509,
"grad_norm": 2.6875,
"learning_rate": 2.3351863266961598e-05,
"loss": 0.525,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6660,
"tokens_per_second_per_gpu": 2265.32
},
{
"epoch": 0.6660675054923108,
"grad_norm": 2.5625,
"learning_rate": 2.3346992346094633e-05,
"loss": 0.5632,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6670,
"tokens_per_second_per_gpu": 2498.8
},
{
"epoch": 0.6670661074495706,
"grad_norm": 4.125,
"learning_rate": 2.3342114747619692e-05,
"loss": 0.5214,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6680,
"tokens_per_second_per_gpu": 2449.85
},
{
"epoch": 0.6680647094068305,
"grad_norm": 3.1875,
"learning_rate": 2.333723047453952e-05,
"loss": 0.5811,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6690,
"tokens_per_second_per_gpu": 2385.67
},
{
"epoch": 0.6690633113640903,
"grad_norm": 3.46875,
"learning_rate": 2.3332339529860956e-05,
"loss": 0.5717,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6700,
"tokens_per_second_per_gpu": 2267.59
},
{
"epoch": 0.6700619133213501,
"grad_norm": 2.71875,
"learning_rate": 2.3327441916594957e-05,
"loss": 0.554,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6710,
"tokens_per_second_per_gpu": 2692.34
},
{
"epoch": 0.67106051527861,
"grad_norm": 2.46875,
"learning_rate": 2.3322537637756592e-05,
"loss": 0.5155,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6720,
"tokens_per_second_per_gpu": 2299.34
},
{
"epoch": 0.6720591172358698,
"grad_norm": 2.765625,
"learning_rate": 2.3317626696365013e-05,
"loss": 0.4726,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6730,
"tokens_per_second_per_gpu": 2434.43
},
{
"epoch": 0.6730577191931296,
"grad_norm": 3.125,
"learning_rate": 2.33127090954435e-05,
"loss": 0.5377,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6740,
"tokens_per_second_per_gpu": 2353.23
},
{
"epoch": 0.6740563211503895,
"grad_norm": 3.125,
"learning_rate": 2.330778483801941e-05,
"loss": 0.5512,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6750,
"tokens_per_second_per_gpu": 2455.87
},
{
"epoch": 0.6750549231076493,
"grad_norm": 3.296875,
"learning_rate": 2.3302853927124203e-05,
"loss": 0.5802,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6760,
"tokens_per_second_per_gpu": 2403.47
},
{
"epoch": 0.6760535250649091,
"grad_norm": 3.0625,
"learning_rate": 2.329791636579344e-05,
"loss": 0.4898,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6770,
"tokens_per_second_per_gpu": 2439.32
},
{
"epoch": 0.677052127022169,
"grad_norm": 3.375,
"learning_rate": 2.3292972157066785e-05,
"loss": 0.5873,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6780,
"tokens_per_second_per_gpu": 2517.71
},
{
"epoch": 0.6780507289794288,
"grad_norm": 3.328125,
"learning_rate": 2.3288021303987974e-05,
"loss": 0.5384,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6790,
"tokens_per_second_per_gpu": 2337.9
},
{
"epoch": 0.6790493309366886,
"grad_norm": 3.0625,
"learning_rate": 2.3283063809604848e-05,
"loss": 0.557,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6800,
"tokens_per_second_per_gpu": 2460.23
},
{
"epoch": 0.6800479328939485,
"grad_norm": 3.046875,
"learning_rate": 2.3278099676969325e-05,
"loss": 0.5572,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6810,
"tokens_per_second_per_gpu": 2422.17
},
{
"epoch": 0.6810465348512084,
"grad_norm": 3.140625,
"learning_rate": 2.327312890913742e-05,
"loss": 0.5814,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6820,
"tokens_per_second_per_gpu": 2366.69
},
{
"epoch": 0.6820451368084681,
"grad_norm": 3.734375,
"learning_rate": 2.3268151509169233e-05,
"loss": 0.6549,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6830,
"tokens_per_second_per_gpu": 2381.57
},
{
"epoch": 0.683043738765728,
"grad_norm": 2.546875,
"learning_rate": 2.3263167480128938e-05,
"loss": 0.4473,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6840,
"tokens_per_second_per_gpu": 2333.71
},
{
"epoch": 0.6840423407229879,
"grad_norm": 3.25,
"learning_rate": 2.32581768250848e-05,
"loss": 0.5636,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6850,
"tokens_per_second_per_gpu": 2466.48
},
{
"epoch": 0.6850409426802476,
"grad_norm": 4.40625,
"learning_rate": 2.325317954710915e-05,
"loss": 0.6141,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6860,
"tokens_per_second_per_gpu": 2504.67
},
{
"epoch": 0.6860395446375075,
"grad_norm": 3.890625,
"learning_rate": 2.3248175649278412e-05,
"loss": 0.5075,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6870,
"tokens_per_second_per_gpu": 2422.32
},
{
"epoch": 0.6870381465947674,
"grad_norm": 3.015625,
"learning_rate": 2.3243165134673074e-05,
"loss": 0.5325,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6880,
"tokens_per_second_per_gpu": 2311.97
},
{
"epoch": 0.6880367485520271,
"grad_norm": 2.84375,
"learning_rate": 2.3238148006377696e-05,
"loss": 0.5173,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6890,
"tokens_per_second_per_gpu": 2550.78
},
{
"epoch": 0.689035350509287,
"grad_norm": 3.09375,
"learning_rate": 2.3233124267480928e-05,
"loss": 0.5285,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6900,
"tokens_per_second_per_gpu": 2353.74
},
{
"epoch": 0.6900339524665469,
"grad_norm": 3.546875,
"learning_rate": 2.322809392107546e-05,
"loss": 0.559,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6910,
"tokens_per_second_per_gpu": 2332.07
},
{
"epoch": 0.6910325544238066,
"grad_norm": 2.953125,
"learning_rate": 2.3223056970258078e-05,
"loss": 0.5789,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6920,
"tokens_per_second_per_gpu": 2275.78
},
{
"epoch": 0.6920311563810665,
"grad_norm": 3.578125,
"learning_rate": 2.3218013418129618e-05,
"loss": 0.5572,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6930,
"tokens_per_second_per_gpu": 2492.71
},
{
"epoch": 0.6930297583383264,
"grad_norm": 2.953125,
"learning_rate": 2.321296326779498e-05,
"loss": 0.5665,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6940,
"tokens_per_second_per_gpu": 2536.83
},
{
"epoch": 0.6940283602955862,
"grad_norm": 2.71875,
"learning_rate": 2.3207906522363133e-05,
"loss": 0.5734,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6950,
"tokens_per_second_per_gpu": 2396.56
},
{
"epoch": 0.695026962252846,
"grad_norm": 3.25,
"learning_rate": 2.3202843184947105e-05,
"loss": 0.457,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6960,
"tokens_per_second_per_gpu": 2154.0
},
{
"epoch": 0.6960255642101059,
"grad_norm": 3.0,
"learning_rate": 2.319777325866397e-05,
"loss": 0.5339,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6970,
"tokens_per_second_per_gpu": 2304.08
},
{
"epoch": 0.6970241661673657,
"grad_norm": 3.515625,
"learning_rate": 2.3192696746634875e-05,
"loss": 0.5745,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6980,
"tokens_per_second_per_gpu": 2525.33
},
{
"epoch": 0.6980227681246255,
"grad_norm": 3.125,
"learning_rate": 2.3187613651985017e-05,
"loss": 0.5382,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 6990,
"tokens_per_second_per_gpu": 2364.81
},
{
"epoch": 0.6990213700818854,
"grad_norm": 2.890625,
"learning_rate": 2.3182523977843637e-05,
"loss": 0.4892,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7000,
"tokens_per_second_per_gpu": 2545.07
},
{
"epoch": 0.7000199720391452,
"grad_norm": 3.625,
"learning_rate": 2.317742772734403e-05,
"loss": 0.5606,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7010,
"tokens_per_second_per_gpu": 2281.94
},
{
"epoch": 0.701018573996405,
"grad_norm": 3.484375,
"learning_rate": 2.317232490362355e-05,
"loss": 0.5464,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7020,
"tokens_per_second_per_gpu": 2285.52
},
{
"epoch": 0.7020171759536649,
"grad_norm": 3.71875,
"learning_rate": 2.3167215509823583e-05,
"loss": 0.5433,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7030,
"tokens_per_second_per_gpu": 2402.18
},
{
"epoch": 0.7030157779109247,
"grad_norm": 3.140625,
"learning_rate": 2.3162099549089566e-05,
"loss": 0.5261,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7040,
"tokens_per_second_per_gpu": 2287.22
},
{
"epoch": 0.7040143798681845,
"grad_norm": 2.9375,
"learning_rate": 2.315697702457098e-05,
"loss": 0.5437,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7050,
"tokens_per_second_per_gpu": 2387.8
},
{
"epoch": 0.7050129818254444,
"grad_norm": 3.03125,
"learning_rate": 2.3151847939421348e-05,
"loss": 0.5964,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7060,
"tokens_per_second_per_gpu": 2282.69
},
{
"epoch": 0.7060115837827042,
"grad_norm": 3.328125,
"learning_rate": 2.3146712296798223e-05,
"loss": 0.5603,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7070,
"tokens_per_second_per_gpu": 2206.5
},
{
"epoch": 0.7070101857399641,
"grad_norm": 2.78125,
"learning_rate": 2.3141570099863206e-05,
"loss": 0.5188,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7080,
"tokens_per_second_per_gpu": 2288.83
},
{
"epoch": 0.7080087876972239,
"grad_norm": 2.921875,
"learning_rate": 2.3136421351781923e-05,
"loss": 0.5354,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7090,
"tokens_per_second_per_gpu": 2443.25
},
{
"epoch": 0.7090073896544837,
"grad_norm": 2.640625,
"learning_rate": 2.3131266055724045e-05,
"loss": 0.5757,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7100,
"tokens_per_second_per_gpu": 2288.69
},
{
"epoch": 0.7100059916117436,
"grad_norm": 2.484375,
"learning_rate": 2.312610421486326e-05,
"loss": 0.5682,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7110,
"tokens_per_second_per_gpu": 2410.45
},
{
"epoch": 0.7110045935690034,
"grad_norm": 3.84375,
"learning_rate": 2.31209358323773e-05,
"loss": 0.5417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7120,
"tokens_per_second_per_gpu": 2399.9
},
{
"epoch": 0.7120031955262632,
"grad_norm": 3.15625,
"learning_rate": 2.3115760911447906e-05,
"loss": 0.4968,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7130,
"tokens_per_second_per_gpu": 2361.98
},
{
"epoch": 0.7130017974835231,
"grad_norm": 2.765625,
"learning_rate": 2.311057945526086e-05,
"loss": 0.4429,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7140,
"tokens_per_second_per_gpu": 2553.69
},
{
"epoch": 0.7140003994407829,
"grad_norm": 3.78125,
"learning_rate": 2.3105391467005967e-05,
"loss": 0.536,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7150,
"tokens_per_second_per_gpu": 2363.76
},
{
"epoch": 0.7149990013980427,
"grad_norm": 2.265625,
"learning_rate": 2.3100196949877044e-05,
"loss": 0.4759,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7160,
"tokens_per_second_per_gpu": 2283.97
},
{
"epoch": 0.7159976033553026,
"grad_norm": 3.21875,
"learning_rate": 2.3094995907071925e-05,
"loss": 0.5372,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7170,
"tokens_per_second_per_gpu": 2337.5
},
{
"epoch": 0.7169962053125624,
"grad_norm": 3.953125,
"learning_rate": 2.3089788341792473e-05,
"loss": 0.5237,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7180,
"tokens_per_second_per_gpu": 2475.47
},
{
"epoch": 0.7179948072698222,
"grad_norm": 3.328125,
"learning_rate": 2.308457425724456e-05,
"loss": 0.5702,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7190,
"tokens_per_second_per_gpu": 2290.77
},
{
"epoch": 0.7189934092270821,
"grad_norm": 2.75,
"learning_rate": 2.307935365663808e-05,
"loss": 0.552,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7200,
"tokens_per_second_per_gpu": 2528.13
},
{
"epoch": 0.719992011184342,
"grad_norm": 2.765625,
"learning_rate": 2.307412654318692e-05,
"loss": 0.4913,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7210,
"tokens_per_second_per_gpu": 2366.5
},
{
"epoch": 0.7209906131416017,
"grad_norm": 3.015625,
"learning_rate": 2.306889292010899e-05,
"loss": 0.6276,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7220,
"tokens_per_second_per_gpu": 2369.08
},
{
"epoch": 0.7219892150988616,
"grad_norm": 2.75,
"learning_rate": 2.3063652790626207e-05,
"loss": 0.5006,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7230,
"tokens_per_second_per_gpu": 2635.17
},
{
"epoch": 0.7229878170561215,
"grad_norm": 3.21875,
"learning_rate": 2.3058406157964485e-05,
"loss": 0.5958,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7240,
"tokens_per_second_per_gpu": 2564.86
},
{
"epoch": 0.7239864190133812,
"grad_norm": 3.265625,
"learning_rate": 2.305315302535376e-05,
"loss": 0.5928,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7250,
"tokens_per_second_per_gpu": 2362.79
},
{
"epoch": 0.7249850209706411,
"grad_norm": 2.359375,
"learning_rate": 2.304789339602794e-05,
"loss": 0.5307,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7260,
"tokens_per_second_per_gpu": 2328.38
},
{
"epoch": 0.725983622927901,
"grad_norm": 3.75,
"learning_rate": 2.3042627273224967e-05,
"loss": 0.5656,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7270,
"tokens_per_second_per_gpu": 2415.82
},
{
"epoch": 0.7269822248851607,
"grad_norm": 2.828125,
"learning_rate": 2.3037354660186754e-05,
"loss": 0.5345,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7280,
"tokens_per_second_per_gpu": 2287.31
},
{
"epoch": 0.7279808268424206,
"grad_norm": 3.046875,
"learning_rate": 2.3032075560159216e-05,
"loss": 0.5189,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7290,
"tokens_per_second_per_gpu": 2276.4
},
{
"epoch": 0.7289794287996805,
"grad_norm": 3.8125,
"learning_rate": 2.302678997639227e-05,
"loss": 0.5787,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7300,
"tokens_per_second_per_gpu": 2392.92
},
{
"epoch": 0.7299780307569402,
"grad_norm": 2.796875,
"learning_rate": 2.3021497912139818e-05,
"loss": 0.5499,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7310,
"tokens_per_second_per_gpu": 2171.75
},
{
"epoch": 0.7309766327142001,
"grad_norm": 3.125,
"learning_rate": 2.3016199370659743e-05,
"loss": 0.4933,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7320,
"tokens_per_second_per_gpu": 2325.68
},
{
"epoch": 0.73197523467146,
"grad_norm": 3.09375,
"learning_rate": 2.3010894355213936e-05,
"loss": 0.6151,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7330,
"tokens_per_second_per_gpu": 2644.43
},
{
"epoch": 0.7329738366287198,
"grad_norm": 3.515625,
"learning_rate": 2.3005582869068258e-05,
"loss": 0.5448,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7340,
"tokens_per_second_per_gpu": 2217.24
},
{
"epoch": 0.7339724385859796,
"grad_norm": 2.5,
"learning_rate": 2.3000264915492558e-05,
"loss": 0.465,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7350,
"tokens_per_second_per_gpu": 2644.47
},
{
"epoch": 0.7349710405432395,
"grad_norm": 3.265625,
"learning_rate": 2.2994940497760665e-05,
"loss": 0.5792,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7360,
"tokens_per_second_per_gpu": 2276.74
},
{
"epoch": 0.7359696425004993,
"grad_norm": 3.921875,
"learning_rate": 2.2989609619150387e-05,
"loss": 0.6336,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7370,
"tokens_per_second_per_gpu": 2352.0
},
{
"epoch": 0.7369682444577591,
"grad_norm": 4.21875,
"learning_rate": 2.2984272282943515e-05,
"loss": 0.5448,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7380,
"tokens_per_second_per_gpu": 2319.48
},
{
"epoch": 0.737966846415019,
"grad_norm": 2.765625,
"learning_rate": 2.297892849242581e-05,
"loss": 0.5394,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7390,
"tokens_per_second_per_gpu": 2427.45
},
{
"epoch": 0.7389654483722788,
"grad_norm": 3.484375,
"learning_rate": 2.2973578250887008e-05,
"loss": 0.5814,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7400,
"tokens_per_second_per_gpu": 2458.18
},
{
"epoch": 0.7399640503295386,
"grad_norm": 2.671875,
"learning_rate": 2.296822156162081e-05,
"loss": 0.4813,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7410,
"tokens_per_second_per_gpu": 2411.48
},
{
"epoch": 0.7409626522867985,
"grad_norm": 3.125,
"learning_rate": 2.29628584279249e-05,
"loss": 0.5994,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7420,
"tokens_per_second_per_gpu": 2220.23
},
{
"epoch": 0.7419612542440583,
"grad_norm": 4.1875,
"learning_rate": 2.295748885310092e-05,
"loss": 0.4731,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7430,
"tokens_per_second_per_gpu": 2265.01
},
{
"epoch": 0.7429598562013181,
"grad_norm": 2.765625,
"learning_rate": 2.2952112840454476e-05,
"loss": 0.5564,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7440,
"tokens_per_second_per_gpu": 2495.87
},
{
"epoch": 0.743958458158578,
"grad_norm": 3.3125,
"learning_rate": 2.2946730393295145e-05,
"loss": 0.5205,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7450,
"tokens_per_second_per_gpu": 2418.8
},
{
"epoch": 0.7449570601158378,
"grad_norm": 3.6875,
"learning_rate": 2.2941341514936454e-05,
"loss": 0.4855,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7460,
"tokens_per_second_per_gpu": 2295.14
},
{
"epoch": 0.7459556620730977,
"grad_norm": 3.90625,
"learning_rate": 2.2935946208695902e-05,
"loss": 0.6616,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7470,
"tokens_per_second_per_gpu": 2271.94
},
{
"epoch": 0.7469542640303575,
"grad_norm": 4.0625,
"learning_rate": 2.2930544477894936e-05,
"loss": 0.5925,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7480,
"tokens_per_second_per_gpu": 2558.26
},
{
"epoch": 0.7479528659876173,
"grad_norm": 3.015625,
"learning_rate": 2.2925136325858962e-05,
"loss": 0.4796,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7490,
"tokens_per_second_per_gpu": 2490.03
},
{
"epoch": 0.7489514679448772,
"grad_norm": 3.890625,
"learning_rate": 2.2919721755917333e-05,
"loss": 0.6282,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7500,
"tokens_per_second_per_gpu": 2555.49
},
{
"epoch": 0.749950069902137,
"grad_norm": 2.453125,
"learning_rate": 2.291430077140337e-05,
"loss": 0.5501,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7510,
"tokens_per_second_per_gpu": 2397.4
},
{
"epoch": 0.7509486718593968,
"grad_norm": 3.34375,
"learning_rate": 2.2908873375654314e-05,
"loss": 0.5462,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7520,
"tokens_per_second_per_gpu": 2462.49
},
{
"epoch": 0.7519472738166567,
"grad_norm": 2.828125,
"learning_rate": 2.290343957201138e-05,
"loss": 0.5414,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7530,
"tokens_per_second_per_gpu": 2351.67
},
{
"epoch": 0.7529458757739165,
"grad_norm": 3.703125,
"learning_rate": 2.2897999363819716e-05,
"loss": 0.5669,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7540,
"tokens_per_second_per_gpu": 2431.55
},
{
"epoch": 0.7539444777311763,
"grad_norm": 3.046875,
"learning_rate": 2.2892552754428414e-05,
"loss": 0.5552,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7550,
"tokens_per_second_per_gpu": 2712.01
},
{
"epoch": 0.7549430796884362,
"grad_norm": 2.84375,
"learning_rate": 2.288709974719051e-05,
"loss": 0.5529,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7560,
"tokens_per_second_per_gpu": 2276.27
},
{
"epoch": 0.755941681645696,
"grad_norm": 3.40625,
"learning_rate": 2.2881640345462968e-05,
"loss": 0.5499,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7570,
"tokens_per_second_per_gpu": 2387.67
},
{
"epoch": 0.7569402836029558,
"grad_norm": 3.203125,
"learning_rate": 2.2876174552606702e-05,
"loss": 0.5291,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7580,
"tokens_per_second_per_gpu": 2293.15
},
{
"epoch": 0.7579388855602157,
"grad_norm": 3.078125,
"learning_rate": 2.2870702371986553e-05,
"loss": 0.497,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7590,
"tokens_per_second_per_gpu": 2539.85
},
{
"epoch": 0.7589374875174756,
"grad_norm": 4.03125,
"learning_rate": 2.2865223806971296e-05,
"loss": 0.5481,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7600,
"tokens_per_second_per_gpu": 2310.48
},
{
"epoch": 0.7599360894747353,
"grad_norm": 3.078125,
"learning_rate": 2.2859738860933637e-05,
"loss": 0.5349,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7610,
"tokens_per_second_per_gpu": 2495.11
},
{
"epoch": 0.7609346914319952,
"grad_norm": 2.203125,
"learning_rate": 2.2854247537250207e-05,
"loss": 0.5207,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7620,
"tokens_per_second_per_gpu": 2522.13
},
{
"epoch": 0.7619332933892551,
"grad_norm": 3.671875,
"learning_rate": 2.284874983930157e-05,
"loss": 0.5002,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7630,
"tokens_per_second_per_gpu": 2432.69
},
{
"epoch": 0.7629318953465148,
"grad_norm": 2.703125,
"learning_rate": 2.2843245770472206e-05,
"loss": 0.556,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7640,
"tokens_per_second_per_gpu": 2412.17
},
{
"epoch": 0.7639304973037747,
"grad_norm": 2.859375,
"learning_rate": 2.2837735334150522e-05,
"loss": 0.5373,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7650,
"tokens_per_second_per_gpu": 2462.07
},
{
"epoch": 0.7649290992610346,
"grad_norm": 3.109375,
"learning_rate": 2.283221853372885e-05,
"loss": 0.5064,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7660,
"tokens_per_second_per_gpu": 2462.24
},
{
"epoch": 0.7659277012182943,
"grad_norm": 3.75,
"learning_rate": 2.2826695372603423e-05,
"loss": 0.5419,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7670,
"tokens_per_second_per_gpu": 2535.41
},
{
"epoch": 0.7669263031755542,
"grad_norm": 2.40625,
"learning_rate": 2.282116585417441e-05,
"loss": 0.5544,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7680,
"tokens_per_second_per_gpu": 2502.21
},
{
"epoch": 0.7679249051328141,
"grad_norm": 3.109375,
"learning_rate": 2.2815629981845876e-05,
"loss": 0.5378,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7690,
"tokens_per_second_per_gpu": 2334.01
},
{
"epoch": 0.7689235070900738,
"grad_norm": 4.28125,
"learning_rate": 2.2810087759025816e-05,
"loss": 0.5439,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7700,
"tokens_per_second_per_gpu": 2448.19
},
{
"epoch": 0.7699221090473337,
"grad_norm": 2.796875,
"learning_rate": 2.2804539189126114e-05,
"loss": 0.5336,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7710,
"tokens_per_second_per_gpu": 2568.44
},
{
"epoch": 0.7709207110045936,
"grad_norm": 2.875,
"learning_rate": 2.279898427556258e-05,
"loss": 0.4611,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7720,
"tokens_per_second_per_gpu": 2355.85
},
{
"epoch": 0.7719193129618535,
"grad_norm": 2.859375,
"learning_rate": 2.279342302175491e-05,
"loss": 0.5298,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7730,
"tokens_per_second_per_gpu": 2523.22
},
{
"epoch": 0.7729179149191132,
"grad_norm": 3.65625,
"learning_rate": 2.2787855431126725e-05,
"loss": 0.5819,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7740,
"tokens_per_second_per_gpu": 2335.88
},
{
"epoch": 0.7739165168763731,
"grad_norm": 3.4375,
"learning_rate": 2.2782281507105536e-05,
"loss": 0.5648,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7750,
"tokens_per_second_per_gpu": 2541.94
},
{
"epoch": 0.774915118833633,
"grad_norm": 3.984375,
"learning_rate": 2.2776701253122746e-05,
"loss": 0.5873,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7760,
"tokens_per_second_per_gpu": 2538.55
},
{
"epoch": 0.7759137207908927,
"grad_norm": 4.15625,
"learning_rate": 2.2771114672613665e-05,
"loss": 0.5665,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7770,
"tokens_per_second_per_gpu": 2249.96
},
{
"epoch": 0.7769123227481526,
"grad_norm": 2.203125,
"learning_rate": 2.27655217690175e-05,
"loss": 0.5088,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7780,
"tokens_per_second_per_gpu": 2460.63
},
{
"epoch": 0.7779109247054125,
"grad_norm": 2.625,
"learning_rate": 2.2759922545777333e-05,
"loss": 0.5716,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7790,
"tokens_per_second_per_gpu": 2432.44
},
{
"epoch": 0.7789095266626722,
"grad_norm": 2.953125,
"learning_rate": 2.2754317006340163e-05,
"loss": 0.5654,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7800,
"tokens_per_second_per_gpu": 2439.62
},
{
"epoch": 0.7799081286199321,
"grad_norm": 3.015625,
"learning_rate": 2.2748705154156854e-05,
"loss": 0.5326,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7810,
"tokens_per_second_per_gpu": 2454.17
},
{
"epoch": 0.780906730577192,
"grad_norm": 2.84375,
"learning_rate": 2.2743086992682168e-05,
"loss": 0.5758,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7820,
"tokens_per_second_per_gpu": 2393.95
},
{
"epoch": 0.7819053325344517,
"grad_norm": 2.421875,
"learning_rate": 2.2737462525374747e-05,
"loss": 0.5469,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7830,
"tokens_per_second_per_gpu": 2462.24
},
{
"epoch": 0.7829039344917116,
"grad_norm": 4.0,
"learning_rate": 2.273183175569712e-05,
"loss": 0.5348,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7840,
"tokens_per_second_per_gpu": 2306.04
},
{
"epoch": 0.7839025364489715,
"grad_norm": 3.140625,
"learning_rate": 2.272619468711569e-05,
"loss": 0.5661,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7850,
"tokens_per_second_per_gpu": 2517.93
},
{
"epoch": 0.7849011384062313,
"grad_norm": 3.578125,
"learning_rate": 2.272055132310074e-05,
"loss": 0.5905,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7860,
"tokens_per_second_per_gpu": 2366.89
},
{
"epoch": 0.7858997403634911,
"grad_norm": 2.828125,
"learning_rate": 2.271490166712643e-05,
"loss": 0.5423,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7870,
"tokens_per_second_per_gpu": 2471.86
},
{
"epoch": 0.786898342320751,
"grad_norm": 3.359375,
"learning_rate": 2.270924572267079e-05,
"loss": 0.579,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7880,
"tokens_per_second_per_gpu": 2515.07
},
{
"epoch": 0.7878969442780108,
"grad_norm": 3.53125,
"learning_rate": 2.2703583493215726e-05,
"loss": 0.5856,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7890,
"tokens_per_second_per_gpu": 2386.43
},
{
"epoch": 0.7888955462352706,
"grad_norm": 2.953125,
"learning_rate": 2.269791498224701e-05,
"loss": 0.5594,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7900,
"tokens_per_second_per_gpu": 2442.99
},
{
"epoch": 0.7898941481925305,
"grad_norm": 2.25,
"learning_rate": 2.2692240193254276e-05,
"loss": 0.4759,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7910,
"tokens_per_second_per_gpu": 2473.48
},
{
"epoch": 0.7908927501497903,
"grad_norm": 3.171875,
"learning_rate": 2.268655912973104e-05,
"loss": 0.5096,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7920,
"tokens_per_second_per_gpu": 2379.89
},
{
"epoch": 0.7918913521070501,
"grad_norm": 2.90625,
"learning_rate": 2.268087179517466e-05,
"loss": 0.4953,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7930,
"tokens_per_second_per_gpu": 2336.16
},
{
"epoch": 0.79288995406431,
"grad_norm": 4.34375,
"learning_rate": 2.267517819308636e-05,
"loss": 0.5291,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7940,
"tokens_per_second_per_gpu": 2167.25
},
{
"epoch": 0.7938885560215698,
"grad_norm": 3.578125,
"learning_rate": 2.266947832697124e-05,
"loss": 0.514,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7950,
"tokens_per_second_per_gpu": 2383.76
},
{
"epoch": 0.7948871579788296,
"grad_norm": 2.71875,
"learning_rate": 2.2663772200338232e-05,
"loss": 0.5056,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7960,
"tokens_per_second_per_gpu": 2417.16
},
{
"epoch": 0.7958857599360895,
"grad_norm": 3.34375,
"learning_rate": 2.2658059816700135e-05,
"loss": 0.6493,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7970,
"tokens_per_second_per_gpu": 2381.21
},
{
"epoch": 0.7968843618933493,
"grad_norm": 2.828125,
"learning_rate": 2.26523411795736e-05,
"loss": 0.5632,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7980,
"tokens_per_second_per_gpu": 2692.3
},
{
"epoch": 0.7978829638506092,
"grad_norm": 3.3125,
"learning_rate": 2.2646616292479123e-05,
"loss": 0.5083,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 7990,
"tokens_per_second_per_gpu": 2592.2
},
{
"epoch": 0.798881565807869,
"grad_norm": 3.40625,
"learning_rate": 2.2640885158941048e-05,
"loss": 0.5337,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8000,
"tokens_per_second_per_gpu": 2286.79
},
{
"epoch": 0.7998801677651288,
"grad_norm": 3.421875,
"learning_rate": 2.2635147782487564e-05,
"loss": 0.5901,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8010,
"tokens_per_second_per_gpu": 2327.43
},
{
"epoch": 0.8008787697223887,
"grad_norm": 3.375,
"learning_rate": 2.2629404166650715e-05,
"loss": 0.5278,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8020,
"tokens_per_second_per_gpu": 2389.36
},
{
"epoch": 0.8018773716796485,
"grad_norm": 2.75,
"learning_rate": 2.2623654314966368e-05,
"loss": 0.5206,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8030,
"tokens_per_second_per_gpu": 2467.61
},
{
"epoch": 0.8028759736369083,
"grad_norm": 3.5625,
"learning_rate": 2.2617898230974243e-05,
"loss": 0.5809,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8040,
"tokens_per_second_per_gpu": 2528.73
},
{
"epoch": 0.8038745755941682,
"grad_norm": 2.5625,
"learning_rate": 2.261213591821789e-05,
"loss": 0.5971,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8050,
"tokens_per_second_per_gpu": 2309.69
},
{
"epoch": 0.804873177551428,
"grad_norm": 2.578125,
"learning_rate": 2.2606367380244688e-05,
"loss": 0.5802,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8060,
"tokens_per_second_per_gpu": 2442.32
},
{
"epoch": 0.8058717795086878,
"grad_norm": 2.796875,
"learning_rate": 2.2600592620605865e-05,
"loss": 0.6028,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8070,
"tokens_per_second_per_gpu": 2379.37
},
{
"epoch": 0.8068703814659477,
"grad_norm": 3.765625,
"learning_rate": 2.259481164285646e-05,
"loss": 0.5985,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8080,
"tokens_per_second_per_gpu": 2443.18
},
{
"epoch": 0.8078689834232075,
"grad_norm": 2.703125,
"learning_rate": 2.2589024450555357e-05,
"loss": 0.5633,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8090,
"tokens_per_second_per_gpu": 2492.64
},
{
"epoch": 0.8088675853804673,
"grad_norm": 3.71875,
"learning_rate": 2.258323104726525e-05,
"loss": 0.5813,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8100,
"tokens_per_second_per_gpu": 2350.55
},
{
"epoch": 0.8098661873377272,
"grad_norm": 3.6875,
"learning_rate": 2.2577431436552676e-05,
"loss": 0.5215,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8110,
"tokens_per_second_per_gpu": 2367.5
},
{
"epoch": 0.8108647892949871,
"grad_norm": 3.25,
"learning_rate": 2.2571625621987973e-05,
"loss": 0.5561,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8120,
"tokens_per_second_per_gpu": 2550.24
},
{
"epoch": 0.8118633912522468,
"grad_norm": 3.046875,
"learning_rate": 2.2565813607145308e-05,
"loss": 0.4721,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8130,
"tokens_per_second_per_gpu": 2424.45
},
{
"epoch": 0.8128619932095067,
"grad_norm": 3.28125,
"learning_rate": 2.255999539560267e-05,
"loss": 0.5161,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8140,
"tokens_per_second_per_gpu": 2517.62
},
{
"epoch": 0.8138605951667666,
"grad_norm": 3.953125,
"learning_rate": 2.255417099094185e-05,
"loss": 0.5448,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8150,
"tokens_per_second_per_gpu": 2434.71
},
{
"epoch": 0.8148591971240263,
"grad_norm": 2.671875,
"learning_rate": 2.254834039674846e-05,
"loss": 0.5371,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8160,
"tokens_per_second_per_gpu": 2211.14
},
{
"epoch": 0.8158577990812862,
"grad_norm": 3.078125,
"learning_rate": 2.2542503616611926e-05,
"loss": 0.5499,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8170,
"tokens_per_second_per_gpu": 2433.32
},
{
"epoch": 0.8168564010385461,
"grad_norm": 3.1875,
"learning_rate": 2.2536660654125467e-05,
"loss": 0.5142,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8180,
"tokens_per_second_per_gpu": 2347.09
},
{
"epoch": 0.8178550029958058,
"grad_norm": 2.515625,
"learning_rate": 2.2530811512886132e-05,
"loss": 0.505,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8190,
"tokens_per_second_per_gpu": 2487.67
},
{
"epoch": 0.8188536049530657,
"grad_norm": 2.859375,
"learning_rate": 2.2524956196494752e-05,
"loss": 0.5634,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8200,
"tokens_per_second_per_gpu": 2507.82
},
{
"epoch": 0.8198522069103256,
"grad_norm": 3.09375,
"learning_rate": 2.2519094708555965e-05,
"loss": 0.5937,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8210,
"tokens_per_second_per_gpu": 2200.55
},
{
"epoch": 0.8208508088675854,
"grad_norm": 3.09375,
"learning_rate": 2.2513227052678216e-05,
"loss": 0.476,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8220,
"tokens_per_second_per_gpu": 2341.89
},
{
"epoch": 0.8218494108248452,
"grad_norm": 3.109375,
"learning_rate": 2.2507353232473738e-05,
"loss": 0.5049,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8230,
"tokens_per_second_per_gpu": 2332.96
},
{
"epoch": 0.8228480127821051,
"grad_norm": 2.375,
"learning_rate": 2.2501473251558568e-05,
"loss": 0.5417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8240,
"tokens_per_second_per_gpu": 2504.4
},
{
"epoch": 0.8238466147393649,
"grad_norm": 2.953125,
"learning_rate": 2.2495587113552525e-05,
"loss": 0.5401,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8250,
"tokens_per_second_per_gpu": 2392.25
},
{
"epoch": 0.8248452166966247,
"grad_norm": 2.90625,
"learning_rate": 2.2489694822079227e-05,
"loss": 0.4868,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8260,
"tokens_per_second_per_gpu": 2354.64
},
{
"epoch": 0.8258438186538846,
"grad_norm": 3.578125,
"learning_rate": 2.2483796380766072e-05,
"loss": 0.5724,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8270,
"tokens_per_second_per_gpu": 2410.26
},
{
"epoch": 0.8268424206111444,
"grad_norm": 3.015625,
"learning_rate": 2.2477891793244257e-05,
"loss": 0.5375,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8280,
"tokens_per_second_per_gpu": 2465.44
},
{
"epoch": 0.8278410225684042,
"grad_norm": 2.765625,
"learning_rate": 2.247198106314875e-05,
"loss": 0.4739,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8290,
"tokens_per_second_per_gpu": 2475.41
},
{
"epoch": 0.8288396245256641,
"grad_norm": 3.078125,
"learning_rate": 2.24660641941183e-05,
"loss": 0.5512,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8300,
"tokens_per_second_per_gpu": 2497.96
},
{
"epoch": 0.8298382264829239,
"grad_norm": 3.03125,
"learning_rate": 2.2460141189795453e-05,
"loss": 0.4953,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8310,
"tokens_per_second_per_gpu": 2393.72
},
{
"epoch": 0.8308368284401837,
"grad_norm": 3.453125,
"learning_rate": 2.2454212053826513e-05,
"loss": 0.5644,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8320,
"tokens_per_second_per_gpu": 2346.15
},
{
"epoch": 0.8318354303974436,
"grad_norm": 2.375,
"learning_rate": 2.244827678986156e-05,
"loss": 0.5691,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8330,
"tokens_per_second_per_gpu": 2324.96
},
{
"epoch": 0.8328340323547034,
"grad_norm": 2.90625,
"learning_rate": 2.244233540155446e-05,
"loss": 0.5595,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8340,
"tokens_per_second_per_gpu": 2500.1
},
{
"epoch": 0.8338326343119633,
"grad_norm": 3.21875,
"learning_rate": 2.2436387892562834e-05,
"loss": 0.5794,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8350,
"tokens_per_second_per_gpu": 2307.01
},
{
"epoch": 0.8348312362692231,
"grad_norm": 3.171875,
"learning_rate": 2.243043426654808e-05,
"loss": 0.4888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8360,
"tokens_per_second_per_gpu": 2174.5
},
{
"epoch": 0.8358298382264829,
"grad_norm": 3.328125,
"learning_rate": 2.2424474527175364e-05,
"loss": 0.557,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8370,
"tokens_per_second_per_gpu": 2300.98
},
{
"epoch": 0.8368284401837428,
"grad_norm": 3.375,
"learning_rate": 2.2418508678113602e-05,
"loss": 0.5397,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8380,
"tokens_per_second_per_gpu": 2091.6
},
{
"epoch": 0.8378270421410026,
"grad_norm": 2.765625,
"learning_rate": 2.2412536723035494e-05,
"loss": 0.4914,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8390,
"tokens_per_second_per_gpu": 2439.76
},
{
"epoch": 0.8388256440982624,
"grad_norm": 3.4375,
"learning_rate": 2.2406558665617472e-05,
"loss": 0.5575,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8400,
"tokens_per_second_per_gpu": 2443.31
},
{
"epoch": 0.8398242460555223,
"grad_norm": 2.828125,
"learning_rate": 2.2400574509539746e-05,
"loss": 0.6244,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8410,
"tokens_per_second_per_gpu": 2517.69
},
{
"epoch": 0.8408228480127821,
"grad_norm": 3.0625,
"learning_rate": 2.239458425848627e-05,
"loss": 0.5573,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8420,
"tokens_per_second_per_gpu": 2431.48
},
{
"epoch": 0.8418214499700419,
"grad_norm": 3.1875,
"learning_rate": 2.2388587916144753e-05,
"loss": 0.5312,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8430,
"tokens_per_second_per_gpu": 2605.73
},
{
"epoch": 0.8428200519273018,
"grad_norm": 4.0625,
"learning_rate": 2.2382585486206656e-05,
"loss": 0.6492,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8440,
"tokens_per_second_per_gpu": 2422.86
},
{
"epoch": 0.8438186538845616,
"grad_norm": 2.5,
"learning_rate": 2.237657697236718e-05,
"loss": 0.5238,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8450,
"tokens_per_second_per_gpu": 2482.93
},
{
"epoch": 0.8448172558418214,
"grad_norm": 3.046875,
"learning_rate": 2.237056237832528e-05,
"loss": 0.5356,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8460,
"tokens_per_second_per_gpu": 2447.26
},
{
"epoch": 0.8458158577990813,
"grad_norm": 2.40625,
"learning_rate": 2.236454170778365e-05,
"loss": 0.5406,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8470,
"tokens_per_second_per_gpu": 2562.36
},
{
"epoch": 0.8468144597563412,
"grad_norm": 3.328125,
"learning_rate": 2.2358514964448727e-05,
"loss": 0.5014,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8480,
"tokens_per_second_per_gpu": 2476.69
},
{
"epoch": 0.8478130617136009,
"grad_norm": 3.0625,
"learning_rate": 2.2352482152030678e-05,
"loss": 0.6257,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8490,
"tokens_per_second_per_gpu": 2559.22
},
{
"epoch": 0.8488116636708608,
"grad_norm": 3.046875,
"learning_rate": 2.2346443274243428e-05,
"loss": 0.4589,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8500,
"tokens_per_second_per_gpu": 2378.61
},
{
"epoch": 0.8498102656281207,
"grad_norm": 2.796875,
"learning_rate": 2.2340398334804607e-05,
"loss": 0.5238,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8510,
"tokens_per_second_per_gpu": 2569.53
},
{
"epoch": 0.8508088675853804,
"grad_norm": 3.1875,
"learning_rate": 2.2334347337435598e-05,
"loss": 0.6102,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8520,
"tokens_per_second_per_gpu": 2400.04
},
{
"epoch": 0.8518074695426403,
"grad_norm": 3.0625,
"learning_rate": 2.2328290285861504e-05,
"loss": 0.6539,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8530,
"tokens_per_second_per_gpu": 2441.67
},
{
"epoch": 0.8528060714999002,
"grad_norm": 3.109375,
"learning_rate": 2.2322227183811156e-05,
"loss": 0.4745,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8540,
"tokens_per_second_per_gpu": 2390.26
},
{
"epoch": 0.8538046734571599,
"grad_norm": 3.109375,
"learning_rate": 2.231615803501711e-05,
"loss": 0.5554,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8550,
"tokens_per_second_per_gpu": 2391.85
},
{
"epoch": 0.8548032754144198,
"grad_norm": 2.921875,
"learning_rate": 2.2310082843215654e-05,
"loss": 0.5477,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8560,
"tokens_per_second_per_gpu": 2475.79
},
{
"epoch": 0.8558018773716797,
"grad_norm": 2.75,
"learning_rate": 2.230400161214678e-05,
"loss": 0.6008,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8570,
"tokens_per_second_per_gpu": 2216.04
},
{
"epoch": 0.8568004793289394,
"grad_norm": 3.21875,
"learning_rate": 2.229791434555421e-05,
"loss": 0.6535,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8580,
"tokens_per_second_per_gpu": 2365.15
},
{
"epoch": 0.8577990812861993,
"grad_norm": 3.46875,
"learning_rate": 2.2291821047185373e-05,
"loss": 0.5401,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8590,
"tokens_per_second_per_gpu": 2466.65
},
{
"epoch": 0.8587976832434592,
"grad_norm": 2.609375,
"learning_rate": 2.228572172079142e-05,
"loss": 0.5552,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8600,
"tokens_per_second_per_gpu": 2427.27
},
{
"epoch": 0.859796285200719,
"grad_norm": 3.34375,
"learning_rate": 2.2279616370127204e-05,
"loss": 0.5406,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8610,
"tokens_per_second_per_gpu": 2549.08
},
{
"epoch": 0.8607948871579788,
"grad_norm": 2.90625,
"learning_rate": 2.227350499895129e-05,
"loss": 0.5832,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8620,
"tokens_per_second_per_gpu": 2311.32
},
{
"epoch": 0.8617934891152387,
"grad_norm": 2.875,
"learning_rate": 2.226738761102596e-05,
"loss": 0.5782,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8630,
"tokens_per_second_per_gpu": 2314.0
},
{
"epoch": 0.8627920910724985,
"grad_norm": 2.84375,
"learning_rate": 2.2261264210117183e-05,
"loss": 0.5472,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8640,
"tokens_per_second_per_gpu": 2419.36
},
{
"epoch": 0.8637906930297583,
"grad_norm": 3.203125,
"learning_rate": 2.2255134799994636e-05,
"loss": 0.4888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8650,
"tokens_per_second_per_gpu": 2460.36
},
{
"epoch": 0.8647892949870182,
"grad_norm": 2.71875,
"learning_rate": 2.2248999384431704e-05,
"loss": 0.4911,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8660,
"tokens_per_second_per_gpu": 2442.51
},
{
"epoch": 0.865787896944278,
"grad_norm": 3.234375,
"learning_rate": 2.2242857967205454e-05,
"loss": 0.5802,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8670,
"tokens_per_second_per_gpu": 2148.28
},
{
"epoch": 0.8667864989015378,
"grad_norm": 2.859375,
"learning_rate": 2.2236710552096655e-05,
"loss": 0.5632,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8680,
"tokens_per_second_per_gpu": 2279.43
},
{
"epoch": 0.8677851008587977,
"grad_norm": 3.484375,
"learning_rate": 2.223055714288978e-05,
"loss": 0.6152,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8690,
"tokens_per_second_per_gpu": 2243.76
},
{
"epoch": 0.8687837028160575,
"grad_norm": 3.6875,
"learning_rate": 2.2224397743372966e-05,
"loss": 0.5889,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8700,
"tokens_per_second_per_gpu": 2294.03
},
{
"epoch": 0.8697823047733173,
"grad_norm": 4.1875,
"learning_rate": 2.2218232357338065e-05,
"loss": 0.5124,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8710,
"tokens_per_second_per_gpu": 2569.88
},
{
"epoch": 0.8707809067305772,
"grad_norm": 4.5,
"learning_rate": 2.221206098858059e-05,
"loss": 0.6064,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8720,
"tokens_per_second_per_gpu": 2492.79
},
{
"epoch": 0.871779508687837,
"grad_norm": 2.53125,
"learning_rate": 2.2205883640899762e-05,
"loss": 0.5355,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8730,
"tokens_per_second_per_gpu": 2384.89
},
{
"epoch": 0.8727781106450969,
"grad_norm": 3.53125,
"learning_rate": 2.219970031809846e-05,
"loss": 0.555,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8740,
"tokens_per_second_per_gpu": 2353.52
},
{
"epoch": 0.8737767126023567,
"grad_norm": 3.328125,
"learning_rate": 2.2193511023983254e-05,
"loss": 0.5759,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8750,
"tokens_per_second_per_gpu": 2463.04
},
{
"epoch": 0.8747753145596165,
"grad_norm": 3.0,
"learning_rate": 2.2187315762364392e-05,
"loss": 0.5653,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8760,
"tokens_per_second_per_gpu": 2370.01
},
{
"epoch": 0.8757739165168764,
"grad_norm": 3.125,
"learning_rate": 2.2181114537055784e-05,
"loss": 0.5424,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8770,
"tokens_per_second_per_gpu": 2537.85
},
{
"epoch": 0.8767725184741362,
"grad_norm": 3.484375,
"learning_rate": 2.217490735187502e-05,
"loss": 0.5139,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8780,
"tokens_per_second_per_gpu": 2351.31
},
{
"epoch": 0.877771120431396,
"grad_norm": 2.9375,
"learning_rate": 2.2168694210643362e-05,
"loss": 0.5176,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8790,
"tokens_per_second_per_gpu": 2283.67
},
{
"epoch": 0.8787697223886559,
"grad_norm": 2.75,
"learning_rate": 2.216247511718573e-05,
"loss": 0.5057,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8800,
"tokens_per_second_per_gpu": 2420.84
},
{
"epoch": 0.8797683243459157,
"grad_norm": 2.578125,
"learning_rate": 2.2156250075330712e-05,
"loss": 0.4969,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8810,
"tokens_per_second_per_gpu": 2627.5
},
{
"epoch": 0.8807669263031755,
"grad_norm": 3.484375,
"learning_rate": 2.2150019088910563e-05,
"loss": 0.5622,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8820,
"tokens_per_second_per_gpu": 2481.65
},
{
"epoch": 0.8817655282604354,
"grad_norm": 3.609375,
"learning_rate": 2.2143782161761182e-05,
"loss": 0.5784,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8830,
"tokens_per_second_per_gpu": 2233.78
},
{
"epoch": 0.8827641302176952,
"grad_norm": 2.75,
"learning_rate": 2.213753929772215e-05,
"loss": 0.5036,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8840,
"tokens_per_second_per_gpu": 2430.47
},
{
"epoch": 0.883762732174955,
"grad_norm": 3.046875,
"learning_rate": 2.2131290500636677e-05,
"loss": 0.5815,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8850,
"tokens_per_second_per_gpu": 2311.83
},
{
"epoch": 0.8847613341322149,
"grad_norm": 3.0625,
"learning_rate": 2.2125035774351647e-05,
"loss": 0.5618,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8860,
"tokens_per_second_per_gpu": 2477.31
},
{
"epoch": 0.8857599360894748,
"grad_norm": 3.5625,
"learning_rate": 2.211877512271758e-05,
"loss": 0.5482,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8870,
"tokens_per_second_per_gpu": 2460.59
},
{
"epoch": 0.8867585380467345,
"grad_norm": 3.484375,
"learning_rate": 2.2112508549588645e-05,
"loss": 0.5454,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8880,
"tokens_per_second_per_gpu": 2306.97
},
{
"epoch": 0.8877571400039944,
"grad_norm": 3.109375,
"learning_rate": 2.2106236058822664e-05,
"loss": 0.5654,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8890,
"tokens_per_second_per_gpu": 2512.2
},
{
"epoch": 0.8887557419612543,
"grad_norm": 3.859375,
"learning_rate": 2.2099957654281094e-05,
"loss": 0.5326,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8900,
"tokens_per_second_per_gpu": 2394.49
},
{
"epoch": 0.889754343918514,
"grad_norm": 3.015625,
"learning_rate": 2.2093673339829044e-05,
"loss": 0.5256,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8910,
"tokens_per_second_per_gpu": 2531.19
},
{
"epoch": 0.8907529458757739,
"grad_norm": 3.828125,
"learning_rate": 2.2087383119335248e-05,
"loss": 0.5619,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8920,
"tokens_per_second_per_gpu": 2409.18
},
{
"epoch": 0.8917515478330338,
"grad_norm": 2.984375,
"learning_rate": 2.2081086996672078e-05,
"loss": 0.4653,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8930,
"tokens_per_second_per_gpu": 2377.94
},
{
"epoch": 0.8927501497902935,
"grad_norm": 3.234375,
"learning_rate": 2.207478497571555e-05,
"loss": 0.6365,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8940,
"tokens_per_second_per_gpu": 2201.82
},
{
"epoch": 0.8937487517475534,
"grad_norm": 3.265625,
"learning_rate": 2.2068477060345298e-05,
"loss": 0.5398,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8950,
"tokens_per_second_per_gpu": 2527.55
},
{
"epoch": 0.8947473537048133,
"grad_norm": 2.9375,
"learning_rate": 2.2062163254444597e-05,
"loss": 0.5552,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8960,
"tokens_per_second_per_gpu": 2453.44
},
{
"epoch": 0.895745955662073,
"grad_norm": 4.0625,
"learning_rate": 2.205584356190034e-05,
"loss": 0.5452,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8970,
"tokens_per_second_per_gpu": 2167.25
},
{
"epoch": 0.8967445576193329,
"grad_norm": 3.09375,
"learning_rate": 2.204951798660304e-05,
"loss": 0.5775,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8980,
"tokens_per_second_per_gpu": 2394.48
},
{
"epoch": 0.8977431595765928,
"grad_norm": 3.4375,
"learning_rate": 2.2043186532446846e-05,
"loss": 0.518,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 8990,
"tokens_per_second_per_gpu": 2521.8
},
{
"epoch": 0.8987417615338527,
"grad_norm": 2.796875,
"learning_rate": 2.2036849203329513e-05,
"loss": 0.4625,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9000,
"tokens_per_second_per_gpu": 2284.03
},
{
"epoch": 0.8997403634911124,
"grad_norm": 2.96875,
"learning_rate": 2.203050600315241e-05,
"loss": 0.5177,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9010,
"tokens_per_second_per_gpu": 2416.24
},
{
"epoch": 0.9007389654483723,
"grad_norm": 3.5,
"learning_rate": 2.2024156935820548e-05,
"loss": 0.6324,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9020,
"tokens_per_second_per_gpu": 2478.94
},
{
"epoch": 0.9017375674056322,
"grad_norm": 3.265625,
"learning_rate": 2.201780200524251e-05,
"loss": 0.5737,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9030,
"tokens_per_second_per_gpu": 2455.49
},
{
"epoch": 0.9027361693628919,
"grad_norm": 3.515625,
"learning_rate": 2.201144121533052e-05,
"loss": 0.525,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9040,
"tokens_per_second_per_gpu": 2193.0
},
{
"epoch": 0.9037347713201518,
"grad_norm": 3.359375,
"learning_rate": 2.200507457000039e-05,
"loss": 0.4754,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9050,
"tokens_per_second_per_gpu": 2320.43
},
{
"epoch": 0.9047333732774117,
"grad_norm": 3.890625,
"learning_rate": 2.1998702073171546e-05,
"loss": 0.5258,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9060,
"tokens_per_second_per_gpu": 2342.26
},
{
"epoch": 0.9057319752346714,
"grad_norm": 3.890625,
"learning_rate": 2.1992323728767015e-05,
"loss": 0.4996,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9070,
"tokens_per_second_per_gpu": 2410.03
},
{
"epoch": 0.9067305771919313,
"grad_norm": 2.59375,
"learning_rate": 2.1985939540713425e-05,
"loss": 0.5307,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9080,
"tokens_per_second_per_gpu": 2339.14
},
{
"epoch": 0.9077291791491912,
"grad_norm": 2.65625,
"learning_rate": 2.1979549512940994e-05,
"loss": 0.5273,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9090,
"tokens_per_second_per_gpu": 2419.89
},
{
"epoch": 0.9087277811064509,
"grad_norm": 3.125,
"learning_rate": 2.1973153649383545e-05,
"loss": 0.6015,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9100,
"tokens_per_second_per_gpu": 2382.63
},
{
"epoch": 0.9097263830637108,
"grad_norm": 2.9375,
"learning_rate": 2.1966751953978494e-05,
"loss": 0.544,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9110,
"tokens_per_second_per_gpu": 2427.62
},
{
"epoch": 0.9107249850209707,
"grad_norm": 4.09375,
"learning_rate": 2.196034443066683e-05,
"loss": 0.5127,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9120,
"tokens_per_second_per_gpu": 2397.15
},
{
"epoch": 0.9117235869782305,
"grad_norm": 2.625,
"learning_rate": 2.1953931083393153e-05,
"loss": 0.6304,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9130,
"tokens_per_second_per_gpu": 2357.27
},
{
"epoch": 0.9127221889354903,
"grad_norm": 2.78125,
"learning_rate": 2.194751191610563e-05,
"loss": 0.5544,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9140,
"tokens_per_second_per_gpu": 2321.5
},
{
"epoch": 0.9137207908927502,
"grad_norm": 3.0,
"learning_rate": 2.1941086932756026e-05,
"loss": 0.546,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9150,
"tokens_per_second_per_gpu": 2600.74
},
{
"epoch": 0.91471939285001,
"grad_norm": 2.3125,
"learning_rate": 2.1934656137299675e-05,
"loss": 0.6103,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9160,
"tokens_per_second_per_gpu": 2372.1
},
{
"epoch": 0.9157179948072698,
"grad_norm": 2.40625,
"learning_rate": 2.192821953369549e-05,
"loss": 0.4995,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9170,
"tokens_per_second_per_gpu": 2437.88
},
{
"epoch": 0.9167165967645297,
"grad_norm": 3.3125,
"learning_rate": 2.1921777125905967e-05,
"loss": 0.5347,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9180,
"tokens_per_second_per_gpu": 2388.22
},
{
"epoch": 0.9177151987217895,
"grad_norm": 2.78125,
"learning_rate": 2.1915328917897167e-05,
"loss": 0.4701,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9190,
"tokens_per_second_per_gpu": 2188.95
},
{
"epoch": 0.9187138006790493,
"grad_norm": 2.65625,
"learning_rate": 2.1908874913638734e-05,
"loss": 0.5864,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9200,
"tokens_per_second_per_gpu": 2365.17
},
{
"epoch": 0.9197124026363092,
"grad_norm": 2.46875,
"learning_rate": 2.1902415117103857e-05,
"loss": 0.4752,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9210,
"tokens_per_second_per_gpu": 2374.77
},
{
"epoch": 0.920711004593569,
"grad_norm": 3.390625,
"learning_rate": 2.189594953226932e-05,
"loss": 0.557,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9220,
"tokens_per_second_per_gpu": 2277.15
},
{
"epoch": 0.9217096065508288,
"grad_norm": 2.90625,
"learning_rate": 2.1889478163115446e-05,
"loss": 0.534,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9230,
"tokens_per_second_per_gpu": 2457.7
},
{
"epoch": 0.9227082085080887,
"grad_norm": 2.046875,
"learning_rate": 2.1883001013626137e-05,
"loss": 0.5348,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9240,
"tokens_per_second_per_gpu": 2464.38
},
{
"epoch": 0.9237068104653485,
"grad_norm": 3.515625,
"learning_rate": 2.1876518087788835e-05,
"loss": 0.6236,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9250,
"tokens_per_second_per_gpu": 2340.68
},
{
"epoch": 0.9247054124226084,
"grad_norm": 3.296875,
"learning_rate": 2.187002938959456e-05,
"loss": 0.5318,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9260,
"tokens_per_second_per_gpu": 2377.95
},
{
"epoch": 0.9257040143798682,
"grad_norm": 4.0625,
"learning_rate": 2.1863534923037865e-05,
"loss": 0.5413,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9270,
"tokens_per_second_per_gpu": 2344.49
},
{
"epoch": 0.926702616337128,
"grad_norm": 3.734375,
"learning_rate": 2.1857034692116866e-05,
"loss": 0.5558,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9280,
"tokens_per_second_per_gpu": 2413.81
},
{
"epoch": 0.9277012182943879,
"grad_norm": 3.640625,
"learning_rate": 2.1850528700833228e-05,
"loss": 0.5688,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9290,
"tokens_per_second_per_gpu": 2539.54
},
{
"epoch": 0.9286998202516477,
"grad_norm": 2.390625,
"learning_rate": 2.1844016953192156e-05,
"loss": 0.5166,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9300,
"tokens_per_second_per_gpu": 2365.44
},
{
"epoch": 0.9296984222089075,
"grad_norm": 3.03125,
"learning_rate": 2.18374994532024e-05,
"loss": 0.5946,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9310,
"tokens_per_second_per_gpu": 2366.15
},
{
"epoch": 0.9306970241661674,
"grad_norm": 3.5625,
"learning_rate": 2.1830976204876253e-05,
"loss": 0.557,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9320,
"tokens_per_second_per_gpu": 2378.9
},
{
"epoch": 0.9316956261234272,
"grad_norm": 2.65625,
"learning_rate": 2.1824447212229552e-05,
"loss": 0.547,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9330,
"tokens_per_second_per_gpu": 2542.11
},
{
"epoch": 0.932694228080687,
"grad_norm": 2.546875,
"learning_rate": 2.1817912479281656e-05,
"loss": 0.4942,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9340,
"tokens_per_second_per_gpu": 2572.56
},
{
"epoch": 0.9336928300379469,
"grad_norm": 2.3125,
"learning_rate": 2.1811372010055476e-05,
"loss": 0.6083,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9350,
"tokens_per_second_per_gpu": 2451.85
},
{
"epoch": 0.9346914319952067,
"grad_norm": 3.15625,
"learning_rate": 2.1804825808577438e-05,
"loss": 0.5453,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9360,
"tokens_per_second_per_gpu": 2369.45
},
{
"epoch": 0.9356900339524665,
"grad_norm": 2.46875,
"learning_rate": 2.179827387887751e-05,
"loss": 0.4714,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9370,
"tokens_per_second_per_gpu": 2535.88
},
{
"epoch": 0.9366886359097264,
"grad_norm": 2.453125,
"learning_rate": 2.1791716224989177e-05,
"loss": 0.5206,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9380,
"tokens_per_second_per_gpu": 2423.01
},
{
"epoch": 0.9376872378669863,
"grad_norm": 3.796875,
"learning_rate": 2.178515285094945e-05,
"loss": 0.4871,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9390,
"tokens_per_second_per_gpu": 2283.89
},
{
"epoch": 0.938685839824246,
"grad_norm": 3.515625,
"learning_rate": 2.1778583760798866e-05,
"loss": 0.5968,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9400,
"tokens_per_second_per_gpu": 2457.73
},
{
"epoch": 0.9396844417815059,
"grad_norm": 2.421875,
"learning_rate": 2.1772008958581473e-05,
"loss": 0.5484,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9410,
"tokens_per_second_per_gpu": 2243.42
},
{
"epoch": 0.9406830437387658,
"grad_norm": 3.75,
"learning_rate": 2.176542844834485e-05,
"loss": 0.4997,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9420,
"tokens_per_second_per_gpu": 2249.54
},
{
"epoch": 0.9416816456960255,
"grad_norm": 2.390625,
"learning_rate": 2.1758842234140067e-05,
"loss": 0.6112,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9430,
"tokens_per_second_per_gpu": 2560.77
},
{
"epoch": 0.9426802476532854,
"grad_norm": 3.59375,
"learning_rate": 2.1752250320021728e-05,
"loss": 0.4795,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9440,
"tokens_per_second_per_gpu": 2440.34
},
{
"epoch": 0.9436788496105453,
"grad_norm": 3.25,
"learning_rate": 2.1745652710047933e-05,
"loss": 0.5529,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9450,
"tokens_per_second_per_gpu": 2426.9
},
{
"epoch": 0.944677451567805,
"grad_norm": 2.671875,
"learning_rate": 2.1739049408280293e-05,
"loss": 0.5781,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9460,
"tokens_per_second_per_gpu": 2385.4
},
{
"epoch": 0.9456760535250649,
"grad_norm": 2.453125,
"learning_rate": 2.173244041878392e-05,
"loss": 0.5026,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9470,
"tokens_per_second_per_gpu": 2223.65
},
{
"epoch": 0.9466746554823248,
"grad_norm": 2.78125,
"learning_rate": 2.1725825745627433e-05,
"loss": 0.5115,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9480,
"tokens_per_second_per_gpu": 2264.93
},
{
"epoch": 0.9476732574395845,
"grad_norm": 2.984375,
"learning_rate": 2.1719205392882945e-05,
"loss": 0.5663,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9490,
"tokens_per_second_per_gpu": 2282.33
},
{
"epoch": 0.9486718593968444,
"grad_norm": 2.921875,
"learning_rate": 2.1712579364626066e-05,
"loss": 0.5415,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9500,
"tokens_per_second_per_gpu": 2309.61
},
{
"epoch": 0.9496704613541043,
"grad_norm": 3.46875,
"learning_rate": 2.17059476649359e-05,
"loss": 0.6023,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9510,
"tokens_per_second_per_gpu": 2457.7
},
{
"epoch": 0.9506690633113641,
"grad_norm": 3.109375,
"learning_rate": 2.1699310297895045e-05,
"loss": 0.5255,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9520,
"tokens_per_second_per_gpu": 2374.72
},
{
"epoch": 0.9516676652686239,
"grad_norm": 2.734375,
"learning_rate": 2.169266726758959e-05,
"loss": 0.5186,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9530,
"tokens_per_second_per_gpu": 2440.49
},
{
"epoch": 0.9526662672258838,
"grad_norm": 3.703125,
"learning_rate": 2.1686018578109095e-05,
"loss": 0.5266,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9540,
"tokens_per_second_per_gpu": 2157.93
},
{
"epoch": 0.9536648691831436,
"grad_norm": 3.296875,
"learning_rate": 2.1679364233546627e-05,
"loss": 0.55,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9550,
"tokens_per_second_per_gpu": 2556.39
},
{
"epoch": 0.9546634711404034,
"grad_norm": 3.34375,
"learning_rate": 2.167270423799872e-05,
"loss": 0.5439,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9560,
"tokens_per_second_per_gpu": 2296.71
},
{
"epoch": 0.9556620730976633,
"grad_norm": 3.078125,
"learning_rate": 2.166603859556539e-05,
"loss": 0.5795,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9570,
"tokens_per_second_per_gpu": 2286.58
},
{
"epoch": 0.9566606750549231,
"grad_norm": 3.265625,
"learning_rate": 2.1659367310350125e-05,
"loss": 0.5357,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9580,
"tokens_per_second_per_gpu": 2493.23
},
{
"epoch": 0.9576592770121829,
"grad_norm": 2.65625,
"learning_rate": 2.1652690386459896e-05,
"loss": 0.4871,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9590,
"tokens_per_second_per_gpu": 2258.97
},
{
"epoch": 0.9586578789694428,
"grad_norm": 3.71875,
"learning_rate": 2.1646007828005133e-05,
"loss": 0.5311,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9600,
"tokens_per_second_per_gpu": 2364.71
},
{
"epoch": 0.9596564809267026,
"grad_norm": 3.65625,
"learning_rate": 2.1639319639099752e-05,
"loss": 0.5891,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9610,
"tokens_per_second_per_gpu": 2369.06
},
{
"epoch": 0.9606550828839625,
"grad_norm": 3.25,
"learning_rate": 2.1632625823861116e-05,
"loss": 0.5249,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9620,
"tokens_per_second_per_gpu": 2427.93
},
{
"epoch": 0.9616536848412223,
"grad_norm": 3.609375,
"learning_rate": 2.1625926386410068e-05,
"loss": 0.515,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9630,
"tokens_per_second_per_gpu": 2442.08
},
{
"epoch": 0.9626522867984821,
"grad_norm": 3.40625,
"learning_rate": 2.16192213308709e-05,
"loss": 0.5134,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9640,
"tokens_per_second_per_gpu": 2447.24
},
{
"epoch": 0.963650888755742,
"grad_norm": 2.171875,
"learning_rate": 2.1612510661371366e-05,
"loss": 0.4723,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9650,
"tokens_per_second_per_gpu": 2490.38
},
{
"epoch": 0.9646494907130018,
"grad_norm": 2.546875,
"learning_rate": 2.1605794382042684e-05,
"loss": 0.5397,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9660,
"tokens_per_second_per_gpu": 2390.34
},
{
"epoch": 0.9656480926702616,
"grad_norm": 3.0625,
"learning_rate": 2.1599072497019514e-05,
"loss": 0.5669,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9670,
"tokens_per_second_per_gpu": 2287.19
},
{
"epoch": 0.9666466946275215,
"grad_norm": 3.703125,
"learning_rate": 2.1592345010439972e-05,
"loss": 0.5543,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9680,
"tokens_per_second_per_gpu": 2241.42
},
{
"epoch": 0.9676452965847813,
"grad_norm": 4.65625,
"learning_rate": 2.1585611926445626e-05,
"loss": 0.5612,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9690,
"tokens_per_second_per_gpu": 2413.12
},
{
"epoch": 0.9686438985420411,
"grad_norm": 2.8125,
"learning_rate": 2.1578873249181484e-05,
"loss": 0.402,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9700,
"tokens_per_second_per_gpu": 2354.2
},
{
"epoch": 0.969642500499301,
"grad_norm": 2.53125,
"learning_rate": 2.1572128982796e-05,
"loss": 0.4903,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9710,
"tokens_per_second_per_gpu": 2437.67
},
{
"epoch": 0.9706411024565608,
"grad_norm": 3.578125,
"learning_rate": 2.1565379131441073e-05,
"loss": 0.6315,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9720,
"tokens_per_second_per_gpu": 2311.3
},
{
"epoch": 0.9716397044138206,
"grad_norm": 2.53125,
"learning_rate": 2.1558623699272024e-05,
"loss": 0.4885,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9730,
"tokens_per_second_per_gpu": 2120.68
},
{
"epoch": 0.9726383063710805,
"grad_norm": 3.125,
"learning_rate": 2.155186269044763e-05,
"loss": 0.5378,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9740,
"tokens_per_second_per_gpu": 2343.54
},
{
"epoch": 0.9736369083283404,
"grad_norm": 3.09375,
"learning_rate": 2.1545096109130092e-05,
"loss": 0.461,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9750,
"tokens_per_second_per_gpu": 2352.78
},
{
"epoch": 0.9746355102856001,
"grad_norm": 3.109375,
"learning_rate": 2.153832395948504e-05,
"loss": 0.5164,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9760,
"tokens_per_second_per_gpu": 2444.0
},
{
"epoch": 0.97563411224286,
"grad_norm": 3.40625,
"learning_rate": 2.1531546245681535e-05,
"loss": 0.4671,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9770,
"tokens_per_second_per_gpu": 2534.83
},
{
"epoch": 0.9766327142001199,
"grad_norm": 4.0625,
"learning_rate": 2.1524762971892065e-05,
"loss": 0.5014,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9780,
"tokens_per_second_per_gpu": 2354.77
},
{
"epoch": 0.9776313161573796,
"grad_norm": 2.515625,
"learning_rate": 2.1517974142292528e-05,
"loss": 0.5193,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9790,
"tokens_per_second_per_gpu": 2571.65
},
{
"epoch": 0.9786299181146395,
"grad_norm": 2.4375,
"learning_rate": 2.1511179761062265e-05,
"loss": 0.542,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9800,
"tokens_per_second_per_gpu": 2262.65
},
{
"epoch": 0.9796285200718994,
"grad_norm": 2.953125,
"learning_rate": 2.1504379832384013e-05,
"loss": 0.6133,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9810,
"tokens_per_second_per_gpu": 2458.48
},
{
"epoch": 0.9806271220291591,
"grad_norm": 3.390625,
"learning_rate": 2.149757436044394e-05,
"loss": 0.5826,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9820,
"tokens_per_second_per_gpu": 2483.31
},
{
"epoch": 0.981625723986419,
"grad_norm": 2.828125,
"learning_rate": 2.1490763349431614e-05,
"loss": 0.5197,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9830,
"tokens_per_second_per_gpu": 2479.53
},
{
"epoch": 0.9826243259436789,
"grad_norm": 2.578125,
"learning_rate": 2.148394680354002e-05,
"loss": 0.4671,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9840,
"tokens_per_second_per_gpu": 2380.68
},
{
"epoch": 0.9836229279009386,
"grad_norm": 3.96875,
"learning_rate": 2.1477124726965553e-05,
"loss": 0.553,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9850,
"tokens_per_second_per_gpu": 2403.42
},
{
"epoch": 0.9846215298581985,
"grad_norm": 3.078125,
"learning_rate": 2.1470297123908e-05,
"loss": 0.5949,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9860,
"tokens_per_second_per_gpu": 2272.79
},
{
"epoch": 0.9856201318154584,
"grad_norm": 3.25,
"learning_rate": 2.1463463998570568e-05,
"loss": 0.5343,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9870,
"tokens_per_second_per_gpu": 2432.03
},
{
"epoch": 0.9866187337727182,
"grad_norm": 2.953125,
"learning_rate": 2.145662535515985e-05,
"loss": 0.5676,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9880,
"tokens_per_second_per_gpu": 2302.1
},
{
"epoch": 0.987617335729978,
"grad_norm": 2.796875,
"learning_rate": 2.1449781197885843e-05,
"loss": 0.5889,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9890,
"tokens_per_second_per_gpu": 2314.78
},
{
"epoch": 0.9886159376872379,
"grad_norm": 3.5625,
"learning_rate": 2.1442931530961935e-05,
"loss": 0.4797,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9900,
"tokens_per_second_per_gpu": 2235.2
},
{
"epoch": 0.9896145396444977,
"grad_norm": 3.734375,
"learning_rate": 2.1436076358604907e-05,
"loss": 0.5592,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9910,
"tokens_per_second_per_gpu": 2326.54
},
{
"epoch": 0.9906131416017575,
"grad_norm": 2.34375,
"learning_rate": 2.1429215685034927e-05,
"loss": 0.5923,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9920,
"tokens_per_second_per_gpu": 2404.87
},
{
"epoch": 0.9916117435590174,
"grad_norm": 3.234375,
"learning_rate": 2.1422349514475558e-05,
"loss": 0.526,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9930,
"tokens_per_second_per_gpu": 2342.15
},
{
"epoch": 0.9926103455162772,
"grad_norm": 3.265625,
"learning_rate": 2.1415477851153734e-05,
"loss": 0.5268,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9940,
"tokens_per_second_per_gpu": 2465.9
},
{
"epoch": 0.993608947473537,
"grad_norm": 3.546875,
"learning_rate": 2.140860069929978e-05,
"loss": 0.4687,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9950,
"tokens_per_second_per_gpu": 2529.89
},
{
"epoch": 0.9946075494307969,
"grad_norm": 2.578125,
"learning_rate": 2.1401718063147395e-05,
"loss": 0.5186,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9960,
"tokens_per_second_per_gpu": 2420.36
},
{
"epoch": 0.9956061513880567,
"grad_norm": 2.171875,
"learning_rate": 2.1394829946933663e-05,
"loss": 0.4581,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9970,
"tokens_per_second_per_gpu": 2447.72
},
{
"epoch": 0.9966047533453165,
"grad_norm": 3.25,
"learning_rate": 2.138793635489903e-05,
"loss": 0.5183,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9980,
"tokens_per_second_per_gpu": 2350.04
},
{
"epoch": 0.9976033553025764,
"grad_norm": 2.390625,
"learning_rate": 2.1381037291287318e-05,
"loss": 0.493,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 9990,
"tokens_per_second_per_gpu": 2590.86
},
{
"epoch": 0.9986019572598362,
"grad_norm": 2.53125,
"learning_rate": 2.1374132760345715e-05,
"loss": 0.4775,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10000,
"tokens_per_second_per_gpu": 2330.85
},
{
"epoch": 0.9996005592170961,
"grad_norm": 4.625,
"learning_rate": 2.1367222766324786e-05,
"loss": 0.6381,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10010,
"tokens_per_second_per_gpu": 2226.06
},
{
"epoch": 1.0005991611743559,
"grad_norm": 2.34375,
"learning_rate": 2.136030731347844e-05,
"loss": 0.4756,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10020,
"tokens_per_second_per_gpu": 2485.65
},
{
"epoch": 1.0015977631316157,
"grad_norm": 2.9375,
"learning_rate": 2.1353386406063962e-05,
"loss": 0.3821,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10030,
"tokens_per_second_per_gpu": 2410.67
},
{
"epoch": 1.0025963650888756,
"grad_norm": 3.703125,
"learning_rate": 2.1346460048341993e-05,
"loss": 0.3844,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10040,
"tokens_per_second_per_gpu": 2553.94
},
{
"epoch": 1.0035949670461355,
"grad_norm": 3.015625,
"learning_rate": 2.1339528244576523e-05,
"loss": 0.3741,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10050,
"tokens_per_second_per_gpu": 2638.91
},
{
"epoch": 1.0045935690033951,
"grad_norm": 2.640625,
"learning_rate": 2.1332590999034896e-05,
"loss": 0.3115,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10060,
"tokens_per_second_per_gpu": 2509.38
},
{
"epoch": 1.005592170960655,
"grad_norm": 3.140625,
"learning_rate": 2.1325648315987805e-05,
"loss": 0.3792,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10070,
"tokens_per_second_per_gpu": 2604.58
},
{
"epoch": 1.0065907729179149,
"grad_norm": 3.40625,
"learning_rate": 2.13187001997093e-05,
"loss": 0.4351,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10080,
"tokens_per_second_per_gpu": 2477.32
},
{
"epoch": 1.0075893748751747,
"grad_norm": 2.8125,
"learning_rate": 2.131174665447677e-05,
"loss": 0.3877,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10090,
"tokens_per_second_per_gpu": 2750.35
},
{
"epoch": 1.0085879768324346,
"grad_norm": 3.015625,
"learning_rate": 2.1304787684570945e-05,
"loss": 0.3788,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10100,
"tokens_per_second_per_gpu": 2480.64
},
{
"epoch": 1.0095865787896945,
"grad_norm": 2.90625,
"learning_rate": 2.1297823294275886e-05,
"loss": 0.4102,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10110,
"tokens_per_second_per_gpu": 2562.26
},
{
"epoch": 1.0105851807469544,
"grad_norm": 2.53125,
"learning_rate": 2.1290853487879015e-05,
"loss": 0.4066,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10120,
"tokens_per_second_per_gpu": 2552.74
},
{
"epoch": 1.011583782704214,
"grad_norm": 3.640625,
"learning_rate": 2.1283878269671056e-05,
"loss": 0.4273,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10130,
"tokens_per_second_per_gpu": 2578.49
},
{
"epoch": 1.0125823846614739,
"grad_norm": 3.9375,
"learning_rate": 2.1276897643946093e-05,
"loss": 0.382,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10140,
"tokens_per_second_per_gpu": 2394.58
},
{
"epoch": 1.0135809866187337,
"grad_norm": 3.125,
"learning_rate": 2.1269911615001524e-05,
"loss": 0.372,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10150,
"tokens_per_second_per_gpu": 2569.02
},
{
"epoch": 1.0145795885759936,
"grad_norm": 3.078125,
"learning_rate": 2.1262920187138073e-05,
"loss": 0.3501,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10160,
"tokens_per_second_per_gpu": 2518.36
},
{
"epoch": 1.0155781905332535,
"grad_norm": 3.046875,
"learning_rate": 2.1255923364659804e-05,
"loss": 0.3756,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10170,
"tokens_per_second_per_gpu": 2605.58
},
{
"epoch": 1.0165767924905134,
"grad_norm": 3.34375,
"learning_rate": 2.1248921151874075e-05,
"loss": 0.4022,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10180,
"tokens_per_second_per_gpu": 2403.47
},
{
"epoch": 1.017575394447773,
"grad_norm": 2.578125,
"learning_rate": 2.1241913553091582e-05,
"loss": 0.3811,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10190,
"tokens_per_second_per_gpu": 2495.42
},
{
"epoch": 1.0185739964050329,
"grad_norm": 3.359375,
"learning_rate": 2.1234900572626337e-05,
"loss": 0.3255,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10200,
"tokens_per_second_per_gpu": 2452.39
},
{
"epoch": 1.0195725983622927,
"grad_norm": 3.0,
"learning_rate": 2.1227882214795654e-05,
"loss": 0.3567,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10210,
"tokens_per_second_per_gpu": 2444.51
},
{
"epoch": 1.0205712003195526,
"grad_norm": 3.203125,
"learning_rate": 2.1220858483920166e-05,
"loss": 0.3636,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10220,
"tokens_per_second_per_gpu": 2307.21
},
{
"epoch": 1.0215698022768125,
"grad_norm": 3.421875,
"learning_rate": 2.1213829384323808e-05,
"loss": 0.3647,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10230,
"tokens_per_second_per_gpu": 2589.07
},
{
"epoch": 1.0225684042340724,
"grad_norm": 3.59375,
"learning_rate": 2.120679492033383e-05,
"loss": 0.4198,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10240,
"tokens_per_second_per_gpu": 2441.03
},
{
"epoch": 1.0235670061913322,
"grad_norm": 2.5625,
"learning_rate": 2.1199755096280765e-05,
"loss": 0.3701,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10250,
"tokens_per_second_per_gpu": 2374.73
},
{
"epoch": 1.0245656081485919,
"grad_norm": 3.65625,
"learning_rate": 2.119270991649847e-05,
"loss": 0.3654,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10260,
"tokens_per_second_per_gpu": 2548.94
},
{
"epoch": 1.0255642101058517,
"grad_norm": 3.3125,
"learning_rate": 2.118565938532408e-05,
"loss": 0.3376,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10270,
"tokens_per_second_per_gpu": 2459.32
},
{
"epoch": 1.0265628120631116,
"grad_norm": 3.078125,
"learning_rate": 2.117860350709803e-05,
"loss": 0.391,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10280,
"tokens_per_second_per_gpu": 2472.67
},
{
"epoch": 1.0275614140203715,
"grad_norm": 2.96875,
"learning_rate": 2.1171542286164064e-05,
"loss": 0.3786,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10290,
"tokens_per_second_per_gpu": 2562.9
},
{
"epoch": 1.0285600159776314,
"grad_norm": 2.84375,
"learning_rate": 2.1164475726869183e-05,
"loss": 0.3793,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10300,
"tokens_per_second_per_gpu": 2413.79
},
{
"epoch": 1.0295586179348912,
"grad_norm": 2.703125,
"learning_rate": 2.1157403833563694e-05,
"loss": 0.3668,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10310,
"tokens_per_second_per_gpu": 2632.13
},
{
"epoch": 1.030557219892151,
"grad_norm": 2.8125,
"learning_rate": 2.1150326610601187e-05,
"loss": 0.3414,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10320,
"tokens_per_second_per_gpu": 2640.21
},
{
"epoch": 1.0315558218494107,
"grad_norm": 3.1875,
"learning_rate": 2.114324406233853e-05,
"loss": 0.3792,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10330,
"tokens_per_second_per_gpu": 2489.73
},
{
"epoch": 1.0325544238066706,
"grad_norm": 3.234375,
"learning_rate": 2.113615619313587e-05,
"loss": 0.3765,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10340,
"tokens_per_second_per_gpu": 2450.85
},
{
"epoch": 1.0335530257639305,
"grad_norm": 2.9375,
"learning_rate": 2.112906300735663e-05,
"loss": 0.3234,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10350,
"tokens_per_second_per_gpu": 2453.46
},
{
"epoch": 1.0345516277211904,
"grad_norm": 2.796875,
"learning_rate": 2.112196450936751e-05,
"loss": 0.3445,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10360,
"tokens_per_second_per_gpu": 2693.76
},
{
"epoch": 1.0355502296784502,
"grad_norm": 3.21875,
"learning_rate": 2.1114860703538465e-05,
"loss": 0.3966,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10370,
"tokens_per_second_per_gpu": 2466.44
},
{
"epoch": 1.03654883163571,
"grad_norm": 2.890625,
"learning_rate": 2.1107751594242738e-05,
"loss": 0.3355,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10380,
"tokens_per_second_per_gpu": 2462.78
},
{
"epoch": 1.0375474335929697,
"grad_norm": 4.15625,
"learning_rate": 2.1100637185856827e-05,
"loss": 0.3952,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10390,
"tokens_per_second_per_gpu": 2437.3
},
{
"epoch": 1.0385460355502296,
"grad_norm": 3.3125,
"learning_rate": 2.1093517482760483e-05,
"loss": 0.4243,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10400,
"tokens_per_second_per_gpu": 2404.75
},
{
"epoch": 1.0395446375074895,
"grad_norm": 2.875,
"learning_rate": 2.1086392489336738e-05,
"loss": 0.3759,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10410,
"tokens_per_second_per_gpu": 2410.51
},
{
"epoch": 1.0405432394647494,
"grad_norm": 2.875,
"learning_rate": 2.107926220997186e-05,
"loss": 0.4327,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10420,
"tokens_per_second_per_gpu": 2532.5
},
{
"epoch": 1.0415418414220092,
"grad_norm": 3.34375,
"learning_rate": 2.1072126649055386e-05,
"loss": 0.3774,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10430,
"tokens_per_second_per_gpu": 2431.3
},
{
"epoch": 1.042540443379269,
"grad_norm": 3.578125,
"learning_rate": 2.10649858109801e-05,
"loss": 0.4126,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10440,
"tokens_per_second_per_gpu": 2400.68
},
{
"epoch": 1.0435390453365287,
"grad_norm": 3.203125,
"learning_rate": 2.1057839700142025e-05,
"loss": 0.3563,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10450,
"tokens_per_second_per_gpu": 2546.33
},
{
"epoch": 1.0445376472937886,
"grad_norm": 3.203125,
"learning_rate": 2.1050688320940447e-05,
"loss": 0.3703,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10460,
"tokens_per_second_per_gpu": 2335.57
},
{
"epoch": 1.0455362492510485,
"grad_norm": 3.484375,
"learning_rate": 2.104353167777788e-05,
"loss": 0.3584,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10470,
"tokens_per_second_per_gpu": 2520.62
},
{
"epoch": 1.0465348512083084,
"grad_norm": 3.4375,
"learning_rate": 2.103636977506009e-05,
"loss": 0.3826,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10480,
"tokens_per_second_per_gpu": 2446.27
},
{
"epoch": 1.0475334531655682,
"grad_norm": 3.09375,
"learning_rate": 2.1029202617196074e-05,
"loss": 0.3371,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10490,
"tokens_per_second_per_gpu": 2393.9
},
{
"epoch": 1.048532055122828,
"grad_norm": 3.03125,
"learning_rate": 2.102203020859806e-05,
"loss": 0.3923,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10500,
"tokens_per_second_per_gpu": 2445.86
},
{
"epoch": 1.049530657080088,
"grad_norm": 2.546875,
"learning_rate": 2.1014852553681527e-05,
"loss": 0.3713,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10510,
"tokens_per_second_per_gpu": 2533.71
},
{
"epoch": 1.0505292590373476,
"grad_norm": 2.890625,
"learning_rate": 2.1007669656865164e-05,
"loss": 0.3394,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10520,
"tokens_per_second_per_gpu": 2500.74
},
{
"epoch": 1.0515278609946075,
"grad_norm": 3.09375,
"learning_rate": 2.1000481522570896e-05,
"loss": 0.3918,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10530,
"tokens_per_second_per_gpu": 2563.53
},
{
"epoch": 1.0525264629518674,
"grad_norm": 2.90625,
"learning_rate": 2.0993288155223868e-05,
"loss": 0.3635,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10540,
"tokens_per_second_per_gpu": 2532.3
},
{
"epoch": 1.0535250649091272,
"grad_norm": 4.78125,
"learning_rate": 2.0986089559252452e-05,
"loss": 0.3708,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10550,
"tokens_per_second_per_gpu": 2270.27
},
{
"epoch": 1.054523666866387,
"grad_norm": 2.875,
"learning_rate": 2.097888573908824e-05,
"loss": 0.3612,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10560,
"tokens_per_second_per_gpu": 2444.8
},
{
"epoch": 1.055522268823647,
"grad_norm": 3.140625,
"learning_rate": 2.0971676699166025e-05,
"loss": 0.4002,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10570,
"tokens_per_second_per_gpu": 2692.28
},
{
"epoch": 1.0565208707809068,
"grad_norm": 3.421875,
"learning_rate": 2.0964462443923837e-05,
"loss": 0.3633,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10580,
"tokens_per_second_per_gpu": 2502.53
},
{
"epoch": 1.0575194727381665,
"grad_norm": 3.796875,
"learning_rate": 2.0957242977802893e-05,
"loss": 0.3816,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10590,
"tokens_per_second_per_gpu": 2478.87
},
{
"epoch": 1.0585180746954264,
"grad_norm": 2.90625,
"learning_rate": 2.095001830524764e-05,
"loss": 0.4351,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10600,
"tokens_per_second_per_gpu": 2647.35
},
{
"epoch": 1.0595166766526862,
"grad_norm": 3.328125,
"learning_rate": 2.0942788430705706e-05,
"loss": 0.3763,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10610,
"tokens_per_second_per_gpu": 2538.81
},
{
"epoch": 1.060515278609946,
"grad_norm": 3.515625,
"learning_rate": 2.0935553358627947e-05,
"loss": 0.3716,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10620,
"tokens_per_second_per_gpu": 2579.16
},
{
"epoch": 1.061513880567206,
"grad_norm": 4.21875,
"learning_rate": 2.0928313093468405e-05,
"loss": 0.4004,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10630,
"tokens_per_second_per_gpu": 2422.37
},
{
"epoch": 1.0625124825244658,
"grad_norm": 2.796875,
"learning_rate": 2.0921067639684315e-05,
"loss": 0.374,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10640,
"tokens_per_second_per_gpu": 2520.12
},
{
"epoch": 1.0635110844817255,
"grad_norm": 3.59375,
"learning_rate": 2.0913817001736112e-05,
"loss": 0.3821,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10650,
"tokens_per_second_per_gpu": 2452.83
},
{
"epoch": 1.0645096864389854,
"grad_norm": 2.90625,
"learning_rate": 2.090656118408743e-05,
"loss": 0.4051,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10660,
"tokens_per_second_per_gpu": 2451.73
},
{
"epoch": 1.0655082883962452,
"grad_norm": 3.484375,
"learning_rate": 2.089930019120507e-05,
"loss": 0.3469,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10670,
"tokens_per_second_per_gpu": 2280.71
},
{
"epoch": 1.066506890353505,
"grad_norm": 3.875,
"learning_rate": 2.0892034027559043e-05,
"loss": 0.3593,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10680,
"tokens_per_second_per_gpu": 2486.52
},
{
"epoch": 1.067505492310765,
"grad_norm": 3.0625,
"learning_rate": 2.0884762697622535e-05,
"loss": 0.3743,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10690,
"tokens_per_second_per_gpu": 2422.05
},
{
"epoch": 1.0685040942680248,
"grad_norm": 2.890625,
"learning_rate": 2.08774862058719e-05,
"loss": 0.3924,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10700,
"tokens_per_second_per_gpu": 2553.78
},
{
"epoch": 1.0695026962252845,
"grad_norm": 2.890625,
"learning_rate": 2.087020455678669e-05,
"loss": 0.3186,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10710,
"tokens_per_second_per_gpu": 2382.4
},
{
"epoch": 1.0705012981825444,
"grad_norm": 3.875,
"learning_rate": 2.0862917754849616e-05,
"loss": 0.417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10720,
"tokens_per_second_per_gpu": 2639.88
},
{
"epoch": 1.0714999001398042,
"grad_norm": 3.265625,
"learning_rate": 2.085562580454657e-05,
"loss": 0.4074,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10730,
"tokens_per_second_per_gpu": 2552.91
},
{
"epoch": 1.072498502097064,
"grad_norm": 3.40625,
"learning_rate": 2.084832871036661e-05,
"loss": 0.39,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10740,
"tokens_per_second_per_gpu": 2434.46
},
{
"epoch": 1.073497104054324,
"grad_norm": 2.859375,
"learning_rate": 2.0841026476801965e-05,
"loss": 0.4252,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10750,
"tokens_per_second_per_gpu": 2686.3
},
{
"epoch": 1.0744957060115838,
"grad_norm": 2.71875,
"learning_rate": 2.0833719108348015e-05,
"loss": 0.3544,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10760,
"tokens_per_second_per_gpu": 2581.04
},
{
"epoch": 1.0754943079688437,
"grad_norm": 2.640625,
"learning_rate": 2.082640660950332e-05,
"loss": 0.3754,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10770,
"tokens_per_second_per_gpu": 2474.54
},
{
"epoch": 1.0764929099261034,
"grad_norm": 2.625,
"learning_rate": 2.0819088984769587e-05,
"loss": 0.39,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10780,
"tokens_per_second_per_gpu": 2333.38
},
{
"epoch": 1.0774915118833632,
"grad_norm": 3.421875,
"learning_rate": 2.0811766238651675e-05,
"loss": 0.4041,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10790,
"tokens_per_second_per_gpu": 2640.59
},
{
"epoch": 1.078490113840623,
"grad_norm": 2.984375,
"learning_rate": 2.0804438375657602e-05,
"loss": 0.3759,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10800,
"tokens_per_second_per_gpu": 2614.4
},
{
"epoch": 1.079488715797883,
"grad_norm": 2.546875,
"learning_rate": 2.0797105400298543e-05,
"loss": 0.3279,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10810,
"tokens_per_second_per_gpu": 2363.96
},
{
"epoch": 1.0804873177551428,
"grad_norm": 2.75,
"learning_rate": 2.078976731708881e-05,
"loss": 0.3981,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10820,
"tokens_per_second_per_gpu": 2583.42
},
{
"epoch": 1.0814859197124027,
"grad_norm": 3.265625,
"learning_rate": 2.078242413054585e-05,
"loss": 0.4142,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10830,
"tokens_per_second_per_gpu": 2524.84
},
{
"epoch": 1.0824845216696626,
"grad_norm": 2.828125,
"learning_rate": 2.077507584519028e-05,
"loss": 0.4036,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10840,
"tokens_per_second_per_gpu": 2518.34
},
{
"epoch": 1.0834831236269222,
"grad_norm": 3.265625,
"learning_rate": 2.076772246554583e-05,
"loss": 0.3966,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10850,
"tokens_per_second_per_gpu": 2368.6
},
{
"epoch": 1.084481725584182,
"grad_norm": 2.3125,
"learning_rate": 2.076036399613938e-05,
"loss": 0.3766,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10860,
"tokens_per_second_per_gpu": 2343.13
},
{
"epoch": 1.085480327541442,
"grad_norm": 3.515625,
"learning_rate": 2.0753000441500937e-05,
"loss": 0.3495,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10870,
"tokens_per_second_per_gpu": 2299.61
},
{
"epoch": 1.0864789294987018,
"grad_norm": 2.5,
"learning_rate": 2.074563180616364e-05,
"loss": 0.3467,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10880,
"tokens_per_second_per_gpu": 2569.65
},
{
"epoch": 1.0874775314559617,
"grad_norm": 2.78125,
"learning_rate": 2.0738258094663758e-05,
"loss": 0.3626,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10890,
"tokens_per_second_per_gpu": 2571.69
},
{
"epoch": 1.0884761334132216,
"grad_norm": 2.828125,
"learning_rate": 2.0730879311540684e-05,
"loss": 0.3679,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10900,
"tokens_per_second_per_gpu": 2519.04
},
{
"epoch": 1.0894747353704812,
"grad_norm": 4.28125,
"learning_rate": 2.0723495461336927e-05,
"loss": 0.442,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10910,
"tokens_per_second_per_gpu": 2490.02
},
{
"epoch": 1.090473337327741,
"grad_norm": 3.109375,
"learning_rate": 2.0716106548598133e-05,
"loss": 0.4164,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10920,
"tokens_per_second_per_gpu": 2561.98
},
{
"epoch": 1.091471939285001,
"grad_norm": 4.25,
"learning_rate": 2.070871257787304e-05,
"loss": 0.3644,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10930,
"tokens_per_second_per_gpu": 2578.14
},
{
"epoch": 1.0924705412422608,
"grad_norm": 3.21875,
"learning_rate": 2.070131355371352e-05,
"loss": 0.4305,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10940,
"tokens_per_second_per_gpu": 2700.82
},
{
"epoch": 1.0934691431995207,
"grad_norm": 2.65625,
"learning_rate": 2.0693909480674545e-05,
"loss": 0.4072,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10950,
"tokens_per_second_per_gpu": 2526.59
},
{
"epoch": 1.0944677451567806,
"grad_norm": 3.109375,
"learning_rate": 2.0686500363314196e-05,
"loss": 0.3517,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10960,
"tokens_per_second_per_gpu": 2573.06
},
{
"epoch": 1.0954663471140402,
"grad_norm": 2.65625,
"learning_rate": 2.0679086206193666e-05,
"loss": 0.3661,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10970,
"tokens_per_second_per_gpu": 2715.48
},
{
"epoch": 1.0964649490713,
"grad_norm": 3.640625,
"learning_rate": 2.067166701387725e-05,
"loss": 0.3936,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10980,
"tokens_per_second_per_gpu": 2321.46
},
{
"epoch": 1.09746355102856,
"grad_norm": 3.46875,
"learning_rate": 2.066424279093233e-05,
"loss": 0.3627,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 10990,
"tokens_per_second_per_gpu": 2623.6
},
{
"epoch": 1.0984621529858198,
"grad_norm": 3.515625,
"learning_rate": 2.0656813541929404e-05,
"loss": 0.4099,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11000,
"tokens_per_second_per_gpu": 2526.87
},
{
"epoch": 1.0994607549430797,
"grad_norm": 2.875,
"learning_rate": 2.064937927144205e-05,
"loss": 0.3461,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11010,
"tokens_per_second_per_gpu": 2485.35
},
{
"epoch": 1.1004593569003396,
"grad_norm": 3.640625,
"learning_rate": 2.0641939984046933e-05,
"loss": 0.3731,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11020,
"tokens_per_second_per_gpu": 2518.93
},
{
"epoch": 1.1014579588575995,
"grad_norm": 3.234375,
"learning_rate": 2.063449568432383e-05,
"loss": 0.3653,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11030,
"tokens_per_second_per_gpu": 2381.33
},
{
"epoch": 1.102456560814859,
"grad_norm": 2.953125,
"learning_rate": 2.062704637685558e-05,
"loss": 0.3393,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11040,
"tokens_per_second_per_gpu": 2523.77
},
{
"epoch": 1.103455162772119,
"grad_norm": 4.21875,
"learning_rate": 2.0619592066228117e-05,
"loss": 0.4006,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11050,
"tokens_per_second_per_gpu": 2464.24
},
{
"epoch": 1.1044537647293788,
"grad_norm": 4.0,
"learning_rate": 2.061213275703045e-05,
"loss": 0.3797,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11060,
"tokens_per_second_per_gpu": 2435.07
},
{
"epoch": 1.1054523666866387,
"grad_norm": 3.4375,
"learning_rate": 2.0604668453854668e-05,
"loss": 0.3935,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11070,
"tokens_per_second_per_gpu": 2459.11
},
{
"epoch": 1.1064509686438986,
"grad_norm": 3.171875,
"learning_rate": 2.059719916129593e-05,
"loss": 0.3557,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11080,
"tokens_per_second_per_gpu": 2593.08
},
{
"epoch": 1.1074495706011585,
"grad_norm": 3.40625,
"learning_rate": 2.0589724883952473e-05,
"loss": 0.3631,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11090,
"tokens_per_second_per_gpu": 2583.35
},
{
"epoch": 1.1084481725584183,
"grad_norm": 2.640625,
"learning_rate": 2.05822456264256e-05,
"loss": 0.3801,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11100,
"tokens_per_second_per_gpu": 2311.85
},
{
"epoch": 1.109446774515678,
"grad_norm": 3.234375,
"learning_rate": 2.057476139331968e-05,
"loss": 0.4028,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11110,
"tokens_per_second_per_gpu": 2648.11
},
{
"epoch": 1.1104453764729378,
"grad_norm": 3.015625,
"learning_rate": 2.056727218924214e-05,
"loss": 0.3596,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11120,
"tokens_per_second_per_gpu": 2455.12
},
{
"epoch": 1.1114439784301977,
"grad_norm": 3.40625,
"learning_rate": 2.055977801880348e-05,
"loss": 0.3667,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11130,
"tokens_per_second_per_gpu": 2489.96
},
{
"epoch": 1.1124425803874576,
"grad_norm": 3.171875,
"learning_rate": 2.055227888661724e-05,
"loss": 0.4134,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11140,
"tokens_per_second_per_gpu": 2642.2
},
{
"epoch": 1.1134411823447175,
"grad_norm": 2.59375,
"learning_rate": 2.054477479730003e-05,
"loss": 0.4068,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11150,
"tokens_per_second_per_gpu": 2507.57
},
{
"epoch": 1.1144397843019773,
"grad_norm": 3.375,
"learning_rate": 2.0537265755471503e-05,
"loss": 0.3853,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11160,
"tokens_per_second_per_gpu": 2595.65
},
{
"epoch": 1.115438386259237,
"grad_norm": 3.640625,
"learning_rate": 2.052975176575436e-05,
"loss": 0.3565,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11170,
"tokens_per_second_per_gpu": 2450.02
},
{
"epoch": 1.1164369882164968,
"grad_norm": 2.59375,
"learning_rate": 2.0522232832774356e-05,
"loss": 0.3698,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11180,
"tokens_per_second_per_gpu": 2692.16
},
{
"epoch": 1.1174355901737567,
"grad_norm": 3.5625,
"learning_rate": 2.0514708961160287e-05,
"loss": 0.3641,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11190,
"tokens_per_second_per_gpu": 2489.54
},
{
"epoch": 1.1184341921310166,
"grad_norm": 3.59375,
"learning_rate": 2.0507180155543978e-05,
"loss": 0.4351,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11200,
"tokens_per_second_per_gpu": 2628.25
},
{
"epoch": 1.1194327940882765,
"grad_norm": 3.40625,
"learning_rate": 2.049964642056031e-05,
"loss": 0.3791,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11210,
"tokens_per_second_per_gpu": 2502.54
},
{
"epoch": 1.1204313960455363,
"grad_norm": 3.96875,
"learning_rate": 2.0492107760847178e-05,
"loss": 0.445,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11220,
"tokens_per_second_per_gpu": 2732.07
},
{
"epoch": 1.121429998002796,
"grad_norm": 3.796875,
"learning_rate": 2.048456418104553e-05,
"loss": 0.4343,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11230,
"tokens_per_second_per_gpu": 2291.46
},
{
"epoch": 1.1224285999600558,
"grad_norm": 3.3125,
"learning_rate": 2.0477015685799326e-05,
"loss": 0.3878,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11240,
"tokens_per_second_per_gpu": 2417.01
},
{
"epoch": 1.1234272019173157,
"grad_norm": 3.25,
"learning_rate": 2.046946227975556e-05,
"loss": 0.3776,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11250,
"tokens_per_second_per_gpu": 2450.72
},
{
"epoch": 1.1244258038745756,
"grad_norm": 3.296875,
"learning_rate": 2.0461903967564248e-05,
"loss": 0.3887,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11260,
"tokens_per_second_per_gpu": 2502.69
},
{
"epoch": 1.1254244058318355,
"grad_norm": 3.171875,
"learning_rate": 2.0454340753878426e-05,
"loss": 0.4002,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11270,
"tokens_per_second_per_gpu": 2606.02
},
{
"epoch": 1.1264230077890953,
"grad_norm": 3.40625,
"learning_rate": 2.0446772643354147e-05,
"loss": 0.3599,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11280,
"tokens_per_second_per_gpu": 2617.9
},
{
"epoch": 1.1274216097463552,
"grad_norm": 3.546875,
"learning_rate": 2.043919964065048e-05,
"loss": 0.3883,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11290,
"tokens_per_second_per_gpu": 2598.42
},
{
"epoch": 1.1284202117036148,
"grad_norm": 2.703125,
"learning_rate": 2.0431621750429504e-05,
"loss": 0.327,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11300,
"tokens_per_second_per_gpu": 2564.09
},
{
"epoch": 1.1294188136608747,
"grad_norm": 3.34375,
"learning_rate": 2.042403897735631e-05,
"loss": 0.4119,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11310,
"tokens_per_second_per_gpu": 2580.92
},
{
"epoch": 1.1304174156181346,
"grad_norm": 3.390625,
"learning_rate": 2.041645132609899e-05,
"loss": 0.368,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11320,
"tokens_per_second_per_gpu": 2489.46
},
{
"epoch": 1.1314160175753945,
"grad_norm": 3.046875,
"learning_rate": 2.040885880132864e-05,
"loss": 0.4276,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11330,
"tokens_per_second_per_gpu": 2517.04
},
{
"epoch": 1.1324146195326543,
"grad_norm": 2.359375,
"learning_rate": 2.0401261407719357e-05,
"loss": 0.3311,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11340,
"tokens_per_second_per_gpu": 2478.82
},
{
"epoch": 1.1334132214899142,
"grad_norm": 2.9375,
"learning_rate": 2.0393659149948246e-05,
"loss": 0.3414,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11350,
"tokens_per_second_per_gpu": 2591.95
},
{
"epoch": 1.134411823447174,
"grad_norm": 2.609375,
"learning_rate": 2.0386052032695384e-05,
"loss": 0.3711,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11360,
"tokens_per_second_per_gpu": 2757.89
},
{
"epoch": 1.1354104254044337,
"grad_norm": 3.84375,
"learning_rate": 2.037844006064386e-05,
"loss": 0.409,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11370,
"tokens_per_second_per_gpu": 2593.42
},
{
"epoch": 1.1364090273616936,
"grad_norm": 2.8125,
"learning_rate": 2.0370823238479746e-05,
"loss": 0.3531,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11380,
"tokens_per_second_per_gpu": 2502.44
},
{
"epoch": 1.1374076293189535,
"grad_norm": 3.5,
"learning_rate": 2.036320157089209e-05,
"loss": 0.3806,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11390,
"tokens_per_second_per_gpu": 2537.08
},
{
"epoch": 1.1384062312762133,
"grad_norm": 3.046875,
"learning_rate": 2.0355575062572933e-05,
"loss": 0.3631,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11400,
"tokens_per_second_per_gpu": 2624.34
},
{
"epoch": 1.1394048332334732,
"grad_norm": 3.078125,
"learning_rate": 2.0347943718217298e-05,
"loss": 0.3321,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11410,
"tokens_per_second_per_gpu": 2527.84
},
{
"epoch": 1.140403435190733,
"grad_norm": 4.0,
"learning_rate": 2.0340307542523178e-05,
"loss": 0.4095,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11420,
"tokens_per_second_per_gpu": 2637.34
},
{
"epoch": 1.141402037147993,
"grad_norm": 3.265625,
"learning_rate": 2.033266654019155e-05,
"loss": 0.4096,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11430,
"tokens_per_second_per_gpu": 2408.25
},
{
"epoch": 1.1424006391052526,
"grad_norm": 2.53125,
"learning_rate": 2.032502071592634e-05,
"loss": 0.4145,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11440,
"tokens_per_second_per_gpu": 2423.87
},
{
"epoch": 1.1433992410625125,
"grad_norm": 3.90625,
"learning_rate": 2.0317370074434468e-05,
"loss": 0.3651,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11450,
"tokens_per_second_per_gpu": 2276.69
},
{
"epoch": 1.1443978430197723,
"grad_norm": 3.59375,
"learning_rate": 2.0309714620425814e-05,
"loss": 0.3676,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11460,
"tokens_per_second_per_gpu": 2499.47
},
{
"epoch": 1.1453964449770322,
"grad_norm": 3.59375,
"learning_rate": 2.0302054358613205e-05,
"loss": 0.3706,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11470,
"tokens_per_second_per_gpu": 2586.36
},
{
"epoch": 1.146395046934292,
"grad_norm": 2.5,
"learning_rate": 2.029438929371245e-05,
"loss": 0.359,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11480,
"tokens_per_second_per_gpu": 2321.0
},
{
"epoch": 1.1473936488915517,
"grad_norm": 3.453125,
"learning_rate": 2.0286719430442295e-05,
"loss": 0.3901,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11490,
"tokens_per_second_per_gpu": 2596.12
},
{
"epoch": 1.1483922508488116,
"grad_norm": 3.375,
"learning_rate": 2.0279044773524454e-05,
"loss": 0.4074,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11500,
"tokens_per_second_per_gpu": 2475.82
},
{
"epoch": 1.1493908528060715,
"grad_norm": 3.328125,
"learning_rate": 2.0271365327683587e-05,
"loss": 0.364,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11510,
"tokens_per_second_per_gpu": 2545.92
},
{
"epoch": 1.1503894547633313,
"grad_norm": 2.765625,
"learning_rate": 2.02636810976473e-05,
"loss": 0.3982,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11520,
"tokens_per_second_per_gpu": 2471.37
},
{
"epoch": 1.1513880567205912,
"grad_norm": 2.625,
"learning_rate": 2.0255992088146147e-05,
"loss": 0.3636,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11530,
"tokens_per_second_per_gpu": 2529.12
},
{
"epoch": 1.152386658677851,
"grad_norm": 3.09375,
"learning_rate": 2.0248298303913628e-05,
"loss": 0.4309,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11540,
"tokens_per_second_per_gpu": 2677.87
},
{
"epoch": 1.153385260635111,
"grad_norm": 2.84375,
"learning_rate": 2.024059974968617e-05,
"loss": 0.3773,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11550,
"tokens_per_second_per_gpu": 2579.77
},
{
"epoch": 1.1543838625923706,
"grad_norm": 2.734375,
"learning_rate": 2.0232896430203148e-05,
"loss": 0.337,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11560,
"tokens_per_second_per_gpu": 2677.88
},
{
"epoch": 1.1553824645496305,
"grad_norm": 4.1875,
"learning_rate": 2.022518835020687e-05,
"loss": 0.3852,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11570,
"tokens_per_second_per_gpu": 2341.58
},
{
"epoch": 1.1563810665068903,
"grad_norm": 3.140625,
"learning_rate": 2.021747551444257e-05,
"loss": 0.4005,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11580,
"tokens_per_second_per_gpu": 2625.04
},
{
"epoch": 1.1573796684641502,
"grad_norm": 2.453125,
"learning_rate": 2.020975792765841e-05,
"loss": 0.3312,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11590,
"tokens_per_second_per_gpu": 2598.25
},
{
"epoch": 1.15837827042141,
"grad_norm": 3.296875,
"learning_rate": 2.0202035594605474e-05,
"loss": 0.412,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11600,
"tokens_per_second_per_gpu": 2500.75
},
{
"epoch": 1.15937687237867,
"grad_norm": 2.859375,
"learning_rate": 2.0194308520037785e-05,
"loss": 0.3422,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11610,
"tokens_per_second_per_gpu": 2524.22
},
{
"epoch": 1.1603754743359298,
"grad_norm": 3.75,
"learning_rate": 2.0186576708712262e-05,
"loss": 0.447,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11620,
"tokens_per_second_per_gpu": 2545.94
},
{
"epoch": 1.1613740762931895,
"grad_norm": 4.125,
"learning_rate": 2.0178840165388753e-05,
"loss": 0.3762,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11630,
"tokens_per_second_per_gpu": 2560.5
},
{
"epoch": 1.1623726782504493,
"grad_norm": 2.75,
"learning_rate": 2.0171098894830016e-05,
"loss": 0.3786,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11640,
"tokens_per_second_per_gpu": 2608.54
},
{
"epoch": 1.1633712802077092,
"grad_norm": 3.140625,
"learning_rate": 2.016335290180172e-05,
"loss": 0.4217,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11650,
"tokens_per_second_per_gpu": 2437.49
},
{
"epoch": 1.164369882164969,
"grad_norm": 3.25,
"learning_rate": 2.015560219107244e-05,
"loss": 0.3945,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11660,
"tokens_per_second_per_gpu": 2586.77
},
{
"epoch": 1.165368484122229,
"grad_norm": 3.8125,
"learning_rate": 2.0147846767413662e-05,
"loss": 0.3557,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11670,
"tokens_per_second_per_gpu": 2425.43
},
{
"epoch": 1.1663670860794888,
"grad_norm": 3.390625,
"learning_rate": 2.014008663559976e-05,
"loss": 0.409,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11680,
"tokens_per_second_per_gpu": 2547.59
},
{
"epoch": 1.1673656880367487,
"grad_norm": 3.78125,
"learning_rate": 2.013232180040801e-05,
"loss": 0.4389,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11690,
"tokens_per_second_per_gpu": 2536.66
},
{
"epoch": 1.1683642899940083,
"grad_norm": 2.953125,
"learning_rate": 2.01245522666186e-05,
"loss": 0.3664,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11700,
"tokens_per_second_per_gpu": 2349.73
},
{
"epoch": 1.1693628919512682,
"grad_norm": 2.6875,
"learning_rate": 2.0116778039014592e-05,
"loss": 0.3896,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11710,
"tokens_per_second_per_gpu": 2623.81
},
{
"epoch": 1.170361493908528,
"grad_norm": 3.0625,
"learning_rate": 2.010899912238194e-05,
"loss": 0.369,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11720,
"tokens_per_second_per_gpu": 2437.79
},
{
"epoch": 1.171360095865788,
"grad_norm": 3.625,
"learning_rate": 2.0101215521509493e-05,
"loss": 0.4155,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11730,
"tokens_per_second_per_gpu": 2578.04
},
{
"epoch": 1.1723586978230478,
"grad_norm": 3.296875,
"learning_rate": 2.0093427241188973e-05,
"loss": 0.3504,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11740,
"tokens_per_second_per_gpu": 2560.81
},
{
"epoch": 1.1733572997803075,
"grad_norm": 2.890625,
"learning_rate": 2.0085634286214994e-05,
"loss": 0.3253,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11750,
"tokens_per_second_per_gpu": 2434.59
},
{
"epoch": 1.1743559017375673,
"grad_norm": 2.953125,
"learning_rate": 2.007783666138504e-05,
"loss": 0.4153,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11760,
"tokens_per_second_per_gpu": 2658.5
},
{
"epoch": 1.1753545036948272,
"grad_norm": 3.609375,
"learning_rate": 2.0070034371499478e-05,
"loss": 0.3487,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11770,
"tokens_per_second_per_gpu": 2557.66
},
{
"epoch": 1.176353105652087,
"grad_norm": 4.40625,
"learning_rate": 2.0062227421361533e-05,
"loss": 0.4175,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11780,
"tokens_per_second_per_gpu": 2538.28
},
{
"epoch": 1.177351707609347,
"grad_norm": 3.390625,
"learning_rate": 2.005441581577731e-05,
"loss": 0.3631,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11790,
"tokens_per_second_per_gpu": 2436.27
},
{
"epoch": 1.1783503095666068,
"grad_norm": 3.3125,
"learning_rate": 2.004659955955578e-05,
"loss": 0.3427,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11800,
"tokens_per_second_per_gpu": 2414.53
},
{
"epoch": 1.1793489115238667,
"grad_norm": 3.15625,
"learning_rate": 2.0038778657508768e-05,
"loss": 0.3972,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11810,
"tokens_per_second_per_gpu": 2447.28
},
{
"epoch": 1.1803475134811263,
"grad_norm": 2.609375,
"learning_rate": 2.0030953114450966e-05,
"loss": 0.3781,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11820,
"tokens_per_second_per_gpu": 2546.67
},
{
"epoch": 1.1813461154383862,
"grad_norm": 2.609375,
"learning_rate": 2.002312293519993e-05,
"loss": 0.3934,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11830,
"tokens_per_second_per_gpu": 2465.55
},
{
"epoch": 1.182344717395646,
"grad_norm": 2.359375,
"learning_rate": 2.0015288124576058e-05,
"loss": 0.3514,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11840,
"tokens_per_second_per_gpu": 2421.48
},
{
"epoch": 1.183343319352906,
"grad_norm": 3.453125,
"learning_rate": 2.0007448687402597e-05,
"loss": 0.3534,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11850,
"tokens_per_second_per_gpu": 2547.46
},
{
"epoch": 1.1843419213101658,
"grad_norm": 3.65625,
"learning_rate": 1.999960462850566e-05,
"loss": 0.3913,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11860,
"tokens_per_second_per_gpu": 2443.77
},
{
"epoch": 1.1853405232674257,
"grad_norm": 3.328125,
"learning_rate": 1.9991755952714186e-05,
"loss": 0.3888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11870,
"tokens_per_second_per_gpu": 2559.85
},
{
"epoch": 1.1863391252246855,
"grad_norm": 3.609375,
"learning_rate": 1.9983902664859964e-05,
"loss": 0.3603,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11880,
"tokens_per_second_per_gpu": 2741.61
},
{
"epoch": 1.1873377271819452,
"grad_norm": 3.21875,
"learning_rate": 1.997604476977763e-05,
"loss": 0.3719,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11890,
"tokens_per_second_per_gpu": 2636.32
},
{
"epoch": 1.188336329139205,
"grad_norm": 3.296875,
"learning_rate": 1.996818227230464e-05,
"loss": 0.3257,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11900,
"tokens_per_second_per_gpu": 2349.74
},
{
"epoch": 1.189334931096465,
"grad_norm": 3.40625,
"learning_rate": 1.9960315177281296e-05,
"loss": 0.3542,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11910,
"tokens_per_second_per_gpu": 2527.22
},
{
"epoch": 1.1903335330537248,
"grad_norm": 3.8125,
"learning_rate": 1.9952443489550725e-05,
"loss": 0.3727,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11920,
"tokens_per_second_per_gpu": 2713.58
},
{
"epoch": 1.1913321350109847,
"grad_norm": 3.125,
"learning_rate": 1.9944567213958888e-05,
"loss": 0.394,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11930,
"tokens_per_second_per_gpu": 2575.67
},
{
"epoch": 1.1923307369682445,
"grad_norm": 3.484375,
"learning_rate": 1.993668635535456e-05,
"loss": 0.4142,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11940,
"tokens_per_second_per_gpu": 2577.41
},
{
"epoch": 1.1933293389255044,
"grad_norm": 3.28125,
"learning_rate": 1.992880091858934e-05,
"loss": 0.3472,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11950,
"tokens_per_second_per_gpu": 2342.02
},
{
"epoch": 1.194327940882764,
"grad_norm": 2.96875,
"learning_rate": 1.9920910908517654e-05,
"loss": 0.3285,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11960,
"tokens_per_second_per_gpu": 2452.88
},
{
"epoch": 1.195326542840024,
"grad_norm": 3.578125,
"learning_rate": 1.9913016329996732e-05,
"loss": 0.3936,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11970,
"tokens_per_second_per_gpu": 2507.86
},
{
"epoch": 1.1963251447972838,
"grad_norm": 3.125,
"learning_rate": 1.9905117187886622e-05,
"loss": 0.3871,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11980,
"tokens_per_second_per_gpu": 2699.43
},
{
"epoch": 1.1973237467545437,
"grad_norm": 3.484375,
"learning_rate": 1.989721348705019e-05,
"loss": 0.4131,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 11990,
"tokens_per_second_per_gpu": 2421.05
},
{
"epoch": 1.1983223487118035,
"grad_norm": 2.96875,
"learning_rate": 1.988930523235309e-05,
"loss": 0.384,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12000,
"tokens_per_second_per_gpu": 2369.98
},
{
"epoch": 1.1993209506690632,
"grad_norm": 2.984375,
"learning_rate": 1.98813924286638e-05,
"loss": 0.3826,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12010,
"tokens_per_second_per_gpu": 2611.23
},
{
"epoch": 1.200319552626323,
"grad_norm": 4.09375,
"learning_rate": 1.9873475080853568e-05,
"loss": 0.4208,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12020,
"tokens_per_second_per_gpu": 2632.17
},
{
"epoch": 1.201318154583583,
"grad_norm": 3.34375,
"learning_rate": 1.9865553193796473e-05,
"loss": 0.3592,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12030,
"tokens_per_second_per_gpu": 2476.77
},
{
"epoch": 1.2023167565408428,
"grad_norm": 4.34375,
"learning_rate": 1.9857626772369377e-05,
"loss": 0.4046,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12040,
"tokens_per_second_per_gpu": 2362.01
},
{
"epoch": 1.2033153584981027,
"grad_norm": 3.265625,
"learning_rate": 1.984969582145192e-05,
"loss": 0.3601,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12050,
"tokens_per_second_per_gpu": 2484.28
},
{
"epoch": 1.2043139604553625,
"grad_norm": 3.359375,
"learning_rate": 1.9841760345926546e-05,
"loss": 0.4243,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12060,
"tokens_per_second_per_gpu": 2397.43
},
{
"epoch": 1.2053125624126224,
"grad_norm": 3.125,
"learning_rate": 1.9833820350678485e-05,
"loss": 0.3798,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12070,
"tokens_per_second_per_gpu": 2499.29
},
{
"epoch": 1.206311164369882,
"grad_norm": 2.984375,
"learning_rate": 1.982587584059574e-05,
"loss": 0.3906,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12080,
"tokens_per_second_per_gpu": 2369.19
},
{
"epoch": 1.207309766327142,
"grad_norm": 3.296875,
"learning_rate": 1.9817926820569093e-05,
"loss": 0.4203,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12090,
"tokens_per_second_per_gpu": 2447.23
},
{
"epoch": 1.2083083682844018,
"grad_norm": 3.015625,
"learning_rate": 1.9809973295492117e-05,
"loss": 0.3454,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12100,
"tokens_per_second_per_gpu": 2507.87
},
{
"epoch": 1.2093069702416617,
"grad_norm": 3.4375,
"learning_rate": 1.9802015270261138e-05,
"loss": 0.3983,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12110,
"tokens_per_second_per_gpu": 2522.79
},
{
"epoch": 1.2103055721989215,
"grad_norm": 3.296875,
"learning_rate": 1.9794052749775265e-05,
"loss": 0.3946,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12120,
"tokens_per_second_per_gpu": 2385.53
},
{
"epoch": 1.2113041741561814,
"grad_norm": 3.921875,
"learning_rate": 1.978608573893638e-05,
"loss": 0.3605,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12130,
"tokens_per_second_per_gpu": 2533.79
},
{
"epoch": 1.2123027761134413,
"grad_norm": 4.09375,
"learning_rate": 1.977811424264911e-05,
"loss": 0.3831,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12140,
"tokens_per_second_per_gpu": 2388.3
},
{
"epoch": 1.213301378070701,
"grad_norm": 3.28125,
"learning_rate": 1.9770138265820863e-05,
"loss": 0.4178,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12150,
"tokens_per_second_per_gpu": 2582.95
},
{
"epoch": 1.2142999800279608,
"grad_norm": 3.828125,
"learning_rate": 1.97621578133618e-05,
"loss": 0.3467,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12160,
"tokens_per_second_per_gpu": 2583.34
},
{
"epoch": 1.2152985819852207,
"grad_norm": 3.328125,
"learning_rate": 1.975417289018482e-05,
"loss": 0.3537,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12170,
"tokens_per_second_per_gpu": 2463.0
},
{
"epoch": 1.2162971839424805,
"grad_norm": 3.734375,
"learning_rate": 1.9746183501205598e-05,
"loss": 0.3984,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12180,
"tokens_per_second_per_gpu": 2424.35
},
{
"epoch": 1.2172957858997404,
"grad_norm": 3.984375,
"learning_rate": 1.9738189651342553e-05,
"loss": 0.4024,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12190,
"tokens_per_second_per_gpu": 2497.42
},
{
"epoch": 1.2182943878570003,
"grad_norm": 2.59375,
"learning_rate": 1.9730191345516836e-05,
"loss": 0.3702,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12200,
"tokens_per_second_per_gpu": 2654.54
},
{
"epoch": 1.2192929898142602,
"grad_norm": 3.53125,
"learning_rate": 1.9722188588652356e-05,
"loss": 0.3358,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12210,
"tokens_per_second_per_gpu": 2431.5
},
{
"epoch": 1.2202915917715198,
"grad_norm": 3.015625,
"learning_rate": 1.971418138567576e-05,
"loss": 0.3658,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12220,
"tokens_per_second_per_gpu": 2449.86
},
{
"epoch": 1.2212901937287797,
"grad_norm": 2.265625,
"learning_rate": 1.970616974151643e-05,
"loss": 0.3403,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12230,
"tokens_per_second_per_gpu": 2307.03
},
{
"epoch": 1.2222887956860395,
"grad_norm": 3.703125,
"learning_rate": 1.9698153661106473e-05,
"loss": 0.3816,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12240,
"tokens_per_second_per_gpu": 2521.98
},
{
"epoch": 1.2232873976432994,
"grad_norm": 3.3125,
"learning_rate": 1.9690133149380742e-05,
"loss": 0.4098,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12250,
"tokens_per_second_per_gpu": 2618.57
},
{
"epoch": 1.2242859996005593,
"grad_norm": 2.25,
"learning_rate": 1.9682108211276808e-05,
"loss": 0.3402,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12260,
"tokens_per_second_per_gpu": 2363.35
},
{
"epoch": 1.225284601557819,
"grad_norm": 4.0,
"learning_rate": 1.9674078851734978e-05,
"loss": 0.3263,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12270,
"tokens_per_second_per_gpu": 2531.37
},
{
"epoch": 1.2262832035150788,
"grad_norm": 3.6875,
"learning_rate": 1.9666045075698265e-05,
"loss": 0.3738,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12280,
"tokens_per_second_per_gpu": 2461.55
},
{
"epoch": 1.2272818054723387,
"grad_norm": 3.03125,
"learning_rate": 1.9658006888112412e-05,
"loss": 0.3872,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12290,
"tokens_per_second_per_gpu": 2490.07
},
{
"epoch": 1.2282804074295985,
"grad_norm": 3.78125,
"learning_rate": 1.9649964293925875e-05,
"loss": 0.3795,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12300,
"tokens_per_second_per_gpu": 2548.95
},
{
"epoch": 1.2292790093868584,
"grad_norm": 2.59375,
"learning_rate": 1.9641917298089826e-05,
"loss": 0.3598,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12310,
"tokens_per_second_per_gpu": 2615.07
},
{
"epoch": 1.2302776113441183,
"grad_norm": 2.71875,
"learning_rate": 1.963386590555814e-05,
"loss": 0.3856,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12320,
"tokens_per_second_per_gpu": 2654.05
},
{
"epoch": 1.2312762133013782,
"grad_norm": 3.21875,
"learning_rate": 1.962581012128741e-05,
"loss": 0.4281,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12330,
"tokens_per_second_per_gpu": 2429.29
},
{
"epoch": 1.2322748152586378,
"grad_norm": 2.90625,
"learning_rate": 1.9617749950236906e-05,
"loss": 0.3792,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12340,
"tokens_per_second_per_gpu": 2331.24
},
{
"epoch": 1.2332734172158977,
"grad_norm": 3.75,
"learning_rate": 1.9609685397368636e-05,
"loss": 0.3907,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12350,
"tokens_per_second_per_gpu": 2449.37
},
{
"epoch": 1.2342720191731575,
"grad_norm": 4.3125,
"learning_rate": 1.9601616467647282e-05,
"loss": 0.3888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12360,
"tokens_per_second_per_gpu": 2437.56
},
{
"epoch": 1.2352706211304174,
"grad_norm": 3.640625,
"learning_rate": 1.9593543166040222e-05,
"loss": 0.3335,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12370,
"tokens_per_second_per_gpu": 2518.86
},
{
"epoch": 1.2362692230876773,
"grad_norm": 3.359375,
"learning_rate": 1.9585465497517536e-05,
"loss": 0.3533,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12380,
"tokens_per_second_per_gpu": 2401.19
},
{
"epoch": 1.2372678250449372,
"grad_norm": 3.53125,
"learning_rate": 1.957738346705198e-05,
"loss": 0.3719,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12390,
"tokens_per_second_per_gpu": 2389.02
},
{
"epoch": 1.238266427002197,
"grad_norm": 2.953125,
"learning_rate": 1.9569297079618997e-05,
"loss": 0.3994,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12400,
"tokens_per_second_per_gpu": 2492.77
},
{
"epoch": 1.2392650289594567,
"grad_norm": 2.515625,
"learning_rate": 1.956120634019672e-05,
"loss": 0.4209,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12410,
"tokens_per_second_per_gpu": 2474.85
},
{
"epoch": 1.2402636309167165,
"grad_norm": 3.34375,
"learning_rate": 1.9553111253765963e-05,
"loss": 0.3419,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12420,
"tokens_per_second_per_gpu": 2461.31
},
{
"epoch": 1.2412622328739764,
"grad_norm": 3.671875,
"learning_rate": 1.95450118253102e-05,
"loss": 0.3986,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12430,
"tokens_per_second_per_gpu": 2366.39
},
{
"epoch": 1.2422608348312363,
"grad_norm": 2.828125,
"learning_rate": 1.9536908059815594e-05,
"loss": 0.3934,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12440,
"tokens_per_second_per_gpu": 2580.87
},
{
"epoch": 1.2432594367884962,
"grad_norm": 3.171875,
"learning_rate": 1.9528799962270972e-05,
"loss": 0.393,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12450,
"tokens_per_second_per_gpu": 2476.25
},
{
"epoch": 1.244258038745756,
"grad_norm": 3.34375,
"learning_rate": 1.952068753766782e-05,
"loss": 0.3963,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12460,
"tokens_per_second_per_gpu": 2296.11
},
{
"epoch": 1.245256640703016,
"grad_norm": 2.953125,
"learning_rate": 1.9512570791000306e-05,
"loss": 0.3903,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12470,
"tokens_per_second_per_gpu": 2375.2
},
{
"epoch": 1.2462552426602755,
"grad_norm": 4.375,
"learning_rate": 1.9504449727265246e-05,
"loss": 0.3919,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12480,
"tokens_per_second_per_gpu": 2347.75
},
{
"epoch": 1.2472538446175354,
"grad_norm": 3.515625,
"learning_rate": 1.9496324351462116e-05,
"loss": 0.3282,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12490,
"tokens_per_second_per_gpu": 2451.43
},
{
"epoch": 1.2482524465747953,
"grad_norm": 3.390625,
"learning_rate": 1.9488194668593044e-05,
"loss": 0.3572,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12500,
"tokens_per_second_per_gpu": 2503.64
},
{
"epoch": 1.2492510485320552,
"grad_norm": 3.28125,
"learning_rate": 1.9480060683662813e-05,
"loss": 0.3386,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12510,
"tokens_per_second_per_gpu": 2346.35
},
{
"epoch": 1.250249650489315,
"grad_norm": 4.09375,
"learning_rate": 1.9471922401678856e-05,
"loss": 0.3793,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12520,
"tokens_per_second_per_gpu": 2339.51
},
{
"epoch": 1.2512482524465747,
"grad_norm": 3.59375,
"learning_rate": 1.9463779827651245e-05,
"loss": 0.4025,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12530,
"tokens_per_second_per_gpu": 2297.27
},
{
"epoch": 1.2522468544038348,
"grad_norm": 2.953125,
"learning_rate": 1.9455632966592703e-05,
"loss": 0.3928,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12540,
"tokens_per_second_per_gpu": 2524.51
},
{
"epoch": 1.2532454563610944,
"grad_norm": 2.375,
"learning_rate": 1.944748182351859e-05,
"loss": 0.3686,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12550,
"tokens_per_second_per_gpu": 2521.32
},
{
"epoch": 1.2542440583183543,
"grad_norm": 3.515625,
"learning_rate": 1.943932640344689e-05,
"loss": 0.3676,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12560,
"tokens_per_second_per_gpu": 2414.97
},
{
"epoch": 1.2552426602756142,
"grad_norm": 3.484375,
"learning_rate": 1.9431166711398234e-05,
"loss": 0.3635,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12570,
"tokens_per_second_per_gpu": 2427.54
},
{
"epoch": 1.256241262232874,
"grad_norm": 3.0625,
"learning_rate": 1.942300275239588e-05,
"loss": 0.3444,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12580,
"tokens_per_second_per_gpu": 2360.92
},
{
"epoch": 1.257239864190134,
"grad_norm": 3.125,
"learning_rate": 1.9414834531465712e-05,
"loss": 0.3798,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12590,
"tokens_per_second_per_gpu": 2686.79
},
{
"epoch": 1.2582384661473935,
"grad_norm": 2.9375,
"learning_rate": 1.940666205363623e-05,
"loss": 0.3988,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12600,
"tokens_per_second_per_gpu": 2389.7
},
{
"epoch": 1.2592370681046534,
"grad_norm": 3.171875,
"learning_rate": 1.9398485323938575e-05,
"loss": 0.362,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12610,
"tokens_per_second_per_gpu": 2307.95
},
{
"epoch": 1.2602356700619133,
"grad_norm": 3.5,
"learning_rate": 1.9390304347406485e-05,
"loss": 0.3888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12620,
"tokens_per_second_per_gpu": 2619.68
},
{
"epoch": 1.2612342720191732,
"grad_norm": 4.15625,
"learning_rate": 1.9382119129076312e-05,
"loss": 0.3932,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12630,
"tokens_per_second_per_gpu": 2450.81
},
{
"epoch": 1.262232873976433,
"grad_norm": 3.78125,
"learning_rate": 1.9373929673987037e-05,
"loss": 0.3585,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12640,
"tokens_per_second_per_gpu": 2462.93
},
{
"epoch": 1.263231475933693,
"grad_norm": 3.953125,
"learning_rate": 1.936573598718024e-05,
"loss": 0.4196,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12650,
"tokens_per_second_per_gpu": 2603.94
},
{
"epoch": 1.2642300778909528,
"grad_norm": 3.515625,
"learning_rate": 1.9357538073700092e-05,
"loss": 0.3967,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12660,
"tokens_per_second_per_gpu": 2602.66
},
{
"epoch": 1.2652286798482124,
"grad_norm": 3.109375,
"learning_rate": 1.93493359385934e-05,
"loss": 0.3779,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12670,
"tokens_per_second_per_gpu": 2480.45
},
{
"epoch": 1.2662272818054723,
"grad_norm": 2.125,
"learning_rate": 1.9341129586909532e-05,
"loss": 0.386,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12680,
"tokens_per_second_per_gpu": 2485.09
},
{
"epoch": 1.2672258837627322,
"grad_norm": 3.40625,
"learning_rate": 1.9332919023700474e-05,
"loss": 0.3976,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12690,
"tokens_per_second_per_gpu": 2536.58
},
{
"epoch": 1.268224485719992,
"grad_norm": 3.109375,
"learning_rate": 1.9324704254020802e-05,
"loss": 0.3673,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12700,
"tokens_per_second_per_gpu": 2557.09
},
{
"epoch": 1.269223087677252,
"grad_norm": 3.984375,
"learning_rate": 1.931648528292768e-05,
"loss": 0.3992,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12710,
"tokens_per_second_per_gpu": 2561.06
},
{
"epoch": 1.2702216896345118,
"grad_norm": 3.6875,
"learning_rate": 1.9308262115480852e-05,
"loss": 0.3766,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12720,
"tokens_per_second_per_gpu": 2502.48
},
{
"epoch": 1.2712202915917716,
"grad_norm": 3.25,
"learning_rate": 1.930003475674266e-05,
"loss": 0.3811,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12730,
"tokens_per_second_per_gpu": 2621.8
},
{
"epoch": 1.2722188935490313,
"grad_norm": 3.125,
"learning_rate": 1.9291803211778005e-05,
"loss": 0.3865,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12740,
"tokens_per_second_per_gpu": 2526.52
},
{
"epoch": 1.2732174955062912,
"grad_norm": 3.234375,
"learning_rate": 1.928356748565439e-05,
"loss": 0.3306,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12750,
"tokens_per_second_per_gpu": 2509.0
},
{
"epoch": 1.274216097463551,
"grad_norm": 3.21875,
"learning_rate": 1.927532758344187e-05,
"loss": 0.3809,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12760,
"tokens_per_second_per_gpu": 2450.5
},
{
"epoch": 1.275214699420811,
"grad_norm": 4.25,
"learning_rate": 1.926708351021309e-05,
"loss": 0.3631,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12770,
"tokens_per_second_per_gpu": 2522.14
},
{
"epoch": 1.2762133013780708,
"grad_norm": 3.296875,
"learning_rate": 1.925883527104324e-05,
"loss": 0.406,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12780,
"tokens_per_second_per_gpu": 2414.96
},
{
"epoch": 1.2772119033353304,
"grad_norm": 3.46875,
"learning_rate": 1.9250582871010104e-05,
"loss": 0.3832,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12790,
"tokens_per_second_per_gpu": 2507.72
},
{
"epoch": 1.2782105052925905,
"grad_norm": 3.59375,
"learning_rate": 1.9242326315194e-05,
"loss": 0.3585,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12800,
"tokens_per_second_per_gpu": 2487.16
},
{
"epoch": 1.2792091072498502,
"grad_norm": 3.5625,
"learning_rate": 1.9234065608677812e-05,
"loss": 0.3769,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12810,
"tokens_per_second_per_gpu": 2544.37
},
{
"epoch": 1.28020770920711,
"grad_norm": 3.046875,
"learning_rate": 1.9225800756546994e-05,
"loss": 0.3898,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12820,
"tokens_per_second_per_gpu": 2551.35
},
{
"epoch": 1.28120631116437,
"grad_norm": 2.59375,
"learning_rate": 1.9217531763889533e-05,
"loss": 0.3136,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12830,
"tokens_per_second_per_gpu": 2350.06
},
{
"epoch": 1.2822049131216298,
"grad_norm": 2.65625,
"learning_rate": 1.920925863579598e-05,
"loss": 0.3604,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12840,
"tokens_per_second_per_gpu": 2511.66
},
{
"epoch": 1.2832035150788896,
"grad_norm": 3.015625,
"learning_rate": 1.9200981377359418e-05,
"loss": 0.3504,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12850,
"tokens_per_second_per_gpu": 2540.24
},
{
"epoch": 1.2842021170361493,
"grad_norm": 3.75,
"learning_rate": 1.9192699993675483e-05,
"loss": 0.3937,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12860,
"tokens_per_second_per_gpu": 2404.07
},
{
"epoch": 1.2852007189934092,
"grad_norm": 2.671875,
"learning_rate": 1.9184414489842345e-05,
"loss": 0.3768,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12870,
"tokens_per_second_per_gpu": 2472.04
},
{
"epoch": 1.286199320950669,
"grad_norm": 3.5,
"learning_rate": 1.9176124870960717e-05,
"loss": 0.4667,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12880,
"tokens_per_second_per_gpu": 2440.02
},
{
"epoch": 1.287197922907929,
"grad_norm": 2.515625,
"learning_rate": 1.9167831142133834e-05,
"loss": 0.3272,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12890,
"tokens_per_second_per_gpu": 2432.44
},
{
"epoch": 1.2881965248651888,
"grad_norm": 3.03125,
"learning_rate": 1.915953330846747e-05,
"loss": 0.398,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12900,
"tokens_per_second_per_gpu": 2403.93
},
{
"epoch": 1.2891951268224486,
"grad_norm": 4.6875,
"learning_rate": 1.9151231375069935e-05,
"loss": 0.4524,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12910,
"tokens_per_second_per_gpu": 2567.85
},
{
"epoch": 1.2901937287797085,
"grad_norm": 3.359375,
"learning_rate": 1.9142925347052036e-05,
"loss": 0.3786,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12920,
"tokens_per_second_per_gpu": 2454.19
},
{
"epoch": 1.2911923307369682,
"grad_norm": 2.703125,
"learning_rate": 1.913461522952712e-05,
"loss": 0.3214,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12930,
"tokens_per_second_per_gpu": 2612.12
},
{
"epoch": 1.292190932694228,
"grad_norm": 3.3125,
"learning_rate": 1.9126301027611058e-05,
"loss": 0.3876,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12940,
"tokens_per_second_per_gpu": 2359.78
},
{
"epoch": 1.293189534651488,
"grad_norm": 2.65625,
"learning_rate": 1.9117982746422214e-05,
"loss": 0.3709,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12950,
"tokens_per_second_per_gpu": 2429.56
},
{
"epoch": 1.2941881366087478,
"grad_norm": 3.1875,
"learning_rate": 1.9109660391081483e-05,
"loss": 0.4212,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12960,
"tokens_per_second_per_gpu": 2623.09
},
{
"epoch": 1.2951867385660076,
"grad_norm": 3.34375,
"learning_rate": 1.9101333966712255e-05,
"loss": 0.3885,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12970,
"tokens_per_second_per_gpu": 2588.93
},
{
"epoch": 1.2961853405232675,
"grad_norm": 2.40625,
"learning_rate": 1.9093003478440434e-05,
"loss": 0.3609,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12980,
"tokens_per_second_per_gpu": 2523.31
},
{
"epoch": 1.2971839424805274,
"grad_norm": 3.109375,
"learning_rate": 1.9084668931394416e-05,
"loss": 0.3779,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 12990,
"tokens_per_second_per_gpu": 2389.39
},
{
"epoch": 1.298182544437787,
"grad_norm": 3.75,
"learning_rate": 1.907633033070511e-05,
"loss": 0.4025,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13000,
"tokens_per_second_per_gpu": 2563.05
},
{
"epoch": 1.299181146395047,
"grad_norm": 3.046875,
"learning_rate": 1.90679876815059e-05,
"loss": 0.4082,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13010,
"tokens_per_second_per_gpu": 2503.0
},
{
"epoch": 1.3001797483523068,
"grad_norm": 3.171875,
"learning_rate": 1.9059640988932686e-05,
"loss": 0.3692,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13020,
"tokens_per_second_per_gpu": 2552.47
},
{
"epoch": 1.3011783503095666,
"grad_norm": 3.59375,
"learning_rate": 1.9051290258123842e-05,
"loss": 0.4099,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13030,
"tokens_per_second_per_gpu": 2343.75
},
{
"epoch": 1.3021769522668265,
"grad_norm": 2.53125,
"learning_rate": 1.9042935494220227e-05,
"loss": 0.3427,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13040,
"tokens_per_second_per_gpu": 2376.17
},
{
"epoch": 1.3031755542240862,
"grad_norm": 3.140625,
"learning_rate": 1.9034576702365192e-05,
"loss": 0.3849,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13050,
"tokens_per_second_per_gpu": 2561.57
},
{
"epoch": 1.3041741561813462,
"grad_norm": 3.140625,
"learning_rate": 1.9026213887704562e-05,
"loss": 0.3827,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13060,
"tokens_per_second_per_gpu": 2259.45
},
{
"epoch": 1.305172758138606,
"grad_norm": 2.71875,
"learning_rate": 1.9017847055386635e-05,
"loss": 0.3458,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13070,
"tokens_per_second_per_gpu": 2406.14
},
{
"epoch": 1.3061713600958658,
"grad_norm": 3.078125,
"learning_rate": 1.9009476210562195e-05,
"loss": 0.3796,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13080,
"tokens_per_second_per_gpu": 2426.36
},
{
"epoch": 1.3071699620531256,
"grad_norm": 3.578125,
"learning_rate": 1.900110135838448e-05,
"loss": 0.3867,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13090,
"tokens_per_second_per_gpu": 2493.91
},
{
"epoch": 1.3081685640103855,
"grad_norm": 2.890625,
"learning_rate": 1.8992722504009214e-05,
"loss": 0.3403,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13100,
"tokens_per_second_per_gpu": 2402.72
},
{
"epoch": 1.3091671659676454,
"grad_norm": 3.46875,
"learning_rate": 1.8984339652594563e-05,
"loss": 0.3518,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13110,
"tokens_per_second_per_gpu": 2309.9
},
{
"epoch": 1.310165767924905,
"grad_norm": 3.296875,
"learning_rate": 1.8975952809301168e-05,
"loss": 0.4081,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13120,
"tokens_per_second_per_gpu": 2413.5
},
{
"epoch": 1.311164369882165,
"grad_norm": 2.8125,
"learning_rate": 1.8967561979292122e-05,
"loss": 0.3865,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13130,
"tokens_per_second_per_gpu": 2584.72
},
{
"epoch": 1.3121629718394248,
"grad_norm": 3.421875,
"learning_rate": 1.895916716773298e-05,
"loss": 0.4143,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13140,
"tokens_per_second_per_gpu": 2494.95
},
{
"epoch": 1.3131615737966846,
"grad_norm": 2.96875,
"learning_rate": 1.8950768379791736e-05,
"loss": 0.419,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13150,
"tokens_per_second_per_gpu": 2453.84
},
{
"epoch": 1.3141601757539445,
"grad_norm": 4.21875,
"learning_rate": 1.8942365620638843e-05,
"loss": 0.4526,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13160,
"tokens_per_second_per_gpu": 2358.73
},
{
"epoch": 1.3151587777112044,
"grad_norm": 2.796875,
"learning_rate": 1.8933958895447193e-05,
"loss": 0.4362,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13170,
"tokens_per_second_per_gpu": 2385.92
},
{
"epoch": 1.3161573796684642,
"grad_norm": 3.359375,
"learning_rate": 1.892554820939212e-05,
"loss": 0.3718,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13180,
"tokens_per_second_per_gpu": 2513.5
},
{
"epoch": 1.317155981625724,
"grad_norm": 2.859375,
"learning_rate": 1.8917133567651397e-05,
"loss": 0.3493,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13190,
"tokens_per_second_per_gpu": 2461.19
},
{
"epoch": 1.3181545835829838,
"grad_norm": 2.484375,
"learning_rate": 1.8908714975405228e-05,
"loss": 0.3829,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13200,
"tokens_per_second_per_gpu": 2549.21
},
{
"epoch": 1.3191531855402436,
"grad_norm": 3.15625,
"learning_rate": 1.8900292437836258e-05,
"loss": 0.4392,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13210,
"tokens_per_second_per_gpu": 2528.05
},
{
"epoch": 1.3201517874975035,
"grad_norm": 2.890625,
"learning_rate": 1.8891865960129558e-05,
"loss": 0.408,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13220,
"tokens_per_second_per_gpu": 2714.31
},
{
"epoch": 1.3211503894547634,
"grad_norm": 3.125,
"learning_rate": 1.8883435547472622e-05,
"loss": 0.3859,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13230,
"tokens_per_second_per_gpu": 2416.22
},
{
"epoch": 1.3221489914120232,
"grad_norm": 3.203125,
"learning_rate": 1.887500120505536e-05,
"loss": 0.463,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13240,
"tokens_per_second_per_gpu": 2612.78
},
{
"epoch": 1.3231475933692831,
"grad_norm": 2.71875,
"learning_rate": 1.8866562938070113e-05,
"loss": 0.3401,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13250,
"tokens_per_second_per_gpu": 2427.68
},
{
"epoch": 1.3241461953265428,
"grad_norm": 3.109375,
"learning_rate": 1.885812075171164e-05,
"loss": 0.417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13260,
"tokens_per_second_per_gpu": 2325.72
},
{
"epoch": 1.3251447972838026,
"grad_norm": 2.671875,
"learning_rate": 1.8849674651177092e-05,
"loss": 0.3658,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13270,
"tokens_per_second_per_gpu": 2566.4
},
{
"epoch": 1.3261433992410625,
"grad_norm": 1.96875,
"learning_rate": 1.884122464166606e-05,
"loss": 0.3704,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13280,
"tokens_per_second_per_gpu": 2464.23
},
{
"epoch": 1.3271420011983224,
"grad_norm": 3.71875,
"learning_rate": 1.8832770728380513e-05,
"loss": 0.3615,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13290,
"tokens_per_second_per_gpu": 2496.78
},
{
"epoch": 1.3281406031555822,
"grad_norm": 3.734375,
"learning_rate": 1.882431291652484e-05,
"loss": 0.3959,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13300,
"tokens_per_second_per_gpu": 2630.52
},
{
"epoch": 1.329139205112842,
"grad_norm": 3.28125,
"learning_rate": 1.8815851211305823e-05,
"loss": 0.3786,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13310,
"tokens_per_second_per_gpu": 2566.44
},
{
"epoch": 1.330137807070102,
"grad_norm": 2.1875,
"learning_rate": 1.8807385617932647e-05,
"loss": 0.3581,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13320,
"tokens_per_second_per_gpu": 2360.4
},
{
"epoch": 1.3311364090273616,
"grad_norm": 2.359375,
"learning_rate": 1.8798916141616886e-05,
"loss": 0.357,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13330,
"tokens_per_second_per_gpu": 2324.39
},
{
"epoch": 1.3321350109846215,
"grad_norm": 2.96875,
"learning_rate": 1.8790442787572505e-05,
"loss": 0.3895,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13340,
"tokens_per_second_per_gpu": 2428.94
},
{
"epoch": 1.3331336129418814,
"grad_norm": 3.234375,
"learning_rate": 1.8781965561015852e-05,
"loss": 0.3629,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13350,
"tokens_per_second_per_gpu": 2364.1
},
{
"epoch": 1.3341322148991412,
"grad_norm": 3.53125,
"learning_rate": 1.8773484467165672e-05,
"loss": 0.3493,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13360,
"tokens_per_second_per_gpu": 2496.25
},
{
"epoch": 1.3351308168564011,
"grad_norm": 3.046875,
"learning_rate": 1.8764999511243078e-05,
"loss": 0.3546,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13370,
"tokens_per_second_per_gpu": 2549.0
},
{
"epoch": 1.3361294188136608,
"grad_norm": 2.59375,
"learning_rate": 1.8756510698471564e-05,
"loss": 0.3576,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13380,
"tokens_per_second_per_gpu": 2282.96
},
{
"epoch": 1.3371280207709206,
"grad_norm": 2.78125,
"learning_rate": 1.8748018034077e-05,
"loss": 0.4062,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13390,
"tokens_per_second_per_gpu": 2567.96
},
{
"epoch": 1.3381266227281805,
"grad_norm": 2.90625,
"learning_rate": 1.8739521523287627e-05,
"loss": 0.3876,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13400,
"tokens_per_second_per_gpu": 2516.37
},
{
"epoch": 1.3391252246854404,
"grad_norm": 2.84375,
"learning_rate": 1.8731021171334055e-05,
"loss": 0.3638,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13410,
"tokens_per_second_per_gpu": 2536.77
},
{
"epoch": 1.3401238266427002,
"grad_norm": 3.953125,
"learning_rate": 1.8722516983449253e-05,
"loss": 0.4169,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13420,
"tokens_per_second_per_gpu": 2340.84
},
{
"epoch": 1.3411224285999601,
"grad_norm": 4.09375,
"learning_rate": 1.871400896486856e-05,
"loss": 0.407,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13430,
"tokens_per_second_per_gpu": 2466.11
},
{
"epoch": 1.34212103055722,
"grad_norm": 3.09375,
"learning_rate": 1.8705497120829664e-05,
"loss": 0.4053,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13440,
"tokens_per_second_per_gpu": 2478.02
},
{
"epoch": 1.3431196325144796,
"grad_norm": 2.984375,
"learning_rate": 1.8696981456572614e-05,
"loss": 0.3844,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13450,
"tokens_per_second_per_gpu": 2316.19
},
{
"epoch": 1.3441182344717395,
"grad_norm": 2.28125,
"learning_rate": 1.8688461977339814e-05,
"loss": 0.3267,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13460,
"tokens_per_second_per_gpu": 2569.43
},
{
"epoch": 1.3451168364289994,
"grad_norm": 3.1875,
"learning_rate": 1.8679938688376003e-05,
"loss": 0.3688,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13470,
"tokens_per_second_per_gpu": 2459.37
},
{
"epoch": 1.3461154383862592,
"grad_norm": 3.1875,
"learning_rate": 1.8671411594928287e-05,
"loss": 0.4114,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13480,
"tokens_per_second_per_gpu": 2405.61
},
{
"epoch": 1.3471140403435191,
"grad_norm": 3.328125,
"learning_rate": 1.866288070224609e-05,
"loss": 0.3695,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13490,
"tokens_per_second_per_gpu": 2508.66
},
{
"epoch": 1.348112642300779,
"grad_norm": 2.9375,
"learning_rate": 1.8654346015581193e-05,
"loss": 0.4449,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13500,
"tokens_per_second_per_gpu": 2392.62
},
{
"epoch": 1.3491112442580389,
"grad_norm": 3.125,
"learning_rate": 1.8645807540187703e-05,
"loss": 0.3977,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13510,
"tokens_per_second_per_gpu": 2739.47
},
{
"epoch": 1.3501098462152985,
"grad_norm": 2.859375,
"learning_rate": 1.8637265281322064e-05,
"loss": 0.3644,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13520,
"tokens_per_second_per_gpu": 2443.57
},
{
"epoch": 1.3511084481725584,
"grad_norm": 3.59375,
"learning_rate": 1.8628719244243047e-05,
"loss": 0.384,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13530,
"tokens_per_second_per_gpu": 2350.35
},
{
"epoch": 1.3521070501298182,
"grad_norm": 3.484375,
"learning_rate": 1.8620169434211752e-05,
"loss": 0.3938,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13540,
"tokens_per_second_per_gpu": 2295.45
},
{
"epoch": 1.3531056520870781,
"grad_norm": 3.484375,
"learning_rate": 1.8611615856491592e-05,
"loss": 0.3747,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13550,
"tokens_per_second_per_gpu": 2464.54
},
{
"epoch": 1.354104254044338,
"grad_norm": 3.578125,
"learning_rate": 1.8603058516348317e-05,
"loss": 0.3998,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13560,
"tokens_per_second_per_gpu": 2509.76
},
{
"epoch": 1.3551028560015976,
"grad_norm": 3.203125,
"learning_rate": 1.859449741904998e-05,
"loss": 0.3876,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13570,
"tokens_per_second_per_gpu": 2473.81
},
{
"epoch": 1.3561014579588577,
"grad_norm": 2.59375,
"learning_rate": 1.858593256986694e-05,
"loss": 0.3959,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13580,
"tokens_per_second_per_gpu": 2517.66
},
{
"epoch": 1.3571000599161174,
"grad_norm": 3.125,
"learning_rate": 1.8577363974071885e-05,
"loss": 0.419,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13590,
"tokens_per_second_per_gpu": 2492.45
},
{
"epoch": 1.3580986618733772,
"grad_norm": 2.59375,
"learning_rate": 1.8568791636939804e-05,
"loss": 0.3713,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13600,
"tokens_per_second_per_gpu": 2336.32
},
{
"epoch": 1.3590972638306371,
"grad_norm": 2.484375,
"learning_rate": 1.8560215563747972e-05,
"loss": 0.37,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13610,
"tokens_per_second_per_gpu": 2402.96
},
{
"epoch": 1.360095865787897,
"grad_norm": 2.859375,
"learning_rate": 1.8551635759775993e-05,
"loss": 0.3525,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13620,
"tokens_per_second_per_gpu": 2539.34
},
{
"epoch": 1.3610944677451569,
"grad_norm": 3.390625,
"learning_rate": 1.8543052230305745e-05,
"loss": 0.3836,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13630,
"tokens_per_second_per_gpu": 2535.64
},
{
"epoch": 1.3620930697024165,
"grad_norm": 3.265625,
"learning_rate": 1.85344649806214e-05,
"loss": 0.3954,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13640,
"tokens_per_second_per_gpu": 2637.16
},
{
"epoch": 1.3630916716596764,
"grad_norm": 3.125,
"learning_rate": 1.8525874016009438e-05,
"loss": 0.4344,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13650,
"tokens_per_second_per_gpu": 2720.16
},
{
"epoch": 1.3640902736169362,
"grad_norm": 3.21875,
"learning_rate": 1.8517279341758603e-05,
"loss": 0.3406,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13660,
"tokens_per_second_per_gpu": 2368.04
},
{
"epoch": 1.3650888755741961,
"grad_norm": 3.890625,
"learning_rate": 1.8508680963159948e-05,
"loss": 0.357,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13670,
"tokens_per_second_per_gpu": 2415.23
},
{
"epoch": 1.366087477531456,
"grad_norm": 3.921875,
"learning_rate": 1.8500078885506782e-05,
"loss": 0.3915,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13680,
"tokens_per_second_per_gpu": 2376.95
},
{
"epoch": 1.3670860794887159,
"grad_norm": 3.0625,
"learning_rate": 1.849147311409471e-05,
"loss": 0.3563,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13690,
"tokens_per_second_per_gpu": 2459.44
},
{
"epoch": 1.3680846814459757,
"grad_norm": 2.96875,
"learning_rate": 1.84828636542216e-05,
"loss": 0.4365,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13700,
"tokens_per_second_per_gpu": 2522.2
},
{
"epoch": 1.3690832834032354,
"grad_norm": 2.890625,
"learning_rate": 1.8474250511187592e-05,
"loss": 0.3843,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13710,
"tokens_per_second_per_gpu": 2474.81
},
{
"epoch": 1.3700818853604952,
"grad_norm": 3.015625,
"learning_rate": 1.8465633690295096e-05,
"loss": 0.3662,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13720,
"tokens_per_second_per_gpu": 2656.01
},
{
"epoch": 1.3710804873177551,
"grad_norm": 3.4375,
"learning_rate": 1.8457013196848785e-05,
"loss": 0.3665,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13730,
"tokens_per_second_per_gpu": 2469.85
},
{
"epoch": 1.372079089275015,
"grad_norm": 3.1875,
"learning_rate": 1.8448389036155596e-05,
"loss": 0.3912,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13740,
"tokens_per_second_per_gpu": 2441.76
},
{
"epoch": 1.3730776912322749,
"grad_norm": 2.703125,
"learning_rate": 1.843976121352472e-05,
"loss": 0.3973,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13750,
"tokens_per_second_per_gpu": 2395.5
},
{
"epoch": 1.3740762931895347,
"grad_norm": 2.46875,
"learning_rate": 1.8431129734267603e-05,
"loss": 0.3535,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13760,
"tokens_per_second_per_gpu": 2453.36
},
{
"epoch": 1.3750748951467946,
"grad_norm": 3.359375,
"learning_rate": 1.8422494603697944e-05,
"loss": 0.3434,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13770,
"tokens_per_second_per_gpu": 2432.84
},
{
"epoch": 1.3760734971040542,
"grad_norm": 3.34375,
"learning_rate": 1.8413855827131678e-05,
"loss": 0.3667,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13780,
"tokens_per_second_per_gpu": 2322.23
},
{
"epoch": 1.3770720990613141,
"grad_norm": 3.328125,
"learning_rate": 1.840521340988701e-05,
"loss": 0.4192,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13790,
"tokens_per_second_per_gpu": 2480.02
},
{
"epoch": 1.378070701018574,
"grad_norm": 3.0,
"learning_rate": 1.839656735728436e-05,
"loss": 0.3488,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13800,
"tokens_per_second_per_gpu": 2582.3
},
{
"epoch": 1.3790693029758339,
"grad_norm": 3.46875,
"learning_rate": 1.8387917674646408e-05,
"loss": 0.4193,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13810,
"tokens_per_second_per_gpu": 2353.88
},
{
"epoch": 1.3800679049330937,
"grad_norm": 3.0625,
"learning_rate": 1.8379264367298043e-05,
"loss": 0.3917,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13820,
"tokens_per_second_per_gpu": 2423.44
},
{
"epoch": 1.3810665068903534,
"grad_norm": 3.109375,
"learning_rate": 1.8370607440566412e-05,
"loss": 0.3822,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13830,
"tokens_per_second_per_gpu": 2481.1
},
{
"epoch": 1.3820651088476135,
"grad_norm": 2.609375,
"learning_rate": 1.8361946899780874e-05,
"loss": 0.3541,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13840,
"tokens_per_second_per_gpu": 2494.78
},
{
"epoch": 1.3830637108048731,
"grad_norm": 2.84375,
"learning_rate": 1.8353282750273016e-05,
"loss": 0.3726,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13850,
"tokens_per_second_per_gpu": 2584.91
},
{
"epoch": 1.384062312762133,
"grad_norm": 5.0,
"learning_rate": 1.8344614997376646e-05,
"loss": 0.4346,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13860,
"tokens_per_second_per_gpu": 2364.65
},
{
"epoch": 1.3850609147193929,
"grad_norm": 4.09375,
"learning_rate": 1.8335943646427803e-05,
"loss": 0.3944,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13870,
"tokens_per_second_per_gpu": 2473.15
},
{
"epoch": 1.3860595166766527,
"grad_norm": 3.46875,
"learning_rate": 1.832726870276472e-05,
"loss": 0.3994,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13880,
"tokens_per_second_per_gpu": 2438.66
},
{
"epoch": 1.3870581186339126,
"grad_norm": 3.53125,
"learning_rate": 1.8318590171727846e-05,
"loss": 0.4132,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13890,
"tokens_per_second_per_gpu": 2510.74
},
{
"epoch": 1.3880567205911722,
"grad_norm": 2.859375,
"learning_rate": 1.8309908058659858e-05,
"loss": 0.3826,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13900,
"tokens_per_second_per_gpu": 2403.19
},
{
"epoch": 1.3890553225484321,
"grad_norm": 3.671875,
"learning_rate": 1.8301222368905624e-05,
"loss": 0.3764,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13910,
"tokens_per_second_per_gpu": 2565.64
},
{
"epoch": 1.390053924505692,
"grad_norm": 3.0625,
"learning_rate": 1.8292533107812202e-05,
"loss": 0.365,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13920,
"tokens_per_second_per_gpu": 2447.79
},
{
"epoch": 1.3910525264629519,
"grad_norm": 3.140625,
"learning_rate": 1.828384028072887e-05,
"loss": 0.3508,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13930,
"tokens_per_second_per_gpu": 2596.01
},
{
"epoch": 1.3920511284202117,
"grad_norm": 3.15625,
"learning_rate": 1.8275143893007092e-05,
"loss": 0.3851,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13940,
"tokens_per_second_per_gpu": 2685.4
},
{
"epoch": 1.3930497303774716,
"grad_norm": 4.0,
"learning_rate": 1.826644395000052e-05,
"loss": 0.3702,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13950,
"tokens_per_second_per_gpu": 2200.93
},
{
"epoch": 1.3940483323347315,
"grad_norm": 3.375,
"learning_rate": 1.8257740457065005e-05,
"loss": 0.3888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13960,
"tokens_per_second_per_gpu": 2421.89
},
{
"epoch": 1.3950469342919911,
"grad_norm": 3.15625,
"learning_rate": 1.824903341955857e-05,
"loss": 0.4062,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13970,
"tokens_per_second_per_gpu": 2428.18
},
{
"epoch": 1.396045536249251,
"grad_norm": 3.46875,
"learning_rate": 1.8240322842841438e-05,
"loss": 0.4059,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13980,
"tokens_per_second_per_gpu": 2494.14
},
{
"epoch": 1.3970441382065109,
"grad_norm": 3.015625,
"learning_rate": 1.8231608732275987e-05,
"loss": 0.3774,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 13990,
"tokens_per_second_per_gpu": 2608.97
},
{
"epoch": 1.3980427401637707,
"grad_norm": 3.140625,
"learning_rate": 1.8222891093226794e-05,
"loss": 0.3585,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14000,
"tokens_per_second_per_gpu": 2368.11
},
{
"epoch": 1.3990413421210306,
"grad_norm": 3.0625,
"learning_rate": 1.8214169931060592e-05,
"loss": 0.3372,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14010,
"tokens_per_second_per_gpu": 2559.87
},
{
"epoch": 1.4000399440782905,
"grad_norm": 3.34375,
"learning_rate": 1.8205445251146292e-05,
"loss": 0.3774,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14020,
"tokens_per_second_per_gpu": 2380.15
},
{
"epoch": 1.4010385460355503,
"grad_norm": 3.359375,
"learning_rate": 1.8196717058854968e-05,
"loss": 0.4067,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14030,
"tokens_per_second_per_gpu": 2424.4
},
{
"epoch": 1.40203714799281,
"grad_norm": 3.453125,
"learning_rate": 1.818798535955985e-05,
"loss": 0.388,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14040,
"tokens_per_second_per_gpu": 2526.65
},
{
"epoch": 1.4030357499500699,
"grad_norm": 2.421875,
"learning_rate": 1.817925015863634e-05,
"loss": 0.3609,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14050,
"tokens_per_second_per_gpu": 2341.9
},
{
"epoch": 1.4040343519073297,
"grad_norm": 3.796875,
"learning_rate": 1.8170511461461983e-05,
"loss": 0.3618,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14060,
"tokens_per_second_per_gpu": 2271.31
},
{
"epoch": 1.4050329538645896,
"grad_norm": 2.625,
"learning_rate": 1.8161769273416487e-05,
"loss": 0.3987,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14070,
"tokens_per_second_per_gpu": 2431.4
},
{
"epoch": 1.4060315558218495,
"grad_norm": 2.578125,
"learning_rate": 1.8153023599881696e-05,
"loss": 0.3864,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14080,
"tokens_per_second_per_gpu": 2401.04
},
{
"epoch": 1.4070301577791091,
"grad_norm": 4.3125,
"learning_rate": 1.8144274446241616e-05,
"loss": 0.3559,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14090,
"tokens_per_second_per_gpu": 2417.15
},
{
"epoch": 1.4080287597363692,
"grad_norm": 3.0,
"learning_rate": 1.813552181788238e-05,
"loss": 0.3558,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14100,
"tokens_per_second_per_gpu": 2652.09
},
{
"epoch": 1.4090273616936289,
"grad_norm": 3.4375,
"learning_rate": 1.8126765720192272e-05,
"loss": 0.3888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14110,
"tokens_per_second_per_gpu": 2313.18
},
{
"epoch": 1.4100259636508887,
"grad_norm": 4.40625,
"learning_rate": 1.8118006158561703e-05,
"loss": 0.3818,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14120,
"tokens_per_second_per_gpu": 2435.47
},
{
"epoch": 1.4110245656081486,
"grad_norm": 4.9375,
"learning_rate": 1.8109243138383225e-05,
"loss": 0.4159,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14130,
"tokens_per_second_per_gpu": 2407.77
},
{
"epoch": 1.4120231675654085,
"grad_norm": 2.296875,
"learning_rate": 1.810047666505151e-05,
"loss": 0.3539,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14140,
"tokens_per_second_per_gpu": 2324.34
},
{
"epoch": 1.4130217695226683,
"grad_norm": 3.75,
"learning_rate": 1.809170674396336e-05,
"loss": 0.4174,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14150,
"tokens_per_second_per_gpu": 2540.51
},
{
"epoch": 1.414020371479928,
"grad_norm": 3.328125,
"learning_rate": 1.8082933380517703e-05,
"loss": 0.3493,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14160,
"tokens_per_second_per_gpu": 2597.51
},
{
"epoch": 1.4150189734371879,
"grad_norm": 4.25,
"learning_rate": 1.807415658011558e-05,
"loss": 0.3734,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14170,
"tokens_per_second_per_gpu": 2397.17
},
{
"epoch": 1.4160175753944477,
"grad_norm": 2.734375,
"learning_rate": 1.8065376348160153e-05,
"loss": 0.3544,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14180,
"tokens_per_second_per_gpu": 2422.33
},
{
"epoch": 1.4170161773517076,
"grad_norm": 3.265625,
"learning_rate": 1.805659269005669e-05,
"loss": 0.3975,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14190,
"tokens_per_second_per_gpu": 2255.13
},
{
"epoch": 1.4180147793089675,
"grad_norm": 4.53125,
"learning_rate": 1.8047805611212582e-05,
"loss": 0.4127,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14200,
"tokens_per_second_per_gpu": 2535.8
},
{
"epoch": 1.4190133812662273,
"grad_norm": 2.953125,
"learning_rate": 1.8039015117037303e-05,
"loss": 0.3867,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14210,
"tokens_per_second_per_gpu": 2256.29
},
{
"epoch": 1.4200119832234872,
"grad_norm": 3.125,
"learning_rate": 1.803022121294245e-05,
"loss": 0.3997,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14220,
"tokens_per_second_per_gpu": 2385.86
},
{
"epoch": 1.4210105851807469,
"grad_norm": 2.28125,
"learning_rate": 1.8021423904341715e-05,
"loss": 0.3652,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14230,
"tokens_per_second_per_gpu": 2463.36
},
{
"epoch": 1.4220091871380067,
"grad_norm": 3.140625,
"learning_rate": 1.801262319665088e-05,
"loss": 0.3556,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14240,
"tokens_per_second_per_gpu": 2540.35
},
{
"epoch": 1.4230077890952666,
"grad_norm": 3.265625,
"learning_rate": 1.8003819095287822e-05,
"loss": 0.3672,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14250,
"tokens_per_second_per_gpu": 2494.33
},
{
"epoch": 1.4240063910525265,
"grad_norm": 2.921875,
"learning_rate": 1.7995011605672508e-05,
"loss": 0.3582,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14260,
"tokens_per_second_per_gpu": 2249.36
},
{
"epoch": 1.4250049930097863,
"grad_norm": 4.46875,
"learning_rate": 1.7986200733226987e-05,
"loss": 0.3847,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14270,
"tokens_per_second_per_gpu": 2399.15
},
{
"epoch": 1.4260035949670462,
"grad_norm": 4.5625,
"learning_rate": 1.79773864833754e-05,
"loss": 0.4352,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14280,
"tokens_per_second_per_gpu": 2500.82
},
{
"epoch": 1.427002196924306,
"grad_norm": 3.3125,
"learning_rate": 1.7968568861543962e-05,
"loss": 0.3915,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14290,
"tokens_per_second_per_gpu": 2629.08
},
{
"epoch": 1.4280007988815657,
"grad_norm": 2.796875,
"learning_rate": 1.7959747873160958e-05,
"loss": 0.3713,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14300,
"tokens_per_second_per_gpu": 2215.62
},
{
"epoch": 1.4289994008388256,
"grad_norm": 3.140625,
"learning_rate": 1.7950923523656755e-05,
"loss": 0.3707,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14310,
"tokens_per_second_per_gpu": 2543.73
},
{
"epoch": 1.4299980027960855,
"grad_norm": 2.8125,
"learning_rate": 1.794209581846378e-05,
"loss": 0.3954,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14320,
"tokens_per_second_per_gpu": 2530.03
},
{
"epoch": 1.4309966047533453,
"grad_norm": 3.0,
"learning_rate": 1.7933264763016537e-05,
"loss": 0.4088,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14330,
"tokens_per_second_per_gpu": 2527.04
},
{
"epoch": 1.4319952067106052,
"grad_norm": 3.421875,
"learning_rate": 1.7924430362751583e-05,
"loss": 0.3975,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14340,
"tokens_per_second_per_gpu": 2648.37
},
{
"epoch": 1.4329938086678649,
"grad_norm": 2.671875,
"learning_rate": 1.791559262310753e-05,
"loss": 0.346,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14350,
"tokens_per_second_per_gpu": 2561.03
},
{
"epoch": 1.433992410625125,
"grad_norm": 3.375,
"learning_rate": 1.7906751549525064e-05,
"loss": 0.3716,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14360,
"tokens_per_second_per_gpu": 2398.54
},
{
"epoch": 1.4349910125823846,
"grad_norm": 3.046875,
"learning_rate": 1.7897907147446907e-05,
"loss": 0.421,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14370,
"tokens_per_second_per_gpu": 2465.9
},
{
"epoch": 1.4359896145396445,
"grad_norm": 2.328125,
"learning_rate": 1.788905942231784e-05,
"loss": 0.3397,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14380,
"tokens_per_second_per_gpu": 2608.94
},
{
"epoch": 1.4369882164969043,
"grad_norm": 3.203125,
"learning_rate": 1.788020837958468e-05,
"loss": 0.385,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14390,
"tokens_per_second_per_gpu": 2346.23
},
{
"epoch": 1.4379868184541642,
"grad_norm": 2.84375,
"learning_rate": 1.7871354024696296e-05,
"loss": 0.4135,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14400,
"tokens_per_second_per_gpu": 2606.42
},
{
"epoch": 1.438985420411424,
"grad_norm": 3.484375,
"learning_rate": 1.7862496363103586e-05,
"loss": 0.4197,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14410,
"tokens_per_second_per_gpu": 2377.24
},
{
"epoch": 1.4399840223686837,
"grad_norm": 3.90625,
"learning_rate": 1.7853635400259494e-05,
"loss": 0.4108,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14420,
"tokens_per_second_per_gpu": 2328.55
},
{
"epoch": 1.4409826243259438,
"grad_norm": 3.75,
"learning_rate": 1.7844771141618996e-05,
"loss": 0.4176,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14430,
"tokens_per_second_per_gpu": 2507.0
},
{
"epoch": 1.4419812262832035,
"grad_norm": 4.0625,
"learning_rate": 1.783590359263908e-05,
"loss": 0.4028,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14440,
"tokens_per_second_per_gpu": 2356.69
},
{
"epoch": 1.4429798282404633,
"grad_norm": 2.59375,
"learning_rate": 1.7827032758778787e-05,
"loss": 0.342,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14450,
"tokens_per_second_per_gpu": 2458.58
},
{
"epoch": 1.4439784301977232,
"grad_norm": 3.625,
"learning_rate": 1.7818158645499157e-05,
"loss": 0.3736,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14460,
"tokens_per_second_per_gpu": 2326.45
},
{
"epoch": 1.444977032154983,
"grad_norm": 2.28125,
"learning_rate": 1.7809281258263253e-05,
"loss": 0.3273,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14470,
"tokens_per_second_per_gpu": 2542.3
},
{
"epoch": 1.445975634112243,
"grad_norm": 3.21875,
"learning_rate": 1.7800400602536165e-05,
"loss": 0.4091,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14480,
"tokens_per_second_per_gpu": 2391.44
},
{
"epoch": 1.4469742360695026,
"grad_norm": 3.90625,
"learning_rate": 1.7791516683784986e-05,
"loss": 0.4127,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14490,
"tokens_per_second_per_gpu": 2525.83
},
{
"epoch": 1.4479728380267625,
"grad_norm": 3.203125,
"learning_rate": 1.7782629507478822e-05,
"loss": 0.3897,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14500,
"tokens_per_second_per_gpu": 2438.37
},
{
"epoch": 1.4489714399840223,
"grad_norm": 3.328125,
"learning_rate": 1.777373907908878e-05,
"loss": 0.3746,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14510,
"tokens_per_second_per_gpu": 2547.23
},
{
"epoch": 1.4499700419412822,
"grad_norm": 4.09375,
"learning_rate": 1.7764845404087973e-05,
"loss": 0.4196,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14520,
"tokens_per_second_per_gpu": 2428.02
},
{
"epoch": 1.450968643898542,
"grad_norm": 3.421875,
"learning_rate": 1.775594848795151e-05,
"loss": 0.3911,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14530,
"tokens_per_second_per_gpu": 2559.47
},
{
"epoch": 1.451967245855802,
"grad_norm": 2.75,
"learning_rate": 1.774704833615649e-05,
"loss": 0.3852,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14540,
"tokens_per_second_per_gpu": 2346.01
},
{
"epoch": 1.4529658478130618,
"grad_norm": 2.234375,
"learning_rate": 1.7738144954182018e-05,
"loss": 0.4694,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14550,
"tokens_per_second_per_gpu": 2335.62
},
{
"epoch": 1.4539644497703215,
"grad_norm": 4.34375,
"learning_rate": 1.772923834750918e-05,
"loss": 0.4106,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14560,
"tokens_per_second_per_gpu": 2558.92
},
{
"epoch": 1.4549630517275813,
"grad_norm": 2.84375,
"learning_rate": 1.7720328521621043e-05,
"loss": 0.377,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14570,
"tokens_per_second_per_gpu": 2495.57
},
{
"epoch": 1.4559616536848412,
"grad_norm": 3.125,
"learning_rate": 1.771141548200266e-05,
"loss": 0.4332,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14580,
"tokens_per_second_per_gpu": 2532.99
},
{
"epoch": 1.456960255642101,
"grad_norm": 3.46875,
"learning_rate": 1.770249923414106e-05,
"loss": 0.3615,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14590,
"tokens_per_second_per_gpu": 2375.0
},
{
"epoch": 1.457958857599361,
"grad_norm": 3.625,
"learning_rate": 1.769357978352526e-05,
"loss": 0.417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14600,
"tokens_per_second_per_gpu": 2435.03
},
{
"epoch": 1.4589574595566206,
"grad_norm": 3.328125,
"learning_rate": 1.7684657135646224e-05,
"loss": 0.3695,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14610,
"tokens_per_second_per_gpu": 2424.58
},
{
"epoch": 1.4599560615138807,
"grad_norm": 3.59375,
"learning_rate": 1.7675731295996906e-05,
"loss": 0.3703,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14620,
"tokens_per_second_per_gpu": 2549.49
},
{
"epoch": 1.4609546634711403,
"grad_norm": 3.109375,
"learning_rate": 1.7666802270072214e-05,
"loss": 0.3275,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14630,
"tokens_per_second_per_gpu": 2504.69
},
{
"epoch": 1.4619532654284002,
"grad_norm": 3.828125,
"learning_rate": 1.7657870063369025e-05,
"loss": 0.3766,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14640,
"tokens_per_second_per_gpu": 2531.29
},
{
"epoch": 1.46295186738566,
"grad_norm": 3.375,
"learning_rate": 1.764893468138616e-05,
"loss": 0.3736,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14650,
"tokens_per_second_per_gpu": 2378.99
},
{
"epoch": 1.46395046934292,
"grad_norm": 3.40625,
"learning_rate": 1.763999612962442e-05,
"loss": 0.3818,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14660,
"tokens_per_second_per_gpu": 2599.83
},
{
"epoch": 1.4649490713001798,
"grad_norm": 3.453125,
"learning_rate": 1.7631054413586526e-05,
"loss": 0.399,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14670,
"tokens_per_second_per_gpu": 2503.43
},
{
"epoch": 1.4659476732574395,
"grad_norm": 2.984375,
"learning_rate": 1.762210953877717e-05,
"loss": 0.4223,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14680,
"tokens_per_second_per_gpu": 2326.02
},
{
"epoch": 1.4669462752146996,
"grad_norm": 3.546875,
"learning_rate": 1.7613161510702984e-05,
"loss": 0.3965,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14690,
"tokens_per_second_per_gpu": 2438.41
},
{
"epoch": 1.4679448771719592,
"grad_norm": 3.390625,
"learning_rate": 1.7604210334872535e-05,
"loss": 0.387,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14700,
"tokens_per_second_per_gpu": 2452.81
},
{
"epoch": 1.468943479129219,
"grad_norm": 3.859375,
"learning_rate": 1.759525601679633e-05,
"loss": 0.3823,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14710,
"tokens_per_second_per_gpu": 2365.33
},
{
"epoch": 1.469942081086479,
"grad_norm": 3.59375,
"learning_rate": 1.758629856198682e-05,
"loss": 0.3737,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14720,
"tokens_per_second_per_gpu": 2322.62
},
{
"epoch": 1.4709406830437388,
"grad_norm": 3.015625,
"learning_rate": 1.7577337975958365e-05,
"loss": 0.3295,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14730,
"tokens_per_second_per_gpu": 2382.1
},
{
"epoch": 1.4719392850009987,
"grad_norm": 3.125,
"learning_rate": 1.7568374264227278e-05,
"loss": 0.4049,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14740,
"tokens_per_second_per_gpu": 2532.38
},
{
"epoch": 1.4729378869582583,
"grad_norm": 3.390625,
"learning_rate": 1.755940743231178e-05,
"loss": 0.3773,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14750,
"tokens_per_second_per_gpu": 2325.55
},
{
"epoch": 1.4739364889155182,
"grad_norm": 3.21875,
"learning_rate": 1.755043748573202e-05,
"loss": 0.3936,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14760,
"tokens_per_second_per_gpu": 2339.45
},
{
"epoch": 1.474935090872778,
"grad_norm": 3.34375,
"learning_rate": 1.7541464430010065e-05,
"loss": 0.347,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14770,
"tokens_per_second_per_gpu": 2491.14
},
{
"epoch": 1.475933692830038,
"grad_norm": 3.71875,
"learning_rate": 1.7532488270669886e-05,
"loss": 0.4098,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14780,
"tokens_per_second_per_gpu": 2424.67
},
{
"epoch": 1.4769322947872978,
"grad_norm": 3.625,
"learning_rate": 1.7523509013237374e-05,
"loss": 0.376,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14790,
"tokens_per_second_per_gpu": 2711.66
},
{
"epoch": 1.4779308967445577,
"grad_norm": 3.328125,
"learning_rate": 1.7514526663240327e-05,
"loss": 0.3495,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14800,
"tokens_per_second_per_gpu": 2447.48
},
{
"epoch": 1.4789294987018176,
"grad_norm": 3.625,
"learning_rate": 1.7505541226208445e-05,
"loss": 0.3946,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14810,
"tokens_per_second_per_gpu": 2500.56
},
{
"epoch": 1.4799281006590772,
"grad_norm": 4.21875,
"learning_rate": 1.7496552707673323e-05,
"loss": 0.3895,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14820,
"tokens_per_second_per_gpu": 2558.36
},
{
"epoch": 1.480926702616337,
"grad_norm": 4.21875,
"learning_rate": 1.7487561113168465e-05,
"loss": 0.4024,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14830,
"tokens_per_second_per_gpu": 2349.8
},
{
"epoch": 1.481925304573597,
"grad_norm": 3.421875,
"learning_rate": 1.7478566448229262e-05,
"loss": 0.3787,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14840,
"tokens_per_second_per_gpu": 2515.28
},
{
"epoch": 1.4829239065308568,
"grad_norm": 3.53125,
"learning_rate": 1.746956871839299e-05,
"loss": 0.401,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14850,
"tokens_per_second_per_gpu": 2328.94
},
{
"epoch": 1.4839225084881167,
"grad_norm": 3.15625,
"learning_rate": 1.746056792919882e-05,
"loss": 0.4052,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14860,
"tokens_per_second_per_gpu": 2512.94
},
{
"epoch": 1.4849211104453763,
"grad_norm": 3.453125,
"learning_rate": 1.7451564086187804e-05,
"loss": 0.4168,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14870,
"tokens_per_second_per_gpu": 2592.81
},
{
"epoch": 1.4859197124026364,
"grad_norm": 4.03125,
"learning_rate": 1.7442557194902868e-05,
"loss": 0.4032,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14880,
"tokens_per_second_per_gpu": 2505.58
},
{
"epoch": 1.486918314359896,
"grad_norm": 3.625,
"learning_rate": 1.7433547260888828e-05,
"loss": 0.3497,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14890,
"tokens_per_second_per_gpu": 2436.34
},
{
"epoch": 1.487916916317156,
"grad_norm": 3.1875,
"learning_rate": 1.742453428969236e-05,
"loss": 0.4103,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14900,
"tokens_per_second_per_gpu": 2397.9
},
{
"epoch": 1.4889155182744158,
"grad_norm": 3.984375,
"learning_rate": 1.7415518286862018e-05,
"loss": 0.4364,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14910,
"tokens_per_second_per_gpu": 2369.47
},
{
"epoch": 1.4899141202316757,
"grad_norm": 2.734375,
"learning_rate": 1.7406499257948218e-05,
"loss": 0.3598,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14920,
"tokens_per_second_per_gpu": 2579.24
},
{
"epoch": 1.4909127221889356,
"grad_norm": 2.46875,
"learning_rate": 1.7397477208503233e-05,
"loss": 0.4074,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14930,
"tokens_per_second_per_gpu": 2455.61
},
{
"epoch": 1.4919113241461952,
"grad_norm": 3.453125,
"learning_rate": 1.738845214408121e-05,
"loss": 0.3815,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14940,
"tokens_per_second_per_gpu": 2273.67
},
{
"epoch": 1.4929099261034553,
"grad_norm": 2.515625,
"learning_rate": 1.7379424070238148e-05,
"loss": 0.3593,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14950,
"tokens_per_second_per_gpu": 2515.66
},
{
"epoch": 1.493908528060715,
"grad_norm": 3.21875,
"learning_rate": 1.7370392992531893e-05,
"loss": 0.44,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14960,
"tokens_per_second_per_gpu": 2592.23
},
{
"epoch": 1.4949071300179748,
"grad_norm": 2.953125,
"learning_rate": 1.7361358916522132e-05,
"loss": 0.3682,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14970,
"tokens_per_second_per_gpu": 2421.75
},
{
"epoch": 1.4959057319752347,
"grad_norm": 3.53125,
"learning_rate": 1.7352321847770422e-05,
"loss": 0.3788,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14980,
"tokens_per_second_per_gpu": 2491.22
},
{
"epoch": 1.4969043339324946,
"grad_norm": 2.96875,
"learning_rate": 1.734328179184014e-05,
"loss": 0.3887,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 14990,
"tokens_per_second_per_gpu": 2465.86
},
{
"epoch": 1.4979029358897544,
"grad_norm": 3.484375,
"learning_rate": 1.7334238754296515e-05,
"loss": 0.3824,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15000,
"tokens_per_second_per_gpu": 2508.14
},
{
"epoch": 1.498901537847014,
"grad_norm": 3.96875,
"learning_rate": 1.732519274070661e-05,
"loss": 0.3574,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15010,
"tokens_per_second_per_gpu": 2321.71
},
{
"epoch": 1.499900139804274,
"grad_norm": 3.234375,
"learning_rate": 1.7316143756639308e-05,
"loss": 0.3824,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15020,
"tokens_per_second_per_gpu": 2506.04
},
{
"epoch": 1.5008987417615338,
"grad_norm": 3.15625,
"learning_rate": 1.730709180766534e-05,
"loss": 0.4269,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15030,
"tokens_per_second_per_gpu": 2398.83
},
{
"epoch": 1.5018973437187937,
"grad_norm": 3.953125,
"learning_rate": 1.7298036899357247e-05,
"loss": 0.3515,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15040,
"tokens_per_second_per_gpu": 2436.11
},
{
"epoch": 1.5028959456760536,
"grad_norm": 2.59375,
"learning_rate": 1.7288979037289398e-05,
"loss": 0.3068,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15050,
"tokens_per_second_per_gpu": 2531.2
},
{
"epoch": 1.5038945476333132,
"grad_norm": 3.25,
"learning_rate": 1.727991822703798e-05,
"loss": 0.393,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15060,
"tokens_per_second_per_gpu": 2419.24
},
{
"epoch": 1.5048931495905733,
"grad_norm": 3.734375,
"learning_rate": 1.7270854474180992e-05,
"loss": 0.3523,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15070,
"tokens_per_second_per_gpu": 2188.69
},
{
"epoch": 1.505891751547833,
"grad_norm": 3.578125,
"learning_rate": 1.7261787784298247e-05,
"loss": 0.3763,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15080,
"tokens_per_second_per_gpu": 2557.32
},
{
"epoch": 1.5068903535050928,
"grad_norm": 3.265625,
"learning_rate": 1.725271816297137e-05,
"loss": 0.4171,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15090,
"tokens_per_second_per_gpu": 2519.01
},
{
"epoch": 1.5078889554623527,
"grad_norm": 3.734375,
"learning_rate": 1.7243645615783786e-05,
"loss": 0.4131,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15100,
"tokens_per_second_per_gpu": 2335.78
},
{
"epoch": 1.5088875574196126,
"grad_norm": 3.4375,
"learning_rate": 1.7234570148320718e-05,
"loss": 0.4079,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15110,
"tokens_per_second_per_gpu": 2540.83
},
{
"epoch": 1.5098861593768724,
"grad_norm": 3.890625,
"learning_rate": 1.72254917661692e-05,
"loss": 0.4316,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15120,
"tokens_per_second_per_gpu": 2702.46
},
{
"epoch": 1.510884761334132,
"grad_norm": 3.8125,
"learning_rate": 1.7216410474918035e-05,
"loss": 0.3961,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15130,
"tokens_per_second_per_gpu": 2403.28
},
{
"epoch": 1.5118833632913922,
"grad_norm": 3.6875,
"learning_rate": 1.7207326280157843e-05,
"loss": 0.4156,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15140,
"tokens_per_second_per_gpu": 2541.55
},
{
"epoch": 1.5128819652486518,
"grad_norm": 3.921875,
"learning_rate": 1.7198239187481023e-05,
"loss": 0.4052,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15150,
"tokens_per_second_per_gpu": 2336.49
},
{
"epoch": 1.5138805672059117,
"grad_norm": 2.8125,
"learning_rate": 1.7189149202481757e-05,
"loss": 0.3502,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15160,
"tokens_per_second_per_gpu": 2466.51
},
{
"epoch": 1.5148791691631716,
"grad_norm": 4.34375,
"learning_rate": 1.7180056330756e-05,
"loss": 0.4123,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15170,
"tokens_per_second_per_gpu": 2307.65
},
{
"epoch": 1.5158777711204314,
"grad_norm": 3.359375,
"learning_rate": 1.7170960577901496e-05,
"loss": 0.3497,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15180,
"tokens_per_second_per_gpu": 2549.81
},
{
"epoch": 1.5168763730776913,
"grad_norm": 2.890625,
"learning_rate": 1.7161861949517753e-05,
"loss": 0.3792,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15190,
"tokens_per_second_per_gpu": 2513.9
},
{
"epoch": 1.517874975034951,
"grad_norm": 3.0625,
"learning_rate": 1.715276045120606e-05,
"loss": 0.3534,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15200,
"tokens_per_second_per_gpu": 2403.09
},
{
"epoch": 1.518873576992211,
"grad_norm": 3.078125,
"learning_rate": 1.7143656088569464e-05,
"loss": 0.3773,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15210,
"tokens_per_second_per_gpu": 2409.16
},
{
"epoch": 1.5198721789494707,
"grad_norm": 3.796875,
"learning_rate": 1.7134548867212784e-05,
"loss": 0.3718,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15220,
"tokens_per_second_per_gpu": 2386.56
},
{
"epoch": 1.5208707809067306,
"grad_norm": 2.828125,
"learning_rate": 1.7125438792742578e-05,
"loss": 0.3777,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15230,
"tokens_per_second_per_gpu": 2354.55
},
{
"epoch": 1.5218693828639904,
"grad_norm": 3.03125,
"learning_rate": 1.7116325870767192e-05,
"loss": 0.4081,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15240,
"tokens_per_second_per_gpu": 2435.87
},
{
"epoch": 1.5228679848212503,
"grad_norm": 3.109375,
"learning_rate": 1.710721010689669e-05,
"loss": 0.3687,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15250,
"tokens_per_second_per_gpu": 2706.45
},
{
"epoch": 1.5238665867785102,
"grad_norm": 2.6875,
"learning_rate": 1.7098091506742926e-05,
"loss": 0.3586,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15260,
"tokens_per_second_per_gpu": 2560.12
},
{
"epoch": 1.5248651887357698,
"grad_norm": 2.59375,
"learning_rate": 1.7088970075919465e-05,
"loss": 0.3695,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15270,
"tokens_per_second_per_gpu": 2251.71
},
{
"epoch": 1.52586379069303,
"grad_norm": 3.203125,
"learning_rate": 1.7079845820041628e-05,
"loss": 0.4037,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15280,
"tokens_per_second_per_gpu": 2287.59
},
{
"epoch": 1.5268623926502896,
"grad_norm": 2.390625,
"learning_rate": 1.7070718744726477e-05,
"loss": 0.3505,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15290,
"tokens_per_second_per_gpu": 2505.53
},
{
"epoch": 1.5278609946075494,
"grad_norm": 2.859375,
"learning_rate": 1.70615888555928e-05,
"loss": 0.3858,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15300,
"tokens_per_second_per_gpu": 2492.08
},
{
"epoch": 1.5288595965648093,
"grad_norm": 2.640625,
"learning_rate": 1.705245615826114e-05,
"loss": 0.3847,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15310,
"tokens_per_second_per_gpu": 2296.69
},
{
"epoch": 1.529858198522069,
"grad_norm": 3.546875,
"learning_rate": 1.7043320658353744e-05,
"loss": 0.3619,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15320,
"tokens_per_second_per_gpu": 2294.72
},
{
"epoch": 1.530856800479329,
"grad_norm": 3.125,
"learning_rate": 1.7034182361494588e-05,
"loss": 0.3842,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15330,
"tokens_per_second_per_gpu": 2523.71
},
{
"epoch": 1.5318554024365887,
"grad_norm": 3.859375,
"learning_rate": 1.7025041273309383e-05,
"loss": 0.388,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15340,
"tokens_per_second_per_gpu": 2586.04
},
{
"epoch": 1.5328540043938486,
"grad_norm": 3.34375,
"learning_rate": 1.7015897399425558e-05,
"loss": 0.3892,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15350,
"tokens_per_second_per_gpu": 2483.57
},
{
"epoch": 1.5338526063511084,
"grad_norm": 3.5,
"learning_rate": 1.7006750745472234e-05,
"loss": 0.3613,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15360,
"tokens_per_second_per_gpu": 2394.34
},
{
"epoch": 1.5348512083083683,
"grad_norm": 4.0,
"learning_rate": 1.699760131708027e-05,
"loss": 0.4003,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15370,
"tokens_per_second_per_gpu": 2340.48
},
{
"epoch": 1.5358498102656282,
"grad_norm": 4.0625,
"learning_rate": 1.6988449119882222e-05,
"loss": 0.4202,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15380,
"tokens_per_second_per_gpu": 2560.43
},
{
"epoch": 1.5368484122228878,
"grad_norm": 3.484375,
"learning_rate": 1.6979294159512342e-05,
"loss": 0.3417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15390,
"tokens_per_second_per_gpu": 2428.61
},
{
"epoch": 1.537847014180148,
"grad_norm": 4.0625,
"learning_rate": 1.6970136441606604e-05,
"loss": 0.408,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15400,
"tokens_per_second_per_gpu": 2555.33
},
{
"epoch": 1.5388456161374076,
"grad_norm": 3.046875,
"learning_rate": 1.6960975971802663e-05,
"loss": 0.349,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15410,
"tokens_per_second_per_gpu": 2448.83
},
{
"epoch": 1.5398442180946674,
"grad_norm": 2.828125,
"learning_rate": 1.695181275573987e-05,
"loss": 0.3828,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15420,
"tokens_per_second_per_gpu": 2488.02
},
{
"epoch": 1.5408428200519273,
"grad_norm": 2.90625,
"learning_rate": 1.694264679905927e-05,
"loss": 0.3796,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15430,
"tokens_per_second_per_gpu": 2451.29
},
{
"epoch": 1.5418414220091872,
"grad_norm": 3.703125,
"learning_rate": 1.69334781074036e-05,
"loss": 0.3655,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15440,
"tokens_per_second_per_gpu": 2449.66
},
{
"epoch": 1.542840023966447,
"grad_norm": 4.03125,
"learning_rate": 1.6924306686417265e-05,
"loss": 0.361,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15450,
"tokens_per_second_per_gpu": 2463.82
},
{
"epoch": 1.5438386259237067,
"grad_norm": 2.578125,
"learning_rate": 1.6915132541746367e-05,
"loss": 0.4395,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15460,
"tokens_per_second_per_gpu": 2592.62
},
{
"epoch": 1.5448372278809668,
"grad_norm": 3.4375,
"learning_rate": 1.690595567903868e-05,
"loss": 0.4053,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15470,
"tokens_per_second_per_gpu": 2595.8
},
{
"epoch": 1.5458358298382264,
"grad_norm": 3.328125,
"learning_rate": 1.689677610394365e-05,
"loss": 0.3858,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15480,
"tokens_per_second_per_gpu": 2285.48
},
{
"epoch": 1.5468344317954863,
"grad_norm": 2.75,
"learning_rate": 1.6887593822112382e-05,
"loss": 0.377,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15490,
"tokens_per_second_per_gpu": 2371.52
},
{
"epoch": 1.5478330337527462,
"grad_norm": 2.765625,
"learning_rate": 1.687840883919767e-05,
"loss": 0.3866,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15500,
"tokens_per_second_per_gpu": 2541.41
},
{
"epoch": 1.548831635710006,
"grad_norm": 2.765625,
"learning_rate": 1.686922116085395e-05,
"loss": 0.376,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15510,
"tokens_per_second_per_gpu": 2453.76
},
{
"epoch": 1.549830237667266,
"grad_norm": 3.203125,
"learning_rate": 1.6860030792737332e-05,
"loss": 0.3924,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15520,
"tokens_per_second_per_gpu": 2531.89
},
{
"epoch": 1.5508288396245256,
"grad_norm": 2.90625,
"learning_rate": 1.685083774050557e-05,
"loss": 0.3417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15530,
"tokens_per_second_per_gpu": 2579.53
},
{
"epoch": 1.5518274415817856,
"grad_norm": 3.5625,
"learning_rate": 1.684164200981808e-05,
"loss": 0.3608,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15540,
"tokens_per_second_per_gpu": 2339.43
},
{
"epoch": 1.5528260435390453,
"grad_norm": 2.5625,
"learning_rate": 1.6832443606335924e-05,
"loss": 0.387,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15550,
"tokens_per_second_per_gpu": 2311.16
},
{
"epoch": 1.5538246454963052,
"grad_norm": 3.484375,
"learning_rate": 1.6823242535721803e-05,
"loss": 0.399,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15560,
"tokens_per_second_per_gpu": 2565.8
},
{
"epoch": 1.554823247453565,
"grad_norm": 2.578125,
"learning_rate": 1.681403880364007e-05,
"loss": 0.4166,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15570,
"tokens_per_second_per_gpu": 2440.27
},
{
"epoch": 1.5558218494108247,
"grad_norm": 3.265625,
"learning_rate": 1.6804832415756706e-05,
"loss": 0.3627,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15580,
"tokens_per_second_per_gpu": 2443.36
},
{
"epoch": 1.5568204513680848,
"grad_norm": 3.828125,
"learning_rate": 1.6795623377739335e-05,
"loss": 0.3837,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15590,
"tokens_per_second_per_gpu": 2357.12
},
{
"epoch": 1.5578190533253444,
"grad_norm": 2.609375,
"learning_rate": 1.678641169525722e-05,
"loss": 0.3512,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15600,
"tokens_per_second_per_gpu": 2593.26
},
{
"epoch": 1.5588176552826045,
"grad_norm": 3.53125,
"learning_rate": 1.677719737398123e-05,
"loss": 0.3627,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15610,
"tokens_per_second_per_gpu": 2327.69
},
{
"epoch": 1.5598162572398642,
"grad_norm": 4.125,
"learning_rate": 1.6767980419583874e-05,
"loss": 0.411,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15620,
"tokens_per_second_per_gpu": 2452.49
},
{
"epoch": 1.560814859197124,
"grad_norm": 3.09375,
"learning_rate": 1.6758760837739275e-05,
"loss": 0.3824,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15630,
"tokens_per_second_per_gpu": 2539.93
},
{
"epoch": 1.561813461154384,
"grad_norm": 3.25,
"learning_rate": 1.674953863412319e-05,
"loss": 0.3425,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15640,
"tokens_per_second_per_gpu": 2477.64
},
{
"epoch": 1.5628120631116436,
"grad_norm": 3.796875,
"learning_rate": 1.674031381441296e-05,
"loss": 0.3683,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15650,
"tokens_per_second_per_gpu": 2301.23
},
{
"epoch": 1.5638106650689036,
"grad_norm": 3.796875,
"learning_rate": 1.6731086384287564e-05,
"loss": 0.4098,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15660,
"tokens_per_second_per_gpu": 2536.25
},
{
"epoch": 1.5648092670261633,
"grad_norm": 3.28125,
"learning_rate": 1.6721856349427582e-05,
"loss": 0.3826,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15670,
"tokens_per_second_per_gpu": 2320.2
},
{
"epoch": 1.5658078689834232,
"grad_norm": 3.390625,
"learning_rate": 1.671262371551518e-05,
"loss": 0.4028,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15680,
"tokens_per_second_per_gpu": 2408.37
},
{
"epoch": 1.566806470940683,
"grad_norm": 3.359375,
"learning_rate": 1.670338848823415e-05,
"loss": 0.3773,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15690,
"tokens_per_second_per_gpu": 2363.74
},
{
"epoch": 1.567805072897943,
"grad_norm": 4.6875,
"learning_rate": 1.669415067326986e-05,
"loss": 0.3719,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15700,
"tokens_per_second_per_gpu": 2523.0
},
{
"epoch": 1.5688036748552028,
"grad_norm": 2.234375,
"learning_rate": 1.668491027630928e-05,
"loss": 0.346,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15710,
"tokens_per_second_per_gpu": 2477.53
},
{
"epoch": 1.5698022768124624,
"grad_norm": 3.609375,
"learning_rate": 1.6675667303040965e-05,
"loss": 0.3545,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15720,
"tokens_per_second_per_gpu": 2294.56
},
{
"epoch": 1.5708008787697225,
"grad_norm": 2.828125,
"learning_rate": 1.6666421759155072e-05,
"loss": 0.3649,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15730,
"tokens_per_second_per_gpu": 2439.22
},
{
"epoch": 1.5717994807269822,
"grad_norm": 3.21875,
"learning_rate": 1.6657173650343315e-05,
"loss": 0.3792,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15740,
"tokens_per_second_per_gpu": 2629.06
},
{
"epoch": 1.572798082684242,
"grad_norm": 2.96875,
"learning_rate": 1.6647922982299004e-05,
"loss": 0.3692,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15750,
"tokens_per_second_per_gpu": 2444.12
},
{
"epoch": 1.573796684641502,
"grad_norm": 2.703125,
"learning_rate": 1.663866976071702e-05,
"loss": 0.3387,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15760,
"tokens_per_second_per_gpu": 2471.75
},
{
"epoch": 1.5747952865987618,
"grad_norm": 3.71875,
"learning_rate": 1.6629413991293815e-05,
"loss": 0.4209,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15770,
"tokens_per_second_per_gpu": 2525.28
},
{
"epoch": 1.5757938885560216,
"grad_norm": 2.984375,
"learning_rate": 1.6620155679727417e-05,
"loss": 0.4158,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15780,
"tokens_per_second_per_gpu": 2492.31
},
{
"epoch": 1.5767924905132813,
"grad_norm": 3.328125,
"learning_rate": 1.6610894831717406e-05,
"loss": 0.369,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15790,
"tokens_per_second_per_gpu": 2576.39
},
{
"epoch": 1.5777910924705414,
"grad_norm": 3.109375,
"learning_rate": 1.6601631452964935e-05,
"loss": 0.3386,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15800,
"tokens_per_second_per_gpu": 2392.57
},
{
"epoch": 1.578789694427801,
"grad_norm": 2.640625,
"learning_rate": 1.6592365549172712e-05,
"loss": 0.2971,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15810,
"tokens_per_second_per_gpu": 2363.93
},
{
"epoch": 1.579788296385061,
"grad_norm": 4.78125,
"learning_rate": 1.6583097126044986e-05,
"loss": 0.4657,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15820,
"tokens_per_second_per_gpu": 2339.79
},
{
"epoch": 1.5807868983423208,
"grad_norm": 3.359375,
"learning_rate": 1.657382618928758e-05,
"loss": 0.4158,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15830,
"tokens_per_second_per_gpu": 2494.56
},
{
"epoch": 1.5817855002995804,
"grad_norm": 4.09375,
"learning_rate": 1.656455274460785e-05,
"loss": 0.4108,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15840,
"tokens_per_second_per_gpu": 2492.41
},
{
"epoch": 1.5827841022568405,
"grad_norm": 3.359375,
"learning_rate": 1.65552767977147e-05,
"loss": 0.3768,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15850,
"tokens_per_second_per_gpu": 2358.51
},
{
"epoch": 1.5837827042141002,
"grad_norm": 2.9375,
"learning_rate": 1.6545998354318573e-05,
"loss": 0.4054,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15860,
"tokens_per_second_per_gpu": 2381.98
},
{
"epoch": 1.5847813061713603,
"grad_norm": 3.59375,
"learning_rate": 1.653671742013145e-05,
"loss": 0.4184,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15870,
"tokens_per_second_per_gpu": 2358.26
},
{
"epoch": 1.58577990812862,
"grad_norm": 3.21875,
"learning_rate": 1.652743400086684e-05,
"loss": 0.3683,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15880,
"tokens_per_second_per_gpu": 2442.09
},
{
"epoch": 1.5867785100858798,
"grad_norm": 3.359375,
"learning_rate": 1.6518148102239792e-05,
"loss": 0.3516,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15890,
"tokens_per_second_per_gpu": 2414.76
},
{
"epoch": 1.5877771120431396,
"grad_norm": 3.671875,
"learning_rate": 1.6508859729966877e-05,
"loss": 0.3894,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15900,
"tokens_per_second_per_gpu": 2349.89
},
{
"epoch": 1.5887757140003993,
"grad_norm": 4.40625,
"learning_rate": 1.6499568889766182e-05,
"loss": 0.4086,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15910,
"tokens_per_second_per_gpu": 2403.98
},
{
"epoch": 1.5897743159576594,
"grad_norm": 3.671875,
"learning_rate": 1.6490275587357323e-05,
"loss": 0.4302,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15920,
"tokens_per_second_per_gpu": 2604.09
},
{
"epoch": 1.590772917914919,
"grad_norm": 3.859375,
"learning_rate": 1.648097982846143e-05,
"loss": 0.4009,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15930,
"tokens_per_second_per_gpu": 2428.34
},
{
"epoch": 1.591771519872179,
"grad_norm": 3.28125,
"learning_rate": 1.6471681618801138e-05,
"loss": 0.3566,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15940,
"tokens_per_second_per_gpu": 2474.83
},
{
"epoch": 1.5927701218294388,
"grad_norm": 3.1875,
"learning_rate": 1.6462380964100597e-05,
"loss": 0.3766,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15950,
"tokens_per_second_per_gpu": 2553.73
},
{
"epoch": 1.5937687237866986,
"grad_norm": 4.03125,
"learning_rate": 1.6453077870085464e-05,
"loss": 0.4132,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15960,
"tokens_per_second_per_gpu": 2416.48
},
{
"epoch": 1.5947673257439585,
"grad_norm": 3.53125,
"learning_rate": 1.644377234248289e-05,
"loss": 0.3492,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15970,
"tokens_per_second_per_gpu": 2423.97
},
{
"epoch": 1.5957659277012182,
"grad_norm": 3.375,
"learning_rate": 1.6434464387021535e-05,
"loss": 0.4313,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15980,
"tokens_per_second_per_gpu": 2272.91
},
{
"epoch": 1.5967645296584783,
"grad_norm": 3.859375,
"learning_rate": 1.6425154009431547e-05,
"loss": 0.4121,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 15990,
"tokens_per_second_per_gpu": 2213.04
},
{
"epoch": 1.597763131615738,
"grad_norm": 2.21875,
"learning_rate": 1.6415841215444556e-05,
"loss": 0.337,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16000,
"tokens_per_second_per_gpu": 2411.7
},
{
"epoch": 1.5987617335729978,
"grad_norm": 3.703125,
"learning_rate": 1.6406526010793704e-05,
"loss": 0.3731,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16010,
"tokens_per_second_per_gpu": 2485.84
},
{
"epoch": 1.5997603355302576,
"grad_norm": 3.03125,
"learning_rate": 1.639720840121359e-05,
"loss": 0.3673,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16020,
"tokens_per_second_per_gpu": 2200.41
},
{
"epoch": 1.6007589374875175,
"grad_norm": 3.984375,
"learning_rate": 1.6387888392440312e-05,
"loss": 0.3673,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16030,
"tokens_per_second_per_gpu": 2526.89
},
{
"epoch": 1.6017575394447774,
"grad_norm": 3.078125,
"learning_rate": 1.637856599021144e-05,
"loss": 0.379,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16040,
"tokens_per_second_per_gpu": 2396.35
},
{
"epoch": 1.602756141402037,
"grad_norm": 2.953125,
"learning_rate": 1.6369241200266007e-05,
"loss": 0.3974,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16050,
"tokens_per_second_per_gpu": 2587.53
},
{
"epoch": 1.6037547433592971,
"grad_norm": 3.109375,
"learning_rate": 1.6359914028344535e-05,
"loss": 0.371,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16060,
"tokens_per_second_per_gpu": 2534.41
},
{
"epoch": 1.6047533453165568,
"grad_norm": 2.96875,
"learning_rate": 1.6350584480189004e-05,
"loss": 0.4175,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16070,
"tokens_per_second_per_gpu": 2456.17
},
{
"epoch": 1.6057519472738166,
"grad_norm": 3.046875,
"learning_rate": 1.6341252561542848e-05,
"loss": 0.378,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16080,
"tokens_per_second_per_gpu": 2438.63
},
{
"epoch": 1.6067505492310765,
"grad_norm": 3.46875,
"learning_rate": 1.6331918278150964e-05,
"loss": 0.351,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16090,
"tokens_per_second_per_gpu": 2432.66
},
{
"epoch": 1.6077491511883362,
"grad_norm": 4.03125,
"learning_rate": 1.6322581635759725e-05,
"loss": 0.4051,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16100,
"tokens_per_second_per_gpu": 2343.49
},
{
"epoch": 1.6087477531455963,
"grad_norm": 3.921875,
"learning_rate": 1.6313242640116923e-05,
"loss": 0.3997,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16110,
"tokens_per_second_per_gpu": 2388.94
},
{
"epoch": 1.609746355102856,
"grad_norm": 5.5625,
"learning_rate": 1.6303901296971826e-05,
"loss": 0.3979,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16120,
"tokens_per_second_per_gpu": 2396.16
},
{
"epoch": 1.610744957060116,
"grad_norm": 3.28125,
"learning_rate": 1.6294557612075136e-05,
"loss": 0.421,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16130,
"tokens_per_second_per_gpu": 2487.29
},
{
"epoch": 1.6117435590173756,
"grad_norm": 3.375,
"learning_rate": 1.6285211591178986e-05,
"loss": 0.3468,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16140,
"tokens_per_second_per_gpu": 2490.75
},
{
"epoch": 1.6127421609746355,
"grad_norm": 3.0625,
"learning_rate": 1.627586324003697e-05,
"loss": 0.3853,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16150,
"tokens_per_second_per_gpu": 2722.54
},
{
"epoch": 1.6137407629318954,
"grad_norm": 3.671875,
"learning_rate": 1.62665125644041e-05,
"loss": 0.4085,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16160,
"tokens_per_second_per_gpu": 2384.13
},
{
"epoch": 1.614739364889155,
"grad_norm": 3.390625,
"learning_rate": 1.6257159570036828e-05,
"loss": 0.4367,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16170,
"tokens_per_second_per_gpu": 2349.83
},
{
"epoch": 1.6157379668464151,
"grad_norm": 2.59375,
"learning_rate": 1.624780426269303e-05,
"loss": 0.3996,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16180,
"tokens_per_second_per_gpu": 2491.19
},
{
"epoch": 1.6167365688036748,
"grad_norm": 3.421875,
"learning_rate": 1.6238446648131996e-05,
"loss": 0.3803,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16190,
"tokens_per_second_per_gpu": 2273.36
},
{
"epoch": 1.6177351707609346,
"grad_norm": 2.28125,
"learning_rate": 1.622908673211445e-05,
"loss": 0.3972,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16200,
"tokens_per_second_per_gpu": 2573.93
},
{
"epoch": 1.6187337727181945,
"grad_norm": 3.578125,
"learning_rate": 1.6219724520402537e-05,
"loss": 0.3444,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16210,
"tokens_per_second_per_gpu": 2334.85
},
{
"epoch": 1.6197323746754544,
"grad_norm": 3.75,
"learning_rate": 1.6210360018759795e-05,
"loss": 0.3539,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16220,
"tokens_per_second_per_gpu": 2392.04
},
{
"epoch": 1.6207309766327143,
"grad_norm": 3.609375,
"learning_rate": 1.6200993232951188e-05,
"loss": 0.4295,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16230,
"tokens_per_second_per_gpu": 2191.63
},
{
"epoch": 1.621729578589974,
"grad_norm": 4.53125,
"learning_rate": 1.6191624168743086e-05,
"loss": 0.4148,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16240,
"tokens_per_second_per_gpu": 2509.97
},
{
"epoch": 1.622728180547234,
"grad_norm": 2.65625,
"learning_rate": 1.618225283190325e-05,
"loss": 0.3348,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16250,
"tokens_per_second_per_gpu": 2371.44
},
{
"epoch": 1.6237267825044936,
"grad_norm": 2.8125,
"learning_rate": 1.617287922820085e-05,
"loss": 0.3699,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16260,
"tokens_per_second_per_gpu": 2487.26
},
{
"epoch": 1.6247253844617535,
"grad_norm": 2.703125,
"learning_rate": 1.6163503363406452e-05,
"loss": 0.3729,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16270,
"tokens_per_second_per_gpu": 2480.94
},
{
"epoch": 1.6257239864190134,
"grad_norm": 2.78125,
"learning_rate": 1.6154125243292e-05,
"loss": 0.3931,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16280,
"tokens_per_second_per_gpu": 2660.41
},
{
"epoch": 1.6267225883762733,
"grad_norm": 2.84375,
"learning_rate": 1.614474487363085e-05,
"loss": 0.3987,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16290,
"tokens_per_second_per_gpu": 2618.38
},
{
"epoch": 1.6277211903335331,
"grad_norm": 4.1875,
"learning_rate": 1.6135362260197728e-05,
"loss": 0.366,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16300,
"tokens_per_second_per_gpu": 2336.9
},
{
"epoch": 1.6287197922907928,
"grad_norm": 3.609375,
"learning_rate": 1.6125977408768737e-05,
"loss": 0.381,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16310,
"tokens_per_second_per_gpu": 2433.41
},
{
"epoch": 1.6297183942480529,
"grad_norm": 2.796875,
"learning_rate": 1.611659032512137e-05,
"loss": 0.3988,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16320,
"tokens_per_second_per_gpu": 2321.94
},
{
"epoch": 1.6307169962053125,
"grad_norm": 3.75,
"learning_rate": 1.6107201015034486e-05,
"loss": 0.3719,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16330,
"tokens_per_second_per_gpu": 2450.81
},
{
"epoch": 1.6317155981625724,
"grad_norm": 3.28125,
"learning_rate": 1.6097809484288314e-05,
"loss": 0.3691,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16340,
"tokens_per_second_per_gpu": 2517.35
},
{
"epoch": 1.6327142001198323,
"grad_norm": 2.75,
"learning_rate": 1.6088415738664458e-05,
"loss": 0.3809,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16350,
"tokens_per_second_per_gpu": 2578.67
},
{
"epoch": 1.633712802077092,
"grad_norm": 3.828125,
"learning_rate": 1.607901978394588e-05,
"loss": 0.3481,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16360,
"tokens_per_second_per_gpu": 2320.32
},
{
"epoch": 1.634711404034352,
"grad_norm": 3.296875,
"learning_rate": 1.6069621625916897e-05,
"loss": 0.3385,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16370,
"tokens_per_second_per_gpu": 2256.75
},
{
"epoch": 1.6357100059916116,
"grad_norm": 2.890625,
"learning_rate": 1.60602212703632e-05,
"loss": 0.418,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16380,
"tokens_per_second_per_gpu": 2485.82
},
{
"epoch": 1.6367086079488717,
"grad_norm": 3.34375,
"learning_rate": 1.6050818723071808e-05,
"loss": 0.3232,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16390,
"tokens_per_second_per_gpu": 2303.64
},
{
"epoch": 1.6377072099061314,
"grad_norm": 2.796875,
"learning_rate": 1.6041413989831105e-05,
"loss": 0.351,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16400,
"tokens_per_second_per_gpu": 2562.14
},
{
"epoch": 1.6387058118633913,
"grad_norm": 3.703125,
"learning_rate": 1.603200707643082e-05,
"loss": 0.3725,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16410,
"tokens_per_second_per_gpu": 2253.22
},
{
"epoch": 1.6397044138206511,
"grad_norm": 3.578125,
"learning_rate": 1.6022597988662026e-05,
"loss": 0.3885,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16420,
"tokens_per_second_per_gpu": 2341.45
},
{
"epoch": 1.6407030157779108,
"grad_norm": 3.671875,
"learning_rate": 1.601318673231712e-05,
"loss": 0.3723,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16430,
"tokens_per_second_per_gpu": 2308.59
},
{
"epoch": 1.6417016177351709,
"grad_norm": 3.328125,
"learning_rate": 1.6003773313189853e-05,
"loss": 0.3951,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16440,
"tokens_per_second_per_gpu": 2468.38
},
{
"epoch": 1.6427002196924305,
"grad_norm": 2.765625,
"learning_rate": 1.59943577370753e-05,
"loss": 0.3699,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16450,
"tokens_per_second_per_gpu": 2404.55
},
{
"epoch": 1.6436988216496904,
"grad_norm": 2.734375,
"learning_rate": 1.5984940009769857e-05,
"loss": 0.4027,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16460,
"tokens_per_second_per_gpu": 2437.55
},
{
"epoch": 1.6446974236069503,
"grad_norm": 2.984375,
"learning_rate": 1.597552013707125e-05,
"loss": 0.3921,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16470,
"tokens_per_second_per_gpu": 2317.36
},
{
"epoch": 1.6456960255642101,
"grad_norm": 3.921875,
"learning_rate": 1.5966098124778528e-05,
"loss": 0.3389,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16480,
"tokens_per_second_per_gpu": 2460.72
},
{
"epoch": 1.64669462752147,
"grad_norm": 3.859375,
"learning_rate": 1.5956673978692054e-05,
"loss": 0.3789,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16490,
"tokens_per_second_per_gpu": 2402.74
},
{
"epoch": 1.6476932294787296,
"grad_norm": 2.765625,
"learning_rate": 1.5947247704613513e-05,
"loss": 0.3663,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16500,
"tokens_per_second_per_gpu": 2576.99
},
{
"epoch": 1.6486918314359897,
"grad_norm": 2.78125,
"learning_rate": 1.5937819308345885e-05,
"loss": 0.3895,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16510,
"tokens_per_second_per_gpu": 2552.62
},
{
"epoch": 1.6496904333932494,
"grad_norm": 3.875,
"learning_rate": 1.5928388795693462e-05,
"loss": 0.3904,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16520,
"tokens_per_second_per_gpu": 2218.15
},
{
"epoch": 1.6506890353505093,
"grad_norm": 2.90625,
"learning_rate": 1.5918956172461852e-05,
"loss": 0.3967,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16530,
"tokens_per_second_per_gpu": 2341.83
},
{
"epoch": 1.6516876373077691,
"grad_norm": 3.046875,
"learning_rate": 1.5909521444457935e-05,
"loss": 0.387,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16540,
"tokens_per_second_per_gpu": 2432.0
},
{
"epoch": 1.652686239265029,
"grad_norm": 4.0,
"learning_rate": 1.5900084617489915e-05,
"loss": 0.417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16550,
"tokens_per_second_per_gpu": 2387.53
},
{
"epoch": 1.6536848412222889,
"grad_norm": 4.125,
"learning_rate": 1.589064569736728e-05,
"loss": 0.3809,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16560,
"tokens_per_second_per_gpu": 2444.95
},
{
"epoch": 1.6546834431795485,
"grad_norm": 3.359375,
"learning_rate": 1.588120468990079e-05,
"loss": 0.3531,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16570,
"tokens_per_second_per_gpu": 2408.9
},
{
"epoch": 1.6556820451368086,
"grad_norm": 3.375,
"learning_rate": 1.587176160090251e-05,
"loss": 0.3421,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16580,
"tokens_per_second_per_gpu": 2414.67
},
{
"epoch": 1.6566806470940683,
"grad_norm": 3.46875,
"learning_rate": 1.5862316436185782e-05,
"loss": 0.4134,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16590,
"tokens_per_second_per_gpu": 2447.99
},
{
"epoch": 1.6576792490513281,
"grad_norm": 3.640625,
"learning_rate": 1.5852869201565212e-05,
"loss": 0.3659,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16600,
"tokens_per_second_per_gpu": 2281.18
},
{
"epoch": 1.658677851008588,
"grad_norm": 4.3125,
"learning_rate": 1.58434199028567e-05,
"loss": 0.4214,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16610,
"tokens_per_second_per_gpu": 2543.05
},
{
"epoch": 1.6596764529658476,
"grad_norm": 3.671875,
"learning_rate": 1.5833968545877414e-05,
"loss": 0.3908,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16620,
"tokens_per_second_per_gpu": 2384.86
},
{
"epoch": 1.6606750549231077,
"grad_norm": 4.375,
"learning_rate": 1.582451513644577e-05,
"loss": 0.3862,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16630,
"tokens_per_second_per_gpu": 2239.67
},
{
"epoch": 1.6616736568803674,
"grad_norm": 2.515625,
"learning_rate": 1.581505968038147e-05,
"loss": 0.3207,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16640,
"tokens_per_second_per_gpu": 2575.84
},
{
"epoch": 1.6626722588376275,
"grad_norm": 2.890625,
"learning_rate": 1.5805602183505465e-05,
"loss": 0.335,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16650,
"tokens_per_second_per_gpu": 2503.88
},
{
"epoch": 1.6636708607948871,
"grad_norm": 3.421875,
"learning_rate": 1.5796142651639957e-05,
"loss": 0.3879,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16660,
"tokens_per_second_per_gpu": 2404.77
},
{
"epoch": 1.664669462752147,
"grad_norm": 3.859375,
"learning_rate": 1.5786681090608416e-05,
"loss": 0.4161,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16670,
"tokens_per_second_per_gpu": 2511.88
},
{
"epoch": 1.6656680647094069,
"grad_norm": 2.546875,
"learning_rate": 1.5777217506235548e-05,
"loss": 0.3874,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16680,
"tokens_per_second_per_gpu": 2527.54
},
{
"epoch": 1.6666666666666665,
"grad_norm": 3.453125,
"learning_rate": 1.5767751904347317e-05,
"loss": 0.3629,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16690,
"tokens_per_second_per_gpu": 2237.65
},
{
"epoch": 1.6676652686239266,
"grad_norm": 3.015625,
"learning_rate": 1.5758284290770914e-05,
"loss": 0.364,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16700,
"tokens_per_second_per_gpu": 2209.74
},
{
"epoch": 1.6686638705811863,
"grad_norm": 2.3125,
"learning_rate": 1.5748814671334776e-05,
"loss": 0.3857,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16710,
"tokens_per_second_per_gpu": 2325.27
},
{
"epoch": 1.6696624725384461,
"grad_norm": 3.265625,
"learning_rate": 1.5739343051868575e-05,
"loss": 0.3689,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16720,
"tokens_per_second_per_gpu": 2341.77
},
{
"epoch": 1.670661074495706,
"grad_norm": 3.578125,
"learning_rate": 1.5729869438203224e-05,
"loss": 0.3742,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16730,
"tokens_per_second_per_gpu": 2231.08
},
{
"epoch": 1.6716596764529659,
"grad_norm": 3.90625,
"learning_rate": 1.572039383617084e-05,
"loss": 0.4107,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16740,
"tokens_per_second_per_gpu": 2485.55
},
{
"epoch": 1.6726582784102257,
"grad_norm": 3.21875,
"learning_rate": 1.5710916251604784e-05,
"loss": 0.4331,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16750,
"tokens_per_second_per_gpu": 2454.2
},
{
"epoch": 1.6736568803674854,
"grad_norm": 3.25,
"learning_rate": 1.5701436690339638e-05,
"loss": 0.452,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16760,
"tokens_per_second_per_gpu": 2433.6
},
{
"epoch": 1.6746554823247455,
"grad_norm": 3.546875,
"learning_rate": 1.5691955158211184e-05,
"loss": 0.3565,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16770,
"tokens_per_second_per_gpu": 2365.04
},
{
"epoch": 1.6756540842820051,
"grad_norm": 2.984375,
"learning_rate": 1.568247166105643e-05,
"loss": 0.349,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16780,
"tokens_per_second_per_gpu": 2299.09
},
{
"epoch": 1.676652686239265,
"grad_norm": 3.8125,
"learning_rate": 1.5672986204713598e-05,
"loss": 0.3882,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16790,
"tokens_per_second_per_gpu": 2208.28
},
{
"epoch": 1.6776512881965249,
"grad_norm": 5.34375,
"learning_rate": 1.5663498795022097e-05,
"loss": 0.4115,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16800,
"tokens_per_second_per_gpu": 2439.72
},
{
"epoch": 1.6786498901537847,
"grad_norm": 3.390625,
"learning_rate": 1.565400943782256e-05,
"loss": 0.3846,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16810,
"tokens_per_second_per_gpu": 2473.8
},
{
"epoch": 1.6796484921110446,
"grad_norm": 3.953125,
"learning_rate": 1.5644518138956807e-05,
"loss": 0.3985,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16820,
"tokens_per_second_per_gpu": 2551.87
},
{
"epoch": 1.6806470940683043,
"grad_norm": 2.890625,
"learning_rate": 1.563502490426786e-05,
"loss": 0.3768,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16830,
"tokens_per_second_per_gpu": 2427.68
},
{
"epoch": 1.6816456960255644,
"grad_norm": 3.5625,
"learning_rate": 1.562552973959992e-05,
"loss": 0.4268,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16840,
"tokens_per_second_per_gpu": 2554.16
},
{
"epoch": 1.682644297982824,
"grad_norm": 3.90625,
"learning_rate": 1.5616032650798397e-05,
"loss": 0.38,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16850,
"tokens_per_second_per_gpu": 2473.33
},
{
"epoch": 1.6836428999400839,
"grad_norm": 3.078125,
"learning_rate": 1.5606533643709865e-05,
"loss": 0.3558,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16860,
"tokens_per_second_per_gpu": 2437.43
},
{
"epoch": 1.6846415018973437,
"grad_norm": 3.265625,
"learning_rate": 1.5597032724182085e-05,
"loss": 0.3928,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16870,
"tokens_per_second_per_gpu": 2344.73
},
{
"epoch": 1.6856401038546034,
"grad_norm": 3.34375,
"learning_rate": 1.5587529898064008e-05,
"loss": 0.387,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16880,
"tokens_per_second_per_gpu": 2487.74
},
{
"epoch": 1.6866387058118635,
"grad_norm": 2.84375,
"learning_rate": 1.5578025171205742e-05,
"loss": 0.4158,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16890,
"tokens_per_second_per_gpu": 2609.74
},
{
"epoch": 1.6876373077691231,
"grad_norm": 2.765625,
"learning_rate": 1.556851854945857e-05,
"loss": 0.4049,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16900,
"tokens_per_second_per_gpu": 2541.82
},
{
"epoch": 1.6886359097263832,
"grad_norm": 2.9375,
"learning_rate": 1.5559010038674947e-05,
"loss": 0.3692,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16910,
"tokens_per_second_per_gpu": 2474.13
},
{
"epoch": 1.6896345116836429,
"grad_norm": 3.5625,
"learning_rate": 1.5549499644708487e-05,
"loss": 0.3821,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16920,
"tokens_per_second_per_gpu": 2335.96
},
{
"epoch": 1.6906331136409027,
"grad_norm": 4.15625,
"learning_rate": 1.5539987373413965e-05,
"loss": 0.413,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16930,
"tokens_per_second_per_gpu": 2518.36
},
{
"epoch": 1.6916317155981626,
"grad_norm": 3.453125,
"learning_rate": 1.55304732306473e-05,
"loss": 0.3526,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16940,
"tokens_per_second_per_gpu": 2397.02
},
{
"epoch": 1.6926303175554223,
"grad_norm": 3.375,
"learning_rate": 1.5520957222265587e-05,
"loss": 0.3766,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16950,
"tokens_per_second_per_gpu": 2398.92
},
{
"epoch": 1.6936289195126824,
"grad_norm": 3.5625,
"learning_rate": 1.551143935412705e-05,
"loss": 0.3504,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16960,
"tokens_per_second_per_gpu": 2427.84
},
{
"epoch": 1.694627521469942,
"grad_norm": 2.84375,
"learning_rate": 1.550191963209106e-05,
"loss": 0.4094,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16970,
"tokens_per_second_per_gpu": 2456.33
},
{
"epoch": 1.6956261234272019,
"grad_norm": 3.8125,
"learning_rate": 1.549239806201813e-05,
"loss": 0.4063,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16980,
"tokens_per_second_per_gpu": 2485.78
},
{
"epoch": 1.6966247253844617,
"grad_norm": 3.65625,
"learning_rate": 1.548287464976993e-05,
"loss": 0.3787,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 16990,
"tokens_per_second_per_gpu": 2438.8
},
{
"epoch": 1.6976233273417216,
"grad_norm": 3.875,
"learning_rate": 1.5473349401209235e-05,
"loss": 0.3975,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17000,
"tokens_per_second_per_gpu": 2492.42
},
{
"epoch": 1.6986219292989815,
"grad_norm": 3.34375,
"learning_rate": 1.5463822322199965e-05,
"loss": 0.346,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17010,
"tokens_per_second_per_gpu": 2476.92
},
{
"epoch": 1.6996205312562411,
"grad_norm": 3.875,
"learning_rate": 1.5454293418607165e-05,
"loss": 0.4391,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17020,
"tokens_per_second_per_gpu": 2477.57
},
{
"epoch": 1.7006191332135012,
"grad_norm": 3.78125,
"learning_rate": 1.5444762696297e-05,
"loss": 0.4231,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17030,
"tokens_per_second_per_gpu": 2519.85
},
{
"epoch": 1.7016177351707609,
"grad_norm": 2.609375,
"learning_rate": 1.543523016113677e-05,
"loss": 0.3371,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17040,
"tokens_per_second_per_gpu": 2523.54
},
{
"epoch": 1.7026163371280207,
"grad_norm": 3.71875,
"learning_rate": 1.5425695818994866e-05,
"loss": 0.3675,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17050,
"tokens_per_second_per_gpu": 2325.39
},
{
"epoch": 1.7036149390852806,
"grad_norm": 3.34375,
"learning_rate": 1.541615967574081e-05,
"loss": 0.3579,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17060,
"tokens_per_second_per_gpu": 2346.2
},
{
"epoch": 1.7046135410425405,
"grad_norm": 3.140625,
"learning_rate": 1.5406621737245226e-05,
"loss": 0.3883,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17070,
"tokens_per_second_per_gpu": 2495.58
},
{
"epoch": 1.7056121429998004,
"grad_norm": 2.625,
"learning_rate": 1.5397082009379846e-05,
"loss": 0.3456,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17080,
"tokens_per_second_per_gpu": 2488.74
},
{
"epoch": 1.70661074495706,
"grad_norm": 3.109375,
"learning_rate": 1.53875404980175e-05,
"loss": 0.3776,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17090,
"tokens_per_second_per_gpu": 2355.56
},
{
"epoch": 1.70760934691432,
"grad_norm": 3.296875,
"learning_rate": 1.5377997209032118e-05,
"loss": 0.3787,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17100,
"tokens_per_second_per_gpu": 2322.53
},
{
"epoch": 1.7086079488715797,
"grad_norm": 3.140625,
"learning_rate": 1.5368452148298727e-05,
"loss": 0.3667,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17110,
"tokens_per_second_per_gpu": 2528.23
},
{
"epoch": 1.7096065508288396,
"grad_norm": 3.328125,
"learning_rate": 1.5358905321693437e-05,
"loss": 0.4305,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17120,
"tokens_per_second_per_gpu": 2625.61
},
{
"epoch": 1.7106051527860995,
"grad_norm": 2.75,
"learning_rate": 1.5349356735093456e-05,
"loss": 0.3983,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17130,
"tokens_per_second_per_gpu": 2477.55
},
{
"epoch": 1.7116037547433594,
"grad_norm": 2.921875,
"learning_rate": 1.533980639437706e-05,
"loss": 0.4107,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17140,
"tokens_per_second_per_gpu": 2471.1
},
{
"epoch": 1.7126023567006192,
"grad_norm": 2.84375,
"learning_rate": 1.533025430542363e-05,
"loss": 0.3888,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17150,
"tokens_per_second_per_gpu": 2560.99
},
{
"epoch": 1.7136009586578789,
"grad_norm": 4.53125,
"learning_rate": 1.5320700474113594e-05,
"loss": 0.3857,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17160,
"tokens_per_second_per_gpu": 2364.04
},
{
"epoch": 1.714599560615139,
"grad_norm": 3.4375,
"learning_rate": 1.531114490632847e-05,
"loss": 0.3667,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17170,
"tokens_per_second_per_gpu": 2337.04
},
{
"epoch": 1.7155981625723986,
"grad_norm": 3.640625,
"learning_rate": 1.530158760795084e-05,
"loss": 0.3554,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17180,
"tokens_per_second_per_gpu": 2369.18
},
{
"epoch": 1.7165967645296585,
"grad_norm": 2.46875,
"learning_rate": 1.529202858486436e-05,
"loss": 0.3681,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17190,
"tokens_per_second_per_gpu": 2572.19
},
{
"epoch": 1.7175953664869184,
"grad_norm": 3.140625,
"learning_rate": 1.528246784295373e-05,
"loss": 0.444,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17200,
"tokens_per_second_per_gpu": 2465.08
},
{
"epoch": 1.718593968444178,
"grad_norm": 3.8125,
"learning_rate": 1.5272905388104724e-05,
"loss": 0.4044,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17210,
"tokens_per_second_per_gpu": 2517.62
},
{
"epoch": 1.719592570401438,
"grad_norm": 3.546875,
"learning_rate": 1.5263341226204166e-05,
"loss": 0.3342,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17220,
"tokens_per_second_per_gpu": 2536.95
},
{
"epoch": 1.7205911723586977,
"grad_norm": 3.34375,
"learning_rate": 1.5253775363139927e-05,
"loss": 0.4,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17230,
"tokens_per_second_per_gpu": 2238.12
},
{
"epoch": 1.7215897743159576,
"grad_norm": 3.265625,
"learning_rate": 1.5244207804800931e-05,
"loss": 0.3735,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17240,
"tokens_per_second_per_gpu": 2523.43
},
{
"epoch": 1.7225883762732175,
"grad_norm": 3.6875,
"learning_rate": 1.523463855707714e-05,
"loss": 0.3502,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17250,
"tokens_per_second_per_gpu": 2440.77
},
{
"epoch": 1.7235869782304774,
"grad_norm": 4.75,
"learning_rate": 1.5225067625859562e-05,
"loss": 0.3676,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17260,
"tokens_per_second_per_gpu": 2511.21
},
{
"epoch": 1.7245855801877372,
"grad_norm": 2.9375,
"learning_rate": 1.5215495017040238e-05,
"loss": 0.3521,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17270,
"tokens_per_second_per_gpu": 2504.83
},
{
"epoch": 1.7255841821449969,
"grad_norm": 3.21875,
"learning_rate": 1.5205920736512238e-05,
"loss": 0.3991,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17280,
"tokens_per_second_per_gpu": 2559.19
},
{
"epoch": 1.726582784102257,
"grad_norm": 4.03125,
"learning_rate": 1.519634479016967e-05,
"loss": 0.3947,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17290,
"tokens_per_second_per_gpu": 2285.31
},
{
"epoch": 1.7275813860595166,
"grad_norm": 3.609375,
"learning_rate": 1.5186767183907658e-05,
"loss": 0.3302,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17300,
"tokens_per_second_per_gpu": 2446.37
},
{
"epoch": 1.7285799880167765,
"grad_norm": 4.3125,
"learning_rate": 1.5177187923622358e-05,
"loss": 0.3661,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17310,
"tokens_per_second_per_gpu": 2478.51
},
{
"epoch": 1.7295785899740364,
"grad_norm": 2.40625,
"learning_rate": 1.5167607015210932e-05,
"loss": 0.3551,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17320,
"tokens_per_second_per_gpu": 2395.42
},
{
"epoch": 1.7305771919312962,
"grad_norm": 2.890625,
"learning_rate": 1.5158024464571575e-05,
"loss": 0.3406,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17330,
"tokens_per_second_per_gpu": 2562.07
},
{
"epoch": 1.731575793888556,
"grad_norm": 2.28125,
"learning_rate": 1.5148440277603468e-05,
"loss": 0.3416,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17340,
"tokens_per_second_per_gpu": 2371.24
},
{
"epoch": 1.7325743958458157,
"grad_norm": 3.734375,
"learning_rate": 1.513885446020682e-05,
"loss": 0.3918,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17350,
"tokens_per_second_per_gpu": 2588.57
},
{
"epoch": 1.7335729978030758,
"grad_norm": 3.53125,
"learning_rate": 1.512926701828283e-05,
"loss": 0.3675,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17360,
"tokens_per_second_per_gpu": 2232.15
},
{
"epoch": 1.7345715997603355,
"grad_norm": 3.34375,
"learning_rate": 1.5119677957733717e-05,
"loss": 0.3741,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17370,
"tokens_per_second_per_gpu": 2270.82
},
{
"epoch": 1.7355702017175954,
"grad_norm": 3.3125,
"learning_rate": 1.511008728446267e-05,
"loss": 0.3669,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17380,
"tokens_per_second_per_gpu": 2260.55
},
{
"epoch": 1.7365688036748552,
"grad_norm": 3.390625,
"learning_rate": 1.510049500437389e-05,
"loss": 0.385,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17390,
"tokens_per_second_per_gpu": 2572.04
},
{
"epoch": 1.737567405632115,
"grad_norm": 3.25,
"learning_rate": 1.509090112337256e-05,
"loss": 0.343,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17400,
"tokens_per_second_per_gpu": 2394.61
},
{
"epoch": 1.738566007589375,
"grad_norm": 2.90625,
"learning_rate": 1.5081305647364846e-05,
"loss": 0.3916,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17410,
"tokens_per_second_per_gpu": 2257.72
},
{
"epoch": 1.7395646095466346,
"grad_norm": 3.9375,
"learning_rate": 1.5071708582257907e-05,
"loss": 0.3529,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17420,
"tokens_per_second_per_gpu": 2346.79
},
{
"epoch": 1.7405632115038947,
"grad_norm": 5.25,
"learning_rate": 1.5062109933959865e-05,
"loss": 0.4323,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17430,
"tokens_per_second_per_gpu": 2243.06
},
{
"epoch": 1.7415618134611544,
"grad_norm": 3.84375,
"learning_rate": 1.5052509708379829e-05,
"loss": 0.4015,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17440,
"tokens_per_second_per_gpu": 2499.24
},
{
"epoch": 1.7425604154184142,
"grad_norm": 4.0625,
"learning_rate": 1.5042907911427872e-05,
"loss": 0.3521,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17450,
"tokens_per_second_per_gpu": 2310.95
},
{
"epoch": 1.743559017375674,
"grad_norm": 3.15625,
"learning_rate": 1.503330454901504e-05,
"loss": 0.3548,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17460,
"tokens_per_second_per_gpu": 2432.9
},
{
"epoch": 1.7445576193329337,
"grad_norm": 2.828125,
"learning_rate": 1.502369962705334e-05,
"loss": 0.3754,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17470,
"tokens_per_second_per_gpu": 2388.42
},
{
"epoch": 1.7455562212901938,
"grad_norm": 3.25,
"learning_rate": 1.5014093151455732e-05,
"loss": 0.3806,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17480,
"tokens_per_second_per_gpu": 2450.92
},
{
"epoch": 1.7465548232474535,
"grad_norm": 2.75,
"learning_rate": 1.5004485128136145e-05,
"loss": 0.3987,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17490,
"tokens_per_second_per_gpu": 2513.24
},
{
"epoch": 1.7475534252047134,
"grad_norm": 3.421875,
"learning_rate": 1.499487556300945e-05,
"loss": 0.3979,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17500,
"tokens_per_second_per_gpu": 2247.5
},
{
"epoch": 1.7485520271619732,
"grad_norm": 3.0,
"learning_rate": 1.4985264461991477e-05,
"loss": 0.3534,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17510,
"tokens_per_second_per_gpu": 2417.18
},
{
"epoch": 1.749550629119233,
"grad_norm": 3.265625,
"learning_rate": 1.4975651830998997e-05,
"loss": 0.3744,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17520,
"tokens_per_second_per_gpu": 2545.88
},
{
"epoch": 1.750549231076493,
"grad_norm": 3.234375,
"learning_rate": 1.4966037675949719e-05,
"loss": 0.3817,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17530,
"tokens_per_second_per_gpu": 2455.09
},
{
"epoch": 1.7515478330337526,
"grad_norm": 3.53125,
"learning_rate": 1.4956422002762293e-05,
"loss": 0.3545,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17540,
"tokens_per_second_per_gpu": 2532.7
},
{
"epoch": 1.7525464349910127,
"grad_norm": 2.828125,
"learning_rate": 1.4946804817356308e-05,
"loss": 0.3963,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17550,
"tokens_per_second_per_gpu": 2378.35
},
{
"epoch": 1.7535450369482724,
"grad_norm": 2.9375,
"learning_rate": 1.4937186125652274e-05,
"loss": 0.366,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17560,
"tokens_per_second_per_gpu": 2497.38
},
{
"epoch": 1.7545436389055322,
"grad_norm": 3.859375,
"learning_rate": 1.4927565933571644e-05,
"loss": 0.4229,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17570,
"tokens_per_second_per_gpu": 2307.21
},
{
"epoch": 1.755542240862792,
"grad_norm": 3.625,
"learning_rate": 1.4917944247036778e-05,
"loss": 0.4262,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17580,
"tokens_per_second_per_gpu": 2361.39
},
{
"epoch": 1.756540842820052,
"grad_norm": 3.90625,
"learning_rate": 1.4908321071970965e-05,
"loss": 0.4157,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17590,
"tokens_per_second_per_gpu": 2503.13
},
{
"epoch": 1.7575394447773118,
"grad_norm": 3.671875,
"learning_rate": 1.489869641429841e-05,
"loss": 0.4054,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17600,
"tokens_per_second_per_gpu": 2399.36
},
{
"epoch": 1.7585380467345715,
"grad_norm": 2.609375,
"learning_rate": 1.4889070279944222e-05,
"loss": 0.3832,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17610,
"tokens_per_second_per_gpu": 2549.11
},
{
"epoch": 1.7595366486918316,
"grad_norm": 3.53125,
"learning_rate": 1.487944267483444e-05,
"loss": 0.3869,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17620,
"tokens_per_second_per_gpu": 2497.77
},
{
"epoch": 1.7605352506490912,
"grad_norm": 3.171875,
"learning_rate": 1.4869813604895982e-05,
"loss": 0.3706,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17630,
"tokens_per_second_per_gpu": 2527.44
},
{
"epoch": 1.761533852606351,
"grad_norm": 3.875,
"learning_rate": 1.4860183076056686e-05,
"loss": 0.374,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17640,
"tokens_per_second_per_gpu": 2320.04
},
{
"epoch": 1.762532454563611,
"grad_norm": 2.5625,
"learning_rate": 1.4850551094245286e-05,
"loss": 0.3598,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17650,
"tokens_per_second_per_gpu": 2408.43
},
{
"epoch": 1.7635310565208708,
"grad_norm": 3.734375,
"learning_rate": 1.4840917665391401e-05,
"loss": 0.3814,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17660,
"tokens_per_second_per_gpu": 2513.46
},
{
"epoch": 1.7645296584781307,
"grad_norm": 4.09375,
"learning_rate": 1.4831282795425546e-05,
"loss": 0.4051,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17670,
"tokens_per_second_per_gpu": 2432.93
},
{
"epoch": 1.7655282604353904,
"grad_norm": 3.296875,
"learning_rate": 1.482164649027913e-05,
"loss": 0.374,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17680,
"tokens_per_second_per_gpu": 2384.39
},
{
"epoch": 1.7665268623926504,
"grad_norm": 3.65625,
"learning_rate": 1.4812008755884438e-05,
"loss": 0.4091,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17690,
"tokens_per_second_per_gpu": 2444.08
},
{
"epoch": 1.76752546434991,
"grad_norm": 2.984375,
"learning_rate": 1.4802369598174634e-05,
"loss": 0.3446,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17700,
"tokens_per_second_per_gpu": 2530.95
},
{
"epoch": 1.76852406630717,
"grad_norm": 3.421875,
"learning_rate": 1.4792729023083765e-05,
"loss": 0.3543,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17710,
"tokens_per_second_per_gpu": 2341.05
},
{
"epoch": 1.7695226682644298,
"grad_norm": 4.03125,
"learning_rate": 1.4783087036546744e-05,
"loss": 0.3558,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17720,
"tokens_per_second_per_gpu": 2275.33
},
{
"epoch": 1.7705212702216895,
"grad_norm": 4.625,
"learning_rate": 1.4773443644499352e-05,
"loss": 0.3676,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17730,
"tokens_per_second_per_gpu": 2361.99
},
{
"epoch": 1.7715198721789496,
"grad_norm": 3.5,
"learning_rate": 1.4763798852878244e-05,
"loss": 0.3568,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17740,
"tokens_per_second_per_gpu": 2428.39
},
{
"epoch": 1.7725184741362092,
"grad_norm": 3.203125,
"learning_rate": 1.4754152667620927e-05,
"loss": 0.3301,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17750,
"tokens_per_second_per_gpu": 2330.25
},
{
"epoch": 1.773517076093469,
"grad_norm": 2.765625,
"learning_rate": 1.474450509466577e-05,
"loss": 0.3201,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17760,
"tokens_per_second_per_gpu": 2475.08
},
{
"epoch": 1.774515678050729,
"grad_norm": 3.90625,
"learning_rate": 1.4734856139952003e-05,
"loss": 0.4158,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17770,
"tokens_per_second_per_gpu": 2371.54
},
{
"epoch": 1.7755142800079888,
"grad_norm": 2.9375,
"learning_rate": 1.472520580941969e-05,
"loss": 0.3417,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17780,
"tokens_per_second_per_gpu": 2321.41
},
{
"epoch": 1.7765128819652487,
"grad_norm": 4.40625,
"learning_rate": 1.471555410900976e-05,
"loss": 0.4249,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17790,
"tokens_per_second_per_gpu": 2539.29
},
{
"epoch": 1.7775114839225084,
"grad_norm": 2.984375,
"learning_rate": 1.4705901044663971e-05,
"loss": 0.3861,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17800,
"tokens_per_second_per_gpu": 2369.51
},
{
"epoch": 1.7785100858797684,
"grad_norm": 3.015625,
"learning_rate": 1.4696246622324933e-05,
"loss": 0.3869,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17810,
"tokens_per_second_per_gpu": 2529.2
},
{
"epoch": 1.779508687837028,
"grad_norm": 3.9375,
"learning_rate": 1.4686590847936083e-05,
"loss": 0.3983,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17820,
"tokens_per_second_per_gpu": 2290.19
},
{
"epoch": 1.780507289794288,
"grad_norm": 4.4375,
"learning_rate": 1.4676933727441694e-05,
"loss": 0.4587,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17830,
"tokens_per_second_per_gpu": 2452.19
},
{
"epoch": 1.7815058917515478,
"grad_norm": 4.46875,
"learning_rate": 1.4667275266786861e-05,
"loss": 0.3588,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17840,
"tokens_per_second_per_gpu": 2594.42
},
{
"epoch": 1.7825044937088077,
"grad_norm": 3.0625,
"learning_rate": 1.4657615471917521e-05,
"loss": 0.3946,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17850,
"tokens_per_second_per_gpu": 2480.87
},
{
"epoch": 1.7835030956660676,
"grad_norm": 3.3125,
"learning_rate": 1.4647954348780416e-05,
"loss": 0.3946,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17860,
"tokens_per_second_per_gpu": 2316.45
},
{
"epoch": 1.7845016976233272,
"grad_norm": 4.59375,
"learning_rate": 1.4638291903323108e-05,
"loss": 0.3916,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17870,
"tokens_per_second_per_gpu": 2470.74
},
{
"epoch": 1.7855002995805873,
"grad_norm": 3.296875,
"learning_rate": 1.462862814149398e-05,
"loss": 0.357,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17880,
"tokens_per_second_per_gpu": 2419.85
},
{
"epoch": 1.786498901537847,
"grad_norm": 3.046875,
"learning_rate": 1.461896306924222e-05,
"loss": 0.3069,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17890,
"tokens_per_second_per_gpu": 2447.47
},
{
"epoch": 1.7874975034951068,
"grad_norm": 3.40625,
"learning_rate": 1.4609296692517824e-05,
"loss": 0.4066,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17900,
"tokens_per_second_per_gpu": 2222.38
},
{
"epoch": 1.7884961054523667,
"grad_norm": 2.71875,
"learning_rate": 1.4599629017271594e-05,
"loss": 0.3298,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17910,
"tokens_per_second_per_gpu": 2479.15
},
{
"epoch": 1.7894947074096266,
"grad_norm": 3.53125,
"learning_rate": 1.4589960049455126e-05,
"loss": 0.3131,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17920,
"tokens_per_second_per_gpu": 2279.97
},
{
"epoch": 1.7904933093668864,
"grad_norm": 2.875,
"learning_rate": 1.4580289795020816e-05,
"loss": 0.3719,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17930,
"tokens_per_second_per_gpu": 2490.91
},
{
"epoch": 1.791491911324146,
"grad_norm": 2.859375,
"learning_rate": 1.4570618259921848e-05,
"loss": 0.3874,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17940,
"tokens_per_second_per_gpu": 2260.57
},
{
"epoch": 1.7924905132814062,
"grad_norm": 3.90625,
"learning_rate": 1.4560945450112198e-05,
"loss": 0.34,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17950,
"tokens_per_second_per_gpu": 2529.84
},
{
"epoch": 1.7934891152386658,
"grad_norm": 2.75,
"learning_rate": 1.4551271371546623e-05,
"loss": 0.349,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17960,
"tokens_per_second_per_gpu": 2510.2
},
{
"epoch": 1.7944877171959257,
"grad_norm": 3.203125,
"learning_rate": 1.4541596030180674e-05,
"loss": 0.3401,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17970,
"tokens_per_second_per_gpu": 2350.56
},
{
"epoch": 1.7954863191531856,
"grad_norm": 3.953125,
"learning_rate": 1.4531919431970658e-05,
"loss": 0.3871,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17980,
"tokens_per_second_per_gpu": 2462.27
},
{
"epoch": 1.7964849211104452,
"grad_norm": 3.5625,
"learning_rate": 1.452224158287367e-05,
"loss": 0.3622,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 17990,
"tokens_per_second_per_gpu": 2383.49
},
{
"epoch": 1.7974835230677053,
"grad_norm": 3.25,
"learning_rate": 1.4512562488847576e-05,
"loss": 0.3328,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18000,
"tokens_per_second_per_gpu": 2332.59
},
{
"epoch": 1.798482125024965,
"grad_norm": 3.15625,
"learning_rate": 1.4502882155850995e-05,
"loss": 0.3452,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18010,
"tokens_per_second_per_gpu": 2597.9
},
{
"epoch": 1.7994807269822248,
"grad_norm": 3.8125,
"learning_rate": 1.4493200589843326e-05,
"loss": 0.342,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18020,
"tokens_per_second_per_gpu": 2358.19
},
{
"epoch": 1.8004793289394847,
"grad_norm": 3.453125,
"learning_rate": 1.4483517796784718e-05,
"loss": 0.3729,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18030,
"tokens_per_second_per_gpu": 2367.87
},
{
"epoch": 1.8014779308967446,
"grad_norm": 3.96875,
"learning_rate": 1.4473833782636075e-05,
"loss": 0.3753,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18040,
"tokens_per_second_per_gpu": 2429.64
},
{
"epoch": 1.8024765328540044,
"grad_norm": 3.171875,
"learning_rate": 1.4464148553359056e-05,
"loss": 0.3823,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18050,
"tokens_per_second_per_gpu": 2368.88
},
{
"epoch": 1.803475134811264,
"grad_norm": 3.53125,
"learning_rate": 1.4454462114916065e-05,
"loss": 0.3652,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18060,
"tokens_per_second_per_gpu": 2438.33
},
{
"epoch": 1.8044737367685242,
"grad_norm": 3.53125,
"learning_rate": 1.4444774473270249e-05,
"loss": 0.4193,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18070,
"tokens_per_second_per_gpu": 2498.69
},
{
"epoch": 1.8054723387257838,
"grad_norm": 3.5,
"learning_rate": 1.4435085634385503e-05,
"loss": 0.3654,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18080,
"tokens_per_second_per_gpu": 2511.51
},
{
"epoch": 1.8064709406830437,
"grad_norm": 3.8125,
"learning_rate": 1.4425395604226449e-05,
"loss": 0.3785,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18090,
"tokens_per_second_per_gpu": 2523.75
},
{
"epoch": 1.8074695426403036,
"grad_norm": 3.015625,
"learning_rate": 1.4415704388758452e-05,
"loss": 0.3805,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18100,
"tokens_per_second_per_gpu": 2338.77
},
{
"epoch": 1.8084681445975634,
"grad_norm": 4.0,
"learning_rate": 1.4406011993947608e-05,
"loss": 0.4201,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18110,
"tokens_per_second_per_gpu": 2290.16
},
{
"epoch": 1.8094667465548233,
"grad_norm": 3.59375,
"learning_rate": 1.4396318425760719e-05,
"loss": 0.3803,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18120,
"tokens_per_second_per_gpu": 2329.67
},
{
"epoch": 1.810465348512083,
"grad_norm": 3.4375,
"learning_rate": 1.4386623690165332e-05,
"loss": 0.4021,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18130,
"tokens_per_second_per_gpu": 2367.53
},
{
"epoch": 1.811463950469343,
"grad_norm": 3.0,
"learning_rate": 1.437692779312971e-05,
"loss": 0.3809,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18140,
"tokens_per_second_per_gpu": 2415.91
},
{
"epoch": 1.8124625524266027,
"grad_norm": 4.21875,
"learning_rate": 1.4367230740622814e-05,
"loss": 0.3717,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18150,
"tokens_per_second_per_gpu": 2358.52
},
{
"epoch": 1.8134611543838626,
"grad_norm": 4.09375,
"learning_rate": 1.4357532538614335e-05,
"loss": 0.3621,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18160,
"tokens_per_second_per_gpu": 2365.65
},
{
"epoch": 1.8144597563411224,
"grad_norm": 3.59375,
"learning_rate": 1.4347833193074667e-05,
"loss": 0.403,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18170,
"tokens_per_second_per_gpu": 2396.34
},
{
"epoch": 1.8154583582983823,
"grad_norm": 2.953125,
"learning_rate": 1.4338132709974902e-05,
"loss": 0.3849,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18180,
"tokens_per_second_per_gpu": 2132.46
},
{
"epoch": 1.8164569602556422,
"grad_norm": 3.265625,
"learning_rate": 1.4328431095286832e-05,
"loss": 0.3841,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18190,
"tokens_per_second_per_gpu": 2537.36
},
{
"epoch": 1.8174555622129018,
"grad_norm": 2.671875,
"learning_rate": 1.4318728354982955e-05,
"loss": 0.3245,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18200,
"tokens_per_second_per_gpu": 2350.41
},
{
"epoch": 1.818454164170162,
"grad_norm": 3.796875,
"learning_rate": 1.4309024495036451e-05,
"loss": 0.3875,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18210,
"tokens_per_second_per_gpu": 2326.13
},
{
"epoch": 1.8194527661274216,
"grad_norm": 3.640625,
"learning_rate": 1.4299319521421201e-05,
"loss": 0.4727,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18220,
"tokens_per_second_per_gpu": 2320.63
},
{
"epoch": 1.8204513680846814,
"grad_norm": 3.953125,
"learning_rate": 1.4289613440111765e-05,
"loss": 0.3595,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18230,
"tokens_per_second_per_gpu": 2557.69
},
{
"epoch": 1.8214499700419413,
"grad_norm": 2.96875,
"learning_rate": 1.4279906257083381e-05,
"loss": 0.3662,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18240,
"tokens_per_second_per_gpu": 2515.88
},
{
"epoch": 1.822448571999201,
"grad_norm": 2.96875,
"learning_rate": 1.4270197978311966e-05,
"loss": 0.332,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18250,
"tokens_per_second_per_gpu": 2506.51
},
{
"epoch": 1.823447173956461,
"grad_norm": 3.03125,
"learning_rate": 1.4260488609774125e-05,
"loss": 0.4097,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18260,
"tokens_per_second_per_gpu": 2447.56
},
{
"epoch": 1.8244457759137207,
"grad_norm": 2.78125,
"learning_rate": 1.4250778157447117e-05,
"loss": 0.3151,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18270,
"tokens_per_second_per_gpu": 2527.94
},
{
"epoch": 1.8254443778709806,
"grad_norm": 3.703125,
"learning_rate": 1.4241066627308874e-05,
"loss": 0.382,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18280,
"tokens_per_second_per_gpu": 2522.73
},
{
"epoch": 1.8264429798282404,
"grad_norm": 3.359375,
"learning_rate": 1.4231354025338001e-05,
"loss": 0.3839,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18290,
"tokens_per_second_per_gpu": 2499.12
},
{
"epoch": 1.8274415817855003,
"grad_norm": 3.359375,
"learning_rate": 1.4221640357513746e-05,
"loss": 0.3357,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18300,
"tokens_per_second_per_gpu": 2440.05
},
{
"epoch": 1.8284401837427602,
"grad_norm": 3.921875,
"learning_rate": 1.4211925629816026e-05,
"loss": 0.3706,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18310,
"tokens_per_second_per_gpu": 2342.95
},
{
"epoch": 1.8294387857000198,
"grad_norm": 2.28125,
"learning_rate": 1.4202209848225397e-05,
"loss": 0.3789,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18320,
"tokens_per_second_per_gpu": 2541.65
},
{
"epoch": 1.83043738765728,
"grad_norm": 4.3125,
"learning_rate": 1.4192493018723085e-05,
"loss": 0.3933,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18330,
"tokens_per_second_per_gpu": 2246.95
},
{
"epoch": 1.8314359896145396,
"grad_norm": 2.828125,
"learning_rate": 1.4182775147290945e-05,
"loss": 0.4187,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18340,
"tokens_per_second_per_gpu": 2467.48
},
{
"epoch": 1.8324345915717994,
"grad_norm": 3.046875,
"learning_rate": 1.4173056239911475e-05,
"loss": 0.3877,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18350,
"tokens_per_second_per_gpu": 2403.09
},
{
"epoch": 1.8334331935290593,
"grad_norm": 2.765625,
"learning_rate": 1.4163336302567815e-05,
"loss": 0.3733,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18360,
"tokens_per_second_per_gpu": 2448.16
},
{
"epoch": 1.8344317954863192,
"grad_norm": 3.53125,
"learning_rate": 1.4153615341243739e-05,
"loss": 0.3858,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18370,
"tokens_per_second_per_gpu": 2405.79
},
{
"epoch": 1.835430397443579,
"grad_norm": 3.796875,
"learning_rate": 1.4143893361923644e-05,
"loss": 0.3944,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18380,
"tokens_per_second_per_gpu": 2325.81
},
{
"epoch": 1.8364289994008387,
"grad_norm": 3.71875,
"learning_rate": 1.4134170370592561e-05,
"loss": 0.376,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18390,
"tokens_per_second_per_gpu": 2470.26
},
{
"epoch": 1.8374276013580988,
"grad_norm": 5.3125,
"learning_rate": 1.4124446373236146e-05,
"loss": 0.3105,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18400,
"tokens_per_second_per_gpu": 2416.3
},
{
"epoch": 1.8384262033153584,
"grad_norm": 4.40625,
"learning_rate": 1.4114721375840667e-05,
"loss": 0.3843,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18410,
"tokens_per_second_per_gpu": 2427.52
},
{
"epoch": 1.8394248052726183,
"grad_norm": 2.8125,
"learning_rate": 1.4104995384393015e-05,
"loss": 0.4416,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18420,
"tokens_per_second_per_gpu": 2430.74
},
{
"epoch": 1.8404234072298782,
"grad_norm": 3.8125,
"learning_rate": 1.4095268404880688e-05,
"loss": 0.3419,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18430,
"tokens_per_second_per_gpu": 2346.33
},
{
"epoch": 1.841422009187138,
"grad_norm": 3.90625,
"learning_rate": 1.408554044329179e-05,
"loss": 0.3963,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18440,
"tokens_per_second_per_gpu": 2381.47
},
{
"epoch": 1.842420611144398,
"grad_norm": 3.5625,
"learning_rate": 1.4075811505615034e-05,
"loss": 0.3895,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18450,
"tokens_per_second_per_gpu": 2405.01
},
{
"epoch": 1.8434192131016576,
"grad_norm": 3.96875,
"learning_rate": 1.4066081597839736e-05,
"loss": 0.3865,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18460,
"tokens_per_second_per_gpu": 2378.02
},
{
"epoch": 1.8444178150589177,
"grad_norm": 3.421875,
"learning_rate": 1.4056350725955806e-05,
"loss": 0.3532,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18470,
"tokens_per_second_per_gpu": 2323.24
},
{
"epoch": 1.8454164170161773,
"grad_norm": 2.625,
"learning_rate": 1.4046618895953741e-05,
"loss": 0.4008,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18480,
"tokens_per_second_per_gpu": 2335.23
},
{
"epoch": 1.8464150189734372,
"grad_norm": 2.796875,
"learning_rate": 1.4036886113824646e-05,
"loss": 0.3998,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18490,
"tokens_per_second_per_gpu": 2477.95
},
{
"epoch": 1.847413620930697,
"grad_norm": 2.90625,
"learning_rate": 1.4027152385560194e-05,
"loss": 0.3606,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18500,
"tokens_per_second_per_gpu": 2326.16
},
{
"epoch": 1.8484122228879567,
"grad_norm": 3.34375,
"learning_rate": 1.4017417717152643e-05,
"loss": 0.3799,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18510,
"tokens_per_second_per_gpu": 2594.54
},
{
"epoch": 1.8494108248452168,
"grad_norm": 3.921875,
"learning_rate": 1.4007682114594842e-05,
"loss": 0.3457,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18520,
"tokens_per_second_per_gpu": 2381.53
},
{
"epoch": 1.8504094268024764,
"grad_norm": 4.03125,
"learning_rate": 1.39979455838802e-05,
"loss": 0.4058,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18530,
"tokens_per_second_per_gpu": 2414.51
},
{
"epoch": 1.8514080287597365,
"grad_norm": 4.25,
"learning_rate": 1.3988208131002715e-05,
"loss": 0.4287,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18540,
"tokens_per_second_per_gpu": 2376.19
},
{
"epoch": 1.8524066307169962,
"grad_norm": 3.84375,
"learning_rate": 1.3978469761956928e-05,
"loss": 0.4282,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18550,
"tokens_per_second_per_gpu": 2610.1
},
{
"epoch": 1.853405232674256,
"grad_norm": 4.03125,
"learning_rate": 1.3968730482737969e-05,
"loss": 0.4451,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18560,
"tokens_per_second_per_gpu": 2470.82
},
{
"epoch": 1.854403834631516,
"grad_norm": 5.03125,
"learning_rate": 1.3958990299341512e-05,
"loss": 0.4628,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18570,
"tokens_per_second_per_gpu": 2370.79
},
{
"epoch": 1.8554024365887756,
"grad_norm": 4.125,
"learning_rate": 1.3949249217763791e-05,
"loss": 0.4089,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18580,
"tokens_per_second_per_gpu": 2400.06
},
{
"epoch": 1.8564010385460357,
"grad_norm": 2.8125,
"learning_rate": 1.3939507244001601e-05,
"loss": 0.3699,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18590,
"tokens_per_second_per_gpu": 2596.1
},
{
"epoch": 1.8573996405032953,
"grad_norm": 2.984375,
"learning_rate": 1.392976438405228e-05,
"loss": 0.3453,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18600,
"tokens_per_second_per_gpu": 2501.55
},
{
"epoch": 1.8583982424605552,
"grad_norm": 3.84375,
"learning_rate": 1.3920020643913706e-05,
"loss": 0.3694,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18610,
"tokens_per_second_per_gpu": 2566.21
},
{
"epoch": 1.859396844417815,
"grad_norm": 3.3125,
"learning_rate": 1.391027602958431e-05,
"loss": 0.4045,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18620,
"tokens_per_second_per_gpu": 2290.04
},
{
"epoch": 1.860395446375075,
"grad_norm": 3.015625,
"learning_rate": 1.3900530547063053e-05,
"loss": 0.3742,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18630,
"tokens_per_second_per_gpu": 2438.26
},
{
"epoch": 1.8613940483323348,
"grad_norm": 3.1875,
"learning_rate": 1.3890784202349433e-05,
"loss": 0.347,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18640,
"tokens_per_second_per_gpu": 2451.16
},
{
"epoch": 1.8623926502895944,
"grad_norm": 3.28125,
"learning_rate": 1.388103700144348e-05,
"loss": 0.3475,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18650,
"tokens_per_second_per_gpu": 2530.67
},
{
"epoch": 1.8633912522468545,
"grad_norm": 3.28125,
"learning_rate": 1.3871288950345752e-05,
"loss": 0.3698,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18660,
"tokens_per_second_per_gpu": 2373.33
},
{
"epoch": 1.8643898542041142,
"grad_norm": 3.28125,
"learning_rate": 1.3861540055057327e-05,
"loss": 0.3856,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18670,
"tokens_per_second_per_gpu": 2185.21
},
{
"epoch": 1.865388456161374,
"grad_norm": 2.640625,
"learning_rate": 1.3851790321579802e-05,
"loss": 0.3471,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18680,
"tokens_per_second_per_gpu": 2376.95
},
{
"epoch": 1.866387058118634,
"grad_norm": 3.71875,
"learning_rate": 1.3842039755915298e-05,
"loss": 0.3425,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18690,
"tokens_per_second_per_gpu": 2400.46
},
{
"epoch": 1.8673856600758938,
"grad_norm": 3.046875,
"learning_rate": 1.3832288364066435e-05,
"loss": 0.3547,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18700,
"tokens_per_second_per_gpu": 2309.22
},
{
"epoch": 1.8683842620331537,
"grad_norm": 3.171875,
"learning_rate": 1.382253615203635e-05,
"loss": 0.3832,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18710,
"tokens_per_second_per_gpu": 2387.77
},
{
"epoch": 1.8693828639904133,
"grad_norm": 3.0,
"learning_rate": 1.3812783125828694e-05,
"loss": 0.3339,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18720,
"tokens_per_second_per_gpu": 2433.82
},
{
"epoch": 1.8703814659476734,
"grad_norm": 3.125,
"learning_rate": 1.3803029291447597e-05,
"loss": 0.3468,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18730,
"tokens_per_second_per_gpu": 2242.52
},
{
"epoch": 1.871380067904933,
"grad_norm": 4.1875,
"learning_rate": 1.3793274654897709e-05,
"loss": 0.4161,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18740,
"tokens_per_second_per_gpu": 2291.83
},
{
"epoch": 1.872378669862193,
"grad_norm": 3.515625,
"learning_rate": 1.3783519222184157e-05,
"loss": 0.3695,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18750,
"tokens_per_second_per_gpu": 2328.45
},
{
"epoch": 1.8733772718194528,
"grad_norm": 3.671875,
"learning_rate": 1.3773762999312564e-05,
"loss": 0.4081,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18760,
"tokens_per_second_per_gpu": 2417.5
},
{
"epoch": 1.8743758737767124,
"grad_norm": 2.796875,
"learning_rate": 1.3764005992289048e-05,
"loss": 0.4027,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18770,
"tokens_per_second_per_gpu": 2512.9
},
{
"epoch": 1.8753744757339725,
"grad_norm": 3.015625,
"learning_rate": 1.375424820712019e-05,
"loss": 0.3513,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18780,
"tokens_per_second_per_gpu": 2450.84
},
{
"epoch": 1.8763730776912322,
"grad_norm": 3.21875,
"learning_rate": 1.3744489649813072e-05,
"loss": 0.3814,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18790,
"tokens_per_second_per_gpu": 2294.03
},
{
"epoch": 1.8773716796484923,
"grad_norm": 3.171875,
"learning_rate": 1.373473032637524e-05,
"loss": 0.3453,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18800,
"tokens_per_second_per_gpu": 2369.03
},
{
"epoch": 1.878370281605752,
"grad_norm": 3.421875,
"learning_rate": 1.3724970242814708e-05,
"loss": 0.3339,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18810,
"tokens_per_second_per_gpu": 2390.22
},
{
"epoch": 1.8793688835630118,
"grad_norm": 3.96875,
"learning_rate": 1.3715209405139967e-05,
"loss": 0.3837,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18820,
"tokens_per_second_per_gpu": 2413.09
},
{
"epoch": 1.8803674855202717,
"grad_norm": 3.265625,
"learning_rate": 1.370544781935997e-05,
"loss": 0.3868,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18830,
"tokens_per_second_per_gpu": 1625.31
},
{
"epoch": 1.8813660874775313,
"grad_norm": 3.5,
"learning_rate": 1.3695685491484123e-05,
"loss": 0.3723,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18840,
"tokens_per_second_per_gpu": 2586.9
},
{
"epoch": 1.8823646894347914,
"grad_norm": 4.625,
"learning_rate": 1.3685922427522296e-05,
"loss": 0.3413,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18850,
"tokens_per_second_per_gpu": 2267.55
},
{
"epoch": 1.883363291392051,
"grad_norm": 2.8125,
"learning_rate": 1.3676158633484818e-05,
"loss": 0.3624,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18860,
"tokens_per_second_per_gpu": 2389.15
},
{
"epoch": 1.884361893349311,
"grad_norm": 2.9375,
"learning_rate": 1.3666394115382447e-05,
"loss": 0.3596,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18870,
"tokens_per_second_per_gpu": 2526.62
},
{
"epoch": 1.8853604953065708,
"grad_norm": 3.359375,
"learning_rate": 1.3656628879226414e-05,
"loss": 0.3614,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18880,
"tokens_per_second_per_gpu": 2243.91
},
{
"epoch": 1.8863590972638307,
"grad_norm": 3.046875,
"learning_rate": 1.364686293102837e-05,
"loss": 0.3308,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18890,
"tokens_per_second_per_gpu": 2574.83
},
{
"epoch": 1.8873576992210905,
"grad_norm": 3.015625,
"learning_rate": 1.3637096276800412e-05,
"loss": 0.3971,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18900,
"tokens_per_second_per_gpu": 2406.56
},
{
"epoch": 1.8883563011783502,
"grad_norm": 3.328125,
"learning_rate": 1.3627328922555071e-05,
"loss": 0.3616,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18910,
"tokens_per_second_per_gpu": 2344.5
},
{
"epoch": 1.8893549031356103,
"grad_norm": 3.578125,
"learning_rate": 1.3617560874305318e-05,
"loss": 0.3575,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18920,
"tokens_per_second_per_gpu": 2432.67
},
{
"epoch": 1.89035350509287,
"grad_norm": 4.0,
"learning_rate": 1.3607792138064534e-05,
"loss": 0.3932,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18930,
"tokens_per_second_per_gpu": 2302.79
},
{
"epoch": 1.8913521070501298,
"grad_norm": 3.265625,
"learning_rate": 1.359802271984654e-05,
"loss": 0.3864,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18940,
"tokens_per_second_per_gpu": 2463.36
},
{
"epoch": 1.8923507090073897,
"grad_norm": 3.125,
"learning_rate": 1.3588252625665566e-05,
"loss": 0.3647,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18950,
"tokens_per_second_per_gpu": 2320.07
},
{
"epoch": 1.8933493109646495,
"grad_norm": 3.34375,
"learning_rate": 1.3578481861536257e-05,
"loss": 0.4094,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18960,
"tokens_per_second_per_gpu": 2423.33
},
{
"epoch": 1.8943479129219094,
"grad_norm": 2.71875,
"learning_rate": 1.356871043347368e-05,
"loss": 0.3978,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18970,
"tokens_per_second_per_gpu": 2531.66
},
{
"epoch": 1.895346514879169,
"grad_norm": 4.21875,
"learning_rate": 1.3558938347493307e-05,
"loss": 0.3547,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18980,
"tokens_per_second_per_gpu": 2391.98
},
{
"epoch": 1.8963451168364291,
"grad_norm": 2.578125,
"learning_rate": 1.3549165609611015e-05,
"loss": 0.4097,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 18990,
"tokens_per_second_per_gpu": 2495.39
},
{
"epoch": 1.8973437187936888,
"grad_norm": 4.0625,
"learning_rate": 1.3539392225843076e-05,
"loss": 0.4157,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19000,
"tokens_per_second_per_gpu": 2498.32
},
{
"epoch": 1.8983423207509487,
"grad_norm": 3.25,
"learning_rate": 1.3529618202206168e-05,
"loss": 0.3746,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19010,
"tokens_per_second_per_gpu": 2223.12
},
{
"epoch": 1.8993409227082085,
"grad_norm": 2.78125,
"learning_rate": 1.3519843544717359e-05,
"loss": 0.4007,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19020,
"tokens_per_second_per_gpu": 2313.29
},
{
"epoch": 1.9003395246654682,
"grad_norm": 2.78125,
"learning_rate": 1.3510068259394106e-05,
"loss": 0.3333,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19030,
"tokens_per_second_per_gpu": 2258.49
},
{
"epoch": 1.9013381266227283,
"grad_norm": 2.796875,
"learning_rate": 1.3500292352254257e-05,
"loss": 0.3844,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19040,
"tokens_per_second_per_gpu": 2426.18
},
{
"epoch": 1.902336728579988,
"grad_norm": 3.09375,
"learning_rate": 1.3490515829316039e-05,
"loss": 0.3405,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19050,
"tokens_per_second_per_gpu": 2349.57
},
{
"epoch": 1.903335330537248,
"grad_norm": 4.0,
"learning_rate": 1.3480738696598063e-05,
"loss": 0.4286,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19060,
"tokens_per_second_per_gpu": 2550.55
},
{
"epoch": 1.9043339324945077,
"grad_norm": 2.96875,
"learning_rate": 1.3470960960119309e-05,
"loss": 0.3374,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19070,
"tokens_per_second_per_gpu": 2378.4
},
{
"epoch": 1.9053325344517675,
"grad_norm": 3.171875,
"learning_rate": 1.3461182625899126e-05,
"loss": 0.3646,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19080,
"tokens_per_second_per_gpu": 2422.24
},
{
"epoch": 1.9063311364090274,
"grad_norm": 3.375,
"learning_rate": 1.3451403699957246e-05,
"loss": 0.3786,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19090,
"tokens_per_second_per_gpu": 2610.91
},
{
"epoch": 1.907329738366287,
"grad_norm": 3.359375,
"learning_rate": 1.3441624188313748e-05,
"loss": 0.3393,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19100,
"tokens_per_second_per_gpu": 2341.12
},
{
"epoch": 1.9083283403235471,
"grad_norm": 3.015625,
"learning_rate": 1.3431844096989082e-05,
"loss": 0.3674,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19110,
"tokens_per_second_per_gpu": 2242.32
},
{
"epoch": 1.9093269422808068,
"grad_norm": 3.25,
"learning_rate": 1.3422063432004056e-05,
"loss": 0.3236,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19120,
"tokens_per_second_per_gpu": 2483.45
},
{
"epoch": 1.9103255442380667,
"grad_norm": 2.890625,
"learning_rate": 1.3412282199379819e-05,
"loss": 0.3704,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19130,
"tokens_per_second_per_gpu": 2394.48
},
{
"epoch": 1.9113241461953265,
"grad_norm": 3.296875,
"learning_rate": 1.3402500405137886e-05,
"loss": 0.4118,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19140,
"tokens_per_second_per_gpu": 2317.7
},
{
"epoch": 1.9123227481525864,
"grad_norm": 2.828125,
"learning_rate": 1.3392718055300105e-05,
"loss": 0.321,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19150,
"tokens_per_second_per_gpu": 2328.8
},
{
"epoch": 1.9133213501098463,
"grad_norm": 3.265625,
"learning_rate": 1.3382935155888668e-05,
"loss": 0.3528,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19160,
"tokens_per_second_per_gpu": 2497.69
},
{
"epoch": 1.914319952067106,
"grad_norm": 3.78125,
"learning_rate": 1.3373151712926112e-05,
"loss": 0.4085,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19170,
"tokens_per_second_per_gpu": 2402.64
},
{
"epoch": 1.915318554024366,
"grad_norm": 4.53125,
"learning_rate": 1.3363367732435306e-05,
"loss": 0.446,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19180,
"tokens_per_second_per_gpu": 2463.06
},
{
"epoch": 1.9163171559816257,
"grad_norm": 2.5,
"learning_rate": 1.3353583220439444e-05,
"loss": 0.3833,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19190,
"tokens_per_second_per_gpu": 2384.65
},
{
"epoch": 1.9173157579388855,
"grad_norm": 3.140625,
"learning_rate": 1.3343798182962056e-05,
"loss": 0.3986,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19200,
"tokens_per_second_per_gpu": 2360.93
},
{
"epoch": 1.9183143598961454,
"grad_norm": 3.09375,
"learning_rate": 1.3334012626026987e-05,
"loss": 0.3504,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19210,
"tokens_per_second_per_gpu": 2487.67
},
{
"epoch": 1.9193129618534053,
"grad_norm": 3.671875,
"learning_rate": 1.3324226555658404e-05,
"loss": 0.3562,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19220,
"tokens_per_second_per_gpu": 2467.01
},
{
"epoch": 1.9203115638106651,
"grad_norm": 3.09375,
"learning_rate": 1.3314439977880799e-05,
"loss": 0.3757,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19230,
"tokens_per_second_per_gpu": 2485.53
},
{
"epoch": 1.9213101657679248,
"grad_norm": 3.078125,
"learning_rate": 1.330465289871896e-05,
"loss": 0.3472,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19240,
"tokens_per_second_per_gpu": 2420.19
},
{
"epoch": 1.9223087677251849,
"grad_norm": 3.90625,
"learning_rate": 1.3294865324198e-05,
"loss": 0.3867,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19250,
"tokens_per_second_per_gpu": 2298.93
},
{
"epoch": 1.9233073696824445,
"grad_norm": 4.96875,
"learning_rate": 1.3285077260343331e-05,
"loss": 0.4261,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19260,
"tokens_per_second_per_gpu": 2265.8
},
{
"epoch": 1.9243059716397044,
"grad_norm": 2.265625,
"learning_rate": 1.3275288713180661e-05,
"loss": 0.3909,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19270,
"tokens_per_second_per_gpu": 2526.39
},
{
"epoch": 1.9253045735969643,
"grad_norm": 3.296875,
"learning_rate": 1.3265499688735999e-05,
"loss": 0.3838,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19280,
"tokens_per_second_per_gpu": 2363.31
},
{
"epoch": 1.926303175554224,
"grad_norm": 2.953125,
"learning_rate": 1.3255710193035652e-05,
"loss": 0.3529,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19290,
"tokens_per_second_per_gpu": 2483.03
},
{
"epoch": 1.927301777511484,
"grad_norm": 2.65625,
"learning_rate": 1.3245920232106212e-05,
"loss": 0.4188,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19300,
"tokens_per_second_per_gpu": 2541.76
},
{
"epoch": 1.9283003794687437,
"grad_norm": 3.828125,
"learning_rate": 1.3236129811974556e-05,
"loss": 0.4059,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19310,
"tokens_per_second_per_gpu": 2557.79
},
{
"epoch": 1.9292989814260038,
"grad_norm": 3.0625,
"learning_rate": 1.3226338938667854e-05,
"loss": 0.3831,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19320,
"tokens_per_second_per_gpu": 2360.33
},
{
"epoch": 1.9302975833832634,
"grad_norm": 3.65625,
"learning_rate": 1.3216547618213542e-05,
"loss": 0.4051,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19330,
"tokens_per_second_per_gpu": 2374.15
},
{
"epoch": 1.9312961853405233,
"grad_norm": 3.765625,
"learning_rate": 1.3206755856639338e-05,
"loss": 0.357,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19340,
"tokens_per_second_per_gpu": 2390.41
},
{
"epoch": 1.9322947872977831,
"grad_norm": 2.6875,
"learning_rate": 1.319696365997323e-05,
"loss": 0.3808,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19350,
"tokens_per_second_per_gpu": 2437.58
},
{
"epoch": 1.9332933892550428,
"grad_norm": 3.40625,
"learning_rate": 1.3187171034243476e-05,
"loss": 0.437,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19360,
"tokens_per_second_per_gpu": 2467.65
},
{
"epoch": 1.9342919912123029,
"grad_norm": 3.359375,
"learning_rate": 1.3177377985478593e-05,
"loss": 0.4186,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19370,
"tokens_per_second_per_gpu": 2390.01
},
{
"epoch": 1.9352905931695625,
"grad_norm": 3.71875,
"learning_rate": 1.3167584519707365e-05,
"loss": 0.3701,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19380,
"tokens_per_second_per_gpu": 2392.6
},
{
"epoch": 1.9362891951268224,
"grad_norm": 2.9375,
"learning_rate": 1.3157790642958833e-05,
"loss": 0.3799,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19390,
"tokens_per_second_per_gpu": 2301.39
},
{
"epoch": 1.9372877970840823,
"grad_norm": 4.15625,
"learning_rate": 1.314799636126228e-05,
"loss": 0.4096,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19400,
"tokens_per_second_per_gpu": 2433.55
},
{
"epoch": 1.9382863990413421,
"grad_norm": 3.375,
"learning_rate": 1.3138201680647252e-05,
"loss": 0.3609,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19410,
"tokens_per_second_per_gpu": 2363.51
},
{
"epoch": 1.939285000998602,
"grad_norm": 3.9375,
"learning_rate": 1.312840660714353e-05,
"loss": 0.4079,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19420,
"tokens_per_second_per_gpu": 2529.1
},
{
"epoch": 1.9402836029558617,
"grad_norm": 2.953125,
"learning_rate": 1.3118611146781144e-05,
"loss": 0.3959,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19430,
"tokens_per_second_per_gpu": 2577.74
},
{
"epoch": 1.9412822049131218,
"grad_norm": 4.8125,
"learning_rate": 1.3108815305590358e-05,
"loss": 0.4253,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19440,
"tokens_per_second_per_gpu": 2435.5
},
{
"epoch": 1.9422808068703814,
"grad_norm": 3.3125,
"learning_rate": 1.3099019089601678e-05,
"loss": 0.4079,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19450,
"tokens_per_second_per_gpu": 2305.29
},
{
"epoch": 1.9432794088276413,
"grad_norm": 3.625,
"learning_rate": 1.3089222504845828e-05,
"loss": 0.3676,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19460,
"tokens_per_second_per_gpu": 2448.48
},
{
"epoch": 1.9442780107849011,
"grad_norm": 3.8125,
"learning_rate": 1.3079425557353763e-05,
"loss": 0.3345,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19470,
"tokens_per_second_per_gpu": 2555.9
},
{
"epoch": 1.945276612742161,
"grad_norm": 3.265625,
"learning_rate": 1.3069628253156672e-05,
"loss": 0.3867,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19480,
"tokens_per_second_per_gpu": 2201.25
},
{
"epoch": 1.9462752146994209,
"grad_norm": 2.328125,
"learning_rate": 1.3059830598285953e-05,
"loss": 0.3546,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19490,
"tokens_per_second_per_gpu": 2334.56
},
{
"epoch": 1.9472738166566805,
"grad_norm": 2.65625,
"learning_rate": 1.3050032598773221e-05,
"loss": 0.4125,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19500,
"tokens_per_second_per_gpu": 2480.26
},
{
"epoch": 1.9482724186139406,
"grad_norm": 3.390625,
"learning_rate": 1.3040234260650302e-05,
"loss": 0.4182,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19510,
"tokens_per_second_per_gpu": 2456.34
},
{
"epoch": 1.9492710205712003,
"grad_norm": 3.359375,
"learning_rate": 1.3030435589949241e-05,
"loss": 0.3713,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19520,
"tokens_per_second_per_gpu": 2479.36
},
{
"epoch": 1.9502696225284601,
"grad_norm": 2.75,
"learning_rate": 1.302063659270228e-05,
"loss": 0.3242,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19530,
"tokens_per_second_per_gpu": 2496.5
},
{
"epoch": 1.95126822448572,
"grad_norm": 2.984375,
"learning_rate": 1.3010837274941848e-05,
"loss": 0.3236,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19540,
"tokens_per_second_per_gpu": 2445.66
},
{
"epoch": 1.9522668264429797,
"grad_norm": 3.59375,
"learning_rate": 1.30010376427006e-05,
"loss": 0.3843,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19550,
"tokens_per_second_per_gpu": 2652.56
},
{
"epoch": 1.9532654284002398,
"grad_norm": 3.265625,
"learning_rate": 1.2991237702011364e-05,
"loss": 0.369,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19560,
"tokens_per_second_per_gpu": 2453.71
},
{
"epoch": 1.9542640303574994,
"grad_norm": 3.6875,
"learning_rate": 1.2981437458907162e-05,
"loss": 0.3912,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19570,
"tokens_per_second_per_gpu": 2221.45
},
{
"epoch": 1.9552626323147595,
"grad_norm": 4.375,
"learning_rate": 1.297163691942121e-05,
"loss": 0.3835,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19580,
"tokens_per_second_per_gpu": 2550.85
},
{
"epoch": 1.9562612342720191,
"grad_norm": 3.09375,
"learning_rate": 1.2961836089586896e-05,
"loss": 0.3808,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19590,
"tokens_per_second_per_gpu": 2335.49
},
{
"epoch": 1.957259836229279,
"grad_norm": 4.625,
"learning_rate": 1.2952034975437787e-05,
"loss": 0.3974,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19600,
"tokens_per_second_per_gpu": 2406.32
},
{
"epoch": 1.9582584381865389,
"grad_norm": 3.53125,
"learning_rate": 1.2942233583007641e-05,
"loss": 0.438,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19610,
"tokens_per_second_per_gpu": 2369.72
},
{
"epoch": 1.9592570401437985,
"grad_norm": 3.40625,
"learning_rate": 1.2932431918330361e-05,
"loss": 0.3863,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19620,
"tokens_per_second_per_gpu": 2398.44
},
{
"epoch": 1.9602556421010586,
"grad_norm": 5.15625,
"learning_rate": 1.292262998744004e-05,
"loss": 0.3962,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19630,
"tokens_per_second_per_gpu": 2325.16
},
{
"epoch": 1.9612542440583183,
"grad_norm": 2.75,
"learning_rate": 1.291282779637093e-05,
"loss": 0.3946,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19640,
"tokens_per_second_per_gpu": 2312.46
},
{
"epoch": 1.9622528460155781,
"grad_norm": 3.421875,
"learning_rate": 1.2903025351157439e-05,
"loss": 0.4268,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19650,
"tokens_per_second_per_gpu": 2424.32
},
{
"epoch": 1.963251447972838,
"grad_norm": 3.46875,
"learning_rate": 1.2893222657834122e-05,
"loss": 0.3717,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19660,
"tokens_per_second_per_gpu": 2505.4
},
{
"epoch": 1.9642500499300979,
"grad_norm": 4.4375,
"learning_rate": 1.288341972243571e-05,
"loss": 0.3648,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19670,
"tokens_per_second_per_gpu": 2468.75
},
{
"epoch": 1.9652486518873578,
"grad_norm": 3.296875,
"learning_rate": 1.2873616550997062e-05,
"loss": 0.3591,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19680,
"tokens_per_second_per_gpu": 2415.91
},
{
"epoch": 1.9662472538446174,
"grad_norm": 2.96875,
"learning_rate": 1.2863813149553197e-05,
"loss": 0.3322,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19690,
"tokens_per_second_per_gpu": 2432.5
},
{
"epoch": 1.9672458558018775,
"grad_norm": 4.09375,
"learning_rate": 1.2854009524139265e-05,
"loss": 0.4131,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19700,
"tokens_per_second_per_gpu": 2378.15
},
{
"epoch": 1.9682444577591371,
"grad_norm": 3.5,
"learning_rate": 1.2844205680790561e-05,
"loss": 0.3607,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19710,
"tokens_per_second_per_gpu": 2339.84
},
{
"epoch": 1.969243059716397,
"grad_norm": 3.015625,
"learning_rate": 1.2834401625542508e-05,
"loss": 0.4145,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19720,
"tokens_per_second_per_gpu": 2402.8
},
{
"epoch": 1.9702416616736569,
"grad_norm": 2.75,
"learning_rate": 1.2824597364430665e-05,
"loss": 0.3871,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19730,
"tokens_per_second_per_gpu": 2680.81
},
{
"epoch": 1.9712402636309168,
"grad_norm": 3.78125,
"learning_rate": 1.2814792903490713e-05,
"loss": 0.339,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19740,
"tokens_per_second_per_gpu": 2342.42
},
{
"epoch": 1.9722388655881766,
"grad_norm": 3.453125,
"learning_rate": 1.2804988248758463e-05,
"loss": 0.2848,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19750,
"tokens_per_second_per_gpu": 2346.68
},
{
"epoch": 1.9732374675454363,
"grad_norm": 3.921875,
"learning_rate": 1.2795183406269833e-05,
"loss": 0.3673,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19760,
"tokens_per_second_per_gpu": 2420.6
},
{
"epoch": 1.9742360695026964,
"grad_norm": 3.015625,
"learning_rate": 1.2785378382060875e-05,
"loss": 0.32,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19770,
"tokens_per_second_per_gpu": 2447.83
},
{
"epoch": 1.975234671459956,
"grad_norm": 3.53125,
"learning_rate": 1.2775573182167735e-05,
"loss": 0.4006,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19780,
"tokens_per_second_per_gpu": 2475.39
},
{
"epoch": 1.9762332734172159,
"grad_norm": 4.71875,
"learning_rate": 1.2765767812626674e-05,
"loss": 0.3822,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19790,
"tokens_per_second_per_gpu": 2312.09
},
{
"epoch": 1.9772318753744758,
"grad_norm": 3.640625,
"learning_rate": 1.2755962279474063e-05,
"loss": 0.3726,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19800,
"tokens_per_second_per_gpu": 2384.89
},
{
"epoch": 1.9782304773317354,
"grad_norm": 3.453125,
"learning_rate": 1.2746156588746364e-05,
"loss": 0.3626,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19810,
"tokens_per_second_per_gpu": 2343.65
},
{
"epoch": 1.9792290792889955,
"grad_norm": 3.890625,
"learning_rate": 1.2736350746480139e-05,
"loss": 0.3677,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19820,
"tokens_per_second_per_gpu": 2454.81
},
{
"epoch": 1.9802276812462551,
"grad_norm": 2.59375,
"learning_rate": 1.272654475871205e-05,
"loss": 0.3489,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19830,
"tokens_per_second_per_gpu": 2239.12
},
{
"epoch": 1.9812262832035152,
"grad_norm": 3.53125,
"learning_rate": 1.2716738631478841e-05,
"loss": 0.39,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19840,
"tokens_per_second_per_gpu": 2462.45
},
{
"epoch": 1.9822248851607749,
"grad_norm": 3.234375,
"learning_rate": 1.2706932370817348e-05,
"loss": 0.3495,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19850,
"tokens_per_second_per_gpu": 2306.96
},
{
"epoch": 1.9832234871180348,
"grad_norm": 4.4375,
"learning_rate": 1.2697125982764477e-05,
"loss": 0.4035,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19860,
"tokens_per_second_per_gpu": 2386.43
},
{
"epoch": 1.9842220890752946,
"grad_norm": 2.34375,
"learning_rate": 1.2687319473357229e-05,
"loss": 0.3414,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19870,
"tokens_per_second_per_gpu": 2185.57
},
{
"epoch": 1.9852206910325543,
"grad_norm": 3.46875,
"learning_rate": 1.2677512848632663e-05,
"loss": 0.352,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19880,
"tokens_per_second_per_gpu": 2255.97
},
{
"epoch": 1.9862192929898144,
"grad_norm": 3.46875,
"learning_rate": 1.2667706114627926e-05,
"loss": 0.3283,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19890,
"tokens_per_second_per_gpu": 2423.27
},
{
"epoch": 1.987217894947074,
"grad_norm": 4.0625,
"learning_rate": 1.2657899277380222e-05,
"loss": 0.3461,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19900,
"tokens_per_second_per_gpu": 2496.33
},
{
"epoch": 1.9882164969043339,
"grad_norm": 3.359375,
"learning_rate": 1.2648092342926821e-05,
"loss": 0.3672,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19910,
"tokens_per_second_per_gpu": 2328.33
},
{
"epoch": 1.9892150988615938,
"grad_norm": 3.75,
"learning_rate": 1.2638285317305043e-05,
"loss": 0.3772,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19920,
"tokens_per_second_per_gpu": 2412.85
},
{
"epoch": 1.9902137008188536,
"grad_norm": 3.171875,
"learning_rate": 1.2628478206552285e-05,
"loss": 0.4296,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19930,
"tokens_per_second_per_gpu": 2376.56
},
{
"epoch": 1.9912123027761135,
"grad_norm": 3.28125,
"learning_rate": 1.2618671016705979e-05,
"loss": 0.3966,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19940,
"tokens_per_second_per_gpu": 2473.32
},
{
"epoch": 1.9922109047333731,
"grad_norm": 3.59375,
"learning_rate": 1.2608863753803615e-05,
"loss": 0.4175,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19950,
"tokens_per_second_per_gpu": 2245.25
},
{
"epoch": 1.9932095066906332,
"grad_norm": 2.515625,
"learning_rate": 1.2599056423882718e-05,
"loss": 0.3791,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19960,
"tokens_per_second_per_gpu": 2420.53
},
{
"epoch": 1.9942081086478929,
"grad_norm": 3.28125,
"learning_rate": 1.2589249032980868e-05,
"loss": 0.3791,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19970,
"tokens_per_second_per_gpu": 2388.55
},
{
"epoch": 1.9952067106051528,
"grad_norm": 3.171875,
"learning_rate": 1.2579441587135674e-05,
"loss": 0.3347,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19980,
"tokens_per_second_per_gpu": 2327.82
},
{
"epoch": 1.9962053125624126,
"grad_norm": 3.828125,
"learning_rate": 1.2569634092384771e-05,
"loss": 0.3912,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 19990,
"tokens_per_second_per_gpu": 2411.73
},
{
"epoch": 1.9972039145196725,
"grad_norm": 3.625,
"learning_rate": 1.2559826554765844e-05,
"loss": 0.3875,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 20000,
"tokens_per_second_per_gpu": 2305.24
},
{
"epoch": 1.9982025164769324,
"grad_norm": 3.703125,
"learning_rate": 1.2550018980316591e-05,
"loss": 0.3579,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 20010,
"tokens_per_second_per_gpu": 2473.31
},
{
"epoch": 1.999201118434192,
"grad_norm": 3.40625,
"learning_rate": 1.2540211375074731e-05,
"loss": 0.3578,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 20020,
"tokens_per_second_per_gpu": 2319.59
},
{
"epoch": 2.000199720391452,
"grad_norm": 2.5,
"learning_rate": 1.2530403745078012e-05,
"loss": 0.3393,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 20030,
"tokens_per_second_per_gpu": 2358.26
},
{
"epoch": 2.0011983223487118,
"grad_norm": 4.09375,
"learning_rate": 1.2520596096364188e-05,
"loss": 0.2728,
"memory/device_reserved (GiB)": 41.27,
"memory/max_active (GiB)": 40.43,
"memory/max_allocated (GiB)": 40.43,
"step": 20040,
"tokens_per_second_per_gpu": 2420.2
}
],
"logging_steps": 10,
"max_steps": 40080,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 20040,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.7895156743562854e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}
Reference in New Issue View Git Blame Copy Permalink