{ "model": "google/gemma-4-E2B-it-qat-mobile-transformers", "args": { "model": "google/gemma-4-E2B-it-qat-mobile-transformers", "steps": 600, "bs": 2, "accum": 4, "lr": 0.0001, "r": 32, "block": 1024, "load_4bit": false, "dequant_qat": true, "targets": "attn_mlp", "corpus": "/home/ubuntu/self-learning-model/strands_tune/corpus.jsonl", "out": "artifacts/strands_gemma4_e2b_lora" }, "nll_before": 2.68935369849205, "nll_after": 1.2602788805961609, "log": [ { "step": 1, "loss": 3.0693, "lr": 9.999931461237134e-05, "tok_s": 2557 }, { "step": 10, "loss": 1.7782, "lr": 9.993147673772868e-05, "tok_s": 2594 }, { "step": 20, "loss": 1.5848, "lr": 9.972609476841367e-05, "tok_s": 2583 }, { "step": 30, "loss": 1.7978, "lr": 9.938441702975687e-05, "tok_s": 2579 }, { "step": 40, "loss": 1.6216, "lr": 9.890738003669028e-05, "tok_s": 2582 }, { "step": 50, "loss": 1.4679, "lr": 9.829629131445343e-05, "tok_s": 2581 }, { "step": 60, "loss": 1.6909, "lr": 9.755282581475769e-05, "tok_s": 2578 }, { "step": 70, "loss": 1.0998, "lr": 9.667902132486009e-05, "tok_s": 2580 }, { "step": 80, "loss": 1.4509, "lr": 9.567727288213006e-05, "tok_s": 2579 }, { "step": 90, "loss": 1.4452, "lr": 9.455032620941841e-05, "tok_s": 2577 }, { "step": 100, "loss": 1.5662, "lr": 9.330127018922195e-05, "tok_s": 2579 }, { "step": 110, "loss": 1.3704, "lr": 9.193352839727123e-05, "tok_s": 2573 }, { "step": 120, "loss": 1.0406, "lr": 9.045084971874741e-05, "tok_s": 2573 }, { "step": 130, "loss": 1.3993, "lr": 8.88572980728486e-05, "tok_s": 2575 }, { "step": 140, "loss": 1.3269, "lr": 8.715724127386976e-05, "tok_s": 2574 }, { "step": 150, "loss": 1.462, "lr": 8.535533905932743e-05, "tok_s": 2573 }, { "step": 160, "loss": 1.2279, "lr": 8.345653031794296e-05, "tok_s": 2575 }, { "step": 170, "loss": 1.2806, "lr": 8.146601955249193e-05, "tok_s": 2575 }, { "step": 180, "loss": 1.1756, "lr": 7.938926261462371e-05, "tok_s": 2574 }, { "step": 190, "loss": 0.918, "lr": 7.723195175075142e-05, "tok_s": 2576 }, { "step": 200, "loss": 1.3074, "lr": 7.500000000000009e-05, "tok_s": 2575 }, { "step": 210, "loss": 1.3411, "lr": 7.269952498697741e-05, "tok_s": 2569 }, { "step": 220, "loss": 1.4813, "lr": 7.033683215379006e-05, "tok_s": 2570 }, { "step": 230, "loss": 1.3284, "lr": 6.791839747726506e-05, "tok_s": 2570 }, { "step": 240, "loss": 1.6288, "lr": 6.545084971874744e-05, "tok_s": 2570 }, { "step": 250, "loss": 1.7878, "lr": 6.294095225512611e-05, "tok_s": 2571 }, { "step": 260, "loss": 1.6825, "lr": 6.039558454088805e-05, "tok_s": 2571 }, { "step": 270, "loss": 1.2099, "lr": 5.7821723252011626e-05, "tok_s": 2570 }, { "step": 280, "loss": 1.3091, "lr": 5.522642316338277e-05, "tok_s": 2571 }, { "step": 290, "loss": 1.4771, "lr": 5.261679781214727e-05, "tok_s": 2571 }, { "step": 300, "loss": 1.0913, "lr": 5.000000000000007e-05, "tok_s": 2571 }, { "step": 310, "loss": 1.5202, "lr": 4.7383202187852876e-05, "tok_s": 2568 }, { "step": 320, "loss": 1.2693, "lr": 4.477357683661741e-05, "tok_s": 2569 }, { "step": 330, "loss": 1.1337, "lr": 4.217827674798853e-05, "tok_s": 2569 }, { "step": 340, "loss": 1.3723, "lr": 3.960441545911209e-05, "tok_s": 2569 }, { "step": 350, "loss": 1.3359, "lr": 3.7059047744874025e-05, "tok_s": 2569 }, { "step": 360, "loss": 1.5499, "lr": 3.454915028125268e-05, "tok_s": 2569 }, { "step": 370, "loss": 1.1808, "lr": 3.208160252273505e-05, "tok_s": 2569 }, { "step": 380, "loss": 1.389, "lr": 2.9663167846210028e-05, "tok_s": 2570 }, { "step": 390, "loss": 1.1763, "lr": 2.73004750130227e-05, "tok_s": 2570 }, { "step": 400, "loss": 1.3025, "lr": 2.500000000000002e-05, "tok_s": 2570 }, { "step": 410, "loss": 1.1334, "lr": 2.2768048249248692e-05, "tok_s": 2568 }, { "step": 420, "loss": 1.2506, "lr": 2.0610737385376383e-05, "tok_s": 2568 }, { "step": 430, "loss": 1.0039, "lr": 1.8533980447508165e-05, "tok_s": 2568 }, { "step": 440, "loss": 1.6885, "lr": 1.6543469682057116e-05, "tok_s": 2568 }, { "step": 450, "loss": 1.3592, "lr": 1.4644660940672653e-05, "tok_s": 2568 }, { "step": 460, "loss": 1.3486, "lr": 1.2842758726130306e-05, "tok_s": 2568 }, { "step": 470, "loss": 1.279, "lr": 1.1142701927151473e-05, "tok_s": 2569 }, { "step": 480, "loss": 1.339, "lr": 9.549150281252647e-06, "tok_s": 2569 }, { "step": 490, "loss": 1.0581, "lr": 8.066471602728818e-06, "tok_s": 2569 }, { "step": 500, "loss": 1.3058, "lr": 6.69872981077809e-06, "tok_s": 2569 }, { "step": 510, "loss": 1.5708, "lr": 5.449673790581622e-06, "tok_s": 2567 }, { "step": 520, "loss": 1.2886, "lr": 4.322727117869972e-06, "tok_s": 2567 }, { "step": 530, "loss": 1.3237, "lr": 3.3209786751399204e-06, "tok_s": 2568 }, { "step": 540, "loss": 1.1265, "lr": 2.4471741852423297e-06, "tok_s": 2568 }, { "step": 550, "loss": 1.3335, "lr": 1.7037086855465882e-06, "tok_s": 2568 }, { "step": 560, "loss": 1.6331, "lr": 1.0926199633097236e-06, "tok_s": 2568 }, { "step": 570, "loss": 1.3864, "lr": 6.155829702431183e-07, "tok_s": 2568 }, { "step": 580, "loss": 1.1831, "lr": 2.7390523158633617e-07, "tok_s": 2568 }, { "step": 590, "loss": 0.9576, "lr": 6.852326227130851e-08, "tok_s": 2568 }, { "step": 600, "loss": 0.9617, "lr": 0.0, "tok_s": 2568 } ] }