{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6365116144500715, "eval_steps": 500, "global_step": 253000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 4.9937103595410075e-05, "loss": 0.4326, "step": 500 }, { "epoch": 0.0, "learning_rate": 4.987420719082014e-05, "loss": 0.3792, "step": 1000 }, { "epoch": 0.0, "learning_rate": 4.981131078623022e-05, "loss": 0.3577, "step": 1500 }, { "epoch": 0.01, "learning_rate": 4.974841438164029e-05, "loss": 0.3538, "step": 2000 }, { "epoch": 0.01, "learning_rate": 4.968551797705036e-05, "loss": 0.3456, "step": 2500 }, { "epoch": 0.01, "learning_rate": 4.962262157246043e-05, "loss": 0.3619, "step": 3000 }, { "epoch": 0.01, "learning_rate": 4.9559725167870506e-05, "loss": 0.3313, "step": 3500 }, { "epoch": 0.01, "learning_rate": 4.949682876328058e-05, "loss": 0.3436, "step": 4000 }, { "epoch": 0.01, "learning_rate": 4.943393235869065e-05, "loss": 0.3339, "step": 4500 }, { "epoch": 0.01, "learning_rate": 4.937103595410072e-05, "loss": 0.3227, "step": 5000 }, { "epoch": 0.01, "learning_rate": 4.93081395495108e-05, "loss": 0.3142, "step": 5500 }, { "epoch": 0.02, "learning_rate": 4.9245243144920864e-05, "loss": 0.3118, "step": 6000 }, { "epoch": 0.02, "learning_rate": 4.918234674033094e-05, "loss": 0.3098, "step": 6500 }, { "epoch": 0.02, "learning_rate": 4.911945033574101e-05, "loss": 0.3005, "step": 7000 }, { "epoch": 0.02, "learning_rate": 4.905655393115108e-05, "loss": 0.295, "step": 7500 }, { "epoch": 0.02, "learning_rate": 4.8993657526561156e-05, "loss": 0.3036, "step": 8000 }, { "epoch": 0.02, "learning_rate": 4.893076112197122e-05, "loss": 0.3196, "step": 8500 }, { "epoch": 0.02, "learning_rate": 4.8867864717381295e-05, "loss": 0.3063, "step": 9000 }, { "epoch": 0.02, "learning_rate": 4.880496831279137e-05, "loss": 0.2949, "step": 9500 }, { "epoch": 0.03, "learning_rate": 4.874207190820144e-05, "loss": 0.2972, "step": 10000 }, { "epoch": 0.03, "learning_rate": 4.8679175503611514e-05, "loss": 0.3006, "step": 10500 }, { "epoch": 0.03, "learning_rate": 4.861627909902158e-05, "loss": 0.2951, "step": 11000 }, { "epoch": 0.03, "learning_rate": 4.855338269443166e-05, "loss": 0.3021, "step": 11500 }, { "epoch": 0.03, "learning_rate": 4.849048628984173e-05, "loss": 0.3026, "step": 12000 }, { "epoch": 0.03, "learning_rate": 4.84275898852518e-05, "loss": 0.2898, "step": 12500 }, { "epoch": 0.03, "learning_rate": 4.836469348066187e-05, "loss": 0.2836, "step": 13000 }, { "epoch": 0.03, "learning_rate": 4.8301797076071945e-05, "loss": 0.2859, "step": 13500 }, { "epoch": 0.04, "learning_rate": 4.823890067148202e-05, "loss": 0.2936, "step": 14000 }, { "epoch": 0.04, "learning_rate": 4.817600426689209e-05, "loss": 0.2935, "step": 14500 }, { "epoch": 0.04, "learning_rate": 4.811310786230216e-05, "loss": 0.2914, "step": 15000 }, { "epoch": 0.04, "learning_rate": 4.805021145771224e-05, "loss": 0.2925, "step": 15500 }, { "epoch": 0.04, "learning_rate": 4.79873150531223e-05, "loss": 0.2844, "step": 16000 }, { "epoch": 0.04, "learning_rate": 4.7924418648532376e-05, "loss": 0.2765, "step": 16500 }, { "epoch": 0.04, "learning_rate": 4.786152224394245e-05, "loss": 0.2866, "step": 17000 }, { "epoch": 0.04, "learning_rate": 4.779862583935252e-05, "loss": 0.2955, "step": 17500 }, { "epoch": 0.05, "learning_rate": 4.7735729434762595e-05, "loss": 0.2901, "step": 18000 }, { "epoch": 0.05, "learning_rate": 4.767283303017267e-05, "loss": 0.2822, "step": 18500 }, { "epoch": 0.05, "learning_rate": 4.7609936625582734e-05, "loss": 0.2806, "step": 19000 }, { "epoch": 0.05, "learning_rate": 4.7547040220992814e-05, "loss": 0.2816, "step": 19500 }, { "epoch": 0.05, "learning_rate": 4.748414381640288e-05, "loss": 0.2823, "step": 20000 }, { "epoch": 0.05, "learning_rate": 4.742124741181295e-05, "loss": 0.2855, "step": 20500 }, { "epoch": 0.05, "learning_rate": 4.7358351007223026e-05, "loss": 0.2848, "step": 21000 }, { "epoch": 0.05, "learning_rate": 4.72954546026331e-05, "loss": 0.2931, "step": 21500 }, { "epoch": 0.06, "learning_rate": 4.723255819804317e-05, "loss": 0.2892, "step": 22000 }, { "epoch": 0.06, "learning_rate": 4.716966179345324e-05, "loss": 0.2755, "step": 22500 }, { "epoch": 0.06, "learning_rate": 4.710676538886331e-05, "loss": 0.2772, "step": 23000 }, { "epoch": 0.06, "learning_rate": 4.7043868984273384e-05, "loss": 0.2654, "step": 23500 }, { "epoch": 0.06, "learning_rate": 4.698097257968346e-05, "loss": 0.268, "step": 24000 }, { "epoch": 0.06, "learning_rate": 4.691807617509353e-05, "loss": 0.2693, "step": 24500 }, { "epoch": 0.06, "learning_rate": 4.6855179770503596e-05, "loss": 0.2763, "step": 25000 }, { "epoch": 0.06, "learning_rate": 4.6792283365913676e-05, "loss": 0.2781, "step": 25500 }, { "epoch": 0.07, "learning_rate": 4.672938696132375e-05, "loss": 0.2792, "step": 26000 }, { "epoch": 0.07, "learning_rate": 4.6666490556733815e-05, "loss": 0.2786, "step": 26500 }, { "epoch": 0.07, "learning_rate": 4.660359415214389e-05, "loss": 0.2903, "step": 27000 }, { "epoch": 0.07, "learning_rate": 4.654069774755396e-05, "loss": 0.2633, "step": 27500 }, { "epoch": 0.07, "learning_rate": 4.6477801342964034e-05, "loss": 0.2675, "step": 28000 }, { "epoch": 0.07, "learning_rate": 4.6414904938374106e-05, "loss": 0.2765, "step": 28500 }, { "epoch": 0.07, "learning_rate": 4.635200853378417e-05, "loss": 0.2741, "step": 29000 }, { "epoch": 0.07, "learning_rate": 4.628911212919425e-05, "loss": 0.272, "step": 29500 }, { "epoch": 0.08, "learning_rate": 4.622621572460432e-05, "loss": 0.2867, "step": 30000 }, { "epoch": 0.08, "learning_rate": 4.616331932001439e-05, "loss": 0.2688, "step": 30500 }, { "epoch": 0.08, "learning_rate": 4.6100422915424464e-05, "loss": 0.2729, "step": 31000 }, { "epoch": 0.08, "learning_rate": 4.603752651083454e-05, "loss": 0.2702, "step": 31500 }, { "epoch": 0.08, "learning_rate": 4.597463010624461e-05, "loss": 0.2718, "step": 32000 }, { "epoch": 0.08, "learning_rate": 4.591173370165468e-05, "loss": 0.2716, "step": 32500 }, { "epoch": 0.08, "learning_rate": 4.584883729706475e-05, "loss": 0.2702, "step": 33000 }, { "epoch": 0.08, "learning_rate": 4.578594089247483e-05, "loss": 0.2667, "step": 33500 }, { "epoch": 0.09, "learning_rate": 4.5723044487884895e-05, "loss": 0.2697, "step": 34000 }, { "epoch": 0.09, "learning_rate": 4.566014808329497e-05, "loss": 0.2796, "step": 34500 }, { "epoch": 0.09, "learning_rate": 4.559725167870504e-05, "loss": 0.2672, "step": 35000 }, { "epoch": 0.09, "learning_rate": 4.5534355274115114e-05, "loss": 0.2671, "step": 35500 }, { "epoch": 0.09, "learning_rate": 4.547145886952519e-05, "loss": 0.2639, "step": 36000 }, { "epoch": 0.09, "learning_rate": 4.5408562464935253e-05, "loss": 0.268, "step": 36500 }, { "epoch": 0.09, "learning_rate": 4.5345666060345326e-05, "loss": 0.2759, "step": 37000 }, { "epoch": 0.09, "learning_rate": 4.52827696557554e-05, "loss": 0.2601, "step": 37500 }, { "epoch": 0.1, "learning_rate": 4.521987325116547e-05, "loss": 0.2613, "step": 38000 }, { "epoch": 0.1, "learning_rate": 4.5156976846575545e-05, "loss": 0.2692, "step": 38500 }, { "epoch": 0.1, "learning_rate": 4.509408044198561e-05, "loss": 0.2666, "step": 39000 }, { "epoch": 0.1, "learning_rate": 4.503118403739569e-05, "loss": 0.2723, "step": 39500 }, { "epoch": 0.1, "learning_rate": 4.4968287632805764e-05, "loss": 0.2647, "step": 40000 }, { "epoch": 0.1, "learning_rate": 4.490539122821583e-05, "loss": 0.2552, "step": 40500 }, { "epoch": 0.1, "learning_rate": 4.48424948236259e-05, "loss": 0.2633, "step": 41000 }, { "epoch": 0.1, "learning_rate": 4.4779598419035976e-05, "loss": 0.2532, "step": 41500 }, { "epoch": 0.11, "learning_rate": 4.471670201444605e-05, "loss": 0.2654, "step": 42000 }, { "epoch": 0.11, "learning_rate": 4.465380560985612e-05, "loss": 0.2534, "step": 42500 }, { "epoch": 0.11, "learning_rate": 4.459090920526619e-05, "loss": 0.2598, "step": 43000 }, { "epoch": 0.11, "learning_rate": 4.452801280067627e-05, "loss": 0.265, "step": 43500 }, { "epoch": 0.11, "learning_rate": 4.4465116396086334e-05, "loss": 0.2628, "step": 44000 }, { "epoch": 0.11, "learning_rate": 4.440221999149641e-05, "loss": 0.2626, "step": 44500 }, { "epoch": 0.11, "learning_rate": 4.433932358690648e-05, "loss": 0.2658, "step": 45000 }, { "epoch": 0.11, "learning_rate": 4.427642718231655e-05, "loss": 0.2592, "step": 45500 }, { "epoch": 0.12, "learning_rate": 4.4213530777726626e-05, "loss": 0.2632, "step": 46000 }, { "epoch": 0.12, "learning_rate": 4.41506343731367e-05, "loss": 0.2606, "step": 46500 }, { "epoch": 0.12, "learning_rate": 4.4087737968546765e-05, "loss": 0.2564, "step": 47000 }, { "epoch": 0.12, "learning_rate": 4.4024841563956845e-05, "loss": 0.2577, "step": 47500 }, { "epoch": 0.12, "learning_rate": 4.396194515936691e-05, "loss": 0.2542, "step": 48000 }, { "epoch": 0.12, "learning_rate": 4.3899048754776984e-05, "loss": 0.2521, "step": 48500 }, { "epoch": 0.12, "learning_rate": 4.383615235018706e-05, "loss": 0.271, "step": 49000 }, { "epoch": 0.12, "learning_rate": 4.377325594559713e-05, "loss": 0.2573, "step": 49500 }, { "epoch": 0.13, "learning_rate": 4.37103595410072e-05, "loss": 0.2593, "step": 50000 }, { "epoch": 0.13, "learning_rate": 4.364746313641727e-05, "loss": 0.2572, "step": 50500 }, { "epoch": 0.13, "learning_rate": 4.358456673182734e-05, "loss": 0.2523, "step": 51000 }, { "epoch": 0.13, "learning_rate": 4.3521670327237415e-05, "loss": 0.2591, "step": 51500 }, { "epoch": 0.13, "learning_rate": 4.345877392264749e-05, "loss": 0.2571, "step": 52000 }, { "epoch": 0.13, "learning_rate": 4.339587751805756e-05, "loss": 0.2474, "step": 52500 }, { "epoch": 0.13, "learning_rate": 4.333298111346763e-05, "loss": 0.2493, "step": 53000 }, { "epoch": 0.13, "learning_rate": 4.327008470887771e-05, "loss": 0.2685, "step": 53500 }, { "epoch": 0.14, "learning_rate": 4.320718830428778e-05, "loss": 0.2679, "step": 54000 }, { "epoch": 0.14, "learning_rate": 4.3144291899697846e-05, "loss": 0.2586, "step": 54500 }, { "epoch": 0.14, "learning_rate": 4.308139549510792e-05, "loss": 0.2474, "step": 55000 }, { "epoch": 0.14, "learning_rate": 4.301849909051799e-05, "loss": 0.2555, "step": 55500 }, { "epoch": 0.14, "learning_rate": 4.2955602685928065e-05, "loss": 0.2478, "step": 56000 }, { "epoch": 0.14, "learning_rate": 4.289270628133814e-05, "loss": 0.2587, "step": 56500 }, { "epoch": 0.14, "learning_rate": 4.2829809876748204e-05, "loss": 0.262, "step": 57000 }, { "epoch": 0.14, "learning_rate": 4.2766913472158284e-05, "loss": 0.2555, "step": 57500 }, { "epoch": 0.15, "learning_rate": 4.270401706756835e-05, "loss": 0.2689, "step": 58000 }, { "epoch": 0.15, "learning_rate": 4.264112066297842e-05, "loss": 0.2556, "step": 58500 }, { "epoch": 0.15, "learning_rate": 4.2578224258388496e-05, "loss": 0.2574, "step": 59000 }, { "epoch": 0.15, "learning_rate": 4.251532785379857e-05, "loss": 0.2644, "step": 59500 }, { "epoch": 0.15, "learning_rate": 4.245243144920864e-05, "loss": 0.2645, "step": 60000 }, { "epoch": 0.15, "learning_rate": 4.2389535044618714e-05, "loss": 0.2559, "step": 60500 }, { "epoch": 0.15, "learning_rate": 4.232663864002878e-05, "loss": 0.2547, "step": 61000 }, { "epoch": 0.15, "learning_rate": 4.226374223543886e-05, "loss": 0.2646, "step": 61500 }, { "epoch": 0.16, "learning_rate": 4.2200845830848927e-05, "loss": 0.2524, "step": 62000 }, { "epoch": 0.16, "learning_rate": 4.2137949426259e-05, "loss": 0.2621, "step": 62500 }, { "epoch": 0.16, "learning_rate": 4.207505302166907e-05, "loss": 0.248, "step": 63000 }, { "epoch": 0.16, "learning_rate": 4.2012156617079145e-05, "loss": 0.2443, "step": 63500 }, { "epoch": 0.16, "learning_rate": 4.194926021248922e-05, "loss": 0.2578, "step": 64000 }, { "epoch": 0.16, "learning_rate": 4.1886363807899285e-05, "loss": 0.2502, "step": 64500 }, { "epoch": 0.16, "learning_rate": 4.182346740330936e-05, "loss": 0.2525, "step": 65000 }, { "epoch": 0.16, "learning_rate": 4.176057099871943e-05, "loss": 0.2571, "step": 65500 }, { "epoch": 0.17, "learning_rate": 4.1697674594129503e-05, "loss": 0.2505, "step": 66000 }, { "epoch": 0.17, "learning_rate": 4.1634778189539576e-05, "loss": 0.2487, "step": 66500 }, { "epoch": 0.17, "learning_rate": 4.157188178494964e-05, "loss": 0.2545, "step": 67000 }, { "epoch": 0.17, "learning_rate": 4.150898538035972e-05, "loss": 0.2482, "step": 67500 }, { "epoch": 0.17, "learning_rate": 4.1446088975769795e-05, "loss": 0.2555, "step": 68000 }, { "epoch": 0.17, "learning_rate": 4.138319257117986e-05, "loss": 0.2433, "step": 68500 }, { "epoch": 0.17, "learning_rate": 4.1320296166589934e-05, "loss": 0.2489, "step": 69000 }, { "epoch": 0.17, "learning_rate": 4.125739976200001e-05, "loss": 0.2363, "step": 69500 }, { "epoch": 0.18, "learning_rate": 4.119450335741008e-05, "loss": 0.27, "step": 70000 }, { "epoch": 0.18, "learning_rate": 4.113160695282015e-05, "loss": 0.2568, "step": 70500 }, { "epoch": 0.18, "learning_rate": 4.106871054823022e-05, "loss": 0.2475, "step": 71000 }, { "epoch": 0.18, "learning_rate": 4.10058141436403e-05, "loss": 0.2562, "step": 71500 }, { "epoch": 0.18, "learning_rate": 4.0942917739050365e-05, "loss": 0.26, "step": 72000 }, { "epoch": 0.18, "learning_rate": 4.088002133446044e-05, "loss": 0.2586, "step": 72500 }, { "epoch": 0.18, "learning_rate": 4.081712492987051e-05, "loss": 0.2464, "step": 73000 }, { "epoch": 0.18, "learning_rate": 4.0754228525280584e-05, "loss": 0.2521, "step": 73500 }, { "epoch": 0.19, "learning_rate": 4.069133212069066e-05, "loss": 0.248, "step": 74000 }, { "epoch": 0.19, "learning_rate": 4.062843571610073e-05, "loss": 0.2555, "step": 74500 }, { "epoch": 0.19, "learning_rate": 4.0565539311510796e-05, "loss": 0.2611, "step": 75000 }, { "epoch": 0.19, "learning_rate": 4.0502642906920876e-05, "loss": 0.2577, "step": 75500 }, { "epoch": 0.19, "learning_rate": 4.043974650233094e-05, "loss": 0.254, "step": 76000 }, { "epoch": 0.19, "learning_rate": 4.0376850097741015e-05, "loss": 0.2489, "step": 76500 }, { "epoch": 0.19, "learning_rate": 4.031395369315109e-05, "loss": 0.2548, "step": 77000 }, { "epoch": 0.19, "learning_rate": 4.0251057288561154e-05, "loss": 0.2478, "step": 77500 }, { "epoch": 0.2, "learning_rate": 4.0188160883971234e-05, "loss": 0.2489, "step": 78000 }, { "epoch": 0.2, "learning_rate": 4.01252644793813e-05, "loss": 0.2491, "step": 78500 }, { "epoch": 0.2, "learning_rate": 4.006236807479137e-05, "loss": 0.2537, "step": 79000 }, { "epoch": 0.2, "learning_rate": 3.9999471670201446e-05, "loss": 0.2533, "step": 79500 }, { "epoch": 0.2, "learning_rate": 3.993657526561152e-05, "loss": 0.2558, "step": 80000 }, { "epoch": 0.2, "learning_rate": 3.987367886102159e-05, "loss": 0.2597, "step": 80500 }, { "epoch": 0.2, "learning_rate": 3.981078245643166e-05, "loss": 0.2463, "step": 81000 }, { "epoch": 0.21, "learning_rate": 3.974788605184173e-05, "loss": 0.2525, "step": 81500 }, { "epoch": 0.21, "learning_rate": 3.968498964725181e-05, "loss": 0.2454, "step": 82000 }, { "epoch": 0.21, "learning_rate": 3.962209324266188e-05, "loss": 0.2649, "step": 82500 }, { "epoch": 0.21, "learning_rate": 3.955919683807195e-05, "loss": 0.2549, "step": 83000 }, { "epoch": 0.21, "learning_rate": 3.949630043348202e-05, "loss": 0.243, "step": 83500 }, { "epoch": 0.21, "learning_rate": 3.9433404028892096e-05, "loss": 0.2472, "step": 84000 }, { "epoch": 0.21, "learning_rate": 3.937050762430217e-05, "loss": 0.2484, "step": 84500 }, { "epoch": 0.21, "learning_rate": 3.9307611219712235e-05, "loss": 0.2691, "step": 85000 }, { "epoch": 0.22, "learning_rate": 3.924471481512231e-05, "loss": 0.2475, "step": 85500 }, { "epoch": 0.22, "learning_rate": 3.918181841053238e-05, "loss": 0.2424, "step": 86000 }, { "epoch": 0.22, "learning_rate": 3.9118922005942454e-05, "loss": 0.2372, "step": 86500 }, { "epoch": 0.22, "learning_rate": 3.905602560135253e-05, "loss": 0.2601, "step": 87000 }, { "epoch": 0.22, "learning_rate": 3.899312919676259e-05, "loss": 0.2504, "step": 87500 }, { "epoch": 0.22, "learning_rate": 3.893023279217267e-05, "loss": 0.2497, "step": 88000 }, { "epoch": 0.22, "learning_rate": 3.8867336387582746e-05, "loss": 0.2458, "step": 88500 }, { "epoch": 0.22, "learning_rate": 3.880443998299281e-05, "loss": 0.2518, "step": 89000 }, { "epoch": 0.23, "learning_rate": 3.8741543578402885e-05, "loss": 0.2499, "step": 89500 }, { "epoch": 0.23, "learning_rate": 3.867864717381296e-05, "loss": 0.256, "step": 90000 }, { "epoch": 0.23, "learning_rate": 3.861575076922303e-05, "loss": 0.2478, "step": 90500 }, { "epoch": 0.23, "learning_rate": 3.8552854364633104e-05, "loss": 0.2499, "step": 91000 }, { "epoch": 0.23, "learning_rate": 3.848995796004317e-05, "loss": 0.2538, "step": 91500 }, { "epoch": 0.23, "learning_rate": 3.842706155545325e-05, "loss": 0.2355, "step": 92000 }, { "epoch": 0.23, "learning_rate": 3.8364165150863316e-05, "loss": 0.2548, "step": 92500 }, { "epoch": 0.23, "learning_rate": 3.830126874627339e-05, "loss": 0.2431, "step": 93000 }, { "epoch": 0.24, "learning_rate": 3.823837234168346e-05, "loss": 0.2553, "step": 93500 }, { "epoch": 0.24, "learning_rate": 3.8175475937093535e-05, "loss": 0.2475, "step": 94000 }, { "epoch": 0.24, "learning_rate": 3.811257953250361e-05, "loss": 0.2371, "step": 94500 }, { "epoch": 0.24, "learning_rate": 3.8049683127913674e-05, "loss": 0.2511, "step": 95000 }, { "epoch": 0.24, "learning_rate": 3.798678672332375e-05, "loss": 0.2446, "step": 95500 }, { "epoch": 0.24, "learning_rate": 3.7923890318733826e-05, "loss": 0.2425, "step": 96000 }, { "epoch": 0.24, "learning_rate": 3.786099391414389e-05, "loss": 0.245, "step": 96500 }, { "epoch": 0.24, "learning_rate": 3.7798097509553965e-05, "loss": 0.2406, "step": 97000 }, { "epoch": 0.25, "learning_rate": 3.773520110496404e-05, "loss": 0.2349, "step": 97500 }, { "epoch": 0.25, "learning_rate": 3.767230470037411e-05, "loss": 0.2405, "step": 98000 }, { "epoch": 0.25, "learning_rate": 3.7609408295784184e-05, "loss": 0.2461, "step": 98500 }, { "epoch": 0.25, "learning_rate": 3.754651189119425e-05, "loss": 0.2426, "step": 99000 }, { "epoch": 0.25, "learning_rate": 3.7483615486604323e-05, "loss": 0.2394, "step": 99500 }, { "epoch": 0.25, "learning_rate": 3.7420719082014396e-05, "loss": 0.2506, "step": 100000 }, { "epoch": 0.25, "learning_rate": 3.735782267742447e-05, "loss": 0.2478, "step": 100500 }, { "epoch": 0.25, "learning_rate": 3.729492627283454e-05, "loss": 0.2477, "step": 101000 }, { "epoch": 0.26, "learning_rate": 3.723202986824461e-05, "loss": 0.2506, "step": 101500 }, { "epoch": 0.26, "learning_rate": 3.716913346365469e-05, "loss": 0.2462, "step": 102000 }, { "epoch": 0.26, "learning_rate": 3.710623705906476e-05, "loss": 0.2501, "step": 102500 }, { "epoch": 0.26, "learning_rate": 3.704334065447483e-05, "loss": 0.2416, "step": 103000 }, { "epoch": 0.26, "learning_rate": 3.69804442498849e-05, "loss": 0.2445, "step": 103500 }, { "epoch": 0.26, "learning_rate": 3.691754784529497e-05, "loss": 0.2514, "step": 104000 }, { "epoch": 0.26, "learning_rate": 3.6854651440705046e-05, "loss": 0.2528, "step": 104500 }, { "epoch": 0.26, "learning_rate": 3.679175503611512e-05, "loss": 0.2459, "step": 105000 }, { "epoch": 0.27, "learning_rate": 3.6728858631525185e-05, "loss": 0.2491, "step": 105500 }, { "epoch": 0.27, "learning_rate": 3.6665962226935265e-05, "loss": 0.2414, "step": 106000 }, { "epoch": 0.27, "learning_rate": 3.660306582234533e-05, "loss": 0.2395, "step": 106500 }, { "epoch": 0.27, "learning_rate": 3.6540169417755404e-05, "loss": 0.2572, "step": 107000 }, { "epoch": 0.27, "learning_rate": 3.647727301316548e-05, "loss": 0.2446, "step": 107500 }, { "epoch": 0.27, "learning_rate": 3.641437660857555e-05, "loss": 0.2357, "step": 108000 }, { "epoch": 0.27, "learning_rate": 3.635148020398562e-05, "loss": 0.2481, "step": 108500 }, { "epoch": 0.27, "learning_rate": 3.628858379939569e-05, "loss": 0.2451, "step": 109000 }, { "epoch": 0.28, "learning_rate": 3.622568739480576e-05, "loss": 0.2366, "step": 109500 }, { "epoch": 0.28, "learning_rate": 3.616279099021584e-05, "loss": 0.2361, "step": 110000 }, { "epoch": 0.28, "learning_rate": 3.609989458562591e-05, "loss": 0.2361, "step": 110500 }, { "epoch": 0.28, "learning_rate": 3.603699818103598e-05, "loss": 0.2404, "step": 111000 }, { "epoch": 0.28, "learning_rate": 3.5974101776446054e-05, "loss": 0.2466, "step": 111500 }, { "epoch": 0.28, "learning_rate": 3.591120537185613e-05, "loss": 0.253, "step": 112000 }, { "epoch": 0.28, "learning_rate": 3.58483089672662e-05, "loss": 0.2248, "step": 112500 }, { "epoch": 0.28, "learning_rate": 3.5785412562676266e-05, "loss": 0.2442, "step": 113000 }, { "epoch": 0.29, "learning_rate": 3.572251615808634e-05, "loss": 0.2456, "step": 113500 }, { "epoch": 0.29, "learning_rate": 3.565961975349641e-05, "loss": 0.2305, "step": 114000 }, { "epoch": 0.29, "learning_rate": 3.5596723348906485e-05, "loss": 0.2462, "step": 114500 }, { "epoch": 0.29, "learning_rate": 3.553382694431656e-05, "loss": 0.2425, "step": 115000 }, { "epoch": 0.29, "learning_rate": 3.5470930539726624e-05, "loss": 0.2414, "step": 115500 }, { "epoch": 0.29, "learning_rate": 3.5408034135136704e-05, "loss": 0.2465, "step": 116000 }, { "epoch": 0.29, "learning_rate": 3.534513773054678e-05, "loss": 0.2511, "step": 116500 }, { "epoch": 0.29, "learning_rate": 3.528224132595684e-05, "loss": 0.2439, "step": 117000 }, { "epoch": 0.3, "learning_rate": 3.5219344921366916e-05, "loss": 0.2394, "step": 117500 }, { "epoch": 0.3, "learning_rate": 3.515644851677699e-05, "loss": 0.2388, "step": 118000 }, { "epoch": 0.3, "learning_rate": 3.509355211218706e-05, "loss": 0.2431, "step": 118500 }, { "epoch": 0.3, "learning_rate": 3.5030655707597135e-05, "loss": 0.2404, "step": 119000 }, { "epoch": 0.3, "learning_rate": 3.49677593030072e-05, "loss": 0.2469, "step": 119500 }, { "epoch": 0.3, "learning_rate": 3.490486289841728e-05, "loss": 0.2414, "step": 120000 }, { "epoch": 0.3, "learning_rate": 3.484196649382735e-05, "loss": 0.2385, "step": 120500 }, { "epoch": 0.3, "learning_rate": 3.477907008923742e-05, "loss": 0.239, "step": 121000 }, { "epoch": 0.31, "learning_rate": 3.471617368464749e-05, "loss": 0.2357, "step": 121500 }, { "epoch": 0.31, "learning_rate": 3.4653277280057566e-05, "loss": 0.236, "step": 122000 }, { "epoch": 0.31, "learning_rate": 3.459038087546764e-05, "loss": 0.2421, "step": 122500 }, { "epoch": 0.31, "learning_rate": 3.4527484470877705e-05, "loss": 0.2346, "step": 123000 }, { "epoch": 0.31, "learning_rate": 3.446458806628778e-05, "loss": 0.2424, "step": 123500 }, { "epoch": 0.31, "learning_rate": 3.440169166169786e-05, "loss": 0.2419, "step": 124000 }, { "epoch": 0.31, "learning_rate": 3.4338795257107924e-05, "loss": 0.2403, "step": 124500 }, { "epoch": 0.31, "learning_rate": 3.4275898852518e-05, "loss": 0.2452, "step": 125000 }, { "epoch": 0.32, "learning_rate": 3.421300244792807e-05, "loss": 0.2418, "step": 125500 }, { "epoch": 0.32, "learning_rate": 3.415010604333814e-05, "loss": 0.2409, "step": 126000 }, { "epoch": 0.32, "learning_rate": 3.4087209638748215e-05, "loss": 0.2372, "step": 126500 }, { "epoch": 0.32, "learning_rate": 3.402431323415828e-05, "loss": 0.23, "step": 127000 }, { "epoch": 0.32, "learning_rate": 3.3961416829568355e-05, "loss": 0.2412, "step": 127500 }, { "epoch": 0.32, "learning_rate": 3.389852042497843e-05, "loss": 0.2433, "step": 128000 }, { "epoch": 0.32, "learning_rate": 3.38356240203885e-05, "loss": 0.2335, "step": 128500 }, { "epoch": 0.32, "learning_rate": 3.3772727615798573e-05, "loss": 0.2355, "step": 129000 }, { "epoch": 0.33, "learning_rate": 3.370983121120864e-05, "loss": 0.2306, "step": 129500 }, { "epoch": 0.33, "learning_rate": 3.364693480661872e-05, "loss": 0.2284, "step": 130000 }, { "epoch": 0.33, "learning_rate": 3.358403840202879e-05, "loss": 0.2416, "step": 130500 }, { "epoch": 0.33, "learning_rate": 3.352114199743886e-05, "loss": 0.2396, "step": 131000 }, { "epoch": 0.33, "learning_rate": 3.345824559284893e-05, "loss": 0.2443, "step": 131500 }, { "epoch": 0.33, "learning_rate": 3.3395349188259004e-05, "loss": 0.2297, "step": 132000 }, { "epoch": 0.33, "learning_rate": 3.333245278366908e-05, "loss": 0.2216, "step": 132500 }, { "epoch": 0.33, "learning_rate": 3.326955637907915e-05, "loss": 0.2356, "step": 133000 }, { "epoch": 0.34, "learning_rate": 3.3206659974489217e-05, "loss": 0.2224, "step": 133500 }, { "epoch": 0.34, "learning_rate": 3.3143763569899296e-05, "loss": 0.2389, "step": 134000 }, { "epoch": 0.34, "learning_rate": 3.308086716530936e-05, "loss": 0.2429, "step": 134500 }, { "epoch": 0.34, "learning_rate": 3.3017970760719435e-05, "loss": 0.2504, "step": 135000 }, { "epoch": 0.34, "learning_rate": 3.295507435612951e-05, "loss": 0.236, "step": 135500 }, { "epoch": 0.34, "learning_rate": 3.289217795153958e-05, "loss": 0.2482, "step": 136000 }, { "epoch": 0.34, "learning_rate": 3.2829281546949654e-05, "loss": 0.2337, "step": 136500 }, { "epoch": 0.34, "learning_rate": 3.276638514235972e-05, "loss": 0.2317, "step": 137000 }, { "epoch": 0.35, "learning_rate": 3.270348873776979e-05, "loss": 0.2344, "step": 137500 }, { "epoch": 0.35, "learning_rate": 3.264059233317987e-05, "loss": 0.2425, "step": 138000 }, { "epoch": 0.35, "learning_rate": 3.257769592858994e-05, "loss": 0.2376, "step": 138500 }, { "epoch": 0.35, "learning_rate": 3.251479952400001e-05, "loss": 0.2324, "step": 139000 }, { "epoch": 0.35, "learning_rate": 3.2451903119410085e-05, "loss": 0.2389, "step": 139500 }, { "epoch": 0.35, "learning_rate": 3.238900671482016e-05, "loss": 0.2369, "step": 140000 }, { "epoch": 0.35, "learning_rate": 3.232611031023023e-05, "loss": 0.2319, "step": 140500 }, { "epoch": 0.35, "learning_rate": 3.22632139056403e-05, "loss": 0.2345, "step": 141000 }, { "epoch": 0.36, "learning_rate": 3.220031750105037e-05, "loss": 0.2443, "step": 141500 }, { "epoch": 0.36, "learning_rate": 3.213742109646044e-05, "loss": 0.2297, "step": 142000 }, { "epoch": 0.36, "learning_rate": 3.2074524691870516e-05, "loss": 0.2375, "step": 142500 }, { "epoch": 0.36, "learning_rate": 3.201162828728059e-05, "loss": 0.2328, "step": 143000 }, { "epoch": 0.36, "learning_rate": 3.1948731882690655e-05, "loss": 0.2408, "step": 143500 }, { "epoch": 0.36, "learning_rate": 3.1885835478100735e-05, "loss": 0.2379, "step": 144000 }, { "epoch": 0.36, "learning_rate": 3.182293907351081e-05, "loss": 0.2267, "step": 144500 }, { "epoch": 0.36, "learning_rate": 3.1760042668920874e-05, "loss": 0.2337, "step": 145000 }, { "epoch": 0.37, "learning_rate": 3.169714626433095e-05, "loss": 0.2363, "step": 145500 }, { "epoch": 0.37, "learning_rate": 3.163424985974102e-05, "loss": 0.2376, "step": 146000 }, { "epoch": 0.37, "learning_rate": 3.157135345515109e-05, "loss": 0.221, "step": 146500 }, { "epoch": 0.37, "learning_rate": 3.1508457050561166e-05, "loss": 0.2369, "step": 147000 }, { "epoch": 0.37, "learning_rate": 3.144556064597123e-05, "loss": 0.2481, "step": 147500 }, { "epoch": 0.37, "learning_rate": 3.138266424138131e-05, "loss": 0.2375, "step": 148000 }, { "epoch": 0.37, "learning_rate": 3.131976783679138e-05, "loss": 0.2247, "step": 148500 }, { "epoch": 0.37, "learning_rate": 3.125687143220145e-05, "loss": 0.2263, "step": 149000 }, { "epoch": 0.38, "learning_rate": 3.1193975027611524e-05, "loss": 0.236, "step": 149500 }, { "epoch": 0.38, "learning_rate": 3.11310786230216e-05, "loss": 0.2209, "step": 150000 }, { "epoch": 0.38, "learning_rate": 3.106818221843167e-05, "loss": 0.2234, "step": 150500 }, { "epoch": 0.38, "learning_rate": 3.1005285813841736e-05, "loss": 0.2381, "step": 151000 }, { "epoch": 0.38, "learning_rate": 3.094238940925181e-05, "loss": 0.2386, "step": 151500 }, { "epoch": 0.38, "learning_rate": 3.087949300466188e-05, "loss": 0.242, "step": 152000 }, { "epoch": 0.38, "learning_rate": 3.0816596600071955e-05, "loss": 0.2368, "step": 152500 }, { "epoch": 0.38, "learning_rate": 3.075370019548203e-05, "loss": 0.2309, "step": 153000 }, { "epoch": 0.39, "learning_rate": 3.06908037908921e-05, "loss": 0.2279, "step": 153500 }, { "epoch": 0.39, "learning_rate": 3.062790738630217e-05, "loss": 0.2373, "step": 154000 }, { "epoch": 0.39, "learning_rate": 3.056501098171225e-05, "loss": 0.2356, "step": 154500 }, { "epoch": 0.39, "learning_rate": 3.0502114577122316e-05, "loss": 0.2342, "step": 155000 }, { "epoch": 0.39, "learning_rate": 3.0439218172532386e-05, "loss": 0.2363, "step": 155500 }, { "epoch": 0.39, "learning_rate": 3.0376321767942455e-05, "loss": 0.2322, "step": 156000 }, { "epoch": 0.39, "learning_rate": 3.031342536335253e-05, "loss": 0.2383, "step": 156500 }, { "epoch": 0.39, "learning_rate": 3.02505289587626e-05, "loss": 0.25, "step": 157000 }, { "epoch": 0.4, "learning_rate": 3.0187632554172674e-05, "loss": 0.2359, "step": 157500 }, { "epoch": 0.4, "learning_rate": 3.0124736149582744e-05, "loss": 0.228, "step": 158000 }, { "epoch": 0.4, "learning_rate": 3.006183974499282e-05, "loss": 0.2356, "step": 158500 }, { "epoch": 0.4, "learning_rate": 2.999894334040289e-05, "loss": 0.2343, "step": 159000 }, { "epoch": 0.4, "learning_rate": 2.9936046935812963e-05, "loss": 0.2276, "step": 159500 }, { "epoch": 0.4, "learning_rate": 2.9873150531223032e-05, "loss": 0.2223, "step": 160000 }, { "epoch": 0.4, "learning_rate": 2.981025412663311e-05, "loss": 0.2224, "step": 160500 }, { "epoch": 0.41, "learning_rate": 2.9747357722043178e-05, "loss": 0.2378, "step": 161000 }, { "epoch": 0.41, "learning_rate": 2.9684461317453248e-05, "loss": 0.2377, "step": 161500 }, { "epoch": 0.41, "learning_rate": 2.962156491286332e-05, "loss": 0.232, "step": 162000 }, { "epoch": 0.41, "learning_rate": 2.9558668508273397e-05, "loss": 0.2272, "step": 162500 }, { "epoch": 0.41, "learning_rate": 2.9495772103683467e-05, "loss": 0.2354, "step": 163000 }, { "epoch": 0.41, "learning_rate": 2.9432875699093536e-05, "loss": 0.2296, "step": 163500 }, { "epoch": 0.41, "learning_rate": 2.936997929450361e-05, "loss": 0.2369, "step": 164000 }, { "epoch": 0.41, "learning_rate": 2.9307082889913685e-05, "loss": 0.2345, "step": 164500 }, { "epoch": 0.42, "learning_rate": 2.9244186485323755e-05, "loss": 0.2288, "step": 165000 }, { "epoch": 0.42, "learning_rate": 2.9181290080733824e-05, "loss": 0.2314, "step": 165500 }, { "epoch": 0.42, "learning_rate": 2.9118393676143897e-05, "loss": 0.2334, "step": 166000 }, { "epoch": 0.42, "learning_rate": 2.905549727155397e-05, "loss": 0.2347, "step": 166500 }, { "epoch": 0.42, "learning_rate": 2.8992600866964043e-05, "loss": 0.2367, "step": 167000 }, { "epoch": 0.42, "learning_rate": 2.8929704462374113e-05, "loss": 0.2364, "step": 167500 }, { "epoch": 0.42, "learning_rate": 2.8866808057784182e-05, "loss": 0.2358, "step": 168000 }, { "epoch": 0.42, "learning_rate": 2.880391165319426e-05, "loss": 0.2296, "step": 168500 }, { "epoch": 0.43, "learning_rate": 2.8741015248604332e-05, "loss": 0.2336, "step": 169000 }, { "epoch": 0.43, "learning_rate": 2.86781188440144e-05, "loss": 0.2265, "step": 169500 }, { "epoch": 0.43, "learning_rate": 2.861522243942447e-05, "loss": 0.23, "step": 170000 }, { "epoch": 0.43, "learning_rate": 2.8552326034834547e-05, "loss": 0.2467, "step": 170500 }, { "epoch": 0.43, "learning_rate": 2.8489429630244617e-05, "loss": 0.2343, "step": 171000 }, { "epoch": 0.43, "learning_rate": 2.842653322565469e-05, "loss": 0.2249, "step": 171500 }, { "epoch": 0.43, "learning_rate": 2.836363682106476e-05, "loss": 0.2364, "step": 172000 }, { "epoch": 0.43, "learning_rate": 2.8300740416474836e-05, "loss": 0.2291, "step": 172500 }, { "epoch": 0.44, "learning_rate": 2.8237844011884905e-05, "loss": 0.243, "step": 173000 }, { "epoch": 0.44, "learning_rate": 2.8174947607294978e-05, "loss": 0.218, "step": 173500 }, { "epoch": 0.44, "learning_rate": 2.8112051202705048e-05, "loss": 0.2356, "step": 174000 }, { "epoch": 0.44, "learning_rate": 2.8049154798115124e-05, "loss": 0.2378, "step": 174500 }, { "epoch": 0.44, "learning_rate": 2.7986258393525194e-05, "loss": 0.2303, "step": 175000 }, { "epoch": 0.44, "learning_rate": 2.7923361988935263e-05, "loss": 0.227, "step": 175500 }, { "epoch": 0.44, "learning_rate": 2.7860465584345336e-05, "loss": 0.2266, "step": 176000 }, { "epoch": 0.44, "learning_rate": 2.7797569179755413e-05, "loss": 0.2384, "step": 176500 }, { "epoch": 0.45, "learning_rate": 2.7734672775165482e-05, "loss": 0.2256, "step": 177000 }, { "epoch": 0.45, "learning_rate": 2.767177637057555e-05, "loss": 0.2358, "step": 177500 }, { "epoch": 0.45, "learning_rate": 2.7608879965985625e-05, "loss": 0.2248, "step": 178000 }, { "epoch": 0.45, "learning_rate": 2.75459835613957e-05, "loss": 0.2335, "step": 178500 }, { "epoch": 0.45, "learning_rate": 2.748308715680577e-05, "loss": 0.2181, "step": 179000 }, { "epoch": 0.45, "learning_rate": 2.742019075221584e-05, "loss": 0.2301, "step": 179500 }, { "epoch": 0.45, "learning_rate": 2.7357294347625913e-05, "loss": 0.2207, "step": 180000 }, { "epoch": 0.45, "learning_rate": 2.7294397943035986e-05, "loss": 0.2367, "step": 180500 }, { "epoch": 0.46, "learning_rate": 2.723150153844606e-05, "loss": 0.2305, "step": 181000 }, { "epoch": 0.46, "learning_rate": 2.716860513385613e-05, "loss": 0.2201, "step": 181500 }, { "epoch": 0.46, "learning_rate": 2.7105708729266198e-05, "loss": 0.2234, "step": 182000 }, { "epoch": 0.46, "learning_rate": 2.7042812324676274e-05, "loss": 0.2391, "step": 182500 }, { "epoch": 0.46, "learning_rate": 2.6979915920086347e-05, "loss": 0.2288, "step": 183000 }, { "epoch": 0.46, "learning_rate": 2.6917019515496417e-05, "loss": 0.2384, "step": 183500 }, { "epoch": 0.46, "learning_rate": 2.6854123110906486e-05, "loss": 0.2233, "step": 184000 }, { "epoch": 0.46, "learning_rate": 2.6791226706316563e-05, "loss": 0.2312, "step": 184500 }, { "epoch": 0.47, "learning_rate": 2.6728330301726632e-05, "loss": 0.2206, "step": 185000 }, { "epoch": 0.47, "learning_rate": 2.6665433897136705e-05, "loss": 0.2293, "step": 185500 }, { "epoch": 0.47, "learning_rate": 2.6602537492546775e-05, "loss": 0.2405, "step": 186000 }, { "epoch": 0.47, "learning_rate": 2.653964108795685e-05, "loss": 0.2242, "step": 186500 }, { "epoch": 0.47, "learning_rate": 2.647674468336692e-05, "loss": 0.2358, "step": 187000 }, { "epoch": 0.47, "learning_rate": 2.6413848278776994e-05, "loss": 0.2468, "step": 187500 }, { "epoch": 0.47, "learning_rate": 2.6350951874187063e-05, "loss": 0.2222, "step": 188000 }, { "epoch": 0.47, "learning_rate": 2.628805546959714e-05, "loss": 0.2262, "step": 188500 }, { "epoch": 0.48, "learning_rate": 2.622515906500721e-05, "loss": 0.226, "step": 189000 }, { "epoch": 0.48, "learning_rate": 2.616226266041728e-05, "loss": 0.223, "step": 189500 }, { "epoch": 0.48, "learning_rate": 2.6099366255827352e-05, "loss": 0.2215, "step": 190000 }, { "epoch": 0.48, "learning_rate": 2.6036469851237428e-05, "loss": 0.2264, "step": 190500 }, { "epoch": 0.48, "learning_rate": 2.5973573446647498e-05, "loss": 0.2455, "step": 191000 }, { "epoch": 0.48, "learning_rate": 2.5910677042057567e-05, "loss": 0.2315, "step": 191500 }, { "epoch": 0.48, "learning_rate": 2.584778063746764e-05, "loss": 0.2327, "step": 192000 }, { "epoch": 0.48, "learning_rate": 2.5784884232877717e-05, "loss": 0.2342, "step": 192500 }, { "epoch": 0.49, "learning_rate": 2.5721987828287786e-05, "loss": 0.2311, "step": 193000 }, { "epoch": 0.49, "learning_rate": 2.5659091423697856e-05, "loss": 0.2294, "step": 193500 }, { "epoch": 0.49, "learning_rate": 2.559619501910793e-05, "loss": 0.2348, "step": 194000 }, { "epoch": 0.49, "learning_rate": 2.5533298614518e-05, "loss": 0.2348, "step": 194500 }, { "epoch": 0.49, "learning_rate": 2.5470402209928074e-05, "loss": 0.2456, "step": 195000 }, { "epoch": 0.49, "learning_rate": 2.5407505805338144e-05, "loss": 0.2339, "step": 195500 }, { "epoch": 0.49, "learning_rate": 2.5344609400748214e-05, "loss": 0.2278, "step": 196000 }, { "epoch": 0.49, "learning_rate": 2.528171299615829e-05, "loss": 0.2268, "step": 196500 }, { "epoch": 0.5, "learning_rate": 2.5218816591568363e-05, "loss": 0.2358, "step": 197000 }, { "epoch": 0.5, "learning_rate": 2.5155920186978432e-05, "loss": 0.2276, "step": 197500 }, { "epoch": 0.5, "learning_rate": 2.5093023782388502e-05, "loss": 0.2324, "step": 198000 }, { "epoch": 0.5, "learning_rate": 2.503012737779858e-05, "loss": 0.2225, "step": 198500 }, { "epoch": 0.5, "learning_rate": 2.4967230973208648e-05, "loss": 0.2239, "step": 199000 }, { "epoch": 0.5, "learning_rate": 2.490433456861872e-05, "loss": 0.2149, "step": 199500 }, { "epoch": 0.5, "learning_rate": 2.4841438164028794e-05, "loss": 0.2314, "step": 200000 }, { "epoch": 0.5, "learning_rate": 2.4778541759438863e-05, "loss": 0.2174, "step": 200500 }, { "epoch": 0.51, "learning_rate": 2.4715645354848936e-05, "loss": 0.2175, "step": 201000 }, { "epoch": 0.51, "learning_rate": 2.465274895025901e-05, "loss": 0.2135, "step": 201500 }, { "epoch": 0.51, "learning_rate": 2.4589852545669082e-05, "loss": 0.2321, "step": 202000 }, { "epoch": 0.51, "learning_rate": 2.4526956141079152e-05, "loss": 0.2426, "step": 202500 }, { "epoch": 0.51, "learning_rate": 2.4464059736489225e-05, "loss": 0.2375, "step": 203000 }, { "epoch": 0.51, "learning_rate": 2.4401163331899294e-05, "loss": 0.2278, "step": 203500 }, { "epoch": 0.51, "learning_rate": 2.433826692730937e-05, "loss": 0.2233, "step": 204000 }, { "epoch": 0.51, "learning_rate": 2.427537052271944e-05, "loss": 0.2356, "step": 204500 }, { "epoch": 0.52, "learning_rate": 2.4212474118129513e-05, "loss": 0.2237, "step": 205000 }, { "epoch": 0.52, "learning_rate": 2.4149577713539583e-05, "loss": 0.2284, "step": 205500 }, { "epoch": 0.52, "learning_rate": 2.4086681308949656e-05, "loss": 0.2253, "step": 206000 }, { "epoch": 0.52, "learning_rate": 2.402378490435973e-05, "loss": 0.2254, "step": 206500 }, { "epoch": 0.52, "learning_rate": 2.39608884997698e-05, "loss": 0.2192, "step": 207000 }, { "epoch": 0.52, "learning_rate": 2.389799209517987e-05, "loss": 0.2282, "step": 207500 }, { "epoch": 0.52, "learning_rate": 2.3835095690589944e-05, "loss": 0.2345, "step": 208000 }, { "epoch": 0.52, "learning_rate": 2.3772199286000017e-05, "loss": 0.2258, "step": 208500 }, { "epoch": 0.53, "learning_rate": 2.370930288141009e-05, "loss": 0.2339, "step": 209000 }, { "epoch": 0.53, "learning_rate": 2.364640647682016e-05, "loss": 0.2344, "step": 209500 }, { "epoch": 0.53, "learning_rate": 2.3583510072230233e-05, "loss": 0.2336, "step": 210000 }, { "epoch": 0.53, "learning_rate": 2.3520613667640302e-05, "loss": 0.2336, "step": 210500 }, { "epoch": 0.53, "learning_rate": 2.345771726305038e-05, "loss": 0.2342, "step": 211000 }, { "epoch": 0.53, "learning_rate": 2.3394820858460448e-05, "loss": 0.2311, "step": 211500 }, { "epoch": 0.53, "learning_rate": 2.333192445387052e-05, "loss": 0.2204, "step": 212000 }, { "epoch": 0.53, "learning_rate": 2.326902804928059e-05, "loss": 0.211, "step": 212500 }, { "epoch": 0.54, "learning_rate": 2.3206131644690664e-05, "loss": 0.2202, "step": 213000 }, { "epoch": 0.54, "learning_rate": 2.3143235240100736e-05, "loss": 0.2302, "step": 213500 }, { "epoch": 0.54, "learning_rate": 2.308033883551081e-05, "loss": 0.2375, "step": 214000 }, { "epoch": 0.54, "learning_rate": 2.301744243092088e-05, "loss": 0.2115, "step": 214500 }, { "epoch": 0.54, "learning_rate": 2.2954546026330952e-05, "loss": 0.2351, "step": 215000 }, { "epoch": 0.54, "learning_rate": 2.2891649621741025e-05, "loss": 0.2228, "step": 215500 }, { "epoch": 0.54, "learning_rate": 2.2828753217151098e-05, "loss": 0.2216, "step": 216000 }, { "epoch": 0.54, "learning_rate": 2.2765856812561167e-05, "loss": 0.2263, "step": 216500 }, { "epoch": 0.55, "learning_rate": 2.270296040797124e-05, "loss": 0.2098, "step": 217000 }, { "epoch": 0.55, "learning_rate": 2.264006400338131e-05, "loss": 0.2137, "step": 217500 }, { "epoch": 0.55, "learning_rate": 2.2577167598791386e-05, "loss": 0.2272, "step": 218000 }, { "epoch": 0.55, "learning_rate": 2.2514271194201456e-05, "loss": 0.2285, "step": 218500 }, { "epoch": 0.55, "learning_rate": 2.245137478961153e-05, "loss": 0.2369, "step": 219000 }, { "epoch": 0.55, "learning_rate": 2.23884783850216e-05, "loss": 0.2274, "step": 219500 }, { "epoch": 0.55, "learning_rate": 2.232558198043167e-05, "loss": 0.2245, "step": 220000 }, { "epoch": 0.55, "learning_rate": 2.2262685575841744e-05, "loss": 0.2282, "step": 220500 }, { "epoch": 0.56, "learning_rate": 2.2199789171251817e-05, "loss": 0.2364, "step": 221000 }, { "epoch": 0.56, "learning_rate": 2.2136892766661887e-05, "loss": 0.2125, "step": 221500 }, { "epoch": 0.56, "learning_rate": 2.207399636207196e-05, "loss": 0.2279, "step": 222000 }, { "epoch": 0.56, "learning_rate": 2.2011099957482033e-05, "loss": 0.2279, "step": 222500 }, { "epoch": 0.56, "learning_rate": 2.1948203552892106e-05, "loss": 0.2225, "step": 223000 }, { "epoch": 0.56, "learning_rate": 2.1885307148302175e-05, "loss": 0.2267, "step": 223500 }, { "epoch": 0.56, "learning_rate": 2.1822410743712248e-05, "loss": 0.2213, "step": 224000 }, { "epoch": 0.56, "learning_rate": 2.1759514339122318e-05, "loss": 0.2317, "step": 224500 }, { "epoch": 0.57, "learning_rate": 2.1696617934532394e-05, "loss": 0.2281, "step": 225000 }, { "epoch": 0.57, "learning_rate": 2.1633721529942464e-05, "loss": 0.226, "step": 225500 }, { "epoch": 0.57, "learning_rate": 2.1570825125352537e-05, "loss": 0.2272, "step": 226000 }, { "epoch": 0.57, "learning_rate": 2.1507928720762606e-05, "loss": 0.2238, "step": 226500 }, { "epoch": 0.57, "learning_rate": 2.144503231617268e-05, "loss": 0.2267, "step": 227000 }, { "epoch": 0.57, "learning_rate": 2.1382135911582752e-05, "loss": 0.2314, "step": 227500 }, { "epoch": 0.57, "learning_rate": 2.131923950699282e-05, "loss": 0.2188, "step": 228000 }, { "epoch": 0.57, "learning_rate": 2.1256343102402895e-05, "loss": 0.2231, "step": 228500 }, { "epoch": 0.58, "learning_rate": 2.1193446697812968e-05, "loss": 0.2262, "step": 229000 }, { "epoch": 0.58, "learning_rate": 2.113055029322304e-05, "loss": 0.2294, "step": 229500 }, { "epoch": 0.58, "learning_rate": 2.106765388863311e-05, "loss": 0.2286, "step": 230000 }, { "epoch": 0.58, "learning_rate": 2.1004757484043183e-05, "loss": 0.2416, "step": 230500 }, { "epoch": 0.58, "learning_rate": 2.0941861079453253e-05, "loss": 0.2296, "step": 231000 }, { "epoch": 0.58, "learning_rate": 2.0878964674863326e-05, "loss": 0.224, "step": 231500 }, { "epoch": 0.58, "learning_rate": 2.08160682702734e-05, "loss": 0.2238, "step": 232000 }, { "epoch": 0.58, "learning_rate": 2.075317186568347e-05, "loss": 0.2254, "step": 232500 }, { "epoch": 0.59, "learning_rate": 2.069027546109354e-05, "loss": 0.2319, "step": 233000 }, { "epoch": 0.59, "learning_rate": 2.0627379056503614e-05, "loss": 0.2193, "step": 233500 }, { "epoch": 0.59, "learning_rate": 2.0564482651913687e-05, "loss": 0.2257, "step": 234000 }, { "epoch": 0.59, "learning_rate": 2.050158624732376e-05, "loss": 0.2168, "step": 234500 }, { "epoch": 0.59, "learning_rate": 2.043868984273383e-05, "loss": 0.2348, "step": 235000 }, { "epoch": 0.59, "learning_rate": 2.0375793438143902e-05, "loss": 0.2292, "step": 235500 }, { "epoch": 0.59, "learning_rate": 2.0312897033553975e-05, "loss": 0.2335, "step": 236000 }, { "epoch": 0.59, "learning_rate": 2.0250000628964048e-05, "loss": 0.2223, "step": 236500 }, { "epoch": 0.6, "learning_rate": 2.0187104224374118e-05, "loss": 0.2085, "step": 237000 }, { "epoch": 0.6, "learning_rate": 2.012420781978419e-05, "loss": 0.2226, "step": 237500 }, { "epoch": 0.6, "learning_rate": 2.006131141519426e-05, "loss": 0.2331, "step": 238000 }, { "epoch": 0.6, "learning_rate": 1.9998415010604333e-05, "loss": 0.2372, "step": 238500 }, { "epoch": 0.6, "learning_rate": 1.9935518606014406e-05, "loss": 0.2242, "step": 239000 }, { "epoch": 0.6, "learning_rate": 1.987262220142448e-05, "loss": 0.2254, "step": 239500 }, { "epoch": 0.6, "learning_rate": 1.980972579683455e-05, "loss": 0.2273, "step": 240000 }, { "epoch": 0.61, "learning_rate": 1.9746829392244622e-05, "loss": 0.2259, "step": 240500 }, { "epoch": 0.61, "learning_rate": 1.9683932987654695e-05, "loss": 0.2236, "step": 241000 }, { "epoch": 0.61, "learning_rate": 1.9621036583064768e-05, "loss": 0.2204, "step": 241500 }, { "epoch": 0.61, "learning_rate": 1.9558140178474837e-05, "loss": 0.2158, "step": 242000 }, { "epoch": 0.61, "learning_rate": 1.949524377388491e-05, "loss": 0.2304, "step": 242500 }, { "epoch": 0.61, "learning_rate": 1.9432347369294983e-05, "loss": 0.2187, "step": 243000 }, { "epoch": 0.61, "learning_rate": 1.9369450964705056e-05, "loss": 0.2283, "step": 243500 }, { "epoch": 0.61, "learning_rate": 1.9306554560115126e-05, "loss": 0.2306, "step": 244000 }, { "epoch": 0.62, "learning_rate": 1.92436581555252e-05, "loss": 0.2311, "step": 244500 }, { "epoch": 0.62, "learning_rate": 1.9180761750935268e-05, "loss": 0.2314, "step": 245000 }, { "epoch": 0.62, "learning_rate": 1.911786534634534e-05, "loss": 0.2115, "step": 245500 }, { "epoch": 0.62, "learning_rate": 1.9054968941755414e-05, "loss": 0.2216, "step": 246000 }, { "epoch": 0.62, "learning_rate": 1.8992072537165487e-05, "loss": 0.2268, "step": 246500 }, { "epoch": 0.62, "learning_rate": 1.8929176132575557e-05, "loss": 0.2192, "step": 247000 }, { "epoch": 0.62, "learning_rate": 1.886627972798563e-05, "loss": 0.2207, "step": 247500 }, { "epoch": 0.62, "learning_rate": 1.8803383323395702e-05, "loss": 0.2203, "step": 248000 }, { "epoch": 0.63, "learning_rate": 1.8740486918805775e-05, "loss": 0.2364, "step": 248500 }, { "epoch": 0.63, "learning_rate": 1.8677590514215845e-05, "loss": 0.2298, "step": 249000 }, { "epoch": 0.63, "learning_rate": 1.8614694109625918e-05, "loss": 0.2236, "step": 249500 }, { "epoch": 0.63, "learning_rate": 1.855179770503599e-05, "loss": 0.2255, "step": 250000 }, { "epoch": 0.63, "learning_rate": 1.8488901300446064e-05, "loss": 0.2184, "step": 250500 }, { "epoch": 0.63, "learning_rate": 1.8426004895856133e-05, "loss": 0.2194, "step": 251000 }, { "epoch": 0.63, "learning_rate": 1.8363108491266206e-05, "loss": 0.2113, "step": 251500 }, { "epoch": 0.63, "learning_rate": 1.8300212086676276e-05, "loss": 0.22, "step": 252000 }, { "epoch": 0.64, "learning_rate": 1.823731568208635e-05, "loss": 0.2269, "step": 252500 }, { "epoch": 0.64, "learning_rate": 1.8174419277496422e-05, "loss": 0.2025, "step": 253000 } ], "logging_steps": 500, "max_steps": 397479, "num_train_epochs": 1, "save_steps": 500, "total_flos": 4.473124678201344e+16, "trial_name": null, "trial_params": null }