Instructions to use lonewolf07/adaption_embedded_systems_qa with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use lonewolf07/adaption_embedded_systems_qa with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("togethercomputer/Meta-Llama-3.3-70B-Instruct-Reference") model = PeftModel.from_pretrained(base_model, "lonewolf07/adaption_embedded_systems_qa") - Notebooks
- Google Colab
- Kaggle
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 108, | |
| "global_step": 540, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005555555555555556, | |
| "grad_norm": 0.23889054358005524, | |
| "learning_rate": 0.0, | |
| "loss": 1.33251953125, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.011111111111111112, | |
| "grad_norm": 0.21150308847427368, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 1.19805908203125, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.016666666666666666, | |
| "grad_norm": 0.3076956272125244, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 1.494140625, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.022222222222222223, | |
| "grad_norm": 0.2740148901939392, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 1.478515625, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.027777777777777776, | |
| "grad_norm": 0.2140553891658783, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 1.4482421875, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03333333333333333, | |
| "grad_norm": 0.16574729979038239, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 1.3896484375, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03888888888888889, | |
| "grad_norm": 0.15761591494083405, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 1.4208984375, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 0.10756171494722366, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 1.23046875, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.10914932936429977, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 1.30859375, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 0.09230244159698486, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.11407470703125, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06111111111111111, | |
| "grad_norm": 0.07531458884477615, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 1.177734375, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06666666666666667, | |
| "grad_norm": 0.07599915564060211, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 1.2685546875, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07222222222222222, | |
| "grad_norm": 0.08924313634634018, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 1.23828125, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07777777777777778, | |
| "grad_norm": 0.09349751472473145, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 1.0948486328125, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 0.09443597495555878, | |
| "learning_rate": 5.185185185185185e-05, | |
| "loss": 1.150390625, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 0.08670377731323242, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 1.20166015625, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09444444444444444, | |
| "grad_norm": 0.07645947486162186, | |
| "learning_rate": 5.925925925925926e-05, | |
| "loss": 1.10595703125, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.06286032497882843, | |
| "learning_rate": 6.296296296296296e-05, | |
| "loss": 1.06494140625, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10555555555555556, | |
| "grad_norm": 0.05197962746024132, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 1.19091796875, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.056515783071517944, | |
| "learning_rate": 7.037037037037038e-05, | |
| "loss": 1.193359375, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11666666666666667, | |
| "grad_norm": 0.04532347992062569, | |
| "learning_rate": 7.407407407407407e-05, | |
| "loss": 1.01318359375, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12222222222222222, | |
| "grad_norm": 0.053678810596466064, | |
| "learning_rate": 7.777777777777778e-05, | |
| "loss": 1.07763671875, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.12777777777777777, | |
| "grad_norm": 0.04876302555203438, | |
| "learning_rate": 8.148148148148148e-05, | |
| "loss": 1.2138671875, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 0.050144899636507034, | |
| "learning_rate": 8.518518518518518e-05, | |
| "loss": 1.10546875, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1388888888888889, | |
| "grad_norm": 0.04767259210348129, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 1.173828125, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14444444444444443, | |
| "grad_norm": 0.04734437167644501, | |
| "learning_rate": 9.25925925925926e-05, | |
| "loss": 0.90924072265625, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.04831893369555473, | |
| "learning_rate": 9.62962962962963e-05, | |
| "loss": 1.10205078125, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.15555555555555556, | |
| "grad_norm": 0.04482392966747284, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0498046875, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.16111111111111112, | |
| "grad_norm": 0.04558224603533745, | |
| "learning_rate": 9.999915618706986e-05, | |
| "loss": 1.0361328125, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 0.04807697981595993, | |
| "learning_rate": 9.999662477992475e-05, | |
| "loss": 1.09814453125, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17222222222222222, | |
| "grad_norm": 0.04267172887921333, | |
| "learning_rate": 9.999240587349952e-05, | |
| "loss": 1.13037109375, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 0.03471946343779564, | |
| "learning_rate": 9.9986499626015e-05, | |
| "loss": 0.8858642578125, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.18333333333333332, | |
| "grad_norm": 0.04485335573554039, | |
| "learning_rate": 9.997890625897194e-05, | |
| "loss": 1.11767578125, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.18888888888888888, | |
| "grad_norm": 0.040208980441093445, | |
| "learning_rate": 9.996962605714287e-05, | |
| "loss": 1.06201171875, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.19444444444444445, | |
| "grad_norm": 0.042576663196086884, | |
| "learning_rate": 9.995865936856132e-05, | |
| "loss": 1.08349609375, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.037157509475946426, | |
| "learning_rate": 9.994600660450875e-05, | |
| "loss": 1.05322265625, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.20555555555555555, | |
| "grad_norm": 0.04873045161366463, | |
| "learning_rate": 9.993166823949923e-05, | |
| "loss": 1.01806640625, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2111111111111111, | |
| "grad_norm": 0.03682542219758034, | |
| "learning_rate": 9.991564481126155e-05, | |
| "loss": 1.0263671875, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.21666666666666667, | |
| "grad_norm": 0.04021592438220978, | |
| "learning_rate": 9.989793692071908e-05, | |
| "loss": 1.0537109375, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.03987225145101547, | |
| "learning_rate": 9.987854523196724e-05, | |
| "loss": 1.01416015625, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.22777777777777777, | |
| "grad_norm": 0.034224290400743484, | |
| "learning_rate": 9.985747047224858e-05, | |
| "loss": 0.873779296875, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.23333333333333334, | |
| "grad_norm": 0.040917523205280304, | |
| "learning_rate": 9.983471343192558e-05, | |
| "loss": 1.01025390625, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2388888888888889, | |
| "grad_norm": 0.04020630940794945, | |
| "learning_rate": 9.981027496445085e-05, | |
| "loss": 1.0263671875, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.24444444444444444, | |
| "grad_norm": 0.038016047328710556, | |
| "learning_rate": 9.978415598633528e-05, | |
| "loss": 0.9833984375, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.04478270933032036, | |
| "learning_rate": 9.975635747711364e-05, | |
| "loss": 1.044921875, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.25555555555555554, | |
| "grad_norm": 0.03915729373693466, | |
| "learning_rate": 9.972688047930772e-05, | |
| "loss": 1.02001953125, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.2611111111111111, | |
| "grad_norm": 0.04099868983030319, | |
| "learning_rate": 9.969572609838744e-05, | |
| "loss": 1.03515625, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.03876965865492821, | |
| "learning_rate": 9.966289550272918e-05, | |
| "loss": 0.97314453125, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.2722222222222222, | |
| "grad_norm": 0.03561818227171898, | |
| "learning_rate": 9.962838992357212e-05, | |
| "loss": 0.962890625, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 0.03729720786213875, | |
| "learning_rate": 9.959221065497195e-05, | |
| "loss": 0.96875, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2833333333333333, | |
| "grad_norm": 0.039395272731781006, | |
| "learning_rate": 9.955435905375249e-05, | |
| "loss": 0.99755859375, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.28888888888888886, | |
| "grad_norm": 0.03935179114341736, | |
| "learning_rate": 9.95148365394546e-05, | |
| "loss": 1.005859375, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.29444444444444445, | |
| "grad_norm": 0.05490157753229141, | |
| "learning_rate": 9.947364459428316e-05, | |
| "loss": 0.99755859375, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.0621793270111084, | |
| "learning_rate": 9.943078476305127e-05, | |
| "loss": 1.01416015625, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3055555555555556, | |
| "grad_norm": 0.0388559065759182, | |
| "learning_rate": 9.938625865312251e-05, | |
| "loss": 0.95947265625, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 0.04379229620099068, | |
| "learning_rate": 9.934006793435052e-05, | |
| "loss": 1.0126953125, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.31666666666666665, | |
| "grad_norm": 0.0424816869199276, | |
| "learning_rate": 9.929221433901645e-05, | |
| "loss": 0.93359375, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.32222222222222224, | |
| "grad_norm": 0.03722338005900383, | |
| "learning_rate": 9.924269966176399e-05, | |
| "loss": 0.96337890625, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3277777777777778, | |
| "grad_norm": 0.041776008903980255, | |
| "learning_rate": 9.919152575953197e-05, | |
| "loss": 0.96826171875, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.04373903572559357, | |
| "learning_rate": 9.91386945514849e-05, | |
| "loss": 0.949462890625, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3388888888888889, | |
| "grad_norm": 0.038092926144599915, | |
| "learning_rate": 9.908420801894083e-05, | |
| "loss": 0.9423828125, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.34444444444444444, | |
| "grad_norm": 0.04326325282454491, | |
| "learning_rate": 9.902806820529712e-05, | |
| "loss": 1.03662109375, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.041222795844078064, | |
| "learning_rate": 9.89702772159538e-05, | |
| "loss": 0.9765625, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.043109845370054245, | |
| "learning_rate": 9.891083721823461e-05, | |
| "loss": 0.94091796875, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.3611111111111111, | |
| "grad_norm": 0.04376206547021866, | |
| "learning_rate": 9.88497504413057e-05, | |
| "loss": 0.9052734375, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.36666666666666664, | |
| "grad_norm": 0.041749946773052216, | |
| "learning_rate": 9.878701917609207e-05, | |
| "loss": 0.86328125, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.37222222222222223, | |
| "grad_norm": 0.04268103092908859, | |
| "learning_rate": 9.872264577519163e-05, | |
| "loss": 0.90576171875, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.37777777777777777, | |
| "grad_norm": 0.04776822775602341, | |
| "learning_rate": 9.865663265278693e-05, | |
| "loss": 0.9892578125, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.38333333333333336, | |
| "grad_norm": 0.09423349797725677, | |
| "learning_rate": 9.85889822845547e-05, | |
| "loss": 0.93896484375, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 0.04379543289542198, | |
| "learning_rate": 9.851969720757299e-05, | |
| "loss": 0.9541015625, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.39444444444444443, | |
| "grad_norm": 0.044626809656620026, | |
| "learning_rate": 9.844878002022593e-05, | |
| "loss": 0.9775390625, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.05369943007826805, | |
| "learning_rate": 9.837623338210641e-05, | |
| "loss": 0.9638671875, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.40555555555555556, | |
| "grad_norm": 0.05514642968773842, | |
| "learning_rate": 9.830206001391626e-05, | |
| "loss": 0.8740234375, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4111111111111111, | |
| "grad_norm": 0.04650852084159851, | |
| "learning_rate": 9.822626269736427e-05, | |
| "loss": 0.9560546875, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 0.06122339144349098, | |
| "learning_rate": 9.814884427506176e-05, | |
| "loss": 0.92431640625, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4222222222222222, | |
| "grad_norm": 0.04741847142577171, | |
| "learning_rate": 9.806980765041614e-05, | |
| "loss": 0.9140625, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.42777777777777776, | |
| "grad_norm": 0.04446836933493614, | |
| "learning_rate": 9.798915578752185e-05, | |
| "loss": 0.92333984375, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.43333333333333335, | |
| "grad_norm": 0.043019894510507584, | |
| "learning_rate": 9.790689171104938e-05, | |
| "loss": 0.92919921875, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4388888888888889, | |
| "grad_norm": 0.045795757323503494, | |
| "learning_rate": 9.782301850613163e-05, | |
| "loss": 0.97705078125, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.041059382259845734, | |
| "learning_rate": 9.773753931824839e-05, | |
| "loss": 0.9169921875, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.06831623613834381, | |
| "learning_rate": 9.76504573531083e-05, | |
| "loss": 0.94921875, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.45555555555555555, | |
| "grad_norm": 0.04747457057237625, | |
| "learning_rate": 9.756177587652856e-05, | |
| "loss": 0.9521484375, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.46111111111111114, | |
| "grad_norm": 0.03961906209588051, | |
| "learning_rate": 9.747149821431264e-05, | |
| "loss": 0.8223876953125, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4666666666666667, | |
| "grad_norm": 0.045329079031944275, | |
| "learning_rate": 9.73796277521253e-05, | |
| "loss": 0.8955078125, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.4722222222222222, | |
| "grad_norm": 0.042129456996917725, | |
| "learning_rate": 9.728616793536588e-05, | |
| "loss": 0.8349609375, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.4777777777777778, | |
| "grad_norm": 0.05354605242609978, | |
| "learning_rate": 9.719112226903889e-05, | |
| "loss": 0.951171875, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.48333333333333334, | |
| "grad_norm": 0.05367351323366165, | |
| "learning_rate": 9.709449431762264e-05, | |
| "loss": 1.041015625, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4888888888888889, | |
| "grad_norm": 0.044667478650808334, | |
| "learning_rate": 9.699628770493558e-05, | |
| "loss": 0.99951171875, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.49444444444444446, | |
| "grad_norm": 0.04973745718598366, | |
| "learning_rate": 9.689650611400033e-05, | |
| "loss": 0.93505859375, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.04364899545907974, | |
| "learning_rate": 9.679515328690568e-05, | |
| "loss": 0.912109375, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5055555555555555, | |
| "grad_norm": 0.09642947465181351, | |
| "learning_rate": 9.669223302466608e-05, | |
| "loss": 0.96142578125, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5111111111111111, | |
| "grad_norm": 0.05173720419406891, | |
| "learning_rate": 9.658774918707925e-05, | |
| "loss": 0.97705078125, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5166666666666667, | |
| "grad_norm": 0.048349179327487946, | |
| "learning_rate": 9.648170569258128e-05, | |
| "loss": 1.01806640625, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5222222222222223, | |
| "grad_norm": 0.045038852840662, | |
| "learning_rate": 9.637410651809985e-05, | |
| "loss": 0.91455078125, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5277777777777778, | |
| "grad_norm": 0.037013258785009384, | |
| "learning_rate": 9.626495569890494e-05, | |
| "loss": 0.7244873046875, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.042244914919137955, | |
| "learning_rate": 9.615425732845755e-05, | |
| "loss": 0.865234375, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5388888888888889, | |
| "grad_norm": 0.04271755740046501, | |
| "learning_rate": 9.604201555825618e-05, | |
| "loss": 0.93505859375, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5444444444444444, | |
| "grad_norm": 0.08087927848100662, | |
| "learning_rate": 9.592823459768113e-05, | |
| "loss": 0.93701171875, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.05718201398849487, | |
| "learning_rate": 9.581291871383669e-05, | |
| "loss": 1.0205078125, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.05421556904911995, | |
| "learning_rate": 9.5696072231391e-05, | |
| "loss": 0.9346923828125, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5611111111111111, | |
| "grad_norm": 0.04943264275789261, | |
| "learning_rate": 9.557769953241395e-05, | |
| "loss": 0.9609375, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5666666666666667, | |
| "grad_norm": 0.05051499605178833, | |
| "learning_rate": 9.545780505621287e-05, | |
| "loss": 0.87255859375, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5722222222222222, | |
| "grad_norm": 0.06603266298770905, | |
| "learning_rate": 9.533639329916591e-05, | |
| "loss": 0.93994140625, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5777777777777777, | |
| "grad_norm": 0.05013374984264374, | |
| "learning_rate": 9.521346881455356e-05, | |
| "loss": 0.82525634765625, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 0.05105440691113472, | |
| "learning_rate": 9.508903621238778e-05, | |
| "loss": 0.91796875, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5888888888888889, | |
| "grad_norm": 0.037478625774383545, | |
| "learning_rate": 9.49631001592392e-05, | |
| "loss": 0.70391845703125, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5944444444444444, | |
| "grad_norm": 0.06024964526295662, | |
| "learning_rate": 9.483566537806202e-05, | |
| "loss": 0.908203125, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.05091103911399841, | |
| "learning_rate": 9.470673664801696e-05, | |
| "loss": 0.93701171875, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_loss": 0.88287353515625, | |
| "eval_runtime": 13.1175, | |
| "eval_samples_per_second": 0.915, | |
| "eval_steps_per_second": 0.152, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6055555555555555, | |
| "grad_norm": 0.05728965625166893, | |
| "learning_rate": 9.4576318804292e-05, | |
| "loss": 1.024169921875, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6111111111111112, | |
| "grad_norm": 0.06394322961568832, | |
| "learning_rate": 9.444441673792107e-05, | |
| "loss": 0.9140625, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6166666666666667, | |
| "grad_norm": 0.055423419922590256, | |
| "learning_rate": 9.431103539560054e-05, | |
| "loss": 1.0439453125, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 0.05242814123630524, | |
| "learning_rate": 9.417617977950381e-05, | |
| "loss": 0.95166015625, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6277777777777778, | |
| "grad_norm": 0.04599849134683609, | |
| "learning_rate": 9.403985494709367e-05, | |
| "loss": 0.943359375, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6333333333333333, | |
| "grad_norm": 0.056768402457237244, | |
| "learning_rate": 9.390206601093261e-05, | |
| "loss": 0.9140625, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6388888888888888, | |
| "grad_norm": 0.04664474353194237, | |
| "learning_rate": 9.376281813849113e-05, | |
| "loss": 0.92333984375, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6444444444444445, | |
| "grad_norm": 0.044502440840005875, | |
| "learning_rate": 9.36221165519539e-05, | |
| "loss": 0.9140625, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.05736678093671799, | |
| "learning_rate": 9.347996652802393e-05, | |
| "loss": 0.94384765625, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6555555555555556, | |
| "grad_norm": 0.047704923897981644, | |
| "learning_rate": 9.333637339772472e-05, | |
| "loss": 0.89697265625, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6611111111111111, | |
| "grad_norm": 0.060899727046489716, | |
| "learning_rate": 9.319134254620026e-05, | |
| "loss": 1.03466796875, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.06606944650411606, | |
| "learning_rate": 9.304487941251313e-05, | |
| "loss": 0.869873046875, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6722222222222223, | |
| "grad_norm": 0.05356299504637718, | |
| "learning_rate": 9.289698948944046e-05, | |
| "loss": 0.892578125, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6777777777777778, | |
| "grad_norm": 0.04922427237033844, | |
| "learning_rate": 9.274767832326802e-05, | |
| "loss": 0.912109375, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.6833333333333333, | |
| "grad_norm": 0.06108797341585159, | |
| "learning_rate": 9.259695151358214e-05, | |
| "loss": 0.8837890625, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6888888888888889, | |
| "grad_norm": 0.04262525588274002, | |
| "learning_rate": 9.244481471305976e-05, | |
| "loss": 0.86181640625, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "grad_norm": 0.05409322679042816, | |
| "learning_rate": 9.229127362725639e-05, | |
| "loss": 0.93017578125, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.05612434446811676, | |
| "learning_rate": 9.213633401439221e-05, | |
| "loss": 0.92724609375, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.7055555555555556, | |
| "grad_norm": 0.07815617322921753, | |
| "learning_rate": 9.198000168513604e-05, | |
| "loss": 0.98974609375, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.05235309526324272, | |
| "learning_rate": 9.182228250238747e-05, | |
| "loss": 0.9091796875, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.7166666666666667, | |
| "grad_norm": 0.04183909296989441, | |
| "learning_rate": 9.166318238105697e-05, | |
| "loss": 0.95654296875, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7222222222222222, | |
| "grad_norm": 0.05235957354307175, | |
| "learning_rate": 9.15027072878441e-05, | |
| "loss": 0.91064453125, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7277777777777777, | |
| "grad_norm": 0.0474664680659771, | |
| "learning_rate": 9.134086324101368e-05, | |
| "loss": 0.9013671875, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7333333333333333, | |
| "grad_norm": 0.04867349565029144, | |
| "learning_rate": 9.117765631017011e-05, | |
| "loss": 0.88720703125, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.7388888888888889, | |
| "grad_norm": 0.04376591369509697, | |
| "learning_rate": 9.101309261602979e-05, | |
| "loss": 0.90576171875, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7444444444444445, | |
| "grad_norm": 0.046063948422670364, | |
| "learning_rate": 9.084717833019151e-05, | |
| "loss": 0.9140625, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.04830019548535347, | |
| "learning_rate": 9.067991967490503e-05, | |
| "loss": 0.90380859375, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7555555555555555, | |
| "grad_norm": 0.05522068962454796, | |
| "learning_rate": 9.051132292283771e-05, | |
| "loss": 0.92333984375, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7611111111111111, | |
| "grad_norm": 0.051822349429130554, | |
| "learning_rate": 9.034139439683935e-05, | |
| "loss": 0.896484375, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7666666666666667, | |
| "grad_norm": 0.04480434209108353, | |
| "learning_rate": 9.01701404697049e-05, | |
| "loss": 0.90478515625, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7722222222222223, | |
| "grad_norm": 0.055487968027591705, | |
| "learning_rate": 8.999756756393563e-05, | |
| "loss": 0.85986328125, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.06759867072105408, | |
| "learning_rate": 8.982368215149817e-05, | |
| "loss": 0.9658203125, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7833333333333333, | |
| "grad_norm": 0.04787890240550041, | |
| "learning_rate": 8.964849075358182e-05, | |
| "loss": 0.90576171875, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7888888888888889, | |
| "grad_norm": 0.041783880442380905, | |
| "learning_rate": 8.947199994035401e-05, | |
| "loss": 0.92333984375, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.7944444444444444, | |
| "grad_norm": 0.04689563438296318, | |
| "learning_rate": 8.929421633071386e-05, | |
| "loss": 0.89453125, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.046710215508937836, | |
| "learning_rate": 8.911514659204394e-05, | |
| "loss": 0.9794921875, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.8055555555555556, | |
| "grad_norm": 0.04201524704694748, | |
| "learning_rate": 8.893479743996034e-05, | |
| "loss": 0.88916015625, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.8111111111111111, | |
| "grad_norm": 0.04641687870025635, | |
| "learning_rate": 8.875317563806064e-05, | |
| "loss": 0.890625, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.8166666666666667, | |
| "grad_norm": 0.045671626925468445, | |
| "learning_rate": 8.857028799767042e-05, | |
| "loss": 0.93115234375, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8222222222222222, | |
| "grad_norm": 0.04884997010231018, | |
| "learning_rate": 8.838614137758767e-05, | |
| "loss": 0.9248046875, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8277777777777777, | |
| "grad_norm": 0.051464494317770004, | |
| "learning_rate": 8.820074268382574e-05, | |
| "loss": 0.953125, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 0.04283425956964493, | |
| "learning_rate": 8.801409886935412e-05, | |
| "loss": 0.79766845703125, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8388888888888889, | |
| "grad_norm": 0.05242162570357323, | |
| "learning_rate": 8.782621693383794e-05, | |
| "loss": 0.923828125, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8444444444444444, | |
| "grad_norm": 0.04432463273406029, | |
| "learning_rate": 8.763710392337522e-05, | |
| "loss": 0.892578125, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.05258423089981079, | |
| "learning_rate": 8.744676693023282e-05, | |
| "loss": 0.89697265625, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.8555555555555555, | |
| "grad_norm": 0.03943394124507904, | |
| "learning_rate": 8.725521309258031e-05, | |
| "loss": 0.76123046875, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.8611111111111112, | |
| "grad_norm": 0.16495642066001892, | |
| "learning_rate": 8.706244959422237e-05, | |
| "loss": 0.9586181640625, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8666666666666667, | |
| "grad_norm": 0.05052783340215683, | |
| "learning_rate": 8.686848366432931e-05, | |
| "loss": 0.90380859375, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8722222222222222, | |
| "grad_norm": 0.04332755506038666, | |
| "learning_rate": 8.667332257716603e-05, | |
| "loss": 0.8125, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.8777777777777778, | |
| "grad_norm": 0.060963306576013565, | |
| "learning_rate": 8.647697365181911e-05, | |
| "loss": 0.935546875, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.8833333333333333, | |
| "grad_norm": 0.049565620720386505, | |
| "learning_rate": 8.627944425192244e-05, | |
| "loss": 0.9248046875, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.04879913479089737, | |
| "learning_rate": 8.6080741785381e-05, | |
| "loss": 0.94970703125, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8944444444444445, | |
| "grad_norm": 0.05664806813001633, | |
| "learning_rate": 8.588087370409303e-05, | |
| "loss": 0.8671875, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.04701758548617363, | |
| "learning_rate": 8.567984750367055e-05, | |
| "loss": 0.89892578125, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.9055555555555556, | |
| "grad_norm": 0.05865476280450821, | |
| "learning_rate": 8.547767072315835e-05, | |
| "loss": 0.89599609375, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.9111111111111111, | |
| "grad_norm": 0.052406009286642075, | |
| "learning_rate": 8.527435094475118e-05, | |
| "loss": 0.94921875, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.9166666666666666, | |
| "grad_norm": 0.04421606659889221, | |
| "learning_rate": 8.506989579350938e-05, | |
| "loss": 0.84521484375, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.9222222222222223, | |
| "grad_norm": 0.05002782866358757, | |
| "learning_rate": 8.486431293707296e-05, | |
| "loss": 0.8828125, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9277777777777778, | |
| "grad_norm": 0.042441803961992264, | |
| "learning_rate": 8.465761008537405e-05, | |
| "loss": 0.861328125, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "grad_norm": 0.05188900604844093, | |
| "learning_rate": 8.44497949903477e-05, | |
| "loss": 0.9039306640625, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9388888888888889, | |
| "grad_norm": 0.04285572096705437, | |
| "learning_rate": 8.42408754456412e-05, | |
| "loss": 0.7958984375, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.9444444444444444, | |
| "grad_norm": 0.05040177330374718, | |
| "learning_rate": 8.403085928632179e-05, | |
| "loss": 0.87890625, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.07656143605709076, | |
| "learning_rate": 8.38197543885829e-05, | |
| "loss": 0.83837890625, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9555555555555556, | |
| "grad_norm": 0.0454406775534153, | |
| "learning_rate": 8.360756866944858e-05, | |
| "loss": 0.92236328125, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9611111111111111, | |
| "grad_norm": 0.0487142838537693, | |
| "learning_rate": 8.339431008647677e-05, | |
| "loss": 0.8642578125, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9666666666666667, | |
| "grad_norm": 0.04474879056215286, | |
| "learning_rate": 8.317998663746079e-05, | |
| "loss": 0.916015625, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.9722222222222222, | |
| "grad_norm": 0.04770045354962349, | |
| "learning_rate": 8.296460636012941e-05, | |
| "loss": 0.9013671875, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 0.09583094716072083, | |
| "learning_rate": 8.274817733184548e-05, | |
| "loss": 0.85546875, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.9833333333333333, | |
| "grad_norm": 0.04274609684944153, | |
| "learning_rate": 8.253070766930282e-05, | |
| "loss": 0.94189453125, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.9888888888888889, | |
| "grad_norm": 0.16147056221961975, | |
| "learning_rate": 8.231220552822205e-05, | |
| "loss": 0.9150390625, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.9944444444444445, | |
| "grad_norm": 0.04417458921670914, | |
| "learning_rate": 8.209267910304462e-05, | |
| "loss": 0.82666015625, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.04557017609477043, | |
| "learning_rate": 8.187213662662538e-05, | |
| "loss": 0.8681640625, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.0055555555555555, | |
| "grad_norm": 0.20462441444396973, | |
| "learning_rate": 8.165058636992411e-05, | |
| "loss": 0.8193359375, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.011111111111111, | |
| "grad_norm": 0.043176133185625076, | |
| "learning_rate": 8.142803664169506e-05, | |
| "loss": 0.734344482421875, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.0166666666666666, | |
| "grad_norm": 0.05020572990179062, | |
| "learning_rate": 8.120449578817548e-05, | |
| "loss": 0.90966796875, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.0222222222222221, | |
| "grad_norm": 0.04717118665575981, | |
| "learning_rate": 8.09799721927726e-05, | |
| "loss": 0.8837890625, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.0277777777777777, | |
| "grad_norm": 0.0437992699444294, | |
| "learning_rate": 8.075447427574924e-05, | |
| "loss": 0.89697265625, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.0333333333333334, | |
| "grad_norm": 0.045845769345760345, | |
| "learning_rate": 8.052801049390793e-05, | |
| "loss": 0.89013671875, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.038888888888889, | |
| "grad_norm": 0.0499701164662838, | |
| "learning_rate": 8.030058934027394e-05, | |
| "loss": 0.90576171875, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.0444444444444445, | |
| "grad_norm": 0.042997606098651886, | |
| "learning_rate": 8.007221934377658e-05, | |
| "loss": 0.8232421875, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.046552494168281555, | |
| "learning_rate": 7.984290906892945e-05, | |
| "loss": 0.9130859375, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.0555555555555556, | |
| "grad_norm": 0.045746274292469025, | |
| "learning_rate": 7.961266711550922e-05, | |
| "loss": 0.76617431640625, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0611111111111111, | |
| "grad_norm": 0.04799175634980202, | |
| "learning_rate": 7.938150211823309e-05, | |
| "loss": 0.85595703125, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.05563824251294136, | |
| "learning_rate": 7.914942274643501e-05, | |
| "loss": 0.94873046875, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.0722222222222222, | |
| "grad_norm": 0.04851698875427246, | |
| "learning_rate": 7.891643770374052e-05, | |
| "loss": 0.8994140625, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.0777777777777777, | |
| "grad_norm": 0.04606899619102478, | |
| "learning_rate": 7.868255572774037e-05, | |
| "loss": 0.7825927734375, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.0833333333333333, | |
| "grad_norm": 0.04620612412691116, | |
| "learning_rate": 7.84477855896628e-05, | |
| "loss": 0.86279296875, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.0888888888888888, | |
| "grad_norm": 0.044590290635824203, | |
| "learning_rate": 7.821213609404459e-05, | |
| "loss": 0.94677734375, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.0944444444444446, | |
| "grad_norm": 0.04122447967529297, | |
| "learning_rate": 7.797561607840094e-05, | |
| "loss": 0.83642578125, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.04792390763759613, | |
| "learning_rate": 7.773823441289396e-05, | |
| "loss": 0.8154296875, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.1055555555555556, | |
| "grad_norm": 0.043234195560216904, | |
| "learning_rate": 7.75e-05, | |
| "loss": 0.92724609375, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.05087165907025337, | |
| "learning_rate": 7.726092177417592e-05, | |
| "loss": 0.91064453125, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.1166666666666667, | |
| "grad_norm": 0.043311040848493576, | |
| "learning_rate": 7.702100870152383e-05, | |
| "loss": 0.77734375, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.1222222222222222, | |
| "grad_norm": 0.04584471881389618, | |
| "learning_rate": 7.678026977945497e-05, | |
| "loss": 0.82177734375, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.1277777777777778, | |
| "grad_norm": 0.04674329236149788, | |
| "learning_rate": 7.653871403635223e-05, | |
| "loss": 0.97119140625, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.1333333333333333, | |
| "grad_norm": 0.049187254160642624, | |
| "learning_rate": 7.629635053123158e-05, | |
| "loss": 0.86376953125, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.1388888888888888, | |
| "grad_norm": 0.04960101097822189, | |
| "learning_rate": 7.605318835340236e-05, | |
| "loss": 0.91650390625, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.1444444444444444, | |
| "grad_norm": 0.05421186983585358, | |
| "learning_rate": 7.580923662212632e-05, | |
| "loss": 0.714599609375, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.04953968524932861, | |
| "learning_rate": 7.556450448627568e-05, | |
| "loss": 0.8740234375, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.1555555555555554, | |
| "grad_norm": 0.04564543813467026, | |
| "learning_rate": 7.531900112399004e-05, | |
| "loss": 0.8359375, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.1611111111111112, | |
| "grad_norm": 0.048400409519672394, | |
| "learning_rate": 7.50727357423321e-05, | |
| "loss": 0.80712890625, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 0.04787645861506462, | |
| "learning_rate": 7.48257175769425e-05, | |
| "loss": 0.8564453125, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1722222222222223, | |
| "grad_norm": 0.057255957275629044, | |
| "learning_rate": 7.45779558916933e-05, | |
| "loss": 0.91064453125, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.1777777777777778, | |
| "grad_norm": 0.07384311407804489, | |
| "learning_rate": 7.432945997834067e-05, | |
| "loss": 0.716552734375, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.1833333333333333, | |
| "grad_norm": 0.04651561751961708, | |
| "learning_rate": 7.408023915617636e-05, | |
| "loss": 0.927734375, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.1888888888888889, | |
| "grad_norm": 0.04706474393606186, | |
| "learning_rate": 7.383030277167826e-05, | |
| "loss": 0.85546875, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.1944444444444444, | |
| "grad_norm": 0.04492956027388573, | |
| "learning_rate": 7.357966019815982e-05, | |
| "loss": 0.8994140625, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.04066196456551552, | |
| "learning_rate": 7.332832083541859e-05, | |
| "loss": 0.87353515625, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.8423919677734375, | |
| "eval_runtime": 13.1063, | |
| "eval_samples_per_second": 0.916, | |
| "eval_steps_per_second": 0.153, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.2055555555555555, | |
| "grad_norm": 0.04753434285521507, | |
| "learning_rate": 7.307629410938363e-05, | |
| "loss": 0.8076171875, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.211111111111111, | |
| "grad_norm": 0.05118556693196297, | |
| "learning_rate": 7.282358947176207e-05, | |
| "loss": 0.83935546875, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.2166666666666668, | |
| "grad_norm": 0.04904546961188316, | |
| "learning_rate": 7.25702163996846e-05, | |
| "loss": 0.86083984375, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.05265267565846443, | |
| "learning_rate": 7.231618439535007e-05, | |
| "loss": 0.80859375, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.2277777777777779, | |
| "grad_norm": 0.04758688434958458, | |
| "learning_rate": 7.206150298566915e-05, | |
| "loss": 0.717529296875, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.2333333333333334, | |
| "grad_norm": 0.05221811681985855, | |
| "learning_rate": 7.180618172190702e-05, | |
| "loss": 0.8048095703125, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.238888888888889, | |
| "grad_norm": 0.05886903032660484, | |
| "learning_rate": 7.15502301793252e-05, | |
| "loss": 0.8515625, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.2444444444444445, | |
| "grad_norm": 0.04534082114696503, | |
| "learning_rate": 7.129365795682238e-05, | |
| "loss": 0.8076171875, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.06249246373772621, | |
| "learning_rate": 7.103647467657452e-05, | |
| "loss": 0.85302734375, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.2555555555555555, | |
| "grad_norm": 0.04579539969563484, | |
| "learning_rate": 7.077868998367395e-05, | |
| "loss": 0.83544921875, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.261111111111111, | |
| "grad_norm": 0.04565001651644707, | |
| "learning_rate": 7.052031354576765e-05, | |
| "loss": 0.865234375, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.2666666666666666, | |
| "grad_norm": 0.05467552691698074, | |
| "learning_rate": 7.026135505269467e-05, | |
| "loss": 0.7939453125, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.2722222222222221, | |
| "grad_norm": 0.045561108738183975, | |
| "learning_rate": 7.000182421612281e-05, | |
| "loss": 0.79345703125, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.2777777777777777, | |
| "grad_norm": 0.045619670301675797, | |
| "learning_rate": 6.97417307691843e-05, | |
| "loss": 0.796875, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.2833333333333332, | |
| "grad_norm": 0.05211074277758598, | |
| "learning_rate": 6.948108446611086e-05, | |
| "loss": 0.82958984375, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.2888888888888888, | |
| "grad_norm": 0.048779115080833435, | |
| "learning_rate": 6.921989508186787e-05, | |
| "loss": 0.830078125, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.2944444444444445, | |
| "grad_norm": 0.046727199107408524, | |
| "learning_rate": 6.895817241178774e-05, | |
| "loss": 0.83349609375, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.04842211678624153, | |
| "learning_rate": 6.869592627120262e-05, | |
| "loss": 0.8369140625, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.3055555555555556, | |
| "grad_norm": 0.04721565917134285, | |
| "learning_rate": 6.843316649507626e-05, | |
| "loss": 0.79443359375, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.3111111111111111, | |
| "grad_norm": 0.09705671668052673, | |
| "learning_rate": 6.816990293763516e-05, | |
| "loss": 0.83984375, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.3166666666666667, | |
| "grad_norm": 0.05335555598139763, | |
| "learning_rate": 6.790614547199907e-05, | |
| "loss": 0.76708984375, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.3222222222222222, | |
| "grad_norm": 0.042442601174116135, | |
| "learning_rate": 6.764190398981063e-05, | |
| "loss": 0.818359375, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.3277777777777777, | |
| "grad_norm": 0.0599215030670166, | |
| "learning_rate": 6.737718840086448e-05, | |
| "loss": 0.8017578125, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.049463603645563126, | |
| "learning_rate": 6.71120086327356e-05, | |
| "loss": 0.78369140625, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.338888888888889, | |
| "grad_norm": 0.047156400978565216, | |
| "learning_rate": 6.684637463040697e-05, | |
| "loss": 0.8037109375, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.3444444444444446, | |
| "grad_norm": 0.055872924625873566, | |
| "learning_rate": 6.658029635589663e-05, | |
| "loss": 0.8681640625, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.050482094287872314, | |
| "learning_rate": 6.631378378788406e-05, | |
| "loss": 0.8076171875, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.3555555555555556, | |
| "grad_norm": 0.04669024422764778, | |
| "learning_rate": 6.604684692133597e-05, | |
| "loss": 0.7978515625, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.3611111111111112, | |
| "grad_norm": 0.055310290306806564, | |
| "learning_rate": 6.577949576713144e-05, | |
| "loss": 0.75439453125, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.3666666666666667, | |
| "grad_norm": 0.058813516050577164, | |
| "learning_rate": 6.551174035168655e-05, | |
| "loss": 0.72607421875, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.3722222222222222, | |
| "grad_norm": 0.05176336318254471, | |
| "learning_rate": 6.524359071657822e-05, | |
| "loss": 0.7578125, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.3777777777777778, | |
| "grad_norm": 0.054734766483306885, | |
| "learning_rate": 6.497505691816774e-05, | |
| "loss": 0.828125, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.3833333333333333, | |
| "grad_norm": 0.047450918704271317, | |
| "learning_rate": 6.470614902722367e-05, | |
| "loss": 0.79345703125, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 0.05016674846410751, | |
| "learning_rate": 6.443687712854397e-05, | |
| "loss": 0.80029296875, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3944444444444444, | |
| "grad_norm": 0.05473002418875694, | |
| "learning_rate": 6.4167251320578e-05, | |
| "loss": 0.8212890625, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.054414138197898865, | |
| "learning_rate": 6.389728171504765e-05, | |
| "loss": 0.8125, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.4055555555555554, | |
| "grad_norm": 0.04979150742292404, | |
| "learning_rate": 6.362697843656823e-05, | |
| "loss": 0.7333984375, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.411111111111111, | |
| "grad_norm": 0.05243462696671486, | |
| "learning_rate": 6.335635162226869e-05, | |
| "loss": 0.80029296875, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.4166666666666667, | |
| "grad_norm": 0.05495977774262428, | |
| "learning_rate": 6.308541142141146e-05, | |
| "loss": 0.77099609375, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.05158250778913498, | |
| "learning_rate": 6.281416799501188e-05, | |
| "loss": 0.7744140625, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.4277777777777778, | |
| "grad_norm": 0.05417507141828537, | |
| "learning_rate": 6.254263151545706e-05, | |
| "loss": 0.780029296875, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.4333333333333333, | |
| "grad_norm": 0.05407530814409256, | |
| "learning_rate": 6.227081216612447e-05, | |
| "loss": 0.7822265625, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.4388888888888889, | |
| "grad_norm": 0.05271197482943535, | |
| "learning_rate": 6.199872014099994e-05, | |
| "loss": 0.82421875, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.051197804510593414, | |
| "learning_rate": 6.172636564429545e-05, | |
| "loss": 0.7763671875, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.0567997470498085, | |
| "learning_rate": 6.145375889006634e-05, | |
| "loss": 0.7958984375, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.4555555555555555, | |
| "grad_norm": 0.05634378641843796, | |
| "learning_rate": 6.118091010182837e-05, | |
| "loss": 0.79833984375, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.4611111111111112, | |
| "grad_norm": 0.04826923832297325, | |
| "learning_rate": 6.090782951217422e-05, | |
| "loss": 0.69891357421875, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.4666666666666668, | |
| "grad_norm": 0.30020812153816223, | |
| "learning_rate": 6.063452736238979e-05, | |
| "loss": 0.76318359375, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.4722222222222223, | |
| "grad_norm": 0.05334388464689255, | |
| "learning_rate": 6.036101390207006e-05, | |
| "loss": 0.7041015625, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.4777777777777779, | |
| "grad_norm": 0.057358499616384506, | |
| "learning_rate": 6.008729938873484e-05, | |
| "loss": 0.794189453125, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.4833333333333334, | |
| "grad_norm": 0.057268984615802765, | |
| "learning_rate": 5.9813394087443876e-05, | |
| "loss": 0.8701171875, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.488888888888889, | |
| "grad_norm": 0.054010648280382156, | |
| "learning_rate": 5.9539308270412086e-05, | |
| "loss": 0.8447265625, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.4944444444444445, | |
| "grad_norm": 0.05055909603834152, | |
| "learning_rate": 5.926505221662416e-05, | |
| "loss": 0.79541015625, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.054270874708890915, | |
| "learning_rate": 5.899063621144922e-05, | |
| "loss": 0.7734375, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.5055555555555555, | |
| "grad_norm": 0.05884323641657829, | |
| "learning_rate": 5.8716070546254966e-05, | |
| "loss": 0.806640625, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.511111111111111, | |
| "grad_norm": 0.05991430953145027, | |
| "learning_rate": 5.844136551802175e-05, | |
| "loss": 0.82177734375, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.5166666666666666, | |
| "grad_norm": 0.055026423186063766, | |
| "learning_rate": 5.816653142895648e-05, | |
| "loss": 0.8759765625, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.5222222222222221, | |
| "grad_norm": 0.05329109728336334, | |
| "learning_rate": 5.789157858610617e-05, | |
| "loss": 0.794921875, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.5277777777777777, | |
| "grad_norm": 0.04542877897620201, | |
| "learning_rate": 5.761651730097142e-05, | |
| "loss": 0.62225341796875, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.5333333333333332, | |
| "grad_norm": 0.05594032257795334, | |
| "learning_rate": 5.734135788911976e-05, | |
| "loss": 0.7392578125, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.5388888888888888, | |
| "grad_norm": 0.04947767034173012, | |
| "learning_rate": 5.7066110669798736e-05, | |
| "loss": 0.8154296875, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.5444444444444443, | |
| "grad_norm": 0.05655859783291817, | |
| "learning_rate": 5.67907859655489e-05, | |
| "loss": 0.79541015625, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.06127001345157623, | |
| "learning_rate": 5.651539410181671e-05, | |
| "loss": 0.8603515625, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.06107491999864578, | |
| "learning_rate": 5.62399454065673e-05, | |
| "loss": 0.78485107421875, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.5611111111111111, | |
| "grad_norm": 0.06380610913038254, | |
| "learning_rate": 5.596445020989715e-05, | |
| "loss": 0.81201171875, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.5666666666666667, | |
| "grad_norm": 0.05685281381011009, | |
| "learning_rate": 5.5688918843646656e-05, | |
| "loss": 0.74072265625, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.5722222222222222, | |
| "grad_norm": 0.06439924985170364, | |
| "learning_rate": 5.5413361641012695e-05, | |
| "loss": 0.7794189453125, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.5777777777777777, | |
| "grad_norm": 0.0554894395172596, | |
| "learning_rate": 5.513778893616106e-05, | |
| "loss": 0.700164794921875, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.5833333333333335, | |
| "grad_norm": 0.05881272628903389, | |
| "learning_rate": 5.486221106383895e-05, | |
| "loss": 0.7861328125, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.588888888888889, | |
| "grad_norm": 0.05062687397003174, | |
| "learning_rate": 5.4586638358987316e-05, | |
| "loss": 0.604644775390625, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.5944444444444446, | |
| "grad_norm": 0.05964241549372673, | |
| "learning_rate": 5.431108115635336e-05, | |
| "loss": 0.7646484375, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.06189911440014839, | |
| "learning_rate": 5.403554979010287e-05, | |
| "loss": 0.7939453125, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.6055555555555556, | |
| "grad_norm": 0.05683770030736923, | |
| "learning_rate": 5.376005459343272e-05, | |
| "loss": 0.8809814453125, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.6111111111111112, | |
| "grad_norm": 0.05130205303430557, | |
| "learning_rate": 5.3484605898183315e-05, | |
| "loss": 0.7978515625, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.6166666666666667, | |
| "grad_norm": 0.0612524151802063, | |
| "learning_rate": 5.320921403445112e-05, | |
| "loss": 0.91015625, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.6222222222222222, | |
| "grad_norm": 0.061944060027599335, | |
| "learning_rate": 5.293388933020127e-05, | |
| "loss": 0.81396484375, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.6277777777777778, | |
| "grad_norm": 0.05926624685525894, | |
| "learning_rate": 5.2658642110880244e-05, | |
| "loss": 0.80810546875, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.6333333333333333, | |
| "grad_norm": 0.06159135326743126, | |
| "learning_rate": 5.23834826990286e-05, | |
| "loss": 0.7763671875, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.6388888888888888, | |
| "grad_norm": 0.05656843259930611, | |
| "learning_rate": 5.210842141389385e-05, | |
| "loss": 0.79345703125, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.6444444444444444, | |
| "grad_norm": 0.05543031916022301, | |
| "learning_rate": 5.183346857104353e-05, | |
| "loss": 0.79150390625, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.06932569295167923, | |
| "learning_rate": 5.1558634481978265e-05, | |
| "loss": 0.80419921875, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.6555555555555554, | |
| "grad_norm": 0.06103169173002243, | |
| "learning_rate": 5.128392945374505e-05, | |
| "loss": 0.771484375, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.661111111111111, | |
| "grad_norm": 0.06873636692762375, | |
| "learning_rate": 5.100936378855079e-05, | |
| "loss": 0.8681640625, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.059117577970027924, | |
| "learning_rate": 5.073494778337584e-05, | |
| "loss": 0.742919921875, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6722222222222223, | |
| "grad_norm": 0.06298115104436874, | |
| "learning_rate": 5.046069172958794e-05, | |
| "loss": 0.7802734375, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.6777777777777778, | |
| "grad_norm": 0.0591348372399807, | |
| "learning_rate": 5.018660591255614e-05, | |
| "loss": 0.787109375, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.6833333333333333, | |
| "grad_norm": 0.059101544320583344, | |
| "learning_rate": 4.9912700611265176e-05, | |
| "loss": 0.7548828125, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.6888888888888889, | |
| "grad_norm": 0.056728918105363846, | |
| "learning_rate": 4.963898609792994e-05, | |
| "loss": 0.75146484375, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.6944444444444444, | |
| "grad_norm": 0.06630321592092514, | |
| "learning_rate": 4.9365472637610235e-05, | |
| "loss": 0.7900390625, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.13757576048374176, | |
| "learning_rate": 4.909217048782579e-05, | |
| "loss": 0.79443359375, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.7055555555555557, | |
| "grad_norm": 0.06553367525339127, | |
| "learning_rate": 4.881908989817163e-05, | |
| "loss": 0.840576171875, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.7111111111111112, | |
| "grad_norm": 0.0583001971244812, | |
| "learning_rate": 4.8546241109933674e-05, | |
| "loss": 0.7802734375, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.7166666666666668, | |
| "grad_norm": 0.052199836820364, | |
| "learning_rate": 4.8273634355704565e-05, | |
| "loss": 0.84423828125, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.7222222222222223, | |
| "grad_norm": 0.06273237615823746, | |
| "learning_rate": 4.800127985900007e-05, | |
| "loss": 0.779296875, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.7277777777777779, | |
| "grad_norm": 0.05880773812532425, | |
| "learning_rate": 4.7729187833875546e-05, | |
| "loss": 0.7734375, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.7333333333333334, | |
| "grad_norm": 0.058469612151384354, | |
| "learning_rate": 4.745736848454295e-05, | |
| "loss": 0.759765625, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.738888888888889, | |
| "grad_norm": 0.0585034117102623, | |
| "learning_rate": 4.718583200498814e-05, | |
| "loss": 0.7783203125, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.7444444444444445, | |
| "grad_norm": 0.05488097667694092, | |
| "learning_rate": 4.691458857858857e-05, | |
| "loss": 0.7880859375, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.05396897345781326, | |
| "learning_rate": 4.664364837773133e-05, | |
| "loss": 0.77197265625, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.7555555555555555, | |
| "grad_norm": 0.3896213173866272, | |
| "learning_rate": 4.6373021563431784e-05, | |
| "loss": 0.780029296875, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.761111111111111, | |
| "grad_norm": 0.06683356314897537, | |
| "learning_rate": 4.610271828495235e-05, | |
| "loss": 0.7724609375, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.7666666666666666, | |
| "grad_norm": 0.05703408643603325, | |
| "learning_rate": 4.5832748679422024e-05, | |
| "loss": 0.781982421875, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.7722222222222221, | |
| "grad_norm": 0.06155144423246384, | |
| "learning_rate": 4.556312287145604e-05, | |
| "loss": 0.7412109375, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.05332708731293678, | |
| "learning_rate": 4.5293850972776364e-05, | |
| "loss": 0.84814453125, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.7833333333333332, | |
| "grad_norm": 0.06812047213315964, | |
| "learning_rate": 4.502494308183226e-05, | |
| "loss": 0.77392578125, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.7888888888888888, | |
| "grad_norm": 0.05285648629069328, | |
| "learning_rate": 4.475640928342182e-05, | |
| "loss": 0.8115234375, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.7944444444444443, | |
| "grad_norm": 0.055206868797540665, | |
| "learning_rate": 4.448825964831347e-05, | |
| "loss": 0.771484375, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.05760844796895981, | |
| "learning_rate": 4.422050423286857e-05, | |
| "loss": 0.8447265625, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "eval_loss": 0.8323516845703125, | |
| "eval_runtime": 13.0952, | |
| "eval_samples_per_second": 0.916, | |
| "eval_steps_per_second": 0.153, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.8055555555555556, | |
| "grad_norm": 0.06074342131614685, | |
| "learning_rate": 4.395315307866405e-05, | |
| "loss": 0.76904296875, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.8111111111111111, | |
| "grad_norm": 0.05502430722117424, | |
| "learning_rate": 4.368621621211596e-05, | |
| "loss": 0.77294921875, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.8166666666666667, | |
| "grad_norm": 0.05855901539325714, | |
| "learning_rate": 4.3419703644103384e-05, | |
| "loss": 0.7998046875, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.8222222222222222, | |
| "grad_norm": 0.057750679552555084, | |
| "learning_rate": 4.3153625369593057e-05, | |
| "loss": 0.80322265625, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.8277777777777777, | |
| "grad_norm": 0.05659996718168259, | |
| "learning_rate": 4.28879913672644e-05, | |
| "loss": 0.82666015625, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 0.05274190381169319, | |
| "learning_rate": 4.262281159913554e-05, | |
| "loss": 0.692291259765625, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.838888888888889, | |
| "grad_norm": 0.06606703996658325, | |
| "learning_rate": 4.2358096010189385e-05, | |
| "loss": 0.79736328125, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.8444444444444446, | |
| "grad_norm": 0.05838339403271675, | |
| "learning_rate": 4.209385452800095e-05, | |
| "loss": 0.77685546875, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.059201326221227646, | |
| "learning_rate": 4.183009706236484e-05, | |
| "loss": 0.77685546875, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.8555555555555556, | |
| "grad_norm": 0.046663038432598114, | |
| "learning_rate": 4.156683350492376e-05, | |
| "loss": 0.676513671875, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.8611111111111112, | |
| "grad_norm": 0.054537270218133926, | |
| "learning_rate": 4.130407372879739e-05, | |
| "loss": 0.83868408203125, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 0.06385454535484314, | |
| "learning_rate": 4.104182758821228e-05, | |
| "loss": 0.786865234375, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.8722222222222222, | |
| "grad_norm": 0.052061136811971664, | |
| "learning_rate": 4.0780104918132144e-05, | |
| "loss": 0.71337890625, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.8777777777777778, | |
| "grad_norm": 0.05896454304456711, | |
| "learning_rate": 4.051891553388916e-05, | |
| "loss": 0.8154296875, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.8833333333333333, | |
| "grad_norm": 0.05938766151666641, | |
| "learning_rate": 4.0258269230815715e-05, | |
| "loss": 0.80322265625, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.05951433256268501, | |
| "learning_rate": 3.9998175783877226e-05, | |
| "loss": 0.8251953125, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.8944444444444444, | |
| "grad_norm": 0.06209597736597061, | |
| "learning_rate": 3.973864494730534e-05, | |
| "loss": 0.74365234375, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.0583549439907074, | |
| "learning_rate": 3.947968645423238e-05, | |
| "loss": 0.78173828125, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.9055555555555554, | |
| "grad_norm": 0.06311269849538803, | |
| "learning_rate": 3.922131001632606e-05, | |
| "loss": 0.76318359375, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.911111111111111, | |
| "grad_norm": 0.060178518295288086, | |
| "learning_rate": 3.8963525323425486e-05, | |
| "loss": 0.82763671875, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.9166666666666665, | |
| "grad_norm": 0.05753462389111519, | |
| "learning_rate": 3.870634204317762e-05, | |
| "loss": 0.73492431640625, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.9222222222222223, | |
| "grad_norm": 0.11456094682216644, | |
| "learning_rate": 3.844976982067482e-05, | |
| "loss": 0.7685546875, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.9277777777777778, | |
| "grad_norm": 0.05636809393763542, | |
| "learning_rate": 3.819381827809299e-05, | |
| "loss": 0.75146484375, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.9333333333333333, | |
| "grad_norm": 0.07920922338962555, | |
| "learning_rate": 3.793849701433088e-05, | |
| "loss": 0.7772216796875, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.9388888888888889, | |
| "grad_norm": 0.054340459406375885, | |
| "learning_rate": 3.768381560464995e-05, | |
| "loss": 0.708251953125, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.9444444444444444, | |
| "grad_norm": 0.056318506598472595, | |
| "learning_rate": 3.742978360031544e-05, | |
| "loss": 0.76904296875, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.0579373836517334, | |
| "learning_rate": 3.717641052823795e-05, | |
| "loss": 0.7314453125, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.9555555555555557, | |
| "grad_norm": 0.056404028087854385, | |
| "learning_rate": 3.692370589061639e-05, | |
| "loss": 0.81103515625, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.9611111111111112, | |
| "grad_norm": 0.05630074441432953, | |
| "learning_rate": 3.667167916458142e-05, | |
| "loss": 0.75537109375, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.9666666666666668, | |
| "grad_norm": 0.06106020510196686, | |
| "learning_rate": 3.6420339801840196e-05, | |
| "loss": 0.79736328125, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.9722222222222223, | |
| "grad_norm": 0.057445403188467026, | |
| "learning_rate": 3.616969722832174e-05, | |
| "loss": 0.78369140625, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.9777777777777779, | |
| "grad_norm": 0.06004713103175163, | |
| "learning_rate": 3.5919760843823667e-05, | |
| "loss": 0.74853515625, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.9833333333333334, | |
| "grad_norm": 0.05260277912020683, | |
| "learning_rate": 3.567054002165936e-05, | |
| "loss": 0.84716796875, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.988888888888889, | |
| "grad_norm": 0.05810333415865898, | |
| "learning_rate": 3.5422044108306716e-05, | |
| "loss": 0.79736328125, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.9944444444444445, | |
| "grad_norm": 0.052693530917167664, | |
| "learning_rate": 3.5174282423057514e-05, | |
| "loss": 0.72314453125, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.05832812190055847, | |
| "learning_rate": 3.492726425766791e-05, | |
| "loss": 0.71728515625, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.0055555555555555, | |
| "grad_norm": 0.06229420378804207, | |
| "learning_rate": 3.468099887600999e-05, | |
| "loss": 0.6767578125, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.011111111111111, | |
| "grad_norm": 0.054630160331726074, | |
| "learning_rate": 3.443549551372434e-05, | |
| "loss": 0.6410369873046875, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.0166666666666666, | |
| "grad_norm": 0.0695592612028122, | |
| "learning_rate": 3.419076337787371e-05, | |
| "loss": 0.79150390625, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.022222222222222, | |
| "grad_norm": 0.06250900775194168, | |
| "learning_rate": 3.394681164659766e-05, | |
| "loss": 0.76953125, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.0277777777777777, | |
| "grad_norm": 0.05870450660586357, | |
| "learning_rate": 3.370364946876844e-05, | |
| "loss": 0.78466796875, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.033333333333333, | |
| "grad_norm": 0.05831899493932724, | |
| "learning_rate": 3.3461285963647805e-05, | |
| "loss": 0.7734375, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.0388888888888888, | |
| "grad_norm": 0.06155839189887047, | |
| "learning_rate": 3.3219730220545064e-05, | |
| "loss": 0.78564453125, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.0444444444444443, | |
| "grad_norm": 0.05363038554787636, | |
| "learning_rate": 3.29789912984762e-05, | |
| "loss": 0.7236328125, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.059053950011730194, | |
| "learning_rate": 3.2739078225824106e-05, | |
| "loss": 0.8056640625, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.0555555555555554, | |
| "grad_norm": 0.058264587074518204, | |
| "learning_rate": 3.250000000000001e-05, | |
| "loss": 0.658477783203125, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.061111111111111, | |
| "grad_norm": 0.053816840052604675, | |
| "learning_rate": 3.2261765587106075e-05, | |
| "loss": 0.75, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.066666666666667, | |
| "grad_norm": 0.05713905394077301, | |
| "learning_rate": 3.2024383921599085e-05, | |
| "loss": 0.8369140625, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.0722222222222224, | |
| "grad_norm": 0.05879650637507439, | |
| "learning_rate": 3.178786390595544e-05, | |
| "loss": 0.774658203125, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.077777777777778, | |
| "grad_norm": 0.06611868739128113, | |
| "learning_rate": 3.1552214410337235e-05, | |
| "loss": 0.6802978515625, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.0833333333333335, | |
| "grad_norm": 0.05817348510026932, | |
| "learning_rate": 3.131744427225965e-05, | |
| "loss": 0.7509765625, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.088888888888889, | |
| "grad_norm": 0.06522732228040695, | |
| "learning_rate": 3.1083562296259495e-05, | |
| "loss": 0.8359375, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.0944444444444446, | |
| "grad_norm": 0.05532973259687424, | |
| "learning_rate": 3.0850577253565014e-05, | |
| "loss": 0.73681640625, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.05453307926654816, | |
| "learning_rate": 3.061849788176693e-05, | |
| "loss": 0.71240234375, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.1055555555555556, | |
| "grad_norm": 0.05531802773475647, | |
| "learning_rate": 3.0387332884490805e-05, | |
| "loss": 0.8291015625, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.111111111111111, | |
| "grad_norm": 0.06268377602100372, | |
| "learning_rate": 3.0157090931070565e-05, | |
| "loss": 0.78564453125, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.1166666666666667, | |
| "grad_norm": 0.05352954939007759, | |
| "learning_rate": 2.9927780656223438e-05, | |
| "loss": 0.681640625, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.1222222222222222, | |
| "grad_norm": 0.056907057762145996, | |
| "learning_rate": 2.9699410659726075e-05, | |
| "loss": 0.715087890625, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.1277777777777778, | |
| "grad_norm": 0.056129924952983856, | |
| "learning_rate": 2.947198950609208e-05, | |
| "loss": 0.85986328125, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.061477258801460266, | |
| "learning_rate": 2.9245525724250793e-05, | |
| "loss": 0.73828125, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.138888888888889, | |
| "grad_norm": 0.06201855093240738, | |
| "learning_rate": 2.9020027807227424e-05, | |
| "loss": 0.79541015625, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.1444444444444444, | |
| "grad_norm": 0.06749982386827469, | |
| "learning_rate": 2.8795504211824543e-05, | |
| "loss": 0.6243896484375, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 0.06140488386154175, | |
| "learning_rate": 2.8571963358304965e-05, | |
| "loss": 0.748046875, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.1555555555555554, | |
| "grad_norm": 0.05919402837753296, | |
| "learning_rate": 2.8349413630075906e-05, | |
| "loss": 0.728759765625, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.161111111111111, | |
| "grad_norm": 0.05981147661805153, | |
| "learning_rate": 2.8127863373374635e-05, | |
| "loss": 0.695556640625, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.1666666666666665, | |
| "grad_norm": 0.06054292991757393, | |
| "learning_rate": 2.7907320896955413e-05, | |
| "loss": 0.744140625, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.172222222222222, | |
| "grad_norm": 0.06435830891132355, | |
| "learning_rate": 2.7687794471777954e-05, | |
| "loss": 0.7861328125, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.1777777777777776, | |
| "grad_norm": 0.05324579402804375, | |
| "learning_rate": 2.7469292330697193e-05, | |
| "loss": 0.6168212890625, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.183333333333333, | |
| "grad_norm": 0.057445455342531204, | |
| "learning_rate": 2.7251822668154546e-05, | |
| "loss": 0.815673828125, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.188888888888889, | |
| "grad_norm": 0.063711978495121, | |
| "learning_rate": 2.70353936398706e-05, | |
| "loss": 0.7333984375, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.1944444444444446, | |
| "grad_norm": 0.05439506098628044, | |
| "learning_rate": 2.6820013362539235e-05, | |
| "loss": 0.7998046875, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.05493047833442688, | |
| "learning_rate": 2.6605689913523263e-05, | |
| "loss": 0.7783203125, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.2055555555555557, | |
| "grad_norm": 0.059178080409765244, | |
| "learning_rate": 2.639243133055145e-05, | |
| "loss": 0.69970703125, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.2111111111111112, | |
| "grad_norm": 0.059976737946271896, | |
| "learning_rate": 2.6180245611417127e-05, | |
| "loss": 0.721923828125, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.216666666666667, | |
| "grad_norm": 0.06078799441456795, | |
| "learning_rate": 2.596914071367822e-05, | |
| "loss": 0.7421875, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.05953007936477661, | |
| "learning_rate": 2.5759124554358826e-05, | |
| "loss": 0.7021484375, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.227777777777778, | |
| "grad_norm": 0.06001618131995201, | |
| "learning_rate": 2.555020500965233e-05, | |
| "loss": 0.62158203125, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.2333333333333334, | |
| "grad_norm": 0.06746778637170792, | |
| "learning_rate": 2.534238991462597e-05, | |
| "loss": 0.675048828125, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.238888888888889, | |
| "grad_norm": 0.05842980369925499, | |
| "learning_rate": 2.513568706292705e-05, | |
| "loss": 0.75048828125, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.2444444444444445, | |
| "grad_norm": 0.05836916342377663, | |
| "learning_rate": 2.4930104206490632e-05, | |
| "loss": 0.695068359375, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 0.07351301610469818, | |
| "learning_rate": 2.4725649055248835e-05, | |
| "loss": 0.7275390625, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.2555555555555555, | |
| "grad_norm": 0.06173352524638176, | |
| "learning_rate": 2.4522329276841663e-05, | |
| "loss": 0.724609375, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.261111111111111, | |
| "grad_norm": 0.05910336226224899, | |
| "learning_rate": 2.4320152496329472e-05, | |
| "loss": 0.7666015625, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.2666666666666666, | |
| "grad_norm": 0.06253387033939362, | |
| "learning_rate": 2.4119126295906998e-05, | |
| "loss": 0.693115234375, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.272222222222222, | |
| "grad_norm": 0.05692410469055176, | |
| "learning_rate": 2.3919258214619008e-05, | |
| "loss": 0.69482421875, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.2777777777777777, | |
| "grad_norm": 0.060253918170928955, | |
| "learning_rate": 2.372055574807756e-05, | |
| "loss": 0.68798828125, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.283333333333333, | |
| "grad_norm": 0.05876357480883598, | |
| "learning_rate": 2.3523026348180902e-05, | |
| "loss": 0.7197265625, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.2888888888888888, | |
| "grad_norm": 0.059474315494298935, | |
| "learning_rate": 2.3326677422833994e-05, | |
| "loss": 0.72216796875, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.2944444444444443, | |
| "grad_norm": 0.05722491815686226, | |
| "learning_rate": 2.3131516335670704e-05, | |
| "loss": 0.72998046875, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.06857947260141373, | |
| "learning_rate": 2.2937550405777647e-05, | |
| "loss": 0.7255859375, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.3055555555555554, | |
| "grad_norm": 0.05790672078728676, | |
| "learning_rate": 2.2744786907419703e-05, | |
| "loss": 0.6865234375, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.311111111111111, | |
| "grad_norm": 0.07045500725507736, | |
| "learning_rate": 2.255323306976719e-05, | |
| "loss": 0.712158203125, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.3166666666666664, | |
| "grad_norm": 0.059966836124658585, | |
| "learning_rate": 2.2362896076624784e-05, | |
| "loss": 0.66015625, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.3222222222222224, | |
| "grad_norm": 0.05525347590446472, | |
| "learning_rate": 2.2173783066162073e-05, | |
| "loss": 0.73046875, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.327777777777778, | |
| "grad_norm": 0.07166549563407898, | |
| "learning_rate": 2.1985901130645882e-05, | |
| "loss": 0.684326171875, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.06499408185482025, | |
| "learning_rate": 2.1799257316174284e-05, | |
| "loss": 0.673828125, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.338888888888889, | |
| "grad_norm": 0.05752599611878395, | |
| "learning_rate": 2.1613858622412337e-05, | |
| "loss": 0.7099609375, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.3444444444444446, | |
| "grad_norm": 0.06391977518796921, | |
| "learning_rate": 2.14297120023296e-05, | |
| "loss": 0.751953125, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.06590812653303146, | |
| "learning_rate": 2.124682436193937e-05, | |
| "loss": 0.6982421875, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.3555555555555556, | |
| "grad_norm": 0.0588902048766613, | |
| "learning_rate": 2.1065202560039677e-05, | |
| "loss": 0.69970703125, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.361111111111111, | |
| "grad_norm": 0.05931660160422325, | |
| "learning_rate": 2.0884853407956064e-05, | |
| "loss": 0.654296875, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.3666666666666667, | |
| "grad_norm": 0.05627746134996414, | |
| "learning_rate": 2.0705783669286157e-05, | |
| "loss": 0.6298828125, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.3722222222222222, | |
| "grad_norm": 0.059777069836854935, | |
| "learning_rate": 2.0528000059645997e-05, | |
| "loss": 0.6552734375, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.3777777777777778, | |
| "grad_norm": 0.06560508161783218, | |
| "learning_rate": 2.0351509246418182e-05, | |
| "loss": 0.7158203125, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.3833333333333333, | |
| "grad_norm": 0.06468797475099564, | |
| "learning_rate": 2.017631784850184e-05, | |
| "loss": 0.69775390625, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.388888888888889, | |
| "grad_norm": 0.06385242938995361, | |
| "learning_rate": 2.0002432436064383e-05, | |
| "loss": 0.6943359375, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.3944444444444444, | |
| "grad_norm": 0.062236085534095764, | |
| "learning_rate": 1.982985953029511e-05, | |
| "loss": 0.7099609375, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.06230605021119118, | |
| "learning_rate": 1.9658605603160668e-05, | |
| "loss": 0.7001953125, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_loss": 0.8336029052734375, | |
| "eval_runtime": 13.0915, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.153, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.4055555555555554, | |
| "grad_norm": 0.06140171363949776, | |
| "learning_rate": 1.9488677077162295e-05, | |
| "loss": 0.630859375, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.411111111111111, | |
| "grad_norm": 0.06601665914058685, | |
| "learning_rate": 1.9320080325094993e-05, | |
| "loss": 0.688232421875, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.4166666666666665, | |
| "grad_norm": 0.06363050639629364, | |
| "learning_rate": 1.9152821669808506e-05, | |
| "loss": 0.66162109375, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.422222222222222, | |
| "grad_norm": 0.061124689877033234, | |
| "learning_rate": 1.8986907383970225e-05, | |
| "loss": 0.6728515625, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.4277777777777776, | |
| "grad_norm": 0.06429571658372879, | |
| "learning_rate": 1.8822343689829896e-05, | |
| "loss": 0.668701171875, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.4333333333333336, | |
| "grad_norm": 0.06824830174446106, | |
| "learning_rate": 1.865913675898633e-05, | |
| "loss": 0.67333984375, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.438888888888889, | |
| "grad_norm": 0.0652233436703682, | |
| "learning_rate": 1.8497292712155905e-05, | |
| "loss": 0.7109375, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 0.06093544885516167, | |
| "learning_rate": 1.833681761894303e-05, | |
| "loss": 0.6748046875, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 0.06681835651397705, | |
| "learning_rate": 1.817771749761254e-05, | |
| "loss": 0.6787109375, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.4555555555555557, | |
| "grad_norm": 0.09045480191707611, | |
| "learning_rate": 1.8019998314863974e-05, | |
| "loss": 0.68359375, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.4611111111111112, | |
| "grad_norm": 0.059314239770174026, | |
| "learning_rate": 1.7863665985607802e-05, | |
| "loss": 0.60302734375, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.466666666666667, | |
| "grad_norm": 0.06481864303350449, | |
| "learning_rate": 1.7708726372743618e-05, | |
| "loss": 0.658203125, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.4722222222222223, | |
| "grad_norm": 0.06287098675966263, | |
| "learning_rate": 1.755518528694025e-05, | |
| "loss": 0.60400390625, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.477777777777778, | |
| "grad_norm": 0.06786705553531647, | |
| "learning_rate": 1.740304848641787e-05, | |
| "loss": 0.672607421875, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.4833333333333334, | |
| "grad_norm": 0.07895544171333313, | |
| "learning_rate": 1.725232167673199e-05, | |
| "loss": 0.73291015625, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.068633072078228, | |
| "learning_rate": 1.710301051055955e-05, | |
| "loss": 0.724609375, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.4944444444444445, | |
| "grad_norm": 0.06754299998283386, | |
| "learning_rate": 1.6955120587486885e-05, | |
| "loss": 0.69287109375, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.07174503803253174, | |
| "learning_rate": 1.6808657453799744e-05, | |
| "loss": 0.6669921875, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.5055555555555555, | |
| "grad_norm": 0.07379809767007828, | |
| "learning_rate": 1.6663626602275288e-05, | |
| "loss": 0.6845703125, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.511111111111111, | |
| "grad_norm": 0.07057462632656097, | |
| "learning_rate": 1.6520033471976077e-05, | |
| "loss": 0.700927734375, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.5166666666666666, | |
| "grad_norm": 0.12075720727443695, | |
| "learning_rate": 1.6377883448046122e-05, | |
| "loss": 0.762451171875, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.522222222222222, | |
| "grad_norm": 0.060440488159656525, | |
| "learning_rate": 1.623718186150889e-05, | |
| "loss": 0.69580078125, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.5277777777777777, | |
| "grad_norm": 0.07302123308181763, | |
| "learning_rate": 1.60979339890674e-05, | |
| "loss": 0.5406494140625, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.533333333333333, | |
| "grad_norm": 0.0633658692240715, | |
| "learning_rate": 1.5960145052906338e-05, | |
| "loss": 0.64111328125, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.5388888888888888, | |
| "grad_norm": 0.05979062244296074, | |
| "learning_rate": 1.5823820220496195e-05, | |
| "loss": 0.732421875, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.5444444444444443, | |
| "grad_norm": 0.07253874093294144, | |
| "learning_rate": 1.568896460439947e-05, | |
| "loss": 0.67919921875, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 0.07385561615228653, | |
| "learning_rate": 1.5555583262078944e-05, | |
| "loss": 0.72998046875, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.5555555555555554, | |
| "grad_norm": 0.06815175712108612, | |
| "learning_rate": 1.5423681195707997e-05, | |
| "loss": 0.66845703125, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.561111111111111, | |
| "grad_norm": 0.07204774767160416, | |
| "learning_rate": 1.5293263351983044e-05, | |
| "loss": 0.689697265625, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.5666666666666664, | |
| "grad_norm": 0.06678230315446854, | |
| "learning_rate": 1.5164334621937992e-05, | |
| "loss": 0.633056640625, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.572222222222222, | |
| "grad_norm": 0.07442482560873032, | |
| "learning_rate": 1.5036899840760815e-05, | |
| "loss": 0.645751953125, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.5777777777777775, | |
| "grad_norm": 0.06967206299304962, | |
| "learning_rate": 1.4910963787612226e-05, | |
| "loss": 0.598236083984375, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.5833333333333335, | |
| "grad_norm": 0.0773426964879036, | |
| "learning_rate": 1.4786531185446454e-05, | |
| "loss": 0.68408203125, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.588888888888889, | |
| "grad_norm": 0.05998453125357628, | |
| "learning_rate": 1.4663606700834101e-05, | |
| "loss": 0.533905029296875, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.5944444444444446, | |
| "grad_norm": 0.0824751928448677, | |
| "learning_rate": 1.4542194943787144e-05, | |
| "loss": 0.656982421875, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.07359018176794052, | |
| "learning_rate": 1.442230046758605e-05, | |
| "loss": 0.6787109375, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.6055555555555556, | |
| "grad_norm": 0.07717732340097427, | |
| "learning_rate": 1.4303927768609015e-05, | |
| "loss": 0.766845703125, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 2.611111111111111, | |
| "grad_norm": 0.07383830100297928, | |
| "learning_rate": 1.4187081286163323e-05, | |
| "loss": 0.70849609375, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.6166666666666667, | |
| "grad_norm": 0.06743068248033524, | |
| "learning_rate": 1.4071765402318873e-05, | |
| "loss": 0.8076171875, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 2.6222222222222222, | |
| "grad_norm": 0.12001865357160568, | |
| "learning_rate": 1.3957984441743832e-05, | |
| "loss": 0.699462890625, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 2.6277777777777778, | |
| "grad_norm": 0.07017767429351807, | |
| "learning_rate": 1.384574267154246e-05, | |
| "loss": 0.69970703125, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 2.6333333333333333, | |
| "grad_norm": 0.11585768312215805, | |
| "learning_rate": 1.3735044301095068e-05, | |
| "loss": 0.6669921875, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 2.638888888888889, | |
| "grad_norm": 0.06867586821317673, | |
| "learning_rate": 1.3625893481900154e-05, | |
| "loss": 0.691650390625, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.6444444444444444, | |
| "grad_norm": 0.06327942758798599, | |
| "learning_rate": 1.3518294307418727e-05, | |
| "loss": 0.6953125, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.07125144451856613, | |
| "learning_rate": 1.3412250812920767e-05, | |
| "loss": 0.68994140625, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 2.6555555555555554, | |
| "grad_norm": 0.0676037147641182, | |
| "learning_rate": 1.330776697533392e-05, | |
| "loss": 0.67236328125, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 2.661111111111111, | |
| "grad_norm": 0.08281464129686356, | |
| "learning_rate": 1.3204846713094326e-05, | |
| "loss": 0.732421875, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.0687449723482132, | |
| "learning_rate": 1.3103493885999663e-05, | |
| "loss": 0.636474609375, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.6722222222222225, | |
| "grad_norm": 0.06657174229621887, | |
| "learning_rate": 1.3003712295064432e-05, | |
| "loss": 0.6875, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 2.677777777777778, | |
| "grad_norm": 0.07291577756404877, | |
| "learning_rate": 1.290550568237737e-05, | |
| "loss": 0.6865234375, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 2.6833333333333336, | |
| "grad_norm": 0.07293625175952911, | |
| "learning_rate": 1.280887773096112e-05, | |
| "loss": 0.65185546875, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 2.688888888888889, | |
| "grad_norm": 0.06640584766864777, | |
| "learning_rate": 1.2713832064634126e-05, | |
| "loss": 0.66455078125, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 2.6944444444444446, | |
| "grad_norm": 0.07449325174093246, | |
| "learning_rate": 1.2620372247874704e-05, | |
| "loss": 0.682373046875, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.06977734714746475, | |
| "learning_rate": 1.2528501785687382e-05, | |
| "loss": 0.689697265625, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 2.7055555555555557, | |
| "grad_norm": 0.08931327611207962, | |
| "learning_rate": 1.2438224123471442e-05, | |
| "loss": 0.7158203125, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 2.7111111111111112, | |
| "grad_norm": 0.06955769658088684, | |
| "learning_rate": 1.2349542646891717e-05, | |
| "loss": 0.681396484375, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 2.716666666666667, | |
| "grad_norm": 0.06726157665252686, | |
| "learning_rate": 1.2262460681751609e-05, | |
| "loss": 0.7607421875, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 2.7222222222222223, | |
| "grad_norm": 0.07009522616863251, | |
| "learning_rate": 1.2176981493868372e-05, | |
| "loss": 0.6748046875, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.727777777777778, | |
| "grad_norm": 0.07001668959856033, | |
| "learning_rate": 1.2093108288950624e-05, | |
| "loss": 0.666015625, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 2.7333333333333334, | |
| "grad_norm": 0.06748469173908234, | |
| "learning_rate": 1.2010844212478138e-05, | |
| "loss": 0.663330078125, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 2.738888888888889, | |
| "grad_norm": 0.07007578760385513, | |
| "learning_rate": 1.1930192349583866e-05, | |
| "loss": 0.677734375, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.7444444444444445, | |
| "grad_norm": 0.06786662340164185, | |
| "learning_rate": 1.1851155724938236e-05, | |
| "loss": 0.693359375, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.0705144852399826, | |
| "learning_rate": 1.177373730263574e-05, | |
| "loss": 0.67041015625, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.7555555555555555, | |
| "grad_norm": 0.0744699165225029, | |
| "learning_rate": 1.1697939986083733e-05, | |
| "loss": 0.67529296875, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.761111111111111, | |
| "grad_norm": 0.06859414279460907, | |
| "learning_rate": 1.1623766617893593e-05, | |
| "loss": 0.6708984375, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.7666666666666666, | |
| "grad_norm": 0.0677483007311821, | |
| "learning_rate": 1.1551219979774077e-05, | |
| "loss": 0.683349609375, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.772222222222222, | |
| "grad_norm": 0.06732510775327682, | |
| "learning_rate": 1.1480302792427021e-05, | |
| "loss": 0.64990234375, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 0.07116559892892838, | |
| "learning_rate": 1.1411017715445298e-05, | |
| "loss": 0.76123046875, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.783333333333333, | |
| "grad_norm": 0.07286809384822845, | |
| "learning_rate": 1.134336734721308e-05, | |
| "loss": 0.6683349609375, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.7888888888888888, | |
| "grad_norm": 0.06269260495901108, | |
| "learning_rate": 1.127735422480838e-05, | |
| "loss": 0.73193359375, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.7944444444444443, | |
| "grad_norm": 0.06778956204652786, | |
| "learning_rate": 1.1212980823907931e-05, | |
| "loss": 0.67822265625, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.08134780079126358, | |
| "learning_rate": 1.1150249558694307e-05, | |
| "loss": 0.740234375, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.8055555555555554, | |
| "grad_norm": 0.06680338084697723, | |
| "learning_rate": 1.1089162781765398e-05, | |
| "loss": 0.67578125, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.811111111111111, | |
| "grad_norm": 0.06948743015527725, | |
| "learning_rate": 1.1029722784046206e-05, | |
| "loss": 0.68408203125, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.8166666666666664, | |
| "grad_norm": 0.07396679371595383, | |
| "learning_rate": 1.0971931794702888e-05, | |
| "loss": 0.695068359375, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.822222222222222, | |
| "grad_norm": 0.07003337144851685, | |
| "learning_rate": 1.0915791981059174e-05, | |
| "loss": 0.7099609375, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.8277777777777775, | |
| "grad_norm": 0.07161819189786911, | |
| "learning_rate": 1.0861305448515102e-05, | |
| "loss": 0.73095703125, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.8333333333333335, | |
| "grad_norm": 0.06442680209875107, | |
| "learning_rate": 1.080847424046803e-05, | |
| "loss": 0.611907958984375, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.838888888888889, | |
| "grad_norm": 0.07292718440294266, | |
| "learning_rate": 1.0757300338236025e-05, | |
| "loss": 0.69873046875, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.07167823612689972, | |
| "learning_rate": 1.0707785660983548e-05, | |
| "loss": 0.6923828125, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 0.07784584164619446, | |
| "learning_rate": 1.0659932065649484e-05, | |
| "loss": 0.6845703125, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.8555555555555556, | |
| "grad_norm": 0.059034563601017, | |
| "learning_rate": 1.0613741346877497e-05, | |
| "loss": 0.615478515625, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.861111111111111, | |
| "grad_norm": 0.06504210084676743, | |
| "learning_rate": 1.056921523694873e-05, | |
| "loss": 0.75335693359375, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.8666666666666667, | |
| "grad_norm": 0.09744574874639511, | |
| "learning_rate": 1.052635540571685e-05, | |
| "loss": 0.703125, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.8722222222222222, | |
| "grad_norm": 0.06383401155471802, | |
| "learning_rate": 1.0485163460545396e-05, | |
| "loss": 0.640625, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.8777777777777778, | |
| "grad_norm": 0.07074729353189468, | |
| "learning_rate": 1.044564094624752e-05, | |
| "loss": 0.7255859375, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.8833333333333333, | |
| "grad_norm": 0.07076207548379898, | |
| "learning_rate": 1.040778934502805e-05, | |
| "loss": 0.7158203125, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 0.07455959171056747, | |
| "learning_rate": 1.03716100764279e-05, | |
| "loss": 0.72998046875, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.8944444444444444, | |
| "grad_norm": 0.07080957293510437, | |
| "learning_rate": 1.0337104497270828e-05, | |
| "loss": 0.650634765625, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.0693393275141716, | |
| "learning_rate": 1.0304273901612566e-05, | |
| "loss": 0.693359375, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.9055555555555554, | |
| "grad_norm": 0.07615337520837784, | |
| "learning_rate": 1.0273119520692275e-05, | |
| "loss": 0.66650390625, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.911111111111111, | |
| "grad_norm": 0.0734100267291069, | |
| "learning_rate": 1.024364252288637e-05, | |
| "loss": 0.73779296875, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.9166666666666665, | |
| "grad_norm": 0.06750860065221786, | |
| "learning_rate": 1.0215844013664718e-05, | |
| "loss": 0.651336669921875, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.9222222222222225, | |
| "grad_norm": 0.06790855526924133, | |
| "learning_rate": 1.0189725035549163e-05, | |
| "loss": 0.6845703125, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.927777777777778, | |
| "grad_norm": 0.06828059256076813, | |
| "learning_rate": 1.0165286568074438e-05, | |
| "loss": 0.67333984375, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.9333333333333336, | |
| "grad_norm": 0.07611801475286484, | |
| "learning_rate": 1.0142529527751419e-05, | |
| "loss": 0.6837158203125, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.938888888888889, | |
| "grad_norm": 0.05909605324268341, | |
| "learning_rate": 1.012145476803277e-05, | |
| "loss": 0.650146484375, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.9444444444444446, | |
| "grad_norm": 0.06827764213085175, | |
| "learning_rate": 1.0102063079280928e-05, | |
| "loss": 0.69580078125, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.06995943933725357, | |
| "learning_rate": 1.0084355188738454e-05, | |
| "loss": 0.656982421875, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.9555555555555557, | |
| "grad_norm": 0.07155478000640869, | |
| "learning_rate": 1.0068331760500774e-05, | |
| "loss": 0.732421875, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.9611111111111112, | |
| "grad_norm": 0.06999050825834274, | |
| "learning_rate": 1.0053993395491255e-05, | |
| "loss": 0.681640625, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.966666666666667, | |
| "grad_norm": 0.06852614879608154, | |
| "learning_rate": 1.0041340631438694e-05, | |
| "loss": 0.71240234375, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.9722222222222223, | |
| "grad_norm": 0.0718725174665451, | |
| "learning_rate": 1.0030373942857137e-05, | |
| "loss": 0.69921875, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.977777777777778, | |
| "grad_norm": 0.06595950573682785, | |
| "learning_rate": 1.0021093741028064e-05, | |
| "loss": 0.67724609375, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.9833333333333334, | |
| "grad_norm": 0.06136660650372505, | |
| "learning_rate": 1.0013500373985012e-05, | |
| "loss": 0.7841796875, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.988888888888889, | |
| "grad_norm": 0.06931949406862259, | |
| "learning_rate": 1.0007594126500479e-05, | |
| "loss": 0.71630859375, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.9944444444444445, | |
| "grad_norm": 0.06364605575799942, | |
| "learning_rate": 1.0003375220075262e-05, | |
| "loss": 0.65283203125, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.07343864440917969, | |
| "learning_rate": 1.0000843812930149e-05, | |
| "loss": 0.611572265625, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.830413818359375, | |
| "eval_runtime": 13.1045, | |
| "eval_samples_per_second": 0.916, | |
| "eval_steps_per_second": 0.153, | |
| "step": 540 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 540, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.469112605928692e+19, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |