| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984301412872841, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0020931449502878076, | |
| "grad_norm": 18.07411787458952, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": -2.6982662677764893, | |
| "logits/rejected": -2.639620304107666, | |
| "logps/chosen": -340.48919677734375, | |
| "logps/rejected": -383.30560302734375, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.020931449502878074, | |
| "grad_norm": 9.9608972008456, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.4643380641937256, | |
| "logits/rejected": -2.41880464553833, | |
| "logps/chosen": -332.1871032714844, | |
| "logps/rejected": -297.4084167480469, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.5277777910232544, | |
| "rewards/chosen": 0.0005977663677185774, | |
| "rewards/margins": 0.0007145923445932567, | |
| "rewards/rejected": -0.00011682594777084887, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04186289900575615, | |
| "grad_norm": 13.745638539501607, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.563354969024658, | |
| "logits/rejected": -2.4802138805389404, | |
| "logps/chosen": -330.6456604003906, | |
| "logps/rejected": -303.16796875, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0026146548334509134, | |
| "rewards/margins": 0.0005810384755022824, | |
| "rewards/rejected": 0.0020336161833256483, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06279434850863422, | |
| "grad_norm": 7.832516309339627, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.532294511795044, | |
| "logits/rejected": -2.4404759407043457, | |
| "logps/chosen": -312.492431640625, | |
| "logps/rejected": -266.53936767578125, | |
| "loss": 0.6904, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.01948517933487892, | |
| "rewards/margins": 0.006482087075710297, | |
| "rewards/rejected": 0.0130030931904912, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0837257980115123, | |
| "grad_norm": 7.9616158798093855, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.431403636932373, | |
| "logits/rejected": -2.379568099975586, | |
| "logps/chosen": -308.171630859375, | |
| "logps/rejected": -279.072998046875, | |
| "loss": 0.6845, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.046426668763160706, | |
| "rewards/margins": 0.019384978339076042, | |
| "rewards/rejected": 0.027041692286729813, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10465724751439037, | |
| "grad_norm": 12.138395522818126, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.467078685760498, | |
| "logits/rejected": -2.3770642280578613, | |
| "logps/chosen": -292.5061950683594, | |
| "logps/rejected": -287.78692626953125, | |
| "loss": 0.6768, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.05952076241374016, | |
| "rewards/margins": 0.031759217381477356, | |
| "rewards/rejected": 0.027761543169617653, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12558869701726844, | |
| "grad_norm": 12.71284691980484, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.5058963298797607, | |
| "logits/rejected": -2.4367785453796387, | |
| "logps/chosen": -269.9830322265625, | |
| "logps/rejected": -261.7590026855469, | |
| "loss": 0.6648, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.054543137550354004, | |
| "rewards/margins": 0.06885507702827454, | |
| "rewards/rejected": -0.01431194506585598, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14652014652014653, | |
| "grad_norm": 8.977266411290348, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -2.4574174880981445, | |
| "logits/rejected": -2.4204437732696533, | |
| "logps/chosen": -323.3358459472656, | |
| "logps/rejected": -313.7773742675781, | |
| "loss": 0.6448, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.02133224718272686, | |
| "rewards/margins": 0.09954627603292465, | |
| "rewards/rejected": -0.12087850272655487, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1674515960230246, | |
| "grad_norm": 15.44090832738, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.5686161518096924, | |
| "logits/rejected": -2.459707260131836, | |
| "logps/chosen": -367.8087463378906, | |
| "logps/rejected": -306.7381286621094, | |
| "loss": 0.6262, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.07196881622076035, | |
| "rewards/margins": 0.20668184757232666, | |
| "rewards/rejected": -0.2786506414413452, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18838304552590268, | |
| "grad_norm": 17.357813817512778, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.484760284423828, | |
| "logits/rejected": -2.4333367347717285, | |
| "logps/chosen": -286.19573974609375, | |
| "logps/rejected": -318.17987060546875, | |
| "loss": 0.6072, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.13759762048721313, | |
| "rewards/margins": 0.2588348686695099, | |
| "rewards/rejected": -0.396432489156723, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20931449502878074, | |
| "grad_norm": 11.555230239819865, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -2.481173038482666, | |
| "logits/rejected": -2.436263084411621, | |
| "logps/chosen": -343.57391357421875, | |
| "logps/rejected": -353.5834655761719, | |
| "loss": 0.6047, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.17556986212730408, | |
| "rewards/margins": 0.37368106842041016, | |
| "rewards/rejected": -0.5492509603500366, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2302459445316588, | |
| "grad_norm": 12.709794889392892, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.3503360748291016, | |
| "logits/rejected": -2.259538173675537, | |
| "logps/chosen": -345.1012268066406, | |
| "logps/rejected": -358.5026550292969, | |
| "loss": 0.6012, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.3211168348789215, | |
| "rewards/margins": 0.30802685022354126, | |
| "rewards/rejected": -0.6291436553001404, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25117739403453687, | |
| "grad_norm": 16.24550776128439, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -1.8585760593414307, | |
| "logits/rejected": -1.812425971031189, | |
| "logps/chosen": -319.6259765625, | |
| "logps/rejected": -346.91424560546875, | |
| "loss": 0.5874, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.20474886894226074, | |
| "rewards/margins": 0.4103702902793884, | |
| "rewards/rejected": -0.6151191592216492, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.272108843537415, | |
| "grad_norm": 18.658464566773528, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -1.4963523149490356, | |
| "logits/rejected": -1.3689486980438232, | |
| "logps/chosen": -381.5474548339844, | |
| "logps/rejected": -392.7295837402344, | |
| "loss": 0.5767, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.6478039622306824, | |
| "rewards/margins": 0.3431198000907898, | |
| "rewards/rejected": -0.9909237623214722, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29304029304029305, | |
| "grad_norm": 14.61768252656386, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -1.2942148447036743, | |
| "logits/rejected": -1.0670334100723267, | |
| "logps/chosen": -374.21197509765625, | |
| "logps/rejected": -370.75750732421875, | |
| "loss": 0.5762, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3039678633213043, | |
| "rewards/margins": 0.4882018566131592, | |
| "rewards/rejected": -0.7921696901321411, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3139717425431711, | |
| "grad_norm": 22.141164123200294, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -1.035481572151184, | |
| "logits/rejected": -0.8223312497138977, | |
| "logps/chosen": -351.9337463378906, | |
| "logps/rejected": -350.9710693359375, | |
| "loss": 0.5711, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.4184446334838867, | |
| "rewards/margins": 0.40214234590530396, | |
| "rewards/rejected": -0.8205869793891907, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3349031920460492, | |
| "grad_norm": 16.775320749910513, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -0.8283090591430664, | |
| "logits/rejected": -0.4383368492126465, | |
| "logps/chosen": -351.2560119628906, | |
| "logps/rejected": -372.83636474609375, | |
| "loss": 0.5738, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.5103927850723267, | |
| "rewards/margins": 0.5579272508621216, | |
| "rewards/rejected": -1.0683200359344482, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35583464154892724, | |
| "grad_norm": 18.321929450201765, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -0.2253287136554718, | |
| "logits/rejected": 0.010831773281097412, | |
| "logps/chosen": -337.71563720703125, | |
| "logps/rejected": -353.23150634765625, | |
| "loss": 0.5681, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5377098917961121, | |
| "rewards/margins": 0.43757739663124084, | |
| "rewards/rejected": -0.9752873182296753, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37676609105180536, | |
| "grad_norm": 17.583054376284373, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -0.036564283072948456, | |
| "logits/rejected": 0.21061678230762482, | |
| "logps/chosen": -348.42608642578125, | |
| "logps/rejected": -363.5872802734375, | |
| "loss": 0.5376, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.5707951784133911, | |
| "rewards/margins": 0.45119351148605347, | |
| "rewards/rejected": -1.0219886302947998, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3976975405546834, | |
| "grad_norm": 23.074661957066084, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": 0.14501239359378815, | |
| "logits/rejected": 0.1892128884792328, | |
| "logps/chosen": -327.35870361328125, | |
| "logps/rejected": -371.1208801269531, | |
| "loss": 0.5554, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.48090630769729614, | |
| "rewards/margins": 0.45326748490333557, | |
| "rewards/rejected": -0.9341737627983093, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4186289900575615, | |
| "grad_norm": 22.069261987026525, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": 0.5265301465988159, | |
| "logits/rejected": 0.6243599653244019, | |
| "logps/chosen": -359.69854736328125, | |
| "logps/rejected": -418.3558044433594, | |
| "loss": 0.5655, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.7841523885726929, | |
| "rewards/margins": 0.46337419748306274, | |
| "rewards/rejected": -1.2475265264511108, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43956043956043955, | |
| "grad_norm": 17.42416201312016, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -0.36963820457458496, | |
| "logits/rejected": 0.10797449201345444, | |
| "logps/chosen": -408.05328369140625, | |
| "logps/rejected": -385.1399841308594, | |
| "loss": 0.573, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.7120479345321655, | |
| "rewards/margins": 0.45911431312561035, | |
| "rewards/rejected": -1.1711622476577759, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4604918890633176, | |
| "grad_norm": 21.650857513168543, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -1.1279886960983276, | |
| "logits/rejected": -1.0101404190063477, | |
| "logps/chosen": -330.86920166015625, | |
| "logps/rejected": -375.27581787109375, | |
| "loss": 0.5521, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.5428093075752258, | |
| "rewards/margins": 0.5479553937911987, | |
| "rewards/rejected": -1.0907647609710693, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48142333856619574, | |
| "grad_norm": 21.29968622459755, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -1.4807205200195312, | |
| "logits/rejected": -1.2653987407684326, | |
| "logps/chosen": -349.1932067871094, | |
| "logps/rejected": -402.3145751953125, | |
| "loss": 0.5591, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.6056884527206421, | |
| "rewards/margins": 0.7848646640777588, | |
| "rewards/rejected": -1.3905531167984009, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023547880690737, | |
| "grad_norm": 21.54037562544609, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -1.4609298706054688, | |
| "logits/rejected": -1.2633426189422607, | |
| "logps/chosen": -361.18109130859375, | |
| "logps/rejected": -385.7818298339844, | |
| "loss": 0.5293, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.4389161169528961, | |
| "rewards/margins": 0.5328649878501892, | |
| "rewards/rejected": -0.9717811346054077, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5232862375719518, | |
| "grad_norm": 19.10437589239632, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -1.1038726568222046, | |
| "logits/rejected": -0.7872940897941589, | |
| "logps/chosen": -367.0468444824219, | |
| "logps/rejected": -402.64837646484375, | |
| "loss": 0.5392, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.618547260761261, | |
| "rewards/margins": 0.7416694760322571, | |
| "rewards/rejected": -1.360216736793518, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.54421768707483, | |
| "grad_norm": 27.812573286232187, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -0.946252703666687, | |
| "logits/rejected": -0.8400171399116516, | |
| "logps/chosen": -360.14691162109375, | |
| "logps/rejected": -391.15576171875, | |
| "loss": 0.5386, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.7270759344100952, | |
| "rewards/margins": 0.5733479261398315, | |
| "rewards/rejected": -1.3004238605499268, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.565149136577708, | |
| "grad_norm": 19.996060331745692, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -1.0181360244750977, | |
| "logits/rejected": -0.8419178128242493, | |
| "logps/chosen": -381.3074645996094, | |
| "logps/rejected": -421.84503173828125, | |
| "loss": 0.5439, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.6155784726142883, | |
| "rewards/margins": 0.7110739946365356, | |
| "rewards/rejected": -1.3266522884368896, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5860805860805861, | |
| "grad_norm": 20.456176579909233, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -0.8568949699401855, | |
| "logits/rejected": -0.6260634660720825, | |
| "logps/chosen": -355.68487548828125, | |
| "logps/rejected": -365.54150390625, | |
| "loss": 0.5521, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.5479617118835449, | |
| "rewards/margins": 0.554092288017273, | |
| "rewards/rejected": -1.1020539999008179, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6070120355834642, | |
| "grad_norm": 21.855087428817388, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -0.9505099058151245, | |
| "logits/rejected": -0.7808119654655457, | |
| "logps/chosen": -334.37713623046875, | |
| "logps/rejected": -388.1415100097656, | |
| "loss": 0.5459, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.6734660863876343, | |
| "rewards/margins": 0.5256973505020142, | |
| "rewards/rejected": -1.1991634368896484, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6279434850863422, | |
| "grad_norm": 20.553743451672336, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -0.9063513875007629, | |
| "logits/rejected": -0.6456217765808105, | |
| "logps/chosen": -422.3011169433594, | |
| "logps/rejected": -447.60308837890625, | |
| "loss": 0.5201, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.6219363212585449, | |
| "rewards/margins": 0.6523629426956177, | |
| "rewards/rejected": -1.2742993831634521, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6488749345892203, | |
| "grad_norm": 19.16702342627125, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -0.8550491333007812, | |
| "logits/rejected": -0.6107693910598755, | |
| "logps/chosen": -392.9826354980469, | |
| "logps/rejected": -404.90277099609375, | |
| "loss": 0.5161, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.6534227132797241, | |
| "rewards/margins": 0.7227972149848938, | |
| "rewards/rejected": -1.3762199878692627, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6698063840920984, | |
| "grad_norm": 20.684056407375124, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -0.7066868543624878, | |
| "logits/rejected": -0.4419567584991455, | |
| "logps/chosen": -337.99554443359375, | |
| "logps/rejected": -378.85528564453125, | |
| "loss": 0.5178, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.5666108131408691, | |
| "rewards/margins": 0.76667320728302, | |
| "rewards/rejected": -1.3332840204238892, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6907378335949764, | |
| "grad_norm": 23.16918692489251, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -0.9136036038398743, | |
| "logits/rejected": -0.8735296130180359, | |
| "logps/chosen": -343.11474609375, | |
| "logps/rejected": -400.88055419921875, | |
| "loss": 0.5291, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.5132071375846863, | |
| "rewards/margins": 0.6378787755966187, | |
| "rewards/rejected": -1.1510859727859497, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7116692830978545, | |
| "grad_norm": 25.09011052065921, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -0.8220704197883606, | |
| "logits/rejected": -0.6656316518783569, | |
| "logps/chosen": -376.2585144042969, | |
| "logps/rejected": -415.30108642578125, | |
| "loss": 0.5298, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.5471745133399963, | |
| "rewards/margins": 0.6663635969161987, | |
| "rewards/rejected": -1.2135381698608398, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7326007326007326, | |
| "grad_norm": 23.913230474173087, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -0.8336887359619141, | |
| "logits/rejected": -0.5391756892204285, | |
| "logps/chosen": -333.2408752441406, | |
| "logps/rejected": -374.46160888671875, | |
| "loss": 0.5046, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.46794453263282776, | |
| "rewards/margins": 0.6849315762519836, | |
| "rewards/rejected": -1.1528760194778442, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7535321821036107, | |
| "grad_norm": 18.671801968846303, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -0.6670598983764648, | |
| "logits/rejected": -0.3578875660896301, | |
| "logps/chosen": -389.22222900390625, | |
| "logps/rejected": -422.08209228515625, | |
| "loss": 0.5407, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.5574816465377808, | |
| "rewards/margins": 0.730495035648346, | |
| "rewards/rejected": -1.287976861000061, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7744636316064888, | |
| "grad_norm": 34.13531738812179, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -0.6234959363937378, | |
| "logits/rejected": -0.37343746423721313, | |
| "logps/chosen": -344.8446960449219, | |
| "logps/rejected": -383.56939697265625, | |
| "loss": 0.5266, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.540058434009552, | |
| "rewards/margins": 0.757440984249115, | |
| "rewards/rejected": -1.297499418258667, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7953950811093669, | |
| "grad_norm": 21.95611942170756, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -0.3392348885536194, | |
| "logits/rejected": -0.2354392111301422, | |
| "logps/chosen": -356.6414794921875, | |
| "logps/rejected": -457.1575622558594, | |
| "loss": 0.5252, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.7903109788894653, | |
| "rewards/margins": 0.8445862531661987, | |
| "rewards/rejected": -1.634897232055664, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 25.808898227359457, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -0.7156884074211121, | |
| "logits/rejected": -0.4770375192165375, | |
| "logps/chosen": -443.89642333984375, | |
| "logps/rejected": -474.5353088378906, | |
| "loss": 0.5136, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.6931928396224976, | |
| "rewards/margins": 0.6563575863838196, | |
| "rewards/rejected": -1.349550485610962, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.837257980115123, | |
| "grad_norm": 20.792087565548005, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -0.577874481678009, | |
| "logits/rejected": -0.3143076002597809, | |
| "logps/chosen": -343.8722839355469, | |
| "logps/rejected": -391.06610107421875, | |
| "loss": 0.528, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.6408818960189819, | |
| "rewards/margins": 0.6500493884086609, | |
| "rewards/rejected": -1.290931224822998, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.858189429618001, | |
| "grad_norm": 24.33600208096771, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -0.859704852104187, | |
| "logits/rejected": -0.572528600692749, | |
| "logps/chosen": -388.897705078125, | |
| "logps/rejected": -415.8624572753906, | |
| "loss": 0.5399, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.6071329712867737, | |
| "rewards/margins": 0.7060071229934692, | |
| "rewards/rejected": -1.3131401538848877, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8791208791208791, | |
| "grad_norm": 24.28511601342855, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -0.64429771900177, | |
| "logits/rejected": -0.47082147002220154, | |
| "logps/chosen": -341.98504638671875, | |
| "logps/rejected": -408.1618957519531, | |
| "loss": 0.5184, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.6186074614524841, | |
| "rewards/margins": 0.6798468828201294, | |
| "rewards/rejected": -1.2984542846679688, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9000523286237572, | |
| "grad_norm": 24.32668129564916, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -0.627988338470459, | |
| "logits/rejected": -0.37057411670684814, | |
| "logps/chosen": -384.9641418457031, | |
| "logps/rejected": -431.1815490722656, | |
| "loss": 0.5035, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.5622997283935547, | |
| "rewards/margins": 0.669585108757019, | |
| "rewards/rejected": -1.2318848371505737, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9209837781266352, | |
| "grad_norm": 20.42397496588787, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -0.8050533533096313, | |
| "logits/rejected": -0.5536060929298401, | |
| "logps/chosen": -338.9949035644531, | |
| "logps/rejected": -367.10296630859375, | |
| "loss": 0.5177, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.5558196306228638, | |
| "rewards/margins": 0.6245731711387634, | |
| "rewards/rejected": -1.1803929805755615, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9419152276295133, | |
| "grad_norm": 22.341141934616147, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -0.513416588306427, | |
| "logits/rejected": -0.3268742263317108, | |
| "logps/chosen": -371.934814453125, | |
| "logps/rejected": -432.7196350097656, | |
| "loss": 0.5312, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.6752878427505493, | |
| "rewards/margins": 0.7238127589225769, | |
| "rewards/rejected": -1.399100661277771, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9628466771323915, | |
| "grad_norm": 24.32627212356958, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -0.7187199592590332, | |
| "logits/rejected": -0.3386891782283783, | |
| "logps/chosen": -361.8067932128906, | |
| "logps/rejected": -383.07525634765625, | |
| "loss": 0.5169, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.6250866651535034, | |
| "rewards/margins": 0.7100397348403931, | |
| "rewards/rejected": -1.335126280784607, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9837781266352695, | |
| "grad_norm": 22.68698991416887, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -0.6799210906028748, | |
| "logits/rejected": -0.4490284025669098, | |
| "logps/chosen": -387.94158935546875, | |
| "logps/rejected": -422.15155029296875, | |
| "loss": 0.5164, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.5430302023887634, | |
| "rewards/margins": 0.6616371273994446, | |
| "rewards/rejected": -1.204667329788208, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9984301412872841, | |
| "eval_logits/chosen": -0.44416600465774536, | |
| "eval_logits/rejected": -0.26567625999450684, | |
| "eval_logps/chosen": -342.55780029296875, | |
| "eval_logps/rejected": -420.65606689453125, | |
| "eval_loss": 0.533210277557373, | |
| "eval_rewards/accuracies": 0.76953125, | |
| "eval_rewards/chosen": -0.568053126335144, | |
| "eval_rewards/margins": 0.7595610618591309, | |
| "eval_rewards/rejected": -1.3276140689849854, | |
| "eval_runtime": 169.453, | |
| "eval_samples_per_second": 11.803, | |
| "eval_steps_per_second": 0.189, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.9984301412872841, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5657099517886244, | |
| "train_runtime": 14255.2941, | |
| "train_samples_per_second": 4.289, | |
| "train_steps_per_second": 0.033 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |