virus_pythia_70_3072 / trainer_state.json
Hack90's picture
Upload folder using huggingface_hub
f745bbf verified
raw
history blame contribute delete
No virus
23 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9776462505945146,
"eval_steps": 500,
"global_step": 18500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.976206841854809e-05,
"loss": 1.337,
"step": 100
},
{
"epoch": 0.01,
"learning_rate": 4.949769999471263e-05,
"loss": 1.2993,
"step": 200
},
{
"epoch": 0.02,
"learning_rate": 4.923333157087718e-05,
"loss": 1.2719,
"step": 300
},
{
"epoch": 0.02,
"learning_rate": 4.896896314704172e-05,
"loss": 1.2761,
"step": 400
},
{
"epoch": 0.03,
"learning_rate": 4.870459472320626e-05,
"loss": 1.271,
"step": 500
},
{
"epoch": 0.03,
"learning_rate": 4.844022629937081e-05,
"loss": 1.2504,
"step": 600
},
{
"epoch": 0.04,
"learning_rate": 4.8175857875535345e-05,
"loss": 1.254,
"step": 700
},
{
"epoch": 0.04,
"learning_rate": 4.791148945169989e-05,
"loss": 1.2508,
"step": 800
},
{
"epoch": 0.05,
"learning_rate": 4.7647121027864434e-05,
"loss": 1.2614,
"step": 900
},
{
"epoch": 0.05,
"learning_rate": 4.738275260402898e-05,
"loss": 1.2447,
"step": 1000
},
{
"epoch": 0.06,
"learning_rate": 4.711838418019352e-05,
"loss": 1.2606,
"step": 1100
},
{
"epoch": 0.06,
"learning_rate": 4.685401575635806e-05,
"loss": 1.2447,
"step": 1200
},
{
"epoch": 0.07,
"learning_rate": 4.6589647332522605e-05,
"loss": 1.2565,
"step": 1300
},
{
"epoch": 0.07,
"learning_rate": 4.632527890868715e-05,
"loss": 1.258,
"step": 1400
},
{
"epoch": 0.08,
"learning_rate": 4.6060910484851693e-05,
"loss": 1.255,
"step": 1500
},
{
"epoch": 0.08,
"learning_rate": 4.579654206101623e-05,
"loss": 1.2449,
"step": 1600
},
{
"epoch": 0.09,
"learning_rate": 4.5532173637180775e-05,
"loss": 1.258,
"step": 1700
},
{
"epoch": 0.1,
"learning_rate": 4.526780521334532e-05,
"loss": 1.2365,
"step": 1800
},
{
"epoch": 0.1,
"learning_rate": 4.5003436789509864e-05,
"loss": 1.2506,
"step": 1900
},
{
"epoch": 0.11,
"learning_rate": 4.473906836567441e-05,
"loss": 1.2561,
"step": 2000
},
{
"epoch": 0.11,
"learning_rate": 4.4474699941838946e-05,
"loss": 1.2538,
"step": 2100
},
{
"epoch": 0.12,
"learning_rate": 4.421033151800349e-05,
"loss": 1.2522,
"step": 2200
},
{
"epoch": 0.12,
"learning_rate": 4.3945963094168035e-05,
"loss": 1.2524,
"step": 2300
},
{
"epoch": 0.13,
"learning_rate": 4.368159467033257e-05,
"loss": 1.2464,
"step": 2400
},
{
"epoch": 0.13,
"learning_rate": 4.3417226246497124e-05,
"loss": 1.2514,
"step": 2500
},
{
"epoch": 0.14,
"learning_rate": 4.315285782266166e-05,
"loss": 1.2429,
"step": 2600
},
{
"epoch": 0.14,
"learning_rate": 4.2888489398826206e-05,
"loss": 1.2481,
"step": 2700
},
{
"epoch": 0.15,
"learning_rate": 4.262412097499075e-05,
"loss": 1.242,
"step": 2800
},
{
"epoch": 0.15,
"learning_rate": 4.235975255115529e-05,
"loss": 1.2489,
"step": 2900
},
{
"epoch": 0.16,
"learning_rate": 4.209538412731984e-05,
"loss": 1.236,
"step": 3000
},
{
"epoch": 0.16,
"learning_rate": 4.183101570348438e-05,
"loss": 1.2447,
"step": 3100
},
{
"epoch": 0.17,
"learning_rate": 4.1566647279648915e-05,
"loss": 1.2449,
"step": 3200
},
{
"epoch": 0.17,
"learning_rate": 4.1302278855813466e-05,
"loss": 1.2452,
"step": 3300
},
{
"epoch": 0.18,
"learning_rate": 4.1037910431978004e-05,
"loss": 1.2367,
"step": 3400
},
{
"epoch": 0.18,
"learning_rate": 4.077354200814255e-05,
"loss": 1.2414,
"step": 3500
},
{
"epoch": 0.19,
"learning_rate": 4.050917358430709e-05,
"loss": 1.2377,
"step": 3600
},
{
"epoch": 0.2,
"learning_rate": 4.024480516047163e-05,
"loss": 1.2365,
"step": 3700
},
{
"epoch": 0.2,
"learning_rate": 3.998043673663618e-05,
"loss": 1.2461,
"step": 3800
},
{
"epoch": 0.21,
"learning_rate": 3.971606831280072e-05,
"loss": 1.2499,
"step": 3900
},
{
"epoch": 0.21,
"learning_rate": 3.9451699888965264e-05,
"loss": 1.2328,
"step": 4000
},
{
"epoch": 0.22,
"learning_rate": 3.918733146512981e-05,
"loss": 1.2347,
"step": 4100
},
{
"epoch": 0.22,
"learning_rate": 3.8922963041294346e-05,
"loss": 1.2482,
"step": 4200
},
{
"epoch": 0.23,
"learning_rate": 3.865859461745889e-05,
"loss": 1.2127,
"step": 4300
},
{
"epoch": 0.23,
"learning_rate": 3.8394226193623434e-05,
"loss": 1.2296,
"step": 4400
},
{
"epoch": 0.24,
"learning_rate": 3.812985776978798e-05,
"loss": 1.2314,
"step": 4500
},
{
"epoch": 0.24,
"learning_rate": 3.786548934595252e-05,
"loss": 1.2111,
"step": 4600
},
{
"epoch": 0.25,
"learning_rate": 3.760112092211707e-05,
"loss": 1.2312,
"step": 4700
},
{
"epoch": 0.25,
"learning_rate": 3.7336752498281605e-05,
"loss": 1.2291,
"step": 4800
},
{
"epoch": 0.26,
"learning_rate": 3.707238407444615e-05,
"loss": 1.2356,
"step": 4900
},
{
"epoch": 0.26,
"learning_rate": 3.6808015650610694e-05,
"loss": 1.2411,
"step": 5000
},
{
"epoch": 0.27,
"learning_rate": 3.654364722677523e-05,
"loss": 1.228,
"step": 5100
},
{
"epoch": 0.27,
"learning_rate": 3.627927880293978e-05,
"loss": 1.2384,
"step": 5200
},
{
"epoch": 0.28,
"learning_rate": 3.601491037910432e-05,
"loss": 1.231,
"step": 5300
},
{
"epoch": 0.29,
"learning_rate": 3.5750541955268865e-05,
"loss": 1.2378,
"step": 5400
},
{
"epoch": 0.29,
"learning_rate": 3.548617353143341e-05,
"loss": 1.2366,
"step": 5500
},
{
"epoch": 0.3,
"learning_rate": 3.522180510759795e-05,
"loss": 1.2356,
"step": 5600
},
{
"epoch": 0.3,
"learning_rate": 3.49574366837625e-05,
"loss": 1.2266,
"step": 5700
},
{
"epoch": 0.31,
"learning_rate": 3.4693068259927036e-05,
"loss": 1.2385,
"step": 5800
},
{
"epoch": 0.31,
"learning_rate": 3.4428699836091574e-05,
"loss": 1.2307,
"step": 5900
},
{
"epoch": 0.32,
"learning_rate": 3.4164331412256125e-05,
"loss": 1.2245,
"step": 6000
},
{
"epoch": 0.32,
"learning_rate": 3.389996298842066e-05,
"loss": 1.2311,
"step": 6100
},
{
"epoch": 0.33,
"learning_rate": 3.363559456458521e-05,
"loss": 1.2214,
"step": 6200
},
{
"epoch": 0.33,
"learning_rate": 3.337122614074975e-05,
"loss": 1.2196,
"step": 6300
},
{
"epoch": 0.34,
"learning_rate": 3.310685771691429e-05,
"loss": 1.2319,
"step": 6400
},
{
"epoch": 0.34,
"learning_rate": 3.284248929307884e-05,
"loss": 1.2291,
"step": 6500
},
{
"epoch": 0.35,
"learning_rate": 3.257812086924338e-05,
"loss": 1.2245,
"step": 6600
},
{
"epoch": 0.35,
"learning_rate": 3.231375244540792e-05,
"loss": 1.2134,
"step": 6700
},
{
"epoch": 0.36,
"learning_rate": 3.204938402157247e-05,
"loss": 1.2256,
"step": 6800
},
{
"epoch": 0.36,
"learning_rate": 3.1785015597737005e-05,
"loss": 1.2235,
"step": 6900
},
{
"epoch": 0.37,
"learning_rate": 3.152064717390155e-05,
"loss": 1.2269,
"step": 7000
},
{
"epoch": 0.38,
"learning_rate": 3.1256278750066093e-05,
"loss": 1.2315,
"step": 7100
},
{
"epoch": 0.38,
"learning_rate": 3.099191032623064e-05,
"loss": 1.2325,
"step": 7200
},
{
"epoch": 0.39,
"learning_rate": 3.072754190239518e-05,
"loss": 1.2165,
"step": 7300
},
{
"epoch": 0.39,
"learning_rate": 3.046317347855972e-05,
"loss": 1.2287,
"step": 7400
},
{
"epoch": 0.4,
"learning_rate": 3.0198805054724268e-05,
"loss": 1.2265,
"step": 7500
},
{
"epoch": 0.4,
"learning_rate": 2.9934436630888805e-05,
"loss": 1.224,
"step": 7600
},
{
"epoch": 0.41,
"learning_rate": 2.9670068207053353e-05,
"loss": 1.217,
"step": 7700
},
{
"epoch": 0.41,
"learning_rate": 2.9405699783217894e-05,
"loss": 1.2181,
"step": 7800
},
{
"epoch": 0.42,
"learning_rate": 2.9141331359382435e-05,
"loss": 1.2297,
"step": 7900
},
{
"epoch": 0.42,
"learning_rate": 2.887696293554698e-05,
"loss": 1.2222,
"step": 8000
},
{
"epoch": 0.43,
"learning_rate": 2.861259451171152e-05,
"loss": 1.2058,
"step": 8100
},
{
"epoch": 0.43,
"learning_rate": 2.834822608787607e-05,
"loss": 1.2119,
"step": 8200
},
{
"epoch": 0.44,
"learning_rate": 2.8086501348278964e-05,
"loss": 1.2316,
"step": 8300
},
{
"epoch": 0.44,
"learning_rate": 2.7822132924443505e-05,
"loss": 1.2241,
"step": 8400
},
{
"epoch": 0.45,
"learning_rate": 2.7557764500608046e-05,
"loss": 1.2189,
"step": 8500
},
{
"epoch": 0.45,
"learning_rate": 2.7293396076772594e-05,
"loss": 1.2146,
"step": 8600
},
{
"epoch": 0.46,
"learning_rate": 2.702902765293713e-05,
"loss": 1.2106,
"step": 8700
},
{
"epoch": 0.47,
"learning_rate": 2.676465922910168e-05,
"loss": 1.2073,
"step": 8800
},
{
"epoch": 0.47,
"learning_rate": 2.650029080526622e-05,
"loss": 1.2197,
"step": 8900
},
{
"epoch": 0.48,
"learning_rate": 2.623592238143076e-05,
"loss": 1.221,
"step": 9000
},
{
"epoch": 0.48,
"learning_rate": 2.5971553957595306e-05,
"loss": 1.2239,
"step": 9100
},
{
"epoch": 0.49,
"learning_rate": 2.5707185533759847e-05,
"loss": 1.2224,
"step": 9200
},
{
"epoch": 0.49,
"learning_rate": 2.5442817109924395e-05,
"loss": 1.2312,
"step": 9300
},
{
"epoch": 0.5,
"learning_rate": 2.5178448686088936e-05,
"loss": 1.2248,
"step": 9400
},
{
"epoch": 0.5,
"learning_rate": 2.4914080262253477e-05,
"loss": 1.2197,
"step": 9500
},
{
"epoch": 0.51,
"learning_rate": 2.464971183841802e-05,
"loss": 1.2362,
"step": 9600
},
{
"epoch": 0.51,
"learning_rate": 2.4385343414582562e-05,
"loss": 1.2029,
"step": 9700
},
{
"epoch": 0.52,
"learning_rate": 2.4120974990747107e-05,
"loss": 1.2212,
"step": 9800
},
{
"epoch": 0.52,
"learning_rate": 2.3856606566911648e-05,
"loss": 1.2217,
"step": 9900
},
{
"epoch": 0.53,
"learning_rate": 2.3592238143076192e-05,
"loss": 1.2089,
"step": 10000
},
{
"epoch": 0.53,
"learning_rate": 2.3327869719240737e-05,
"loss": 1.2174,
"step": 10100
},
{
"epoch": 0.54,
"learning_rate": 2.3063501295405278e-05,
"loss": 1.2269,
"step": 10200
},
{
"epoch": 0.54,
"learning_rate": 2.2801776555808173e-05,
"loss": 1.2177,
"step": 10300
},
{
"epoch": 0.55,
"learning_rate": 2.2537408131972718e-05,
"loss": 1.23,
"step": 10400
},
{
"epoch": 0.55,
"learning_rate": 2.2273039708137262e-05,
"loss": 1.2235,
"step": 10500
},
{
"epoch": 0.56,
"learning_rate": 2.2008671284301803e-05,
"loss": 1.2291,
"step": 10600
},
{
"epoch": 0.57,
"learning_rate": 2.1744302860466348e-05,
"loss": 1.2199,
"step": 10700
},
{
"epoch": 0.57,
"learning_rate": 2.147993443663089e-05,
"loss": 1.2101,
"step": 10800
},
{
"epoch": 0.58,
"learning_rate": 2.1215566012795433e-05,
"loss": 1.2239,
"step": 10900
},
{
"epoch": 0.58,
"learning_rate": 2.0951197588959974e-05,
"loss": 1.2105,
"step": 11000
},
{
"epoch": 0.59,
"learning_rate": 2.068682916512452e-05,
"loss": 1.2249,
"step": 11100
},
{
"epoch": 0.59,
"learning_rate": 2.0422460741289063e-05,
"loss": 1.2032,
"step": 11200
},
{
"epoch": 0.6,
"learning_rate": 2.0158092317453604e-05,
"loss": 1.2297,
"step": 11300
},
{
"epoch": 0.6,
"learning_rate": 1.9893723893618145e-05,
"loss": 1.2161,
"step": 11400
},
{
"epoch": 0.61,
"learning_rate": 1.962935546978269e-05,
"loss": 1.2199,
"step": 11500
},
{
"epoch": 0.61,
"learning_rate": 1.9364987045947234e-05,
"loss": 1.2221,
"step": 11600
},
{
"epoch": 0.62,
"learning_rate": 1.9100618622111775e-05,
"loss": 1.2171,
"step": 11700
},
{
"epoch": 0.62,
"learning_rate": 1.883625019827632e-05,
"loss": 1.2129,
"step": 11800
},
{
"epoch": 0.63,
"learning_rate": 1.857188177444086e-05,
"loss": 1.2165,
"step": 11900
},
{
"epoch": 0.63,
"learning_rate": 1.8307513350605405e-05,
"loss": 1.2229,
"step": 12000
},
{
"epoch": 0.64,
"learning_rate": 1.8043144926769946e-05,
"loss": 1.2186,
"step": 12100
},
{
"epoch": 0.64,
"learning_rate": 1.777877650293449e-05,
"loss": 1.2142,
"step": 12200
},
{
"epoch": 0.65,
"learning_rate": 1.7514408079099035e-05,
"loss": 1.207,
"step": 12300
},
{
"epoch": 0.66,
"learning_rate": 1.7250039655263576e-05,
"loss": 1.2164,
"step": 12400
},
{
"epoch": 0.66,
"learning_rate": 1.6985671231428117e-05,
"loss": 1.2245,
"step": 12500
},
{
"epoch": 0.67,
"learning_rate": 1.672130280759266e-05,
"loss": 1.2131,
"step": 12600
},
{
"epoch": 0.67,
"learning_rate": 1.645957806799556e-05,
"loss": 1.2144,
"step": 12700
},
{
"epoch": 0.68,
"learning_rate": 1.6195209644160105e-05,
"loss": 1.22,
"step": 12800
},
{
"epoch": 0.68,
"learning_rate": 1.5930841220324646e-05,
"loss": 1.2168,
"step": 12900
},
{
"epoch": 0.69,
"learning_rate": 1.5666472796489187e-05,
"loss": 1.2251,
"step": 13000
},
{
"epoch": 0.69,
"learning_rate": 1.540210437265373e-05,
"loss": 1.2185,
"step": 13100
},
{
"epoch": 0.7,
"learning_rate": 1.5137735948818274e-05,
"loss": 1.2222,
"step": 13200
},
{
"epoch": 0.7,
"learning_rate": 1.4873367524982817e-05,
"loss": 1.1998,
"step": 13300
},
{
"epoch": 0.71,
"learning_rate": 1.4608999101147361e-05,
"loss": 1.2237,
"step": 13400
},
{
"epoch": 0.71,
"learning_rate": 1.4344630677311902e-05,
"loss": 1.2265,
"step": 13500
},
{
"epoch": 0.72,
"learning_rate": 1.4080262253476445e-05,
"loss": 1.2253,
"step": 13600
},
{
"epoch": 0.72,
"learning_rate": 1.3815893829640988e-05,
"loss": 1.2144,
"step": 13700
},
{
"epoch": 0.73,
"learning_rate": 1.3551525405805532e-05,
"loss": 1.2199,
"step": 13800
},
{
"epoch": 0.73,
"learning_rate": 1.3287156981970075e-05,
"loss": 1.2057,
"step": 13900
},
{
"epoch": 0.74,
"learning_rate": 1.3022788558134616e-05,
"loss": 1.2192,
"step": 14000
},
{
"epoch": 0.75,
"learning_rate": 1.2758420134299159e-05,
"loss": 1.231,
"step": 14100
},
{
"epoch": 0.75,
"learning_rate": 1.2494051710463703e-05,
"loss": 1.2189,
"step": 14200
},
{
"epoch": 0.76,
"learning_rate": 1.2229683286628246e-05,
"loss": 1.2234,
"step": 14300
},
{
"epoch": 0.76,
"learning_rate": 1.1965314862792788e-05,
"loss": 1.2117,
"step": 14400
},
{
"epoch": 0.77,
"learning_rate": 1.1700946438957331e-05,
"loss": 1.2118,
"step": 14500
},
{
"epoch": 0.77,
"learning_rate": 1.1436578015121874e-05,
"loss": 1.2267,
"step": 14600
},
{
"epoch": 0.78,
"learning_rate": 1.1172209591286417e-05,
"loss": 1.2286,
"step": 14700
},
{
"epoch": 0.78,
"learning_rate": 1.0907841167450961e-05,
"loss": 1.2125,
"step": 14800
},
{
"epoch": 0.79,
"learning_rate": 1.0643472743615502e-05,
"loss": 1.2132,
"step": 14900
},
{
"epoch": 0.79,
"learning_rate": 1.0379104319780047e-05,
"loss": 1.2215,
"step": 15000
},
{
"epoch": 0.8,
"learning_rate": 1.011473589594459e-05,
"loss": 1.2182,
"step": 15100
},
{
"epoch": 0.8,
"learning_rate": 9.850367472109132e-06,
"loss": 1.2192,
"step": 15200
},
{
"epoch": 0.81,
"learning_rate": 9.585999048273675e-06,
"loss": 1.2185,
"step": 15300
},
{
"epoch": 0.81,
"learning_rate": 9.321630624438218e-06,
"loss": 1.2298,
"step": 15400
},
{
"epoch": 0.82,
"learning_rate": 9.05726220060276e-06,
"loss": 1.2297,
"step": 15500
},
{
"epoch": 0.82,
"learning_rate": 8.792893776767305e-06,
"loss": 1.2076,
"step": 15600
},
{
"epoch": 0.83,
"learning_rate": 8.528525352931846e-06,
"loss": 1.2244,
"step": 15700
},
{
"epoch": 0.83,
"learning_rate": 8.26415692909639e-06,
"loss": 1.2168,
"step": 15800
},
{
"epoch": 0.84,
"learning_rate": 8.002432189499287e-06,
"loss": 1.2245,
"step": 15900
},
{
"epoch": 0.85,
"learning_rate": 7.738063765663828e-06,
"loss": 1.2088,
"step": 16000
},
{
"epoch": 0.85,
"learning_rate": 7.473695341828373e-06,
"loss": 1.2214,
"step": 16100
},
{
"epoch": 0.86,
"learning_rate": 7.209326917992915e-06,
"loss": 1.2149,
"step": 16200
},
{
"epoch": 0.86,
"learning_rate": 6.944958494157458e-06,
"loss": 1.2143,
"step": 16300
},
{
"epoch": 0.87,
"learning_rate": 6.680590070322001e-06,
"loss": 1.2175,
"step": 16400
},
{
"epoch": 0.87,
"learning_rate": 6.416221646486544e-06,
"loss": 1.2132,
"step": 16500
},
{
"epoch": 0.88,
"learning_rate": 6.1518532226510865e-06,
"loss": 1.2299,
"step": 16600
},
{
"epoch": 0.88,
"learning_rate": 5.887484798815629e-06,
"loss": 1.2156,
"step": 16700
},
{
"epoch": 0.89,
"learning_rate": 5.623116374980172e-06,
"loss": 1.2037,
"step": 16800
},
{
"epoch": 0.89,
"learning_rate": 5.358747951144716e-06,
"loss": 1.2134,
"step": 16900
},
{
"epoch": 0.9,
"learning_rate": 5.094379527309258e-06,
"loss": 1.2253,
"step": 17000
},
{
"epoch": 0.9,
"learning_rate": 4.830011103473801e-06,
"loss": 1.2136,
"step": 17100
},
{
"epoch": 0.91,
"learning_rate": 4.565642679638344e-06,
"loss": 1.2039,
"step": 17200
},
{
"epoch": 0.91,
"learning_rate": 4.3012742558028865e-06,
"loss": 1.2187,
"step": 17300
},
{
"epoch": 0.92,
"learning_rate": 4.03690583196743e-06,
"loss": 1.2113,
"step": 17400
},
{
"epoch": 0.92,
"learning_rate": 3.772537408131973e-06,
"loss": 1.2124,
"step": 17500
},
{
"epoch": 0.93,
"learning_rate": 3.5081689842965156e-06,
"loss": 1.215,
"step": 17600
},
{
"epoch": 0.94,
"learning_rate": 3.2438005604610583e-06,
"loss": 1.2161,
"step": 17700
},
{
"epoch": 0.94,
"learning_rate": 2.9794321366256015e-06,
"loss": 1.224,
"step": 17800
},
{
"epoch": 0.95,
"learning_rate": 2.7150637127901447e-06,
"loss": 1.2201,
"step": 17900
},
{
"epoch": 0.95,
"learning_rate": 2.4506952889546874e-06,
"loss": 1.2173,
"step": 18000
},
{
"epoch": 0.96,
"learning_rate": 2.1863268651192306e-06,
"loss": 1.2074,
"step": 18100
},
{
"epoch": 0.96,
"learning_rate": 1.9219584412837733e-06,
"loss": 1.2178,
"step": 18200
},
{
"epoch": 0.97,
"learning_rate": 1.6575900174483158e-06,
"loss": 1.2064,
"step": 18300
},
{
"epoch": 0.97,
"learning_rate": 1.393221593612859e-06,
"loss": 1.2258,
"step": 18400
},
{
"epoch": 0.98,
"learning_rate": 1.128853169777402e-06,
"loss": 1.2236,
"step": 18500
}
],
"logging_steps": 100,
"max_steps": 18923,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"total_flos": 6.451372228608e+16,
"train_batch_size": 10,
"trial_name": null,
"trial_params": null
}