trident-ec / g1c_no_opfield_s0.json
farguney's picture
Upload g1c_no_opfield_s0.json with huggingface_hub
f54b68a verified
Raw
History Blame Contribute Delete
16.9 kB
{
"screen": "G1-FS16",
"arm": "no_opfield",
"seeds": [
0
],
"nonce": 0,
"steps": 24000,
"events": 2,
"m_records": 16,
"d": 64,
"layers": 3,
"carrier": 256,
"train_depth": 10,
"closure": 16,
"lam_mid": 0.0,
"lam_cpu": 0.02,
"lam_a": 0.02,
"mid_warmup": 0.0,
"curriculum": 0.0,
"straight_through": true,
"st_after": 0.0,
"act_init": 4.0,
"sampled": true,
"eps": 0.05,
"device": "cuda",
"runs": [
{
"seed": 0,
"history": [
{
"acc_d11": 0.002,
"exact_d11": 0.0,
"acc_d12": 0.0049,
"exact_d12": 0.0,
"acc_d13": 0.0059,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.0029,
"exact_d15": 0.0,
"acc_d16": 0.0068,
"exact_d16": 0.0,
"heldout_byte_acc": 0.004231770833333333,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.016158431768417,
"heldout_min_exact": 0.0,
"step": 0,
"loss": 4100.154296875,
"train_bits": 4100.15,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.836,
"wall_s": 0.8
},
{
"acc_d11": 0.0156,
"exact_d11": 0.0,
"acc_d12": 0.0059,
"exact_d12": 0.0,
"acc_d13": 0.0,
"exact_d13": 0.0,
"acc_d14": 0.002,
"exact_d14": 0.0,
"acc_d15": 0.0039,
"exact_d15": 0.0,
"acc_d16": 0.0059,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005533854166666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.62705972790718,
"heldout_min_exact": 0.0,
"step": 1000,
"loss": 2286.45361328125,
"train_bits": 2286.45,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.024,
"wall_s": 24.0
},
{
"acc_d11": 0.0059,
"exact_d11": 0.0,
"acc_d12": 0.0049,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.002,
"exact_d14": 0.0,
"acc_d15": 0.0029,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.00390625,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.863070785999298,
"heldout_min_exact": 0.0,
"step": 2000,
"loss": 2274.506103515625,
"train_bits": 2274.51,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0237,
"wall_s": 47.5
},
{
"acc_d11": 0.001,
"exact_d11": 0.0,
"acc_d12": 0.001,
"exact_d12": 0.0,
"acc_d13": 0.0029,
"exact_d13": 0.0,
"acc_d14": 0.0059,
"exact_d14": 0.0,
"acc_d15": 0.002,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0026041666666666665,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.891308387120565,
"heldout_min_exact": 0.0,
"step": 3000,
"loss": 2248.608642578125,
"train_bits": 2248.61,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0235,
"wall_s": 70.5
},
{
"acc_d11": 0.0049,
"exact_d11": 0.0,
"acc_d12": 0.0039,
"exact_d12": 0.0,
"acc_d13": 0.0059,
"exact_d13": 0.0,
"acc_d14": 0.0039,
"exact_d14": 0.0,
"acc_d15": 0.0059,
"exact_d15": 0.0,
"acc_d16": 0.0049,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0048828125,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 9.105519513289133,
"heldout_min_exact": 0.0,
"step": 4000,
"loss": 2153.90478515625,
"train_bits": 2153.9,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0235,
"wall_s": 93.8
},
{
"acc_d11": 0.001,
"exact_d11": 0.0,
"acc_d12": 0.0059,
"exact_d12": 0.0,
"acc_d13": 0.0029,
"exact_d13": 0.0,
"acc_d14": 0.0068,
"exact_d14": 0.0,
"acc_d15": 0.0049,
"exact_d15": 0.0,
"acc_d16": 0.0098,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005208333333333333,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.943342685699463,
"heldout_min_exact": 0.0,
"step": 5000,
"loss": 2028.9041748046875,
"train_bits": 2028.9,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0235,
"wall_s": 117.4
},
{
"acc_d11": 0.0039,
"exact_d11": 0.0,
"acc_d12": 0.0029,
"exact_d12": 0.0,
"acc_d13": 0.0039,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.0029,
"exact_d15": 0.0,
"acc_d16": 0.0049,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0035807291666666665,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.620874206225077,
"heldout_min_exact": 0.0,
"step": 6000,
"loss": 1838.67041015625,
"train_bits": 1838.67,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0234,
"wall_s": 140.6
},
{
"acc_d11": 0.0029,
"exact_d11": 0.0,
"acc_d12": 0.0039,
"exact_d12": 0.0,
"acc_d13": 0.0029,
"exact_d13": 0.0,
"acc_d14": 0.0127,
"exact_d14": 0.0,
"acc_d15": 0.002,
"exact_d15": 0.0,
"acc_d16": 0.0039,
"exact_d16": 0.0,
"heldout_byte_acc": 0.004720052083333333,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.80271648367246,
"heldout_min_exact": 0.0,
"step": 7000,
"loss": 1470.447021484375,
"train_bits": 1470.45,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0234,
"wall_s": 163.7
},
{
"acc_d11": 0.0107,
"exact_d11": 0.0,
"acc_d12": 0.0068,
"exact_d12": 0.0,
"acc_d13": 0.0098,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.001,
"exact_d15": 0.0,
"acc_d16": 0.0049,
"exact_d16": 0.0,
"heldout_byte_acc": 0.006022135416666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.817947417497635,
"heldout_min_exact": 0.0,
"step": 8000,
"loss": 1249.823974609375,
"train_bits": 1249.82,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0234,
"wall_s": 186.9
},
{
"acc_d11": 0.001,
"exact_d11": 0.0,
"acc_d12": 0.0039,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.0029,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0030924479166666665,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.857758621374765,
"heldout_min_exact": 0.0,
"step": 9000,
"loss": 1129.73779296875,
"train_bits": 1129.74,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0233,
"wall_s": 209.8
},
{
"acc_d11": 0.0039,
"exact_d11": 0.0,
"acc_d12": 0.0039,
"exact_d12": 0.0,
"acc_d13": 0.0039,
"exact_d13": 0.0,
"acc_d14": 0.0059,
"exact_d14": 0.0,
"acc_d15": 0.0059,
"exact_d15": 0.0,
"acc_d16": 0.0039,
"exact_d16": 0.0,
"heldout_byte_acc": 0.004557291666666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.519497702519098,
"heldout_min_exact": 0.0,
"step": 10000,
"loss": 1094.3458251953125,
"train_bits": 1094.35,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0233,
"wall_s": 232.7
},
{
"acc_d11": 0.0039,
"exact_d11": 0.0,
"acc_d12": 0.0049,
"exact_d12": 0.0,
"acc_d13": 0.0039,
"exact_d13": 0.0,
"acc_d14": 0.0049,
"exact_d14": 0.0,
"acc_d15": 0.0049,
"exact_d15": 0.0,
"acc_d16": 0.0039,
"exact_d16": 0.0,
"heldout_byte_acc": 0.00439453125,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.331887116034826,
"heldout_min_exact": 0.0,
"step": 11000,
"loss": 1116.6572265625,
"train_bits": 1116.66,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0233,
"wall_s": 255.8
},
{
"acc_d11": 0.0049,
"exact_d11": 0.0,
"acc_d12": 0.0049,
"exact_d12": 0.0,
"acc_d13": 0.0029,
"exact_d13": 0.0,
"acc_d14": 0.0039,
"exact_d14": 0.0,
"acc_d15": 0.002,
"exact_d15": 0.0,
"acc_d16": 0.0068,
"exact_d16": 0.0,
"heldout_byte_acc": 0.004231770833333333,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.09334409236908,
"heldout_min_exact": 0.0,
"step": 12000,
"loss": 2243.900146484375,
"train_bits": 2243.9,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 278.7
},
{
"acc_d11": 0.0039,
"exact_d11": 0.0,
"acc_d12": 0.001,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.0049,
"exact_d14": 0.0,
"acc_d15": 0.0049,
"exact_d15": 0.0,
"acc_d16": 0.001,
"exact_d16": 0.0,
"heldout_byte_acc": 0.00341796875,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.15005138516426,
"heldout_min_exact": 0.0,
"step": 13000,
"loss": 2163.795166015625,
"train_bits": 2163.8,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 301.7
},
{
"acc_d11": 0.0137,
"exact_d11": 0.0,
"acc_d12": 0.0039,
"exact_d12": 0.0,
"acc_d13": 0.001,
"exact_d13": 0.0,
"acc_d14": 0.0098,
"exact_d14": 0.0,
"acc_d15": 0.0098,
"exact_d15": 0.0,
"acc_d16": 0.0039,
"exact_d16": 0.0,
"heldout_byte_acc": 0.006998697916666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.278612116972605,
"heldout_min_exact": 0.0,
"step": 14000,
"loss": 2170.215576171875,
"train_bits": 2170.22,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 324.8
},
{
"acc_d11": 0.0039,
"exact_d11": 0.0,
"acc_d12": 0.0029,
"exact_d12": 0.0,
"acc_d13": 0.001,
"exact_d13": 0.0,
"acc_d14": 0.002,
"exact_d14": 0.0,
"acc_d15": 0.001,
"exact_d15": 0.0,
"acc_d16": 0.002,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0021158854166666665,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.59625118970871,
"heldout_min_exact": 0.0,
"step": 15000,
"loss": 1973.7393798828125,
"train_bits": 1973.74,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 347.7
},
{
"acc_d11": 0.0078,
"exact_d11": 0.0,
"acc_d12": 0.002,
"exact_d12": 0.0,
"acc_d13": 0.0029,
"exact_d13": 0.0,
"acc_d14": 0.0039,
"exact_d14": 0.0,
"acc_d15": 0.0029,
"exact_d15": 0.0,
"acc_d16": 0.0078,
"exact_d16": 0.0,
"heldout_byte_acc": 0.004557291666666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.647134840488434,
"heldout_min_exact": 0.0,
"step": 16000,
"loss": 1666.372802734375,
"train_bits": 1666.37,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 370.6
},
{
"acc_d11": 0.002,
"exact_d11": 0.0,
"acc_d12": 0.0059,
"exact_d12": 0.0,
"acc_d13": 0.0088,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.0039,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.00439453125,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.88350823521614,
"heldout_min_exact": 0.0,
"step": 17000,
"loss": 1464.0228271484375,
"train_bits": 1464.02,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0231,
"wall_s": 393.2
},
{
"acc_d11": 0.0117,
"exact_d11": 0.0,
"acc_d12": 0.0039,
"exact_d12": 0.0,
"acc_d13": 0.0068,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.0039,
"exact_d15": 0.0,
"acc_d16": 0.0,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0048828125,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.35377949476242,
"heldout_min_exact": 0.0,
"step": 18000,
"loss": 1211.3798828125,
"train_bits": 1211.38,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0231,
"wall_s": 416.7
},
{
"acc_d11": 0.0059,
"exact_d11": 0.0,
"acc_d12": 0.0059,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.0039,
"exact_d14": 0.0,
"acc_d15": 0.0039,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.004557291666666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.358020315567652,
"heldout_min_exact": 0.0,
"step": 19000,
"loss": 1133.9022216796875,
"train_bits": 1133.9,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0231,
"wall_s": 439.8
},
{
"acc_d11": 0.0078,
"exact_d11": 0.0,
"acc_d12": 0.0029,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.002,
"exact_d14": 0.0,
"acc_d15": 0.0039,
"exact_d15": 0.0,
"acc_d16": 0.0039,
"exact_d16": 0.0,
"heldout_byte_acc": 0.004231770833333333,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.295753786961237,
"heldout_min_exact": 0.0,
"step": 20000,
"loss": 1105.0831298828125,
"train_bits": 1105.08,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 463.2
},
{
"acc_d11": 0.0029,
"exact_d11": 0.0,
"acc_d12": 0.002,
"exact_d12": 0.0,
"acc_d13": 0.002,
"exact_d13": 0.0,
"acc_d14": 0.0039,
"exact_d14": 0.0,
"acc_d15": 0.001,
"exact_d15": 0.0,
"acc_d16": 0.002,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0022786458333333335,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.445337603489557,
"heldout_min_exact": 0.0,
"step": 21000,
"loss": 1081.3262939453125,
"train_bits": 1081.33,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 486.6
},
{
"acc_d11": 0.0127,
"exact_d11": 0.0,
"acc_d12": 0.0039,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.001,
"exact_d15": 0.0,
"acc_d16": 0.0059,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005208333333333333,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.416951010624567,
"heldout_min_exact": 0.0,
"step": 22000,
"loss": 1071.9287109375,
"train_bits": 1071.93,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 510.7
},
{
"acc_d11": 0.0098,
"exact_d11": 0.0,
"acc_d12": 0.0068,
"exact_d12": 0.0,
"acc_d13": 0.002,
"exact_d13": 0.0,
"acc_d14": 0.0049,
"exact_d14": 0.0,
"acc_d15": 0.002,
"exact_d15": 0.0,
"acc_d16": 0.0039,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0048828125,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.462893078724543,
"heldout_min_exact": 0.0,
"step": 23000,
"loss": 2197.716796875,
"train_bits": 2197.72,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0232,
"wall_s": 534.6
},
{
"acc_d11": 0.0049,
"exact_d11": 0.0,
"acc_d12": 0.0049,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.0059,
"exact_d14": 0.0,
"acc_d15": 0.0068,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005045572916666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.498466481765112,
"heldout_min_exact": 0.0,
"step": 23999,
"loss": 1956.5263671875,
"train_bits": 1956.53,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0233,
"wall_s": 558.3
}
],
"final": {
"acc_d11": 0.0049,
"exact_d11": 0.0,
"acc_d12": 0.0049,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.0059,
"exact_d14": 0.0,
"acc_d15": 0.0068,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005045572916666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.498466481765112,
"heldout_min_exact": 0.0,
"step": 23999,
"loss": 1956.5263671875,
"train_bits": 1956.53,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.0233,
"wall_s": 558.3
}
}
],
"partial": false
}