trident-ec / scratch /g1marg_field_s2.json
farguney's picture
Upload scratch/g1marg_field_s2.json with huggingface_hub
02cd5cd verified
Raw
History Blame Contribute Delete
9.54 kB
{
"screen": "G1-FS16",
"arm": "field",
"seeds": [
2
],
"nonce": 0,
"steps": 4000,
"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": false,
"eps": 0.05,
"device": "cuda",
"runs": [
{
"seed": 2,
"oracle": {
"worst_acc": 1.0,
"worst_depth": -1,
"max_mid_entropy": 0.0010785266058519483,
"pass": true
},
"history": [
{
"acc_d11": 0.002,
"exact_d11": 0.0,
"acc_d12": 0.0078,
"exact_d12": 0.0,
"acc_d13": 0.0049,
"exact_d13": 0.0,
"acc_d14": 0.0049,
"exact_d14": 0.0,
"acc_d15": 0.0029,
"exact_d15": 0.0,
"acc_d16": 0.0,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0037434895833333335,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.051346133152643,
"heldout_min_exact": 0.0,
"singleton_len1": 0.5,
"distinct_codes": 0.0,
"code_map": "[None, 1]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 1.251,
"clip_actions": 10.448,
"clip_spine": 3942.616,
"step": 0,
"loss": 4072.126708984375,
"train_bits": 4028.18,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 1.1305,
"wall_s": 1.1
},
{
"acc_d11": 0.0039,
"exact_d11": 0.0,
"acc_d12": 0.0029,
"exact_d12": 0.0,
"acc_d13": 0.0098,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.0049,
"exact_d15": 0.0,
"acc_d16": 0.002,
"exact_d16": 0.0,
"heldout_byte_acc": 0.00439453125,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.247412701447805,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 10.63,
"clip_actions": 15.918,
"clip_spine": 729.545,
"step": 500,
"loss": 2223.355712890625,
"train_bits": 2203.15,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.3408,
"wall_s": 170.4
},
{
"acc_d11": 0.0039,
"exact_d11": 0.0,
"acc_d12": 0.0068,
"exact_d12": 0.0,
"acc_d13": 0.001,
"exact_d13": 0.0,
"acc_d14": 0.001,
"exact_d14": 0.0,
"acc_d15": 0.0059,
"exact_d15": 0.0,
"acc_d16": 0.0049,
"exact_d16": 0.0,
"heldout_byte_acc": 0.00390625,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.306143502394358,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 2.529,
"clip_actions": 0.0,
"clip_spine": 633.048,
"step": 1000,
"loss": 2281.115234375,
"train_bits": 2280.17,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.3373,
"wall_s": 337.3
},
{
"acc_d11": 0.0039,
"exact_d11": 0.0,
"acc_d12": 0.001,
"exact_d12": 0.0,
"acc_d13": 0.002,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.0029,
"exact_d15": 0.0,
"acc_d16": 0.001,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0022786458333333335,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.295728385448456,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 2.967,
"clip_actions": 0.0,
"clip_spine": 713.351,
"step": 1500,
"loss": 2094.02490234375,
"train_bits": 2092.79,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.3358,
"wall_s": 503.6
},
{
"acc_d11": 0.0049,
"exact_d11": 0.0,
"acc_d12": 0.0029,
"exact_d12": 0.0,
"acc_d13": 0.002,
"exact_d13": 0.0,
"acc_d14": 0.0029,
"exact_d14": 0.0,
"acc_d15": 0.002,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.0029296875,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.394025584061941,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 5.191,
"clip_actions": 0.0,
"clip_spine": 762.887,
"step": 2000,
"loss": 2307.814208984375,
"train_bits": 2305.64,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.335,
"wall_s": 670.1
},
{
"acc_d11": 0.0107,
"exact_d11": 0.0,
"acc_d12": 0.002,
"exact_d12": 0.0,
"acc_d13": 0.0117,
"exact_d13": 0.0,
"acc_d14": 0.001,
"exact_d14": 0.0,
"acc_d15": 0.0049,
"exact_d15": 0.0,
"acc_d16": 0.0049,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005859375,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.441218674182892,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 9.574,
"clip_actions": 32.119,
"clip_spine": 554.457,
"step": 2500,
"loss": 1979.371337890625,
"train_bits": 1969.32,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.3346,
"wall_s": 836.5
},
{
"acc_d11": 0.0059,
"exact_d11": 0.0,
"acc_d12": 0.002,
"exact_d12": 0.0,
"acc_d13": 0.0068,
"exact_d13": 0.0,
"acc_d14": 0.0039,
"exact_d14": 0.0,
"acc_d15": 0.0039,
"exact_d15": 0.0,
"acc_d16": 0.0049,
"exact_d16": 0.0,
"heldout_byte_acc": 0.004557291666666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.392551402250925,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 4.851,
"clip_actions": 0.0,
"clip_spine": 915.092,
"step": 3000,
"loss": 2289.947021484375,
"train_bits": 2288.45,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.3342,
"wall_s": 1002.6
},
{
"acc_d11": 0.0078,
"exact_d11": 0.0,
"acc_d12": 0.0049,
"exact_d12": 0.0,
"acc_d13": 0.0117,
"exact_d13": 0.0,
"acc_d14": 0.002,
"exact_d14": 0.0,
"acc_d15": 0.0039,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005533854166666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.471528689066568,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 30.471,
"clip_actions": 47.183,
"clip_spine": 397.678,
"step": 3500,
"loss": 1600.5592041015625,
"train_bits": 1581.74,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.334,
"wall_s": 1168.8
},
{
"acc_d11": 0.0088,
"exact_d11": 0.0,
"acc_d12": 0.002,
"exact_d12": 0.0,
"acc_d13": 0.0127,
"exact_d13": 0.0,
"acc_d14": 0.001,
"exact_d14": 0.0,
"acc_d15": 0.0059,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005533854166666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.458107014497122,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 7.555,
"clip_actions": 0.0,
"clip_spine": 211.669,
"step": 3999,
"loss": 2151.280029296875,
"train_bits": 2149.39,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.3338,
"wall_s": 1334.7
}
],
"final": {
"acc_d11": 0.0088,
"exact_d11": 0.0,
"acc_d12": 0.002,
"exact_d12": 0.0,
"acc_d13": 0.0127,
"exact_d13": 0.0,
"acc_d14": 0.001,
"exact_d14": 0.0,
"acc_d15": 0.0059,
"exact_d15": 0.0,
"acc_d16": 0.0029,
"exact_d16": 0.0,
"heldout_byte_acc": 0.005533854166666667,
"heldout_record_exact": 0.0,
"heldout_answer_bpb": 8.458107014497122,
"heldout_min_exact": 0.0,
"singleton_len1": 0.0,
"distinct_codes": 0.0,
"code_map": "[None, None]",
"empty_word_rate": 1.0,
"word_purity": 0.0,
"word_len_match": 0.0,
"clip_compiler": 7.555,
"clip_actions": 0.0,
"clip_spine": 211.669,
"step": 3999,
"loss": 2151.280029296875,
"train_bits": 2149.39,
"lam_mid": 0.0,
"cur_depth": 10,
"st": 1,
"s_per_step": 0.3338,
"wall_s": 1334.7
}
}
],
"partial": false,
"result_file": "scratch/g1marg_field_s2.json",
"off_law": {},
"certification_geometry": true
}