Upload folder using huggingface_hub

m16/benchmark_results.json

@@ -0,0 +1,540 @@
+{
+  "model_id": "m16",
+  "checkpoint": "results/m16/final.pt",
+  "trained_step": 15258,
+  "num_fewshot": 0,
+  "limit": null,
+  "results": {
+    "lambada_openai": {
+      "name": "lambada_openai",
+      "alias": "lambada_openai",
+      "sample_len": 5153,
+      "perplexity,none": 488.6423869861212,
+      "perplexity_stderr,none": 27.277809731578422,
+      "acc,none": 0.18183582379196583,
+      "acc_stderr,none": 0.005373682797180109
+    },
+    "hellaswag": {
+      "name": "hellaswag",
+      "alias": "hellaswag",
+      "sample_len": 10042,
+      "acc,none": 0.27813184624576776,
+      "acc_stderr,none": 0.0044716295468955935,
+      "acc_norm,none": 0.2913762198765186,
+      "acc_norm_stderr,none": 0.004534677750103097
+    },
+    "piqa": {
+      "name": "piqa",
+      "alias": "piqa",
+      "sample_len": 1838,
+      "acc,none": 0.6202393906420022,
+      "acc_stderr,none": 0.011323483504715666,
+      "acc_norm,none": 0.6082698585418934,
+      "acc_norm_stderr,none": 0.011389038597720108
+    },
+    "sciq": {
+      "name": "sciq",
+      "alias": "sciq",
+      "sample_len": 1000,
+      "acc,none": 0.679,
+      "acc_stderr,none": 0.01477082181793475,
+      "acc_norm,none": 0.584,
+      "acc_norm_stderr,none": 0.015594460144140522
+    },
+    "arc_easy": {
+      "name": "arc_easy",
+      "alias": "arc_easy",
+      "sample_len": 2376,
+      "acc,none": 0.48484848484848486,
+      "acc_stderr,none": 0.01025507179453148,
+      "acc_norm,none": 0.43686868686868685,
+      "acc_norm_stderr,none": 0.0101776729281577
+    },
+    "winogrande": {
+      "name": "winogrande",
+      "alias": "winogrande",
+      "sample_len": 1267,
+      "acc,none": 0.5303867403314917,
+      "acc_stderr,none": 0.014026510839428848
+    },
+    "blimp_adjunct_island": {
+      "name": "blimp_adjunct_island",
+      "alias": "blimp_adjunct_island",
+      "sample_len": 1000,
+      "acc,none": 0.814,
+      "acc_stderr,none": 0.012310790208412926
+    },
+    "blimp_anaphor_gender_agreement": {
+      "name": "blimp_anaphor_gender_agreement",
+      "alias": "blimp_anaphor_gender_agreement",
+      "sample_len": 1000,
+      "acc,none": 0.928,
+      "acc_stderr,none": 0.008178195576218687
+    },
+    "blimp_anaphor_number_agreement": {
+      "name": "blimp_anaphor_number_agreement",
+      "alias": "blimp_anaphor_number_agreement",
+      "sample_len": 1000,
+      "acc,none": 0.969,
+      "acc_stderr,none": 0.005483527064679181
+    },
+    "blimp_animate_subject_passive": {
+      "name": "blimp_animate_subject_passive",
+      "alias": "blimp_animate_subject_passive",
+      "sample_len": 1000,
+      "acc,none": 0.694,
+      "acc_stderr,none": 0.014580006055436898
+    },
+    "blimp_animate_subject_trans": {
+      "name": "blimp_animate_subject_trans",
+      "alias": "blimp_animate_subject_trans",
+      "sample_len": 1000,
+      "acc,none": 0.907,
+      "acc_stderr,none": 0.009188875634996669
+    },
+    "blimp_causative": {
+      "name": "blimp_causative",
+      "alias": "blimp_causative",
+      "sample_len": 1000,
+      "acc,none": 0.708,
+      "acc_stderr,none": 0.014385511563477433
+    },
+    "blimp_complex_NP_island": {
+      "name": "blimp_complex_NP_island",
+      "alias": "blimp_complex_NP_island",
+      "sample_len": 1000,
+      "acc,none": 0.521,
+      "acc_stderr,none": 0.015805341148131185
+    },
+    "blimp_coordinate_structure_constraint_complex_left_branch": {
+      "name": "blimp_coordinate_structure_constraint_complex_left_branch",
+      "alias": "blimp_coordinate_structure_constraint_complex_left_branch",
+      "sample_len": 1000,
+      "acc,none": 0.487,
+      "acc_stderr,none": 0.015813952101896636
+    },
+    "blimp_coordinate_structure_constraint_object_extraction": {
+      "name": "blimp_coordinate_structure_constraint_object_extraction",
+      "alias": "blimp_coordinate_structure_constraint_object_extraction",
+      "sample_len": 1000,
+      "acc,none": 0.786,
+      "acc_stderr,none": 0.012975838021968847
+    },
+    "blimp_determiner_noun_agreement_1": {
+      "name": "blimp_determiner_noun_agreement_1",
+      "alias": "blimp_determiner_noun_agreement_1",
+      "sample_len": 1000,
+      "acc,none": 0.959,
+      "acc_stderr,none": 0.006273624021118764
+    },
+    "blimp_determiner_noun_agreement_2": {
+      "name": "blimp_determiner_noun_agreement_2",
+      "alias": "blimp_determiner_noun_agreement_2",
+      "sample_len": 1000,
+      "acc,none": 0.933,
+      "acc_stderr,none": 0.007910345983177575
+    },
+    "blimp_determiner_noun_agreement_irregular_1": {
+      "name": "blimp_determiner_noun_agreement_irregular_1",
+      "alias": "blimp_determiner_noun_agreement_irregular_1",
+      "sample_len": 1000,
+      "acc,none": 0.9,
+      "acc_stderr,none": 0.00949157995752499
+    },
+    "blimp_determiner_noun_agreement_irregular_2": {
+      "name": "blimp_determiner_noun_agreement_irregular_2",
+      "alias": "blimp_determiner_noun_agreement_irregular_2",
+      "sample_len": 1000,
+      "acc,none": 0.938,
+      "acc_stderr,none": 0.007629823996280269
+    },
+    "blimp_determiner_noun_agreement_with_adj_2": {
+      "name": "blimp_determiner_noun_agreement_with_adj_2",
+      "alias": "blimp_determiner_noun_agreement_with_adj_2",
+      "sample_len": 1000,
+      "acc,none": 0.889,
+      "acc_stderr,none": 0.009938701010583716
+    },
+    "blimp_determiner_noun_agreement_with_adj_irregular_1": {
+      "name": "blimp_determiner_noun_agreement_with_adj_irregular_1",
+      "alias": "blimp_determiner_noun_agreement_with_adj_irregular_1",
+      "sample_len": 1000,
+      "acc,none": 0.856,
+      "acc_stderr,none": 0.01110798754893916
+    },
+    "blimp_determiner_noun_agreement_with_adj_irregular_2": {
+      "name": "blimp_determiner_noun_agreement_with_adj_irregular_2",
+      "alias": "blimp_determiner_noun_agreement_with_adj_irregular_2",
+      "sample_len": 1000,
+      "acc,none": 0.879,
+      "acc_stderr,none": 0.010318210380946179
+    },
+    "blimp_determiner_noun_agreement_with_adjective_1": {
+      "name": "blimp_determiner_noun_agreement_with_adjective_1",
+      "alias": "blimp_determiner_noun_agreement_with_adjective_1",
+      "sample_len": 1000,
+      "acc,none": 0.919,
+      "acc_stderr,none": 0.008632121032139986
+    },
+    "blimp_distractor_agreement_relational_noun": {
+      "name": "blimp_distractor_agreement_relational_noun",
+      "alias": "blimp_distractor_agreement_relational_noun",
+      "sample_len": 1000,
+      "acc,none": 0.833,
+      "acc_stderr,none": 0.011800434324644593
+    },
+    "blimp_distractor_agreement_relative_clause": {
+      "name": "blimp_distractor_agreement_relative_clause",
+      "alias": "blimp_distractor_agreement_relative_clause",
+      "sample_len": 1000,
+      "acc,none": 0.744,
+      "acc_stderr,none": 0.013807775152234145
+    },
+    "blimp_drop_argument": {
+      "name": "blimp_drop_argument",
+      "alias": "blimp_drop_argument",
+      "sample_len": 1000,
+      "acc,none": 0.761,
+      "acc_stderr,none": 0.013493000446937704
+    },
+    "blimp_ellipsis_n_bar_1": {
+      "name": "blimp_ellipsis_n_bar_1",
+      "alias": "blimp_ellipsis_n_bar_1",
+      "sample_len": 1000,
+      "acc,none": 0.817,
+      "acc_stderr,none": 0.012233587399477745
+    },
+    "blimp_ellipsis_n_bar_2": {
+      "name": "blimp_ellipsis_n_bar_2",
+      "alias": "blimp_ellipsis_n_bar_2",
+      "sample_len": 1000,
+      "acc,none": 0.832,
+      "acc_stderr,none": 0.011828605831454352
+    },
+    "blimp_existential_there_object_raising": {
+      "name": "blimp_existential_there_object_raising",
+      "alias": "blimp_existential_there_object_raising",
+      "sample_len": 1000,
+      "acc,none": 0.805,
+      "acc_stderr,none": 0.012535235623319256
+    },
+    "blimp_existential_there_quantifiers_1": {
+      "name": "blimp_existential_there_quantifiers_1",
+      "alias": "blimp_existential_there_quantifiers_1",
+      "sample_len": 1000,
+      "acc,none": 0.969,
+      "acc_stderr,none": 0.005483527064679181
+    },
+    "blimp_existential_there_quantifiers_2": {
+      "name": "blimp_existential_there_quantifiers_2",
+      "alias": "blimp_existential_there_quantifiers_2",
+      "sample_len": 1000,
+      "acc,none": 0.121,
+      "acc_stderr,none": 0.010318210380946179
+    },
+    "blimp_existential_there_subject_raising": {
+      "name": "blimp_existential_there_subject_raising",
+      "alias": "blimp_existential_there_subject_raising",
+      "sample_len": 1000,
+      "acc,none": 0.832,
+      "acc_stderr,none": 0.011828605831454352
+    },
+    "blimp_expletive_it_object_raising": {
+      "name": "blimp_expletive_it_object_raising",
+      "alias": "blimp_expletive_it_object_raising",
+      "sample_len": 1000,
+      "acc,none": 0.728,
+      "acc_stderr,none": 0.014078856992462642
+    },
+    "blimp_inchoative": {
+      "name": "blimp_inchoative",
+      "alias": "blimp_inchoative",
+      "sample_len": 1000,
+      "acc,none": 0.702,
+      "acc_stderr,none": 0.014470846741134585
+    },
+    "blimp_intransitive": {
+      "name": "blimp_intransitive",
+      "alias": "blimp_intransitive",
+      "sample_len": 1000,
+      "acc,none": 0.767,
+      "acc_stderr,none": 0.01337497251921999
+    },
+    "blimp_irregular_past_participle_adjectives": {
+      "name": "blimp_irregular_past_participle_adjectives",
+      "alias": "blimp_irregular_past_participle_adjectives",
+      "sample_len": 1000,
+      "acc,none": 0.813,
+      "acc_stderr,none": 0.012336254828074168
+    },
+    "blimp_irregular_past_participle_verbs": {
+      "name": "blimp_irregular_past_participle_verbs",
+      "alias": "blimp_irregular_past_participle_verbs",
+      "sample_len": 1000,
+      "acc,none": 0.815,
+      "acc_stderr,none": 0.012285191326386656
+    },
+    "blimp_irregular_plural_subject_verb_agreement_1": {
+      "name": "blimp_irregular_plural_subject_verb_agreement_1",
+      "alias": "blimp_irregular_plural_subject_verb_agreement_1",
+      "sample_len": 1000,
+      "acc,none": 0.914,
+      "acc_stderr,none": 0.008870325962594761
+    },
+    "blimp_irregular_plural_subject_verb_agreement_2": {
+      "name": "blimp_irregular_plural_subject_verb_agreement_2",
+      "alias": "blimp_irregular_plural_subject_verb_agreement_2",
+      "sample_len": 1000,
+      "acc,none": 0.87,
+      "acc_stderr,none": 0.010640169792499236
+    },
+    "blimp_left_branch_island_echo_question": {
+      "name": "blimp_left_branch_island_echo_question",
+      "alias": "blimp_left_branch_island_echo_question",
+      "sample_len": 1000,
+      "acc,none": 0.536,
+      "acc_stderr,none": 0.015778243024904673
+    },
+    "blimp_left_branch_island_simple_question": {
+      "name": "blimp_left_branch_island_simple_question",
+      "alias": "blimp_left_branch_island_simple_question",
+      "sample_len": 1000,
+      "acc,none": 0.609,
+      "acc_stderr,none": 0.015438826294681775
+    },
+    "blimp_matrix_question_npi_licensor_present": {
+      "name": "blimp_matrix_question_npi_licensor_present",
+      "alias": "blimp_matrix_question_npi_licensor_present",
+      "sample_len": 1000,
+      "acc,none": 0.548,
+      "acc_stderr,none": 0.01574623586588058
+    },
+    "blimp_npi_present_1": {
+      "name": "blimp_npi_present_1",
+      "alias": "blimp_npi_present_1",
+      "sample_len": 1000,
+      "acc,none": 0.578,
+      "acc_stderr,none": 0.015625625112620622
+    },
+    "blimp_npi_present_2": {
+      "name": "blimp_npi_present_2",
+      "alias": "blimp_npi_present_2",
+      "sample_len": 1000,
+      "acc,none": 0.623,
+      "acc_stderr,none": 0.0153331701257798
+    },
+    "blimp_only_npi_licensor_present": {
+      "name": "blimp_only_npi_licensor_present",
+      "alias": "blimp_only_npi_licensor_present",
+      "sample_len": 1000,
+      "acc,none": 0.978,
+      "acc_stderr,none": 0.004640855259274689
+    },
+    "blimp_only_npi_scope": {
+      "name": "blimp_only_npi_scope",
+      "alias": "blimp_only_npi_scope",
+      "sample_len": 1000,
+      "acc,none": 0.732,
+      "acc_stderr,none": 0.014013292702729519
+    },
+    "blimp_passive_1": {
+      "name": "blimp_passive_1",
+      "alias": "blimp_passive_1",
+      "sample_len": 1000,
+      "acc,none": 0.849,
+      "acc_stderr,none": 0.011328165223341657
+    },
+    "blimp_passive_2": {
+      "name": "blimp_passive_2",
+      "alias": "blimp_passive_2",
+      "sample_len": 1000,
+      "acc,none": 0.805,
+      "acc_stderr,none": 0.012535235623319256
+    },
+    "blimp_principle_A_c_command": {
+      "name": "blimp_principle_A_c_command",
+      "alias": "blimp_principle_A_c_command",
+      "sample_len": 1000,
+      "acc,none": 0.717,
+      "acc_stderr,none": 0.01425181090648182
+    },
+    "blimp_principle_A_case_1": {
+      "name": "blimp_principle_A_case_1",
+      "alias": "blimp_principle_A_case_1",
+      "sample_len": 1000,
+      "acc,none": 1.0,
+      "acc_stderr,none": 0.0
+    },
+    "blimp_principle_A_case_2": {
+      "name": "blimp_principle_A_case_2",
+      "alias": "blimp_principle_A_case_2",
+      "sample_len": 1000,
+      "acc,none": 0.884,
+      "acc_stderr,none": 0.010131468138756922
+    },
+    "blimp_principle_A_domain_1": {
+      "name": "blimp_principle_A_domain_1",
+      "alias": "blimp_principle_A_domain_1",
+      "sample_len": 1000,
+      "acc,none": 0.906,
+      "acc_stderr,none": 0.009233052000787672
+    },
+    "blimp_principle_A_domain_2": {
+      "name": "blimp_principle_A_domain_2",
+      "alias": "blimp_principle_A_domain_2",
+      "sample_len": 1000,
+      "acc,none": 0.672,
+      "acc_stderr,none": 0.01485384248727031
+    },
+    "blimp_principle_A_domain_3": {
+      "name": "blimp_principle_A_domain_3",
+      "alias": "blimp_principle_A_domain_3",
+      "sample_len": 1000,
+      "acc,none": 0.612,
+      "acc_stderr,none": 0.015417317979911216
+    },
+    "blimp_principle_A_reconstruction": {
+      "name": "blimp_principle_A_reconstruction",
+      "alias": "blimp_principle_A_reconstruction",
+      "sample_len": 1000,
+      "acc,none": 0.319,
+      "acc_stderr,none": 0.014746404865473364
+    },
+    "blimp_regular_plural_subject_verb_agreement_1": {
+      "name": "blimp_regular_plural_subject_verb_agreement_1",
+      "alias": "blimp_regular_plural_subject_verb_agreement_1",
+      "sample_len": 1000,
+      "acc,none": 0.811,
+      "acc_stderr,none": 0.01238678458811779
+    },
+    "blimp_regular_plural_subject_verb_agreement_2": {
+      "name": "blimp_regular_plural_subject_verb_agreement_2",
+      "alias": "blimp_regular_plural_subject_verb_agreement_2",
+      "sample_len": 1000,
+      "acc,none": 0.842,
+      "acc_stderr,none": 0.011539894677559635
+    },
+    "blimp_sentential_negation_npi_licensor_present": {
+      "name": "blimp_sentential_negation_npi_licensor_present",
+      "alias": "blimp_sentential_negation_npi_licensor_present",
+      "sample_len": 1000,
+      "acc,none": 0.978,
+      "acc_stderr,none": 0.004640855259274689
+    },
+    "blimp_sentential_negation_npi_scope": {
+      "name": "blimp_sentential_negation_npi_scope",
+      "alias": "blimp_sentential_negation_npi_scope",
+      "sample_len": 1000,
+      "acc,none": 0.566,
+      "acc_stderr,none": 0.01568087656637504
+    },
+    "blimp_sentential_subject_island": {
+      "name": "blimp_sentential_subject_island",
+      "alias": "blimp_sentential_subject_island",
+      "sample_len": 1000,
+      "acc,none": 0.415,
+      "acc_stderr,none": 0.015589035185604594
+    },
+    "blimp_superlative_quantifiers_1": {
+      "name": "blimp_superlative_quantifiers_1",
+      "alias": "blimp_superlative_quantifiers_1",
+      "sample_len": 1000,
+      "acc,none": 0.937,
+      "acc_stderr,none": 0.007687007876286461
+    },
+    "blimp_superlative_quantifiers_2": {
+      "name": "blimp_superlative_quantifiers_2",
+      "alias": "blimp_superlative_quantifiers_2",
+      "sample_len": 1000,
+      "acc,none": 0.765,
+      "acc_stderr,none": 0.013414729030247124
+    },
+    "blimp_tough_vs_raising_1": {
+      "name": "blimp_tough_vs_raising_1",
+      "alias": "blimp_tough_vs_raising_1",
+      "sample_len": 1000,
+      "acc,none": 0.676,
+      "acc_stderr,none": 0.014806864733738986
+    },
+    "blimp_tough_vs_raising_2": {
+      "name": "blimp_tough_vs_raising_2",
+      "alias": "blimp_tough_vs_raising_2",
+      "sample_len": 1000,
+      "acc,none": 0.891,
+      "acc_stderr,none": 0.009859828407037124
+    },
+    "blimp_transitive": {
+      "name": "blimp_transitive",
+      "alias": "blimp_transitive",
+      "sample_len": 1000,
+      "acc,none": 0.818,
+      "acc_stderr,none": 0.0122075806376622
+    },
+    "blimp_wh_island": {
+      "name": "blimp_wh_island",
+      "alias": "blimp_wh_island",
+      "sample_len": 1000,
+      "acc,none": 0.758,
+      "acc_stderr,none": 0.013550631705556003
+    },
+    "blimp_wh_questions_object_gap": {
+      "name": "blimp_wh_questions_object_gap",
+      "alias": "blimp_wh_questions_object_gap",
+      "sample_len": 1000,
+      "acc,none": 0.749,
+      "acc_stderr,none": 0.013718133516888775
+    },
+    "blimp_wh_questions_subject_gap": {
+      "name": "blimp_wh_questions_subject_gap",
+      "alias": "blimp_wh_questions_subject_gap",
+      "sample_len": 1000,
+      "acc,none": 0.905,
+      "acc_stderr,none": 0.009276910103103353
+    },
+    "blimp_wh_questions_subject_gap_long_distance": {
+      "name": "blimp_wh_questions_subject_gap_long_distance",
+      "alias": "blimp_wh_questions_subject_gap_long_distance",
+      "sample_len": 1000,
+      "acc,none": 0.838,
+      "acc_stderr,none": 0.011657267771304384
+    },
+    "blimp_wh_vs_that_no_gap": {
+      "name": "blimp_wh_vs_that_no_gap",
+      "alias": "blimp_wh_vs_that_no_gap",
+      "sample_len": 1000,
+      "acc,none": 0.956,
+      "acc_stderr,none": 0.006488921798427387
+    },
+    "blimp_wh_vs_that_no_gap_long_distance": {
+      "name": "blimp_wh_vs_that_no_gap_long_distance",
+      "alias": "blimp_wh_vs_that_no_gap_long_distance",
+      "sample_len": 1000,
+      "acc,none": 0.927,
+      "acc_stderr,none": 0.008230354715243976
+    },
+    "blimp_wh_vs_that_with_gap": {
+      "name": "blimp_wh_vs_that_with_gap",
+      "alias": "blimp_wh_vs_that_with_gap",
+      "sample_len": 1000,
+      "acc,none": 0.609,
+      "acc_stderr,none": 0.015438826294681775
+    },
+    "blimp_wh_vs_that_with_gap_long_distance": {
+      "name": "blimp_wh_vs_that_with_gap_long_distance",
+      "alias": "blimp_wh_vs_that_with_gap_long_distance",
+      "sample_len": 1000,
+      "acc,none": 0.374,
+      "acc_stderr,none": 0.015308767369006505
+    },
+    "blimp": {
+      "alias": "blimp",
+      "name": "blimp",
+      "sample_len": 67000,
+      "acc,none": 0.7700447761194029,
+      "acc_stderr,none": 0.0014815512338215082,
+      "sample_count": {
+        "acc,none": 67000
+      }
+    }
+  }
+}

m16/config.json

@@ -0,0 +1,52 @@
+{
+  "model_id": "m16",
+  "run_name": "m16-mha-rope-fat-shared-moe-muon",
+  "data_dir": "data/fineweb-edu",
+  "vocab_size": 50257,
+  "block_size": 1024,
+  "d_model": 512,
+  "num_layers": 12,
+  "num_heads": 8,
+  "ffn_mult": 4,
+  "dropout": 0.0,
+  "bias": false,
+  "norm_type": "rmsnorm",
+  "tie_embeddings": true,
+  "attention_type": "mha",
+  "gqa_num_kv_heads": 2,
+  "mla_latent_dim": 64,
+  "pos_type": "rope",
+  "rope_base": 10000,
+  "ffn_type": "shared_moe",
+  "moe_num_experts": 8,
+  "moe_top_k": 2,
+  "optimizer": "muon",
+  "lr": 0.0006,
+  "min_lr": 6e-05,
+  "weight_decay": 0.1,
+  "beta1": 0.9,
+  "beta2": 0.95,
+  "grad_clip": 1.0,
+  "muon_lr": 0.02,
+  "muon_momentum": 0.95,
+  "training_tokens": 1000000000,
+  "warmup_tokens": 6000000,
+  "lr_schedule": "cosine",
+  "batch_size": 2,
+  "grad_accum_steps": 32,
+  "device": "cuda",
+  "dtype": "bfloat16",
+  "compile": true,
+  "seed": 1337,
+  "eval_interval": 250,
+  "eval_iters": 100,
+  "log_interval": 10,
+  "checkpoint_interval": 1000,
+  "results_dir": "results",
+  "mlflow_experiment": "slm-ablation",
+  "mlflow_tracking_uri": "sqlite:///mlflow.db",
+  "mlflow_artifact_location": "./mlartifacts",
+  "moe_num_shared_experts": 1,
+  "moe_shared_ffn_mult": 8,
+  "moe_aux_loss_coef": 0.01
+}

m16/eval_results.json

@@ -0,0 +1,223 @@
+{
+  "model_id": "m16",
+  "checkpoint": "results/m16/final.pt",
+  "trained_step": 15258,
+  "total_params": 290034688,
+  "active_params": 139039744,
+  "flops_per_forward": 278079488.0,
+  "val_loss": 3.45664686858654,
+  "val_perplexity": 31.710468678414106,
+  "bits_per_token": 4.986887295414166,
+  "ppl_tokens_evaluated": 655360,
+  "prefill_tok_per_sec": 57482.46046879679,
+  "inference_tok_per_sec": 41.88420904730769,
+  "inference_tok_per_sec_kvcache": 70.19531699040982,
+  "kvcache_speedup": 1.6759375093158637,
+  "gen_tokens": 128,
+  "peak_infer_vram_gb": 1.754,
+  "peak_train_vram_gb": 4.338,
+  "closeness_answer_ppl": 14.486,
+  "closeness_hit_rate": 0.65,
+  "closeness_token_f1": 0.064,
+  "closeness_n": 20,
+  "closeness_items": [
+    {
+      "prompt": "The capital of France is",
+      "answer": " Paris",
+      "generated": " Paris, and the capital of",
+      "answer_nll": 1.6583,
+      "answer_ppl": 5.25,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The capital of Japan is",
+      "answer": " Tokyo",
+      "generated": " Tokyo.\nThe city is",
+      "answer_nll": 1.607,
+      "answer_ppl": 4.988,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The largest planet in the solar system is",
+      "answer": " Jupiter",
+      "generated": " Jupiter, which is about the",
+      "answer_nll": 1.1213,
+      "answer_ppl": 3.069,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Water is made of hydrogen and",
+      "answer": " oxygen",
+      "generated": " oxygen.\nThe hydrogen is",
+      "answer_nll": 0.3562,
+      "answer_ppl": 1.428,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The opposite of hot is",
+      "answer": " cold",
+      "generated": " cold, and the cold is",
+      "answer_nll": 1.8448,
+      "answer_ppl": 6.327,
+      "hit": true,
+      "token_f1": 0.333
+    },
+    {
+      "prompt": "The opposite of up is",
+      "answer": " down",
+      "generated": " the fact that the two are",
+      "answer_nll": 7.4933,
+      "answer_ppl": 1795.958,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Two plus two equals",
+      "answer": " four",
+      "generated": " two.\n- The ratio",
+      "answer_nll": 3.8241,
+      "answer_ppl": 45.792,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The sun rises in the",
+      "answer": " east",
+      "generated": " east and sets in the west",
+      "answer_nll": 1.4976,
+      "answer_ppl": 4.471,
+      "hit": true,
+      "token_f1": 0.286
+    },
+    {
+      "prompt": "The first president of the United States was George",
+      "answer": " Washington",
+      "generated": " Washington, who was born in",
+      "answer_nll": 0.1429,
+      "answer_ppl": 1.154,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The speed of light is approximately 300,000 kilometers per",
+      "answer": " second",
+      "generated": " second.\nThe speed of",
+      "answer_nll": 0.0854,
+      "answer_ppl": 1.089,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "A triangle has three",
+      "answer": " sides",
+      "generated": " sides, and the sides of",
+      "answer_nll": 0.6984,
+      "answer_ppl": 2.011,
+      "hit": true,
+      "token_f1": 0.333
+    },
+    {
+      "prompt": "The chemical symbol for gold is",
+      "answer": " Au",
+      "generated": " gold. Gold is a precious",
+      "answer_nll": 5.97,
+      "answer_ppl": 391.516,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The freezing point of water in Celsius is",
+      "answer": " zero",
+      "generated": " the freezing point of water in",
+      "answer_nll": 6.6724,
+      "answer_ppl": 790.297,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The author of Romeo and Juliet is William",
+      "answer": " Shakespeare",
+      "generated": " Shakespeare.\nThe author of",
+      "answer_nll": 0.7232,
+      "answer_ppl": 2.061,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The Earth revolves around the",
+      "answer": " Sun",
+      "generated": " Sun, and the Earth is",
+      "answer_nll": 0.5369,
+      "answer_ppl": 1.711,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The plural of mouse is",
+      "answer": " mice",
+      "generated": " a word that is used to",
+      "answer_nll": 5.9564,
+      "answer_ppl": 386.222,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Photosynthesis occurs in the",
+      "answer": " leaves",
+      "generated": " form of photosynthesis.\n",
+      "answer_nll": 5.4259,
+      "answer_ppl": 227.222,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The human heart pumps",
+      "answer": " blood",
+      "generated": " blood through the body, and",
+      "answer_nll": 1.5057,
+      "answer_ppl": 4.507,
+      "hit": true,
+      "token_f1": 0.333
+    },
+    {
+      "prompt": "The currency used in the United States is the",
+      "answer": " dollar",
+      "generated": " U.S. dollar.",
+      "answer_nll": 3.8917,
+      "answer_ppl": 48.992,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The tallest animal in the world is the",
+      "answer": " giraffe",
+      "generated": " elephant.\nThe elephant is",
+      "answer_nll": 2.4518,
+      "answer_ppl": 11.609,
+      "hit": false,
+      "token_f1": 0.0
+    }
+  ],
+  "moe/aux_loss": 0.12057399628683925,
+  "moe/routing_entropy": 0.9932721691826979,
+  "moe/max_load_frac": 0.1440912882486979,
+  "moe/expert_0_prob": 0.12709741964936255,
+  "moe/expert_1_prob": 0.12377416417002678,
+  "moe/expert_2_prob": 0.12426500096917152,
+  "moe/expert_3_prob": 0.10939359031617642,
+  "moe/expert_4_prob": 0.13168378099799155,
+  "moe/expert_5_prob": 0.13501964434981345,
+  "moe/expert_6_prob": 0.12264941483736039,
+  "moe/expert_7_prob": 0.1261169958859682,
+  "generation_samples": {
+    "The most important thing about sleep is": "The most important thing about sleep is that it\u2019s a combination of many things.\n- It\u2019s not a matter of a single event.\n- It\u2019s about getting the right amount of sleep each night.\n- It\u2019s about the right amount of sleep each night.\n- It can",
+    "Scientists have discovered that": "Scientists have discovered that, in many ways, the human brain is actually a computer that is \"wired\" to \"learn\" to become a computer.\n\"These findings are the first of many to have a positive impact on how we think about the way we think about computers, how we understand how computers work, and",
+    "In the history of mathematics,": "In the history of mathematics, the term arithmetic began to be used as a synonym for arithmetic.\nThe word \u201calgebra\u201d has its origins in the Greek word \u201calgebra,\u201d (plus, minus) meaning problem solving. The term algebra is from the Greek word algebra, or \u201cal",
+    "Python code for prime number": "Python code for prime number to 5 decimal number. However, the program does have two solutions for prime and prime operations. One is for prime numbers (25) and the other for prime numbers (50).\nThe program can be run in almost any programming language. It includes both the\n- Python library (which allows for",
+    "The best way to learn a new language is": "The best way to learn a new language is to learn it in a fun and fun way. When you are learning a language, you should have a wide variety of skills and tools. You should be able to read a language without having to master the language in your head. Learning a new language is a great way to learn new things.\n"
+  }
+}

m16/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44293dbf2ffd26d5705179b90163359d4b5ef39e7a0f3750844a73d086fe7bf2
+size 1160171192