Upload folder using huggingface_hub

m02/benchmark_results.json

@@ -0,0 +1,540 @@
+{
+  "model_id": "m02",
+  "checkpoint": "results/m02/final.pt",
+  "trained_step": 15258,
+  "num_fewshot": 0,
+  "limit": null,
+  "results": {
+    "lambada_openai": {
+      "name": "lambada_openai",
+      "alias": "lambada_openai",
+      "sample_len": 5153,
+      "perplexity,none": 500.5320741005073,
+      "perplexity_stderr,none": 27.427331535429097,
+      "acc,none": 0.1785367746943528,
+      "acc_stderr,none": 0.005335436779871464
+    },
+    "hellaswag": {
+      "name": "hellaswag",
+      "alias": "hellaswag",
+      "sample_len": 10042,
+      "acc,none": 0.2753435570603465,
+      "acc_stderr,none": 0.004457743287380272,
+      "acc_norm,none": 0.2848038239394543,
+      "acc_norm_stderr,none": 0.004503985839042443
+    },
+    "piqa": {
+      "name": "piqa",
+      "alias": "piqa",
+      "sample_len": 1838,
+      "acc,none": 0.5995647442872688,
+      "acc_stderr,none": 0.01143219488356234,
+      "acc_norm,none": 0.5963003264417845,
+      "acc_norm_stderr,none": 0.011447407541749126
+    },
+    "sciq": {
+      "name": "sciq",
+      "alias": "sciq",
+      "sample_len": 1000,
+      "acc,none": 0.702,
+      "acc_stderr,none": 0.014470846741134585,
+      "acc_norm,none": 0.61,
+      "acc_norm_stderr,none": 0.01543172505386673
+    },
+    "arc_easy": {
+      "name": "arc_easy",
+      "alias": "arc_easy",
+      "sample_len": 2376,
+      "acc,none": 0.46380471380471383,
+      "acc_stderr,none": 0.01023286555034659,
+      "acc_norm,none": 0.42003367003367004,
+      "acc_norm_stderr,none": 0.010127718838529311
+    },
+    "winogrande": {
+      "name": "winogrande",
+      "alias": "winogrande",
+      "sample_len": 1267,
+      "acc,none": 0.5082872928176796,
+      "acc_stderr,none": 0.01405055532282417
+    },
+    "blimp_adjunct_island": {
+      "name": "blimp_adjunct_island",
+      "alias": "blimp_adjunct_island",
+      "sample_len": 1000,
+      "acc,none": 0.808,
+      "acc_stderr,none": 0.012461592646660028
+    },
+    "blimp_anaphor_gender_agreement": {
+      "name": "blimp_anaphor_gender_agreement",
+      "alias": "blimp_anaphor_gender_agreement",
+      "sample_len": 1000,
+      "acc,none": 0.894,
+      "acc_stderr,none": 0.00973955126578524
+    },
+    "blimp_anaphor_number_agreement": {
+      "name": "blimp_anaphor_number_agreement",
+      "alias": "blimp_anaphor_number_agreement",
+      "sample_len": 1000,
+      "acc,none": 0.971,
+      "acc_stderr,none": 0.005309160685756948
+    },
+    "blimp_animate_subject_passive": {
+      "name": "blimp_animate_subject_passive",
+      "alias": "blimp_animate_subject_passive",
+      "sample_len": 1000,
+      "acc,none": 0.731,
+      "acc_stderr,none": 0.014029819522568203
+    },
+    "blimp_animate_subject_trans": {
+      "name": "blimp_animate_subject_trans",
+      "alias": "blimp_animate_subject_trans",
+      "sample_len": 1000,
+      "acc,none": 0.908,
+      "acc_stderr,none": 0.009144376393151129
+    },
+    "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": 0.537,
+      "acc_stderr,none": 0.015775927227262426
+    },
+    "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": 0.442,
+      "acc_stderr,none": 0.015712507211864152
+    },
+    "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": 0.808,
+      "acc_stderr,none": 0.012461592646660028
+    },
+    "blimp_determiner_noun_agreement_1": {
+      "name": "blimp_determiner_noun_agreement_1",
+      "alias": "blimp_determiner_noun_agreement_1",
+      "sample_len": 1000,
+      "acc,none": 0.963,
+      "acc_stderr,none": 0.0059721576223896395
+    },
+    "blimp_determiner_noun_agreement_2": {
+      "name": "blimp_determiner_noun_agreement_2",
+      "alias": "blimp_determiner_noun_agreement_2",
+      "sample_len": 1000,
+      "acc,none": 0.94,
+      "acc_stderr,none": 0.007513751157474971
+    },
+    "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": 0.875,
+      "acc_stderr,none": 0.010463483381956722
+    },
+    "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": 0.93,
+      "acc_stderr,none": 0.008072494358323525
+    },
+    "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": 0.892,
+      "acc_stderr,none": 0.009820001651345708
+    },
+    "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": 0.855,
+      "acc_stderr,none": 0.0111399775178902
+    },
+    "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": 0.882,
+      "acc_stderr,none": 0.010206869264381718
+    },
+    "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": 0.908,
+      "acc_stderr,none": 0.009144376393151129
+    },
+    "blimp_distractor_agreement_relational_noun": {
+      "name": "blimp_distractor_agreement_relational_noun",
+      "alias": "blimp_distractor_agreement_relational_noun",
+      "sample_len": 1000,
+      "acc,none": 0.75,
+      "acc_stderr,none": 0.013699915608779773
+    },
+    "blimp_distractor_agreement_relative_clause": {
+      "name": "blimp_distractor_agreement_relative_clause",
+      "alias": "blimp_distractor_agreement_relative_clause",
+      "sample_len": 1000,
+      "acc,none": 0.533,
+      "acc_stderr,none": 0.015784807891138876
+    },
+    "blimp_drop_argument": {
+      "name": "blimp_drop_argument",
+      "alias": "blimp_drop_argument",
+      "sample_len": 1000,
+      "acc,none": 0.764,
+      "acc_stderr,none": 0.013434451402438602
+    },
+    "blimp_ellipsis_n_bar_1": {
+      "name": "blimp_ellipsis_n_bar_1",
+      "alias": "blimp_ellipsis_n_bar_1",
+      "sample_len": 1000,
+      "acc,none": 0.782,
+      "acc_stderr,none": 0.013063179040595235
+    },
+    "blimp_ellipsis_n_bar_2": {
+      "name": "blimp_ellipsis_n_bar_2",
+      "alias": "blimp_ellipsis_n_bar_2",
+      "sample_len": 1000,
+      "acc,none": 0.831,
+      "acc_stderr,none": 0.011856625977890152
+    },
+    "blimp_existential_there_object_raising": {
+      "name": "blimp_existential_there_object_raising",
+      "alias": "blimp_existential_there_object_raising",
+      "sample_len": 1000,
+      "acc,none": 0.78,
+      "acc_stderr,none": 0.013106173040661862
+    },
+    "blimp_existential_there_quantifiers_1": {
+      "name": "blimp_existential_there_quantifiers_1",
+      "alias": "blimp_existential_there_quantifiers_1",
+      "sample_len": 1000,
+      "acc,none": 0.97,
+      "acc_stderr,none": 0.0053971408290991564
+    },
+    "blimp_existential_there_quantifiers_2": {
+      "name": "blimp_existential_there_quantifiers_2",
+      "alias": "blimp_existential_there_quantifiers_2",
+      "sample_len": 1000,
+      "acc,none": 0.262,
+      "acc_stderr,none": 0.013912208651021217
+    },
+    "blimp_existential_there_subject_raising": {
+      "name": "blimp_existential_there_subject_raising",
+      "alias": "blimp_existential_there_subject_raising",
+      "sample_len": 1000,
+      "acc,none": 0.834,
+      "acc_stderr,none": 0.011772110370812133
+    },
+    "blimp_expletive_it_object_raising": {
+      "name": "blimp_expletive_it_object_raising",
+      "alias": "blimp_expletive_it_object_raising",
+      "sample_len": 1000,
+      "acc,none": 0.713,
+      "acc_stderr,none": 0.014312087053810001
+    },
+    "blimp_inchoative": {
+      "name": "blimp_inchoative",
+      "alias": "blimp_inchoative",
+      "sample_len": 1000,
+      "acc,none": 0.668,
+      "acc_stderr,none": 0.014899597242811567
+    },
+    "blimp_intransitive": {
+      "name": "blimp_intransitive",
+      "alias": "blimp_intransitive",
+      "sample_len": 1000,
+      "acc,none": 0.758,
+      "acc_stderr,none": 0.013550631705556003
+    },
+    "blimp_irregular_past_participle_adjectives": {
+      "name": "blimp_irregular_past_participle_adjectives",
+      "alias": "blimp_irregular_past_participle_adjectives",
+      "sample_len": 1000,
+      "acc,none": 0.937,
+      "acc_stderr,none": 0.007687007876286461
+    },
+    "blimp_irregular_past_participle_verbs": {
+      "name": "blimp_irregular_past_participle_verbs",
+      "alias": "blimp_irregular_past_participle_verbs",
+      "sample_len": 1000,
+      "acc,none": 0.865,
+      "acc_stderr,none": 0.010811655372416006
+    },
+    "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": 0.907,
+      "acc_stderr,none": 0.009188875634996669
+    },
+    "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": 0.907,
+      "acc_stderr,none": 0.009188875634996669
+    },
+    "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": 0.661,
+      "acc_stderr,none": 0.014976758771620224
+    },
+    "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": 0.529,
+      "acc_stderr,none": 0.015792669451628764
+    },
+    "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": 0.44,
+      "acc_stderr,none": 0.015704987954361718
+    },
+    "blimp_npi_present_1": {
+      "name": "blimp_npi_present_1",
+      "alias": "blimp_npi_present_1",
+      "sample_len": 1000,
+      "acc,none": 0.452,
+      "acc_stderr,none": 0.01574623586588058
+    },
+    "blimp_npi_present_2": {
+      "name": "blimp_npi_present_2",
+      "alias": "blimp_npi_present_2",
+      "sample_len": 1000,
+      "acc,none": 0.553,
+      "acc_stderr,none": 0.015730176046009098
+    },
+    "blimp_only_npi_licensor_present": {
+      "name": "blimp_only_npi_licensor_present",
+      "alias": "blimp_only_npi_licensor_present",
+      "sample_len": 1000,
+      "acc,none": 0.877,
+      "acc_stderr,none": 0.010391293421849803
+    },
+    "blimp_only_npi_scope": {
+      "name": "blimp_only_npi_scope",
+      "alias": "blimp_only_npi_scope",
+      "sample_len": 1000,
+      "acc,none": 0.461,
+      "acc_stderr,none": 0.015771104201283047
+    },
+    "blimp_passive_1": {
+      "name": "blimp_passive_1",
+      "alias": "blimp_passive_1",
+      "sample_len": 1000,
+      "acc,none": 0.875,
+      "acc_stderr,none": 0.010463483381956722
+    },
+    "blimp_passive_2": {
+      "name": "blimp_passive_2",
+      "alias": "blimp_passive_2",
+      "sample_len": 1000,
+      "acc,none": 0.872,
+      "acc_stderr,none": 0.010570133761108595
+    },
+    "blimp_principle_A_c_command": {
+      "name": "blimp_principle_A_c_command",
+      "alias": "blimp_principle_A_c_command",
+      "sample_len": 1000,
+      "acc,none": 0.595,
+      "acc_stderr,none": 0.01553113699045296
+    },
+    "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": 0.932,
+      "acc_stderr,none": 0.007964887911291624
+    },
+    "blimp_principle_A_domain_1": {
+      "name": "blimp_principle_A_domain_1",
+      "alias": "blimp_principle_A_domain_1",
+      "sample_len": 1000,
+      "acc,none": 0.933,
+      "acc_stderr,none": 0.007910345983177575
+    },
+    "blimp_principle_A_domain_2": {
+      "name": "blimp_principle_A_domain_2",
+      "alias": "blimp_principle_A_domain_2",
+      "sample_len": 1000,
+      "acc,none": 0.79,
+      "acc_stderr,none": 0.012886662332274642
+    },
+    "blimp_principle_A_domain_3": {
+      "name": "blimp_principle_A_domain_3",
+      "alias": "blimp_principle_A_domain_3",
+      "sample_len": 1000,
+      "acc,none": 0.651,
+      "acc_stderr,none": 0.015080663991563052
+    },
+    "blimp_principle_A_reconstruction": {
+      "name": "blimp_principle_A_reconstruction",
+      "alias": "blimp_principle_A_reconstruction",
+      "sample_len": 1000,
+      "acc,none": 0.358,
+      "acc_stderr,none": 0.015167928865407633
+    },
+    "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": 0.889,
+      "acc_stderr,none": 0.009938701010583716
+    },
+    "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": 0.87,
+      "acc_stderr,none": 0.010640169792499236
+    },
+    "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": 0.981,
+      "acc_stderr,none": 0.004319451082910613
+    },
+    "blimp_sentential_negation_npi_scope": {
+      "name": "blimp_sentential_negation_npi_scope",
+      "alias": "blimp_sentential_negation_npi_scope",
+      "sample_len": 1000,
+      "acc,none": 0.581,
+      "acc_stderr,none": 0.015610338967577953
+    },
+    "blimp_sentential_subject_island": {
+      "name": "blimp_sentential_subject_island",
+      "alias": "blimp_sentential_subject_island",
+      "sample_len": 1000,
+      "acc,none": 0.465,
+      "acc_stderr,none": 0.015780495050030086
+    },
+    "blimp_superlative_quantifiers_1": {
+      "name": "blimp_superlative_quantifiers_1",
+      "alias": "blimp_superlative_quantifiers_1",
+      "sample_len": 1000,
+      "acc,none": 0.893,
+      "acc_stderr,none": 0.009779910359847203
+    },
+    "blimp_superlative_quantifiers_2": {
+      "name": "blimp_superlative_quantifiers_2",
+      "alias": "blimp_superlative_quantifiers_2",
+      "sample_len": 1000,
+      "acc,none": 0.837,
+      "acc_stderr,none": 0.011686212712746913
+    },
+    "blimp_tough_vs_raising_1": {
+      "name": "blimp_tough_vs_raising_1",
+      "alias": "blimp_tough_vs_raising_1",
+      "sample_len": 1000,
+      "acc,none": 0.61,
+      "acc_stderr,none": 0.01543172505386673
+    },
+    "blimp_tough_vs_raising_2": {
+      "name": "blimp_tough_vs_raising_2",
+      "alias": "blimp_tough_vs_raising_2",
+      "sample_len": 1000,
+      "acc,none": 0.9,
+      "acc_stderr,none": 0.00949157995752499
+    },
+    "blimp_transitive": {
+      "name": "blimp_transitive",
+      "alias": "blimp_transitive",
+      "sample_len": 1000,
+      "acc,none": 0.82,
+      "acc_stderr,none": 0.012155153135512022
+    },
+    "blimp_wh_island": {
+      "name": "blimp_wh_island",
+      "alias": "blimp_wh_island",
+      "sample_len": 1000,
+      "acc,none": 0.793,
+      "acc_stderr,none": 0.012818553557844009
+    },
+    "blimp_wh_questions_object_gap": {
+      "name": "blimp_wh_questions_object_gap",
+      "alias": "blimp_wh_questions_object_gap",
+      "sample_len": 1000,
+      "acc,none": 0.674,
+      "acc_stderr,none": 0.014830507204541049
+    },
+    "blimp_wh_questions_subject_gap": {
+      "name": "blimp_wh_questions_subject_gap",
+      "alias": "blimp_wh_questions_subject_gap",
+      "sample_len": 1000,
+      "acc,none": 0.873,
+      "acc_stderr,none": 0.010534798620855644
+    },
+    "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": 0.813,
+      "acc_stderr,none": 0.012336254828074168
+    },
+    "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": 0.975,
+      "acc_stderr,none": 0.004939574819698493
+    },
+    "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": 0.986,
+      "acc_stderr,none": 0.003717232548256541
+    },
+    "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": 0.516,
+      "acc_stderr,none": 0.015811198373114985
+    },
+    "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": 0.217,
+      "acc_stderr,none": 0.013041513757270706
+    },
+    "blimp": {
+      "alias": "blimp",
+      "name": "blimp",
+      "sample_len": 67000,
+      "acc,none": 0.7609701492537313,
+      "acc_stderr,none": 0.0014798341123625034,
+      "sample_count": {
+        "acc,none": 67000
+      }
+    }
+  }
+}

m02/config.json

@@ -0,0 +1,50 @@
+{
+  "model_id": "m02",
+  "run_name": "m02-mha-learned-moe-adamw",
+  "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": "learned",
+  "rope_base": 10000,
+  "ffn_type": "moe",
+  "moe_num_experts": 8,
+  "moe_top_k": 2,
+  "optimizer": "adamw",
+  "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_aux_loss_coef": 0.01
+}

m02/eval_results.json

@@ -0,0 +1,223 @@
+{
+  "model_id": "m02",
+  "checkpoint": "results/m02/final.pt",
+  "trained_step": 15258,
+  "total_params": 240227328,
+  "active_params": 89232384,
+  "flops_per_forward": 178464768.0,
+  "val_loss": 3.5109181255102158,
+  "val_perplexity": 33.47899159386848,
+  "bits_per_token": 5.065184168640764,
+  "ppl_tokens_evaluated": 655360,
+  "prefill_tok_per_sec": 77578.77939608434,
+  "inference_tok_per_sec": 53.35199797114284,
+  "inference_tok_per_sec_kvcache": null,
+  "kvcache_speedup": null,
+  "gen_tokens": 128,
+  "peak_infer_vram_gb": 1.451,
+  "peak_train_vram_gb": 3.604,
+  "closeness_answer_ppl": 18.621,
+  "closeness_hit_rate": 0.55,
+  "closeness_token_f1": 0.05,
+  "closeness_n": 20,
+  "closeness_items": [
+    {
+      "prompt": "The capital of France is",
+      "answer": " Paris",
+      "generated": " the city of Paris.\n",
+      "answer_nll": 2.7305,
+      "answer_ppl": 15.341,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The capital of Japan is",
+      "answer": " Tokyo",
+      "generated": " Tokyo.\nThe capital of",
+      "answer_nll": 2.1854,
+      "answer_ppl": 8.894,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The largest planet in the solar system is",
+      "answer": " Jupiter",
+      "generated": " the Sun, which is about",
+      "answer_nll": 2.3684,
+      "answer_ppl": 10.68,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Water is made of hydrogen and",
+      "answer": " oxygen",
+      "generated": " oxygen.\nThe hydrogen in",
+      "answer_nll": 0.6836,
+      "answer_ppl": 1.981,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The opposite of hot is",
+      "answer": " cold",
+      "generated": " the hot.\nThe hot",
+      "answer_nll": 4.1665,
+      "answer_ppl": 64.489,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The opposite of up is",
+      "answer": " down",
+      "generated": " the one that is the most",
+      "answer_nll": 4.8774,
+      "answer_ppl": 131.285,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Two plus two equals",
+      "answer": " four",
+      "generated": " two.\n- The first",
+      "answer_nll": 4.1175,
+      "answer_ppl": 61.404,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The sun rises in the",
+      "answer": " east",
+      "generated": " sky, and the moon rises",
+      "answer_nll": 2.3503,
+      "answer_ppl": 10.489,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The first president of the United States was George",
+      "answer": " Washington",
+      "generated": " Washington. He was the first",
+      "answer_nll": 0.2887,
+      "answer_ppl": 1.335,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The speed of light is approximately 300,000 kilometers per",
+      "answer": " second",
+      "generated": " second. The speed of light",
+      "answer_nll": 0.2093,
+      "answer_ppl": 1.233,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "A triangle has three",
+      "answer": " sides",
+      "generated": " sides, and the sides are",
+      "answer_nll": 1.3377,
+      "answer_ppl": 3.81,
+      "hit": true,
+      "token_f1": 0.333
+    },
+    {
+      "prompt": "The chemical symbol for gold is",
+      "answer": " Au",
+      "generated": " gold. Gold is a color",
+      "answer_nll": 8.8131,
+      "answer_ppl": 6721.845,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The freezing point of water in Celsius is",
+      "answer": " zero",
+      "generated": " about 1.5 degrees Celsius",
+      "answer_nll": 5.6079,
+      "answer_ppl": 272.579,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The author of Romeo and Juliet is William",
+      "answer": " Shakespeare",
+      "generated": " Shakespeare. He is the author",
+      "answer_nll": 0.1459,
+      "answer_ppl": 1.157,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The Earth revolves around the",
+      "answer": " Sun",
+      "generated": " sun, and the sun is",
+      "answer_nll": 1.5873,
+      "answer_ppl": 4.89,
+      "hit": true,
+      "token_f1": 0.333
+    },
+    {
+      "prompt": "The plural of mouse is",
+      "answer": " mice",
+      "generated": " \u201cmice\u201d",
+      "answer_nll": 5.4431,
+      "answer_ppl": 231.157,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Photosynthesis occurs in the",
+      "answer": " leaves",
+      "generated": " absence of oxygen.\nThe",
+      "answer_nll": 4.9299,
+      "answer_ppl": 138.36,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The human heart pumps",
+      "answer": " blood",
+      "generated": " blood to the lungs, where",
+      "answer_nll": 1.1339,
+      "answer_ppl": 3.108,
+      "hit": true,
+      "token_f1": 0.333
+    },
+    {
+      "prompt": "The currency used in the United States is the",
+      "answer": " dollar",
+      "generated": " United States dollar. The United",
+      "answer_nll": 3.3442,
+      "answer_ppl": 28.337,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The tallest animal in the world is the",
+      "answer": " giraffe",
+      "generated": " elephant, which is the largest",
+      "answer_nll": 2.1657,
+      "answer_ppl": 8.721,
+      "hit": false,
+      "token_f1": 0.0
+    }
+  ],
+  "moe/aux_loss": 0.12001463086344302,
+  "moe/routing_entropy": 0.9973604219655196,
+  "moe/max_load_frac": 0.13966166178385417,
+  "moe/expert_0_prob": 0.12155533321201802,
+  "moe/expert_1_prob": 0.1292680971324444,
+  "moe/expert_2_prob": 0.12275821752846242,
+  "moe/expert_3_prob": 0.12823809832334518,
+  "moe/expert_4_prob": 0.12794600427150726,
+  "moe/expert_5_prob": 0.11847541145980359,
+  "moe/expert_6_prob": 0.1259801872074604,
+  "moe/expert_7_prob": 0.1257786512374878,
+  "generation_samples": {
+    "The most important thing about sleep is": "The most important thing about sleep is that it helps you sleep better.\n- Sleep Apnea is a condition in which the muscles in your mouth relax during sleep. The muscles in the mouth relax during sleep while you sleep.\n- Sleep Apnea is also called an atrial fibrillation. This happens when the heart rate is",
+    "Scientists have discovered that": "Scientists have discovered that when their muscles contract, it releases an amino acid called acetylcholine, which is why acetylcholine is involved in making muscles contract. This is why it's also used in many other applications, such as cooking and energy.\nTreatments & Supplements\nThere are two major",
+    "In the history of mathematics,": "In the history of mathematics, it is very much as one of the most successful and well-funded mathematical experiments in modern times. It has been proven to be very effective. It has been proved that the best combination of mathematics and mathematical reasoning is the best in terms of accuracy, completeness, and efficiency. Mathematics plays an immense",
+    "Python code for prime number": "Python code for prime number, and also for the array and the line of code for the whole number, and for the whole number of prime numbers, etc.\n- Get the vector representation of the data in the array.\n- Get the line of code that will be used to find all the prime numbers in one line",
+    "The best way to learn a new language is": "The best way to learn a new language is to start learning it, and then get the hang of it.\nCognitive behavioral therapy (CBT) is a form of therapy where a therapist focuses on a specific problem or condition to help them figure out ways to solve a problem. The therapist uses cognitive behavioral therapy to help patients learn new problem"
+  }
+}

m02/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84b20f0343e9f3e58a6fc68f505b273f34bed22d35130560f635c7a19da9d5c9
+size 960938984