Add m21 weights and results

m21/benchmark_results.json

@@ -0,0 +1,540 @@
+{
+  "model_id": "m21",
+  "checkpoint": "results/m21/final.pt",
+  "trained_step": 15258,
+  "num_fewshot": 0,
+  "limit": null,
+  "results": {
+    "lambada_openai": {
+      "name": "lambada_openai",
+      "alias": "lambada_openai",
+      "sample_len": 5153,
+      "perplexity,none": 770.8014820424523,
+      "perplexity_stderr,none": 41.84263892067875,
+      "acc,none": 0.16572870172714924,
+      "acc_stderr,none": 0.005180415430035788
+    },
+    "hellaswag": {
+      "name": "hellaswag",
+      "alias": "hellaswag",
+      "sample_len": 10042,
+      "acc,none": 0.269169488149771,
+      "acc_stderr,none": 0.00442621765491807,
+      "acc_norm,none": 0.27434773949412466,
+      "acc_norm_stderr,none": 0.004452731272840024
+    },
+    "piqa": {
+      "name": "piqa",
+      "alias": "piqa",
+      "sample_len": 1838,
+      "acc,none": 0.5930359085963003,
+      "acc_stderr,none": 0.011462093919190149,
+      "acc_norm,none": 0.5810663764961915,
+      "acc_norm_stderr,none": 0.01151147311999497
+    },
+    "sciq": {
+      "name": "sciq",
+      "alias": "sciq",
+      "sample_len": 1000,
+      "acc,none": 0.68,
+      "acc_stderr,none": 0.014758652303574938,
+      "acc_norm,none": 0.594,
+      "acc_norm_stderr,none": 0.01553722643863458
+    },
+    "arc_easy": {
+      "name": "arc_easy",
+      "alias": "arc_easy",
+      "sample_len": 2376,
+      "acc,none": 0.45496632996632996,
+      "acc_stderr,none": 0.010218084454602478,
+      "acc_norm,none": 0.41708754208754206,
+      "acc_norm_stderr,none": 0.010117738967782014
+    },
+    "winogrande": {
+      "name": "winogrande",
+      "alias": "winogrande",
+      "sample_len": 1267,
+      "acc,none": 0.5217048145224941,
+      "acc_stderr,none": 0.014039239216484667
+    },
+    "blimp_adjunct_island": {
+      "name": "blimp_adjunct_island",
+      "alias": "blimp_adjunct_island",
+      "sample_len": 1000,
+      "acc,none": 0.79,
+      "acc_stderr,none": 0.012886662332274642
+    },
+    "blimp_anaphor_gender_agreement": {
+      "name": "blimp_anaphor_gender_agreement",
+      "alias": "blimp_anaphor_gender_agreement",
+      "sample_len": 1000,
+      "acc,none": 0.96,
+      "acc_stderr,none": 0.006199874066337028
+    },
+    "blimp_anaphor_number_agreement": {
+      "name": "blimp_anaphor_number_agreement",
+      "alias": "blimp_anaphor_number_agreement",
+      "sample_len": 1000,
+      "acc,none": 0.989,
+      "acc_stderr,none": 0.0032999833166078114
+    },
+    "blimp_animate_subject_passive": {
+      "name": "blimp_animate_subject_passive",
+      "alias": "blimp_animate_subject_passive",
+      "sample_len": 1000,
+      "acc,none": 0.734,
+      "acc_stderr,none": 0.013979965645145179
+    },
+    "blimp_animate_subject_trans": {
+      "name": "blimp_animate_subject_trans",
+      "alias": "blimp_animate_subject_trans",
+      "sample_len": 1000,
+      "acc,none": 0.915,
+      "acc_stderr,none": 0.008823426366942316
+    },
+    "blimp_causative": {
+      "name": "blimp_causative",
+      "alias": "blimp_causative",
+      "sample_len": 1000,
+      "acc,none": 0.717,
+      "acc_stderr,none": 0.01425181090648182
+    },
+    "blimp_complex_NP_island": {
+      "name": "blimp_complex_NP_island",
+      "alias": "blimp_complex_NP_island",
+      "sample_len": 1000,
+      "acc,none": 0.549,
+      "acc_stderr,none": 0.015743152379585443
+    },
+    "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": 0.45,
+      "acc_stderr,none": 0.015740004693383918
+    },
+    "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": 0.838,
+      "acc_stderr,none": 0.011657267771304384
+    },
+    "blimp_determiner_noun_agreement_1": {
+      "name": "blimp_determiner_noun_agreement_1",
+      "alias": "blimp_determiner_noun_agreement_1",
+      "sample_len": 1000,
+      "acc,none": 0.979,
+      "acc_stderr,none": 0.0045364721513065165
+    },
+    "blimp_determiner_noun_agreement_2": {
+      "name": "blimp_determiner_noun_agreement_2",
+      "alias": "blimp_determiner_noun_agreement_2",
+      "sample_len": 1000,
+      "acc,none": 0.947,
+      "acc_stderr,none": 0.00708810561724649
+    },
+    "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": 0.888,
+      "acc_stderr,none": 0.009977753031397177
+    },
+    "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": 0.929,
+      "acc_stderr,none": 0.008125578442487959
+    },
+    "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": 0.918,
+      "acc_stderr,none": 0.008680515615523732
+    },
+    "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": 0.851,
+      "acc_stderr,none": 0.011266140684632246
+    },
+    "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": 0.926,
+      "acc_stderr,none": 0.0082820645127041
+    },
+    "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": 0.924,
+      "acc_stderr,none": 0.008384169266796462
+    },
+    "blimp_distractor_agreement_relational_noun": {
+      "name": "blimp_distractor_agreement_relational_noun",
+      "alias": "blimp_distractor_agreement_relational_noun",
+      "sample_len": 1000,
+      "acc,none": 0.845,
+      "acc_stderr,none": 0.011450157470799522
+    },
+    "blimp_distractor_agreement_relative_clause": {
+      "name": "blimp_distractor_agreement_relative_clause",
+      "alias": "blimp_distractor_agreement_relative_clause",
+      "sample_len": 1000,
+      "acc,none": 0.692,
+      "acc_stderr,none": 0.01460648312734278
+    },
+    "blimp_drop_argument": {
+      "name": "blimp_drop_argument",
+      "alias": "blimp_drop_argument",
+      "sample_len": 1000,
+      "acc,none": 0.751,
+      "acc_stderr,none": 0.013681600278702275
+    },
+    "blimp_ellipsis_n_bar_1": {
+      "name": "blimp_ellipsis_n_bar_1",
+      "alias": "blimp_ellipsis_n_bar_1",
+      "sample_len": 1000,
+      "acc,none": 0.81,
+      "acc_stderr,none": 0.012411851354816254
+    },
+    "blimp_ellipsis_n_bar_2": {
+      "name": "blimp_ellipsis_n_bar_2",
+      "alias": "blimp_ellipsis_n_bar_2",
+      "sample_len": 1000,
+      "acc,none": 0.826,
+      "acc_stderr,none": 0.011994493230973447
+    },
+    "blimp_existential_there_object_raising": {
+      "name": "blimp_existential_there_object_raising",
+      "alias": "blimp_existential_there_object_raising",
+      "sample_len": 1000,
+      "acc,none": 0.831,
+      "acc_stderr,none": 0.011856625977890152
+    },
+    "blimp_existential_there_quantifiers_1": {
+      "name": "blimp_existential_there_quantifiers_1",
+      "alias": "blimp_existential_there_quantifiers_1",
+      "sample_len": 1000,
+      "acc,none": 0.973,
+      "acc_stderr,none": 0.005128089049275305
+    },
+    "blimp_existential_there_quantifiers_2": {
+      "name": "blimp_existential_there_quantifiers_2",
+      "alias": "blimp_existential_there_quantifiers_2",
+      "sample_len": 1000,
+      "acc,none": 0.188,
+      "acc_stderr,none": 0.01236158601510376
+    },
+    "blimp_existential_there_subject_raising": {
+      "name": "blimp_existential_there_subject_raising",
+      "alias": "blimp_existential_there_subject_raising",
+      "sample_len": 1000,
+      "acc,none": 0.858,
+      "acc_stderr,none": 0.011043457699378305
+    },
+    "blimp_expletive_it_object_raising": {
+      "name": "blimp_expletive_it_object_raising",
+      "alias": "blimp_expletive_it_object_raising",
+      "sample_len": 1000,
+      "acc,none": 0.75,
+      "acc_stderr,none": 0.013699915608779773
+    },
+    "blimp_inchoative": {
+      "name": "blimp_inchoative",
+      "alias": "blimp_inchoative",
+      "sample_len": 1000,
+      "acc,none": 0.66,
+      "acc_stderr,none": 0.014987482264363972
+    },
+    "blimp_intransitive": {
+      "name": "blimp_intransitive",
+      "alias": "blimp_intransitive",
+      "sample_len": 1000,
+      "acc,none": 0.765,
+      "acc_stderr,none": 0.013414729030247124
+    },
+    "blimp_irregular_past_participle_adjectives": {
+      "name": "blimp_irregular_past_participle_adjectives",
+      "alias": "blimp_irregular_past_participle_adjectives",
+      "sample_len": 1000,
+      "acc,none": 0.812,
+      "acc_stderr,none": 0.012361586015103758
+    },
+    "blimp_irregular_past_participle_verbs": {
+      "name": "blimp_irregular_past_participle_verbs",
+      "alias": "blimp_irregular_past_participle_verbs",
+      "sample_len": 1000,
+      "acc,none": 0.83,
+      "acc_stderr,none": 0.011884495834541793
+    },
+    "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": 0.909,
+      "acc_stderr,none": 0.009099549538400337
+    },
+    "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": 0.871,
+      "acc_stderr,none": 0.010605256784796466
+    },
+    "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": 0.535,
+      "acc_stderr,none": 0.015780495050030086
+    },
+    "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": 0.555,
+      "acc_stderr,none": 0.015723301886761007
+    },
+    "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": 0.481,
+      "acc_stderr,none": 0.015807874268505835
+    },
+    "blimp_npi_present_1": {
+      "name": "blimp_npi_present_1",
+      "alias": "blimp_npi_present_1",
+      "sample_len": 1000,
+      "acc,none": 0.503,
+      "acc_stderr,none": 0.01581901517924682
+    },
+    "blimp_npi_present_2": {
+      "name": "blimp_npi_present_2",
+      "alias": "blimp_npi_present_2",
+      "sample_len": 1000,
+      "acc,none": 0.609,
+      "acc_stderr,none": 0.015438826294681775
+    },
+    "blimp_only_npi_licensor_present": {
+      "name": "blimp_only_npi_licensor_present",
+      "alias": "blimp_only_npi_licensor_present",
+      "sample_len": 1000,
+      "acc,none": 0.954,
+      "acc_stderr,none": 0.006627814717380679
+    },
+    "blimp_only_npi_scope": {
+      "name": "blimp_only_npi_scope",
+      "alias": "blimp_only_npi_scope",
+      "sample_len": 1000,
+      "acc,none": 0.691,
+      "acc_stderr,none": 0.014619600977206347
+    },
+    "blimp_passive_1": {
+      "name": "blimp_passive_1",
+      "alias": "blimp_passive_1",
+      "sample_len": 1000,
+      "acc,none": 0.889,
+      "acc_stderr,none": 0.009938701010583716
+    },
+    "blimp_passive_2": {
+      "name": "blimp_passive_2",
+      "alias": "blimp_passive_2",
+      "sample_len": 1000,
+      "acc,none": 0.886,
+      "acc_stderr,none": 0.010055103435823278
+    },
+    "blimp_principle_A_c_command": {
+      "name": "blimp_principle_A_c_command",
+      "alias": "blimp_principle_A_c_command",
+      "sample_len": 1000,
+      "acc,none": 0.656,
+      "acc_stderr,none": 0.015029633724408919
+    },
+    "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": 0.96,
+      "acc_stderr,none": 0.006199874066337028
+    },
+    "blimp_principle_A_domain_1": {
+      "name": "blimp_principle_A_domain_1",
+      "alias": "blimp_principle_A_domain_1",
+      "sample_len": 1000,
+      "acc,none": 0.892,
+      "acc_stderr,none": 0.009820001651345708
+    },
+    "blimp_principle_A_domain_2": {
+      "name": "blimp_principle_A_domain_2",
+      "alias": "blimp_principle_A_domain_2",
+      "sample_len": 1000,
+      "acc,none": 0.753,
+      "acc_stderr,none": 0.013644675781314012
+    },
+    "blimp_principle_A_domain_3": {
+      "name": "blimp_principle_A_domain_3",
+      "alias": "blimp_principle_A_domain_3",
+      "sample_len": 1000,
+      "acc,none": 0.64,
+      "acc_stderr,none": 0.015186527932039985
+    },
+    "blimp_principle_A_reconstruction": {
+      "name": "blimp_principle_A_reconstruction",
+      "alias": "blimp_principle_A_reconstruction",
+      "sample_len": 1000,
+      "acc,none": 0.524,
+      "acc_stderr,none": 0.01580106558665186
+    },
+    "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": 0.891,
+      "acc_stderr,none": 0.009859828407037124
+    },
+    "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": 0.905,
+      "acc_stderr,none": 0.009276910103103353
+    },
+    "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": 0.993,
+      "acc_stderr,none": 0.002637794146243781
+    },
+    "blimp_sentential_negation_npi_scope": {
+      "name": "blimp_sentential_negation_npi_scope",
+      "alias": "blimp_sentential_negation_npi_scope",
+      "sample_len": 1000,
+      "acc,none": 0.748,
+      "acc_stderr,none": 0.013736254390651213
+    },
+    "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": 0.792,
+      "acc_stderr,none": 0.012841374572097058
+    },
+    "blimp_superlative_quantifiers_2": {
+      "name": "blimp_superlative_quantifiers_2",
+      "alias": "blimp_superlative_quantifiers_2",
+      "sample_len": 1000,
+      "acc,none": 0.688,
+      "acc_stderr,none": 0.014658474370509057
+    },
+    "blimp_tough_vs_raising_1": {
+      "name": "blimp_tough_vs_raising_1",
+      "alias": "blimp_tough_vs_raising_1",
+      "sample_len": 1000,
+      "acc,none": 0.553,
+      "acc_stderr,none": 0.015730176046009098
+    },
+    "blimp_tough_vs_raising_2": {
+      "name": "blimp_tough_vs_raising_2",
+      "alias": "blimp_tough_vs_raising_2",
+      "sample_len": 1000,
+      "acc,none": 0.867,
+      "acc_stderr,none": 0.010743669132397295
+    },
+    "blimp_transitive": {
+      "name": "blimp_transitive",
+      "alias": "blimp_transitive",
+      "sample_len": 1000,
+      "acc,none": 0.838,
+      "acc_stderr,none": 0.011657267771304384
+    },
+    "blimp_wh_island": {
+      "name": "blimp_wh_island",
+      "alias": "blimp_wh_island",
+      "sample_len": 1000,
+      "acc,none": 0.65,
+      "acc_stderr,none": 0.015090650341444127
+    },
+    "blimp_wh_questions_object_gap": {
+      "name": "blimp_wh_questions_object_gap",
+      "alias": "blimp_wh_questions_object_gap",
+      "sample_len": 1000,
+      "acc,none": 0.731,
+      "acc_stderr,none": 0.014029819522568203
+    },
+    "blimp_wh_questions_subject_gap": {
+      "name": "blimp_wh_questions_subject_gap",
+      "alias": "blimp_wh_questions_subject_gap",
+      "sample_len": 1000,
+      "acc,none": 0.914,
+      "acc_stderr,none": 0.008870325962594761
+    },
+    "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": 0.85,
+      "acc_stderr,none": 0.011297239823409418
+    },
+    "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": 0.964,
+      "acc_stderr,none": 0.00589395781616553
+    },
+    "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": 0.966,
+      "acc_stderr,none": 0.0057338361396953854
+    },
+    "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": 0.221,
+      "acc_stderr,none": 0.01312750285969618
+    },
+    "blimp": {
+      "alias": "blimp",
+      "name": "blimp",
+      "sample_len": 67000,
+      "acc,none": 0.7724626865671642,
+      "acc_stderr,none": 0.0014624167781541737,
+      "sample_count": {
+        "acc,none": 67000
+      }
+    }
+  }
+}

m21/config.json

@@ -0,0 +1,49 @@
+{
+  "model_id": "m21",
+  "run_name": "m21-gqa-rope-dense-muonv2",
+  "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": "gqa",
+  "gqa_num_kv_heads": 2,
+  "mla_latent_dim": 64,
+  "pos_type": "rope",
+  "rope_base": 10000,
+  "ffn_type": "dense",
+  "moe_num_experts": 8,
+  "moe_top_k": 2,
+  "optimizer": "muon_v2",
+  "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": 4,
+  "grad_accum_steps": 16,
+  "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"
+}

m21/eval_results.json

@@ -0,0 +1,212 @@
+{
+  "model_id": "m21",
+  "checkpoint": "/home/akshat/codemindv2/results/m21/final.pt",
+  "trained_step": 15258,
+  "total_params": 58774528,
+  "active_params": 58774528,
+  "flops_per_forward": 117549056.0,
+  "val_loss": 3.483479607105255,
+  "val_perplexity": 32.572865889283186,
+  "bits_per_token": 5.025598754208586,
+  "ppl_tokens_evaluated": 655360,
+  "prefill_tok_per_sec": 181761.25333696578,
+  "inference_tok_per_sec": 200.5319340782315,
+  "inference_tok_per_sec_kvcache": 215.48350833704265,
+  "kvcache_speedup": 1.0745595674202106,
+  "gen_tokens": 128,
+  "peak_infer_vram_gb": 0.366,
+  "peak_train_vram_gb": 3.993,
+  "closeness_answer_ppl": 34.072,
+  "closeness_hit_rate": 0.4,
+  "closeness_token_f1": 0.033,
+  "closeness_n": 20,
+  "closeness_items": [
+    {
+      "prompt": "The capital of France is",
+      "answer": " Paris",
+      "generated": " the capital of the United States",
+      "answer_nll": 4.6488,
+      "answer_ppl": 104.46,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The capital of Japan is",
+      "answer": " Tokyo",
+      "generated": " the capital of the United States",
+      "answer_nll": 2.6744,
+      "answer_ppl": 14.504,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The largest planet in the solar system is",
+      "answer": " Jupiter",
+      "generated": " the Sun.\nThe Sun",
+      "answer_nll": 3.9389,
+      "answer_ppl": 51.363,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Water is made of hydrogen and",
+      "answer": " oxygen",
+      "generated": " oxygen.\nThe water is",
+      "answer_nll": 0.6682,
+      "answer_ppl": 1.951,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The opposite of hot is",
+      "answer": " cold",
+      "generated": " the hot side.\nThe",
+      "answer_nll": 5.55,
+      "answer_ppl": 257.24,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The opposite of up is",
+      "answer": " down",
+      "generated": " the one that is not in",
+      "answer_nll": 5.9849,
+      "answer_ppl": 397.382,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Two plus two equals",
+      "answer": " four",
+      "generated": " one.\nThe first is",
+      "answer_nll": 2.8717,
+      "answer_ppl": 17.667,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The sun rises in the",
+      "answer": " east",
+      "generated": " sky and the moon rises in",
+      "answer_nll": 3.2926,
+      "answer_ppl": 26.913,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The first president of the United States was George",
+      "answer": " Washington",
+      "generated": " Washington. He was a man",
+      "answer_nll": 0.5498,
+      "answer_ppl": 1.733,
+      "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": 0.285,
+      "answer_ppl": 1.33,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "A triangle has three",
+      "answer": " sides",
+      "generated": " sides, the triangle is a",
+      "answer_nll": 1.6668,
+      "answer_ppl": 5.295,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The chemical symbol for gold is",
+      "answer": " Au",
+      "generated": " the gold symbol of the sun",
+      "answer_nll": 8.503,
+      "answer_ppl": 4929.396,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The freezing point of water in Celsius is",
+      "answer": " zero",
+      "generated": " the temperature of the water in",
+      "answer_nll": 5.5915,
+      "answer_ppl": 268.131,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The author of Romeo and Juliet is William",
+      "answer": " Shakespeare",
+      "generated": " Shakespeare\u2019s most famous",
+      "answer_nll": 0.6613,
+      "answer_ppl": 1.937,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The Earth revolves around the",
+      "answer": " Sun",
+      "generated": " sun, and the sun is",
+      "answer_nll": 1.2002,
+      "answer_ppl": 3.321,
+      "hit": true,
+      "token_f1": 0.333
+    },
+    {
+      "prompt": "The plural of mouse is",
+      "answer": " mice",
+      "generated": " the plural of the word mouse",
+      "answer_nll": 11.3772,
+      "answer_ppl": 87309.905,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "Photosynthesis occurs in the",
+      "answer": " leaves",
+      "generated": " plant, and the plant produces",
+      "answer_nll": 3.3627,
+      "answer_ppl": 28.868,
+      "hit": false,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The human heart pumps",
+      "answer": " blood",
+      "generated": " blood to the brain, and",
+      "answer_nll": 2.1132,
+      "answer_ppl": 8.275,
+      "hit": true,
+      "token_f1": 0.333
+    },
+    {
+      "prompt": "The currency used in the United States is the",
+      "answer": " dollar",
+      "generated": " US dollar.\nThe US",
+      "answer_nll": 3.0413,
+      "answer_ppl": 20.932,
+      "hit": true,
+      "token_f1": 0.0
+    },
+    {
+      "prompt": "The tallest animal in the world is the",
+      "answer": " giraffe",
+      "generated": " elephant.\nThe elephant is",
+      "answer_nll": 2.588,
+      "answer_ppl": 13.303,
+      "hit": false,
+      "token_f1": 0.0
+    }
+  ],
+  "generation_samples": {
+    "The most important thing about sleep is": "The most important thing about sleep is that it\u2019s a good idea to take some time to relax.\nIt\u2019s also a great idea to set aside time to focus on the day, because if you\u2019re having a bad night, it also helps to put back together the whole night or even go outside to",
+    "Scientists have discovered that": "Scientists have discovered that, in a recent study of about 20,000 New England men in India, the number of people with cancer in the US has doubled, with about one in five of those with a cancer diagnosis. The results of the study are based on studies conducted by the National Cancer Institute (NCI) at",
+    "In the history of mathematics,": "In the history of mathematics, the term has been used to describe the practice of algebraic thinking.\nThe word \u201cmath\u201d has become a term used in the United States, Canada, and the United Kingdom. It refers to the practice of solving problems involving algebra by means of rational algebra.\nThe term \u201c",
+    "Python code for prime number": "Python code for prime number input, for example:\n- I can see the value I want to be as a result of the value from the input. For example, I will write a python with a parameter of 1. And I will make a function from the output (and the\n- I can be just a function",
+    "The best way to learn a new language is": "The best way to learn a new language is to learn about it.\nRead more: How to Make a New Language\nRead more: How to Make a New Language\nThe best way to prepare your child for a new language is to start the conversation in your child\u2019s first language. This language can help learn new words, it"
+  }
+}

m21/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eae8f38483dd8526708a90ffdb826a300272b8014e544da5c7fcc32802247436
+size 235107624