Push classification fine-tuned model

README.md

@@ -0,0 +1,98 @@
+---
+library_name: transformers
+license: apache-2.0
+base_model: distilbert/distilgpt2
+tags:
+- generated_from_trainer
+model-index:
+- name: clinical-field-mapper-classification
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# clinical-field-mapper-classification
+
+This model is a fine-tuned version of [distilbert/distilgpt2](https://huggingface.co/distilbert/distilgpt2) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.2783
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.0005
+- train_batch_size: 512
+- eval_batch_size: 512
+- seed: 42
+- distributed_type: multi-GPU
+- optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.01
+- num_epochs: 50
+- mixed_precision_training: Native AMP
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 7.615         | 1.0   | 18   | 5.8320          |
+| 5.7769        | 2.0   | 36   | 5.5586          |
+| 5.099         | 3.0   | 54   | 3.5             |
+| 2.3325        | 4.0   | 72   | 1.2383          |
+| 0.7197        | 5.0   | 90   | 0.6011          |
+| 0.2701        | 6.0   | 108  | 0.4207          |
+| 0.1317        | 7.0   | 126  | 0.3694          |
+| 0.0744        | 8.0   | 144  | 0.3362          |
+| 0.0522        | 9.0   | 162  | 0.3208          |
+| 0.039         | 10.0  | 180  | 0.2910          |
+| 0.0301        | 11.0  | 198  | 0.2922          |
+| 0.0239        | 12.0  | 216  | 0.2900          |
+| 0.0198        | 13.0  | 234  | 0.2935          |
+| 0.0174        | 14.0  | 252  | 0.2825          |
+| 0.0149        | 15.0  | 270  | 0.2788          |
+| 0.0135        | 16.0  | 288  | 0.2842          |
+| 0.0122        | 17.0  | 306  | 0.2825          |
+| 0.0103        | 18.0  | 324  | 0.2788          |
+| 0.0097        | 19.0  | 342  | 0.2776          |
+| 0.0083        | 20.0  | 360  | 0.2798          |
+| 0.0075        | 21.0  | 378  | 0.2793          |
+| 0.0071        | 22.0  | 396  | 0.2788          |
+| 0.0067        | 23.0  | 414  | 0.2756          |
+| 0.0059        | 24.0  | 432  | 0.2754          |
+| 0.0059        | 25.0  | 450  | 0.2788          |
+| 0.0067        | 26.0  | 468  | 0.2864          |
+| 0.0058        | 27.0  | 486  | 0.2769          |
+| 0.0046        | 28.0  | 504  | 0.2742          |
+| 0.0044        | 29.0  | 522  | 0.2786          |
+| 0.0038        | 30.0  | 540  | 0.2776          |
+| 0.0039        | 31.0  | 558  | 0.2764          |
+| 0.0034        | 32.0  | 576  | 0.2764          |
+| 0.0032        | 33.0  | 594  | 0.2705          |
+| 0.0029        | 34.0  | 612  | 0.2766          |
+| 0.0029        | 35.0  | 630  | 0.2742          |
+| 0.0026        | 36.0  | 648  | 0.2751          |
+| 0.0027        | 37.0  | 666  | 0.2771          |
+| 0.0024        | 38.0  | 684  | 0.2783          |
+
+
+### Framework versions
+
+- Transformers 4.51.3
+- Pytorch 2.6.0+cu124
+- Datasets 3.5.1
+- Tokenizers 0.21.1

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<|pad|>": 50257
+}

config.json

@@ -0,0 +1,721 @@
+{
+  "_num_labels": 1,
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2ForSequenceClassification"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 50256,
+  "embd_pdrop": 0.1,
+  "eos_token_id": 50256,
+  "id2label": {
+    "0": "a1at_status",
+    "1": "acq_score",
+    "2": "act_score",
+    "3": "adas_cog_score",
+    "4": "admission_date",
+    "5": "admission_id",
+    "6": "admission_reason",
+    "7": "admission_source",
+    "8": "adverse_event_flag",
+    "9": "adverse_event_management_end_date",
+    "10": "adverse_event_management_start_date",
+    "11": "adverse_event_management_type",
+    "12": "adverse_event_permanent_discontinuation_flag",
+    "13": "adverse_event_type",
+    "14": "age",
+    "15": "alcohol_status",
+    "16": "allergy_reaction_type",
+    "17": "allergy_severity",
+    "18": "ami_event_date",
+    "19": "ana_result_status",
+    "20": "anesthesia_type",
+    "21": "animal_verbal_fluency_score",
+    "22": "ann_arbor_stage",
+    "23": "anti_ccp_level",
+    "24": "anti_ccp_status",
+    "25": "anxiety_depression_flag",
+    "26": "asa_physical_status",
+    "27": "asdas_score",
+    "28": "asthma_flag",
+    "29": "atrial_fibrillation_flag",
+    "30": "autoimmune_condition_reported",
+    "31": "basdai_score",
+    "32": "basfi_score",
+    "33": "benzodiazepine_use_flag",
+    "34": "best_response_date",
+    "35": "beta2_microglobulin_level",
+    "36": "biopsy_results_note",
+    "37": "biopsy_type",
+    "38": "birth_date",
+    "39": "bmi",
+    "40": "bmi_category",
+    "41": "bone_marrow_involvement_flag",
+    "42": "bone_marrow_plasma_cell_percent",
+    "43": "breast_cancer_flag",
+    "44": "bronchiectasis_flag",
+    "45": "cancer_history_flag",
+    "46": "cardiovascular_disease_flag",
+    "47": "care_site_id",
+    "48": "cart_manufacturing_completion_date",
+    "49": "cart_manufacturing_delay_reason",
+    "50": "cart_manufacturing_start_date",
+    "51": "cart_target_antigen",
+    "52": "cat_score",
+    "53": "cause_of_death_code",
+    "54": "cdr_score",
+    "55": "charlson_comorbidity_index",
+    "56": "chronic_bronchitis_flag",
+    "57": "chronic_kidney_disease_flag",
+    "58": "ck_mb_level",
+    "59": "clinical_response",
+    "60": "cns_risk_stratification_category",
+    "61": "comorbidities_reported",
+    "62": "concomitant_corticosteroid_flag",
+    "63": "concomitant_csdmard_flag",
+    "64": "condition_status",
+    "65": "copd_flag",
+    "66": "copd_severity_stage",
+    "67": "covid19_severity",
+    "68": "covid19_test_date",
+    "69": "covid19_test_result",
+    "70": "covid19_vaccination_date",
+    "71": "covid19_vaccination_status",
+    "72": "cpet_assessment",
+    "73": "crab_features_reported",
+    "74": "creatinine_level",
+    "75": "crp_level",
+    "76": "crs_grade",
+    "77": "cytogenetics_result",
+    "78": "dactylitis_severity_score",
+    "79": "dapsa_score",
+    "80": "death_date",
+    "81": "death_flag",
+    "82": "dementia_flag",
+    "83": "diabetes_flag",
+    "84": "diagnosis_code",
+    "85": "diagnosis_date",
+    "86": "diastolic_bp",
+    "87": "discharge_date",
+    "88": "discharge_destination",
+    "89": "discharge_medications_reported",
+    "90": "discontinuation_reason",
+    "91": "disease_activity_score_das28",
+    "92": "dlco_percent_predicted",
+    "93": "drug_administration_frequency",
+    "94": "drug_dose",
+    "95": "drug_name",
+    "96": "drug_regimen_name",
+    "97": "drug_route",
+    "98": "dyslipidemia_flag",
+    "99": "ecog_performance_status",
+    "100": "education_years",
+    "101": "egfr_level",
+    "102": "emphysema_flag",
+    "103": "eortc_qlq_c30_emotional_score",
+    "104": "eortc_qlq_c30_fatigue_score",
+    "105": "eortc_qlq_c30_global_health_score",
+    "106": "eortc_qlq_c30_nausea_vomiting_score",
+    "107": "eortc_qlq_c30_pain_score",
+    "108": "eortc_qlq_c30_physical_score",
+    "109": "eortc_qlq_c30_role_score",
+    "110": "eosinophil_category",
+    "111": "eosinophilic_inflammation_v1_flag",
+    "112": "eosinophilic_inflammation_v2_flag",
+    "113": "eq5d_score",
+    "114": "er_status",
+    "115": "erosion_score",
+    "116": "esr_level",
+    "117": "exercise_intolerance_6mwt_flag",
+    "118": "exertional_desaturation_flag",
+    "119": "extra_articular_manifestation_flag",
+    "120": "extra_articular_manifestation_reported",
+    "121": "extranodal_sites_reported",
+    "122": "fact_lym_emotional_score",
+    "123": "fact_lym_lymphoma_subscale_score",
+    "124": "fact_lym_physical_score",
+    "125": "fact_lym_total_score",
+    "126": "family_history_reported",
+    "127": "fasting_glucose_level",
+    "128": "feno_above_threshold_flag",
+    "129": "feno_level",
+    "130": "ferritin_level",
+    "131": "fev1_gold_stage",
+    "132": "fev1_percent_predicted",
+    "133": "follow_up_appointment_scheduled_flag",
+    "134": "follow_up_specialty",
+    "135": "frailty_status",
+    "136": "frequent_exacerbator_flag",
+    "137": "frequent_readmissions_flag",
+    "138": "functional_status",
+    "139": "fvc_percent_predicted",
+    "140": "gerd_flag",
+    "141": "global_disease_vas_patient",
+    "142": "global_disease_vas_physician",
+    "143": "gold_group",
+    "144": "hamd_score",
+    "145": "haq_di_score",
+    "146": "hba1c_level",
+    "147": "hdl_cholesterol_level",
+    "148": "heart_failure_etiology",
+    "149": "heart_failure_event_type",
+    "150": "heart_failure_flag",
+    "151": "heart_failure_primary_cause",
+    "152": "heart_failure_primary_cause_other",
+    "153": "heart_failure_type",
+    "154": "heart_rate",
+    "155": "height",
+    "156": "hemoglobin_level",
+    "157": "hepatitis_b_status",
+    "158": "hepatitis_c_status",
+    "159": "her2_status",
+    "160": "hiv_status",
+    "161": "hla_b27_status",
+    "162": "hospitalization_heart_failure_flag",
+    "163": "hpylori_status",
+    "164": "hsct_details_note",
+    "165": "hyperinflation_rv_tlc_flag",
+    "166": "hyperinflation_rv_tlc_tlc_flag",
+    "167": "hypertension_flag",
+    "168": "icans_grade",
+    "169": "imaging_extramedullary_involvement_sites_reported",
+    "170": "imaging_joint_inflammation_flag",
+    "171": "imaging_joint_space_narrowing_grade",
+    "172": "imaging_leukemia_involvement_flag",
+    "173": "imaging_lymphoma_involvement_flag",
+    "174": "imaging_myeloma_findings_reported",
+    "175": "imaging_progressive_disease_sites_reported",
+    "176": "imaging_reporting_standard_flag",
+    "177": "imaging_suv_max",
+    "178": "imaging_type",
+    "179": "immunophenotyping_marker_reported",
+    "180": "infusion_reaction_flag",
+    "181": "inhaler_technique_assessment",
+    "182": "initial_diagnostic_method",
+    "183": "insurance_type",
+    "184": "ischemic_heart_disease_flag",
+    "185": "iv_diuretic_use_flag",
+    "186": "last_visit_date",
+    "187": "ldh_level_category",
+    "188": "ldl_cholesterol_level",
+    "189": "leeds_enthesitis_index",
+    "190": "length_of_stay_days",
+    "191": "leukemia_risk_stratification",
+    "192": "long_covid_symptoms_reported",
+    "193": "long_term_oxygen_therapy_flag",
+    "194": "long_term_toxicity_reported",
+    "195": "loss_of_response_date",
+    "196": "lost_followup_flag",
+    "197": "lugano_classification",
+    "198": "lvef_percent",
+    "199": "madrs_score",
+    "200": "malnutrition_flag",
+    "201": "manufacturing_issue_note",
+    "202": "measurement_date",
+    "203": "measurement_lower_limit",
+    "204": "measurement_unit",
+    "205": "measurement_upper_limit",
+    "206": "medication_adherence_assessment",
+    "207": "mild_exacerbation_count",
+    "208": "minicog_score",
+    "209": "mmrc_dyspnea_score",
+    "210": "mmse_score",
+    "211": "moca_score",
+    "212": "moderate_exacerbation_count",
+    "213": "molecular_genetic_marker_reported",
+    "214": "mortality_30day_flag",
+    "215": "niv_bipap_flag",
+    "216": "nodal_mass_size",
+    "217": "note_date",
+    "218": "note_text",
+    "219": "note_type",
+    "220": "nutritional_status",
+    "221": "nyha_class",
+    "222": "obesity_flag",
+    "223": "opioid_use_flag",
+    "224": "osteoporosis_flag",
+    "225": "other_inhaled_exposure_status",
+    "226": "oxygen_requirement_status",
+    "227": "pacu_discharge_intermediate_care_flag",
+    "228": "pain_vas",
+    "229": "pasdas_score",
+    "230": "pasqal_tool_used_flag",
+    "231": "past_surgical_history_reported",
+    "232": "peripheral_arterial_disease_flag",
+    "233": "person_id",
+    "234": "phq9_score",
+    "235": "physical_activity_flag",
+    "236": "platelet_count",
+    "237": "polypharmacy_status",
+    "238": "postoperative_complication_flag",
+    "239": "postoperative_delirium_flag",
+    "240": "postoperative_delirium_history_flag",
+    "241": "pr_status",
+    "242": "prescription_days_supply",
+    "243": "prescription_dispense_date",
+    "244": "pretreatment_screening_reported",
+    "245": "prior_corticosteroids_count",
+    "246": "prior_dmards_count",
+    "247": "prior_nsaids_count",
+    "248": "procedure_code",
+    "249": "procedure_date",
+    "250": "procedure_notes",
+    "251": "provider_specialty",
+    "252": "psa_subtype",
+    "253": "pulmonary_hypertension_flag",
+    "254": "ra_functional_class",
+    "255": "race",
+    "256": "radiation_fraction_count",
+    "257": "radiation_fraction_dose",
+    "258": "radiation_total_dose",
+    "259": "radiation_treated_sites_reported",
+    "260": "radiographic_stage",
+    "261": "random_glucose_level",
+    "262": "readmission_date",
+    "263": "readmission_flag",
+    "264": "readmission_prevention_measures_reported",
+    "265": "readmission_reason",
+    "266": "readmission_risk_score",
+    "267": "relapse_confirmation_method",
+    "268": "relapse_date",
+    "269": "remission_flag",
+    "270": "reported_allergy_drug",
+    "271": "respiratory_insufficiency_flag",
+    "272": "respiratory_insufficiency_type",
+    "273": "respiratory_rate",
+    "274": "rheumatoid_factor_level",
+    "275": "rheumatoid_factor_status",
+    "276": "rso2_baseline_percent",
+    "277": "rv_percent_predicted",
+    "278": "rv_tlc_ratio",
+    "279": "serum_albumin_level",
+    "280": "serum_calcium_level",
+    "281": "serum_free_light_chain_level",
+    "282": "severe_exacerbation_count",
+    "283": "sex",
+    "284": "sf36_score",
+    "285": "sleep_apnea_flag",
+    "286": "smoking_status",
+    "287": "social_determinants_reported",
+    "288": "social_support_assessment",
+    "289": "sparcc_score",
+    "290": "spirometry_reversibility_flag",
+    "291": "steroid_equivalent_dose",
+    "292": "steroid_taper_duration",
+    "293": "substance_use_history_reported",
+    "294": "support_system_assessment",
+    "295": "supportive_medication_reported",
+    "296": "surgical_risk_classification",
+    "297": "surgical_specialty",
+    "298": "symptoms_at_presentation_reported",
+    "299": "synovial_fluid_analysis_reported",
+    "300": "systolic_bp",
+    "301": "temperature",
+    "302": "therapeutic_adherence_assessment",
+    "303": "therapeutic_class",
+    "304": "time_from_diagnosis_to_biologic",
+    "305": "tlc_percent_predicted",
+    "306": "total_cholesterol_level",
+    "307": "toxicity_grade",
+    "308": "treatment_cycle_count",
+    "309": "treatment_end_date",
+    "310": "treatment_inefficacy_flag",
+    "311": "treatment_intent",
+    "312": "treatment_line",
+    "313": "treatment_modification_date",
+    "314": "treatment_modification_reason",
+    "315": "treatment_modification_type",
+    "316": "treatment_start_date",
+    "317": "treatment_switch_count",
+    "318": "treatment_type",
+    "319": "triglycerides_level",
+    "320": "troponin_level",
+    "321": "tumor_grade",
+    "322": "tumor_stage_m",
+    "323": "tumor_stage_n",
+    "324": "tumor_stage_t",
+    "325": "ultrasound_doppler_grade",
+    "326": "unknown_field",
+    "327": "urine_albumin_creatinine_ratio",
+    "328": "visit_end_date",
+    "329": "visit_id",
+    "330": "visit_start_date",
+    "331": "visit_type",
+    "332": "wbc_count",
+    "333": "weight",
+    "334": "worsening_heart_failure_episode_order",
+    "335": "worsening_heart_failure_event_type",
+    "336": "worsening_heart_failure_flag",
+    "337": "worsening_heart_failure_start_date"
+  },
+  "initializer_range": 0.02,
+  "label2id": {
+    "a1at_status": 0,
+    "acq_score": 1,
+    "act_score": 2,
+    "adas_cog_score": 3,
+    "admission_date": 4,
+    "admission_id": 5,
+    "admission_reason": 6,
+    "admission_source": 7,
+    "adverse_event_flag": 8,
+    "adverse_event_management_end_date": 9,
+    "adverse_event_management_start_date": 10,
+    "adverse_event_management_type": 11,
+    "adverse_event_permanent_discontinuation_flag": 12,
+    "adverse_event_type": 13,
+    "age": 14,
+    "alcohol_status": 15,
+    "allergy_reaction_type": 16,
+    "allergy_severity": 17,
+    "ami_event_date": 18,
+    "ana_result_status": 19,
+    "anesthesia_type": 20,
+    "animal_verbal_fluency_score": 21,
+    "ann_arbor_stage": 22,
+    "anti_ccp_level": 23,
+    "anti_ccp_status": 24,
+    "anxiety_depression_flag": 25,
+    "asa_physical_status": 26,
+    "asdas_score": 27,
+    "asthma_flag": 28,
+    "atrial_fibrillation_flag": 29,
+    "autoimmune_condition_reported": 30,
+    "basdai_score": 31,
+    "basfi_score": 32,
+    "benzodiazepine_use_flag": 33,
+    "best_response_date": 34,
+    "beta2_microglobulin_level": 35,
+    "biopsy_results_note": 36,
+    "biopsy_type": 37,
+    "birth_date": 38,
+    "bmi": 39,
+    "bmi_category": 40,
+    "bone_marrow_involvement_flag": 41,
+    "bone_marrow_plasma_cell_percent": 42,
+    "breast_cancer_flag": 43,
+    "bronchiectasis_flag": 44,
+    "cancer_history_flag": 45,
+    "cardiovascular_disease_flag": 46,
+    "care_site_id": 47,
+    "cart_manufacturing_completion_date": 48,
+    "cart_manufacturing_delay_reason": 49,
+    "cart_manufacturing_start_date": 50,
+    "cart_target_antigen": 51,
+    "cat_score": 52,
+    "cause_of_death_code": 53,
+    "cdr_score": 54,
+    "charlson_comorbidity_index": 55,
+    "chronic_bronchitis_flag": 56,
+    "chronic_kidney_disease_flag": 57,
+    "ck_mb_level": 58,
+    "clinical_response": 59,
+    "cns_risk_stratification_category": 60,
+    "comorbidities_reported": 61,
+    "concomitant_corticosteroid_flag": 62,
+    "concomitant_csdmard_flag": 63,
+    "condition_status": 64,
+    "copd_flag": 65,
+    "copd_severity_stage": 66,
+    "covid19_severity": 67,
+    "covid19_test_date": 68,
+    "covid19_test_result": 69,
+    "covid19_vaccination_date": 70,
+    "covid19_vaccination_status": 71,
+    "cpet_assessment": 72,
+    "crab_features_reported": 73,
+    "creatinine_level": 74,
+    "crp_level": 75,
+    "crs_grade": 76,
+    "cytogenetics_result": 77,
+    "dactylitis_severity_score": 78,
+    "dapsa_score": 79,
+    "death_date": 80,
+    "death_flag": 81,
+    "dementia_flag": 82,
+    "diabetes_flag": 83,
+    "diagnosis_code": 84,
+    "diagnosis_date": 85,
+    "diastolic_bp": 86,
+    "discharge_date": 87,
+    "discharge_destination": 88,
+    "discharge_medications_reported": 89,
+    "discontinuation_reason": 90,
+    "disease_activity_score_das28": 91,
+    "dlco_percent_predicted": 92,
+    "drug_administration_frequency": 93,
+    "drug_dose": 94,
+    "drug_name": 95,
+    "drug_regimen_name": 96,
+    "drug_route": 97,
+    "dyslipidemia_flag": 98,
+    "ecog_performance_status": 99,
+    "education_years": 100,
+    "egfr_level": 101,
+    "emphysema_flag": 102,
+    "eortc_qlq_c30_emotional_score": 103,
+    "eortc_qlq_c30_fatigue_score": 104,
+    "eortc_qlq_c30_global_health_score": 105,
+    "eortc_qlq_c30_nausea_vomiting_score": 106,
+    "eortc_qlq_c30_pain_score": 107,
+    "eortc_qlq_c30_physical_score": 108,
+    "eortc_qlq_c30_role_score": 109,
+    "eosinophil_category": 110,
+    "eosinophilic_inflammation_v1_flag": 111,
+    "eosinophilic_inflammation_v2_flag": 112,
+    "eq5d_score": 113,
+    "er_status": 114,
+    "erosion_score": 115,
+    "esr_level": 116,
+    "exercise_intolerance_6mwt_flag": 117,
+    "exertional_desaturation_flag": 118,
+    "extra_articular_manifestation_flag": 119,
+    "extra_articular_manifestation_reported": 120,
+    "extranodal_sites_reported": 121,
+    "fact_lym_emotional_score": 122,
+    "fact_lym_lymphoma_subscale_score": 123,
+    "fact_lym_physical_score": 124,
+    "fact_lym_total_score": 125,
+    "family_history_reported": 126,
+    "fasting_glucose_level": 127,
+    "feno_above_threshold_flag": 128,
+    "feno_level": 129,
+    "ferritin_level": 130,
+    "fev1_gold_stage": 131,
+    "fev1_percent_predicted": 132,
+    "follow_up_appointment_scheduled_flag": 133,
+    "follow_up_specialty": 134,
+    "frailty_status": 135,
+    "frequent_exacerbator_flag": 136,
+    "frequent_readmissions_flag": 137,
+    "functional_status": 138,
+    "fvc_percent_predicted": 139,
+    "gerd_flag": 140,
+    "global_disease_vas_patient": 141,
+    "global_disease_vas_physician": 142,
+    "gold_group": 143,
+    "hamd_score": 144,
+    "haq_di_score": 145,
+    "hba1c_level": 146,
+    "hdl_cholesterol_level": 147,
+    "heart_failure_etiology": 148,
+    "heart_failure_event_type": 149,
+    "heart_failure_flag": 150,
+    "heart_failure_primary_cause": 151,
+    "heart_failure_primary_cause_other": 152,
+    "heart_failure_type": 153,
+    "heart_rate": 154,
+    "height": 155,
+    "hemoglobin_level": 156,
+    "hepatitis_b_status": 157,
+    "hepatitis_c_status": 158,
+    "her2_status": 159,
+    "hiv_status": 160,
+    "hla_b27_status": 161,
+    "hospitalization_heart_failure_flag": 162,
+    "hpylori_status": 163,
+    "hsct_details_note": 164,
+    "hyperinflation_rv_tlc_flag": 165,
+    "hyperinflation_rv_tlc_tlc_flag": 166,
+    "hypertension_flag": 167,
+    "icans_grade": 168,
+    "imaging_extramedullary_involvement_sites_reported": 169,
+    "imaging_joint_inflammation_flag": 170,
+    "imaging_joint_space_narrowing_grade": 171,
+    "imaging_leukemia_involvement_flag": 172,
+    "imaging_lymphoma_involvement_flag": 173,
+    "imaging_myeloma_findings_reported": 174,
+    "imaging_progressive_disease_sites_reported": 175,
+    "imaging_reporting_standard_flag": 176,
+    "imaging_suv_max": 177,
+    "imaging_type": 178,
+    "immunophenotyping_marker_reported": 179,
+    "infusion_reaction_flag": 180,
+    "inhaler_technique_assessment": 181,
+    "initial_diagnostic_method": 182,
+    "insurance_type": 183,
+    "ischemic_heart_disease_flag": 184,
+    "iv_diuretic_use_flag": 185,
+    "last_visit_date": 186,
+    "ldh_level_category": 187,
+    "ldl_cholesterol_level": 188,
+    "leeds_enthesitis_index": 189,
+    "length_of_stay_days": 190,
+    "leukemia_risk_stratification": 191,
+    "long_covid_symptoms_reported": 192,
+    "long_term_oxygen_therapy_flag": 193,
+    "long_term_toxicity_reported": 194,
+    "loss_of_response_date": 195,
+    "lost_followup_flag": 196,
+    "lugano_classification": 197,
+    "lvef_percent": 198,
+    "madrs_score": 199,
+    "malnutrition_flag": 200,
+    "manufacturing_issue_note": 201,
+    "measurement_date": 202,
+    "measurement_lower_limit": 203,
+    "measurement_unit": 204,
+    "measurement_upper_limit": 205,
+    "medication_adherence_assessment": 206,
+    "mild_exacerbation_count": 207,
+    "minicog_score": 208,
+    "mmrc_dyspnea_score": 209,
+    "mmse_score": 210,
+    "moca_score": 211,
+    "moderate_exacerbation_count": 212,
+    "molecular_genetic_marker_reported": 213,
+    "mortality_30day_flag": 214,
+    "niv_bipap_flag": 215,
+    "nodal_mass_size": 216,
+    "note_date": 217,
+    "note_text": 218,
+    "note_type": 219,
+    "nutritional_status": 220,
+    "nyha_class": 221,
+    "obesity_flag": 222,
+    "opioid_use_flag": 223,
+    "osteoporosis_flag": 224,
+    "other_inhaled_exposure_status": 225,
+    "oxygen_requirement_status": 226,
+    "pacu_discharge_intermediate_care_flag": 227,
+    "pain_vas": 228,
+    "pasdas_score": 229,
+    "pasqal_tool_used_flag": 230,
+    "past_surgical_history_reported": 231,
+    "peripheral_arterial_disease_flag": 232,
+    "person_id": 233,
+    "phq9_score": 234,
+    "physical_activity_flag": 235,
+    "platelet_count": 236,
+    "polypharmacy_status": 237,
+    "postoperative_complication_flag": 238,
+    "postoperative_delirium_flag": 239,
+    "postoperative_delirium_history_flag": 240,
+    "pr_status": 241,
+    "prescription_days_supply": 242,
+    "prescription_dispense_date": 243,
+    "pretreatment_screening_reported": 244,
+    "prior_corticosteroids_count": 245,
+    "prior_dmards_count": 246,
+    "prior_nsaids_count": 247,
+    "procedure_code": 248,
+    "procedure_date": 249,
+    "procedure_notes": 250,
+    "provider_specialty": 251,
+    "psa_subtype": 252,
+    "pulmonary_hypertension_flag": 253,
+    "ra_functional_class": 254,
+    "race": 255,
+    "radiation_fraction_count": 256,
+    "radiation_fraction_dose": 257,
+    "radiation_total_dose": 258,
+    "radiation_treated_sites_reported": 259,
+    "radiographic_stage": 260,
+    "random_glucose_level": 261,
+    "readmission_date": 262,
+    "readmission_flag": 263,
+    "readmission_prevention_measures_reported": 264,
+    "readmission_reason": 265,
+    "readmission_risk_score": 266,
+    "relapse_confirmation_method": 267,
+    "relapse_date": 268,
+    "remission_flag": 269,
+    "reported_allergy_drug": 270,
+    "respiratory_insufficiency_flag": 271,
+    "respiratory_insufficiency_type": 272,
+    "respiratory_rate": 273,
+    "rheumatoid_factor_level": 274,
+    "rheumatoid_factor_status": 275,
+    "rso2_baseline_percent": 276,
+    "rv_percent_predicted": 277,
+    "rv_tlc_ratio": 278,
+    "serum_albumin_level": 279,
+    "serum_calcium_level": 280,
+    "serum_free_light_chain_level": 281,
+    "severe_exacerbation_count": 282,
+    "sex": 283,
+    "sf36_score": 284,
+    "sleep_apnea_flag": 285,
+    "smoking_status": 286,
+    "social_determinants_reported": 287,
+    "social_support_assessment": 288,
+    "sparcc_score": 289,
+    "spirometry_reversibility_flag": 290,
+    "steroid_equivalent_dose": 291,
+    "steroid_taper_duration": 292,
+    "substance_use_history_reported": 293,
+    "support_system_assessment": 294,
+    "supportive_medication_reported": 295,
+    "surgical_risk_classification": 296,
+    "surgical_specialty": 297,
+    "symptoms_at_presentation_reported": 298,
+    "synovial_fluid_analysis_reported": 299,
+    "systolic_bp": 300,
+    "temperature": 301,
+    "therapeutic_adherence_assessment": 302,
+    "therapeutic_class": 303,
+    "time_from_diagnosis_to_biologic": 304,
+    "tlc_percent_predicted": 305,
+    "total_cholesterol_level": 306,
+    "toxicity_grade": 307,
+    "treatment_cycle_count": 308,
+    "treatment_end_date": 309,
+    "treatment_inefficacy_flag": 310,
+    "treatment_intent": 311,
+    "treatment_line": 312,
+    "treatment_modification_date": 313,
+    "treatment_modification_reason": 314,
+    "treatment_modification_type": 315,
+    "treatment_start_date": 316,
+    "treatment_switch_count": 317,
+    "treatment_type": 318,
+    "triglycerides_level": 319,
+    "troponin_level": 320,
+    "tumor_grade": 321,
+    "tumor_stage_m": 322,
+    "tumor_stage_n": 323,
+    "tumor_stage_t": 324,
+    "ultrasound_doppler_grade": 325,
+    "unknown_field": 326,
+    "urine_albumin_creatinine_ratio": 327,
+    "visit_end_date": 328,
+    "visit_id": 329,
+    "visit_start_date": 330,
+    "visit_type": 331,
+    "wbc_count": 332,
+    "weight": 333,
+    "worsening_heart_failure_episode_order": 334,
+    "worsening_heart_failure_event_type": 335,
+    "worsening_heart_failure_flag": 336,
+    "worsening_heart_failure_start_date": 337
+  },
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 768,
+  "n_head": 12,
+  "n_inner": null,
+  "n_layer": 6,
+  "n_positions": 1024,
+  "pad_token_id": 50257,
+  "problem_type": "single_label_classification",
+  "reorder_and_upcast_attn": false,
+  "resid_pdrop": 0.1,
+  "scale_attn_by_inverse_layer_idx": false,
+  "scale_attn_weights": true,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "task_specific_params": {
+    "text-generation": {
+      "do_sample": true,
+      "max_length": 50
+    }
+  },
+  "torch_dtype": "float16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "vocab_size": 50258
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4946b1321e11372c1c8714b6ac4a251744bb552ecf521617fb828a6291c3411
+size 164353504

special_tokens_map.json

@@ -0,0 +1,12 @@
+{
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "pad_token": {
+    "content": "<|pad|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": "<|endoftext|>"
+}

tokenizer_config.json

@@ -0,0 +1,29 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "50256": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "50257": {
+      "content": "<|pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": {},
+  "model_max_length": 1024,
+  "pad_token": "<|pad|>",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7697c4e9b1007db58fda5e695eceb61866ceb9d5c73b48287d7bd931be45a4d9
+size 7352