crfm-helm 0.5.3__py3-none-any.whl → 0.5.5__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Potentially problematic release.
This version of crfm-helm might be problematic. Click here for more details.
- crfm_helm-0.5.5.dist-info/METADATA +413 -0
- crfm_helm-0.5.5.dist-info/RECORD +894 -0
- {crfm_helm-0.5.3.dist-info → crfm_helm-0.5.5.dist-info}/WHEEL +1 -1
- helm/benchmark/adaptation/adapter_spec.py +13 -1
- helm/benchmark/adaptation/adapters/adapter_factory.py +15 -1
- helm/benchmark/adaptation/adapters/binary_ranking_adapter.py +1 -1
- helm/benchmark/adaptation/adapters/chat_adapter.py +49 -0
- helm/benchmark/adaptation/adapters/ehr_instruction_adapter.py +108 -0
- helm/benchmark/adaptation/adapters/generation_adapter.py +1 -1
- helm/benchmark/adaptation/adapters/in_context_learning_adapter.py +1 -1
- helm/benchmark/adaptation/adapters/language_modeling_adapter.py +1 -1
- helm/benchmark/adaptation/adapters/multimodal/generation_multimodal_adapter.py +4 -2
- helm/benchmark/adaptation/adapters/multimodal/in_context_learning_multimodal_adapter.py +1 -1
- helm/benchmark/adaptation/adapters/multimodal/multiple_choice_joint_multimodal_adapter.py +1 -1
- helm/benchmark/adaptation/adapters/multimodal/test_in_context_learning_multimodal_adapter.py +4 -2
- helm/benchmark/adaptation/adapters/multimodal/test_multimodal_prompt.py +1 -1
- helm/benchmark/adaptation/adapters/multiple_choice_calibrated_adapter.py +1 -1
- helm/benchmark/adaptation/adapters/multiple_choice_joint_adapter.py +2 -2
- helm/benchmark/adaptation/adapters/multiple_choice_joint_chain_of_thought_adapter.py +87 -0
- helm/benchmark/adaptation/adapters/multiple_choice_separate_adapter.py +1 -1
- helm/benchmark/adaptation/adapters/test_generation_adapter.py +3 -3
- helm/benchmark/adaptation/adapters/test_language_modeling_adapter.py +2 -2
- helm/benchmark/adaptation/adapters/test_multiple_choice_joint_adapter.py +2 -2
- helm/benchmark/adaptation/common_adapter_specs.py +69 -4
- helm/benchmark/adaptation/prompt.py +1 -1
- helm/benchmark/annotation/aci_bench_annotator.py +95 -0
- helm/benchmark/annotation/air_bench_annotator.py +20 -5
- helm/benchmark/annotation/annotator.py +5 -0
- helm/benchmark/annotation/annotator_factory.py +3 -20
- helm/benchmark/annotation/anthropic_red_team_annotator.py +11 -24
- helm/benchmark/annotation/autobencher_capabilities_annotator.py +107 -0
- helm/benchmark/annotation/autobencher_safety_annotator.py +98 -0
- helm/benchmark/annotation/bigcodebench_annotator.py +108 -0
- helm/benchmark/annotation/bird_sql_annotator.py +58 -0
- helm/benchmark/annotation/call_center_annotator.py +22 -11
- helm/benchmark/annotation/chw_care_plan_annotator.py +98 -0
- helm/benchmark/annotation/czech_bank_qa_annotator.py +78 -0
- helm/benchmark/annotation/dischargeme_annotator.py +107 -0
- helm/benchmark/annotation/ehr_sql_annotator.py +87 -0
- helm/benchmark/annotation/harm_bench_annotator.py +11 -24
- helm/benchmark/annotation/helpdesk_call_summarization_annotator.py +131 -0
- helm/benchmark/annotation/image2struct/image_compiler_annotator.py +6 -1
- helm/benchmark/annotation/live_qa_annotator.py +10 -5
- helm/benchmark/annotation/med_dialog_annotator.py +99 -0
- helm/benchmark/annotation/medalign_annotator.py +100 -0
- helm/benchmark/annotation/medi_qa_annotator.py +98 -0
- helm/benchmark/annotation/medication_qa_annotator.py +90 -61
- helm/benchmark/annotation/mental_health_annotator.py +98 -0
- helm/benchmark/annotation/mimic_rrs_annotator.py +100 -0
- helm/benchmark/annotation/model_as_judge.py +281 -18
- helm/benchmark/annotation/mtsamples_procedures_annotator.py +98 -0
- helm/benchmark/annotation/mtsamples_replicate_annotator.py +101 -0
- helm/benchmark/annotation/omni_math/gpt_evaluation_template.txt +152 -0
- helm/benchmark/annotation/omni_math/gpt_evaluation_zero_shot_template.txt +36 -0
- helm/benchmark/annotation/omni_math_annotator.py +132 -0
- helm/benchmark/annotation/simple_safety_tests_annotator.py +11 -25
- helm/benchmark/annotation/spider_annotator.py +18 -0
- helm/benchmark/annotation/starr_patient_instructions_annotator.py +98 -0
- helm/benchmark/annotation/wildbench/eval_template.pairwise.v2.md +75 -0
- helm/benchmark/annotation/wildbench/eval_template.score.v2.md +66 -0
- helm/benchmark/annotation/wildbench_annotator.py +119 -0
- helm/benchmark/annotation/xstest_annotator.py +20 -30
- helm/benchmark/annotation_executor.py +35 -15
- helm/benchmark/augmentations/cleva_perturbation.py +9 -8
- helm/benchmark/augmentations/contraction_expansion_perturbation.py +2 -2
- helm/benchmark/augmentations/contrast_sets_perturbation.py +2 -2
- helm/benchmark/augmentations/dialect_perturbation.py +4 -5
- helm/benchmark/augmentations/extra_space_perturbation.py +2 -2
- helm/benchmark/augmentations/filler_words_perturbation.py +2 -2
- helm/benchmark/augmentations/gender_perturbation.py +2 -2
- helm/benchmark/augmentations/lowercase_perturbation.py +2 -2
- helm/benchmark/augmentations/mild_mix_perturbation.py +6 -6
- helm/benchmark/augmentations/misspelling_perturbation.py +2 -2
- helm/benchmark/augmentations/person_name_perturbation.py +4 -5
- helm/benchmark/augmentations/perturbation.py +1 -1
- helm/benchmark/augmentations/space_perturbation.py +2 -2
- helm/benchmark/augmentations/suffix_perturbation.py +2 -2
- helm/benchmark/augmentations/synonym_perturbation.py +4 -3
- helm/benchmark/augmentations/test_perturbation.py +16 -13
- helm/benchmark/augmentations/translate_perturbation.py +2 -2
- helm/benchmark/augmentations/typos_perturbation.py +2 -2
- helm/benchmark/data_preprocessor.py +2 -2
- helm/benchmark/huggingface_registration.py +2 -7
- helm/benchmark/metrics/aci_bench_metrics.py +34 -0
- helm/benchmark/metrics/basic_metrics.py +6 -6
- helm/benchmark/metrics/bbq_metrics.py +2 -2
- helm/benchmark/metrics/bias_metrics.py +12 -3
- helm/benchmark/metrics/bigcodebench_metrics.py +25 -0
- helm/benchmark/metrics/bird_sql_metrics.py +28 -0
- helm/benchmark/metrics/chw_care_plan_metrics.py +34 -0
- helm/benchmark/metrics/classification_metrics.py +76 -12
- helm/benchmark/metrics/cleva_harms_metrics.py +8 -7
- helm/benchmark/metrics/code_metrics.py +5 -5
- helm/benchmark/metrics/comet_metric.py +125 -0
- helm/benchmark/metrics/common_metric_specs.py +9 -2
- helm/benchmark/metrics/conv_fin_qa_calc_metrics.py +72 -0
- helm/benchmark/metrics/copyright_metrics.py +4 -4
- helm/benchmark/metrics/czech_bank_qa_metrics.py +29 -0
- helm/benchmark/metrics/decodingtrust_fairness_metrics.py +2 -2
- helm/benchmark/metrics/decodingtrust_privacy_metrics.py +2 -2
- helm/benchmark/metrics/decodingtrust_stereotype_bias_metrics.py +2 -2
- helm/benchmark/metrics/dischargeme_metrics.py +34 -0
- helm/benchmark/metrics/disinformation_metrics.py +4 -4
- helm/benchmark/metrics/dry_run_metrics.py +5 -5
- helm/benchmark/metrics/efficiency_metrics.py +3 -3
- helm/benchmark/metrics/ehr_sql_metrics.py +103 -0
- helm/benchmark/metrics/evaluate_instances_metric.py +3 -3
- helm/benchmark/metrics/evaluate_reference_metrics.py +144 -16
- helm/benchmark/metrics/gpqa_chain_of_thought_metric.py +103 -0
- helm/benchmark/metrics/gpt4_audio_critique_metrics.py +167 -0
- helm/benchmark/metrics/helpdesk_call_summarization_metrics.py +36 -0
- helm/benchmark/metrics/ifeval/instructions.py +1574 -0
- helm/benchmark/metrics/ifeval/instructions_registry.py +182 -0
- helm/benchmark/metrics/ifeval/instructions_registry.pyi +3 -0
- helm/benchmark/metrics/ifeval/instructions_util.py +153 -0
- helm/benchmark/metrics/ifeval_metrics.py +55 -0
- helm/benchmark/metrics/image_generation/aesthetics_metrics.py +1 -1
- helm/benchmark/metrics/image_generation/detection_metrics.py +1 -1
- helm/benchmark/metrics/image_generation/detectors/vitdet.py +1 -1
- helm/benchmark/metrics/image_generation/fractal_dimension/test_fractal_dimension_util.py +1 -1
- helm/benchmark/metrics/image_generation/fractal_dimension_metric.py +1 -1
- helm/benchmark/metrics/image_generation/nsfw_metrics.py +1 -1
- helm/benchmark/metrics/image_generation/q16/test_q16.py +3 -1
- helm/benchmark/metrics/image_generation/q16_toxicity_metrics.py +1 -1
- helm/benchmark/metrics/image_generation/skin_tone_metrics.py +2 -2
- helm/benchmark/metrics/image_generation/watermark/test_watermark_detector.py +1 -1
- helm/benchmark/metrics/image_generation/watermark_metrics.py +1 -1
- helm/benchmark/metrics/instruction_following_critique_metrics.py +4 -4
- helm/benchmark/metrics/language_modeling_metrics.py +4 -4
- helm/benchmark/metrics/machine_translation_metrics.py +2 -2
- helm/benchmark/metrics/med_dialog_metrics.py +34 -0
- helm/benchmark/metrics/medalign_metrics.py +34 -0
- helm/benchmark/metrics/medcalc_bench_metrics.py +124 -0
- helm/benchmark/metrics/medec_metrics.py +101 -0
- helm/benchmark/metrics/medi_qa_metrics.py +34 -0
- helm/benchmark/metrics/medication_qa_metrics.py +15 -4
- helm/benchmark/metrics/mental_health_metrics.py +34 -0
- helm/benchmark/metrics/metric.py +3 -3
- helm/benchmark/metrics/mimic_rrs_metrics.py +34 -0
- helm/benchmark/metrics/mimiciv_billing_code_metrics.py +96 -0
- helm/benchmark/metrics/mtsamples_procedures_metrics.py +34 -0
- helm/benchmark/metrics/mtsamples_replicate_metrics.py +34 -0
- helm/benchmark/metrics/nltk_helper.py +32 -0
- helm/benchmark/metrics/numeracy_metrics.py +4 -4
- helm/benchmark/metrics/omni_math_metrics.py +32 -0
- helm/benchmark/metrics/output_processing_metric.py +60 -0
- helm/benchmark/metrics/output_processors.py +15 -0
- helm/benchmark/metrics/paraphrase_generation_metrics.py +2 -2
- helm/benchmark/metrics/ranking_metrics.py +3 -3
- helm/benchmark/metrics/reference_metric.py +3 -3
- helm/benchmark/metrics/safety_metrics.py +39 -17
- helm/benchmark/metrics/{bhasa_metrics.py → seahelm_metrics.py} +3 -3
- helm/benchmark/metrics/seahelm_metrics_specs.py +10 -0
- helm/benchmark/metrics/spider_metrics.py +7 -0
- helm/benchmark/metrics/starr_patient_instructions_metrics.py +34 -0
- helm/benchmark/metrics/statistic.py +1 -1
- helm/benchmark/metrics/summac/model_summac.py +1 -1
- helm/benchmark/metrics/summarization_critique_metrics.py +4 -4
- helm/benchmark/metrics/summarization_metrics.py +19 -9
- helm/benchmark/metrics/test_bias_metrics.py +5 -1
- helm/benchmark/metrics/test_classification_metrics.py +140 -68
- helm/benchmark/metrics/test_evaluate_reference_metrics.py +15 -0
- helm/benchmark/metrics/test_metric.py +1 -1
- helm/benchmark/metrics/test_statistic.py +2 -2
- helm/benchmark/metrics/tokens/ai21_token_cost_estimator.py +1 -1
- helm/benchmark/metrics/tokens/auto_token_cost_estimator.py +6 -6
- helm/benchmark/metrics/tokens/cohere_token_cost_estimator.py +1 -1
- helm/benchmark/metrics/tokens/free_token_cost_estimator.py +1 -1
- helm/benchmark/metrics/tokens/gooseai_token_cost_estimator.py +1 -1
- helm/benchmark/metrics/tokens/openai_token_cost_estimator.py +1 -1
- helm/benchmark/metrics/tokens/test_ai21_token_cost_estimator.py +1 -1
- helm/benchmark/metrics/tokens/test_openai_token_cost_estimator.py +1 -1
- helm/benchmark/metrics/toxicity_metrics.py +4 -4
- helm/benchmark/metrics/unitxt_metrics.py +21 -4
- helm/benchmark/metrics/vision_language/image_metrics.py +7 -3
- helm/benchmark/metrics/wildbench_metrics.py +34 -0
- helm/benchmark/model_metadata_registry.py +16 -0
- helm/benchmark/presentation/create_plots.py +1 -1
- helm/benchmark/presentation/schema.py +3 -0
- helm/benchmark/presentation/summarize.py +119 -256
- helm/benchmark/presentation/test_summarize.py +145 -3
- helm/benchmark/presentation/torr_robustness_summarizer.py +178 -0
- helm/benchmark/reeval_run.py +203 -0
- helm/benchmark/reeval_runner.py +355 -0
- helm/benchmark/run.py +8 -17
- helm/benchmark/run_expander.py +105 -8
- helm/benchmark/run_spec_factory.py +12 -0
- helm/benchmark/run_specs/air_bench_run_specs.py +21 -3
- helm/benchmark/run_specs/audio_run_specs.py +613 -0
- helm/benchmark/run_specs/call_center_run_specs.py +49 -0
- helm/benchmark/run_specs/capabilities_run_specs.py +308 -0
- helm/benchmark/run_specs/classic_run_specs.py +1 -69
- helm/benchmark/run_specs/enem_challenge_specs.py +31 -0
- helm/benchmark/run_specs/enterprise_run_specs.py +260 -0
- helm/benchmark/run_specs/experimental_run_specs.py +112 -3
- helm/benchmark/run_specs/finance_run_specs.py +6 -2
- helm/benchmark/run_specs/imdb_ptbr_run_specs.py +30 -0
- helm/benchmark/run_specs/lite_run_specs.py +2 -2
- helm/benchmark/run_specs/long_context_run_specs.py +89 -0
- helm/benchmark/run_specs/medhelm_run_specs.py +1155 -0
- helm/benchmark/run_specs/mmlu_clinical_afr_run_specs.py +49 -0
- helm/benchmark/run_specs/oab_exams_specs.py +32 -0
- helm/benchmark/run_specs/safety_run_specs.py +37 -0
- helm/benchmark/run_specs/{bhasa_run_specs.py → seahelm_run_specs.py} +66 -52
- helm/benchmark/run_specs/sql_run_specs.py +54 -0
- helm/benchmark/run_specs/tweetsentbr_run_specs.py +32 -0
- helm/benchmark/run_specs/unitxt_run_specs.py +14 -5
- helm/benchmark/run_specs/vlm_run_specs.py +83 -5
- helm/benchmark/run_specs/winogrande_afr_run_specs.py +47 -0
- helm/benchmark/scenarios/aci_bench_scenario.py +120 -0
- helm/benchmark/scenarios/air_bench_scenario.py +6 -1
- helm/benchmark/scenarios/anthropic_hh_rlhf_scenario.py +5 -3
- helm/benchmark/scenarios/anthropic_red_team_scenario.py +1 -1
- helm/benchmark/scenarios/audio_language/__init__.py +0 -0
- helm/benchmark/scenarios/audio_language/air_bench_chat_scenario.py +128 -0
- helm/benchmark/scenarios/audio_language/air_bench_foundation_scenario.py +154 -0
- helm/benchmark/scenarios/audio_language/ami_scenario.py +96 -0
- helm/benchmark/scenarios/audio_language/audio_mnist_scenario.py +62 -0
- helm/benchmark/scenarios/audio_language/audio_pairs_scenario.py +62 -0
- helm/benchmark/scenarios/audio_language/audiocaps_scenario.py +59 -0
- helm/benchmark/scenarios/audio_language/casual_conversations2_scenario.py +152 -0
- helm/benchmark/scenarios/audio_language/common_voice_15_scenario.py +99 -0
- helm/benchmark/scenarios/audio_language/covost2_scenario.py +163 -0
- helm/benchmark/scenarios/audio_language/fleurs_fairness_scenario.py +83 -0
- helm/benchmark/scenarios/audio_language/fleurs_scenario.py +312 -0
- helm/benchmark/scenarios/audio_language/iemocap_audio_scenario.py +83 -0
- helm/benchmark/scenarios/audio_language/librispeech_fairness_scenario.py +96 -0
- helm/benchmark/scenarios/audio_language/librispeech_scenario.py +80 -0
- helm/benchmark/scenarios/audio_language/meld_audio_scenario.py +113 -0
- helm/benchmark/scenarios/audio_language/multilingual_librispeech_scenario.py +80 -0
- helm/benchmark/scenarios/audio_language/mustard_scenario.py +142 -0
- helm/benchmark/scenarios/audio_language/mutox_scenario.py +254 -0
- helm/benchmark/scenarios/audio_language/parade_scenario.py +97 -0
- helm/benchmark/scenarios/audio_language/speech_robust_bench_scenario.py +124 -0
- helm/benchmark/scenarios/audio_language/vocal_sound_scenario.py +69 -0
- helm/benchmark/scenarios/audio_language/voice_jailbreak_attacks_scenario.py +87 -0
- helm/benchmark/scenarios/audio_language/voxceleb2_scenario.py +106 -0
- helm/benchmark/scenarios/autobencher_capabilities_scenario.py +68 -0
- helm/benchmark/scenarios/autobencher_safety_scenario.py +51 -0
- helm/benchmark/scenarios/babi_qa_scenario.py +1 -1
- helm/benchmark/scenarios/banking77_scenario.py +6 -1
- helm/benchmark/scenarios/bbq_scenario.py +1 -1
- helm/benchmark/scenarios/big_bench_scenario.py +11 -1
- helm/benchmark/scenarios/bigcodebench_scenario.py +58 -0
- helm/benchmark/scenarios/bird_sql_scenario.py +94 -0
- helm/benchmark/scenarios/bird_sql_scenario_helper.py +118 -0
- helm/benchmark/scenarios/blimp_scenario.py +1 -1
- helm/benchmark/scenarios/bold_scenario.py +1 -1
- helm/benchmark/scenarios/boolq_scenario.py +1 -1
- helm/benchmark/scenarios/casehold_scenario.py +79 -0
- helm/benchmark/scenarios/chw_care_plan_scenario.py +105 -0
- helm/benchmark/scenarios/civil_comments_scenario.py +1 -1
- helm/benchmark/scenarios/clear_scenario.py +153 -0
- helm/benchmark/scenarios/cleva_scenario.py +2 -2
- helm/benchmark/scenarios/code_scenario.py +17 -4
- helm/benchmark/scenarios/commonsense_scenario.py +1 -1
- helm/benchmark/scenarios/conv_fin_qa_calc_scenario.py +97 -0
- helm/benchmark/scenarios/copyright_scenario.py +1 -1
- helm/benchmark/scenarios/covid_dialog_scenario.py +10 -1
- helm/benchmark/scenarios/cti_to_mitre_scenario.py +240 -0
- helm/benchmark/scenarios/custom_mcqa_scenario.py +1 -1
- helm/benchmark/scenarios/czech_bank_qa_scenario.py +130 -0
- helm/benchmark/scenarios/decodingtrust_adv_demonstration_scenario.py +1 -1
- helm/benchmark/scenarios/decodingtrust_privacy_scenario.py +1 -1
- helm/benchmark/scenarios/decodingtrust_stereotype_bias_scenario.py +1 -1
- helm/benchmark/scenarios/decodingtrust_toxicity_prompts_scenario.py +1 -1
- helm/benchmark/scenarios/dialogue_scenarios.py +13 -2
- helm/benchmark/scenarios/dischargeme_scenario.py +157 -0
- helm/benchmark/scenarios/disinformation_scenario.py +10 -1
- helm/benchmark/scenarios/dyck_language_scenario.py +10 -1
- helm/benchmark/scenarios/echr_judgment_classification_scenario.py +113 -0
- helm/benchmark/scenarios/ehr_sql_scenario.py +131 -0
- helm/benchmark/scenarios/ehrshot_scenario.py +1546 -0
- helm/benchmark/scenarios/enem_challenge_scenario.py +58 -0
- helm/benchmark/scenarios/entity_data_imputation_scenario.py +11 -1
- helm/benchmark/scenarios/entity_matching_scenario.py +12 -2
- helm/benchmark/scenarios/financial_phrasebank_scenario.py +94 -0
- helm/benchmark/scenarios/gold_commodity_news_scenario.py +124 -0
- helm/benchmark/scenarios/gpqa_scenario.py +80 -0
- helm/benchmark/scenarios/grammar_scenario.py +2 -2
- helm/benchmark/scenarios/gsm_scenario.py +10 -1
- helm/benchmark/scenarios/harm_bench_gcg_transfer_scenario.py +50 -0
- helm/benchmark/scenarios/harm_bench_scenario.py +1 -1
- helm/benchmark/scenarios/headqa_scenario.py +131 -0
- helm/benchmark/scenarios/helpdesk_call_summarization_scenario.py +37 -0
- helm/benchmark/scenarios/ice_scenario.py +8 -4
- helm/benchmark/scenarios/ifeval_scenario.py +53 -0
- helm/benchmark/scenarios/imdb_ptbr_scenario.py +60 -0
- helm/benchmark/scenarios/imdb_scenario.py +11 -2
- helm/benchmark/scenarios/infinite_bench_sum_scenario.py +82 -0
- helm/benchmark/scenarios/interactive_qa_mmlu_scenario.py +2 -2
- helm/benchmark/scenarios/koala_scenario.py +1 -1
- helm/benchmark/scenarios/legal_contract_summarization_scenario.py +129 -0
- helm/benchmark/scenarios/legal_opinion_sentiment_classification_scenario.py +77 -0
- helm/benchmark/scenarios/legal_summarization_scenario.py +11 -1
- helm/benchmark/scenarios/legal_support_scenario.py +11 -1
- helm/benchmark/scenarios/legalbench_scenario.py +22 -3
- helm/benchmark/scenarios/lex_glue_scenario.py +12 -2
- helm/benchmark/scenarios/lextreme_scenario.py +11 -1
- helm/benchmark/scenarios/live_qa_scenario.py +1 -1
- helm/benchmark/scenarios/lm_entry_scenario.py +1 -1
- helm/benchmark/scenarios/lsat_qa_scenario.py +1 -1
- helm/benchmark/scenarios/math_scenario.py +9 -1
- helm/benchmark/scenarios/me_q_sum_scenario.py +10 -1
- helm/benchmark/scenarios/med_dialog_scenario.py +22 -24
- helm/benchmark/scenarios/med_mcqa_scenario.py +10 -1
- helm/benchmark/scenarios/med_paragraph_simplification_scenario.py +10 -1
- helm/benchmark/scenarios/med_qa_scenario.py +10 -1
- helm/benchmark/scenarios/medalign_scenario.py +88 -0
- helm/benchmark/scenarios/medalign_scenario_helper.py +429 -0
- helm/benchmark/scenarios/medbullets_scenario.py +140 -0
- helm/benchmark/scenarios/medcalc_bench_scenario.py +125 -0
- helm/benchmark/scenarios/medec_scenario.py +120 -0
- helm/benchmark/scenarios/medhallu_scenario.py +66 -0
- helm/benchmark/scenarios/medi_qa_scenario.py +105 -0
- helm/benchmark/scenarios/medication_qa_scenario.py +2 -2
- helm/benchmark/scenarios/mental_health_scenario.py +112 -0
- helm/benchmark/scenarios/mimic_bhc_scenario.py +98 -0
- helm/benchmark/scenarios/mimic_rrs_scenario.py +89 -0
- helm/benchmark/scenarios/mimiciv_billing_code_scenario.py +71 -0
- helm/benchmark/scenarios/mmlu_clinical_afr_scenario.py +74 -0
- helm/benchmark/scenarios/mmlu_pro_scenario.py +95 -0
- helm/benchmark/scenarios/mmlu_scenario.py +11 -1
- helm/benchmark/scenarios/msmarco_scenario.py +1 -1
- helm/benchmark/scenarios/mtsamples_procedures_scenario.py +141 -0
- helm/benchmark/scenarios/mtsamples_replicate_scenario.py +141 -0
- helm/benchmark/scenarios/n2c2_ct_matching_scenario.py +271 -0
- helm/benchmark/scenarios/narrativeqa_scenario.py +1 -1
- helm/benchmark/scenarios/natural_qa_scenario.py +1 -1
- helm/benchmark/scenarios/newsqa_scenario.py +1 -1
- helm/benchmark/scenarios/numeracy_scenario.py +10 -1
- helm/benchmark/scenarios/oab_exams_scenario.py +57 -0
- helm/benchmark/scenarios/omni_math_scenario.py +53 -0
- helm/benchmark/scenarios/open_assistant_scenario.py +11 -2
- helm/benchmark/scenarios/opinions_qa_scenario.py +1 -1
- helm/benchmark/scenarios/pubmed_qa_scenario.py +54 -43
- helm/benchmark/scenarios/quac_scenario.py +10 -1
- helm/benchmark/scenarios/race_based_med_scenario.py +142 -0
- helm/benchmark/scenarios/raft_scenario.py +18 -3
- helm/benchmark/scenarios/real_toxicity_prompts_scenario.py +1 -1
- helm/benchmark/scenarios/ruler_qa_scenario_helper.py +171 -0
- helm/benchmark/scenarios/ruler_qa_scenarios.py +88 -0
- helm/benchmark/scenarios/scenario.py +9 -1
- helm/benchmark/scenarios/{bhasa_scenario.py → seahelm_scenario.py} +233 -84
- helm/benchmark/scenarios/self_instruct_scenario.py +1 -1
- helm/benchmark/scenarios/shc_bmt_scenario.py +69 -0
- helm/benchmark/scenarios/shc_cdi_scenario.py +70 -0
- helm/benchmark/scenarios/shc_conf_scenario.py +70 -0
- helm/benchmark/scenarios/shc_ent_scenario.py +72 -0
- helm/benchmark/scenarios/shc_gip_scenario.py +66 -0
- helm/benchmark/scenarios/shc_ptbm_scenario.py +76 -0
- helm/benchmark/scenarios/shc_sei_scenario.py +89 -0
- helm/benchmark/scenarios/shc_sequoia_scenario.py +69 -0
- helm/benchmark/scenarios/simple_safety_tests_scenario.py +1 -1
- helm/benchmark/scenarios/spider_scenario.py +91 -0
- helm/benchmark/scenarios/starr_patient_instructions_scenario.py +90 -0
- helm/benchmark/scenarios/summarization_scenario.py +11 -1
- helm/benchmark/scenarios/sumosum_scenario.py +157 -0
- helm/benchmark/scenarios/synthetic_efficiency_scenario.py +1 -1
- helm/benchmark/scenarios/synthetic_reasoning_natural_scenario.py +11 -1
- helm/benchmark/scenarios/synthetic_reasoning_scenario.py +11 -1
- helm/benchmark/scenarios/test_bigcodebench_scenario.py +26 -0
- helm/benchmark/scenarios/test_czech_bank_qa_scenario.py +18 -0
- helm/benchmark/scenarios/test_enem_challenge_scenario.py +53 -0
- helm/benchmark/scenarios/test_ewok_scenario.py +6 -2
- helm/benchmark/scenarios/test_gold_commodity_news_scenario.py +18 -0
- helm/benchmark/scenarios/test_gpqa_scenario.py +44 -0
- helm/benchmark/scenarios/test_ifeval_scenario.py +36 -0
- helm/benchmark/scenarios/test_imdb_ptbr_scenario.py +27 -0
- helm/benchmark/scenarios/test_infinite_bench_sum_scenario.py +46 -0
- helm/benchmark/scenarios/test_math_scenario.py +1 -0
- helm/benchmark/scenarios/test_mmlu_clinical_afr_scenario.py +21 -0
- helm/benchmark/scenarios/test_mmlu_pro_scenario.py +53 -0
- helm/benchmark/scenarios/test_oab_exams_scenario.py +51 -0
- helm/benchmark/scenarios/test_omni_math_scenario.py +27 -0
- helm/benchmark/scenarios/test_tweetsentbr_scenario.py +24 -0
- helm/benchmark/scenarios/test_wildbench_scenario.py +15 -0
- helm/benchmark/scenarios/test_winogrande_afr_scenario.py +19 -0
- helm/benchmark/scenarios/thai_exam_scenario.py +10 -1
- helm/benchmark/scenarios/the_pile_scenario.py +1 -1
- helm/benchmark/scenarios/truthful_qa_scenario.py +10 -1
- helm/benchmark/scenarios/tweetsentbr_scenario.py +66 -0
- helm/benchmark/scenarios/twitter_aae_scenario.py +1 -1
- helm/benchmark/scenarios/unitxt_scenario.py +8 -2
- helm/benchmark/scenarios/verifiability_judgment_scenario.py +1 -1
- helm/benchmark/scenarios/vicuna_scenario.py +1 -1
- helm/benchmark/scenarios/vision_language/blink_scenario.py +140 -0
- helm/benchmark/scenarios/vision_language/mm_star_scenario.py +95 -0
- helm/benchmark/scenarios/vision_language/vqa_rad_scenario.py +88 -0
- helm/benchmark/scenarios/wikifact_scenario.py +11 -1
- helm/benchmark/scenarios/wikitext_103_scenario.py +1 -1
- helm/benchmark/scenarios/wildbench_scenario.py +83 -0
- helm/benchmark/scenarios/winogrande_afr_scenario.py +78 -0
- helm/benchmark/scenarios/wmt_14_scenario.py +14 -2
- helm/benchmark/scenarios/xstest_scenario.py +1 -1
- helm/benchmark/server.py +11 -0
- helm/benchmark/slurm_runner.py +1 -1
- helm/benchmark/static/schema_audio.yaml +752 -0
- helm/benchmark/static/schema_autobencher.yaml +150 -0
- helm/benchmark/static/schema_call_center.yaml +97 -60
- helm/benchmark/static/schema_capabilities.yaml +254 -0
- helm/benchmark/static/schema_czech_bank.yaml +148 -0
- helm/benchmark/static/schema_enem_challenge.yaml +146 -0
- helm/benchmark/static/schema_enterprise.yaml +298 -0
- helm/benchmark/static/schema_finance.yaml +14 -12
- helm/benchmark/static/schema_heim.yaml +1389 -0
- helm/benchmark/static/schema_legal.yaml +566 -0
- helm/benchmark/static/{schema_medical.yaml → schema_long_context.yaml} +67 -82
- helm/benchmark/static/schema_medhelm.yaml +1081 -0
- helm/benchmark/static/schema_mmlu_winogrande_afr.yaml +1045 -0
- helm/benchmark/static/schema_safety.yaml +42 -6
- helm/benchmark/static/{schema_bhasa.yaml → schema_seahelm.yaml} +40 -26
- helm/benchmark/static/schema_social_audio.yaml +224 -0
- helm/benchmark/static/schema_sql.yaml +171 -0
- helm/benchmark/static/{schema_tables.yaml → schema_torr.yaml} +187 -30
- helm/benchmark/static/schema_tweetsentbr.yaml +146 -0
- helm/benchmark/static/schema_vhelm.yaml +151 -47
- helm/benchmark/static_build/assets/helm-safety-2907a7b6.png +0 -0
- helm/benchmark/static_build/assets/index-262903c1.js +10 -0
- helm/benchmark/static_build/assets/index-42060d71.css +1 -0
- helm/benchmark/static_build/assets/medhelm-overview-3ddfcd65.png +0 -0
- helm/benchmark/static_build/assets/{react-d4a0b69b.js → react-f82877fd.js} +1 -1
- helm/benchmark/static_build/assets/{recharts-6d337683.js → recharts-4037aff0.js} +1 -1
- helm/benchmark/static_build/assets/{tremor-54a99cc4.js → tremor-9cefc3c5.js} +1 -1
- helm/benchmark/static_build/assets/vhelm-aspects-1437d673.png +0 -0
- helm/benchmark/static_build/assets/vhelm-framework-a1ca3f3f.png +0 -0
- helm/benchmark/static_build/assets/vhelm-model-8afb7616.png +0 -0
- helm/benchmark/static_build/config.js +1 -1
- helm/benchmark/static_build/index.html +5 -5
- helm/benchmark/window_services/default_window_service.py +1 -1
- helm/benchmark/window_services/encoder_decoder_window_service.py +1 -1
- helm/benchmark/window_services/ice_window_service.py +1 -1
- helm/benchmark/window_services/image_generation/lexica_search_window_service.py +1 -1
- helm/benchmark/window_services/image_generation/openai_dalle_window_service.py +1 -1
- helm/benchmark/window_services/local_window_service.py +2 -2
- helm/benchmark/window_services/test_anthropic_window_service.py +3 -3
- helm/benchmark/window_services/test_bloom_window_service.py +3 -3
- helm/benchmark/window_services/test_gpt2_window_service.py +7 -2
- helm/benchmark/window_services/test_gpt4_window_service.py +8 -3
- helm/benchmark/window_services/test_gptj_window_service.py +8 -3
- helm/benchmark/window_services/test_gptneox_window_service.py +3 -3
- helm/benchmark/window_services/test_openai_window_service.py +8 -3
- helm/benchmark/window_services/test_opt_window_service.py +3 -3
- helm/benchmark/window_services/test_palmyra_window_service.py +3 -3
- helm/benchmark/window_services/test_t0pp_window_service.py +3 -3
- helm/benchmark/window_services/test_t511b_window_service.py +3 -3
- helm/benchmark/window_services/test_ul2_window_service.py +3 -3
- helm/benchmark/window_services/test_utils.py +1 -1
- helm/benchmark/window_services/test_yalm_window_service.py +3 -3
- helm/benchmark/window_services/tokenizer_service.py +0 -5
- helm/benchmark/window_services/yalm_window_service.py +1 -1
- helm/clients/ai21_client.py +3 -3
- helm/clients/aleph_alpha_client.py +1 -1
- helm/clients/audio_language/__init__.py +0 -0
- helm/clients/audio_language/diva_llama_client.py +118 -0
- helm/clients/audio_language/llama_omni_client.py +198 -0
- helm/clients/audio_language/qwen2_audiolm_client.py +188 -0
- helm/clients/audio_language/qwen_audiolm_client.py +150 -0
- helm/clients/auto_client.py +4 -2
- helm/clients/azure_openai_client.py +55 -0
- helm/clients/bedrock_client.py +201 -7
- helm/clients/bedrock_utils.py +33 -0
- helm/clients/clip_scorers/clip_scorer.py +1 -1
- helm/clients/clip_scorers/multilingual_clip_scorer.py +1 -1
- helm/clients/cohere_client.py +3 -3
- helm/clients/google_client.py +1 -1
- helm/clients/http_model_client.py +1 -1
- helm/clients/huggingface_client.py +10 -18
- helm/clients/ibm_client.py +267 -0
- helm/clients/image_generation/adobe_vision_client.py +1 -1
- helm/clients/image_generation/aleph_alpha_image_generation_client.py +1 -1
- helm/clients/image_generation/cogview2/sr_pipeline/__init__.py +3 -3
- helm/clients/image_generation/cogview2/sr_pipeline/direct_sr.py +5 -2
- helm/clients/image_generation/cogview2/sr_pipeline/iterative_sr.py +5 -2
- helm/clients/image_generation/cogview2/sr_pipeline/sr_group.py +2 -2
- helm/clients/image_generation/cogview2_client.py +1 -1
- helm/clients/image_generation/dalle2_client.py +1 -1
- helm/clients/image_generation/dalle3_client.py +2 -2
- helm/clients/image_generation/dalle_mini/__init__.py +1 -1
- helm/clients/image_generation/dalle_mini/data.py +1 -1
- helm/clients/image_generation/dalle_mini/model/__init__.py +5 -5
- helm/clients/image_generation/dalle_mini/model/configuration.py +1 -1
- helm/clients/image_generation/dalle_mini/model/modeling.py +2 -2
- helm/clients/image_generation/dalle_mini/model/processor.py +4 -4
- helm/clients/image_generation/dalle_mini/model/tokenizer.py +1 -1
- helm/clients/image_generation/dalle_mini/vqgan_jax/__init__.py +1 -1
- helm/clients/image_generation/dalle_mini/vqgan_jax/convert_pt_model_to_jax.py +2 -2
- helm/clients/image_generation/dalle_mini/vqgan_jax/modeling_flax_vqgan.py +1 -1
- helm/clients/image_generation/dalle_mini_client.py +1 -1
- helm/clients/image_generation/deep_floyd_client.py +1 -1
- helm/clients/image_generation/huggingface_diffusers_client.py +1 -1
- helm/clients/image_generation/lexica_client.py +1 -1
- helm/clients/image_generation/mindalle/models/__init__.py +6 -6
- helm/clients/image_generation/mindalle/models/stage1/vqgan.py +1 -1
- helm/clients/image_generation/mindalle/models/stage2/transformer.py +1 -1
- helm/clients/image_generation/mindalle/utils/__init__.py +3 -3
- helm/clients/image_generation/mindalle_client.py +1 -1
- helm/clients/image_generation/together_image_generation_client.py +1 -1
- helm/clients/lit_gpt_client.py +2 -2
- helm/clients/mistral_client.py +62 -18
- helm/clients/nvidia_nim_client.py +0 -3
- helm/clients/openai_client.py +255 -21
- helm/clients/palmyra_client.py +2 -6
- helm/clients/reka_client.py +1 -1
- helm/clients/stanfordhealthcare_azure_openai_client.py +58 -0
- helm/clients/stanfordhealthcare_claude_client.py +31 -0
- helm/clients/stanfordhealthcare_google_client.py +43 -0
- helm/clients/stanfordhealthcare_http_model_client.py +93 -0
- helm/clients/stanfordhealthcare_openai_client.py +62 -0
- helm/clients/stanfordhealthcare_shc_openai_client.py +42 -0
- helm/clients/test_client.py +1 -1
- helm/clients/test_together_client.py +6 -1
- helm/clients/together_client.py +69 -7
- helm/clients/upstage_client.py +23 -0
- helm/clients/vertexai_client.py +39 -13
- helm/clients/vision_language/open_flamingo/__init__.py +2 -2
- helm/clients/vision_language/open_flamingo/src/factory.py +3 -3
- helm/clients/vision_language/open_flamingo/src/flamingo.py +2 -2
- helm/clients/vision_language/open_flamingo/src/flamingo_lm.py +2 -2
- helm/clients/vision_language/qwen2_vlm_client.py +175 -0
- helm/clients/vllm_client.py +4 -6
- helm/clients/yi_client.py +0 -3
- helm/common/audio_utils.py +111 -0
- helm/common/cache.py +8 -30
- helm/common/file_caches/local_file_cache.py +1 -1
- helm/common/file_caches/test_local_file_cache.py +1 -1
- helm/common/images_utils.py +2 -2
- helm/common/key_value_store.py +9 -9
- helm/common/media_object.py +2 -2
- helm/common/mongo_key_value_store.py +3 -3
- helm/common/multimodal_request_utils.py +26 -0
- helm/common/reeval_parameters.py +12 -0
- helm/common/request.py +6 -2
- helm/common/response_format.py +18 -0
- helm/common/test_cache.py +1 -48
- helm/common/test_media_object.py +1 -1
- helm/common/tokenization_request.py +0 -9
- helm/config/model_deployments.yaml +1258 -33
- helm/config/model_metadata.yaml +1110 -41
- helm/config/tokenizer_configs.yaml +403 -3
- helm/proxy/cli.py +2 -2
- helm/proxy/example_queries.py +1 -1
- helm/proxy/server.py +11 -13
- helm/proxy/services/remote_service.py +1 -7
- helm/proxy/services/server_service.py +6 -19
- helm/proxy/services/service.py +0 -6
- helm/proxy/services/test_remote_service.py +2 -2
- helm/proxy/services/test_service.py +1 -1
- helm/proxy/static/general.js +122 -0
- helm/proxy/static/help.html +99 -0
- helm/proxy/static/index.css +57 -0
- helm/proxy/static/index.html +40 -0
- helm/proxy/static/index.js +456 -0
- helm/proxy/static/info-icon.png +0 -0
- helm/proxy/test_retry.py +1 -1
- helm/proxy/token_counters/auto_token_counter.py +1 -1
- helm/tokenizers/aleph_alpha_tokenizer.py +1 -1
- helm/tokenizers/caching_tokenizer.py +2 -30
- helm/tokenizers/http_model_tokenizer.py +1 -1
- helm/tokenizers/huggingface_tokenizer.py +2 -2
- helm/tokenizers/lit_gpt_tokenizer.py +1 -1
- helm/tokenizers/test_anthropic_tokenizer.py +6 -2
- helm/tokenizers/test_huggingface_tokenizer.py +1 -1
- helm/tokenizers/test_yalm_tokenizer.py +1 -1
- helm/tokenizers/tiktoken_tokenizer.py +1 -1
- helm/tokenizers/tokenizer.py +3 -1
- helm/tokenizers/yalm_tokenizer.py +3 -3
- helm/tokenizers/yalm_tokenizer_data/test_yalm_tokenizer.py +1 -1
- crfm_helm-0.5.3.dist-info/METADATA +0 -355
- crfm_helm-0.5.3.dist-info/RECORD +0 -699
- helm/benchmark/data_overlap/data_overlap_spec.py +0 -86
- helm/benchmark/data_overlap/export_scenario_text.py +0 -119
- helm/benchmark/data_overlap/light_scenario.py +0 -60
- helm/benchmark/metrics/bhasa_metrics_specs.py +0 -10
- helm/benchmark/static_build/assets/01-694cb9b7.png +0 -0
- helm/benchmark/static_build/assets/accenture-6f97eeda.png +0 -0
- helm/benchmark/static_build/assets/ai21-0eb91ec3.png +0 -0
- helm/benchmark/static_build/assets/aisingapore-6dfc9acf.png +0 -0
- helm/benchmark/static_build/assets/aleph-alpha-7ce10034.png +0 -0
- helm/benchmark/static_build/assets/anthropic-70d8bc39.png +0 -0
- helm/benchmark/static_build/assets/bigscience-7f0400c0.png +0 -0
- helm/benchmark/static_build/assets/cohere-3550c6cb.png +0 -0
- helm/benchmark/static_build/assets/cresta-9e22b983.png +0 -0
- helm/benchmark/static_build/assets/cuhk-8c5631e9.png +0 -0
- helm/benchmark/static_build/assets/eleutherai-b9451114.png +0 -0
- helm/benchmark/static_build/assets/google-06d997ad.png +0 -0
- helm/benchmark/static_build/assets/index-05c76bb1.css +0 -1
- helm/benchmark/static_build/assets/index-58f97dcd.js +0 -10
- helm/benchmark/static_build/assets/meta-5580e9f1.png +0 -0
- helm/benchmark/static_build/assets/microsoft-f5ee5016.png +0 -0
- helm/benchmark/static_build/assets/mistral-18e1be23.png +0 -0
- helm/benchmark/static_build/assets/nvidia-86fa75c1.png +0 -0
- helm/benchmark/static_build/assets/openai-3f8653e4.png +0 -0
- helm/benchmark/static_build/assets/scb10x-204bd786.png +0 -0
- helm/benchmark/static_build/assets/tii-24de195c.png +0 -0
- helm/benchmark/static_build/assets/together-a665a35b.png +0 -0
- helm/benchmark/static_build/assets/tsinghua-keg-97d4b395.png +0 -0
- helm/benchmark/static_build/assets/vhelm-framework-cde7618a.png +0 -0
- helm/benchmark/static_build/assets/vhelm-model-6d812526.png +0 -0
- helm/benchmark/static_build/assets/wellsfargo-a86a6c4a.png +0 -0
- helm/benchmark/static_build/assets/yandex-38e09d70.png +0 -0
- helm/tokenizers/anthropic_tokenizer.py +0 -52
- {crfm_helm-0.5.3.dist-info → crfm_helm-0.5.5.dist-info}/entry_points.txt +0 -0
- {crfm_helm-0.5.3.dist-info → crfm_helm-0.5.5.dist-info/licenses}/LICENSE +0 -0
- {crfm_helm-0.5.3.dist-info → crfm_helm-0.5.5.dist-info}/top_level.txt +0 -0
- /helm/benchmark/{data_overlap → metrics/ifeval}/__init__.py +0 -0
helm/common/test_cache.py
CHANGED
|
@@ -3,9 +3,7 @@ import tempfile
|
|
|
3
3
|
import unittest
|
|
4
4
|
import threading
|
|
5
5
|
|
|
6
|
-
from helm.common.cache import Cache, SqliteCacheConfig,
|
|
7
|
-
|
|
8
|
-
from sqlitedict import SqliteDict
|
|
6
|
+
from helm.common.cache import Cache, SqliteCacheConfig, cache_stats, get_all_from_sqlite
|
|
9
7
|
|
|
10
8
|
|
|
11
9
|
class TestCache(unittest.TestCase):
|
|
@@ -99,51 +97,6 @@ class TestCache(unittest.TestCase):
|
|
|
99
97
|
assert cache_stats.num_computes[self.cache_path] >= num_items
|
|
100
98
|
assert cache_stats.num_computes[self.cache_path] <= num_items * num_threads
|
|
101
99
|
|
|
102
|
-
def test_follower(self):
|
|
103
|
-
cache = Cache(SqliteCacheConfig(self.cache_path))
|
|
104
|
-
request_1 = {"name": "request1"}
|
|
105
|
-
compute_1 = lambda: {"response": "response1"}
|
|
106
|
-
|
|
107
|
-
response, cached = cache.get(request_1, compute_1)
|
|
108
|
-
assert response == {"response": "response1"}
|
|
109
|
-
assert not cached
|
|
110
|
-
assert cache_stats.num_queries[self.cache_path] == 1
|
|
111
|
-
assert cache_stats.num_computes[self.cache_path] == 1
|
|
112
|
-
|
|
113
|
-
follower_cache_file = tempfile.NamedTemporaryFile(delete=False)
|
|
114
|
-
follower_cache_path = follower_cache_file.name
|
|
115
|
-
with follower_cache_file:
|
|
116
|
-
cache_with_follower_config = WithFollowerCacheConfig(
|
|
117
|
-
main=SqliteCacheConfig(self.cache_path),
|
|
118
|
-
follower=SqliteCacheConfig(follower_cache_path),
|
|
119
|
-
)
|
|
120
|
-
cache_with_follower = Cache(cache_with_follower_config)
|
|
121
|
-
|
|
122
|
-
response, cached = cache_with_follower.get(request_1, compute_1)
|
|
123
|
-
assert response == {"response": "response1"}
|
|
124
|
-
assert cached
|
|
125
|
-
assert cache_stats.num_queries[self.cache_path] == 2
|
|
126
|
-
assert cache_stats.num_computes[self.cache_path] == 1
|
|
127
|
-
assert cache_stats.num_queries[follower_cache_path] == 0
|
|
128
|
-
assert cache_stats.num_computes[follower_cache_path] == 0
|
|
129
|
-
|
|
130
|
-
request_2 = {"name": "request2"}
|
|
131
|
-
compute_2 = lambda: {"response": "response2"}
|
|
132
|
-
|
|
133
|
-
response, cached = cache_with_follower.get(request_2, compute_2)
|
|
134
|
-
assert response == {"response": "response2"}
|
|
135
|
-
assert not cached
|
|
136
|
-
assert cache_stats.num_queries[self.cache_path] == 3
|
|
137
|
-
assert cache_stats.num_computes[self.cache_path] == 2
|
|
138
|
-
assert cache_stats.num_queries[follower_cache_path] == 0
|
|
139
|
-
assert cache_stats.num_computes[follower_cache_path] == 0
|
|
140
|
-
|
|
141
|
-
expected_dict = {
|
|
142
|
-
'{"name": "request1"}': {"response": "response1"},
|
|
143
|
-
'{"name": "request2"}': {"response": "response2"},
|
|
144
|
-
}
|
|
145
|
-
self.assertCountEqual(SqliteDict(follower_cache_path).items(), expected_dict.items())
|
|
146
|
-
|
|
147
100
|
def test_get_all_from_sqlite(self):
|
|
148
101
|
cache = Cache(SqliteCacheConfig(self.cache_path))
|
|
149
102
|
num_items = 10 # TODO: Inrcrease to 100
|
helm/common/test_media_object.py
CHANGED
|
@@ -2,15 +2,6 @@ from dataclasses import dataclass
|
|
|
2
2
|
from typing import List, Optional, Union
|
|
3
3
|
|
|
4
4
|
|
|
5
|
-
@dataclass(frozen=True)
|
|
6
|
-
class WindowServiceInfo:
|
|
7
|
-
tokenizer_name: str
|
|
8
|
-
max_sequence_length: int
|
|
9
|
-
max_request_length: int
|
|
10
|
-
end_of_text_token: str
|
|
11
|
-
prefix_token: str
|
|
12
|
-
|
|
13
|
-
|
|
14
5
|
@dataclass(frozen=True)
|
|
15
6
|
class TokenizationRequest:
|
|
16
7
|
"""A `TokenizationRequest` specifies how to tokenize some text."""
|