├── .github ├── ISSUE_TEMPLATE │ ├── config.yml │ ├── eval_request.md │ ├── faithfulness_gap.md │ └── perf_regression.md └── workflows │ └── ci.yml ├── .gitignore ├── CHANGELOG.md ├── LICENSE ├── README.md ├── TODO.md ├── configs ├── ablations │ └── cms_sparse.yaml ├── data │ ├── continual_segments_sample.yaml │ ├── refinedweb_mixture.yaml │ ├── refinedweb_mixture_filtered.yaml │ ├── refinedweb_mixture_full.yaml │ └── refinedweb_mixture_sample.yaml ├── deepspeed │ └── zero3.json ├── hope │ ├── mid.yaml │ ├── mid_fsdp.yaml │ ├── pilot.yaml │ ├── target.yaml │ └── target_fsdp.yaml ├── mid_smoke.yaml ├── mid_stage2.yaml ├── mid_stage2_smoke.yaml ├── mid_titan_baseline.yaml ├── pilot.yaml ├── pilot_smoke.yaml └── resolved │ └── cms_sparse_eval.yaml ├── docs ├── FSDP_SCALING_GUIDE.md ├── PHASE_2_PLAN.md ├── compute_plan.md ├── continual_eval.md ├── data_pipeline.md ├── env_matrix.md ├── experiments_report.md ├── future_directions.md ├── guide.md ├── planner_convo_01.md ├── release_checklist.md ├── release_plan.md ├── scaling_guidance.md ├── spec_interfaces.md ├── sprint_next_plan.md ├── stability_journal.md ├── stage1_plan.md ├── stage2_plan.md ├── stage2_progress.md ├── templates │ └── checkpoint_report.md └── zeroshot_eval.md ├── eval ├── continual_dummy.json ├── continual_mid_stage2.json ├── continual_mid_stage2_smoke.json ├── continual_mid_stage2_ts10.json ├── continual_mid_stage2_ts10_single120_clip.json ├── continual_mid_stage2_ts10_single140_schedC.json ├── continual_mid_stage2_ts10_single220_schedD.json ├── continual_mid_stage2_ts10_single80.json ├── continual_mid_stage2_ts10_single80lr2e5.json ├── continual_mid_stage2_ts20.json ├── continual_mid_titan_baseline.json ├── continual_pilot.json ├── continual_pilot_cms_nochunk_step5000.json ├── continual_pilot_cms_sparse_step5000.json ├── continual_pilot_multi.json ├── continual_pilot_opt_adamw_step5000.json ├── continual_pilot_opt_muon_step5000.json ├── continual_pilot_selfmod_off_step5000.json ├── continual_pilot_step22000.json ├── continual_pilot_step230000.json ├── continual_pilot_teach05_long_step25000.json ├── continual_pilot_teach05_step2000.json ├── continual_pilot_teach15_long_step25000.json ├── continual_pilot_teach15_step2000.json ├── continual_smoke.json ├── continual_titan.json ├── continual_titan_relaunch_step001000.json ├── continual_titan_step25000.json ├── niah_dummy.json ├── niah_mid_stage2.json ├── niah_mid_stage2_smoke.json ├── niah_mid_stage2_ts10.json ├── niah_mid_stage2_ts10_single120_clip.json ├── niah_mid_stage2_ts10_single140_schedC.json ├── niah_mid_stage2_ts10_single220_schedD.json ├── niah_mid_stage2_ts10_single80.json ├── niah_mid_stage2_ts10_single80lr2e5.json ├── niah_mid_stage2_ts20.json ├── niah_mid_titan_baseline.json ├── niah_pilot.json ├── niah_pilot_cms_nochunk_step5000.json ├── niah_pilot_cms_sparse_step5000.json ├── niah_pilot_opt_adamw_step5000.json ├── niah_pilot_opt_muon_step5000.json ├── niah_pilot_selfmod_off_step5000.json ├── niah_pilot_step22000.json ├── niah_pilot_step230000.json ├── niah_pilot_teach05_long_step25000.json ├── niah_pilot_teach05_step2000.json ├── niah_pilot_teach15_long_step25000.json ├── niah_pilot_teach15_step2000.json ├── niah_smoke.json ├── niah_titan.json ├── niah_titan_relaunch_step001000.json ├── niah_titan_step25000.json ├── passkey_pilot.json ├── passkey_pilot_step230000.json ├── passkey_titan_relaunch_step001000.json ├── passkey_titan_step25000.json ├── pg19_pilot.json ├── pg19_pilot_step230000.json ├── pg19_titan_relaunch_step001000.json ├── pg19_titan_step25000.json ├── zeroshot_full_smoke.json ├── zeroshot_mid_stage2.json ├── zeroshot_mid_stage2_smoke.json ├── zeroshot_mid_stage2_smoke_piqa_baseline.json ├── zeroshot_mid_stage2_smoke_piqa_mem.json ├── zeroshot_mid_stage2_ts10.json ├── zeroshot_mid_stage2_ts10_single120_clip.json ├── zeroshot_mid_stage2_ts10_single140_schedC.json ├── zeroshot_mid_stage2_ts10_single220_schedD.json ├── zeroshot_mid_stage2_ts10_single80.json ├── zeroshot_mid_stage2_ts10_single80lr2e5.json ├── zeroshot_mid_stage2_ts20.json ├── zeroshot_mid_titan_baseline.json ├── zeroshot_pilot.json ├── zeroshot_pilot_cms_nochunk_step5000.json ├── zeroshot_pilot_cms_sparse_step5000.json ├── zeroshot_pilot_debug.json ├── zeroshot_pilot_dummy_piqa.json ├── zeroshot_pilot_opt_adamw_step5000.json ├── zeroshot_pilot_opt_muon_step5000.json ├── zeroshot_pilot_selfmod_off_step5000.json ├── zeroshot_pilot_step22000.json ├── zeroshot_pilot_step230000.json ├── zeroshot_pilot_teach05_long_step25000.json ├── zeroshot_pilot_teach05_step2000.json ├── zeroshot_pilot_teach15_long_step25000.json ├── zeroshot_pilot_teach15_step2000.json ├── zeroshot_smoke.json ├── zeroshot_titan.json ├── zeroshot_titan_relaunch_step001000.json └── zeroshot_titan_step25000.json ├── google_papers ├── Nested_Learning.pdf ├── Nested_Learning │ ├── Nested_Learning.json │ └── Nested_Learning.md ├── TITANs.pdf └── TITANs │ ├── TITANs.json │ └── TITANs.md ├── pyproject.toml ├── reports ├── ablations.md ├── plots │ └── continual_pilot_refinedweb.png └── stage2_smoke.md ├── scripts ├── __init__.py ├── checkpoint │ └── verify.py ├── checks │ └── tokenizer_coverage_guard.py ├── compute │ └── create_reservations.sh ├── data │ ├── check_tokenizer.py │ ├── check_tokenizer_coverage.py │ ├── filter_corpus.py │ ├── process_mixture.py │ ├── run_full.sh │ ├── run_sample.sh │ ├── shard_corpus.py │ ├── train_tokenizer.py │ └── validate_mixture.py ├── eval │ ├── __init__.py │ ├── continual.py │ ├── niah.py │ ├── passkey.py │ ├── pg19_perplexity.py │ ├── plot_forgetting.py │ ├── run_pilot_suite.sh │ └── zeroshot.py ├── package_pilot_release.sh ├── run_cpu_ddp_smoke.sh ├── run_e2e_smoke.sh ├── run_smoke.sh └── tests │ └── run_passkey_smoke.sh ├── src └── nested_learning │ ├── __init__.py │ ├── assoc_memory.py │ ├── backbones.py │ ├── cms.py │ ├── data.py │ ├── hope │ ├── __init__.py │ ├── block.py │ └── self_mod.py │ ├── instrumentation.py │ ├── levels.py │ ├── logging_utils.py │ ├── memorize.py │ ├── model.py │ ├── optim │ ├── __init__.py │ ├── deep.py │ ├── factory.py │ └── manager.py │ ├── titan │ ├── __init__.py │ ├── memory.py │ └── model.py │ ├── tokenizer.py │ ├── tokenizer_coverage.py │ └── training.py ├── tests ├── conftest.py ├── data │ ├── passkey_corpus.txt │ ├── tiny_tokenizer.model │ └── tiny_tokenizer.vocab ├── test_cms.py ├── test_eval_builders.py ├── test_hope_block.py ├── test_levels.py ├── test_memorization.py ├── test_model.py ├── test_optim.py └── test_teach_signal.py ├── train.py ├── train_deepspeed.py ├── train_dist.py ├── train_fsdp.py └── uv.lock /.github/ISSUE_TEMPLATE/config.yml: -------------------------------------------------------------------------------- 1 | blank_issues_enabled: false 2 | contact_links: [] 3 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/eval_request.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/.github/ISSUE_TEMPLATE/eval_request.md -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/faithfulness_gap.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/.github/ISSUE_TEMPLATE/faithfulness_gap.md -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/perf_regression.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/.github/ISSUE_TEMPLATE/perf_regression.md -------------------------------------------------------------------------------- /.github/workflows/ci.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/.github/workflows/ci.yml -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/.gitignore -------------------------------------------------------------------------------- /CHANGELOG.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/CHANGELOG.md -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/LICENSE -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/README.md -------------------------------------------------------------------------------- /TODO.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/TODO.md -------------------------------------------------------------------------------- /configs/ablations/cms_sparse.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/ablations/cms_sparse.yaml -------------------------------------------------------------------------------- /configs/data/continual_segments_sample.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/data/continual_segments_sample.yaml -------------------------------------------------------------------------------- /configs/data/refinedweb_mixture.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/data/refinedweb_mixture.yaml -------------------------------------------------------------------------------- /configs/data/refinedweb_mixture_filtered.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/data/refinedweb_mixture_filtered.yaml -------------------------------------------------------------------------------- /configs/data/refinedweb_mixture_full.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/data/refinedweb_mixture_full.yaml -------------------------------------------------------------------------------- /configs/data/refinedweb_mixture_sample.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/data/refinedweb_mixture_sample.yaml -------------------------------------------------------------------------------- /configs/deepspeed/zero3.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/deepspeed/zero3.json -------------------------------------------------------------------------------- /configs/hope/mid.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/hope/mid.yaml -------------------------------------------------------------------------------- /configs/hope/mid_fsdp.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/hope/mid_fsdp.yaml -------------------------------------------------------------------------------- /configs/hope/pilot.yaml: -------------------------------------------------------------------------------- 1 | defaults: 2 | - /pilot 3 | -------------------------------------------------------------------------------- /configs/hope/target.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/hope/target.yaml -------------------------------------------------------------------------------- /configs/hope/target_fsdp.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/hope/target_fsdp.yaml -------------------------------------------------------------------------------- /configs/mid_smoke.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/mid_smoke.yaml -------------------------------------------------------------------------------- /configs/mid_stage2.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/mid_stage2.yaml -------------------------------------------------------------------------------- /configs/mid_stage2_smoke.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/mid_stage2_smoke.yaml -------------------------------------------------------------------------------- /configs/mid_titan_baseline.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/mid_titan_baseline.yaml -------------------------------------------------------------------------------- /configs/pilot.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/pilot.yaml -------------------------------------------------------------------------------- /configs/pilot_smoke.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/pilot_smoke.yaml -------------------------------------------------------------------------------- /configs/resolved/cms_sparse_eval.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/configs/resolved/cms_sparse_eval.yaml -------------------------------------------------------------------------------- /docs/FSDP_SCALING_GUIDE.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/FSDP_SCALING_GUIDE.md -------------------------------------------------------------------------------- /docs/PHASE_2_PLAN.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/PHASE_2_PLAN.md -------------------------------------------------------------------------------- /docs/compute_plan.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/compute_plan.md -------------------------------------------------------------------------------- /docs/continual_eval.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/continual_eval.md -------------------------------------------------------------------------------- /docs/data_pipeline.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/data_pipeline.md -------------------------------------------------------------------------------- /docs/env_matrix.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/env_matrix.md -------------------------------------------------------------------------------- /docs/experiments_report.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/experiments_report.md -------------------------------------------------------------------------------- /docs/future_directions.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/future_directions.md -------------------------------------------------------------------------------- /docs/guide.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/guide.md -------------------------------------------------------------------------------- /docs/planner_convo_01.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/planner_convo_01.md -------------------------------------------------------------------------------- /docs/release_checklist.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/release_checklist.md -------------------------------------------------------------------------------- /docs/release_plan.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/release_plan.md -------------------------------------------------------------------------------- /docs/scaling_guidance.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/scaling_guidance.md -------------------------------------------------------------------------------- /docs/spec_interfaces.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/spec_interfaces.md -------------------------------------------------------------------------------- /docs/sprint_next_plan.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/sprint_next_plan.md -------------------------------------------------------------------------------- /docs/stability_journal.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/stability_journal.md -------------------------------------------------------------------------------- /docs/stage1_plan.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/stage1_plan.md -------------------------------------------------------------------------------- /docs/stage2_plan.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/stage2_plan.md -------------------------------------------------------------------------------- /docs/stage2_progress.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/stage2_progress.md -------------------------------------------------------------------------------- /docs/templates/checkpoint_report.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/templates/checkpoint_report.md -------------------------------------------------------------------------------- /docs/zeroshot_eval.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/docs/zeroshot_eval.md -------------------------------------------------------------------------------- /eval/continual_dummy.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_dummy.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2_smoke.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2_smoke.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2_ts10.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2_ts10.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2_ts10_single120_clip.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2_ts10_single120_clip.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2_ts10_single140_schedC.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2_ts10_single140_schedC.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2_ts10_single220_schedD.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2_ts10_single220_schedD.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2_ts10_single80.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2_ts10_single80.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2_ts10_single80lr2e5.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2_ts10_single80lr2e5.json -------------------------------------------------------------------------------- /eval/continual_mid_stage2_ts20.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_stage2_ts20.json -------------------------------------------------------------------------------- /eval/continual_mid_titan_baseline.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_mid_titan_baseline.json -------------------------------------------------------------------------------- /eval/continual_pilot.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot.json -------------------------------------------------------------------------------- /eval/continual_pilot_cms_nochunk_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_cms_nochunk_step5000.json -------------------------------------------------------------------------------- /eval/continual_pilot_cms_sparse_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_cms_sparse_step5000.json -------------------------------------------------------------------------------- /eval/continual_pilot_multi.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_multi.json -------------------------------------------------------------------------------- /eval/continual_pilot_opt_adamw_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_opt_adamw_step5000.json -------------------------------------------------------------------------------- /eval/continual_pilot_opt_muon_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_opt_muon_step5000.json -------------------------------------------------------------------------------- /eval/continual_pilot_selfmod_off_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_selfmod_off_step5000.json -------------------------------------------------------------------------------- /eval/continual_pilot_step22000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_step22000.json -------------------------------------------------------------------------------- /eval/continual_pilot_step230000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_step230000.json -------------------------------------------------------------------------------- /eval/continual_pilot_teach05_long_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_teach05_long_step25000.json -------------------------------------------------------------------------------- /eval/continual_pilot_teach05_step2000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_teach05_step2000.json -------------------------------------------------------------------------------- /eval/continual_pilot_teach15_long_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_teach15_long_step25000.json -------------------------------------------------------------------------------- /eval/continual_pilot_teach15_step2000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_pilot_teach15_step2000.json -------------------------------------------------------------------------------- /eval/continual_smoke.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_smoke.json -------------------------------------------------------------------------------- /eval/continual_titan.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_titan.json -------------------------------------------------------------------------------- /eval/continual_titan_relaunch_step001000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_titan_relaunch_step001000.json -------------------------------------------------------------------------------- /eval/continual_titan_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/continual_titan_step25000.json -------------------------------------------------------------------------------- /eval/niah_dummy.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_dummy.json -------------------------------------------------------------------------------- /eval/niah_mid_stage2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_mid_stage2.json -------------------------------------------------------------------------------- /eval/niah_mid_stage2_smoke.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_mid_stage2_smoke.json -------------------------------------------------------------------------------- /eval/niah_mid_stage2_ts10.json: -------------------------------------------------------------------------------- 1 | { 2 | "niah_2048": 0.0 3 | } -------------------------------------------------------------------------------- /eval/niah_mid_stage2_ts10_single120_clip.json: -------------------------------------------------------------------------------- 1 | { 2 | "niah_2048": 0.0 3 | } -------------------------------------------------------------------------------- /eval/niah_mid_stage2_ts10_single140_schedC.json: -------------------------------------------------------------------------------- 1 | { 2 | "niah_2048": 0.4 3 | } -------------------------------------------------------------------------------- /eval/niah_mid_stage2_ts10_single220_schedD.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_mid_stage2_ts10_single220_schedD.json -------------------------------------------------------------------------------- /eval/niah_mid_stage2_ts10_single80.json: -------------------------------------------------------------------------------- 1 | { 2 | "niah_2048": 0.0 3 | } -------------------------------------------------------------------------------- /eval/niah_mid_stage2_ts10_single80lr2e5.json: -------------------------------------------------------------------------------- 1 | { 2 | "niah_2048": 0.0 3 | } -------------------------------------------------------------------------------- /eval/niah_mid_stage2_ts20.json: -------------------------------------------------------------------------------- 1 | { 2 | "niah_2048": 0.0 3 | } -------------------------------------------------------------------------------- /eval/niah_mid_titan_baseline.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_mid_titan_baseline.json -------------------------------------------------------------------------------- /eval/niah_pilot.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot.json -------------------------------------------------------------------------------- /eval/niah_pilot_cms_nochunk_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_cms_nochunk_step5000.json -------------------------------------------------------------------------------- /eval/niah_pilot_cms_sparse_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_cms_sparse_step5000.json -------------------------------------------------------------------------------- /eval/niah_pilot_opt_adamw_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_opt_adamw_step5000.json -------------------------------------------------------------------------------- /eval/niah_pilot_opt_muon_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_opt_muon_step5000.json -------------------------------------------------------------------------------- /eval/niah_pilot_selfmod_off_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_selfmod_off_step5000.json -------------------------------------------------------------------------------- /eval/niah_pilot_step22000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_step22000.json -------------------------------------------------------------------------------- /eval/niah_pilot_step230000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_step230000.json -------------------------------------------------------------------------------- /eval/niah_pilot_teach05_long_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_teach05_long_step25000.json -------------------------------------------------------------------------------- /eval/niah_pilot_teach05_step2000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_teach05_step2000.json -------------------------------------------------------------------------------- /eval/niah_pilot_teach15_long_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_teach15_long_step25000.json -------------------------------------------------------------------------------- /eval/niah_pilot_teach15_step2000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_pilot_teach15_step2000.json -------------------------------------------------------------------------------- /eval/niah_smoke.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_smoke.json -------------------------------------------------------------------------------- /eval/niah_titan.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_titan.json -------------------------------------------------------------------------------- /eval/niah_titan_relaunch_step001000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_titan_relaunch_step001000.json -------------------------------------------------------------------------------- /eval/niah_titan_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/niah_titan_step25000.json -------------------------------------------------------------------------------- /eval/passkey_pilot.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/passkey_pilot.json -------------------------------------------------------------------------------- /eval/passkey_pilot_step230000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/passkey_pilot_step230000.json -------------------------------------------------------------------------------- /eval/passkey_titan_relaunch_step001000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/passkey_titan_relaunch_step001000.json -------------------------------------------------------------------------------- /eval/passkey_titan_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/passkey_titan_step25000.json -------------------------------------------------------------------------------- /eval/pg19_pilot.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/pg19_pilot.json -------------------------------------------------------------------------------- /eval/pg19_pilot_step230000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/pg19_pilot_step230000.json -------------------------------------------------------------------------------- /eval/pg19_titan_relaunch_step001000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/pg19_titan_relaunch_step001000.json -------------------------------------------------------------------------------- /eval/pg19_titan_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/pg19_titan_step25000.json -------------------------------------------------------------------------------- /eval/zeroshot_full_smoke.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_full_smoke.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_smoke.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_smoke.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_smoke_piqa_baseline.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_smoke_piqa_baseline.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_smoke_piqa_mem.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_smoke_piqa_mem.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_ts10.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_ts10.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_ts10_single120_clip.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_ts10_single120_clip.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_ts10_single140_schedC.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_ts10_single140_schedC.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_ts10_single220_schedD.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_ts10_single220_schedD.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_ts10_single80.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_ts10_single80.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_ts10_single80lr2e5.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_ts10_single80lr2e5.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_stage2_ts20.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_stage2_ts20.json -------------------------------------------------------------------------------- /eval/zeroshot_mid_titan_baseline.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_mid_titan_baseline.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_cms_nochunk_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_cms_nochunk_step5000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_cms_sparse_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_cms_sparse_step5000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_debug.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_debug.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_dummy_piqa.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_dummy_piqa.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_opt_adamw_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_opt_adamw_step5000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_opt_muon_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_opt_muon_step5000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_selfmod_off_step5000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_selfmod_off_step5000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_step22000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_step22000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_step230000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_step230000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_teach05_long_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_teach05_long_step25000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_teach05_step2000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_teach05_step2000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_teach15_long_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_teach15_long_step25000.json -------------------------------------------------------------------------------- /eval/zeroshot_pilot_teach15_step2000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_pilot_teach15_step2000.json -------------------------------------------------------------------------------- /eval/zeroshot_smoke.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_smoke.json -------------------------------------------------------------------------------- /eval/zeroshot_titan.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_titan.json -------------------------------------------------------------------------------- /eval/zeroshot_titan_relaunch_step001000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_titan_relaunch_step001000.json -------------------------------------------------------------------------------- /eval/zeroshot_titan_step25000.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/eval/zeroshot_titan_step25000.json -------------------------------------------------------------------------------- /google_papers/Nested_Learning.pdf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/google_papers/Nested_Learning.pdf -------------------------------------------------------------------------------- /google_papers/Nested_Learning/Nested_Learning.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/google_papers/Nested_Learning/Nested_Learning.json -------------------------------------------------------------------------------- /google_papers/Nested_Learning/Nested_Learning.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/google_papers/Nested_Learning/Nested_Learning.md -------------------------------------------------------------------------------- /google_papers/TITANs.pdf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/google_papers/TITANs.pdf -------------------------------------------------------------------------------- /google_papers/TITANs/TITANs.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/google_papers/TITANs/TITANs.json -------------------------------------------------------------------------------- /google_papers/TITANs/TITANs.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/google_papers/TITANs/TITANs.md -------------------------------------------------------------------------------- /pyproject.toml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/pyproject.toml -------------------------------------------------------------------------------- /reports/ablations.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/reports/ablations.md -------------------------------------------------------------------------------- /reports/plots/continual_pilot_refinedweb.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/reports/plots/continual_pilot_refinedweb.png -------------------------------------------------------------------------------- /reports/stage2_smoke.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/reports/stage2_smoke.md -------------------------------------------------------------------------------- /scripts/__init__.py: -------------------------------------------------------------------------------- 1 | # Makes `scripts` a package for intra-eval imports. 2 | -------------------------------------------------------------------------------- /scripts/checkpoint/verify.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/checkpoint/verify.py -------------------------------------------------------------------------------- /scripts/checks/tokenizer_coverage_guard.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/checks/tokenizer_coverage_guard.py -------------------------------------------------------------------------------- /scripts/compute/create_reservations.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/compute/create_reservations.sh -------------------------------------------------------------------------------- /scripts/data/check_tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/check_tokenizer.py -------------------------------------------------------------------------------- /scripts/data/check_tokenizer_coverage.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/check_tokenizer_coverage.py -------------------------------------------------------------------------------- /scripts/data/filter_corpus.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/filter_corpus.py -------------------------------------------------------------------------------- /scripts/data/process_mixture.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/process_mixture.py -------------------------------------------------------------------------------- /scripts/data/run_full.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/run_full.sh -------------------------------------------------------------------------------- /scripts/data/run_sample.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/run_sample.sh -------------------------------------------------------------------------------- /scripts/data/shard_corpus.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/shard_corpus.py -------------------------------------------------------------------------------- /scripts/data/train_tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/train_tokenizer.py -------------------------------------------------------------------------------- /scripts/data/validate_mixture.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/data/validate_mixture.py -------------------------------------------------------------------------------- /scripts/eval/__init__.py: -------------------------------------------------------------------------------- 1 | # Eval utilities package marker. 2 | -------------------------------------------------------------------------------- /scripts/eval/continual.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/eval/continual.py -------------------------------------------------------------------------------- /scripts/eval/niah.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/eval/niah.py -------------------------------------------------------------------------------- /scripts/eval/passkey.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/eval/passkey.py -------------------------------------------------------------------------------- /scripts/eval/pg19_perplexity.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/eval/pg19_perplexity.py -------------------------------------------------------------------------------- /scripts/eval/plot_forgetting.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/eval/plot_forgetting.py -------------------------------------------------------------------------------- /scripts/eval/run_pilot_suite.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/eval/run_pilot_suite.sh -------------------------------------------------------------------------------- /scripts/eval/zeroshot.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/eval/zeroshot.py -------------------------------------------------------------------------------- /scripts/package_pilot_release.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/package_pilot_release.sh -------------------------------------------------------------------------------- /scripts/run_cpu_ddp_smoke.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/run_cpu_ddp_smoke.sh -------------------------------------------------------------------------------- /scripts/run_e2e_smoke.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/run_e2e_smoke.sh -------------------------------------------------------------------------------- /scripts/run_smoke.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/run_smoke.sh -------------------------------------------------------------------------------- /scripts/tests/run_passkey_smoke.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/scripts/tests/run_passkey_smoke.sh -------------------------------------------------------------------------------- /src/nested_learning/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/__init__.py -------------------------------------------------------------------------------- /src/nested_learning/assoc_memory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/assoc_memory.py -------------------------------------------------------------------------------- /src/nested_learning/backbones.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/backbones.py -------------------------------------------------------------------------------- /src/nested_learning/cms.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/cms.py -------------------------------------------------------------------------------- /src/nested_learning/data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/data.py -------------------------------------------------------------------------------- /src/nested_learning/hope/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/nested_learning/hope/block.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/hope/block.py -------------------------------------------------------------------------------- /src/nested_learning/hope/self_mod.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/hope/self_mod.py -------------------------------------------------------------------------------- /src/nested_learning/instrumentation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/instrumentation.py -------------------------------------------------------------------------------- /src/nested_learning/levels.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/levels.py -------------------------------------------------------------------------------- /src/nested_learning/logging_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/logging_utils.py -------------------------------------------------------------------------------- /src/nested_learning/memorize.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/memorize.py -------------------------------------------------------------------------------- /src/nested_learning/model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/model.py -------------------------------------------------------------------------------- /src/nested_learning/optim/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/nested_learning/optim/deep.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/optim/deep.py -------------------------------------------------------------------------------- /src/nested_learning/optim/factory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/optim/factory.py -------------------------------------------------------------------------------- /src/nested_learning/optim/manager.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/optim/manager.py -------------------------------------------------------------------------------- /src/nested_learning/titan/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/nested_learning/titan/memory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/titan/memory.py -------------------------------------------------------------------------------- /src/nested_learning/titan/model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/titan/model.py -------------------------------------------------------------------------------- /src/nested_learning/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/tokenizer.py -------------------------------------------------------------------------------- /src/nested_learning/tokenizer_coverage.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/tokenizer_coverage.py -------------------------------------------------------------------------------- /src/nested_learning/training.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/src/nested_learning/training.py -------------------------------------------------------------------------------- /tests/conftest.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/conftest.py -------------------------------------------------------------------------------- /tests/data/passkey_corpus.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/data/passkey_corpus.txt -------------------------------------------------------------------------------- /tests/data/tiny_tokenizer.model: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/data/tiny_tokenizer.model -------------------------------------------------------------------------------- /tests/data/tiny_tokenizer.vocab: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/data/tiny_tokenizer.vocab -------------------------------------------------------------------------------- /tests/test_cms.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/test_cms.py -------------------------------------------------------------------------------- /tests/test_eval_builders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/test_eval_builders.py -------------------------------------------------------------------------------- /tests/test_hope_block.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/test_hope_block.py -------------------------------------------------------------------------------- /tests/test_levels.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/test_levels.py -------------------------------------------------------------------------------- /tests/test_memorization.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/test_memorization.py -------------------------------------------------------------------------------- /tests/test_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/test_model.py -------------------------------------------------------------------------------- /tests/test_optim.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/test_optim.py -------------------------------------------------------------------------------- /tests/test_teach_signal.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/tests/test_teach_signal.py -------------------------------------------------------------------------------- /train.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/train.py -------------------------------------------------------------------------------- /train_deepspeed.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/train_deepspeed.py -------------------------------------------------------------------------------- /train_dist.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/train_dist.py -------------------------------------------------------------------------------- /train_fsdp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/train_fsdp.py -------------------------------------------------------------------------------- /uv.lock: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/kmccleary3301/nested_learning/HEAD/uv.lock --------------------------------------------------------------------------------