{ "schema_version": 1, "builder_plan": { "scenario": "A narrow, repeated product task may not need your largest model.", "change": "Check whether a released student model fits the task before generating training data.", "baseline": "Your current model and an off-the-shelf small model on the same held-out tasks.", "steps": [ "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.", "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.", "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.", "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test." ], "measure": "Student quality, serving cost, teacher calls, training cost, and failures on unusual inputs.", "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation." }, "builder_review": null, "arxiv_id": "2609.19213", "title": "Layer-wise Curriculum Learning for Efficient LLM Compression", "summary": "The method replaces monolithic full-model knowledge distillation (and, on larger models, training-free pruning baselines) with segmented layer-wise distillation ordered by a curriculum from easy to hard optimization tasks, plus feature caching with multi-threading to align layer outputs. The cost is additional implementation complexity: segment partitioning, a curriculum schedule, and a cache management layer, and it still requires a fine-tuning run on GPU rather than being training-free. It can fail if cached features drift out of alignment with the current student during training, if the curriculum schedule or segment granularity does not transfer to the reader's model family, and because the strongest efficiency evidence is on BERT and GPT-2 while LLaMA/Qwen gains are comparative rather than absolute.", "rationale": "The technique targets exactly the reader's constraint—compression under limited memory and hours—and the >50% memory and time reduction fits a 24 GB GPU, but the claimed gains are workload- and model-specific, so small-scale validation on the reader's own models is required before production adoption.", "technique": "Layer-wise curriculum distillation for LLM compression", "family": "destilacao", "recommendation": "testar", "published": "2026-09-16", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.19213", "independent_implementations": 0, "total_implementations": 0, "stars": 0, "citations": 0, "claims": [ { "claim_id": "claim-01", "statement": "The method replaces monolithic full-model knowledge distillation (and, on larger models, training-free pruning baselines) with segmented layer-wise distillation ordered by a curriculum from easy to hard optimization tasks, plus feature caching with multi-threading to align layer outputs. The cost is additional implementation complexity: segment partitioning, a curriculum schedule, and a cache management layer, and it still requires a fine-tuning run on GPU rather than being training-free. It can fail if cached features drift out of alignment with the current student during training, if the curriculum schedule or segment granularity does not transfer to the reader's model family, and because the strongest efficiency evidence is on BERT and GPT-2 while LLaMA/Qwen gains are comparative rather than absolute.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.19213", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "The paper reports reducing GPU memory usage and training hours by more than 50% on BERT and GPT-2 relative to prior compression methods, and outperforming other pruning methods on LLaMA-family and Qwen models under equal training-hour budgets with lower memory footprint.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.19213", "source_page": null, "source_excerpt": "" } ], "exposure_map": [ { "dimension": "quality", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compute", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "latency", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "operations", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compatibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "security", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "data_and_training", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "reproducibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" } ], "risks": [], "minimum_test": [], "open_questions": [], "independent_tests": [], "lab_reproductions": [], "report_available": false, "validation_tier": "unknown", "evidence_tier": "unknown", "infrastructure_basis": "unknown", "equations": [], "equations_status": "not_fetched", "core_kind": "", "equations_fetched_at": "", "equations_source_url": "", "equations_source_sha256": "", "exposures_reviewed_at": "", "exposures_source_sha256": "" }