Runtime (Fine-tuning to convergence, average per pruned model) of DNABERT-2 117M, one layer pruned, fine-tuned to convergence (Hegde et al. 2025) on Enformer eQTL causal-variant dataset, Genomics Long-Range Benchmark version (Hegde et al. 2025)
around 1 Runtime (Fine-tuning to convergence, average per pruned model)
Methods
- Tested configuration
- DNABERT-2 117M, one layer pruned, fine-tuned to convergence (Hegde et al. 2025)
- Protocol
- Fine-tuning wall time to convergence on the Enformer eQTL dataset (Hegde et al. 2025 Methods)
- Dataset
- Enformer eQTL causal-variant dataset, Genomics Long-Range Benchmark version (Hegde et al. 2025)
- Procedure
- Fine-tuning wall time to convergence on the Enformer eQTL dataset (Hegde et al. 2025 Methods)
- Evaluation
- DNABERT-2 pruned models, fine-tuning time to convergence (Hegde et al. 2025)
- Coverage
- Not reported scored / Not reported eligible
- Uncertainty
- Not reported by the source
- Evidence
- Independent external evaluation · source checkedReaping the Fruits of LLM Pruning: Towards Small Language Models for Efficient Non-Coding Variant Effect Prediction · Methods paragraph 3: 'taking around 1 h per pruned model on average'
A source-checked result verifies the numerical transcription, not every model or protocol detail. Evaluation metadata: source checked. Source checked does not mean independently reproduced.
Reproduction
- Split
- Benchmark train split for fine-tuning; evaluation split for evaluation
- Adaptation
- LoRA parameter-efficient fine-tuning (peft; r=1, lora_alpha=32, lora_dropout=0.1)
- Scoring implementation
- Not reported
No execution recipe has been verified for this exact configuration and evaluation. A benchmark's general instructions may use different inputs, splits or model settings.
Reproducing this published result requires matching its model configuration, data, split and scorer. Source checking or a successful smoke test does not establish score reproduction.
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
1 evidence row matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Reported result around 1 Individual claims | Reaping the Fruits of LLM Pruning: Towards Small Language Models for Efficient Non-Coding Variant Effect Prediction Methods paragraph 3: 'taking around 1 h per pruned model on average' Version: Genes 16(11):1358, 2025; PMC12652071 full-text XML | source checked ["source-hash-verification","independent-cell-check"] · 2026-10-14 independent paper Audit detailsValue matches, including the printed hedge word. Transcribed, not reproduced. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record Extraction artifact SHA-256: |
Sources and history
Release 2026-10-22-1b7cf027e4f6 · Record review: source checked
1 source record and release history
- Reaping the Fruits of LLM Pruning: Towards Small Language Models for Efficient Non-Coding Variant Effect Prediction · Original source · Genes 16(11):1358, 2025; PMC12652071 full-text XML
Technical metadata and extraction receipts
Stable ID: gfm-guide-20261014-result-hegde-dnabert-2-pruned-convergence-time
- metric
- runtime
- metric direction
- lower
- unit
- hour
- printed value
- around 1
- numeric value
- 1
- source locator
- Methods paragraph 3: 'taking around 1 h per pruned model on average'
- missing metadata
- uncertainty: reason: unreported
- review
- method: source-hash-verification; independent-cell-check; method note: Methods paragraph 3 sentence asserted verbatim in the reviewer's ElementTree text of the XML (review/check_gfm_guide.py); XML SHA-256 matches.; reviewer: claude; reviewer note: Independent Claude (Opus 5.5) review agent for PR #150; did not extract this batch; no human review claimed; date: 2026-10-14; artifact sha256: 9ea68b688637b40f0aeced7c905822f6b14c352a3a935c61c407945f3d66ef47; retrieval url: https://www.ebi.ac.uk/europepmc/webservices/rest/PMC12652071/fullTextXML; note: Value matches, including the printed hedge word. Transcribed, not reproduced.
- metric qualifier
- Fine-tuning to convergence, average per pruned model
- unit detail
- Printed as 'around 1 h'
- approximate
- true
- hardware
- 2 × Quadro RTX 8000