https://jahanzaibawan.com/ 2026-06-30 weekly 1.0 https://jahanzaibawan.com/assets/profile.jpg https://jahanzaibawan.com/assets/og.png https://jahanzaibawan.com/about.html2026-06-30monthly0.9 https://jahanzaibawan.com/blog.html2026-10-20weekly0.7 https://jahanzaibawan.com/blog/data-science-vs-data-analytics.html2026-10-20monthly0.7 https://jahanzaibawan.com/blog/evaluation-versus-threat-research.html2026-10-20monthly0.7 https://jahanzaibawan.com/blog/what-counts-as-llm.html2026-10-19monthly0.7 https://jahanzaibawan.com/blog/computer-vision-evergreen-skills.html2026-10-19monthly0.7 https://jahanzaibawan.com/blog/ml-syllabi-vs-the-job.html2026-10-18monthly0.7 https://jahanzaibawan.com/blog/internship-lessons-courses-miss.html2026-10-18monthly0.7 https://jahanzaibawan.com/blog/certification-versus-prompt-portfolio.html2026-10-17monthly0.7 https://jahanzaibawan.com/blog/pixels-to-predictions-cv.html2026-10-17monthly0.7 https://jahanzaibawan.com/blog/algorithms-before-the-library.html2026-10-16monthly0.7 https://jahanzaibawan.com/blog/degree-versus-minor-data-science.html2026-10-16monthly0.7 https://jahanzaibawan.com/blog/structuring-a-prompt-from-scratch.html2026-10-15monthly0.7 https://jahanzaibawan.com/blog/evaluate-ml-definitions.html2026-10-15monthly0.7 https://jahanzaibawan.com/blog/entry-level-computer-vision-jobs.html2026-10-14monthly0.7 https://jahanzaibawan.com/blog/ml-engineer-salary-bands-explained.html2026-10-14monthly0.7 https://jahanzaibawan.com/blog/what-prompt-courses-skip.html2026-10-13monthly0.7 https://jahanzaibawan.com/blog/prompt-engineering-salary-claims.html2026-10-13monthly0.7 https://jahanzaibawan.com/blog/ml-algorithms-by-problem.html2026-10-12monthly0.7 https://jahanzaibawan.com/blog/prompt-engineering-transferable-skill.html2026-10-12monthly0.7 https://jahanzaibawan.com/blog/what-training-actually-learns.html2026-10-11monthly0.7 https://jahanzaibawan.com/blog/rag-as-context-engineering.html2026-10-11monthly0.7 https://jahanzaibawan.com/blog/choosing-a-deep-learning-book.html2026-10-10monthly0.7 https://jahanzaibawan.com/blog/deep-learning-versus-classical-statistics.html2026-10-10monthly0.7 https://jahanzaibawan.com/blog/ml-engineer-job-titles-decoded.html2026-10-09monthly0.7 https://jahanzaibawan.com/blog/context-engineering-not-just-length.html2026-10-09monthly0.7 https://jahanzaibawan.com/blog/prompt-engineering-certifications.html2026-10-08monthly0.7 https://jahanzaibawan.com/blog/fine-tuning-means-three-things.html2026-10-08monthly0.7 https://jahanzaibawan.com/blog/fine-tuning-cosmology-vs-ml.html2026-10-07monthly0.7 https://jahanzaibawan.com/blog/what-goodfellow-still-gets-right.html2026-10-07monthly0.7 https://jahanzaibawan.com/blog/ml-definitions-textbooks-disagree.html2026-10-06monthly0.7 https://jahanzaibawan.com/blog/genetic-perturbation-models-deep-learning.html2026-10-06monthly0.7 https://jahanzaibawan.com/blog/context-engineering-vs-prompt-engineering.html2026-10-05monthly0.7 https://jahanzaibawan.com/blog/first-nlp-project-choice.html2026-10-05monthly0.7 https://jahanzaibawan.com/blog/true-cost-of-fine-tuning.html2026-10-04monthly0.7 https://jahanzaibawan.com/blog/model-versus-its-own-baseline.html2026-10-04monthly0.7 https://jahanzaibawan.com/blog/how-cv-and-ml-engineers-read-papers.html2026-10-03monthly0.7 https://jahanzaibawan.com/blog/overfitting-the-leaderboard.html2026-10-03monthly0.7 https://jahanzaibawan.com/blog/vocabulary-trap-ml-job-titles.html2026-10-02monthly0.7 https://jahanzaibawan.com/blog/threat-aware-evaluation-harness.html2026-10-02monthly0.7 https://jahanzaibawan.com/blog/metric-versus-decision.html2026-10-01monthly0.7 https://jahanzaibawan.com/blog/offline-metrics-mislead-recsys.html2026-10-01monthly0.7 https://jahanzaibawan.com/blog/cv-book-versus-course.html2026-09-30monthly0.7 https://jahanzaibawan.com/blog/data-annotation-quality-nlp.html2026-09-30monthly0.7 https://jahanzaibawan.com/blog/rule-based-roots-of-nlp.html2026-09-29monthly0.7 https://jahanzaibawan.com/blog/full-vs-parameter-efficient-tuning.html2026-09-29monthly0.7 https://jahanzaibawan.com/blog/fine-tuning-a-borrowed-word.html2026-09-28monthly0.7 https://jahanzaibawan.com/blog/curriculum-learning-ordering-data.html2026-09-28monthly0.7 https://jahanzaibawan.com/blog/explaining-models-to-skeptical-stakeholders.html2026-09-27monthly0.7 https://jahanzaibawan.com/blog/data-contracts-for-ml-pipelines.html2026-09-27monthly0.7 https://jahanzaibawan.com/blog/multimodal-fusion-text-image.html2026-09-26monthly0.7 https://jahanzaibawan.com/blog/choosing-a-batch-size.html2026-09-26monthly0.7 https://jahanzaibawan.com/blog/feature-engineering-beats-neural-nets.html2026-09-25monthly0.7 https://jahanzaibawan.com/blog/zero-shot-classification-limits.html2026-09-25monthly0.7 https://jahanzaibawan.com/blog/exploration-exploitation-tradeoff-explained.html2026-09-24monthly0.7 https://jahanzaibawan.com/blog/synthetic-data-domain-gap.html2026-09-24monthly0.7 https://jahanzaibawan.com/blog/active-learning-what-to-label.html2026-09-23monthly0.7 https://jahanzaibawan.com/blog/recommenders-beyond-accuracy.html2026-09-23monthly0.7 https://jahanzaibawan.com/blog/skewed-multiclass-imbalance.html2026-09-22monthly0.7 https://jahanzaibawan.com/blog/leakage-lag-features-time-series.html2026-09-22monthly0.7 https://jahanzaibawan.com/blog/grid-random-bayesian-search-compared.html2026-09-21monthly0.7 https://jahanzaibawan.com/blog/explaining-transformers-without-analogy.html2026-09-21monthly0.7 https://jahanzaibawan.com/blog/what-eval-frameworks-assume.html2026-09-20monthly0.7 https://jahanzaibawan.com/blog/shortlisting-mlops-tools-small-team.html2026-09-20monthly0.7 https://jahanzaibawan.com/blog/data-science-salary-bands-skills.html2026-09-19monthly0.7 https://jahanzaibawan.com/blog/personal-mlops-toolkit-before-platforms.html2026-09-19monthly0.7 https://jahanzaibawan.com/blog/beyond-the-held-out-test-set.html2026-09-18monthly0.7 https://jahanzaibawan.com/blog/evaluation-metrics-interview-vs-practice.html2026-09-18monthly0.7 https://jahanzaibawan.com/blog/internship-advert-vs-reality.html2026-09-17monthly0.7 https://jahanzaibawan.com/blog/data-science-modules-real-jobs.html2026-09-17monthly0.7 https://jahanzaibawan.com/blog/transformer-word-two-meanings.html2026-09-16monthly0.7 https://jahanzaibawan.com/blog/weekly-life-of-mlops-engineer.html2026-09-16monthly0.7 https://jahanzaibawan.com/blog/dashboards-hide-silent-failures.html2026-09-15monthly0.7 https://jahanzaibawan.com/blog/reading-mlops-certification-syllabus.html2026-09-15monthly0.7 https://jahanzaibawan.com/blog/bootcamp-vs-self-taught-portfolios.html2026-09-14monthly0.7 https://jahanzaibawan.com/blog/read-evaluation-section-like-reviewer.html2026-09-14monthly0.7 https://jahanzaibawan.com/blog/knowledge-distillation-without-breaking-it.html2026-09-13monthly0.7 https://jahanzaibawan.com/blog/multi-task-learning-shared-backbones.html2026-09-13monthly0.7 https://jahanzaibawan.com/blog/warm-restarts-learning-rate-schedules.html2026-09-12monthly0.7 https://jahanzaibawan.com/blog/weight-decay-vs-dropout.html2026-09-12monthly0.7 https://jahanzaibawan.com/blog/regression-versus-classification-framing.html2026-09-11monthly0.7 https://jahanzaibawan.com/blog/cleaning-corpus-before-model.html2026-09-11monthly0.7 https://jahanzaibawan.com/blog/vision-transformers-from-patches.html2026-09-10monthly0.7 https://jahanzaibawan.com/blog/explaining-attention-to-non-technical.html2026-09-10monthly0.7 https://jahanzaibawan.com/blog/nlp-techniques-small-datasets.html2026-09-09monthly0.7 https://jahanzaibawan.com/blog/pairwise-preference-vs-accuracy.html2026-09-09monthly0.7 https://jahanzaibawan.com/blog/degree-course-portfolio-fine-tuning.html2026-09-08monthly0.7 https://jahanzaibawan.com/blog/llm-benchmark-scores-decoded.html2026-09-08monthly0.7 https://jahanzaibawan.com/blog/what-fine-tuning-actually-means.html2026-09-07monthly0.7 https://jahanzaibawan.com/blog/confidence-intervals-model-evaluation.html2026-09-07monthly0.7 https://jahanzaibawan.com/blog/first-computer-vision-job.html2026-09-06monthly0.7 https://jahanzaibawan.com/blog/years-of-experience-decoded.html2026-09-06monthly0.7 https://jahanzaibawan.com/blog/evaluating-safety-critical-models.html2026-09-05monthly0.7 https://jahanzaibawan.com/blog/strong-evaluation-harness-matters.html2026-09-05monthly0.7 https://jahanzaibawan.com/blog/finetune-vs-retrieval-small-projects.html2026-09-04monthly0.7 https://jahanzaibawan.com/blog/cv-papers-versus-deployment.html2026-09-04monthly0.7 https://jahanzaibawan.com/blog/fair-comparison-checklist-finetuning.html2026-09-03monthly0.7 https://jahanzaibawan.com/blog/managed-platform-dashboard-metrics.html2026-09-03monthly0.7 https://jahanzaibawan.com/blog/failed-attempts-behind-success-stories.html2026-09-02monthly0.7 https://jahanzaibawan.com/blog/cv-engineer-vs-ml-engineer.html2026-09-02monthly0.7 https://jahanzaibawan.com/blog/reading-certifications-honestly.html2026-09-01monthly0.7 https://jahanzaibawan.com/blog/data-science-internships-worth-taking.html2026-09-01monthly0.7 https://jahanzaibawan.com/blog/quantisation-how-low-can-you-go.html2026-08-31monthly0.7 https://jahanzaibawan.com/blog/cold-start-problems-machine-learning.html2026-08-29monthly0.7 https://jahanzaibawan.com/blog/ab-testing-ml-models-production.html2026-08-29monthly0.7 https://jahanzaibawan.com/blog/reading-a-validation-curve.html2026-08-28monthly0.7 https://jahanzaibawan.com/blog/choosing-the-right-loss-function.html2026-08-28monthly0.7 https://jahanzaibawan.com/blog/honest-computer-vision-benchmarking.html2026-08-27monthly0.7 https://jahanzaibawan.com/blog/data-versioning-beyond-git.html2026-08-27monthly0.7 https://jahanzaibawan.com/blog/feature-importance-limits.html2026-08-26monthly0.7 https://jahanzaibawan.com/blog/bagging-boosting-stacking-explained.html2026-08-26monthly0.7 https://jahanzaibawan.com/blog/first-end-to-end-ml-project.html2026-08-25monthly0.7 https://jahanzaibawan.com/blog/maths-before-deep-learning-book.html2026-08-25monthly0.7 https://jahanzaibawan.com/blog/reading-ml-job-adverts.html2026-08-24monthly0.7 https://jahanzaibawan.com/blog/choosing-a-machine-learning-course.html2026-08-24monthly0.7 https://jahanzaibawan.com/blog/what-deep-learning-adds.html2026-08-23monthly0.7 https://jahanzaibawan.com/blog/rmse-mae-when-they-lie.html2026-08-23monthly0.7 https://jahanzaibawan.com/blog/segmentation-metrics-beyond-accuracy.html2026-08-22monthly0.7 https://jahanzaibawan.com/blog/annotation-quality-computer-vision-ceiling.html2026-08-22monthly0.7 https://jahanzaibawan.com/blog/video-versus-image-classification.html2026-08-21monthly0.7 https://jahanzaibawan.com/blog/ocr-classical-to-deep-learning.html2026-08-21monthly0.7 https://jahanzaibawan.com/blog/one-class-anomaly-detection-vision.html2026-08-20monthly0.7 https://jahanzaibawan.com/blog/cnns-versus-vision-transformers.html2026-08-20monthly0.7 https://jahanzaibawan.com/blog/catastrophic-forgetting-fine-tuning.html2026-08-19monthly0.7 https://jahanzaibawan.com/blog/fine-tuning-small-dataset-size.html2026-08-19monthly0.7 https://jahanzaibawan.com/blog/instruction-tuning-vs-fine-tuning.html2026-08-18monthly0.7 https://jahanzaibawan.com/blog/scheduled-retraining-wrong-idea.html2026-08-18monthly0.7 https://jahanzaibawan.com/blog/mlops-for-a-team-of-one.html2026-08-17monthly0.7 https://jahanzaibawan.com/blog/mlops-on-your-laptop-first.html2026-08-17monthly0.7 https://jahanzaibawan.com/blog/mlops-interview-questions-explained.html2026-08-16monthly0.7 https://jahanzaibawan.com/blog/prompt-evaluation-harnesses.html2026-08-16monthly0.7 https://jahanzaibawan.com/blog/chain-of-thought-prompting-tradeoffs.html2026-08-15monthly0.7 https://jahanzaibawan.com/blog/system-prompts-versus-user-prompts.html2026-08-15monthly0.7 https://jahanzaibawan.com/blog/how-many-few-shot-examples.html2026-08-14monthly0.7 https://jahanzaibawan.com/blog/structured-prompt-templates.html2026-08-14monthly0.7 https://jahanzaibawan.com/blog/case-study-hiring-managers-read.html2026-08-13monthly0.7 https://jahanzaibawan.com/blog/pretraining-data-shapes-transformers.html2026-08-13monthly0.7 https://jahanzaibawan.com/blog/segmentation-vs-object-detection.html2026-08-12monthly0.7 https://jahanzaibawan.com/blog/data-scientist-vs-ml-engineer.html2026-08-12monthly0.7 https://jahanzaibawan.com/blog/prompts-that-survive-upgrades.html2026-08-11monthly0.7 https://jahanzaibawan.com/blog/realistic-mlops-course-syllabus.html2026-08-11monthly0.7 https://jahanzaibawan.com/blog/choosing-your-first-cv-project.html2026-08-10monthly0.7 https://jahanzaibawan.com/blog/salary-numbers-data-science-work.html2026-08-10monthly0.7 https://jahanzaibawan.com/blog/degrees-versus-portfolios-interviews.html2026-08-09monthly0.7 https://jahanzaibawan.com/blog/reading-confusion-matrix-by-class.html2026-08-09monthly0.7 https://jahanzaibawan.com/blog/computer-vision-syndrome-vs-cv-evaluation.html2026-08-08monthly0.7 https://jahanzaibawan.com/blog/prompt-engineering-what-transfers.html2026-08-08monthly0.7 https://jahanzaibawan.com/blog/mlops-tools-ten-vs-ten-thousand.html2026-08-07monthly0.7 https://jahanzaibawan.com/blog/why-transformers-need-positional-encodings.html2026-08-07monthly0.7 https://jahanzaibawan.com/blog/sequence-labelling-metrics-accuracy.html2026-08-05monthly0.7 https://jahanzaibawan.com/blog/domain-adaptation-vs-fine-tuning.html2026-08-05monthly0.7 https://jahanzaibawan.com/blog/minimal-model-card-checklist.html2026-08-04monthly0.7 https://jahanzaibawan.com/blog/skills-that-get-you-hired-mle.html2026-08-04monthly0.7 https://jahanzaibawan.com/blog/fine-tuning-vs-few-shot-prompting.html2026-08-03monthly0.7 https://jahanzaibawan.com/blog/macro-micro-weighted-averages.html2026-08-03monthly0.7 https://jahanzaibawan.com/blog/choosing-a-pretrained-transformer.html2026-08-02monthly0.7 https://jahanzaibawan.com/blog/grouped-data-cross-validation-leakage.html2026-08-02monthly0.7 https://jahanzaibawan.com/blog/ml-engineer-week-to-week.html2026-08-01monthly0.7 https://jahanzaibawan.com/blog/nlp-before-deep-learning-refresher.html2026-08-01monthly0.7 https://jahanzaibawan.com/blog/evaluation-harness-unseen-job.html2026-07-31monthly0.7 https://jahanzaibawan.com/blog/what-red-teaming-actually-tests.html2026-07-31monthly0.7 https://jahanzaibawan.com/blog/transformers-without-the-maths.html2026-07-30monthly0.7 https://jahanzaibawan.com/blog/parameter-efficient-fine-tuning-explained.html2026-07-30monthly0.7 https://jahanzaibawan.com/blog/cost-matrix-nobody-writes-down.html2026-07-29monthly0.7 https://jahanzaibawan.com/blog/shap-lime-honest-interpretability.html2026-07-29monthly0.7 https://jahanzaibawan.com/blog/thresholds-beyond-accuracy.html2026-07-28monthly0.7 https://jahanzaibawan.com/blog/synthetic-benchmarks-vs-real-data.html2026-07-28monthly0.7 https://jahanzaibawan.com/blog/what-learning-curves-reveal.html2026-07-27monthly0.7 https://jahanzaibawan.com/blog/bayesian-model-comparison-example.html2026-07-27monthly0.7 https://jahanzaibawan.com/blog/time-series-cross-validation-leakage.html2026-07-26monthly0.7 https://jahanzaibawan.com/blog/multi-label-vs-multi-class-metrics.html2026-07-26monthly0.7 https://jahanzaibawan.com/blog/ablation-studies-isolating-gains.html2026-07-25monthly0.7 https://jahanzaibawan.com/blog/always-try-baseline-models-first.html2026-07-25monthly0.7 https://jahanzaibawan.com/blog/stationarity-assumption-evaluation-metrics.html2026-07-24monthly0.7 https://jahanzaibawan.com/blog/data-drift-vs-concept-drift.html2026-07-24monthly0.7 https://jahanzaibawan.com/blog/nested-cross-validation-model-selection.html2026-07-23monthly0.7 https://jahanzaibawan.com/blog/how-large-should-test-sets-be.html2026-07-23monthly0.7 https://jahanzaibawan.com/blog/threat-modelling-for-ml-systems.html2026-07-22monthly0.7 https://jahanzaibawan.com/blog/modelling-is-the-smaller-half.html2026-07-22monthly0.7 https://jahanzaibawan.com/blog/mlops-engineer-day-to-day.html2026-07-21monthly0.7 https://jahanzaibawan.com/blog/mlops-roadmap-for-researchers.html2026-07-21monthly0.7 https://jahanzaibawan.com/blog/what-fine-tuning-changes.html2026-07-20monthly0.7 https://jahanzaibawan.com/blog/finetuning-vs-prompt-design.html2026-07-20monthly0.7 https://jahanzaibawan.com/blog/build-your-own-eval-harness.html2026-07-19monthly0.7 https://jahanzaibawan.com/blog/model-evaluation-frameworks-compared.html2026-07-19monthly0.7 https://jahanzaibawan.com/blog/classical-vs-deep-vision.html2026-07-18monthly0.7 https://jahanzaibawan.com/blog/vision-models-fail-silently-factory.html2026-07-18monthly0.7 https://jahanzaibawan.com/blog/cv-defect-detection-evaluation.html2026-07-17monthly0.7 https://jahanzaibawan.com/blog/what-a-model-actually-is.html2026-07-17monthly0.7 https://jahanzaibawan.com/blog/machine-learning-versus-ai.html2026-07-16monthly0.7 https://jahanzaibawan.com/blog/ml-versus-deep-learning.html2026-07-16monthly0.7 https://jahanzaibawan.com/blog/significance-testing-two-models.html2026-07-15monthly0.7 https://jahanzaibawan.com/blog/vector-databases-and-rag-explained.html2026-07-15monthly0.7 https://jahanzaibawan.com/blog/prompt-engineering-vs-fine-tuning.html2026-07-14monthly0.7 https://jahanzaibawan.com/blog/ner-rule-statistical-neural.html2026-07-14monthly0.7 https://jahanzaibawan.com/blog/monitoring-model-drift.html2026-07-13monthly0.7 https://jahanzaibawan.com/blog/notebook-to-production-checklist.html2026-07-13monthly0.7 https://jahanzaibawan.com/blog/image-augmentation-what-generalises.html2026-07-12monthly0.7 https://jahanzaibawan.com/blog/map-explained-without-hand-waving.html2026-07-12monthly0.7 https://jahanzaibawan.com/blog/cv-pipelines-distribution-shift.html2026-07-11monthly0.7 https://jahanzaibawan.com/blog/evaluating-llms-beyond-accuracy.html2026-07-11monthly0.7 https://jahanzaibawan.com/blog/llm-leaderboards-hidden-costs.html2026-07-11monthly0.7 https://jahanzaibawan.com/blog/feature-stores-real-problem.html2026-07-11monthly0.7 https://jahanzaibawan.com/blog/model-registries-beyond-weights.html2026-07-10monthly0.7 https://jahanzaibawan.com/blog/before-you-need-kubernetes.html2026-07-10monthly0.7 https://jahanzaibawan.com/blog/smote-synthetic-data-backfires.html2026-07-10monthly0.7 https://jahanzaibawan.com/blog/overfitting-the-validation-set.html2026-07-09monthly0.7 https://jahanzaibawan.com/blog/reproducibility-seeds-configs-runs.html2026-07-09monthly0.7 https://jahanzaibawan.com/blog/augmentation-that-leaks.html2026-07-09monthly0.7 https://jahanzaibawan.com/blog/reading-a-confusion-matrix.html2026-07-08monthly0.7 https://jahanzaibawan.com/blog/bleu-and-its-limits.html2026-07-08monthly0.7 https://jahanzaibawan.com/blog/perplexity-rewards-and-ignores.html2026-07-07monthly0.7 https://jahanzaibawan.com/blog/tokenization-choices-change-results.html2026-07-07monthly0.7 https://jahanzaibawan.com/blog/freeze-fine-tune-or-lora.html2026-07-06monthly0.7 https://jahanzaibawan.com/blog/gradient-clipping-exploding-gradients.html2026-07-06monthly0.7 https://jahanzaibawan.com/blog/learning-rate-first-hyperparameter.html2026-07-06monthly0.7 https://jahanzaibawan.com/blog/dropout-as-an-ensemble.html2026-07-06monthly0.7 https://jahanzaibawan.com/blog/batch-normalization-helps-and-hurts.html2026-07-05monthly0.7 https://jahanzaibawan.com/blog/attention-queries-keys-values.html2026-07-05monthly0.7 https://jahanzaibawan.com/blog/cosine-similarity-embeddings.html2026-07-04monthly0.7 https://jahanzaibawan.com/blog/train-validation-test-splits.html2026-07-04monthly0.7 https://jahanzaibawan.com/blog/optimizing-the-wrong-metric.html2026-07-04monthly0.7 https://jahanzaibawan.com/blog/feature-scaling-which-models-care.html2026-07-03monthly0.7 https://jahanzaibawan.com/blog/class-weights-vs-resampling.html2026-07-03monthly0.7 https://jahanzaibawan.com/blog/early-stopping-vs-regularization.html2026-07-02monthly0.7 https://jahanzaibawan.com/blog/gru-vs-gpt2.html2026-06-20monthly0.7 https://jahanzaibawan.com/blog/uav-data-leakage.html2026-06-12monthly0.7 https://jahanzaibawan.com/projects/automation.html2026-09-22monthly0.8 https://jahanzaibawan.com/projects/uav.html2026-06-30monthly0.8 https://jahanzaibawan.com/projects/nlp.html2026-06-30monthly0.8 https://jahanzaibawan.com/projects/robot.html2026-06-30monthly0.8 https://jahanzaibawan.com/projects/pso.html2026-06-30monthly0.8 https://jahanzaibawan.com/projects/fuzzy.html2026-06-30monthly0.8