VLAgeBench: Benchmarking Large Vision-Language Models for Zero-Shot Human Age Estimation
Paper • 2603.26015 • Published
AI & ML interests
A Tour through the 🤗 Hub • Build and Host ML Demos with Gradio & 🤗 • Getting Started with Transformers
Recent Activity
kishormorol
authored 7
papers 23 days ago
An Explainable Ensemble Learning Framework for Crop Classification with Optimized Feature Pyramids and Deep Networks
Paper • 2603.25070 • Published
Can Multi-turn Self-refined Single Agent LMs with Retrieval Solve Hard Coding Problems?
Paper • 2509.00629 • Published
Outcome-Based Education: Evaluating Students' Perspectives Using Transformer
Paper • 2506.17223 • Published
Xolver: Multi-Agent Reasoning with Holistic Experience Learning Just Like an Olympiad Team
Paper • 2506.14234 • Published • 41
Small Language Models: Architectures, Techniques, Evaluation, Problems and Future Adaptation
Paper • 2505.19529 • Published • 5
Multilingual Question Answering in Low-Resource Settings: A Dzongkha-English Benchmark for Foundation Models
Paper • 2505.18638 • Published • 1
BramVanroy
posted an update 2 months ago
Post
2805
**I benchmarked HF buckets against https access for Common Crawl.**
Took me a while to get round to do this but I benchmarked access to Common Crawl via https vs hf buckets. Both experiments were run at night in Europe. I do not think other hardware problems were impacting the speeds since CPU processing time of the non-download pipeline components were highly similar (within 2% identical) and below only the WarcReader speeds of datatrove are used.
Experiment: selected 5 disjoint samples of 64 files each (randomly from the latest crawl; 20,499 docs/file). Those five batches were then processed by 32 single-core tasks with 4GB/core (five batches to calculate CIs). Paired experiment between using https and hf bucket.
- https: 40.0 [39.3-40.6] (seconds per WARC file)
- hf bucket: 172.0 [122.7-221.2]
That is a difference of about 4x in streaming speed. You'll see that https is also more stable (smaller CI).
I also ran raw throughput tests to the endpoints to measure rate limiting (64MiB transfer at 8/32/128/256 concurrent readers) and rate limiting seems not an issue for either: at any of those parallel reader numbers, their respective speeds stay about the same.
Note that, given CC scale, this is still a small test. Rate limiting may become more obvious when processing a full crawl. I do not know whether the https endpoint vs HF bucket will shut you out earlier with which limits.
Took me a while to get round to do this but I benchmarked access to Common Crawl via https vs hf buckets. Both experiments were run at night in Europe. I do not think other hardware problems were impacting the speeds since CPU processing time of the non-download pipeline components were highly similar (within 2% identical) and below only the WarcReader speeds of datatrove are used.
Experiment: selected 5 disjoint samples of 64 files each (randomly from the latest crawl; 20,499 docs/file). Those five batches were then processed by 32 single-core tasks with 4GB/core (five batches to calculate CIs). Paired experiment between using https and hf bucket.
- https: 40.0 [39.3-40.6] (seconds per WARC file)
- hf bucket: 172.0 [122.7-221.2]
That is a difference of about 4x in streaming speed. You'll see that https is also more stable (smaller CI).
I also ran raw throughput tests to the endpoints to measure rate limiting (64MiB transfer at 8/32/128/256 concurrent readers) and rate limiting seems not an issue for either: at any of those parallel reader numbers, their respective speeds stay about the same.
Note that, given CC scale, this is still a small test. Rate limiting may become more obvious when processing a full crawl. I do not know whether the https endpoint vs HF bucket will shut you out earlier with which limits.
gsarti
authored 5
papers 3 months ago
Beyond the Commitment Boundary: Probing Epiphenomenal Chain-of-Thought in Large Reasoning Models
Paper • 2606.13603 • Published
Distilling Formal Logic into Neural Spaces: A Kernel Alignment Approach for Signal Temporal Logic
Paper • 2603.05198 • Published
Bridging Logic and Learning: Decoding Temporal Logic Embeddings via Transformers
Paper • 2507.07808 • Published
Predicting Future Behaviors in Reasoning Models Enables Better Steering
Paper • 2606.11172 • Published • 1
Interpreto: An Explainability Library for Transformers
Paper • 2512.09730 • Published • 1
gsarti
authored 2
papers 7 months ago
lewtun
submitted 2
papers to Daily Papers 8 months ago
kishormorol
authored a
paper 9 months ago
BramVanroy
posted an update about 1 year ago
Post
801
What are currently the best multilingual models with at most 72B parameters? Are Llama 3.3 70B and Qwen 2.5 72B still king?
- 1 reply
BramVanroy
posted an update about 1 year ago
Post
1182
Thanks to popular request, I've just added two subsets to the CommonCrawl-Creative Commons Corpus (C5; BramVanroy/CommonCrawl-CreativeCommons) so that you do not have to do filtering manually
- C5f ( BramVanroy/CommonCrawl-CreativeCommons-fine): only retains high-quality samples that are also present in FineWeb or FineWeb-2;
- C5r (https://huggingface.co/datasets/BramVanroy/CommonCrawl-CreativeCommons-recommended): additional strict filtering that removes samples with license disagreement, non-commercial licenses, and Wikipedia samples. The latter because you should probably get those from a more reliable source that provides better parsed content.
It goes without saying that these filters lead to a massive reduction in quantity. Doc and token counts are given on the dataset pages.
- C5f ( BramVanroy/CommonCrawl-CreativeCommons-fine): only retains high-quality samples that are also present in FineWeb or FineWeb-2;
- C5r (https://huggingface.co/datasets/BramVanroy/CommonCrawl-CreativeCommons-recommended): additional strict filtering that removes samples with license disagreement, non-commercial licenses, and Wikipedia samples. The latter because you should probably get those from a more reliable source that provides better parsed content.
It goes without saying that these filters lead to a massive reduction in quantity. Doc and token counts are given on the dataset pages.