Rapidata/10k-human-favorite-number
Viewer • Updated • 11.6k • 61 • 8
Marian Kannwischer
canwiper
AI & ML interests
RLHF & Computer Vision
Recent Activity
liked a dataset 1 day ago
Rapidata/10k-human-favorite-number reacted to jasoncorkill's post with 🚀 29 days ago
Most public benchmarks collapse model performance into one broad preference signal.
That makes it hard to understand which capabilities differentiate between models. It's also almost impossible to inspect the evidence behind it. So @RapidataAI is releasing Benchmark.AI.
We started with an SVG generation benchmark including 42 models, 500 prompts, 1.9M+ human judgements, 300K+ match-ups.
We evaluate models separately on Preference, Alignment and Coherence, while making the prompts, outputs, match-ups and methodology public.
Full dataset: https://huggingface.co/datasets/Rapidata/svg-benchmark
Full benchmark: https://www.benchmark.ai/svg
Methodology feedback and benchmark suggestions very welcome!
liked a dataset 3 months ago
Rapidata/svg-benchmarkOrganizations
liked a
dataset 1 day ago
reacted to jasoncorkill's post with 🚀 29 days ago
Post
2595
Most public benchmarks collapse model performance into one broad preference signal.
That makes it hard to understand which capabilities differentiate between models. It's also almost impossible to inspect the evidence behind it. So @RapidataAI is releasing Benchmark.AI.
We started with an SVG generation benchmark including 42 models, 500 prompts, 1.9M+ human judgements, 300K+ match-ups.
We evaluate models separately on Preference, Alignment and Coherence, while making the prompts, outputs, match-ups and methodology public.
Full dataset: Rapidata/svg-benchmark
Full benchmark: https://www.benchmark.ai/svg
Methodology feedback and benchmark suggestions very welcome!
That makes it hard to understand which capabilities differentiate between models. It's also almost impossible to inspect the evidence behind it. So @RapidataAI is releasing Benchmark.AI.
We started with an SVG generation benchmark including 42 models, 500 prompts, 1.9M+ human judgements, 300K+ match-ups.
We evaluate models separately on Preference, Alignment and Coherence, while making the prompts, outputs, match-ups and methodology public.
Full dataset: Rapidata/svg-benchmark
Full benchmark: https://www.benchmark.ai/svg
Methodology feedback and benchmark suggestions very welcome!
- 13 replies
liked a
dataset 3 months ago
liked a
dataset 9 months ago
liked 2
datasets 10 months ago
liked 2
datasets 11 months ago
liked a
dataset 12 months ago
upvoted 2
articles about 1 year ago
Article
(LoRA) Fine-Tuning FLUX.1-dev on Consumer Hardware
- +3
derekl35, marcsun13, sayakpaul, merve, linoyts
• • 107 Article
Using LoRA for Efficient Stable Diffusion Fine-Tuning
pcuenq, sayakpaul
• • 84 liked 7
datasets about 1 year ago
Rapidata/Recraft-v3-24-7-25_t2i_human_preference
Viewer • Updated • 65.9k • 246 • 10
Rapidata/Imagen-4-ultra-24-7-25_t2i_human_preference
Viewer • Updated • 55.9k • 315 • 8
Rapidata/image-to-video-human-preference-hailuo-02-marey
Viewer • Updated • 98 • 91 • 11
Rapidata/text-2-video-human-preferences-seedance-1-pro
Viewer • Updated • 1.09k • 697 • 9
Rapidata/text-2-video-human-preferences-kling-v2.1-master
Viewer • Updated • 1.19k • 65 • 10
Rapidata/text-2-video-human-preferences-genmo-mochi-1
Viewer • Updated • 1.1k • 127 • 9
Rapidata/multilingual-llm-jokes-4o-claude-gemini
Viewer • Updated • 9.98k • 65 • 14
reacted to jasoncorkill's post with 🚀 over 1 year ago
Post
2448
Imagine you could have an Image Arena score equivalent at each checkpoint during training. We released the first version of just that:
Crowd-Eval
Add one line of code to your training loop and you will have a new real human loss curve in your W&B dashboard.
Thousands of real humans from around the world rating your model in real time at the cost of a few dollars per checkpoint is a game changer.
Check it out here: https://github.com/RapidataAI/crowd-eval
First 5 people to put it in their loop get 100'000 human responses for free! (ping me)
Crowd-Eval
Add one line of code to your training loop and you will have a new real human loss curve in your W&B dashboard.
Thousands of real humans from around the world rating your model in real time at the cost of a few dollars per checkpoint is a game changer.
Check it out here: https://github.com/RapidataAI/crowd-eval
First 5 people to put it in their loop get 100'000 human responses for free! (ping me)
reacted to jasoncorkill's post with 👀 over 1 year ago
Post
4031
Benchmark Update: @google Veo3 (Text-to-Video)
Two months ago, we benchmarked @google ’s Veo2 model. It fell short, struggling with style consistency and temporal coherence, trailing behind Runway, Pika, @tencent , and even @alibaba-pai .
That’s changed.
We just wrapped up benchmarking Veo3, and the improvements are substantial. It outperformed every other model by a wide margin across all key metrics. Not just better, dominating across style, coherence, and prompt adherence. It's rare to see such a clear lead in today’s hyper-competitive T2V landscape.
Dataset coming soon. Stay tuned.
Two months ago, we benchmarked @google ’s Veo2 model. It fell short, struggling with style consistency and temporal coherence, trailing behind Runway, Pika, @tencent , and even @alibaba-pai .
That’s changed.
We just wrapped up benchmarking Veo3, and the improvements are substantial. It outperformed every other model by a wide margin across all key metrics. Not just better, dominating across style, coherence, and prompt adherence. It's rare to see such a clear lead in today’s hyper-competitive T2V landscape.
Dataset coming soon. Stay tuned.
- 5 replies