As America turns 250, we put together 250 open AI milestones from the US: open models, datasets, demos, papers, and tools that helped shape the field. They go from attention is all you need, pytorch, gpt2, ULMFIT, llama, imagenet, Lora and hundreds more.

They are a reminder of what made America the world’s engine of innovation:

• Open science
• Open competition
• Open ecosystems

Builders and scientists building on each other’s work, challenging each other, remixing ideas, and pushing the frontier forward. That is America at its best!

And that philosophy is at risk right now in AI.

In the coming months, scientists and AI builders will have to decide what side of history they want to be on: an AI future shaped by openness, transparency, participation, and competition, or one increasingly controlled behind closed doors by a few actors optimizing for money, secrecy, and gatekeeping.

Let’s make the next 250 even more open!

https://huggingface.co/collections/clem/250-years-of-america-250-open-ai-milestones…

250 Years of America, 250 Open AI Milestones - a clem Collection

Source: https://huggingface.co/collections/clem/250-years-of-america-250-open-ai-milestones

clem’s Collections

For the 250th birthday of the USA: the 250 most important American open models, datasets, papers and Spaces, ranked from #1 down. Happy 4th of July!

• — #### Attention Is All You Need Paper•1706.03762•PublishedJun 12, 2017 • 128 Note#1 — Attention Is All You Need (Google, 2017). The paper that introduced the transformer.
• — #### PyTorch: An Imperative Style, High-Performance Deep Learning Library Paper•1912.01703•PublishedDec 3, 2019 • 2 Note#2 — PyTorch (Meta, 2019). The framework open AI runs on.
• — #### google-bert/bert-base-uncased Fill-Mask• 0.1B• UpdatedFeb 19, 2024 • 66.3M • • 2.7k Note#3 — BERT (Google, 2018). The model that kicked off the open model era and the Hub itself.
• — #### openai-community/gpt2 Text Generation• 0.1B• UpdatedFeb 19, 2024 • 13.4M • 3.33k Note#4 — GPT-2 (OpenAI, 2019). The original open LLM.
• — #### Efficient Estimation of Word Representations in Vector Space Paper•1301.3781•PublishedJan 16, 2013 • 8 Note#5 — word2vec: Efficient Estimation of Word Representations (Google, 2013).
• — #### HuggingFace’s Transformers: State-of-the-art Natural Language Processing Paper•1910.03771•PublishedOct 9, 2019 • 23 Note#6 — HuggingFace’s Transformers: State-of-the-art NLP (2019). The library open AI runs on — with its siblings Datasets, Hub, and TGI.
• — #### Universal Language Model Fine-tuning for Text Classification Paper•1801.06146•PublishedJan 18, 2018 • 8 Note#7 — ULMFiT (Howard & Ruder, fast.ai, 2018). Universal Language Model Fine-tuning — transfer learning for NLP before BERT.
• — #### meta-llama/Llama-2-7b Text Generation• UpdatedApr 17, 2024 • 113 • 4.51k Note#8 — Llama 2 (Meta, 2023). Open weights went commercial and the ecosystem exploded.
• — #### openai/whisper-large-v3 Automatic Speech Recognition• 2B• UpdatedAug 12, 2024 • 5.89M • • 5.92k Note#9 — Whisper (OpenAI). Open speech recognition for the whole world.
• — #### openai/gpt-oss-120b Text Generation• 120B• UpdatedAug 26, 2025 • 4.3M • • 4.95k Note#10 — gpt-oss-120b (OpenAI, 2025). OpenAI’s return to open weights, Apache 2.0.
• — #### ILSVRC/imagenet-1k Viewer• UpdatedSep 17, 2025 • 1.43M • 112k • 849 Note#11 — ImageNet (Fei-Fei Li, Princeton/Stanford). The dataset that launched deep learning.
• — #### TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems Paper•1603.04467•PublishedMar 14, 2016 Note#12 — TensorFlow (Google, 2016). Large-scale ML, open sourced.
• — #### openai/clip-vit-large-patch14 Zero-Shot Image Classification• 0.4B• UpdatedSep 15, 2023 • 12.7M • 2.05k Note#13 — CLIP (OpenAI, 2021). Connected vision and language; powers image generation everywhere.
• — #### Language Models are Few-Shot Learners Paper•2005.14165•PublishedMay 28, 2020 • 20 Note#14 — GPT-3: Language Models are Few-Shot Learners (OpenAI, 2020). In-context learning arrives.
• — #### Deep contextualized word representations Paper•1802.05365•PublishedFeb 15, 2018 • 1 Note#15 — ELMo: Deep Contextualized Word Representations (AI2/UW, 2018). Contextual embeddings arrive.
• — #### stanfordnlp/glove Note#16 — GloVe (Stanford, 2014). Global word vectors; with word2vec, the embeddings that started it all.
• — #### meta-llama/Meta-Llama-3-8B Text Generation• 8B• UpdatedSep 27, 2024 • 1.32M • • 6.59k Note#17 — Llama 3 (Meta, 2024). The most downloaded LLM family in Hub history.
• — #### rajpurkar/squad Viewer• UpdatedMar 4, 2024 • 98.2k • 188k • 368 Note#18 — SQuAD (Stanford, 2016). The QA benchmark that defined an era of NLP.
• — #### LLaMA: Open and Efficient Foundation Language Models Paper•2302.13971•PublishedFeb 27, 2023 • 25 Note#19 — LLaMA: Open and Efficient Foundation Language Models (Meta, 2023).
• — #### distilbert/distilbert-base-uncased Fill-Mask• 67M• UpdatedMay 6, 2024 • 8.62M • • 908 Note#20 — DistilBERT (Hugging Face). Distillation for the masses; billions of downloads.
• — #### Distilling the Knowledge in a Neural Network Paper•1503.02531•PublishedMar 9, 2015 • 2 Note#21 — Distilling the Knowledge in a Neural Network (Hinton et al., Google, 2015).
• — #### google-t5/t5-base Translation• 0.2B• UpdatedFeb 14, 2024 • 1.85M • • 778 Note#22 — T5 (Google, 2019). Text-to-text everything.
• — #### nyu-mll/glue Viewer• UpdatedJan 30, 2024 • 1.49M • 402k • 512 Note#23 — GLUE (NYU/UW/DeepMind-adjacent, led from NYC). The benchmark suite of the BERT era.
• — #### FacebookAI/roberta-base Fill-Mask• 0.1B• UpdatedFeb 19, 2024 • 10.3M • • 621 Note#24 — RoBERTa (Meta, 2019). BERT, robustly optimized; still a workhorse.
• — #### allenai/c4 Viewer• UpdatedJan 9, 2024 • 10.4B • 1.36M • 607 Note#25 — C4 (Google/AI2). The Colossal Clean Crawled Corpus behind T5 and countless LLMs.
• — #### Scaling Laws for Neural Language Models Paper•2001.08361•PublishedJan 23, 2020 • 10 Note#26 — Scaling Laws for Neural Language Models (Kaplan et al., OpenAI, 2020).
• — #### Adam: A Method for Stochastic Optimization Paper•1412.6980•PublishedDec 22, 2014 • 5 Note#27 — Adam: A Method for Stochastic Optimization (2015). The optimizer that trains almost everything on this list.
• — #### Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift Paper•1502.03167•PublishedFeb 11, 2015 • 2 Note#28 — Batch Normalization (Google, 2015). The trick that made very deep networks trainable.
• — #### Generative Adversarial Networks Paper•1406.2661•PublishedJun 10, 2014 • 5 Note#29 — Generative Adversarial Networks (Goodfellow et al., 2014). The paper that launched generative modeling.
• — #### A Style-Based Generator Architecture for Generative Adversarial Networks Paper•1812.04948•PublishedDec 12, 2018 • 3 Note#30 — StyleGAN (NVIDIA, 2018). The GAN era’s crown jewel; ‘this person does not exist.’
• — #### Denoising Diffusion Probabilistic Models Paper•2006.11239•PublishedJun 19, 2020 • 9 Note#31 — Denoising Diffusion Probabilistic Models (UC Berkeley, 2020). DDPM.
• — #### Scalable Diffusion Models with Transformers Paper•2212.09748•PublishedDec 19, 2022 • 17 Note#32 — Scalable Diffusion Models with Transformers — DiT (Berkeley/NYU, 2022). The backbone of modern image and video generation.
• — #### Deep Residual Learning for Image Recognition Paper•1512.03385•PublishedDec 10, 2015 • 17 Note#33 — Deep Residual Learning for Image Recognition (Microsoft, 2015).
• — #### EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks Paper•1905.11946•PublishedMay 28, 2019 • 4 Note#34 — EfficientNet (Google, 2019). Compound scaling for vision.
• — Track, rank and evaluate open LLMs and chatbots Note#35 — The Open LLM Leaderboard. Set the pace of the open model race.
• — #### EleutherAI/pile UpdatedMay 3, 2023 • 6.22k • 500 Note#36 — The Pile (EleutherAI, 2020). 825GB that trained a generation of open LLMs.
• — #### Training language models to follow instructions with human feedback Paper•2203.02155•PublishedMar 4, 2022 • 24 Note#37 — InstructGPT (OpenAI, 2022). RLHF as we know it.
• — #### GPT-4 Technical Report Paper•2303.08774•PublishedMar 15, 2023 • 7 Note#38 — GPT-4 Technical Report (OpenAI, 2023).
• — #### cais/mmlu Viewer• UpdatedMar 8, 2024 • 231k • 433k • 783 Note#39 — MMLU (Hendrycks et al., Berkeley). The default knowledge benchmark for years.
• — #### bigcode/starcoder Text Generation• 16B• UpdatedOct 8, 2024 • 23k • 2.97k Note#40 — StarCoder (BigCode: Hugging Face + ServiceNow). Open code generation at scale.
• — #### EleutherAI/gpt-j-6b Text Generation• UpdatedJun 21, 2023 • 258k • 1.53k Note#41 — GPT-J (EleutherAI, 2021). The grassroots open GPT-3 rival.
• — #### meta-llama/Llama-3.1-405B Text Generation• 406B• UpdatedSep 25, 2024 • 209k • • 978 Note#42 — Llama 3.1 405B (Meta, 2024). The first open frontier-scale model.
• — #### bigscience/bloom Text Generation• 176B• UpdatedJul 28, 2023 • 5.32k • 5.02k Note#43 — BLOOM (BigScience, 2022). 176B params, 1000+ researchers, led by Hugging Face. The first open LLM at GPT-3 scale.
• — #### Evaluate & Evaluation on the Hub: Better Best Practices for Data and Model Measurements Paper•2210.01970•PublishedSep 30, 2022 • 14 Note#44 — Evaluate & Evaluation on the Hub (Hugging Face, 2022). Open evaluation infrastructure for the Hub.
• — #### Model Cards for Model Reporting Paper•1810.03993•PublishedOct 5, 2018 • 7 Note#45 — Model Cards for Model Reporting (Mitchell et al., Google, 2018). The documentation standard on every Hub repo.
• — #### Gradio: Hassle-Free Sharing and Testing of ML Models in the Wild Paper•1906.02569•PublishedJun 6, 2019 • 2 Note#46 — Gradio (2019). The interface layer of open ML — every Space runs on it.
• — View the LMArena leaderboard in full‑screen Note#47 — Chatbot Arena / LMArena (LMSYS, UC Berkeley). Community model rankings via millions of human votes.
• — #### google/gemma-7b Text Generation• 9B• UpdatedJun 27, 2024 • 29.4k • • 3.37k Note#48 — Gemma (Google, 2024). Google returns to open weights.
• — #### Learning Transferable Visual Models From Natural Language Supervision Paper•2103.00020•PublishedFeb 26, 2021 • 22 Note#49 — Learning Transferable Visual Models From Natural Language Supervision — CLIP (OpenAI).
• — #### Datasets: A Community Library for Natural Language Processing Paper•2109.02846•PublishedSep 7, 2021 • 15 Note#50 — Datasets: A Community Library for NLP (Hugging Face, 2021). The library serving every dataset on this list.
• — #### xai-org/grok-1 Text Generation• UpdatedMar 28, 2024 • 177 • 2.42k Note#51 — Grok-1 (xAI, 2024). 314B parameters, the largest open release of its time.
• — #### HuggingFaceFW/fineweb Viewer• UpdatedJul 11, 2025 • 52.5B • 446k • 2.92k Note#52 — FineWeb (Hugging Face, NYC-HQ). The modern open pretraining corpus.
• — #### bigcode/the-stack Viewer• UpdatedApr 13, 2023 • 546M • 17.7k • 1.03k Note#53 — The Stack (BigCode). 6TB of permissively licensed source code.
• — #### Chain-of-Thought Prompting Elicits Reasoning in Large Language Models Paper•2201.11903•PublishedJan 28, 2022 • 15 Note#54 — Chain-of-Thought Prompting (Google, 2022).
• — #### nvidia/Llama-3.1-Nemotron-70B-Instruct-HF Text Generation• 71B• UpdatedApr 13, 2025 • 37.6k • • 2.07k Note#55 — Nemotron 70B (NVIDIA, 2024). Open post-training done right.
• — #### LoRA: Low-Rank Adaptation of Large Language Models Paper•2106.09685•PublishedJun 17, 2021 • 63 Note#56 — LoRA: Low-Rank Adaptation (Microsoft, 2021). Made fine-tuning affordable.
• — #### Anthropic/hh-rlhf Viewer• UpdatedMay 26, 2023 • 169k • 29.6k • 1.81k Note#57 — HH-RLHF (Anthropic). The canonical open human preference dataset.
• — #### facebook/bart-large Feature Extraction• UpdatedJun 3, 2022 • 136k • • 201 Note#58 — BART (Meta, 2019). Denoising seq2seq; summarization staple.
• — #### openai/gsm8k Benchmark• UpdatedMar 23 • 17.6k • 955k • 1.43k Note#59 — GSM8K (OpenAI). Grade school math that humbled LLMs for years.
• — #### FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness Paper•2205.14135•PublishedMay 27, 2022 • 15 Note#60 — FlashAttention (Stanford, 2022). Made long context practical.
• — #### EleutherAI/gpt-neox-20b Text Generation• 21B• UpdatedJan 31, 2024 • 440k • 584 Note#61 — GPT-NeoX-20B (EleutherAI, 2022). Open scale before it was cool.
• — #### togethercomputer/RedPajama-Data-1T Viewer• UpdatedJun 17, 2024 • 1.73M • 2.08k • 1.18k Note#62 — RedPajama 1T (Together AI). Open reproduction of the LLaMA training data.
• — #### Direct Preference Optimization: Your Language Model is Secretly a Reward Model Paper•2305.18290•PublishedMay 29, 2023 • 66 Note#63 — Direct Preference Optimization (Stanford, 2023). RLHF without the RL.
• — #### microsoft/phi-2 Text Generation• 3B• UpdatedDec 8, 2025 • 849k • 3.48k Note#64 — Phi-2 (Microsoft, 2023). Small models, textbook data, big results.
• — #### openai/gpt-oss-20b Text Generation• 22B• UpdatedAug 26, 2025 • 7.01M • • 4.77k Note#65 — gpt-oss-20b (OpenAI, 2025). Frontier reasoning on a laptop.
• — #### openai/openai_humaneval Viewer• UpdatedJan 4, 2024 • 164 • 227k • 395 Note#66 — HumanEval (OpenAI). THE code generation benchmark.
• — #### Segment Anything Paper•2304.02643•PublishedApr 5, 2023 • 6 Note#67 — Segment Anything (Meta, 2023).
• — #### facebook/sam-vit-huge Mask Generation• 0.6B• UpdatedJan 11, 2024 • 445k • 198 Note#68 — SAM (Meta, 2023). Segment anything, literally.
• — #### google/vit-base-patch16-224 Image Classification• 86.6M• UpdatedSep 5, 2023 • 5.49M • • 981 Note#69 — ViT (Google). Transformers eat computer vision.
• — #### Skylion007/openwebtext Viewer• UpdatedDec 26, 2025 • 8.01M • 69.7k • 523 Note#70 — OpenWebText. The open recreation of GPT-2’s training data.
• — #### lmsys/vicuna-13b-v1.5 Text Generation• UpdatedMar 17, 2024 • 15.1k • • 243 Note#71 — Vicuna (LMSYS, Berkeley). The chatbot that showed open models could chat.
• — #### tatsu-lab/alpaca Viewer• UpdatedMay 22, 2023 • 52k • 77.9k • 1k Note#72 — Alpaca (Stanford, 2023). 52K instructions that launched the instruction-tuning boom.
• — #### Constitutional AI: Harmlessness from AI Feedback Paper•2212.08073•PublishedDec 15, 2022 • 4 Note#73 — Constitutional AI (Anthropic, 2022).
• — #### In-context Learning and Induction Heads Paper•2209.11895•PublishedSep 24, 2022 • 2 Note#74 — In-context Learning and Induction Heads (Anthropic, 2022). Mechanistic interpretability arrives.
• — #### allenai/OLMo-2-1124-7B 7B• UpdatedJan 6, 2025 • 90.6k • 68 Note#75 — OLMo 2 (AI2, Seattle). Truly open: weights, data, code, logs.
• — #### allenai/dolma UpdatedApr 17, 2024 • 4.12k • 1.05k Note#76 — Dolma (AI2). 3T tokens, fully documented open pretraining data.
• — #### Salesforce/codegen-16B-mono Text Generation• UpdatedJan 31, 2025 • 232 • 126 Note#77 — CodeGen (Salesforce, 2022). The first big open code LLM, pre-StarCoder.
• — #### databricks/databricks-dolly-15k Viewer• UpdatedJun 30, 2023 • 15k • 41.5k • 994 Note#78 — dolly-15k (Databricks). Human-written instructions, employee-sourced.
• — #### HuggingFaceH4/zephyr-7b-beta Text Generation• 7B• UpdatedOct 16, 2024 • 186k • • 1.85k Note#79 — Zephyr (Hugging Face H4). DPO goes mainstream.
• — #### Efficient Memory Management for Large Language Model Serving with PagedAttention Paper•2309.06180•PublishedSep 12, 2023 • 61 Note#80 — PagedAttention / vLLM (UC Berkeley, 2023). How open models get served.
• — #### meta-llama/Llama-3.1-8B-Instruct Text Generation• 8B• UpdatedSep 25, 2024 • 9M • • 6.24k Note#81 — Llama 3.1 8B Instruct (Meta). The default open chat model of 2024-25.
• — #### togethercomputer/RedPajama-INCITE-7B-Base Text Generation• UpdatedJun 6, 2023 • 386 • 92 Note#82 — RedPajama-INCITE (Together AI, 2023). Fully open reproduction, trained on US supercomputers.
• — #### HuggingFaceFW/fineweb-edu Viewer• UpdatedJul 11, 2025 • 3.5B • 382k • 1.18k Note#83 — FineWeb-Edu (Hugging Face). Education-filtered web data, new pretraining default.
• — #### microsoft/deberta-v3-base Fill-Mask• UpdatedSep 22, 2022 • 2.55M • • 429 Note#84 — DeBERTa v3 (Microsoft). Still the encoder to beat on GLUE-style tasks.
• — #### OpenAI Gym Paper•1606.01540•PublishedJun 5, 2016 Note#85 — OpenAI Gym (2016). The playground that standardized reinforcement learning.
• — #### facebook/wav2vec2-base-960h Automatic Speech Recognition• 94.4M• UpdatedNov 14, 2022 • 1.51M • 398 Note#86 — wav2vec 2.0 (Meta, 2020). Self-supervised speech revolution.
• — #### openslr/librispeech_asr Viewer• UpdatedJul 25, 2025 • 585k • 75.6k • 229 Note#87 — LibriSpeech (Johns Hopkins). The canonical ASR benchmark.
• — #### facebook/dinov2-base Image Feature Extraction• 86.6M• UpdatedJan 17, 2024 • 1.47M • 182 Note#88 — DINOv2 (Meta). Self-supervised vision features for everything.
• — #### Masked Autoencoders Are Scalable Vision Learners Paper•2111.06377•PublishedNov 11, 2021 • 6 Note#89 — Masked Autoencoders Are Scalable Vision Learners (Meta, 2021).
• — #### facebook/detr-resnet-50 Object Detection• 41.6M• UpdatedApr 10, 2024 • 2.06M • • 959 Note#90 — DETR (Meta, 2020). Object detection, transformerized.
• — #### HuggingFaceM4/COCO UpdatedDec 15, 2022 • 4.45k • 33 Note#91 — COCO (Microsoft, 2014). Common Objects in Context; vision’s other bedrock.
• — #### microsoft/resnet-50 Image Classification• 25.6M• UpdatedFeb 13, 2024 • 580k • • 498 Note#92 — ResNet-50 (Microsoft, 2015). The backbone of a decade of computer vision.
• — #### Ultralytics/YOLOv8 Object Detection• Updated11 days ago • 10.2k • 374 Note#93 — YOLOv8 (Ultralytics). Real-time detection, descended from Joseph Redmon’s YOLO (UW).
• — #### You Only Look Once: Unified, Real-Time Object Detection Paper•1506.02640•PublishedJun 8, 2015 • 3 Note#94 — You Only Look Once (Joseph Redmon, UW, 2015). Real-time object detection.
• — #### facebook/bart-large-mnli Zero-Shot Classification• 0.4B• UpdatedSep 5, 2023 • 3.31M • • 1.58k Note#95 — BART-MNLI (Meta). The zero-shot classification workhorse of the Hub.
• — #### stanfordnlp/imdb Viewer• UpdatedJan 4, 2024 • 100k • 183k • 389 Note#96 — IMDB (Stanford, 2011). The hello-world of sentiment analysis.
• — #### ylecun/mnist Viewer• UpdatedAug 8, 2024 • 70k • 133k • 253 Note#97 — MNIST (Yann LeCun, AT&T Bell Labs/NYU). Where it all started.
• — #### xlnet/xlnet-base-cased Text Generation• UpdatedJan 24, 2023 • 438k • 82 Note#98 — XLNet (Google + CMU, 2019).
• — #### Sequence to Sequence Learning with Neural Networks Paper•1409.3215•PublishedSep 10, 2014 • 4 Note#99 — Sequence to Sequence Learning (Sutskever et al., Google, 2014).
• — #### XGBoost: A Scalable Tree Boosting System Paper•1603.02754•PublishedMar 9, 2016 • 2 Note#100 — XGBoost (University of Washington, 2016). The workhorse of practical ML for a decade.
• — #### google/electra-base-discriminator UpdatedFeb 29, 2024 • 39M • 134 Note#101 — ELECTRA (Google + Stanford, 2020). Pretraining efficiency pioneer.
• — #### Salesforce/wikitext Viewer• UpdatedJan 4, 2024 • 3.71M • 1.34M • 729 Note#102 — WikiText (Salesforce, 2016). The classic language modeling benchmark.
• — #### ptb-text-only/ptb_text_only UpdatedJan 18, 2024 • 21k • 20 Note#103 — Penn Treebank (UPenn, 1993). The dataset NLP grew up on.
• — #### Bag of Tricks for Efficient Text Classification Paper•1607.01759•PublishedJul 6, 2016 Note#104 — fastText (Meta FAIR, 2016). Fast, practical open NLP for everyone.
• — #### stanfordnlp/sst2 Viewer• UpdatedJan 4, 2024 • 70k • 30.8k • 162 Note#105 — SST-2 (Stanford). Sentiment treebank, GLUE staple.
• — #### openai-community/openai-gpt Text Generation• 0.1B• UpdatedFeb 19, 2024 • 214k • 296 Note#106 — GPT-1 (OpenAI, 2018). The one that started the G-P-T.
• — #### google/flan-t5-xxl 11B• UpdatedJul 27, 2023 • 13k • 1.29k Note#107 — Flan-T5 (Google, 2022). Instruction tuning at scale, open.
• — #### facebook/opt-6.7b Text Generation• UpdatedJan 24, 2023 • 143k • 120 Note#108 — OPT (Meta, 2022). Open Pre-trained Transformers, with logbooks and all.
• — #### mozilla-foundation/common_voice_17_0 UpdatedOct 24, 2025 • 4.79k • 29 Note#109 — Common Voice (Mozilla). Crowdsourced speech in 100+ languages.
• — #### suno/bark Text-to-Speech• UpdatedOct 4, 2023 • 18.7k • 1.54k Note#110 — Bark (Suno, 2023). Open generative text-to-audio, pre-Suno-the-app.
• — Generate realistic speech and sounds from typed text Note#111 — The Bark demo Space that made TTS fun.
• — #### facebook/musicgen-small Text-to-Audio• 0.6B• UpdatedNov 17, 2023 • 277k • 496 Note#112 — MusicGen (Meta, 2023). Open music generation.
• — #### openai/jukebox-1b-lyrics Feature Extraction• UpdatedNov 10, 2022 • 76 • 21 Note#113 — Jukebox (OpenAI, 2020). Music generation with singing, years early.
• — Generate music from a text description and optional melody Note#114 — MusicGen demo (Meta). Everyone’s first AI song.
• — #### keithito/lj_speech UpdatedAug 14, 2024 • 1.11k • 62 Note#115 — LJ Speech. The public-domain voice behind a thousand TTS models.
• — #### microsoft/speecht5_tts Text-to-Speech• UpdatedNov 8, 2023 • 78.8k • 836 Note#116 — SpeechT5 (Microsoft). Unified speech-text pretraining.
• — #### sesame/csm-1b Text-to-Speech• 2B• UpdatedDec 1, 2025 • 282k • 2.41k Note#117 — CSM (Sesame, 2025). Open conversational speech that sounded human.
• — #### meta-llama/Llama-4-Scout-17B-16E-Instruct Image-Text-to-Text• 109B• UpdatedMay 22, 2025 • 722k • • 1.32k Note#118 — Llama 4 Scout (Meta, 2025). Natively multimodal MoE.
• — #### nvidia/HelpSteer2 Viewer• UpdatedDec 18, 2024 • 21.4k • 5.48k • 452 Note#119 — HelpSteer2 (NVIDIA). Open preference data for reward models.
• — #### microsoft/phi-4 Text Generation• 15B• UpdatedNov 24, 2025 • 897k • • 2.27k Note#120 — Phi-4 (Microsoft, 2024). Small model, big reasoning.
• — #### Textbooks Are All You Need Paper•2306.11644•PublishedJun 20, 2023 • 159 Note#121 — Textbooks Are All You Need (Microsoft, 2023). The phi thesis.
• — #### roneneldan/TinyStories Viewer• UpdatedAug 12, 2024 • 2.14M • 84.6k • 1.05k Note#122 — TinyStories (Microsoft). How small can a coherent LM be?
• — #### EleutherAI/pythia-12b Text Generation• 12B• UpdatedJul 9, 2024 • 142k • 145 Note#123 — Pythia (EleutherAI). The scaling suite for LLM science.
• — #### EleutherAI/gpt-neo-2.7B Text Generation• 3B• UpdatedJul 9, 2023 • 41.5k • 503 Note#124 — GPT-Neo (EleutherAI, 2021). The first open GPT-3-style replication.
• — #### cerebras/Cerebras-GPT-13B Text Generation• UpdatedNov 22, 2023 • 539 • 652 Note#125 — Cerebras-GPT (2023). Chinchilla-optimal, wafer-scale trained.
• — #### togethercomputer/RedPajama-Data-V2 UpdatedNov 21, 2024 • 9.39k • 404 Note#126 — RedPajama V2 (Together AI). 30T tokens of open web data.
• — #### Snowflake/snowflake-arctic-instruct Text Generation• 479B• UpdatedMay 21, 2024 • 35.1k • 362 Note#127 — Arctic (Snowflake, 2024). Enterprise-grade open MoE.
• — #### Mamba: Linear-Time Sequence Modeling with Selective State Spaces Paper•2312.00752•PublishedDec 1, 2023 • 152 Note#128 — Mamba: Linear-Time Sequence Modeling (CMU + Princeton, 2023).
• — #### Snowflake/snowflake-arctic-embed-m Sentence Similarity• 0.1B• UpdatedDec 13, 2024 • 471k • • 166 Note#129 — Arctic Embed (Snowflake). Open retrieval embeddings.
• — #### ibm-granite/granite-3.3-8b-instruct Text Generation• 8B• UpdatedMay 12, 2025 • 97k • 158 Note#130 — Granite (IBM). A century-old American company, Apache 2.0 models.
• — #### apple/DCLM-7B 7B• UpdatedJul 26, 2024 • 194 • 833 Note#131 — DCLM (Apple, 2024). DataComp-LM, open data-centric science.
• — #### apple/OpenELM-3B Text Generation• 3B• UpdatedFeb 28, 2025 • 210 • 130 Note#132 — OpenELM (Apple, 2024). Apple ships open weights.
• — #### amazon/chronos-t5-large Time Series Forecasting• 0.7B• UpdatedNov 21, 2025 • 96.6k • 178 Note#133 — Chronos (Amazon). LLMs for time series forecasting.
• — #### google/timesfm-1.0-200m Time Series Forecasting• UpdatedMay 17, 2024 • 144 • 828 Note#134 — TimesFM (Google). Foundation model for forecasting.
• — #### allenai/Molmo-7B-D-0924 Image-Text-to-Text• 8B• UpdatedDec 15, 2025 • 29.8k • 567 Note#135 — Molmo (AI2, 2024). Open VLM matching closed ones, with open data.
• — #### adept/fuyu-8b Image-Text-to-Text• 9B• UpdatedNov 4, 2023 • 134k • 1.02k Note#136 — Fuyu (Adept, 2023). Multimodal without an image encoder.
• — #### vikhyatk/moondream2 Image-Text-to-Text• 2B• UpdatedSep 23, 2025 • 1.67M • 1.42k Note#137 — Moondream (2024). The beloved tiny vision-language model.
• — #### microsoft/Florence-2-large Image-Text-to-Text• 0.8B• UpdatedAug 4, 2025 • 712k • 1.83k Note#138 — Florence-2 (Microsoft, 2024). One small model, every vision task.
• — #### BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models Paper•2301.12597•PublishedJan 30, 2023 • 3 Note#139 — BLIP-2 (Salesforce, 2023). Bootstrapping VLMs from frozen parts.
• — #### Salesforce/blip2-opt-2.7b Image-Text-to-Text• 4B• UpdatedFeb 3, 2025 • 655k • 446 Note#140 — BLIP-2 (Salesforce). Vision-language bridge.
• — #### Salesforce/blip-image-captioning-base Image-to-Text• UpdatedFeb 3, 2025 • 1.73M • 865 Note#141 — BLIP (Salesforce). The Hub’s default image captioner for years.
• — #### Visual Instruction Tuning Paper•2304.08485•PublishedApr 17, 2023 • 21 Note#142 — Visual Instruction Tuning — LLaVA (UW-Madison + Microsoft, 2023).
• — #### llava-hf/llava-1.5-7b-hf Image-Text-to-Text• 7B• UpdatedJun 6, 2025 • 3.16M • 367 Note#143 — LLaVA 1.5 (UW-Madison/Microsoft). Open multimodal chat for everyone.
• — #### facebook/chameleon-7b Image-Text-to-Text• 7B• UpdatedJul 23, 2024 • 229k • 202 Note#144 — Chameleon (Meta, 2024). Early-fusion multimodal.
• — #### google/paligemma-3b-pt-224 Image-Text-to-Text• 3B• UpdatedSep 21, 2024 • 274k • 495 Note#145 — PaliGemma (Google, 2024). Open VLM building block.
• — #### google/siglip-so400m-patch14-384 Zero-Shot Image Classification• 0.9B• UpdatedSep 26, 2024 • 1.75M • 680 Note#146 — SigLIP (Google). The CLIP successor inside most modern VLMs.
• — #### MMMU/MMMU Viewer• UpdatedApr 21 • 11.6k • 55.5k • 330 Note#147 — MMMU (Ohio State-led). The multimodal exam.
• — #### lmsys/lmsys-chat-1m Viewer• UpdatedJul 27, 2024 • 1M • 6.11k • 933 Note#148 — LMSYS-Chat-1M (Berkeley). One million real conversations.
• — #### Judging LLM-as-a-judge with MT-Bench and Chatbot Arena Paper•2306.05685•PublishedJun 9, 2023 • 43 Note#149 — Judging LLM-as-a-Judge with MT-Bench (LMSYS, Berkeley, 2023).
• — #### NousResearch/Hermes-3-Llama-3.1-8B Text Generation• 8B• UpdatedSep 8, 2024 • 359k • • 465 Note#150 — Hermes 3 (Nous Research). The American fine-tuning scene at its best.
• — #### teknium/OpenHermes-2.5 Viewer• UpdatedApr 15, 2024 • 1M • 15.6k • 867 Note#151 — OpenHermes 2.5 (Teknium). The community SFT dataset.
• — #### Open-Orca/OpenOrca Viewer• UpdatedFeb 19, 2025 • 2.94M • 17.1k • 1.56k Note#152 — OpenOrca. Open reproduction of Microsoft’s Orca data.
• — #### berkeley-nest/Starling-LM-7B-alpha Text Generation• 7B• UpdatedMar 20, 2024 • 1.83k • • 560 Note#153 — Starling (Berkeley). Open RLAIF.
• — #### allenai/Llama-3.1-Tulu-3-8B Text Generation• 8B• UpdatedFeb 13, 2025 • 3.81k • • 179 Note#154 — Tulu 3 (AI2). Fully open post-training recipe.
• — #### allenai/tulu-3-sft-mixture Viewer• UpdatedDec 2, 2024 • 939k • 19k • 251 Note#155 — Tulu 3 SFT mixture (AI2). The open post-training data stack.
• — #### QLoRA: Efficient Finetuning of Quantized LLMs Paper•2305.14314•PublishedMay 23, 2023 • 62 Note#156 — QLoRA (University of Washington, 2023). Fine-tune 65B on one GPU.
• — #### AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration Paper•2306.00978•PublishedJun 1, 2023 • 13 Note#157 — AWQ (MIT, 2023). Activation-aware quantization, everywhere now.
• — #### bigcode/starcoder2-15b Text Generation• 16B• UpdatedJun 5, 2024 • 10.9k • 674 Note#158 — StarCoder2 (BigCode, 2024).
• — #### codellama/CodeLlama-7b-hf Text Generation• 7B• UpdatedApr 12, 2024 • 244k • 377 Note#159 — Code Llama (Meta, 2023). Open code models at every size.
• — #### Evaluating Large Language Models Trained on Code Paper•2107.03374•PublishedJul 7, 2021 • 11 Note#160 — Evaluating LLMs Trained on Code — Codex/HumanEval (OpenAI, 2021).
• — #### google-research-datasets/mbpp Viewer• UpdatedJan 4, 2024 • 1.4k • 165k • 232 Note#161 — MBPP (Google). Mostly Basic Python Problems.
• — #### princeton-nlp/SWE-bench Viewer• UpdatedMar 3, 2025 • 21.5k • 44.1k • 143 Note#162 — SWE-bench (Princeton). Can models fix real GitHub issues?
• — #### SWE-bench: Can Language Models Resolve Real-World GitHub Issues? Paper•2310.06770•PublishedOct 10, 2023 • 12 Note#163 — SWE-bench paper (Princeton, 2023). Defined agentic coding evaluation.
• — #### openai/summarize_from_feedback Viewer• UpdatedJan 3, 2023 • 194k • 1.89k • 220 Note#164 — Summarize-from-feedback (OpenAI). Foundational RLHF data.
• — #### Learning to summarize from human feedback Paper•2009.01325•PublishedSep 2, 2020 • 4 Note#165 — Learning to Summarize from Human Feedback (OpenAI, 2020).
• — #### Deep reinforcement learning from human preferences Paper•1706.03741•PublishedJun 12, 2017 • 4 Note#166 — Deep RL from Human Preferences (Christiano et al., OpenAI, 2017). RLHF’s origin.
• — #### Proximal Policy Optimization Algorithms Paper•1707.06347•PublishedJul 20, 2017 • 11 Note#167 — Proximal Policy Optimization (OpenAI, 2017). The algorithm behind RLHF.
• — #### openai/webgpt_comparisons Viewer• UpdatedDec 19, 2022 • 19.6k • 693 • 241 Note#168 — WebGPT comparisons (OpenAI). Early human feedback data.
• — #### Let’s Verify Step by Step Paper•2305.20050•PublishedMay 31, 2023 • 11 Note#169 — Let’s Verify Step by Step (OpenAI, 2023). Process reward models.
• — #### HuggingFaceH4/MATH-500 Viewer• UpdatedDec 15, 2025 • 500 • 136k • 318 Note#170 — MATH-500. The competition math eval, from OpenAI’s split of Hendrycks MATH.
• — #### PaLM: Scaling Language Modeling with Pathways Paper•2204.02311•PublishedApr 5, 2022 • 3 Note#171 — PaLM: Scaling Language Modeling with Pathways (Google, 2022).
• — #### Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity Paper•2101.03961•PublishedJan 11, 2021 • 13 Note#172 — Switch Transformers (Google, 2021). MoE at trillion scale.
• — #### Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer Paper•1701.06538•PublishedJan 23, 2017 • 7 Note#173 — Outrageously Large Neural Networks (Shazeer et al., Google, 2017). The mixture-of-experts layer behind today’s MoEs.
• — #### Emergent Abilities of Large Language Models Paper•2206.07682•PublishedJun 15, 2022 • 3 Note#174 — Emergent Abilities of Large Language Models (Google, 2022).
• — #### Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback Paper•2204.05862•PublishedApr 12, 2022 • 4 Note#175 — Training a Helpful and Harmless Assistant with RLHF (Anthropic, 2022).
• — #### Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks Paper•2005.11401•PublishedMay 22, 2020 • 14 Note#176 — Retrieval-Augmented Generation (Meta, 2020). RAG.
• — #### ReAct: Synergizing Reasoning and Acting in Language Models Paper•2210.03629•PublishedOct 6, 2022 • 35 Note#177 — ReAct: Synergizing Reasoning and Acting (Princeton + Google, 2022).
• — #### Toolformer: Language Models Can Teach Themselves to Use Tools Paper•2302.04761•PublishedFeb 9, 2023 • 12 Note#178 — Toolformer (Meta, 2023). LLMs learn to use tools.
• — #### Generative Agents: Interactive Simulacra of Human Behavior Paper•2304.03442•PublishedApr 7, 2023 • 16 Note#179 — Generative Agents: Interactive Simulacra (Stanford, 2023). The Smallville paper.
• — #### allenai/WildChat-1M Viewer• UpdatedOct 17, 2024 • 838k • 32.1k • 446 Note#180 — WildChat (AI2). Real user-LLM conversations, opened.
• — #### facebook/flores Viewer• UpdatedMay 29 • 1.23M • 8.82k • 107 Note#181 — FLORES (Meta). The multilingual translation benchmark.
• — #### facebook/nllb-200-distilled-600M Translation• UpdatedFeb 14, 2024 • 2M • 929 Note#182 — NLLB (Meta, 2022). No Language Left Behind: 200 languages.
• — #### facebook/seamless-m4t-v2-large Automatic Speech Recognition• 2B• UpdatedJan 4, 2024 • 117k • 991 Note#183 — SeamlessM4T (Meta, 2023). Universal speech translation.
• — #### facebook/mms-1b-all Automatic Speech Recognition• 1.0B• UpdatedJun 15, 2023 • 312k • 202 Note#184 — MMS (Meta, 2023). Speech tech for 1,000+ languages.
• — #### distil-whisper/distil-large-v3 Automatic Speech Recognition• 0.8B• UpdatedApr 21 • 826k • 376 Note#185 — Distil-Whisper (Hugging Face). Whisper, 6x faster.
• — #### nvidia/parakeet-tdt-0.6b-v2 Automatic Speech Recognition• Updated8 days ago • 370k • 1.51k Note#186 — Parakeet (NVIDIA, 2025). Top of the open ASR leaderboard.
• — #### nvidia/canary-1b Automatic Speech Recognition• UpdatedDec 3, 2025 • 2.27k • 457 Note#187 — Canary (NVIDIA). Multilingual ASR and translation.
• — #### MLCommons/peoples_speech Viewer• UpdatedNov 20, 2024 • 8.05M • 37.6k • 272 Note#188 — The People’s Speech (MLCommons). 30K+ hours of open ASR data.
• — Explore speech model performance benchmarks Note#189 — Open ASR Leaderboard. Keeping speech recognition honest.
• — Transcribe audio files into text instantly Note#190 — The Whisper demo Space. Transcription for everyone.
• — #### nvidia/NV-Embed-v2 Feature Extraction• 8B• UpdatedJul 21, 2025 • 24.5k • 513 Note#191 — NV-Embed (NVIDIA). Top open embedding model of its day.
• — #### nomic-ai/nomic-embed-text-v1.5 Sentence Similarity• 0.1B• UpdatedApr 7 • 15.9M • 860 Note#192 — Nomic Embed (NYC). Open weights, open data embeddings.
• — https://huggingface.co/spaces/mteb/leaderboard Note#193 — MTEB Leaderboard. Where embedding models compete.
• — #### answerdotai/ModernBERT-base Fill-Mask• 0.1B• UpdatedJan 15, 2025 • 10M • 1.07k Note#194 — ModernBERT (Answer.AI, 2024). BERT, rebuilt for the 2020s.
• — #### allenai/longformer-base-4096 UpdatedApr 5, 2023 • 1.48M • 229 Note#195 — Longformer (AI2, 2020). Long documents before long context.
• — #### microsoft/biogpt Text Generation• UpdatedFeb 3, 2023 • 124k • 307 Note#196 — BioGPT (Microsoft). Biomedical text generation.
• — #### facebook/esm2_t33_650M_UR50D Fill-Mask• 0.7B• UpdatedMar 21, 2023 • 1.16M • • 82 Note#197 — ESM-2 (Meta). Protein language models, open sourced — Meta’s open answer to AlphaFold.
• — #### facebook/esmfold_v1 UpdatedMar 22, 2023 • 1.58M • 50 Note#198 — ESMFold (Meta, 2022). Open protein structure prediction — Meta’s open answer to AlphaFold.
• — #### allenai/peS2o UpdatedOct 13, 2024 • 11.5k • 198 Note#199 — peS2o (AI2). Open academic papers corpus.
• — #### facebook/galactica-6.7b Text Generation• UpdatedJan 24, 2023 • 657 • 104 Note#200 — Galactica (Meta, 2022). Ahead of its time, in every sense.
• — #### allenai/objaverse UpdatedMar 31, 2023 • 262k • 454 Note#201 — Objaverse (AI2). 800K+ annotated 3D objects.
• — #### openai/shap-e Text-to-3D• UpdatedDec 11, 2023 • 3.26k • 277 Note#202 — Shap-E (OpenAI, 2023). Open text-to-3D.
• — #### NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis Paper•2003.08934•PublishedMar 19, 2020 • 2 Note#203 — NeRF: Neural Radiance Fields (Berkeley + Google, 2020).
• — #### genmo/mochi-1-preview Text-to-Video• UpdatedSep 4, 2025 • 2.86k • • 1.33k Note#204 — Mochi (Genmo, 2024). Open video generation from the Bay.
• — #### playgroundai/playground-v2.5-1024px-aesthetic Text-to-Image• UpdatedMar 15, 2024 • 297k • • 767 Note#205 — Playground v2.5. American open text-to-image.
• — #### fal/AuraFlow Text-to-Image• UpdatedJul 18, 2024 • 311 • • 654 Note#206 — AuraFlow (fal). Open flow-based image generation.
• — Generate images from any text prompt Note#207 — DALL-E mini (Boris Dayma, Texas). The Space that broke the internet in 2022.
• — #### Hierarchical Text-Conditional Image Generation with CLIP Latents Paper•2204.06125•PublishedApr 13, 2022 • 3 Note#208 — DALL-E 2: Hierarchical Text-Conditional Image Generation (OpenAI, 2022).
• — #### Diffusion Models Beat GANs on Image Synthesis Paper•2105.05233•PublishedMay 11, 2021 • 2 Note#209 — Diffusion Models Beat GANs on Image Synthesis (OpenAI, 2021).
• — #### google-research-datasets/conceptual_captions Viewer• UpdatedJun 17, 2024 • 5.34M • 16.3k • 108 Note#210 — Conceptual Captions (Google). Image-text pretraining pioneer.
• — #### poloclub/diffusiondb UpdatedJan 22, 2024 • 11.1k • 632 Note#211 — DiffusionDB (Georgia Tech). 14M images from Stable Diffusion users.
• — #### QR Code AI Art Generator 📱 QR Code AI Art Generator Blend QR codes with AI Art Note#212 — QR Code AI Art. Functional QR codes as art; a 2023 moment.
• — Create your own AI comic with a single prompt Note#213 — AI Comic Factory. One of the most-liked Spaces ever.
• — Open HuggingChat web chat interface Note#214 — HuggingChat. Open source ChatGPT alternative.
• — #### HuggingFaceTB/SmolLM3-3B Text Generation• 3B• UpdatedSep 10, 2025 • 703k • 981 Note#215 — SmolLM3 (Hugging Face, 2025). Fully open small model, trained in the open.
• — #### HuggingFaceTB/cosmopedia Viewer• UpdatedAug 12, 2024 • 31.1M • 16.6k • 723 Note#216 — Cosmopedia (Hugging Face). Largest open synthetic pretraining dataset.
• — #### LiquidAI/LFM2-1.2B Text Generation• 1B• UpdatedFeb 12 • 146k • 355 Note#217 — LFM2 (Liquid AI, MIT spinoff, 2025). Post-transformer architectures, open.
• — #### PrimeIntellect/INTELLECT-1 Text Generation• 10B• UpdatedNov 29, 2024 • 128 • 67 Note#218 — INTELLECT-1 (Prime Intellect, 2024). First decentralized-trained 10B model.
• — #### microsoft/bitnet-b1.58-2B-4T Text Generation• 0.8B• UpdatedDec 17, 2025 • 8.76k • 1.47k Note#219 — BitNet (Microsoft, 2025). 1.58-bit LLMs, for real.
• — #### nvidia/Llama-3_1-Nemotron-Ultra-253B-v1 Text Generation• 253B• UpdatedOct 15, 2025 • 990 • 352 Note#220 — Nemotron Ultra 253B (NVIDIA, 2025). Open reasoning at scale.
• — #### nvidia/GR00T-N1-2B Robotics• 2B• UpdatedSep 2, 2025 • 202 • 356 Note#221 — GR00T N1 (NVIDIA, 2025). Open foundation model for humanoid robots.
• — #### lerobot/smolvla_base Robotics• 0.5B• UpdatedJan 22 • 49.2k • 397 Note#222 — SmolVLA (Hugging Face LeRobot, 2025). Open vision-language-action model for robotics.
• — #### openvla/openvla-7b Robotics• 8B• UpdatedFeb 17 • 1.19M • 237 Note#223 — OpenVLA (Stanford, 2024). Open vision-language-action for robots.
• — #### meta-llama/Llama-3.3-70B-Instruct Text Generation• 71B• UpdatedDec 21, 2024 • 785k • • 2.88k Note#224 — Llama 3.3 70B (Meta). 405B-class quality, 70B price.
• — #### meta-llama/Llama-Guard-3-8B Text Generation• 8B• UpdatedOct 11, 2024 • 242k • • 308 Note#225 — Llama Guard (Meta). Open safety classification.
• — #### google/gemma-3-27b-it Image-Text-to-Text• 27B• UpdatedMar 21, 2025 • 855k • • 1.99k Note#226 — Gemma 3 (Google, 2025). Multimodal, 140+ languages, single GPU.
• — #### Gemma: Open Models Based on Gemini Research and Technology Paper•2403.08295•PublishedMar 13, 2024 • 51 Note#227 — Gemma: Open Models Based on Gemini Research (Google, 2024).
• — #### abisee/cnn_dailymail Viewer• UpdatedJan 18, 2024 • 936k • 261k • 346 Note#228 — CNN/DailyMail (Stanford/Google). The summarization benchmark.
• — #### microsoft/ms_marco Viewer• UpdatedJan 4, 2024 • 1.11M • 22.3k • 244 Note#229 — MS MARCO (Microsoft). The dataset behind neural search.
• — #### google-research-datasets/natural_questions Viewer• UpdatedMar 11, 2024 • 26.3k • 19.5k • 125 Note#230 — Natural Questions (Google). Real queries, real answers.
• — #### Rowan/hellaswag Viewer• UpdatedJul 10, 2025 • 60k • 246k • 183 Note#231 — HellaSwag (UW/AI2). Commonsense that fooled models for years.
• — #### allenai/winogrande Viewer• UpdatedJul 11, 2025 • 81.4k • 206k • 83 Note#232 — WinoGrande (AI2). Winograd schemas at scale.
• — #### allenai/ai2_arc Viewer• UpdatedDec 21, 2023 • 7.79k • 418k • 363 Note#233 — ARC (AI2). The AI2 Reasoning Challenge.
• — #### aps/super_glue Viewer• UpdatedMay 16, 2025 • 196k • 151k • 188 Note#234 — SuperGLUE (NYU/UW). GLUE’s harder sibling.
• — #### truthfulqa/truthful_qa Viewer• UpdatedJan 4, 2024 • 1.63k • 78k • 286 Note#235 — TruthfulQA (OpenAI-affiliated, 2021). Measuring imitative falsehoods.
• — #### google/IFEval Viewer• UpdatedAug 14, 2024 • 541 • 87.9k • 155 Note#236 — IFEval (Google). Instruction-following, verifiable.
• — #### Idavidrein/gpqa Benchmark• UpdatedMar 5 • 1.25k • 94.2k • 475 Note#237 — GPQA (NYU). Google-proof graduate-level science questions.
• — #### cais/hle Benchmark• UpdatedJan 20 • 2.5k • 29.3k • 855 Note#238 — Humanity’s Last Exam (CAIS + Scale AI, 2025). The frontier benchmark.
• — #### gaia-benchmark/GAIA Viewer• UpdatedOct 28, 2025 • 932 • 16.3k • 714 Note#239 — GAIA (Meta + Hugging Face). The agent benchmark.
• — #### Reward Bench Leaderboard 📐 Explore and compare model scores on RewardBench benchmarks Note#240 — RewardBench (AI2). Evaluating reward models.
• — #### Big Code Models Leaderboard 📈 Explore and compare code model performance on a leaderboard Note#241 — BigCode Models Leaderboard. Open code model rankings.
• — #### Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism Paper•1909.08053•PublishedSep 17, 2019 • 5 Note#242 — Megatron-LM (NVIDIA, 2019). Model parallelism playbook.
• — #### ZeRO: Memory Optimizations Toward Training Trillion Parameter Models Paper•1910.02054•PublishedOct 4, 2019 • 11 Note#243 — ZeRO (Microsoft, 2019). DeepSpeed’s memory magic.
• — #### The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale Paper•2406.17557•PublishedJun 25, 2024 • 105 Note#244 — The FineWeb Datasets (Hugging Face, 2024).
• — #### OpenAI o1 System Card Paper•2412.16720•PublishedDec 21, 2024 • 38 Note#245 — OpenAI o1 System Card (2024). Reasoning models arrive.
• — #### allenai/OLMo-7B Text Generation• 7B• UpdatedOct 9, 2025 • 55k • 653 Note#246 — OLMo 1 (AI2, 2024). The first fully open LLM pipeline.
• — #### allenai/olmOCR-7B-0225-preview Image-Text-to-Text• 8B• UpdatedAug 19, 2025 • 2.4k • 708 Note#247 — olmOCR (AI2, 2025). Open document understanding hit.
• — #### microsoft/layoutlmv3-base 0.1B• UpdatedApr 10, 2024 • 883k • 502 Note#248 — LayoutLMv3 (Microsoft). Document AI foundation.
• — Generate any application by Vibe Coding it Note#249 — DeepSite. Vibe coding in a Space; 2025 phenomenon.
• — #### gpt-oss-120b & gpt-oss-20b Model Card Paper•2508.10925•PublishedAug 8, 2025 • 22 Note#250 — gpt-oss-120b & gpt-oss-20b Model Card (OpenAI, 2025).