| [ | |
| "not-for-all-audiences", | |
| "endpoints_compatible", | |
| "transformers", | |
| "pytorch", | |
| "autotrain_compatible", | |
| "text-generation-inference", | |
| "tensorboard", | |
| "safetensors", | |
| "bert", | |
| "generated_from_trainer", | |
| "model-index", | |
| "gpt2", | |
| "gguf", | |
| "tf", | |
| "diffusers", | |
| "roberta", | |
| "wav2vec2", | |
| "mergekit", | |
| "distilbert", | |
| "t5", | |
| "llama", | |
| "huggingtweets", | |
| "lora", | |
| "autotrain", | |
| "imatrix", | |
| "marian", | |
| "deep-reinforcement-learning", | |
| "mistral", | |
| "evaluation", | |
| "stable-diffusion", | |
| "xlm-roberta", | |
| "stable-baselines3", | |
| "text-embeddings-inference", | |
| "bart", | |
| "custom_code", | |
| "onnx", | |
| "sentence-transformers", | |
| "qwen2", | |
| "flux", | |
| "generated_from_keras_callback", | |
| "LunarLander-v2", | |
| "custom-implementation", | |
| "pretraining", | |
| "unsloth", | |
| "electra", | |
| "mt5", | |
| "vision", | |
| "chat", | |
| "speech", | |
| "trl", | |
| "medical", | |
| "q-learning", | |
| "replicate", | |
| "co2_eq_emissions", | |
| "lazymergekit", | |
| "llama-cpp", | |
| "4-bit", | |
| "vit", | |
| "gguf-my-repo", | |
| "hf-asr-leaderboard", | |
| "biology", | |
| "gemma2", | |
| "lyrics", | |
| "huggingartists", | |
| "albert", | |
| "legal", | |
| "sft", | |
| "robust-speech-event", | |
| "stable-diffusion-xl", | |
| "espnet", | |
| "llama-3", | |
| "tf-keras", | |
| "camembert", | |
| "causal-lm", | |
| "meta", | |
| "roleplay", | |
| "pegasus", | |
| "deberta-v2", | |
| "multiberts", | |
| "facebook", | |
| "mbart", | |
| "stable-diffusion-diffusers", | |
| "exbert", | |
| "peft", | |
| "Taxi-v3", | |
| "anime", | |
| "whisper", | |
| "FrozenLake-v1-4x4-no_slippery", | |
| "8-bit", | |
| "gpt_neo", | |
| "xlsr-fine-tuning-week", | |
| "instruct", | |
| "encoder-decoder", | |
| "qwen", | |
| "chemistry", | |
| "music", | |
| "m2m_100", | |
| "GGUF", | |
| "spacy", | |
| "lm-head", | |
| "bitsandbytes", | |
| "nsfw", | |
| "huggingpics", | |
| "autonlp", | |
| "nlp", | |
| "multimodal", | |
| "vision-encoder-decoder", | |
| "uncensored", | |
| "mteb", | |
| "adapter-transformers", | |
| "axolotl", | |
| "deep-rl-class", | |
| "SpaceInvadersNoFrameskip-v4", | |
| "deberta", | |
| "reinforce", | |
| "transformers.js", | |
| "math", | |
| "gptq", | |
| "dpo", | |
| "question generation", | |
| "deep-narrow", | |
| "llm", | |
| "mpnet", | |
| "fashion", | |
| "speech-encoder-decoder", | |
| "llava", | |
| "nemo", | |
| "mixtral", | |
| "openvino", | |
| "clip", | |
| "roleplaying", | |
| "rp", | |
| "ecommerce", | |
| "rust", | |
| "llama-factory", | |
| "chatml", | |
| "exl2", | |
| "qwen-coder", | |
| "abliterated", | |
| "image-generation", | |
| "climate", | |
| "hubert", | |
| "codeqwen", | |
| "timm", | |
| "keras", | |
| "segformer", | |
| "webscraping", | |
| "fashion product", | |
| "fashion image", | |
| "gptj", | |
| "multi-modal", | |
| "multi-task", | |
| "PyTorch", | |
| "xlnet", | |
| "creative", | |
| "reasoning", | |
| "spanish", | |
| "ml-agents", | |
| "awq", | |
| "5-bit", | |
| "unispeech", | |
| "imitation-learning", | |
| "offline-reinforcement-learning", | |
| "3-bit", | |
| "6-bit", | |
| "CartPole-v1", | |
| "gpt4", | |
| "japanese", | |
| "unity-ml-agents", | |
| "NER", | |
| "bloom", | |
| "stable-diffusion-xl-diffusers", | |
| "classification", | |
| "human-feedback", | |
| "diffusers-training", | |
| "model_for_talk", | |
| "opus-mt-tc", | |
| "llama-2", | |
| "distillation", | |
| "phi3", | |
| "seq2seq", | |
| "2-bit", | |
| "TensorBlock", | |
| "longformer", | |
| "pony", | |
| "rf100", | |
| "writing", | |
| "masked-lm", | |
| "swin", | |
| "roberta-base", | |
| "bert_causalm", | |
| "story", | |
| "synthetic data", | |
| "sentiment", | |
| "benchmark", | |
| "LLM", | |
| "alpaca", | |
| "style", | |
| "embeddings", | |
| "speechbrain", | |
| "Transformer", | |
| "fiction", | |
| "bfloat16", | |
| "opt", | |
| "fastai", | |
| "huggan", | |
| "tabular_classification", | |
| "coreml", | |
| "controlnet", | |
| "stanza", | |
| "binary_classification", | |
| "sequence-tagger-model", | |
| "question-generation", | |
| "ML-Agents-Pyramids", | |
| "language", | |
| "storytelling", | |
| "wavlm", | |
| "layoutlmv2", | |
| "fuego", | |
| "qwen2_vl", | |
| "flair", | |
| "dpr", | |
| "CTC", | |
| "dependency-parsing", | |
| "lstm_causalm", | |
| "gpt2_causalm", | |
| "roberta_causalm", | |
| "DFP", | |
| "french prompts", | |
| "science", | |
| "RLHF", | |
| "sequence-to-sequence", | |
| "generic", | |
| "function calling", | |
| "migrated", | |
| "instruction-finetuning", | |
| "ocr", | |
| "internvl_chat", | |
| "full", | |
| "pytorch_model_hub_mixin", | |
| "unispeech-sat", | |
| "creative-writing", | |
| "fluxgym", | |
| "vllm", | |
| "teach-my-agent-parkour", | |
| "resnet", | |
| "model_hub_mixin", | |
| "big_bird", | |
| "creative writing", | |
| "realistic", | |
| "finetuned", | |
| "fiction writing", | |
| "italian", | |
| "NeMo", | |
| "rlfh", | |
| "UCI", | |
| "scene continue", | |
| "NLP", | |
| "horror", | |
| "tflite", | |
| "gpt", | |
| "qa", | |
| "sparsity", | |
| "conditional-text-generation", | |
| "story generation", | |
| "nvidia", | |
| "open_clip", | |
| "fiction story", | |
| "text generation", | |
| "rlhf", | |
| "Transformers", | |
| "rlaif", | |
| "plot generation", | |
| "sub-plot generation", | |
| "chatglm", | |
| "photorealistic", | |
| "plbart", | |
| "florence2", | |
| "tapas", | |
| "AntBulletEnv-v0", | |
| "Pixelcopter-PLE-v0", | |
| "swearing", | |
| "english", | |
| "gpt_neox", | |
| "images", | |
| "physics", | |
| "convnext", | |
| "oBERT", | |
| "internvl", | |
| "vits", | |
| "joblib", | |
| "compression", | |
| "chinese", | |
| "ultra-realistic", | |
| "compressed-tensors", | |
| "alignment-handbook", | |
| "finetuning", | |
| "pruning", | |
| "qwen2.5", | |
| "orpo", | |
| "llava_llama", | |
| "llama-3.1", | |
| "optimum_graphcore", | |
| "mms", | |
| "stable diffusion", | |
| "instruction-following", | |
| "sillytavern", | |
| "layoutlmv3", | |
| "tensorflow", | |
| "storywriting", | |
| "Llama3", | |
| "science fiction", | |
| "yolo", | |
| "korean", | |
| "pixel", | |
| "speech-recognition", | |
| "sv-SE", | |
| "zh-CN", | |
| "sdxl", | |
| "sentiment analysis", | |
| "biomedical", | |
| "python", | |
| "FrozenLake-v1-4x4", | |
| "all genres", | |
| "lm", | |
| "russian", | |
| "concept", | |
| "text classification", | |
| "flux-diffusers", | |
| "dataset", | |
| "granite", | |
| "forecasting", | |
| "roboflow", | |
| "esc", | |
| "romance", | |
| "vivid writing", | |
| "Math", | |
| "falcon3", | |
| "turkish", | |
| "french", | |
| "fairseq", | |
| "Conformer", | |
| "question answering", | |
| "mllama", | |
| "vivid prosing", | |
| "generation", | |
| "clinical", | |
| "persian", | |
| "esb", | |
| "diffusion", | |
| "time-series", | |
| "sklearn", | |
| "English", | |
| "frankenmoe", | |
| "model_stock", | |
| "layoutlm", | |
| "voice", | |
| "whisper-event", | |
| "mobilebert", | |
| "paligemma", | |
| "remote-sensing", | |
| "function-calling", | |
| "int8", | |
| "QA", | |
| "code-generation", | |
| "MountainCar-v0", | |
| "roboflow2huggingface", | |
| "Flux", | |
| "glm", | |
| "computer vision", | |
| "machine-translation", | |
| "mathematics", | |
| "DPO", | |
| "simpletuner", | |
| "gliner", | |
| "sentence_embedding", | |
| "german", | |
| "fine-tuning", | |
| "long-form-generation", | |
| "exaone", | |
| "lg-ai", | |
| "deberta-v3", | |
| "healthcare", | |
| "nft", | |
| "longt5", | |
| "instruction", | |
| "ai-toolkit", | |
| "Stable Diffusion", | |
| "time series", | |
| "convbert", | |
| "farsi", | |
| "security", | |
| "huggingnft", | |
| "safety", | |
| "long context", | |
| "Quality Estimation", | |
| "detr", | |
| "stable-diffusion-api", | |
| "phi", | |
| "role-play", | |
| "beit", | |
| "regnet", | |
| "named entity recognition", | |
| "retrieval", | |
| "Llama-3", | |
| "ctranslate2", | |
| "foundation models", | |
| "fasttext", | |
| "computer-vision", | |
| "politics", | |
| "segmentation", | |
| "lycoris", | |
| "health", | |
| "woman", | |
| "cohere", | |
| "ultralytics", | |
| "style-transfer", | |
| "pretrained", | |
| "speecht5", | |
| "cybersecurity", | |
| "RAG", | |
| "asteroid", | |
| "Instruct", | |
| "internlm2", | |
| "multiclass_classification", | |
| "thudm", | |
| "exaone-3.5", | |
| "json mode", | |
| "pretrained models", | |
| "summary", | |
| "girls", | |
| "finnish", | |
| "large language model", | |
| "reward model", | |
| "Text Classification", | |
| "Long Context", | |
| "celebrity", | |
| "voxpopuli-v2", | |
| "french law", | |
| "droit fran\u00e7ais", | |
| "modernbert", | |
| "Llama", | |
| "BERT", | |
| "portuguese-speech-corpus", | |
| "luke", | |
| "multiberts-seed_0", | |
| "multiberts-seed_1", | |
| "multiberts-seed_2", | |
| "multiberts-seed_3", | |
| "multiberts-seed_4", | |
| "Abstractive Summarization", | |
| "CarRacing-v0", | |
| "fp8", | |
| "instruction-tuning", | |
| "time series foundation models", | |
| "multiberts-seed-0", | |
| "speech-translation", | |
| "chatbot", | |
| "codegen", | |
| "NSFW", | |
| "multiberts-seed-1", | |
| "multiberts-seed-2", | |
| "multiberts-seed-3", | |
| "multiberts-seed-4", | |
| "siglip", | |
| "LoRA", | |
| "3D", | |
| "reddit", | |
| "Athene", | |
| "Attention", | |
| "SFT", | |
| "pruna-ai", | |
| "splinter", | |
| "gary109/AI_Light_Dance", | |
| "granite-3.1", | |
| "CoT", | |
| "yolov8", | |
| "android", | |
| "granite-3.0", | |
| "Nexusflow", | |
| "Qwen2.5", | |
| "speech_to_text", | |
| "language-model", | |
| "pt-br", | |
| "Transducer", | |
| "artistic", | |
| "historical", | |
| "llamafile", | |
| "Mistral", | |
| "torchdistill", | |
| "earth-observation", | |
| "croissant", | |
| "Chinese", | |
| "esm", | |
| "education", | |
| "ChatGPT", | |
| "reflection", | |
| "zle", | |
| "FrozenLake-v1-8x8", | |
| "Pong-PLE-v0", | |
| "any-to-any", | |
| "ollama", | |
| "Ollama", | |
| "NLU", | |
| "zero-shot", | |
| "transformer", | |
| "paraphrasing", | |
| "fnet", | |
| "Arabic", | |
| "agriculture", | |
| "catalan", | |
| "information retrieval", | |
| "alignment", | |
| "preference", | |
| "economics", | |
| "optimum_habana", | |
| "game", | |
| "12b", | |
| "funnel", | |
| "whispering", | |
| "infinite-dataset-hub", | |
| "jw", | |
| "deepseek_v2", | |
| "speaker-diarization", | |
| "olmo", | |
| "protein", | |
| "xglm", | |
| "punctuation", | |
| "rvc", | |
| "linktransformer", | |
| "fantasy", | |
| "tweets", | |
| "newspaper", | |
| "conversation", | |
| "Visual novel", | |
| "BipedalWalker-v3", | |
| "FrozenLake-v1-8x8-no_slippery", | |
| "7b", | |
| "blenderbot", | |
| "pyannote-audio", | |
| "pyannote", | |
| "megatron-bert", | |
| "gpt_bigcode", | |
| "molecules", | |
| "research", | |
| "hybrid-clip", | |
| "sil-ai/bloom-speech", | |
| "danish", | |
| "information-retrieval", | |
| "food", | |
| "engineering", | |
| "ctrl", | |
| "Intel\u00ae Neural Compressor", | |
| "Text-to-Image", | |
| "super-resolution", | |
| "classical chinese", | |
| "literary chinese", | |
| "ancient chinese", | |
| "emotion-classification", | |
| "canine", | |
| "tagalog", | |
| "filipino", | |
| "abstractive summarization", | |
| "portuguese", | |
| "text-to-sql", | |
| "architecture", | |
| "solar", | |
| "torch", | |
| "Not-for-all-Audiences", | |
| "Reasoning", | |
| "ultralyticsplus", | |
| "llama.cpp", | |
| "trocr", | |
| "cross-encoder", | |
| "passage-retrieval", | |
| "insurance", | |
| "vietnamese", | |
| "Anime", | |
| "crypto", | |
| "fine-tuned", | |
| "toxicity", | |
| "fsmt", | |
| "htr", | |
| "ASR", | |
| "baichuan", | |
| "cute", | |
| "machine learning", | |
| "roformer", | |
| "Roleplay", | |
| "history", | |
| "national library of spain", | |
| "satellite-imagery", | |
| "structure-prediction", | |
| "data-juicer", | |
| "Pytorch", | |
| "speaker", | |
| "Sentence Transformers", | |
| "OCR", | |
| "VLM", | |
| "embedding", | |
| "character", | |
| "norwegian", | |
| "paraphrase", | |
| "MSA", | |
| "language model", | |
| "xlm", | |
| "grammar", | |
| "reranking", | |
| "endpoints-template", | |
| "dreambooth", | |
| "literature", | |
| "sdxl-sliders", | |
| "ntcai.xyz-sliders", | |
| "FelixChao/WestSeverus-7B-DPO-v2", | |
| "prompts", | |
| "sample-factory", | |
| "bioinformatics", | |
| "Mr Porter", | |
| "Diffusers", | |
| "Code", | |
| "financial", | |
| "tool-use", | |
| "v8", | |
| "stablediffusionapi.com", | |
| "realism", | |
| "Chat Model", | |
| "perceiver", | |
| "RoBERTa", | |
| "reviews", | |
| "zhs", | |
| "knowledge-distillation", | |
| "paddlepaddle", | |
| "deutsch", | |
| "hate speech", | |
| "business", | |
| "monotransquest", | |
| "sew-d", | |
| "corenlp", | |
| "speechmix", | |
| "data-to-text", | |
| "phi3_v", | |
| "vidore", | |
| "falcon", | |
| "keypoint-detection", | |
| "Safetensors", | |
| "Text Generation", | |
| "8b", | |
| "relation-extraction", | |
| "indicnlp", | |
| "mT5", | |
| "image-super-resolution", | |
| "speech-to-speech-translation", | |
| "gpt3", | |
| "genomics", | |
| "decision_transformer", | |
| "swinv2", | |
| "xclip", | |
| "aiart", | |
| "handwritten", | |
| "arabic", | |
| "ggml", | |
| "styles", | |
| "text adventure", | |
| "web", | |
| "Biology", | |
| "glove", | |
| "experimental", | |
| "hate-speech", | |
| "ConvTasNet", | |
| "fnet-bert-base-comparison", | |
| "ilgiornale", | |
| "repubblica", | |
| "psychology", | |
| "Net", | |
| "parler_tts", | |
| "annotation", | |
| "video-generation", | |
| "colpali", | |
| "dino", | |
| "llava_next", | |
| "maskformer", | |
| "sagemaker", | |
| "tensorflowtts", | |
| "natural-language-understanding", | |
| "LLaMA", | |
| "astronomy", | |
| "atr", | |
| "stablelm", | |
| "poetry", | |
| "4bit", | |
| "vlm", | |
| "Art", | |
| "agent", | |
| "vqa", | |
| "fse", | |
| "Walker2DBulletEnv-v0", | |
| "text-mining", | |
| "3b", | |
| "financial-sentiment-analysis", | |
| "minicpmv", | |
| "relation extraction", | |
| "bitnet", | |
| "social media", | |
| "yolos", | |
| "dinov2", | |
| "donut", | |
| "mamba", | |
| "photo", | |
| "chocolatine", | |
| "books", | |
| "Spanish", | |
| "protein language model", | |
| "super-image", | |
| "data2vec-audio", | |
| "cyber security", | |
| "deberta-mnli", | |
| "ernie", | |
| "text-embedding", | |
| "midjourney", | |
| "ai", | |
| "7B", | |
| "bardsai/jaskier-7b-dpo-v5.6", | |
| "skops", | |
| "open-source", | |
| "token classification", | |
| "green", | |
| "valiant", | |
| "valiant-labs", | |
| "illustrious", | |
| "coding", | |
| "sd3.5-large", | |
| "stance-detection", | |
| "thai", | |
| "gensim", | |
| "SEAD", | |
| "trajectory_transformer", | |
| "word-segmentation", | |
| "diffusion-single-file", | |
| "mixture of experts", | |
| "legal liability", | |
| "convAI", | |
| "vision-language", | |
| "rwkv", | |
| "Multimodal", | |
| "depth", | |
| "modelslab.com", | |
| "sd3", | |
| "Dutch", | |
| "Sentence Similarity", | |
| "document-expansion", | |
| "uncased", | |
| "Riva", | |
| "chatgpt", | |
| "Composer", | |
| "MosaicML", | |
| "llm-foundry", | |
| "ai-safety", | |
| "qlora", | |
| "Decoder", | |
| "AI", | |
| "deepseek", | |
| "environment", | |
| "imagepipeline", | |
| "imagepipeline.io", | |
| "SQL", | |
| "base", | |
| "Logic", | |
| "technical", | |
| "chess", | |
| "Chemistry", | |
| "satellite", | |
| "next-sentence-prediction", | |
| "allennlp", | |
| "PostTrainingStatic", | |
| "dk", | |
| "HalfCheetahBulletEnv-v0", | |
| "medicine", | |
| "synthetic-dataset", | |
| "Physics", | |
| "lrgb", | |
| "ControlNet", | |
| "e-commerce", | |
| "stocks", | |
| "information extraction", | |
| "encoder", | |
| "T5", | |
| "nomic_bert", | |
| "Summarization", | |
| "LLMs", | |
| "drug-discovery", | |
| "CLIP", | |
| "summarisation", | |
| "semantic role labeling", | |
| "tapex", | |
| "NLI", | |
| "Russian", | |
| "audio-source-separation", | |
| "generative", | |
| "scifi", | |
| "RLAIF", | |
| "pathology", | |
| "Python", | |
| "chain-of-thought", | |
| "programming", | |
| "llama-3.2", | |
| "llama-3-instruct", | |
| "computer-science", | |
| "gguf-my-lora", | |
| "Arknights", | |
| "\u660e\u65e5\u65b9\u821f", | |
| "timelms", | |
| "resnetd", | |
| "neural-compressor", | |
| "paraphrase-identification", | |
| "wiki", | |
| "sentinel-2", | |
| "image-text-dataset", | |
| "training", | |
| "musicgen", | |
| "background-removal", | |
| "roa", | |
| "custom-research", | |
| "8B", | |
| "logic", | |
| "argumentation", | |
| "nllb", | |
| "movie", | |
| "Turkish", | |
| "2D", | |
| "Language", | |
| "TDNN", | |
| "people", | |
| "SmallMoleculeMultiView", | |
| "binding-affinity-prediction", | |
| "bio-medical", | |
| "drug-target-interaction", | |
| "molecular-property-prediction", | |
| "moleculenet", | |
| "multi-view", | |
| "small-molecules", | |
| "virtual-screening", | |
| "text2sql", | |
| "search", | |
| "Arabic Text Summarization", | |
| "indobenchmark", | |
| "object detection", | |
| "question", | |
| "booksum", | |
| "long-document", | |
| "SepFormer", | |
| "squeezebert", | |
| "bag-of-words", | |
| "TTS", | |
| "mobilevit", | |
| "Question Answering", | |
| "beir", | |
| "furry", | |
| "privacy", | |
| "FastConformer", | |
| "sail-rvc", | |
| "unity-sentis", | |
| "preferences", | |
| "int4", | |
| "UnfilteredAI", | |
| "sd3-diffusers", | |
| "spectrum", | |
| "recommendation", | |
| "youtube", | |
| "compsci", | |
| "Llama-cpp", | |
| "MountainCarContinuous-v0", | |
| "3B", | |
| "knowledge-graph", | |
| "k2", | |
| "tf2.0", | |
| "icelandic", | |
| "codemix", | |
| "ukrainian", | |
| "text-classfication", | |
| "ML-Agents-Worm", | |
| "pythae", | |
| "reproducibility", | |
| "fine-tune", | |
| "Dataset", | |
| "Culture", | |
| "LeRobot", | |
| "blip-2", | |
| "idefics3", | |
| "inc", | |
| "general-purpose", | |
| "cogvideox", | |
| "cancer", | |
| "fluently-lm", | |
| "thinking", | |
| "Identification", | |
| "commercial use", | |
| "code generation", | |
| "flux-dev", | |
| "HelpingAI", | |
| "Code Generation", | |
| "Czech", | |
| "bigbird_pegasus", | |
| "detection", | |
| "dcgan", | |
| "decision-transformer", | |
| "gym-continous-control", | |
| "CaText", | |
| "Catalan Textual Corpus", | |
| "videomae", | |
| "Multilingual", | |
| "text-clustering", | |
| "text-semantic-similarity", | |
| "text-evaluation", | |
| "text-reranking", | |
| "awesome-yolov8-models", | |
| "language-identification", | |
| "controlnet-v1-1", | |
| "photography", | |
| "self-instruct", | |
| "speech-to-text", | |
| "animals", | |
| "jais", | |
| "Information Extraction", | |
| "philosophy", | |
| "illustration", | |
| "mlc-llm", | |
| "nemotron", | |
| "2.5D", | |
| "extreme swearing", | |
| "mistral nemo", | |
| "design", | |
| "shining-valiant", | |
| "AstralFusion", | |
| "TIES", | |
| "Pendulum-v1", | |
| "indie", | |
| "xlm-roberta-large", | |
| "formality-style-transfer", | |
| "textual-entailment", | |
| "movies", | |
| "visual_bert", | |
| "lexical normalization", | |
| "email", | |
| "bygpt5", | |
| "datacraft", | |
| "medium", | |
| "language modelling", | |
| "multi-modal-qa", | |
| "math-qa", | |
| "language-agent", | |
| "long-context", | |
| "tokenization", | |
| "zh-tw", | |
| "overlapped-speech-detection", | |
| "arctic", | |
| "speaker-segmentation", | |
| "Vision", | |
| "Text", | |
| "ColBERT", | |
| "prompt", | |
| "idefics2", | |
| "ONNX", | |
| "sapiens", | |
| "MLLM", | |
| "aimv2", | |
| "big_vision", | |
| "Llama-Cpp", | |
| "MIDI", | |
| "vilt", | |
| "bert-base", | |
| "reformer", | |
| "speech-emotion-recognition", | |
| "deberta-v1", | |
| "Flemish", | |
| "RobBERT", | |
| "long-form", | |
| "text-to-mel", | |
| "semantic-search", | |
| "keyphrase-extraction", | |
| "spelling", | |
| "setfit", | |
| "paraphrase-generation", | |
| "aiartchan", | |
| "LLaVA", | |
| "internlm", | |
| "astrophysics", | |
| "h2o-llmstudio", | |
| "blockchain", | |
| "xl", | |
| "IE", | |
| "jailbreak", | |
| "Portuguese", | |
| "Q&A", | |
| "langchain", | |
| "scenery", | |
| "city", | |
| "phishing", | |
| "clip_vision_model", | |
| "web-llm", | |
| "8bit", | |
| "WizardLM/WizardMath-7B-V1.1", | |
| "Mixtral", | |
| "fusechat", | |
| "binary", | |
| "bggpt", | |
| "insait", | |
| "efficient", | |
| "chameleon", | |
| "16-bit", | |
| "quantization", | |
| "lifescience", | |
| "social", | |
| "bangla", | |
| "Embodied AI", | |
| "shining-valiant-2", | |
| "ANDSystem", | |
| "bengali", | |
| "sparsh", | |
| "DarkStock", | |
| "Aspire", | |
| "Storm", | |
| "DarkEnigma", | |
| "vicgalle/Configurable-Hermes-2-Pro-Llama-3-8B", | |
| "conversational-ai", | |
| "graphic horror", | |
| "novel", | |
| "transfo-xl", | |
| "nystromformer", | |
| "answer extraction", | |
| "commonsense-reasoning", | |
| "newspapers", | |
| "GPT-4", | |
| "chart", | |
| "common crawl", | |
| "Gucci", | |
| "gguf-comfy", | |
| "pyannote-audio-pipeline", | |
| "ud", | |
| "mbart-50", | |
| "pyannote-audio-model", | |
| "Chat", | |
| "entity recognition", | |
| "Sentiment", | |
| "Architecture", | |
| "DML", | |
| "ONNXRuntime", | |
| "SDXL", | |
| "ethics", | |
| "Agent", | |
| "rubert", | |
| "multiclass", | |
| "tiny", | |
| "xls_r", | |
| "keytotext", | |
| "indobert", | |
| "Source Separation", | |
| "Speech Separation", | |
| "sequence-classification", | |
| "baseline-trainer", | |
| "text2image", | |
| "mask2former", | |
| "yolov5", | |
| "Named Entity Recognition", | |
| "DNA", | |
| "sam", | |
| "deep-learning", | |
| "CV", | |
| "moderation", | |
| "midi", | |
| "tax", | |
| "Event Extraction", | |
| "Relation Extraction", | |
| "wav2vec2-bert", | |
| "model-fusion", | |
| "automerger", | |
| "hindi", | |
| "materials science", | |
| "enhanced", | |
| "retrieval-augmented-generation", | |
| "SMILES", | |
| "regression", | |
| "medical imaging", | |
| "jais-family", | |
| "RP", | |
| "book", | |
| "o1", | |
| "sexy", | |
| "COCO", | |
| "mlx-my-repo", | |
| "swedish", | |
| "COT", | |
| "adaption", | |
| "recycled", | |
| "DA", | |
| "commonsense", | |
| "icefall", | |
| "data2vec-text", | |
| "headline-generation", | |
| "codeswitching", | |
| "natural language understanding", | |
| "social-media", | |
| "wav2vec2-conformer", | |
| "fact checking", | |
| "QbertNoFrameskip-v4", | |
| "ddpm_diffusion", | |
| "tweet", | |
| "ANDDigest", | |
| "conversations-summarization", | |
| "verification", | |
| "PDF", | |
| "LM", | |
| "hate speech detection", | |
| "wikidata", | |
| "bio-chem", | |
| "molnet", | |
| "molecule-net", | |
| "biophysics", | |
| "godot-rl", | |
| "environments", | |
| "video-games", | |
| "regularization-images", | |
| "class-instance", | |
| "preservation-loss-training", | |
| "protein structure", | |
| "spacerunner", | |
| "blip", | |
| "moondream1", | |
| "multi_modality", | |
| "Taiwan", | |
| "pdf", | |
| "speaker-change-detection", | |
| "idol", | |
| "inpainting", | |
| "Inference Endpoints", | |
| "Medical", | |
| "MiniCPM", | |
| "vla", | |
| "reranker", | |
| "Photography", | |
| "splade", | |
| "SPO", | |
| "model2vec", | |
| "chatqa", | |
| "QwQ", | |
| "sana", | |
| "Sana", | |
| "GPT-2", | |
| "music-generation", | |
| "xlm-roberta-xl", | |
| "Keywords to Sentences", | |
| "aspect-based-sentiment-analysis", | |
| "radiology", | |
| "GAN", | |
| "image-text", | |
| "mobilenet_v2", | |
| "dreambooth-hackathon", | |
| "painting", | |
| "pix2struct", | |
| "face", | |
| "labse", | |
| "histology", | |
| "flan", | |
| "Named Entity Recogniton", | |
| "mistral-7b", | |
| "lmm", | |
| "Alpaca", | |
| "gpt2-medium", | |
| "orca", | |
| "buildings", | |
| "cartoon", | |
| "augmentation", | |
| "espa\u00f1ol", | |
| "cyber", | |
| "CultriX/Wernicke-7B-v9", | |
| "Audio", | |
| "VQA", | |
| "liminerity/merge", | |
| "gemma_torch", | |
| "keras-hub", | |
| "on-device language model", | |
| "trained", | |
| "instructions", | |
| "materials", | |
| "openelm", | |
| "lab", | |
| "heathcare", | |
| "Pharmaceutical", | |
| "Pharma", | |
| "llama-3-ko", | |
| "Science", | |
| "parallel", | |
| "multi-turn", | |
| "ov", | |
| "LLM Agent", | |
| "gptqmodel", | |
| "modelcloud", | |
| "transliteration", | |
| "autoquant", | |
| "image classification", | |
| "llava_onevision", | |
| "auto-gptq", | |
| "safe-for-work", | |
| "moshi", | |
| "128k context", | |
| "theprint", | |
| "multiturn", | |
| "llama-3.1-instruct", | |
| "tactile", | |
| "technical-assistance", | |
| "structured-output", | |
| "Llama3.2", | |
| "natural-language-processing", | |
| "vision-text-dual-encoder", | |
| "lxmert", | |
| "realm", | |
| "machine translation", | |
| "prophetnet", | |
| "cnlpt", | |
| "pos-tagging", | |
| "kd", | |
| "sparql", | |
| "BipedalWalkerHardcore-v3", | |
| "BreakoutNoFrameskip-v4", | |
| "BeamRiderNoFrameskip-v4", | |
| "dutch", | |
| "math-word-problems", | |
| "grounding", | |
| "safe", | |
| "Benchmark", | |
| "example", | |
| "touch rugby", | |
| "self-supervised-pretraining", | |
| "dual_ar", | |
| "Flux.1-Dev", | |
| "background", | |
| "sound language model", | |
| "chain_of_thought", | |
| "stock market", | |
| "ROC", | |
| "multi-image", | |
| "molmo", | |
| "immich", | |
| "hunyuan", | |
| "machine tranlsation", | |
| "O1-like model", | |
| "saelens", | |
| "Realism", | |
| "bitcoin", | |
| "pythia", | |
| "starcoder2", | |
| "open", | |
| "stablelm_epoch", | |
| "granitemoe", | |
| "tulu", | |
| "Italian", | |
| "gender", | |
| "toxic comments classification", | |
| "Arabic T5", | |
| "k2t", | |
| "GPT", | |
| "spam", | |
| "dpt", | |
| "punctuation prediction", | |
| "pre-training", | |
| "pysentimiento", | |
| "speech-enhancement", | |
| "Audio Source Separation", | |
| "ranking", | |
| "tc", | |
| "query-expansion", | |
| "Finance", | |
| "named-entity-linking", | |
| "Citrinet", | |
| "retrieval augmented generation", | |
| "core-ml", | |
| "Tensorflow", | |
| "Vietnamese", | |
| "analysis", | |
| "quant", | |
| "vicuna", | |
| "dialogue-summarization", | |
| "Cybersecurity", | |
| "t5-small", | |
| "shot", | |
| "zxx", | |
| "prk", | |
| "tag2", | |
| "slerp", | |
| "q4_k_m", | |
| "q8_0", | |
| "bias", | |
| "remote sensing", | |
| "openchat", | |
| "deep learning", | |
| "t5x", | |
| "computer science", | |
| "review", | |
| "seedbox", | |
| "Maths", | |
| "natural language processing", | |
| "remyx", | |
| "farming", | |
| "Equall/Saul-Base", | |
| "brazil", | |
| "knowledge", | |
| "instruction tuning", | |
| "AWQ", | |
| "Enterprise LLM", | |
| "Enterprise", | |
| "Enterprise ready", | |
| "multimodal large language model", | |
| "Qwen2", | |
| "machine-learning", | |
| "Llama3.1", | |
| "generative-ai", | |
| "monet", | |
| "fast-apply", | |
| "instant-apply", | |
| "TIES_merge", | |
| "TheSpice", | |
| "Yggdrasil", | |
| "Bluuwhale", | |
| "magical-realism", | |
| "sd3.5", | |
| "sd3.5-diffusers", | |
| "decisions", | |
| "TLA", | |
| "pegasus_x", | |
| "F16", | |
| "sailor", | |
| "Translation", | |
| "casual-lm", | |
| "deit", | |
| "microtransquest", | |
| "hter", | |
| "siamesetransquest", | |
| "diarization", | |
| "allenai", | |
| "testing", | |
| "xlm-prophetnet", | |
| "entailment", | |
| "qarib", | |
| "minds14", | |
| "projected-gan", | |
| "rita", | |
| "coptic", | |
| "opt_metasq", | |
| "HopperBulletEnv-v0", | |
| "LunarLanderContinuous-v2", | |
| "ML-Agents-PushBlock", | |
| "spectrograms", | |
| "mlconsole", | |
| "qa-nli", | |
| "games", | |
| "sports", | |
| "table", | |
| "government", | |
| "figure-qa", | |
| "geometry-diagram", | |
| "scientific-figure", | |
| "Imitation Learning", | |
| "Farfetch", | |
| "corpus", | |
| "amazon", | |
| "synthetic-captions", | |
| "planning", | |
| "reward", | |
| "lichess", | |
| "dialog-response-generation", | |
| "github-stars", | |
| "Prada", | |
| "deepseek_v3", | |
| "coqui", | |
| "video-to-video", | |
| "f5-tts", | |
| "llama_for_causal_lm", | |
| "GLiNER", | |
| "brainstorm 40x", | |
| "ECAPA-TDNN", | |
| "Interior", | |
| "Food", | |
| "llava_mistral", | |
| "falcon_mamba", | |
| "sql", | |
| "Function Calling", | |
| "Extraction", | |
| "intern_vit_6b", | |
| "RVC", | |
| "Arabic Dialect", | |
| "Arabic Machine Translation", | |
| "Arabic News Title and Question Generation", | |
| "Arabic Paraphrasing and Transliteration", | |
| "Arabic Code-Switched Translation", | |
| "vit_mae", | |
| "xls_r_translation", | |
| "distilroberta", | |
| "election2020", | |
| "speaker-recognition", | |
| "speaker-verification", | |
| "emotion-recognition", | |
| "gmq", | |
| "speech-synthesis", | |
| "HalfCheetah-v3", | |
| "entity-linking", | |
| "error-correction", | |
| "CNN", | |
| "span-marker", | |
| "relation classification", | |
| "txtai", | |
| "intent", | |
| "building", | |
| "landscape", | |
| "diffusion-models-class", | |
| "checkpoint", | |
| "Nijijourney", | |
| "text2vec", | |
| "dreamshaper", | |
| "stablediffusion", | |
| "keywords", | |
| "openai", | |
| "simplification", | |
| "foundation model", | |
| "histopathology", | |
| "minecraft", | |
| "Network Intrusion Detection", | |
| "GPT2", | |
| "absa", | |
| "media", | |
| "retail", | |
| "solidity", | |
| "transcription", | |
| "mel", | |
| "tks", | |
| "tag1", | |
| "Sentiment Analysis", | |
| "q2_k", | |
| "q3_k_m", | |
| "q5_k_m", | |
| "q6_k", | |
| "text-to-music", | |
| "female", | |
| "UNA", | |
| "generated_from_setfit_trainer", | |
| "scientific AI", | |
| "polish", | |
| "hacking", | |
| "topic", | |
| "depth_anything", | |
| "ModelBest", | |
| "THUNLP", | |
| "yolov10", | |
| "tracking", | |
| "function", | |
| "gaming", | |
| "VILA", | |
| "decompile", | |
| "CorticalStack/pastiche-crown-clown-7b-dare-dpo", | |
| "leaderboard", | |
| "Law", | |
| "labrador", | |
| "VoxCeleb", | |
| "static", | |
| "zhtw", | |
| "16bit", | |
| "llava_qwen", | |
| "java", | |
| "autoawq", | |
| "olmoe", | |
| "nncf", | |
| "Computer Vision", | |
| "llmware-chat", | |
| "cosmology", | |
| "droit", | |
| "role play", | |
| "mobile", | |
| "llama-3.1-instruct-8b", | |
| "llama-3-instruct-8b", | |
| "paper", | |
| "actress", | |
| "Video", | |
| "mobilellm", | |
| "Q4", | |
| "Q5", | |
| "Q8", | |
| "AutoRound", | |
| "indonesian", | |
| "maths", | |
| "synthetic-data", | |
| "Earth Observation", | |
| "accounting", | |
| "stock", | |
| "llama-ti", | |
| "fluently-merge", | |
| "Mathematics", | |
| "NuSLERP", | |
| "India", | |
| "hentai", | |
| "zlw", | |
| "metric", | |
| "audacity", | |
| "pretrain", | |
| "miniLM", | |
| "spoken language understanding", | |
| "cvt", | |
| "keyphrase-generation", | |
| "multi-displinary", | |
| "levit", | |
| "Walker2d-v3", | |
| "Swimmer-v3", | |
| "Hopper-v3", | |
| "ReacherBulletEnv-v0", | |
| "Ant-v3", | |
| "MedicalNet", | |
| "medical images", | |
| "Med3D", | |
| "clustering", | |
| "email generation", | |
| "tau/sled", | |
| "afro-digits-speech", | |
| "geography", | |
| "research papers", | |
| "explanation", | |
| "graphic design", | |
| "linguistics", | |
| "instruct-tune", | |
| "fairness", | |
| "visual-reasoning", | |
| "energy", | |
| "NeRF", | |
| "3D Vision", | |
| "GPT-4V", | |
| "Image", | |
| "biology`", | |
| "Burberry", | |
| "video generation", | |
| "Deepsync", | |
| "image generation", | |
| "candlesticks", | |
| "option trading", | |
| "captioning", | |
| "minicpm-v", | |
| "snowflake-arctic-embed", | |
| "animatediff", | |
| "sam2", | |
| "Design", | |
| "aria", | |
| "emotions", | |
| "ultravox", | |
| "Midjourney", | |
| "fluently-sets", | |
| "Cinematic", | |
| "Landscape", | |
| "Car", | |
| "Wildlife", | |
| "jamba", | |
| "audiocraft", | |
| "static-embeddings", | |
| "vidore-experimental", | |
| "gte", | |
| "mesh-generation", | |
| "chatqa-1.5", | |
| "Llama3 MOE", | |
| "autogenerated-modelcard", | |
| "historic", | |
| "BERTje", | |
| "go-emotion", | |
| "negation", | |
| "flaubert-base", | |
| "fanpage", | |
| "ilpost", | |
| "anomaly detection", | |
| "audio-frame-classification", | |
| "pokemon", | |
| "deformable_detr", | |
| "byt5", | |
| "WHAM!", | |
| "enformer", | |
| "Arabic News Title Generation", | |
| "Arabic Paraphrasing", | |
| "named-entity-disambiguation", | |
| "entity-disambiguation", | |
| "markuplm", | |
| "Token Classification", | |
| "grammar synthesis", | |
| "invoices", | |
| "conditional_detr", | |
| "timesformer", | |
| "cyberpunk", | |
| "urdu", | |
| "Automatic Speech Recognition", | |
| "colbert", | |
| "hate", | |
| "offensive language", | |
| "efficientnet", | |
| "semantic search", | |
| "text-image", | |
| "sentence-embedding", | |
| "g2p", | |
| "characters", | |
| "kandinsky", | |
| "bertopic", | |
| "KoAlpaca", | |
| "safe-rlhf", | |
| "girl", | |
| "Classification", | |
| "natural language generation", | |
| "japanese-stablelm", | |
| "mgpt", | |
| "monai", | |
| "foundation", | |
| "infosec", | |
| "cybersec", | |
| "Instruction", | |
| "optimum", | |
| "prompt-injection", | |
| "dolphin", | |
| "stripedhyena", | |
| "knowledge graph", | |
| "phi-msft", | |
| "information-extraction", | |
| "phi-2", | |
| "Yi", | |
| "mlabonne/Marcoro14-7B-slerp", | |
| "Image Classification", | |
| "MoE", | |
| "decruz07/kellemar-DPO-Orca-Distilled-7B-SLERP", | |
| "vulnerability", | |
| "GPT-SoVITS", | |
| "controlnet model", | |
| "turbo", | |
| "reading comprehension", | |
| "human feedback", | |
| "aqlm", | |
| "fluetnly-xl", | |
| "fluently", | |
| "AIGC", | |
| "portugues", | |
| "Rust", | |
| "1bit", | |
| "labradorite", | |
| "interpretability", | |
| "bilingual", | |
| "biological materials", | |
| "bioinspiration", | |
| "steerlm", | |
| "sarcasm", | |
| "marketing", | |
| "imat", | |
| "GLiClass", | |
| "small language models", | |
| "large video-language model", | |
| "relative depth", | |
| "6bit", | |
| "5bit", | |
| "3bit", | |
| "2bit", | |
| "HelpSteer2", | |
| "graph", | |
| "p1", | |
| "zamba2", | |
| "StableDiffusion", | |
| "legalkit", | |
| "Machine Learning", | |
| "manga", | |
| "ML-Agents-Huggy", | |
| "asian", | |
| "tools", | |
| "pixtral", | |
| "llama 3.1", | |
| "IBM", | |
| "caption", | |
| "code-instruct", | |
| "Segmentation", | |
| "gutenberg", | |
| "Psychology", | |
| "rewiz", | |
| "doge", | |
| "Coder", | |
| "Small", | |
| "edge", | |
| "mochi-1-preview", | |
| "bamba", | |
| "Conversational", | |
| "metamotivo", | |
| "novel view synthesis", | |
| "Autonomous Driving", | |
| "textnet", | |
| "porn", | |
| "anthro", | |
| "OpenCENIA", | |
| "irish", | |
| "Fill-Mask", | |
| "gpt2-small", | |
| "zls", | |
| "deberta-v2`", | |
| "rembert", | |
| "mrc", | |
| "authorship-identification", | |
| "fire2020", | |
| "pan2020", | |
| "audio-xvector", | |
| "dependency parsing", | |
| "normalization", | |
| "speech recognition", | |
| "basque", | |
| "sibyl", | |
| "common-voice", | |
| "TapasModel", | |
| "text2text-question-answering", | |
| "twitch", | |
| "Electra", | |
| "bertin", | |
| "masked_bert", | |
| "answer-extraction", | |
| "mel-to-wav", | |
| "News", | |
| "yoso", | |
| "speech_to_text_transformer", | |
| "malaysia", | |
| "SLU", | |
| "understanding", | |
| "AMRBART", | |
| "mctct", | |
| "FrozenLake-v1", | |
| "mbart-cc25", | |
| "seals/Walker2d-v0", | |
| "Acrobot-v1", | |
| "object-tracking", | |
| "MaCoCu", | |
| "doe2vec", | |
| "exploratory-landscape-analysis", | |
| "autoencoders", | |
| "test", | |
| "long-documents", | |
| "coreference-nli", | |
| "ratings", | |
| "abstractive-summarization", | |
| "ML", | |
| "3D shapes", | |
| "prompt engineering", | |
| "Manufacturing", | |
| "logical reasoning", | |
| "OOD", | |
| "language modeling", | |
| "gpt-3.5", | |
| "QnA", | |
| "CodeSearchNet", | |
| "fictitious dialogues", | |
| "prototyping", | |
| "custom-format", | |
| "spirituality", | |
| "geometry-reasoning", | |
| "function-plot", | |
| "Expert Trajectory", | |
| "Louis Vuitton", | |
| "land-cover-classification", | |
| "change-detection", | |
| "telecom", | |
| "exams", | |
| "multi-modal dialogue", | |
| "dalle3", | |
| "Natural Language Processing", | |
| "clothing", | |
| "dictionary", | |
| "poem", | |
| "metadata", | |
| "large", | |
| "speechbox_punc", | |
| "keras-dreambooth", | |
| "Logical-Reasoning", | |
| "code-style", | |
| "Saint Laurent", | |
| "Loro Piana", | |
| "Fendi", | |
| "cohere2", | |
| "Diffusion", | |
| "Flux.1-dev", | |
| "remove background", | |
| "codepy", | |
| "vae", | |
| "bark", | |
| "upscaler", | |
| "SLM", | |
| "video-understanding", | |
| "voice-conversion", | |
| "Healthcare & Lifesciences", | |
| "BioMed", | |
| "Salient Object Detection", | |
| "olmo2", | |
| "timesfm", | |
| "demo", | |
| "seamless_communication", | |
| "in-the-wild", | |
| "Intel", | |
| "Qwen", | |
| "llamacpp", | |
| "photorealism", | |
| "audio-generation", | |
| "ovis", | |
| "llava_phi", | |
| "Quantization", | |
| "flux.1", | |
| "flux.1-dev", | |
| "4x8B", | |
| "baai-nova", | |
| "math_lingo", | |
| "openai-gpt", | |
| "propaganda", | |
| "codet5", | |
| "xlm-r-distilroberta-base-paraphrase-v1", | |
| "multilabel", | |
| "distilgpt2", | |
| "multi-speaker", | |
| "KKY", | |
| "FAV", | |
| "cased", | |
| "BART", | |
| "autoencoder", | |
| "emotion-detection", | |
| "mental health", | |
| "distilbart", | |
| "vn", | |
| "speaker-embedding", | |
| "ancient", | |
| "disease", | |
| "dialog", | |
| "Emotion", | |
| "Speaker", | |
| "Verification", | |
| "stylegan", | |
| "stylegan3", | |
| "coreference resolution", | |
| "entity-retrieval", | |
| "pubmedbert", | |
| "NLG", | |
| "jukebox", | |
| "self-supervised learning", | |
| "lean_albert", | |
| "sarcasm-detection", | |
| "SentimentAnalysis", | |
| "not-for-all-eyes", | |
| "instruction fine-tuning", | |
| "Arabic BERT", | |
| "Masked Langauge Model", | |
| "RefinedWebModel", | |
| "table extraction", | |
| "tokenizer", | |
| "deepspeed", | |
| "taxonomy", | |
| "artwork", | |
| "Network Packets", | |
| "NxtWave-GenAI-Webinar", | |
| "finetuner", | |
| "text-to-code", | |
| "wizard", | |
| "tokenizers", | |
| "wizardlm", | |
| "midi_model", | |
| "smart contract", | |
| "animation", | |
| "decoder", | |
| "future stock prediction", | |
| "trends prediction", | |
| "nature", | |
| "image recognition", | |
| "opensource", | |
| "microsoft", | |
| "spellcheck", | |
| "pixel art", | |
| "earth science", | |
| "vision-language model", | |
| "RyzenAI", | |
| "modern", | |
| "flowers", | |
| "proteins", | |
| "bnb", | |
| "australia", | |
| "maywell/PiVoT-0.1-Starling-LM-RP", | |
| "Personality", | |
| "German", | |
| "argilla/distilabeled-Marcoro14-7B-slerp", | |
| "atari", | |
| "TDT", | |
| "IGN", | |
| "Table", | |
| "smiles", | |
| "deep signal processing", | |
| "hybrid", | |
| "graph problem", | |
| "tinyllava", | |
| "cosmopedia", | |
| "patent", | |
| "PII", | |
| "learned sparse", | |
| "MaziyarPanahi/Calme-7B-Instruct-v0.1.1", | |
| "similarity", | |
| "llms", | |
| "jeiku/Eros_Prodigadigm_7B", | |
| "Misral", | |
| "instruction-tuned", | |
| "rkllm", | |
| "emoji", | |
| "large language models", | |
| "phi-3", | |
| "Chatbot", | |
| "Graph", | |
| "Java", | |
| "Research", | |
| "hunyuan-dit", | |
| "wildfire", | |
| "cogvlm2", | |
| "flash", | |
| "hinglish", | |
| "32 bit upscale", | |
| "mllm", | |
| "openvla", | |
| "AI safety", | |
| "Education", | |
| "EQ", | |
| "mobilenet_v3", | |
| "translate", | |
| "Japanese", | |
| "OpenHermes-2.5", | |
| "Orca", | |
| "SlimOrca", | |
| "aimo", | |
| "electronics", | |
| "neural-vocoder", | |
| "AI4Science", | |
| "Any2Any", | |
| "bash", | |
| "diffusionkit", | |
| "OpenAI", | |
| "anthracite-org/magnum-12b-v2", | |
| "Huggy", | |
| "misinformation", | |
| "GUI", | |
| "nynorsk", | |
| "tibetan", | |
| "llama-3.2-instruct", | |
| "llama-3.2-instruct-3b", | |
| "llama-3-instruct-3b", | |
| "openthaigpt", | |
| "azure", | |
| "pruned", | |
| "small", | |
| "moonshine", | |
| "Theresa", | |
| "\u7279\u857e\u897f\u5a05", | |
| "\u9b54\u738b", | |
| "context 128k", | |
| "slm", | |
| "img2img", | |
| "shuttle", | |
| "backyard", | |
| "BharatGPT", | |
| "CoRover", | |
| "chatbots", | |
| "frankenmerge", | |
| "Reddit", | |
| "Neumind", | |
| "recipe", | |
| "feedback", | |
| "article", | |
| "Long-CoT", | |
| "RL", | |
| "dinov2_with_registers", | |
| "Keras", | |
| "multi-label", | |
| "program-synthesis", | |
| "abusive language", | |
| "bat", | |
| "bert-base-uncased", | |
| "Sinhala", | |
| "t5-new-failed", | |
| "t5-new-hf-not-loaded", | |
| "transducer", | |
| "ibert", | |
| "blenderbot-small", | |
| "clip-vision-bert", | |
| "code_synthesis", | |
| "t5-lm-adapt", | |
| "opennmt", | |
| "semantic-role-labeling", | |
| "osanseviero/hubert_base", | |
| "xlm-roberta-base", | |
| "sentence classification", | |
| "mbert", | |
| "QARiB", | |
| "poolformer", | |
| "dense-passage-retrieval", | |
| "questions", | |
| "modularNMT", | |
| "fake-news", | |
| "data2vec-vision", | |
| "PostTrainingDynamic", | |
| "ofa", | |
| "PongNoFrameskip-v4", | |
| "EnduroNoFrameskip-v4", | |
| "AsteroidsNoFrameskip-v4", | |
| "omnivore", | |
| "Financial Language Modelling", | |
| "MRC", | |
| "video-object-segmentation", | |
| "contrastive learning", | |
| "metric learning", | |
| "MONAI", | |
| "vit_msn", | |
| "claim-detection", | |
| "image segmentation", | |
| "hierarchical-transformer", | |
| "context-sql", | |
| "sqlglot", | |
| "sociology", | |
| "grammatical-error-correction", | |
| "text-editing", | |
| "quran", | |
| "noisy-speech-recognition", | |
| "Construction", | |
| "Logistics", | |
| "Detection", | |
| "Protein", | |
| "3D scenes", | |
| "Instruction Response", | |
| "parliament", | |
| "sentiment analysis, Twitter, tweets", | |
| "captions", | |
| "simulation", | |
| "table-understanding", | |
| "GPT-3.5", | |
| "Claude", | |
| "mask", | |
| "geometry-qa", | |
| "logical-reasoning", | |
| "synthetic-scene", | |
| "plot", | |
| "puzzle-test", | |
| "Compilation", | |
| "electricity", | |
| "maltese", | |
| "3D Gaussian", | |
| "Novel View Synthesis", | |
| "Text to 3D", | |
| "Image to 3D", | |
| "Lilac", | |
| "arXiv", | |
| "problem solving", | |
| "Preference", | |
| "eval", | |
| "photos", | |
| "text data", | |
| "jokes", | |
| "world", | |
| "russia", | |
| "other-image-super-resolution", | |
| "readability", | |
| "multiword-expression-detection", | |
| "coreference", | |
| "geo", | |
| "ciempiess", | |
| "mexican spanish", | |
| "ciempiess project", | |
| "ciempiess-unam project", | |
| "anthrope", | |
| "Multi-Step-Deductive-Reasoning", | |
| "spotlight", | |
| "ghana", | |
| "ltx-video", | |
| "virtual try-on", | |
| "creative-writer", | |
| "multiplicative-lora", | |
| "deep-think", | |
| "qwen2_audio", | |
| "world model", | |
| "torchtune", | |
| "stable-audio-tools", | |
| "multi-class", | |
| "apollo", | |
| "pixmo", | |
| "mdeberta", | |
| "resegmentation", | |
| "facial-recognition", | |
| "sentiment extraction", | |
| "Text2Text Generation", | |
| "sentencepiece", | |
| "audio-spectrogram-transformer", | |
| "Dichotomous Image Segmentation", | |
| "Camouflaged Object Detection", | |
| "ethereum", | |
| "table-transformer", | |
| "multitask", | |
| "tasksource", | |
| "owlv2", | |
| "monocular depth estimation", | |
| "playground", | |
| "automation", | |
| "assistant", | |
| "clickbait", | |
| "llava-qwen2", | |
| "codegeex", | |
| "sound", | |
| "FLUX.1-dev", | |
| "LLM2CLIP", | |
| "comfyui", | |
| "Prompt_Enhance", | |
| "Multi-language", | |
| "mingru", | |
| "piano", | |
| "fluxpipeline", | |
| "genre", | |
| "xlsr-fine-tuning", | |
| "Biomedical", | |
| "bluebert", | |
| "bionlp", | |
| "entity linking", | |
| "fairlex", | |
| "dallebart", | |
| "historic german", | |
| "phoneme-recognition", | |
| "emotion-analysis", | |
| "flue", | |
| "indogpt", | |
| "indonlg", | |
| "tags", | |
| "imagegpt", | |
| "classical", | |
| "chemical", | |
| "rna", | |
| "industry", | |
| "sentence-bert", | |
| "distilbart-mnli", | |
| "dit", | |
| "linkbert", | |
| "biobart", | |
| "javascript", | |
| "Deepspeed", | |
| "Megatron", | |
| "job postings", | |
| "ESG", | |
| "bert-base-cased", | |
| "biodiversity", | |
| "Tacotron2", | |
| "Vocoder", | |
| "ddim_diffusion", | |
| "nezha", | |
| "denoising", | |
| "patents", | |
| "document encoder", | |
| "loss", | |
| "Tensor Flow", | |
| "romanian", | |
| "gpt-neo", | |
| "protein embedding", | |
| "chinese_clip", | |
| "oneformer", | |
| "STS", | |
| "biogpt", | |
| "prompt-generator", | |
| "depression", | |
| "prompt-retrieval", | |
| "stable-diffusion-1.5", | |
| "paddlenlp", | |
| "geolocalization", | |
| "urban", | |
| "clap", | |
| "Norwegian", | |
| "vector search", | |
| "gmp-dev", | |
| "if", | |
| "guanaco", | |
| "generated_from_span_marker_trainer", | |
| "laion", | |
| "face-generation", | |
| "StreamingDatasets", | |
| "tfjs", | |
| "word2vec", | |
| "instructblip", | |
| "float16", | |
| "api", | |
| "fish", | |
| "table structure recognition", | |
| "RefinedWeb", | |
| "Text-to-Video", | |
| "musicgen_melody", | |
| "discord", | |
| "voice model", | |
| "species", | |
| "rare species", | |
| "endangered species", | |
| "evolutionary biology", | |
| "knowledge-guided", | |
| "faster-whisper", | |
| "latin", | |
| "spellchecking", | |
| "hezar", | |
| "indonesia", | |
| "captcha", | |
| "openskyml", | |
| "diffusion model", | |
| "music generation", | |
| "code llama", | |
| "suicide", | |
| "rpg", | |
| "t2i-adapter", | |
| "portrait", | |
| "nougat", | |
| "grounding-dino", | |
| "sales", | |
| "tinyllama", | |
| "comedy", | |
| "Llama 2", | |
| "Hate Speech Detection", | |
| "Mathematical Reasoning", | |
| "mplug_owl2", | |
| "hermes", | |
| "hearthstone", | |
| "funny", | |
| "person", | |
| "rwkv5", | |
| "dalle-3", | |
| "dalle", | |
| "ultrafeedback", | |
| "vibrant", | |
| "comic", | |
| "cinematic", | |
| "drawing", | |
| "space", | |
| "injection", | |
| "kurdish", | |
| "LCM", | |
| "juanako", | |
| "video understanding", | |
| "YOLO", | |
| "C-RLFT", | |
| "threat", | |
| "MBTI", | |
| "beowolx/CodeNinja-1.0-OpenChat-7B", | |
| "TensorFlow", | |
| "samir-fama/SamirGPT-v1", | |
| "abacusai/Slerp-CM-mist-dpo", | |
| "Large Language Models", | |
| "Image-to-Text", | |
| "recognition", | |
| "uzbek", | |
| "dvilasuero/DistilabelBeagle14-7B", | |
| "MemGPT", | |
| "YOLOv8", | |
| "AiMavenAi/AiMaven-Prometheus", | |
| "taiwan", | |
| "\ud83e\udd68", | |
| "laser", | |
| "generative_ai", | |
| "qwen2_moe", | |
| "eren23/ogno-monarch-jaskier-merge-7b", | |
| "mobileclip", | |
| "opensearch", | |
| "Equall/Saul-Instruct-v1", | |
| "speech-to-speech", | |
| "calme", | |
| "Logical Reasoning", | |
| "lightning", | |
| "Undi95/LewdMistral-7B-0.2", | |
| "brasil", | |
| "freeze", | |
| "svbrdf", | |
| "arcee-ai/sec-mistral-7b-instruct-1.6-epoch", | |
| "Cyber-Series", | |
| "aesthetic", | |
| "rockchip", | |
| "rk3588", | |
| "Legal", | |
| "Ruby", | |
| "allknowingroger/MultiverseEx26-7B-slerp", | |
| "llama 3", | |
| "ruslanmv", | |
| "agents", | |
| "traditional_chinese", | |
| "mlabonne/Zebrafish-7B", | |
| "mlp_speculator", | |
| "pets", | |
| "pose", | |
| "Fine-tuning", | |
| "Cantonese", | |
| "xgenmm", | |
| "Text-to-Speech", | |
| "rwkv6", | |
| "fast", | |
| "quality", | |
| "Quantized", | |
| "zero-shot NER", | |
| "npu", | |
| "amd", | |
| "autonomous driving", | |
| "xlm-token", | |
| "cambrian_llama", | |
| "llava_qwen2", | |
| "Blockchain", | |
| "cypher", | |
| "neo4j", | |
| "regmix", | |
| "deepfake", | |
| "in-context learning", | |
| "Question-Answer", | |
| "fbgemm_fp8", | |
| "sample-generation", | |
| "stable-audio", | |
| "transfer learning", | |
| "nai", | |
| "mental_health", | |
| "Whisper", | |
| "Speech", | |
| "multimodal retrieval", | |
| "aetherwiing/MN-12B-Starcannon-v3", | |
| "Politics", | |
| "affiliations", | |
| "NovelAI", | |
| "ijepa", | |
| "gemma-2", | |
| "matting", | |
| "rknn", | |
| "banglaLLM", | |
| "banglaNLP", | |
| "LLama", | |
| "Foundation Model", | |
| "Uncensored", | |
| "dev-ops", | |
| "developer", | |
| "Locutusque/Hercules-6.1-Llama-3.1-8B", | |
| "yolo11", | |
| "purz", | |
| "flux1.d", | |
| "NeuralMahou", | |
| "NIHAPPY", | |
| "Flux-Dev", | |
| "taxation", | |
| "fiscalit\u00e9", | |
| "GenAI", | |
| "conversational ai", | |
| "p3", | |
| "Datasets", | |
| "70b", | |
| "public domain", | |
| "computervision", | |
| "housing", | |
| "inference", | |
| "fine-grained", | |
| "1024px_based_image_size", | |
| "GPTSoVITS", | |
| "loi", | |
| "So-VITS-SVC", | |
| "Modeling", | |
| "low-resource-language", | |
| "DefinitionGeneration", | |
| "definition-modeling", | |
| "anvita", | |
| "homer", | |
| "Qandora", | |
| "Sentient", | |
| "tech", | |
| "Rosmontis", | |
| "\u8ff7\u8fed\u9999", | |
| "Llama-CPP", | |
| "bokm\u00e5l", | |
| "qwq", | |
| "bluesky", | |
| "vtuber", | |
| "unlearning", | |
| "MilkDrop", | |
| "Korean", | |
| "notes", | |
| "reconstruction", | |
| "Chain-of-Thought Activation", | |
| "tagger", | |
| "rule34", | |
| "ERP", | |
| "humor", | |
| "French", | |
| "RobBERTje", | |
| "transformer_vae", | |
| "cel", | |
| "fiu", | |
| "gmw", | |
| "torch==1.8.0", | |
| "\u00e6l\u00e6ctra", | |
| "ELECTRA-Small", | |
| "replaced token detection", | |
| "collaborative", | |
| "answering", | |
| "data2text", | |
| "semeval2020", | |
| "comve", | |
| "hierarchical_model", | |
| "parlaspeech", | |
| "stateless transducer", | |
| "Greek", | |
| "TextClassification", | |
| "speech2text2", | |
| "character_bert", | |
| "xlmindic", | |
| "indoaryan", | |
| "iso15919", | |
| "jira", | |
| "ga-IE", | |
| "reinforcement learning", | |
| "gpt-2", | |
| "satflow", | |
| "other-image-classification", | |
| "npc-engine", | |
| "nowcasting", | |
| "llama-leaderboard", | |
| "encoder_decoder", | |
| "xls_r_repro_common_voice_tr", | |
| "phongdtd/VinDataVLSP", | |
| "html", | |
| "LSTM", | |
| "structured-data-classification", | |
| "sentence-embeddings", | |
| "spanish-english", | |
| "RUDOLPH", | |
| "morphology", | |
| "SplinterModel", | |
| "mongolian", | |
| "distilt5", | |
| "distilt5-qg", | |
| "text-2-text-generation", | |
| "question_answering", | |
| "medicalimaging", | |
| "belarusian", | |
| "oral", | |
| "roformer-v2", | |
| "serbian", | |
| "MTee", | |
| "general", | |
| "liltrobertalike", | |
| "greaselm", | |
| "multilingual model", | |
| "non-commercial", | |
| "seals/Ant-v0", | |
| "seals/Hopper-v0", | |
| "seals/Humanoid-v0", | |
| "RoadRunnerNoFrameskip-v4", | |
| "SeaquestNoFrameskip-v4", | |
| "Humanoid-v3", | |
| "deep-rl-course", | |
| "seals/HalfCheetah-v0", | |
| "seals/Swimmer-v0", | |
| "vison", | |
| "tranception", | |
| "decoding", | |
| "Poet", | |
| "MsPacmanNoFrameskip-v4", | |
| "document-understanding", | |
| "tfhub", | |
| "UAV", | |
| "erzya", | |
| "mordovian", | |
| "plagiarism", | |
| "conversations", | |
| "humanities", | |
| "social_science", | |
| "computer_science", | |
| "claude", | |
| "observers", | |
| "customer service", | |
| "multi-domain", | |
| "evaluating-dialogue-systems", | |
| "knowledge-verification", | |
| "multi-hop", | |
| "query-based-summarization", | |
| "optical-character-recognition", | |
| "markets", | |
| "academic", | |
| "scientific text", | |
| "Web3", | |
| "research paper", | |
| "stackoverflow", | |
| "en-atc", | |
| "stacked summaries", | |
| "personality", | |
| "Damage Risk", | |
| "sec", | |
| "SimpleAI", | |
| "mxeval", | |
| "transcripts", | |
| "knowledge-base-qa", | |
| "mental", | |
| "dolly", | |
| "automatic speech recognition", | |
| "Spatial-Temporal", | |
| "sharegpt", | |
| "theorem-proving", | |
| "geology", | |
| "fraud", | |
| "Genomics", | |
| "hallucination", | |
| "spotify", | |
| "table-structure-recognition", | |
| "a.", | |
| "haiku", | |
| "red teaming", | |
| "parallel data", | |
| "Histopathology", | |
| "Histology", | |
| "Digital Pathology", | |
| "prompt injection", | |
| "Bard", | |
| "LLaMA-2", | |
| "Vicuna", | |
| "PaLM-2", | |
| "math-word-problem", | |
| "arithmetic-reasoning", | |
| "algebraic-reasoning", | |
| "numeric-common-sense", | |
| "scientific-reasoning", | |
| "abstract-scene", | |
| "Environement", | |
| "mt-evaluation", | |
| "WMT", | |
| "sentinel-1", | |
| "data science", | |
| "galgame", | |
| "website", | |
| "Safety", | |
| "image captioning", | |
| "document", | |
| "Multimodal Learning", | |
| "biomedicine", | |
| "Livre des proc\u00e9dures fiscales", | |
| "Code du travail", | |
| "Code de commerce", | |
| "Code mon\u00e9taire et financier", | |
| "Code de la construction et de l'habitation", | |
| "Code civil", | |
| "Code de la consommation", | |
| "Code des assurances", | |
| "Code de la propri\u00e9t\u00e9 intellectuelle", | |
| "Code de la commande publique", | |
| "Code p\u00e9nal", | |
| "mozilla", | |
| "news articles", | |
| "spatial", | |
| "recipes", | |
| "driving", | |
| "Human", | |
| "country", | |
| "stabeldiffusion", | |
| "facial recognition", | |
| "documentation", | |
| "offensive-language", | |
| "code-mixed", | |
| "creature-dataset", | |
| "validation", | |
| "policy", | |
| "samromur", | |
| "word frequencies", | |
| "AIvtuber", | |
| "VirtuaReal", | |
| "web-text", | |
| "plain language", | |
| "easy-to-read language", | |
| "semantics", | |
| "esa", | |
| "uci", | |
| "dhivehi", | |
| "fiscal", | |
| "web-agent", | |
| "Chloe", | |
| "Bottega Veneta", | |
| "Balenciaga", | |
| "Chanel", | |
| "flux dev", | |
| "Llama 3.2 MOE", | |
| "FLUX", | |
| "deepseek_vl_v2", | |
| "Super-Realism", | |
| "code-solve", | |
| "algorithm", | |
| "qwen_base", | |
| "bfcl", | |
| "pattern recognition", | |
| "chart reader", | |
| "muiltimodal", | |
| "unified-model", | |
| "grok-1", | |
| "cross-modal", | |
| "Patronus AI", | |
| "hallucination-detection", | |
| "triangulum_1b", | |
| "text_to_image", | |
| "mambavision", | |
| "rdt", | |
| "Emu3", | |
| "Qwen2-VL", | |
| "Pixart-\u03b1", | |
| "Kolors", | |
| "Photorealistic", | |
| "stable-diffusion-3.5-large", | |
| "v6", | |
| "erax", | |
| "memes", | |
| "LMM", | |
| "Code Summarization", | |
| "cryptocurrency", | |
| "cantonese", | |
| "vivit", | |
| "gpt-neox", | |
| "pipeline", | |
| "extreme-multi-task", | |
| "extreme-mtl", | |
| "seamless_m4t_v2", | |
| "vmistral", | |
| "home", | |
| "noticia", | |
| "reward_model", | |
| "phimoe", | |
| "Inpainting", | |
| "ncsoft", | |
| "varco", | |
| "hymba", | |
| "flux.1-schnell", | |
| "flux-merge", | |
| "Wav2Vec2", | |
| "DDSP-SVC", | |
| "PartAI", | |
| "ablated", | |
| "ablation", | |
| "nani", | |
| "sentiment_analysis", | |
| "multimodal-retrieval", | |
| "music ai", | |
| "grammatical error correction", | |
| "text2text", | |
| "kenlm", | |
| "xls_r_pretrained", | |
| "kb", | |
| "tp", | |
| "masked-token-prediction", | |
| "bert-base-multilingual-cased", | |
| "drugs", | |
| "sentence_50agree", | |
| "\u6587\u8a00\u6587", | |
| "bioinfomatics", | |
| "Speech Enhancement", | |
| "Spoken language understanding", | |
| "ECAPA", | |
| "Sentence-Embedding", | |
| "Similarity", | |
| "financial-text-analysis", | |
| "Business", | |
| "Financial News", | |
| "Sustainability", | |
| "HiFIGAN", | |
| "BERT2BERT", | |
| "FLAN", | |
| "prompting", | |
| "license-plate-detection", | |
| "math-aware", | |
| "robots", | |
| "lilt", | |
| "maxim", | |
| "VAE", | |
| "Data Augmentation", | |
| "handwriting-recognition", | |
| "entity typing", | |
| "galactica", | |
| "text2text generation", | |
| "textual-inversion", | |
| "puli", | |
| "textual inversion", | |
| "graphs", | |
| "LoRa", | |
| "extractive", | |
| "abstractive", | |
| "Image Generation", | |
| "STT", | |
| "animal", | |
| "geolocation", | |
| "text retrieval", | |
| "sentence-boundary-detection", | |
| "latent diffusion", | |
| "xmod", | |
| "human_genome", | |
| "matcha", | |
| "baize", | |
| "instructional", | |
| "generative-adversarial-network", | |
| "albertina-pt*", | |
| "Text2Text-Generation", | |
| "trading", | |
| "encodec", | |
| "beaver", | |
| "locon", | |
| "shap-e", | |
| "idefics", | |
| "upstage", | |
| "music-captioning", | |
| "Geospatial", | |
| "synthetic instruction", | |
| "kollama", | |
| "llama-2-ko", | |
| "Foundation model", | |
| "multilingual-code-generation", | |
| "plants", | |
| "document analysis", | |
| "unstructured document", | |
| "computer_vision", | |
| "fairseq2", | |
| "SeamlessM4T", | |
| "functions", | |
| "icons", | |
| "midi generation", | |
| "automatic-speech-translation", | |
| "foundation-model", | |
| "vitmatte", | |
| "PyLaia", | |
| "artist", | |
| "sparse sparsity quantized onnx embeddings int8", | |
| "audio-captioning", | |
| "streaming", | |
| "m42", | |
| "clinical-llm", | |
| "Bode", | |
| "Social Media", | |
| "BlueLM", | |
| "distilled-model", | |
| "StableLM", | |
| "causallm", | |
| "openhermes", | |
| "m2_bert", | |
| "\u5c01\u795e\u699c", | |
| "layout-to-image", | |
| "wespeaker", | |
| "orca2", | |
| "nemotron-3", | |
| "surrealism", | |
| "kawaii", | |
| "silly", | |
| "magic", | |
| "digital art", | |
| "Storywriter", | |
| "expressionism", | |
| "landscapes", | |
| "share4v", | |
| "art style", | |
| "pruna-engine", | |
| "cities", | |
| "full body", | |
| "sorani", | |
| "ultra", | |
| "sentinel2", | |
| "landsat", | |
| "arabic ", | |
| "audio2text", | |
| "shadow", | |
| "Megatron-LM", | |
| "ctc", | |
| "UBC", | |
| "DLNLP", | |
| "MachineMindset", | |
| "cookinai/CatMacaroni-Slerp", | |
| "hf_olmo", | |
| "MRI", | |
| "Weyaxi/OpenHermes-2.5-neural-chat-v3-3-openchat-3.5-1210-Slerp", | |
| "EmbeddedLLM/Mistral-7B-Merge-14-v0.1", | |
| "AI-B/UTENA-7B-UNA-V2", | |
| "AI-B/UTENA-7B-NSFW-V2", | |
| "orion", | |
| "liminerity/Blur-7b-slerp-v0.1", | |
| "has_space", | |
| "liminerity/Blured-Ties-7B", | |
| "freecs/ThetaWave-7B", | |
| "Manga", | |
| "Object Detection", | |
| "Scepter", | |
| "Solar", | |
| "abideen/DareVox-7B", | |
| "senseable/Westlake-7B-v2", | |
| "vistral", | |
| "meta-math/MetaMath-Mistral-7B", | |
| "CultriX/MergeTrix-7B-v2", | |
| "snorkelai/Snorkel-Mistral-PairRM-DPO", | |
| "myanmar", | |
| "cybersecwithai", | |
| "ai4security", | |
| "llmsecurity", | |
| "VAGOsolutions/SauerkrautLM-7b-v1-mistral", | |
| "chargoddard/loyal-piano-m7", | |
| "free", | |
| "freeai", | |
| "SanjiWatsuki/Loyal-Toppy-Bruins-Maid-7B-DARE", | |
| "ND911/EE-LMaid-7B-Slerp", | |
| "flemmingmiguel/MBX-7B", | |
| "chargoddard/loyal-piano-m7-cdpo", | |
| "athirdpath/NSFW_DPO_vmgb-7b", | |
| "kaitchup/Mayonnaise-4in1-022", | |
| "judge", | |
| "rishiraj/CatPPT-base", | |
| "latex-ocr", | |
| "Kukedlc/NeuTrixOmniBe-7B-model-remix", | |
| "machinists/Mistral-7B-SQL", | |
| "Zenith-7B-dpo-v2", | |
| "Eclipse-13B-dpo", | |
| "scholarly", | |
| "traditional chinese", | |
| "zh-hant", | |
| "Kukedlc/Neural-4-ARC-7B-slerp", | |
| "Kukedlc/Neural-4-GSM8K-7B-slerp", | |
| "paulml/OGNO-7B", | |
| "bunny-phi", | |
| "\ud83c\udf7b", | |
| "9B", | |
| "marathi", | |
| "gordicaleksa/YugoGPT", | |
| "khanacademy", | |
| "vc", | |
| "dac", | |
| "logo", | |
| "Kukedlc/NeoCortex-7B-slerp", | |
| "cognitivecomputations/samantha-mistral-7b", | |
| "CorticalStack/shadow-clown-7B-dare", | |
| "BryanSwk/LaserPipe-7B-SLERP", | |
| "function-call", | |
| "Kukedlc/NeuralGlitch-Yam-Peleg-7B-DT", | |
| "StableDiffusionXL", | |
| "voice conversion", | |
| "block expansion", | |
| "progressive mistral", | |
| "arcee cpt", | |
| "unity", | |
| "rwitz/experiment26-truthy-iter-0", | |
| "Hypersniper/The_Philosopher_Zephyr_7B", | |
| "sayhan/OpenHermes-2.5-Strix-Philosophy-Mistral-7B-LoRA", | |
| "AetherResearch/Cerebrum-1.0-7b", | |
| "Kukedlc/SuperMente-7B-v2", | |
| "nlpguy/AlloyIngotNeoX", | |
| "automerger/YamShadow-7B", | |
| "self-supervised", | |
| "vortexmergekit", | |
| "Problem Solving", | |
| "AI Programming Assistant", | |
| "mipha_phi", | |
| "unsloth/mistral-7b-v0.2", | |
| "Yuma42/KangalKhan-RawRuby-7B", | |
| "nlpguy/T3QM7", | |
| "PocketDoc/Dans-AdventurousWinds-Mk2-7b", | |
| "Smuggling1710/Ak4ri-7b", | |
| "Base Model", | |
| "dbrx", | |
| "all-in-one", | |
| "lemonilia/LimaRP-Mistral-7B-v0.1", | |
| "Weyaxi/Einstein-v5-v0.2-7B", | |
| "preference_learning", | |
| "educational", | |
| "feature_extraction", | |
| "hatespeech", | |
| "R", | |
| "Julia", | |
| "chaoyi-wu/MedLLaMA_13B", | |
| "lmsys/vicuna-13b-v1.3", | |
| "Text-to-sql", | |
| "lex-hue/Delexa-7b", | |
| "Diffusion Models", | |
| "sprites", | |
| "yleo/EmertonMonarch-7B", | |
| "keyword-extraction", | |
| "merlinite", | |
| "pyannet", | |
| "bunny-llama", | |
| "xtuner", | |
| "vitpose", | |
| "Phi-3", | |
| "snowflake", | |
| "style-bert-vits2", | |
| "style-bert-vits2-jp-extra", | |
| "childish", | |
| "childish voice", | |
| "text2audio", | |
| "text to audio", | |
| "64k", | |
| "iMat", | |
| "zero shot", | |
| "docvqa", | |
| "visual question answering", | |
| "Mental Health", | |
| "llama-3-8b", | |
| "ddh0/OrcaMaid-v3-13b-32k", | |
| "DavidAU/D_AU-Orac-13B-Tiefighter-slerp", | |
| "Jayant9928/orpo_med_v3", | |
| "skumar9/Llama-medx_v3", | |
| "mantis", | |
| "Uncensored conversation", | |
| "Uncensored jokes", | |
| "Uncensored romance", | |
| "merges", | |
| "PHP", | |
| "phi3small", | |
| "sdxl-flash", | |
| "hyper", | |
| "fast-sdxl", | |
| "sd-community", | |
| "general tasks", | |
| "selm", | |
| "finbert", | |
| "dora", | |
| "detoxification", | |
| "llava_next_video", | |
| "hpc", | |
| "axonn", | |
| "large_language_model", | |
| "sec_data", | |
| "continual_pre_training", | |
| "penetration testing", | |
| "texture", | |
| "entity", | |
| "entity disambiguation", | |
| "depth-anything-v2", | |
| "editing", | |
| "openlm", | |
| "OmnicromsBrain/StoryFusion-7B", | |
| "Healthcare", | |
| "Reward Model", | |
| "TaMeR", | |
| "groq", | |
| "Solana", | |
| "SmartContract", | |
| "semi-realistic", | |
| "FinGPT", | |
| "Banks", | |
| "Wealth Management", | |
| "evf", | |
| "Video Generative Model Training", | |
| "Text-to-Video Diffusion Model Training", | |
| "GDPR", | |
| "Spatial Understanding", | |
| "Self-supervised Learning", | |
| "datadreamer", | |
| "audio classification", | |
| "Sequence-Classification", | |
| "transliteration system", | |
| "entity-relationship-extraction", | |
| "news-analysis", | |
| "autogptq", | |
| "Knowledge Graphs", | |
| "light-embed", | |
| "scene generation", | |
| "shell", | |
| "BiRefNet", | |
| "text-correction", | |
| "word-sense-linking", | |
| "lexical-semantics", | |
| "high-quality", | |
| "OpenAssistant", | |
| "lighting", | |
| "Deep Learning", | |
| "goldfish", | |
| "debugging", | |
| "mplugowl3", | |
| "network", | |
| "videollama2_qwen2", | |
| "SmolLM", | |
| "ghost", | |
| "foundation-models", | |
| "golang", | |
| "go", | |
| "eagle_llama", | |
| "Eagle", | |
| "sd.cpp", | |
| "stable-diffusion.cpp", | |
| "Retrieval Augmented Generation", | |
| "role-playing", | |
| "novel-generation", | |
| "minimind", | |
| "float8_e4m3fn", | |
| "HotpotQA", | |
| "RNA", | |
| "text2cypher", | |
| "pii", | |
| "lm-judge", | |
| "anatomy", | |
| "litgpt", | |
| "litdata", | |
| "oryx_qwen", | |
| "lerobot", | |
| "novelwriting", | |
| "NASA", | |
| "Named entity recognition", | |
| "Articles", | |
| "neuralmagic", | |
| "llmcompressor", | |
| "TW", | |
| "jina-judge", | |
| "brain", | |
| "slang", | |
| "esper", | |
| "terraform", | |
| "aws", | |
| "gcp", | |
| "architect", | |
| "engineer", | |
| "opus-mt-tc-bible", | |
| "cryptology", | |
| "cipher", | |
| "emerald", | |
| "welding", | |
| "PHI", | |
| "glm4", | |
| "longwriter", | |
| "cambrian_qwen", | |
| "medembed", | |
| "medical-embedding", | |
| "clinical-embedding", | |
| "4-bit precision", | |
| "Fine-Tuning with LoRA", | |
| "NT GenAI", | |
| "ntgenai", | |
| "lahnmah", | |
| "NT Thai GPT", | |
| "ntthaigpt", | |
| "\u0e2b\u0e25\u0e32\u0e19\u0e21\u0e48\u0e32", | |
| "NT Academy", | |
| "VPTQ", | |
| "vptq", | |
| "LCARS", | |
| "gpt_optimized", | |
| "conversational skill", | |
| "Pneuma", | |
| "Allades", | |
| "mergekit-community/L3.1-Pneuma-8B-v1", | |
| "marvel", | |
| "model_fusion", | |
| "coding_assistant", | |
| "creative_writing", | |
| "latent_diffusion", | |
| "long_context", | |
| "agentic_AI", | |
| "multi_domain", | |
| "technical_reasoning", | |
| "task_generalization", | |
| "AI_tools", | |
| "llmware-fx", | |
| "llama 3.2", | |
| "CohereForAI", | |
| "Music", | |
| "Generation", | |
| "SmolLM2", | |
| "webrl", | |
| "webarena-lite", | |
| "denoiser", | |
| "Lappland", | |
| "\u62c9\u666e\u5170\u5fb7", | |
| "qiskit", | |
| "hydra-project/ChatHercules-2.5-Mistral-7B", | |
| "Nitral-Archive/Prima-Pastacles-7b", | |
| "coherence", | |
| "retro", | |
| "segment anything", | |
| "zero-shot matting", | |
| "medical-embeddings", | |
| "512px_based_image_size", | |
| "llama-2-chat", | |
| "SmolTalk", | |
| "95b", | |
| "encyclopedia", | |
| "meme", | |
| "Role-play", | |
| "google-earth", | |
| "3DXL", | |
| "influencer", | |
| "anomaly-detection", | |
| "Magpie", | |
| "Tulu3", | |
| "Smollm", | |
| "SLMs", | |
| "Huggingface", | |
| "Allenai", | |
| "empathetic", | |
| "masked-auto-encoding", | |
| "turing", | |
| "Priya", | |
| "Teen-AI", | |
| "Promptist-Instruct", | |
| "DELLA-linear", | |
| "Visualizations", | |
| "dnotitia", | |
| "MATH", | |
| "higgs", | |
| "llama-3.3", | |
| "render", | |
| "CAD", | |
| "Blender", | |
| "plonk", | |
| "text to image", | |
| "peer", | |
| "automatic", | |
| "conference", | |
| "manuscript", | |
| "openreview", | |
| "model-stock", | |
| "Qwen with Questions", | |
| "summaries", | |
| "DucHaiten", | |
| "inappropriate", | |
| "stories", | |
| "sex", | |
| "BeaverAI/mistral-doryV2-12b", | |
| "for-entertainment-purposes-only", | |
| "do-not-use-seriously", | |
| "cannabis", | |
| "HELVETE-X", | |
| "Unfiltered-AI", | |
| "Advanced-EI", | |
| "character_creation", | |
| "world-building", | |
| "lore_writing", | |
| "misogyny", | |
| "Axon", | |
| "Elixir", | |
| "DistilBert", | |
| "Dialect", | |
| "Tweets", | |
| "rm-vallader", | |
| "Russian-speech-corpus", | |
| "afa", | |
| "cpp", | |
| "iir", | |
| "ine", | |
| "itc", | |
| "sem", | |
| "sla", | |
| "urj", | |
| "sentence segmentation", | |
| "beto", | |
| "Slovak GPT-J", | |
| "fake", | |
| "translation Cszech Deustch model", | |
| "translation Cszech English model", | |
| "translation Cszech Spanish model", | |
| "translation Cszech French model", | |
| "translation Cszech Italian model", | |
| "translation Cszech Swedish model", | |
| "translation Deustch English model", | |
| "translation Deustch Spanish model", | |
| "translation Deustch French model", | |
| "translation Deustch Italian model", | |
| "translation Deustch Swedish model", | |
| "translation English Cszech model", | |
| "translation English Deustch model", | |
| "translation English Italian model", | |
| "translation French Cszech model", | |
| "translation French English model", | |
| "translation French Spanish model", | |
| "translation French Italian model", | |
| "translation French Swedish model", | |
| "translation Italian Cszech model", | |
| "translation Italian Deustch model", | |
| "translation Italian English model", | |
| "translation Italian Spanish model", | |
| "translation Italian French model", | |
| "translation Italian Swedish model", | |
| "translation Swedish Cszech model", | |
| "translation Swedish Deustch model", | |
| "translation Swedish English model", | |
| "translation Swedish Spanish model", | |
| "translation Swedish French model", | |
| "translation Swedish Italian model", | |
| "ChatBot", | |
| "openslr_SLR53", | |
| "qg", | |
| "wolof", | |
| "Bible", | |
| "simcls", | |
| "torchscript", | |
| "FastNN", | |
| "malayalam", | |
| "social_media", | |
| "dacy", | |
| "pos tagging", | |
| "morphological analysis", | |
| "named entity linking", | |
| "named entity disambiguation", | |
| "openslr_SLR66", | |
| "cyrillic", | |
| "historic french", | |
| "COVID-19", | |
| "irony", | |
| "conditional-image-generation", | |
| "clip-vision-marian", | |
| "elasticbert", | |
| "Multi-exit-BERT", | |
| "azureml", | |
| "intent classification", | |
| "pt_BR", | |
| "wobert", | |
| "lassl", | |
| "sci", | |
| "nyc", | |
| "films", | |
| "banking", | |
| "mudes", | |
| "mental-health", | |
| "mobile app descriptions", | |
| "playstore", | |
| "ehr", | |
| "wavlm_libri_finetune", | |
| "mozilla-foundation/common_voice_3_0", | |
| "casual language modeling", | |
| "parsbert", | |
| "query-paraphrasing", | |
| "reading-comprehension", | |
| "BabyBERTa", | |
| "RNN", | |
| "lfqa", | |
| "hindi-english", | |
| "semantic-similarity", | |
| "classics", | |
| "simcse", | |
| "sbert", | |
| "faroese", | |
| "distractor", | |
| "dstc10", | |
| "singapore", | |
| "singlish", | |
| "manglish", | |
| "masked-image-modeling", | |
| "m2m100-12B", | |
| "abdusahmbzuai/arabic_speech_massive_300hrs", | |
| "cyberbullying", | |
| "hunflair", | |
| "text-to-rating", | |
| "QG", | |
| "xgboost", | |
| "r3m", | |
| "quality estimation", | |
| "document sections", | |
| "document classification", | |
| "language classification", | |
| "groupvit", | |
| "domain adaptation", | |
| "collaborative-filtering", | |
| "recommender", | |
| "mT5_multilingual_XLSum", | |
| "ImageClassification", | |
| "molecule-generation", | |
| "cheminformatics", | |
| "biochemical-language-models", | |
| "xlmroberta", | |
| "Caribbean dialect", | |
| "paraphrase-detection", | |
| "owlvit", | |
| "formality transfer", | |
| "PROP", | |
| "Pretrain4IR", | |
| "btcv", | |
| "Hindi", | |
| "ZEN", | |
| "HTR", | |
| "face alignment", | |
| "facial landmark point", | |
| "pose estimation", | |
| "tokenizer only", | |
| "luxembourgish", | |
| "l\u00ebtzebuergesch", | |
| "adverse-drug-events", | |
| "social-media-mining-for-health", | |
| "SMM4H", | |
| "\u4e2d\u6587", | |
| "masked language modeling", | |
| "albert_act", | |
| "software", | |
| "rejection-sampling", | |
| "large-language-model", | |
| "screenshots", | |
| "Satellite", | |
| "Asia", | |
| "identification", | |
| "idioms", | |
| "conversational-qa", | |
| "meaning-representation-to-text", | |
| "knowledge-base", | |
| "probing", | |
| "fake-news-detection", | |
| "superglue", | |
| "audio-slot-filling", | |
| "code-switching", | |
| "cleaned", | |
| "long-texts", | |
| "children education", | |
| "oxford", | |
| "citation", | |
| "text-detection", | |
| "image-text pairs", | |
| "stackexchange", | |
| "commonsense reasoning", | |
| "academic integrity", | |
| "theses", | |
| "acl", | |
| "headline", | |
| "gender bias", | |
| "social bias", | |
| "edgar", | |
| "filings", | |
| "10K", | |
| "10-K", | |
| "Self Driving", | |
| "public", | |
| "selfies", | |
| "Gaming", | |
| "Retail", | |
| "Aerial", | |
| "layout-segmentation", | |
| "DocLayNet", | |
| "Financial-Reports", | |
| "Manuals", | |
| "Scientific-Articles", | |
| "Laws", | |
| "Regulations", | |
| "Patents", | |
| "Government-Tenders", | |
| "preference model", | |
| "microbiome", | |
| "argument mining", | |
| "television", | |
| "steam", | |
| "gpt-4", | |
| "nutrition", | |
| "Jailbreak", | |
| "Last-mile Delivery", | |
| "molecule", | |
| "balanced", | |
| "websites", | |
| "music-to-text", | |
| "function call", | |
| "ehartford", | |
| "embodied ai", | |
| "unit test", | |
| "shakespeare", | |
| "arithmetics", | |
| "logos", | |
| "rdf", | |
| "ontology", | |
| "harmless", | |
| "occultism", | |
| "statistical-reasoning", | |
| "crawl", | |
| "papers", | |
| "question-answer", | |
| "haerae", | |
| "4d", | |
| "History", | |
| "aerial", | |
| "typescript", | |
| "outdoor", | |
| "weather", | |
| "mathematical-reasoning", | |
| "LVLM", | |
| "statistics", | |
| "culture", | |
| "multilabel classification", | |
| "vdf", | |
| "vector-io", | |
| "vector-dataset", | |
| "vector-embeddings", | |
| "Retrieval", | |
| "retrieval-augmented generation", | |
| "Synthetic", | |
| "dialogue-system", | |
| "internet", | |
| "sar", | |
| "Benchmarks", | |
| "conversational QA", | |
| "multi-turn QA", | |
| "QA with context", | |
| "Alignment-Lab-AI", | |
| "Document Retrieval", | |
| "documents", | |
| "RGB", | |
| "LLaVA-NeXt", | |
| "refusal", | |
| "Entertainment", | |
| "articles", | |
| "droit-fran\u00e7ais", | |
| "code-civil", | |
| "juris", | |
| "Code g\u00e9n\u00e9ral des imp\u00f4ts", | |
| "Code de la d\u00e9fense", | |
| "Code de l'action sociale et des familles", | |
| "Code du cin\u00e9ma et de l'image anim\u00e9e", | |
| "Code des impositions sur les biens et services", | |
| "Ocean", | |
| "visual novel", | |
| "adversarial robustness", | |
| "yelp", | |
| "semisynthetic", | |
| "africa", | |
| "anti-spoofing", | |
| "arknights", | |
| "fma", | |
| "free-music-archive", | |
| "humour", | |
| "dialogues", | |
| "face recognition", | |
| "chat-instruct", | |
| "llama-3.1-405b", | |
| "lezghian", | |
| "Software Analysis", | |
| "Human Pose and Shape Estimation", | |
| "Synthetic Data", | |
| "devops", | |
| "lidar", | |
| "manuscripts", | |
| "blood", | |
| "so100", | |
| "tutorial", | |
| "india", | |
| "github", | |
| "drone", | |
| "computer vison", | |
| "temporal", | |
| "vectors", | |
| "clusterisation", | |
| "flux1.1", | |
| "flux1", | |
| "imagen3", | |
| "jsonl", | |
| "radio", | |
| "patent-summarization", | |
| "news-category-classification", | |
| "gender-bias", | |
| "constituency-parsing", | |
| "causal-reasoning", | |
| "speech-modeling", | |
| "argument-mining", | |
| "abbreviation-detection", | |
| "conservation", | |
| "movie reviews", | |
| "sentence similarity", | |
| "metaphor-classification", | |
| "toxic comments", | |
| "unsplash", | |
| "kids", | |
| "pretraining-with-human-feedback", | |
| "short answer feedback", | |
| "maps", | |
| "masri", | |
| "masri-project", | |
| "malta", | |
| "crowd-sourced icelandic", | |
| "icelandic speech", | |
| "iceland", | |
| "demographic", | |
| "doc2query--", | |
| "low-resource", | |
| "Reverse Engineered", | |
| "faiss", | |
| "adult", | |
| "occlusion", | |
| "face detection", | |
| "vision-and-language", | |
| "fake news", | |
| "student performance", | |
| "arXiv.org", | |
| "publication", | |
| "preprint", | |
| "section", | |
| "wikification", | |
| "dogs", | |
| "anaphora", | |
| "mode classification", | |
| "renumics", | |
| "shopee", | |
| "complex", | |
| "stopwords", | |
| "keyword2", | |
| "keyword1", | |
| "natsql", | |
| "sql finetune", | |
| "langchain-docs", | |
| "binary-sentiment-analysis", | |
| "mgame", | |
| "machine reading", | |
| "morphological-inflection", | |
| "Non-fungible Tokens", | |
| "Crypto", | |
| "keywords-extraction", | |
| "spectrogram", | |
| "dialogue segmentation", | |
| "Dior", | |
| "Hermes", | |
| "Blickers", | |
| "taiga", | |
| "tayga", | |
| "ghana-news", | |
| "pinyin", | |
| "mining", | |
| "UltraRealism", | |
| "RAW", | |
| "4K", | |
| "virtual try-off", | |
| "triangulum_10b", | |
| "video-to-audio", | |
| "triangulum_5b", | |
| "social-media-analysis", | |
| "customer-feedback", | |
| "product-reviews", | |
| "brand-monitoring", | |
| "intel/auto-round", | |
| "instant-voice-cloning", | |
| "kolors", | |
| "GOT", | |
| "ocr2.0", | |
| "jina_clip", | |
| "eva02", | |
| "Contact Doctor", | |
| "Llama 3", | |
| "nvembed", | |
| "qwen2vl", | |
| "4x7B", | |
| "32 bit enhanced", | |
| "float 32 quants", | |
| "mistral MOE", | |
| "granite-tsfm", | |
| "tinytimemixer", | |
| "RAGatouille", | |
| "minicpm3", | |
| "template", | |
| "erax-vl-2B", | |
| "4x3B", | |
| "Cute", | |
| "Girls", | |
| "Barons", | |
| "juaner0211589", | |
| "clinicaltrials", | |
| "PyLate", | |
| "Explain code", | |
| "sentiment classification", | |
| "Object detection", | |
| "tvlt", | |
| "deberta-v3-large", | |
| "dghs-imgutils", | |
| "pixelart", | |
| "single image depth estimation", | |
| "real_time", | |
| "whisperkit", | |
| "diffusion distillation", | |
| "lrm_generator", | |
| "IP-Adapter", | |
| "kosmos-2.5", | |
| "audiogen", | |
| "reality", | |
| "ko_leaderboard", | |
| "SVDQuant", | |
| "INT4", | |
| "FLUX.1", | |
| "Retro", | |
| "computer use", | |
| "PG-13", | |
| "LLM-as-a-Judge", | |
| "OBJ", | |
| "Programming", | |
| "Adapter", | |
| "2Kpx_based_image_size", | |
| "mochi", | |
| "optimizer_states", | |
| "minGRU", | |
| "hf_integration", | |
| "embedding-model", | |
| "Applio", | |
| "monster", | |
| "coder", | |
| "library", | |
| "jap", | |
| "trk", | |
| "bert-fa", | |
| "bert-persian", | |
| "persian-lm", | |
| "music-modeling", | |
| "Question Generation", | |
| "herbert", | |
| "Diseases", | |
| "passage reranking", | |
| "aragpt2", | |
| "t5-base", | |
| "perplexity", | |
| "n-gram", | |
| "kneser-ney", | |
| "bigscience", | |
| "ai-msgbot", | |
| "EDSR", | |
| "bert-large", | |
| "flaubert-large", | |
| "recipe-generation", | |
| "threat hunting", | |
| "threat intelligence", | |
| "array", | |
| "of", | |
| "crosslingual", | |
| "chemist", | |
| "drug design", | |
| "common sense", | |
| "\u53e4\u6587", | |
| "gene", | |
| "Recognition", | |
| "WSJ02Mix", | |
| "xvectors", | |
| "math learning", | |
| "\u00daFAL", | |
| "biolinkbert", | |
| "camembert-base", | |
| "PyABSA", | |
| "cyclegan", | |
| "RE", | |
| "entity mention detection", | |
| "EMD", | |
| "financial-emotion-analysis", | |
| "extra_trees", | |
| "hupd", | |
| "synthesis", | |
| "coqui.ai", | |
| "Grapheme-to-Phoneme", | |
| "vehicle-detection", | |
| "DeBERTa", | |
| "unified model", | |
| "SpeechT5", | |
| "Voice Conversion", | |
| "grammar-correction", | |
| "switch_transformers", | |
| "waifu-diffusion", | |
| "ai4bharat", | |
| "\ud55c\uad6d\uc5b4", | |
| "wildcard", | |
| "synthwave", | |
| "kurrent", | |
| "swin2sr", | |
| "stable-Diffusion", | |
| "AraBERT", | |
| "safety-checker", | |
| "sygil-devs", | |
| "protogen", | |
| "disco-diffusion", | |
| "graphormer", | |
| "fp16", | |
| "document summary", | |
| "Woman", | |
| "text2img", | |
| "SD", | |
| "Lora", | |
| " stable-diffusion", | |
| "Grapefruit", | |
| "reward-model", | |
| "hifigan", | |
| "hate_speech", | |
| "Document Question Answering", | |
| "truecasing", | |
| "InvertedPendulum-v2", | |
| "veterinary", | |
| "OpenNiji", | |
| "Niji", | |
| "Stylised", | |
| "NorBERT", | |
| "webui", | |
| "true-casing", | |
| "wrime", | |
| "mert_model", | |
| "stable_diffusion", | |
| "pcdet", | |
| "spelling correction", | |
| "Google", | |
| "Amazon", | |
| "gpt_neox_reward_model", | |
| "FunASR", | |
| "moss", | |
| "video-captioning", | |
| "jax-diffusers-event", | |
| "doctor", | |
| "scat", | |
| "textual_inversion", | |
| "real", | |
| "loha", | |
| "albertina-ptpt", | |
| "albertina-ptbr", | |
| "uav", | |
| "mplug-owl", | |
| "chibi", | |
| "lokr", | |
| "traits", | |
| "table detection", | |
| "archive", | |
| "KoLLaVA", | |
| "glaucoma", | |
| "unet", | |
| "reinforcement-learning-from-human-feedback", | |
| "qrcode", | |
| "RVC v2", | |
| "Audio-to-Audio", | |
| "punctuation restoration", | |
| "mmsegmentation", | |
| "Sentinel-2", | |
| "XLSR", | |
| "sentence transformers", | |
| "VITS", | |
| "tensorrt", | |
| "ted-deberta-v2", | |
| "\u751f\u6210", | |
| "blacklight", | |
| "neon", | |
| "absolute-realism", | |
| "automatic-audio-captioning", | |
| "audio-based-storytelling", | |
| "speech-audio-coreasoning", | |
| "auditory understanding", | |
| "seamless_m4t", | |
| "speech-generation", | |
| "hack", | |
| "D&D", | |
| "portraits", | |
| "42dot_llm", | |
| "cotracker", | |
| "entity-recognition", | |
| "printed", | |
| "rock", | |
| "fauna", | |
| "flora", | |
| "startups", | |
| "passage-reranking", | |
| "flan-t5", | |
| "plamo", | |
| "InternLMXComposer", | |
| "gptneo", | |
| "ecology", | |
| "mixformer-sequential", | |
| "Voice2Voice", | |
| "leolm", | |
| "metaclip", | |
| "instance segmentation", | |
| "retnet", | |
| "pickle", | |
| "GO", | |
| "CHECKPOINT", | |
| "financial sentiment analysis", | |
| "Entity Linking", | |
| "SD1.5", | |
| "spherical linear interpolation merge", | |
| "zephyr", | |
| "naberius", | |
| "FLOR", | |
| "roboflow-universe", | |
| "human-detection", | |
| "OpenCLIP", | |
| "reco", | |
| "cards", | |
| "sdxl style lora", | |
| "programming-language", | |
| "Yi-34B-200K", | |
| "language-detection", | |
| "comic book", | |
| "sticker", | |
| "spooky", | |
| "oil painting", | |
| "stars", | |
| "planets", | |
| "colorful", | |
| "watercolor", | |
| "digital painting", | |
| "deepvision", | |
| "semirealistic", | |
| "impressionism", | |
| "abstract", | |
| "miniatures", | |
| "floods", | |
| "femr", | |
| "Medicine", | |
| "FinBERT", | |
| "FinTwitBERT", | |
| "financial-analysis", | |
| "water", | |
| "foreground", | |
| "japan", | |
| "weather-forecasting", | |
| "solar-ko", | |
| "African", | |
| "Chaeetah", | |
| "dcase-challenge", | |
| "vision language", | |
| "ps1 style", | |
| "llm-agent", | |
| "openchat/openchat-3.5-1210", | |
| "defect detection", | |
| "electrical engineering", | |
| "invest", | |
| "telechat", | |
| "slimsam", | |
| "medical-imaging", | |
| "genome", | |
| "udkai/Garrulus", | |
| "Deutsch", | |
| "Clustering", | |
| "Diarisation", | |
| "liminerity/Blur-7b-v1.21", | |
| "Solar Moe", | |
| "222gate/BrurryDog-7b-v0.1", | |
| "SanjiWatsuki/Lelantos-DPO-7B", | |
| "uonlp", | |
| "Viet-Mistral", | |
| "internlmxcomposer2", | |
| "distil-whisper", | |
| "burmese", | |
| "zysec.ai", | |
| "malware analysis", | |
| "exploitdev", | |
| "ai4good", | |
| "aisecurity", | |
| "mlabonne/DareBeagle-7B-v2", | |
| "\u8349\u8599\u5be7\u3005", | |
| "pjsk", | |
| "\u8349\u8599\u5b81\u5b81", | |
| "\u30d7\u30ed\u30bb\u30ab", | |
| "prsk", | |
| "\u5b81\u5b81", | |
| "image restoration", | |
| "pose-estimation", | |
| "obb", | |
| "yolov3", | |
| "yolov9", | |
| "aerial imagery", | |
| "watermark", | |
| "Aditya685/upshot-sih", | |
| "ponyxl", | |
| "safetyguard", | |
| "komodo", | |
| "speech separation", | |
| "trendyol", | |
| "jondurbin/bagel-dpo-34b-v0.2", | |
| "abacusai/MetaMath-Bagel-DPO-34B", | |
| "tamil", | |
| "common-canvas", | |
| "Voyage", | |
| "RJuro/munin-neuralbeagle-7b", | |
| "sparsetral", | |
| "Eric111/AlphaMayo", | |
| "paulml/DPOB-INMTOB-7B", | |
| "LORA", | |
| "Chatgpt", | |
| "liminerity/Neurotic-Momningojask-slerp", | |
| "liminerity/Omningotex-7b-slerp", | |
| "backbone", | |
| "liminerity/1", | |
| "Inv/Konstanta-7B", | |
| "allstax/AI-G-Expander-v6", | |
| "voice-cloning", | |
| "DAC", | |
| "ui", | |
| "Kukedlc/NeuralMarioMonarch-7B-slerp", | |
| "kviai", | |
| "Image-to-Text Retrieval", | |
| "Text-To-Image Retrieval", | |
| "CultriX/NeuralTrixlaser-bf16", | |
| "indic", | |
| "vendor-names", | |
| "zip-codes", | |
| "rl", | |
| "eren23/ogno-monarch-jaskier-merge-7b-OH-PREF-DPO-v3", | |
| "Kukedlc/NeuralExperiment-7b-MagicCoder-v7", | |
| "ser", | |
| "msp-podcast", | |
| "Corianas/Neural-Mistral-7B", | |
| "Text-To-Image", | |
| "MediaTek-Research/Breeze-7B-Instruct-v1_0", | |
| "ammarali32/multi_verse_model", | |
| "berturk", | |
| "lstm", | |
| "BioMistral/BioMistral-7B-DARE", | |
| "meowgpt", | |
| "splittic", | |
| "samir-fama/FernandoGPT-v1", | |
| "mahiatlinux/MasherAI-7B-v4", | |
| "mohsinmubaraksk/Beast-Mixed", | |
| "image-synthesis", | |
| "wavelet-transform", | |
| "mvpmaster/PearlMath-lafted-7b-slerp", | |
| "Locutusque/NeuralHyperion-2.0-Mistral-7B", | |
| "mvpmaster/NeuralMaths-lafted-7b-slerp", | |
| "mvpmaster/nddmp-kellemar-KrishnaHercules-7b-slerp", | |
| "mossformer2", | |
| "eldogbbhed/NeuralMonarchCoderPearlBeagle", | |
| "Gille/StrangeMerges_42-7B-dare_ties", | |
| "swallow", | |
| "dhyay/mistral_codev2_3k", | |
| "Kukedlc/Neural-Krishna-Multiverse-7b-v3", | |
| "Kukedlc/NeuTrixOmniBe-DPO", | |
| "MatthieuJ/Jason1903_SLERP", | |
| "Gille/StrangeMerges_30-7B-slerp", | |
| "mlabonne/UltraMerge-7B", | |
| "Smuggling1710/WestLakev2-IreneRP-Neural-7B-slerp", | |
| "Nitral-AI/Eris_PrimeV3.075-Vision-7B", | |
| "ml-4m", | |
| "bobofrut/ladybird-base-7B-v8", | |
| "Q-bert/MetaMath-Cybertron-Starling", | |
| "ozayezerceli/BetterSaul-7B-slerp", | |
| "kettleguts/zephyr-7b-beta_sparse05", | |
| "kaist-ai/mistral-orpo-beta", | |
| "Film", | |
| "Photo", | |
| "wildzzz", | |
| "agritech", | |
| "mplug_docowl", | |
| "llm_content_detection", | |
| "AI_detection", | |
| "AI-Sweden-Models/tyr", | |
| "neph1/bellman-7b-mistral-instruct-v0.2", | |
| "matcha-tts", | |
| "acoustic modelling", | |
| "multispeaker", | |
| "medialbertina-ptpt", | |
| "european portuguese", | |
| "ESCO", | |
| "occupation coding", | |
| "kz", | |
| "MaziyarPanahi/Calme-7B-Instruct-v0.2", | |
| "allknowingroger/LadybirdGonzo-7B-slerp", | |
| "Ksgk-fy/M7Percival_010.14-0.33-0.6-0.72-0.02-0.65-7B", | |
| "persona", | |
| "foredoomed", | |
| "passthrough_merge", | |
| "starling", | |
| "erebus", | |
| "cockatrice", | |
| "holodeck", | |
| "limarp", | |
| "koboldai", | |
| "Clickbait", | |
| "musilingo", | |
| "young children", | |
| "MaziyarPanahi/openchat_3.5-16k-Mistral-7B-Instruct-v0.2-slerp", | |
| "fterry/FofoNet-DolphinChat-slerp", | |
| "vgorce/MarcoroNeuralChat-7B-slerp", | |
| "img2latex", | |
| "openba", | |
| "allknowingroger/AutoLimmy-7B-slerp", | |
| "allknowingroger/StarlingDolphin-7B-slerp", | |
| "recurrent_gemma", | |
| "PixArt-\u03a3", | |
| "FPHam/Karen_TheEditor_V2_STRICT_Mistral_7B", | |
| "allknowingroger/RasGullaINEX12-7B-slerp", | |
| "cstr/Spaetzle-v61-7b", | |
| "cstr/Spaetzle-v53-7b", | |
| "AlekseiPravdin/KSI-RP-NSK-128k-7B", | |
| "grimjim/fireblossom-32K-7B", | |
| "two stage dpo", | |
| "phobert", | |
| "llm-security", | |
| "llama38b", | |
| "NorskGPT", | |
| "Phi-3-mini", | |
| "video LLM", | |
| "storm", | |
| "Human Preference Alignment", | |
| "PEFT", | |
| "pythorch", | |
| "llamantino", | |
| "bunny-phi3", | |
| "Female", | |
| "ViT", | |
| "Cpp", | |
| "JS", | |
| "MySql", | |
| "imputation", | |
| "Academic", | |
| "Papers", | |
| "Arxiv", | |
| "japanese input", | |
| "kana kanji conversion", | |
| "enko", | |
| "openpose", | |
| "gender-classification", | |
| "Theta Scaling", | |
| "Big Patents", | |
| "RoPE", | |
| "sustainability", | |
| "governance", | |
| "bf16", | |
| "ssd-1b", | |
| "Stock Market", | |
| "local", | |
| "float32", | |
| "space whale", | |
| "speaker-separation", | |
| "speech-separation", | |
| "rerank", | |
| "ibm-granite-code", | |
| "tool-using", | |
| "latex", | |
| "telegram", | |
| "drug discovery", | |
| "mdlm", | |
| "TCM", | |
| "proprime", | |
| "videollama2_mistral", | |
| "value alignment", | |
| "material", | |
| "roberta-large", | |
| "Mamba", | |
| "Mamba-2", | |
| "SSM", | |
| "Emotionally Intelligent", | |
| "classifier", | |
| "tool-calling", | |
| "agentic", | |
| "personal", | |
| "Sa\u011fl\u0131k", | |
| "tensorart", | |
| "violence-detection", | |
| "apple", | |
| "ane", | |
| "gta", | |
| "gta6", | |
| "online video understanding", | |
| "Time-series", | |
| "PyTorch, Transformers", | |
| "Yuna AI", | |
| "EliTA Enhanced", | |
| "companion", | |
| "girlfriend", | |
| "dust3r", | |
| "Automated Peer Reviewing", | |
| "OG_finetune_merge", | |
| "cosplay", | |
| "qwen-2", | |
| "cogvlm--video", | |
| "Lynx", | |
| "llmware-rag", | |
| "lineart", | |
| "set-encoder", | |
| "semiconductor", | |
| "Data Protection", | |
| "Phi-2", | |
| "sciences", | |
| "bunny", | |
| "hebrew", | |
| "masked-language-model", | |
| "Knesset", | |
| "composed image retrieval", | |
| "image retrieval", | |
| "acmmm2024", | |
| "Mistral_Star", | |
| "Mistral_Quiet", | |
| "Token-Classification", | |
| "SpydazWeb-AI", | |
| "punjabi shahmukhi", | |
| "punjabi gurmukhi", | |
| "punjabi transliteration", | |
| "punjabi gur to shahmukhi", | |
| "punjabi transliteration system", | |
| "Hyper-SDXL 8steps", | |
| "multi-step merge", | |
| "emotion classification", | |
| "tabular data", | |
| "py-feat", | |
| "product", | |
| "nl2bash", | |
| "make-a-shape", | |
| "vision transformer", | |
| "spam-detection", | |
| "Medusa", | |
| "Speculative Decoding", | |
| "shape_opt", | |
| "Image-generation", | |
| "pytoch", | |
| "semantic parsing", | |
| "Generative AI", | |
| "3dtopia-xl", | |
| "backgrounds", | |
| "collaborative filtering", | |
| "Tanuki", | |
| "instruction following", | |
| "nostalgia", | |
| "story telling", | |
| "ultra high precision", | |
| "Bilingual", | |
| "animagine", | |
| "skin", | |
| "natural language", | |
| "Meta", | |
| "taivisionlm", | |
| "rene", | |
| "cartesia", | |
| "Seq2Seq", | |
| "cats", | |
| "trigger", | |
| "covid", | |
| "font", | |
| "the protoart", | |
| "pencil", | |
| "chatqa-2", | |
| "control-vector", | |
| "qwen2vlm", | |
| "VLLM", | |
| "gpt-neo-x", | |
| "llmware-encoder", | |
| "vit_b", | |
| "stable-diffusion-xl-lightning", | |
| "p7", | |
| "llava_jamba", | |
| "contracts", | |
| "Ultrachat", | |
| "motion planning", | |
| "flux schnell", | |
| "low-resource-languages", | |
| "OneGen", | |
| "prototype", | |
| "llama1", | |
| "esthetic", | |
| "fintech", | |
| "gLM2", | |
| "optical-flow-estimation", | |
| "cleverboi", | |
| "Conversational AI", | |
| "Mistral-7B", | |
| "cogvideox-fun", | |
| "alibaba-pai", | |
| "Weather & Climate", | |
| "Web", | |
| "llava_cohere", | |
| "openCLIP", | |
| "sami", | |
| "nepali", | |
| "politica", | |
| "music llama", | |
| "TextGeneration", | |
| "eeg", | |
| "image-processing", | |
| "neuroscience", | |
| "genz", | |
| "NVLM_D", | |
| "NVLM", | |
| "Abliterated", | |
| "llamaify", | |
| "esper-2", | |
| "1b", | |
| "cogvideox-diffusers", | |
| "fire", | |
| "tost", | |
| "ComfyUI", | |
| "Animation", | |
| "wala", | |
| "genshin-impact", | |
| "hoyoverse", | |
| "80s", | |
| "1980s", | |
| "sign", | |
| "light", | |
| "llmware-vision", | |
| "colonoscopy", | |
| "polyp", | |
| "ethereal", | |
| "occult", | |
| "vintage", | |
| "COCOM", | |
| "glif", | |
| "hi-res", | |
| "yolov11", | |
| "p2", | |
| "vjepa", | |
| "medtech", | |
| "HealthGPT", | |
| "Star-Trek", | |
| "128k-Context", | |
| "drugdiscovery", | |
| "llama3.2 1B", | |
| "llama3.1 8B", | |
| "llama3.2 3B", | |
| "celeb", | |
| "code-llama", | |
| "hf_aigcodexmoe", | |
| "actor", | |
| "diff_llama", | |
| "200K", | |
| "Tool-Calling", | |
| "tango", | |
| "Non-Autoregressive", | |
| "Fake News", | |
| "Diffusion Transformer", | |
| "Image Editing", | |
| "ACE", | |
| "llama-3.1-instruct-70b", | |
| "llama-3-instruct-70b", | |
| "rationality", | |
| "advanced", | |
| "southpark", | |
| "vidore-exclude", | |
| "covenants", | |
| "property", | |
| "deed", | |
| "racial-covenant", | |
| "kazakh", | |
| "pop", | |
| "hook", | |
| "music-ai", | |
| "music-transformer", | |
| "Trappu/Nemo-Picaro-12B", | |
| "multi-document", | |
| "AtAndDev/CapybaraMarcoroni-7B", | |
| "eren23/DistilHermes-2.5-Mistral-7B", | |
| "qaic", | |
| "qaicrt", | |
| "scientific", | |
| "serverless", | |
| "mistral-ft-optimized", | |
| "neural-hermes", | |
| "song", | |
| "Kukedlc/neuronal-7b-Mlab", | |
| "phonelm", | |
| "Illusion", | |
| "eren23/dpo-binarized-NeutrixOmnibe-7B", | |
| "fourier", | |
| "task addition", | |
| "Flare", | |
| "kl3m", | |
| "enterprise", | |
| "jaychou", | |
| "Lappland the Decadenza", | |
| "\u8352\u829c\u62c9\u666e\u5170\u5fb7", | |
| "poses", | |
| "OpenCoder", | |
| "Art-Free", | |
| "upper_sorbian", | |
| "kannada", | |
| "diseases", | |
| "Locutusque/StockQwen-2.5-7B", | |
| "allknowingroger/QwenSlerp8-7B", | |
| "full precision", | |
| "quants", | |
| "fm4bio", | |
| "FLUXv1-schnell", | |
| "Pruned", | |
| "link", | |
| "versality", | |
| "stability", | |
| "FC", | |
| "zim-anything", | |
| "Llama-Doctor", | |
| "audio-visual", | |
| "STEM", | |
| "ChatGLM", | |
| "real-world", | |
| "topic-relatedness", | |
| "semantic-relatedness", | |
| "GEMM", | |
| "text_generation", | |
| "conversation_summarization", | |
| "dermatology", | |
| "full-finetune", | |
| "egyptian", | |
| "TEST", | |
| "giux78/zefiro-7b-beta-ITA-v0.1", | |
| "azale-ai/Starstreak-7b-beta", | |
| "gagan3012/Mistral_arabic_dpo", | |
| "davidkim205/komt-mistral-7b-v1", | |
| "OpenBuddy/openbuddy-zephyr-7b-v14.1", | |
| "manishiitg/open-aditi-hi-v1", | |
| "Mixtral 8x7B", | |
| "Virt-io/Erebus-Holodeck-7B", | |
| "video depth estimation", | |
| "wizardmath", | |
| "mique", | |
| "Marco-o1", | |
| "tryon", | |
| "vto", | |
| "T3Q-ko-solar-sft-v3.0", | |
| "kyujinpy/KoCommercial-NoSSL", | |
| "rqwen", | |
| "ehristoforu", | |
| "kphi3", | |
| "llama_enc", | |
| "historical person", | |
| "archival", | |
| "llava_phi3", | |
| "fran\u00e7ais", | |
| "Prompt", | |
| "foundation_model", | |
| "MIDI-to-text", | |
| "MIDI-classification", | |
| "Style", | |
| "Isabelia", | |
| "Character-3D", | |
| "NFT", | |
| "mochi-1-preview-diffusers", | |
| "ShinojiResearch/Senku-70B-Full", | |
| "cognitivecomputations/TinyDolphin-2.8.2-1.1b-laser", | |
| "cognitivecomputations/TinyDolphin-2.8.1-1.1b", | |
| "TinyLlama/TinyLlama-1.1B-intermediate-step-715k-1.5T", | |
| "SVECTOR", | |
| "upstage/SOLAR-10.7B-Instruct-v1.0", | |
| "heavytail/kullm-solar", | |
| "SOLAR", | |
| "MathPILE", | |
| "one-shot", | |
| "smol", | |
| "SLERP", | |
| "pmod_llava_llama", | |
| "ContaLLM", | |
| "ContaAI", | |
| "euclid_qwen2", | |
| "prompt_enhancement", | |
| "short_prompt", | |
| "Prompts", | |
| "simple-math", | |
| "v-pred", | |
| "embedding_space_map", | |
| "colorization", | |
| "vaporwave", | |
| "image_classification", | |
| "align3r", | |
| "Unified Medical Language System", | |
| "National Library of Medicine", | |
| "umls", | |
| "aviation", | |
| "autoround", | |
| "intel-autoround", | |
| "woq", | |
| "intel", | |
| "long-CoT", | |
| "Lyra_Qwen2VL_SpeechGenerator", | |
| "smol-course", | |
| "module_1", | |
| "Minecraft", | |
| "vita-Qwen2", | |
| "anaphora-resolution", | |
| "mentions-linking", | |
| "literary-texts", | |
| "nested-entities", | |
| "BookNLP-fr", | |
| "image2text", | |
| "camera", | |
| "Urban Scenario", | |
| "Synthethic data", | |
| "RLVR", | |
| "Extended-Stream", | |
| "chat-summary", | |
| "Text classification", | |
| "Question answering", | |
| "SER", | |
| "W4A16", | |
| "monolithic", | |
| "wavymulder", | |
| "luau", | |
| "roblox", | |
| "automotive", | |
| "3D Cartoon", | |
| "Objects", | |
| "Game Assets", | |
| "proximasan", | |
| "model merge basket", | |
| "AIGurukul", | |
| "image_captioning", | |
| "gguf-node", | |
| "videorefer_qwen2", | |
| "erotic", | |
| "booru", | |
| "imagebooru", | |
| "imageboard", | |
| "gelbooru", | |
| "guofeng", | |
| "visual", | |
| "LID", | |
| "BERTa", | |
| "Sarcasm", | |
| "Egyptian", | |
| "fill-in-the-blanks", | |
| "superresolution", | |
| "pa-IN", | |
| "Title-Generation", | |
| "nl_BE", | |
| "nl_NL", | |
| "aav", | |
| "aed", | |
| "alv", | |
| "cpf", | |
| "csg", | |
| "csn", | |
| "cus", | |
| "dra", | |
| "euq", | |
| "grk", | |
| "mkh", | |
| "nic", | |
| "pqe", | |
| "sal", | |
| "mfs", | |
| "prl", | |
| "FewCLUE", | |
| "4ulan", | |
| "DPRNNTasNet", | |
| "DPTNet", | |
| "VAD", | |
| "Voice Activity Detection", | |
| "Openslr Multilingual", | |
| "Lithuanian", | |
| "text reranking", | |
| "botxo", | |
| "Certainly", | |
| "sequence classification", | |
| "Songs", | |
| "competition", | |
| "nb-NO", | |
| "code completion", | |
| "translation English Spanish model", | |
| "translation English French model", | |
| "translation English Swedish model", | |
| "translation Spanish Cszech model", | |
| "translation Spanish Deustch model", | |
| "translation Spanish English model", | |
| "translation Spanish French model", | |
| "translation Spanish Italian model", | |
| "translation Spanish Swedish model", | |
| "translation Deustch Cszech model", | |
| "translation French Deustch model", | |
| "semantic", | |
| "text Classification", | |
| "fluency", | |
| "Conditional Generation", | |
| "EManuals", | |
| "customer support", | |
| "Europarl", | |
| "reference_recognizer", | |
| "megatron", | |
| "license", | |
| "sentence-classification", | |
| "scancode", | |
| "license-compliance", | |
| "BJP", | |
| "Congress", | |
| "AAP", | |
| "sentence_transformers", | |
| "bert2bert", | |
| "BioNLP", | |
| "topic classification", | |
| "topic labeling", | |
| "microsoft/deberta-v3-xsmall", | |
| "RNN-T", | |
| "tinybert", | |
| "loodos-bert-base", | |
| "CARN", | |
| "DRLN", | |
| "MDSR", | |
| "MSRN", | |
| "PAN", | |
| "wmt21", | |
| "Slovak", | |
| "f_t5", | |
| "flax", | |
| "indonesian-roberta-base", | |
| "indonesian-roberta-large", | |
| "diff generation", | |
| "update summarization", | |
| "speech to text", | |
| "sentence-order-prediction", | |
| "issue", | |
| "development", | |
| "rm-sursilv", | |
| "fastspeech2", | |
| "google/pegasus-reddit_tifu", | |
| "glycebert", | |
| "autobert", | |
| "rotary position embedding", | |
| "multimodal-entailment", | |
| "question-answer generation", | |
| "polarity", | |
| "minuscule", | |
| "bert-large-portuguese-cased", | |
| "al", | |
| "ELECTRA", | |
| "MIT", | |
| "Dialectal Arabic", | |
| "metallurgy", | |
| "spanish gpt2", | |
| "restaurant", | |
| "Long documents", | |
| "bros", | |
| "SequenceClassification", | |
| "glpn", | |
| "deidentification", | |
| "medical notes", | |
| "buy-intent", | |
| "sell-intent", | |
| "consumer-intent", | |
| "multitask-model", | |
| "asr_seq2esq", | |
| "wikibert", | |
| "part of speech tagging", | |
| "semantic textual similarity", | |
| "sts-ca", | |
| "tecla", | |
| "teca", | |
| "askscience", | |
| "qarib60_1790k", | |
| "clinical trial", | |
| "Universal Sentence Encoder", | |
| "indonesian-roberta-base-sentiment-classifier", | |
| "salesken", | |
| "passage-ranking", | |
| "neural-search-query-classification", | |
| "neural-search", | |
| "autocomplete", | |
| "Keywords", | |
| "Keyword Spotting", | |
| "Command Recognition", | |
| "question_generation", | |
| "named-entities", | |
| "BIO", | |
| "counterspeech", | |
| "irt", | |
| "awesome", | |
| "cog_view", | |
| "retribert", | |
| "vilbert", | |
| "roberta-wwm", | |
| "text2text_generation", | |
| "microscopy", | |
| "ubertext", | |
| "Darknet", | |
| "emoberta", | |
| "wikibio", | |
| "MEDIA", | |
| "shared_bart", | |
| "bulgarian", | |
| "macedonian", | |
| "hate-speech-classification", | |
| "suicidio", | |
| "kgqa", | |
| "STILT", | |
| "retraining", | |
| "multi-task learning", | |
| "anglicisms", | |
| "loanwords", | |
| "borrowing", | |
| "multi_class_classification", | |
| "crisis", | |
| "Fake News Detection", | |
| "QuantizationAwareTraining", | |
| "eHR", | |
| "detectron2", | |
| "/workspace/datasets/datasets/MIR_ST500/MIR_ST500.py", | |
| "mutlimodal", | |
| "3DBall", | |
| "floret", | |
| "integration", | |
| "translation evaluation", | |
| "sentence-summarization", | |
| "recall", | |
| "language models", | |
| "fleurs-asr", | |
| "speech-enhancement-recognition", | |
| "geezlab", | |
| "bertweet", | |
| "discourse-marker-prediction", | |
| "pragmatics", | |
| "discourse", | |
| "biobert", | |
| "Indonesian", | |
| "pix2seq", | |
| "transfer-learning", | |
| "covid19", | |
| "capitalization", | |
| "forward-looking-statement", | |
| "text segmentation", | |
| "deepflash2", | |
| "ALE/SpaceInvaders-v5", | |
| "custom-license", | |
| "auto-complete", | |
| "generative qa", | |
| "Taxi-v3-4x4-no_slippery", | |
| "3D Object Detection", | |
| "MusicGeneration", | |
| "hf_diffuse", | |
| "climate change", | |
| "gpt-j", | |
| "gpt-j-6b", | |
| "data-augmentation", | |
| "structured-data", | |
| "tabular-data", | |
| "arabert", | |
| "idt5", | |
| "ldmbert", | |
| "consistency-regularization", | |
| "Clinical notes", | |
| "Discharge summaries", | |
| "mbert2mbert", | |
| "face-stylization", | |
| "Arat5-base", | |
| "targeted-drug-design", | |
| "gen_ffa", | |
| "PyTorch Lightning", | |
| "Image Translation", | |
| "binary-classification", | |
| "seals/CartPole-v0", | |
| "ML-Agents-Walker", | |
| "mobilenet_v1", | |
| "Trinidadian Creole", | |
| "multiclass-classification", | |
| "TyDiQA", | |
| "SQuAD 1.1", | |
| "text style transfer", | |
| "graph_nystromformer", | |
| "covid-19", | |
| "DialoGPT", | |
| "ShiftVit", | |
| "character-level", | |
| "t5-large-summarization", | |
| "Video Transformers", | |
| "medical documents", | |
| "biomedical papers", | |
| "MarianNMT", | |
| "BERTovski", | |
| "MaltBERTa", | |
| "hungarian", | |
| "spoken-language-understanding", | |
| "Natural Questions List", | |
| "text summarization", | |
| "chinese poem", | |
| "visual-novel", | |
| "answer-aware-question-generation", | |
| "image matting", | |
| "neuspell", | |
| "spell-correction", | |
| "factuality", | |
| "universal", | |
| "multiset", | |
| "internet-culture", | |
| "art_theory", | |
| "manage", | |
| "basic_medical_science", | |
| "pharmacy", | |
| "public_health", | |
| "energy_and_power", | |
| "mechanical_engineering", | |
| "tool", | |
| "mindfulness", | |
| "android-apps", | |
| "Replete-AI", | |
| "benchmarks", | |
| "spatial-reasoning", | |
| "aspect-based-summarization", | |
| "simplification-evaluation", | |
| "word-tokenization", | |
| "digital-humanities-research", | |
| "judgement-prediction", | |
| "compositionality", | |
| "concepts-to-text", | |
| "medical-personal-protective-equipment-detection", | |
| "dialog-act-classification", | |
| "rationale-extraction", | |
| "legal-topic-classification", | |
| "intent-to-text", | |
| "disambiguation", | |
| "reddit-posts-summarization", | |
| "text-search", | |
| "movielens", | |
| "dialogue-summary", | |
| "one-liner-summary", | |
| "meeting-title", | |
| "email-subject", | |
| "named-entity-recognition-and-classification", | |
| "1800-1900", | |
| "offensive-language-classification", | |
| "conversational-question-answering", | |
| "story-generation", | |
| "robust-speech-recognition", | |
| "news, mafand, masakhane", | |
| "wallstreet", | |
| "hedgefunds", | |
| "text-recognition", | |
| "topics", | |
| "poems", | |
| "scholarly text", | |
| "DAO", | |
| "legalnlp", | |
| "class action", | |
| "dialogue safety", | |
| "social norm", | |
| "rules-of-thumb", | |
| "technical questions", | |
| "texture-synthesis", | |
| "non-infringing", | |
| "doc retrieval", | |
| "Utilities", | |
| "Ppe", | |
| "Assembly Line", | |
| "Warehouse", | |
| "Factory", | |
| "econlp", | |
| "manufacturing", | |
| "Transportation", | |
| "Documents", | |
| "cross domain", | |
| "metaphor", | |
| "figurative language", | |
| "earth-science", | |
| "vocabulary", | |
| "natural-science", | |
| "comments", | |
| "forum", | |
| "flan 2022", | |
| "flan v2", | |
| "alexa", | |
| "Latin", | |
| "gut-microbiota", | |
| "project gutenberg", | |
| "e-book", | |
| "gutenberg.org", | |
| "fine-grained ner", | |
| "multi-humaneval", | |
| "humaneval", | |
| "subtitles", | |
| "Anomaly Detection", | |
| "common", | |
| "multi30k", | |
| "Out Of Distribution", | |
| "video games", | |
| "gamedev", | |
| "databricks", | |
| "research abstracts", | |
| "AbeShinzo", | |
| "FormerJapanesePrimeMinister", | |
| "applescript", | |
| "medical consultation", | |
| "news-topic", | |
| "masakhanews", | |
| "masakhane", | |
| "words", | |
| "casual", | |
| "MARC", | |
| "CommonsenseQA", | |
| "as2", | |
| "answer sentence selection", | |
| "air traffic management", | |
| "atm", | |
| "kaggle", | |
| "scientific paper", | |
| "atcosim", | |
| "social science", | |
| "large language modeling", | |
| "inverse rendering", | |
| "material decomposition", | |
| "atc", | |
| "deepfakes", | |
| "LAION", | |
| "2023", | |
| "islam", | |
| "image-text-bounding-box pairs", | |
| "egypt", | |
| "text-image-matching", | |
| "languages", | |
| "syllable", | |
| "Affective Captioning", | |
| "Emotions", | |
| "commit", | |
| "patch", | |
| "self-driving", | |
| "robotics navigation", | |
| "semantic segmentation", | |
| "procedural-generation", | |
| "Evaluation", | |
| "answer", | |
| "customer-support", | |
| "subset", | |
| "airoboros", | |
| "math world problems", | |
| "Network Security", | |
| "harm", | |
| "damage assessment", | |
| "Podcast", | |
| "SARFish", | |
| "Illegal Fishing", | |
| "Complex-Valued", | |
| "Synthetic Aperture Radar", | |
| "Colorectal Pancer", | |
| "webpages", | |
| "parlament", | |
| "fraud-detection", | |
| "deception-detection", | |
| "opinion-spam", | |
| "owl", | |
| "Persian", | |
| "religion", | |
| "storybook", | |
| "textbook-qa", | |
| "document-image", | |
| "explainability", | |
| "oriented-bounding-boxes", | |
| "SVG", | |
| "vector", | |
| "Robotics", | |
| "flytech", | |
| "behavioral", | |
| "codegeneration", | |
| "layout-generation", | |
| "hackernoon", | |
| "\u65e5\u672c\u8a9e", | |
| "audio-text", | |
| "medi", | |
| "human_feedback", | |
| "regex", | |
| "pronunciation-scoring", | |
| "helpsteer", | |
| "ai-feedback", | |
| "visual-chat", | |
| "multimodal-chat", | |
| "graphics", | |
| "Morocco", | |
| "scene-classification", | |
| "brain tumor", | |
| "tumor", | |
| "instruct-tuning", | |
| "3D Gaussian Splatting", | |
| "passports", | |
| "3D vision", | |
| "nifty", | |
| "stock-movement", | |
| "news-and-events", | |
| "RLMF", | |
| "video-summarization", | |
| "tajik", | |
| "names", | |
| "representation learning", | |
| "Code Efficiency", | |
| "heritage", | |
| "novelai", | |
| "knowledge-graphs", | |
| "document-type objects", | |
| "smart contracts", | |
| "singing voice", | |
| "entity extraction", | |
| "natural disasters", | |
| "MMLU", | |
| "multi-spectral", | |
| "SKG", | |
| "LLM Safety", | |
| "Multiple Choice", | |
| "language-learning", | |
| "caselaw", | |
| "ICL", | |
| "patient summary", | |
| "Forest", | |
| "children", | |
| "counseling", | |
| "breast cancer", | |
| "coding preferences", | |
| "llm-as-a-judge", | |
| "leetcode", | |
| "automatic_speech_recognition", | |
| "Caption", | |
| "technology", | |
| "simple", | |
| "\u0641\u0627\u0631\u0633\u06cc", | |
| "instrument", | |
| "leyes", | |
| "peru", | |
| "constitucion", | |
| "system prompt", | |
| "speaker diarization", | |
| "opus", | |
| "anthropic", | |
| "igbo", | |
| "mediabias", | |
| "media-bias", | |
| "product search", | |
| "Knowledge", | |
| "3D medical", | |
| "document-level", | |
| "toxicity detection", | |
| "clinical medicine", | |
| "my little pony", | |
| "pony preservation project", | |
| "singing", | |
| "asl", | |
| "palestine", | |
| "CEFR", | |
| "epigenetics", | |
| "life sciences", | |
| "data", | |
| "disaster", | |
| "distill", | |
| "Probability", | |
| "Liner Algebra", | |
| "Algebra", | |
| "Differential Equations", | |
| "Calculus", | |
| "over-alignment", | |
| "freebase", | |
| "IT", | |
| "Web Scraping", | |
| "Media", | |
| "autonomous_driving", | |
| "hallucinations", | |
| "indian", | |
| "Instruction Tuning", | |
| "image-to-4d", | |
| "esoterism", | |
| "VisualQA", | |
| "time series analysis", | |
| "IR", | |
| "Artificial Intelligence", | |
| "probability", | |
| "neuron", | |
| "posts", | |
| "information", | |
| "slot filling", | |
| "temporal-reasoning", | |
| "Role-Playing", | |
| "cti", | |
| "Small Chunks", | |
| "Scientific", | |
| "Scientific Wikipedia", | |
| "Science Wikipedia", | |
| "Emotional Intelligence", | |
| "engineering design", | |
| "T\u00fcrk\u00e7e", | |
| "keypoints", | |
| "pairwise", | |
| "Information Retrieval", | |
| "tcm", | |
| "mscoco", | |
| "relation-classification", | |
| "multimodality", | |
| "photographs", | |
| "image-data", | |
| "image-caption pairs", | |
| "personalization", | |
| "music images", | |
| "group", | |
| "robustness", | |
| "fstar", | |
| "popai", | |
| "fruit", | |
| "transcript", | |
| "answers", | |
| "docker", | |
| "medieval", | |
| "constitution", | |
| "fineweb", | |
| "time", | |
| "home-assistant", | |
| "synthethic", | |
| "extraction", | |
| "ibeta", | |
| "InternVL2", | |
| "human red teaming", | |
| "Bioinformatics", | |
| "truthfulqa", | |
| "azerbaijan", | |
| "Human-Language", | |
| "Customer", | |
| "MIDI images", | |
| "MIDI music", | |
| "tv", | |
| "aya-23", | |
| "command-r", | |
| "cloth", | |
| "fraud detection", | |
| "liveness detection", | |
| "smart meter", | |
| "Human Caption", | |
| "Face Caption", | |
| "surgery", | |
| "vehicles", | |
| "cars", | |
| "ilsvrc-2012", | |
| "celestia", | |
| "dicom", | |
| "cxr", | |
| "Software", | |
| "Frameworks", | |
| "wikimedia", | |
| "stockfish", | |
| "event-detection", | |
| "ARC", | |
| "ptsd", | |
| "veterans", | |
| "buddhism", | |
| "dharma", | |
| "exploits", | |
| "canon", | |
| "Sports", | |
| "multi-agent", | |
| "diverse", | |
| "Max_length = 8180", | |
| "truthfulness", | |
| "debiasing", | |
| "bias-detection", | |
| "LLaMA-3.2b", | |
| "domain-specific", | |
| "rosbag", | |
| "ros2", | |
| "vehicle", | |
| "canbus", | |
| "autonomous_vehicles", | |
| "gesture", | |
| "fastdata", | |
| "Gemini", | |
| "civil engineering", | |
| "gifteval", | |
| "logical", | |
| "DIGIT", | |
| "slip detection", | |
| "YOLO11", | |
| "vqasynth", | |
| "Dialog", | |
| "fandom", | |
| "songs", | |
| "Ancient Greek", | |
| "postocr", | |
| "nom-script", | |
| "visualization", | |
| "ir", | |
| "topic-modeling", | |
| "context", | |
| "dialects", | |
| "overlai.ai", | |
| "c2pa", | |
| "contentcredentials", | |
| "opt-in", | |
| "podcasts", | |
| "Prediction", | |
| "definition", | |
| "ethical dilemmas", | |
| "breast-cancer", | |
| "causal", | |
| "data-mining", | |
| "parallel-corpus", | |
| "ct scans", | |
| "xray", | |
| "mammography", | |
| "radar", | |
| "formal-methods", | |
| "coq", | |
| "fineweb2", | |
| "sign language", | |
| "multi-lingual", | |
| "Indonesion", | |
| "Bahasa", | |
| "smart city", | |
| "genetics", | |
| "quiz", | |
| "united-states", | |
| "real-time", | |
| "current-events", | |
| "political-discourse", | |
| "media-analysis", | |
| "Gujarati", | |
| "blogs", | |
| "adversarial attack", | |
| "NSFW benchmark", | |
| "fanfiction", | |
| "bias-evaluation", | |
| "dialogue-act-classification", | |
| "text-to-structured", | |
| "explanations-in-question-answering", | |
| "stereotype-detection", | |
| "synonyms", | |
| "mobility", | |
| "auto-generated", | |
| "other-keyword-spotting", | |
| "stuctured-to-text", | |
| "Enzyme Commission", | |
| "semantic similarity", | |
| "grug", | |
| "greentext", | |
| "code-mixing", | |
| "mgt", | |
| "magic-card-game", | |
| "LAM", | |
| "middle english", | |
| "dialogue-response-generation", | |
| "metaphor-frame-classification", | |
| "consumer", | |
| "consumer goods", | |
| "jobs", | |
| "yandex", | |
| "ICD-10", | |
| "wordnet", | |
| "quotes", | |
| "ordinal", | |
| "cloze", | |
| "toloka", | |
| "knowledge base", | |
| "camera trap data", | |
| "wildlife monitoring", | |
| "sentence", | |
| "ms-coco", | |
| "gpt-3", | |
| "bills", | |
| "semantic-parsing", | |
| "K-MHaS", | |
| "Korean NLP", | |
| "Coling2022", | |
| "czech NER", | |
| "riddles", | |
| "children's speech", | |
| "icelandic children", | |
| "icelandic kids", | |
| "gigafida", | |
| "samr\u00f3mur", | |
| "bible", | |
| "National Corpus of Polish", | |
| "Narodowy Korpus J\u0119zyka Polskiego", | |
| "rating prediction", | |
| "central kurdish", | |
| "kurdi", | |
| "fluentui", | |
| "images ", | |
| "Aviation", | |
| "humanitarian", | |
| "analytical-framework", | |
| "humset", | |
| "humbert", | |
| "instagram", | |
| "gradio-theme", | |
| "price", | |
| "Health", | |
| "podcast", | |
| "Relation Classification", | |
| "Relation extraction", | |
| "Scientific papers", | |
| "Research papers", | |
| "20newsgroups", | |
| "lam ", | |
| "model cards", | |
| "WIP", | |
| "instructor embeddings", | |
| "vector stores", | |
| "comics", | |
| "duskfallcrew", | |
| "sentence simplification", | |
| "document simplification", | |
| "explainability-judgment-prediction", | |
| "joke", | |
| "virtual assistant", | |
| "compas", | |
| "ned", | |
| "mbxp", | |
| "celebFaces attributes", | |
| "SugaYoshihide", | |
| "\u83c5\u7fa9\u5049", | |
| "consentful", | |
| "instruction generation", | |
| "breast", | |
| "psycholinguistics", | |
| "Nucleotide", | |
| "unsupervised", | |
| "SketchyCOCO", | |
| "musk", | |
| "teaser", | |
| "news snippets", | |
| "headline generation", | |
| "teaser generation", | |
| "keyword generation", | |
| "tweet generation", | |
| "news snippet generation", | |
| "Image Segmentation", | |
| "hackernews", | |
| "enriched", | |
| "cross-lingual", | |
| "haskell", | |
| "script", | |
| "contextual-mt", | |
| "document-mt", | |
| "sotu", | |
| "CN", | |
| "gpt35-alpha", | |
| "docugami", | |
| "dfm-csl", | |
| "xml-knowledge-graphs", | |
| "heterogeneous sources", | |
| "conversational recommendation", | |
| "Machine Translation", | |
| "butterfly", | |
| "dorsal", | |
| "imbalanced", | |
| "mimicry", | |
| "coaching", | |
| "webcams", | |
| "indoor", | |
| "Logistic", | |
| "Spider", | |
| "Hello", | |
| "spider-eval", | |
| "synthetic speech", | |
| "Timex", | |
| "Timexs", | |
| "Temporal Expression", | |
| "Temporal Expressions", | |
| "Temporal Information", | |
| "Timex Identification", | |
| "Timex Classification", | |
| "Timex Extraction", | |
| "minigpt4", | |
| "common_sense", | |
| "political", | |
| "bundestag", | |
| "narrated", | |
| "stars-classification", | |
| "variant-effect-prediction", | |
| "jpg", | |
| "livestream", | |
| "stream", | |
| "messages", | |
| "vtubers", | |
| "corpora", | |
| "audioset", | |
| "fatwa", | |
| "mufti", | |
| "cricket", | |
| "bees", | |
| "coordinates", | |
| "geo-tagged", | |
| "geographic-data", | |
| "random", | |
| "single", | |
| "mixed", | |
| "Discourse", | |
| "Discourse Evaluation", | |
| "brand", | |
| "comparison", | |
| "sim-to-real", | |
| "Bengaluru", | |
| "disparity maps", | |
| "depth dataset", | |
| "ai art", | |
| "semantic web", | |
| "dbpedia", | |
| "NIL", | |
| "\u5cb8\u7530\u6587\u96c4", | |
| "KishidaFumio", | |
| "embryo", | |
| "acting", | |
| "source code", | |
| "code readability", | |
| "autogen", | |
| "web-browsing", | |
| "prodigy", | |
| "clouds", | |
| "bank", | |
| "X", | |
| "dentistry", | |
| "radiation protection", | |
| "EU", | |
| "Celine", | |
| "My Theresa", | |
| "Ounass", | |
| "tweet-classification", | |
| "tone", | |
| "bank-of-ghana", | |
| "exchange-rates", | |
| "ghana data", | |
| "therapy", | |
| "gender identity", | |
| "steganography", | |
| "hpi", | |
| "workshop", | |
| "GEITje", | |
| "cgi", | |
| "jsonifize", | |
| "v1.2.1", | |
| "TimeSeries", | |
| "CanariaView", | |
| "CENIA" | |
| ] |