Brief-details: HelpingAI2.5-5B is an emotionally intelligent language model with 2B parameters, achieving 94.28 on EI tests, optimized for empathetic conversations.
Brief-details: A quantized version of DeepSeek-R1-Distill-Qwen-7B model optimized for efficiency, featuring multiple quantization options from Q2 to F16, with sizes ranging from 2.82GB to 15.24GB
BRIEF-DETAILS: 4-bit quantized version of Qwen2.5-0.5B using Unsloth's Dynamic Quantization, offering 70% less memory usage and 2x faster finetuning capabilities.
Brief-details: A fine-tuned 7B parameter Japanese language model based on Stable LM Base Gamma, optimized for enhanced Japanese language capabilities and performance.
Brief-details: FLAN-T5 Large model fine-tuned on AG News subset, optimized for news classification and text generation tasks using LoRA adaptation
BRIEF-DETAILS: Compressed version of Meta's Llama 3 70B Instruct model using AWQ quantization for improved efficiency while maintaining performance
Brief-details: hakoMay is a Stable Diffusion 2.1 fine-tuning model featuring multiple versions (A/B/C/D/Boy) with custom embeddings, released under FAIPL 1.0-SD license.
Brief Details: 4-bit quantized version of LLaVA-13B multimodal model, optimized for efficient deployment with GPTQ, featuring 128-group size quantization.
Brief Details: A specialized AI model focused on accurately representing Japanese skin tones, created by syaimu and hosted on Hugging Face.
Brief-details: A highly optimized GGUF quantization of Sky-T1-32B model with multiple compression variants, offering flexible deployment options from 9GB to 65GB
Brief-details: Fine-tuned Qwen2.5-14B-Instruct model specialized for role-play scenarios, offering 32K context window and optimized for creative dialogue generation.
Brief-details: Qwen2.5-32B-AGI is a fine-tuned version of Qwen2.5 32B, specifically designed to address hypercensorship issues while maintaining model capabilities.
Brief Details: Lightweight 8B parameter LLaMA3 model variant optimized for GGUF format, requiring only 10GB RAM. Ideal for efficient local deployment with llama.cpp.
Brief-details: GGUF-formatted models using innovative 2-bit quantization approach, optimized for llama.cpp with improved efficiency and reduced size
Brief-details: ct5-small is a compact T5-based language model developed by lemon234071, optimized for efficient text-to-text transformation tasks while maintaining performance.
Brief-details: BERT-based model pre-trained on 152M StackOverflow sentences, specialized for code and named entity recognition in technical discussions
Brief-details: ALBERT base model optimized for Korean language processing, trained on 70GB dataset with 42K subwords vocabulary. Implements efficient BERT-like architecture.
Brief-details: A Tensorflow/Keras implementation of ConvLSTM for video frame prediction, trained on Moving MNIST to predict next frames from previous ones.
BRIEF-DETAILS: Fine-tuned DistilRoBERTa model for reaction GIF classification, achieving 26.62% accuracy. Trained for 3 epochs with Adam optimizer and linear learning rate schedule.
Brief Details: A binary image classification model inspired by Silicon Valley's "hotdog-not-hotdog" app, built using HuggingPics for detecting hotdogs in images.
BRIEF-DETAILS: FastText language identification model based on efficient text classification, capable of detecting languages from text samples with high accuracy
