Brief Details: A 1.5B parameter GPT-2 based model specialized in Python code generation, achieving 3.99% pass@1 on HumanEval benchmark, trained on cleaned CodeParrot dataset.
Brief-details: CryptoBERT - Specialized sentiment analysis model for cryptocurrency social media posts, trained on 3.2M posts with 3-class classification (Bearish/Neutral/Bullish).
Brief-details: StarChat2 15B - Advanced coding assistant with 16B parameters, fine-tuned from StarCoder2. Excels in programming and chat capabilities, supporting 600+ programming languages.
Brief-details: A compact 1.7B parameter instruction-tuned LLM optimized for everyday conversations and general tasks, featuring efficient BF16 precision and Apache 2.0 license.
Brief-details: OLMo-1B is a 1.18B parameter open language model trained on 3T tokens, focusing on research accessibility and reproducibility in LLM science.
Brief Details: CodeNinja-1.0-OpenChat-7B: A 7.24B parameter coding assistant fine-tuned on 400k+ coding instructions, supporting multiple programming languages with 8192 context window.
Brief-details: A powerful 8x7B GPTQ-quantized Mixtral model optimized for coding and general tasks, featuring 6.09B parameters and multiple quantization options for different hardware setups.
BRIEF DETAILS: 7B parameter Hindi/English language model based on Llama2, designed for text generation. Requires fine-tuning for specific tasks.
Brief Details: First open-source Chinese-English bilingual Stable Diffusion model, trained on 20M filtered image-text pairs. Specializes in generating images from Chinese/English prompts using 1B parameters.
Brief-details: CodeQwen1.5-7B-Chat is a powerful code-specialized LLM with 7.25B parameters, supporting 92 programming languages and 64K context length. GGUF-optimized for efficient deployment.
Brief Details: T5-base model fine-tuned on SQuAD for question generation tasks. 297M parameters, Apache 2.0 licensed. Generates questions from answer-context pairs.
Brief Details: A Stable Diffusion model fine-tuned with Dreambooth to replicate Disco Diffusion artistic style. MIT licensed, 105 likes, specialized for creative AI art generation.
Brief Details: ELLA is an innovative text-to-image diffusion model adapter that integrates LLMs to enhance semantic alignment, particularly excelling at dense prompt interpretation.
Brief Details: Qwen1.5-32B-Chat is a powerful 32.5B parameter chat model from the Qwen1.5 series, featuring 32K context length and improved multilingual capabilities.
Brief-details: Advanced language model combining Mistral-7B with iterative DPO alignment and PairRM scoring, achieving impressive Alpaca-Eval 2.0 scores of 30.22 (base) and 34.86 (with post-processing).
BRIEF-DETAILS: Portuguese language model with 6.7B parameters, based on LLaMA architecture. Optimized for Portuguese text generation with strong performance on native benchmarks.
Brief Details: 70B parameter Python-specialized LLM from Meta's Code Llama family. Optimized for code completion and Python development with 16k context window.
BRIEF-DETAILS: HelixNet: A novel 3-LLM architecture using actor-critic-regenerator framework based on Mistral-7B, achieving enhanced response quality through multi-stage refinement.
Brief-details: A specialized ControlNet model for SDXL focused on inpainting/outpainting tasks, using white-masked areas for generation control. Early alpha version with experimental features.
Brief-details: 13B parameter LLaMa-based model trained on synthetic data, achieving 98.087 GPT-3.5 adjusted score. Research-only, non-commercial use.
Brief Details: 4-bit quantized version of WizardLM-7B offering multiple GPTQ parameter options, optimized for GPU inference with 1.13B parameters and SafeTensors support.
