# Efficient fine-tuning

This model is a dialogue turn detection model fine-tuned based on Qwen3-0.6B. It uses the Unsloth and Huggingface TRL libraries to accelerate training, achieving an accuracy of 96.22% in the dialogue turn detection task.

This is a transformers model published on Hub. The specific functions and detailed information are to be supplemented.

This is a transformers model that has been uploaded to the Hugging Face Hub. Specific information is to be supplemented.

A large language model fine-tuned on a specific dataset based on the ByteDance-Seed/Seed-Coder-8B-Base model, supporting long sequence input and efficient training strategies.

Falcon-E-1B-Base is an efficient 1.58-bit language model developed by TII, featuring a pure Transformer architecture and optimized for edge devices.

This is an Arabic model trained on the MediaSpeech dataset. Specific uses and functionalities require further information for confirmation.

French text generation model fine-tuned based on canopylabs/orpheus-3b-0.1-pretrained

This is a Turkish function calling model fine-tuned based on the Turkish-Llama-8b-DPO-v0.1 model, specifically designed for executing Turkish function calling tasks.

This model is fine-tuned based on Turkish-Llama-8b-DPO-v0.1, specifically designed for Turkish function calling tasks.

Croatian text generation model based on Mistral architecture, trained with Unsloth acceleration

This model is a fine-tuned version of DistilBERT-base-uncased on the CoLA (Corpus of Linguistic Acceptability) dataset, designed for grammatical acceptability judgment tasks.

This is an audio processing model fine-tuned based on facebook/wav2vec2-base, mainly used for speech-related tasks

A distilled model based on the transformers library, developed by DeepSeek-AI through knowledge distillation from the Qwen-14B model

DeepSeek-R1 is the first-generation inference model launched by the DeepSeek team. Through large-scale reinforcement learning training, it does not require supervised fine-tuning (SFT) as an initial step and demonstrates excellent inference capabilities.

SmolMaestra is a lightweight translation model that focuses on text translation between English and Italian and can efficiently and directly output translation results.

Large language model fine-tuned based on BitNet 1.58b architecture, with Llama-3-8B-Instruct as the base model, employing extreme quantization techniques

Mobius is a groundbreaking diffusion model that achieves high-quality cross-style image generation through domain-agnostic debiasing technology.

Yi-1.5 is an upgraded version of the Yi model, excelling in programming, mathematics, reasoning, and instruction-following capabilities while maintaining outstanding language understanding, commonsense reasoning, and reading comprehension abilities.

Phi-3 4k Instruct is a lightweight yet powerful language model, processed with 4-bit quantization to reduce resource requirements.

Llama-3-SauerkrautLM-8b-Instruct is an improved version based on Meta-Llama-3-8B-Instruct jointly developed by VAGO Solutions and Hyperspace.ai. It is optimized through DPO alignment and supports German and English.

A Korean large language model extended from SOLAR-10.7B-v1.0, optimized for Korean understanding through vocabulary expansion and parameter-frozen training

This model is a zero-shot classification model fine-tuned on an economics dataset based on DistilBERT, suitable for classification tasks related to economics texts.

A 32k-context fine-tuned version based on TinyLlama-1.1B, achieving long-context processing capability by increasing rope theta

This is a SetFit model trained on the sst2 dataset for English text classification tasks, utilizing efficient few-shot learning techniques.

TinyLlama-1.1B is a lightweight language model based on the Llama 2 architecture with 1.1B parameters, designed for computation and memory-constrained applications.

CodeFuse-13B is a 13-billion-parameter code generation model trained on the GPT-NeoX framework, supporting over 40 programming languages and capable of processing code sequences up to 4096 characters long.

OpenChat is a series of open-source language models that have been fine-tuned on diverse and high-quality multi-round dialogue datasets, aiming to achieve high performance with limited data.

LERT is a linguistics theory-driven pre-trained language model designed to enhance model performance through linguistic knowledge.

This is a tiny version of the BERT model, case-insensitive, suitable for natural language processing tasks in resource-constrained environments.

A Q&A model fine-tuned on the SQuAD v2 dataset based on DistilRoBERTa-base, featuring lightweight and efficient characteristics.

RoBERTa series models pretrained on datasets of varying scales (1M-1B tokens), including BASE and MED-SMALL specifications

DistilBart-MNLI is a lightweight version distilled from bart-large-mnli using teacher-free distillation technology, maintaining high accuracy while reducing model complexity.

This model is a fine-tuned version of distilbert-base-uncased on an unspecified dataset, primarily used for text-related tasks.

DistilBart-MNLI is a distilled version of BART-large-MNLI, using teacher-free distillation technology, significantly reducing model size while maintaining high performance.

A streamlined version based on google/mt5-base, optimized for Russian and English with 58% fewer parameters

RoBERTa variants pre-trained on datasets ranging from 1M to 1B tokens, including BASE and MED-SMALL specifications, suitable for natural language processing tasks in resource-limited scenarios

DistilRoBERTa is a lightweight distilled version of the RoBERTa model, retaining most of its performance while being smaller and faster.

DistilBart-MNLI is a distilled version of bart-large-mnli using teacher-free distillation techniques, achieving performance close to the original model while being more lightweight.