Search FoxAI
AI Models 84999
- SaluNet↗
Mourad-zaied
SaluNet replaces feature normalization with geometric methods, using SALU, SWALU, and GALU modules, to achieve convolutional neural networks and Transformers without normalization.
- turingtest-checkpoints↗
Suqin-Test
This repository contains fine-tuned weights based on Meta Llama 3, Qwen2.5, and Microsoft DeBERTa. Each directory is a separate checkpoint, and does not include training data or experiment logs.
- MyAwesomeModel-TestRepo↗
ASDCX1DXSXA
MyAwesomeModel is a model that has undergone a major version upgrade, significantly enhancing its reasoning and reasoning capabilities, with its overall performance approaching that of other leading models.
- MyAwesomeModel-TestRepo↗
sadD12DSZX
MyAwesomeModel is a model that has undergone a major version upgrade, significantly enhancing its reasoning and reasoning capabilities, with its overall performance approaching that of other leading models.
- Qwen3.8-27B-DiM-heretic↗
richardyoung
This is the uncensored version of Qwen/Qwen3.8-27B, created using Heretic v2.0.0.
- MyAwesomeModel-TestRepo↗
david123TJP
MyAwesomeModel is a model that has undergone a major version upgrade, significantly enhancing its reasoning and reasoning capabilities, with its overall performance approaching that of other leading models.
- novel-reranker-gpl-base-v1↗
liudev
A CrossEncoder model based on hfl/chinese-roberta-wwm-ext, used for text re-ranking and semantic search.
- mtds↗
janniskirschner
This tool is a fine-tuned model released together with a research project, with each model corresponding to a folder, and the folder names are opaque IDs.
- KaiLu3↗
KVCHub
KaiLu3 is a large language model built from scratch by KvcHub, featuring a hybrid architecture that includes 16 fast gated short convolutional layers and 8 attention layers.
- ocrrag-lfm3-plain-gguf-v7↗
p4ulbr4dl3y
An ocrrag plain answerer based on the LFM3-1.7B model and LoRA v7, used to extract answers from retrieved text blocks or refuse to answer.