Star 总榜

GLM-5.2 on 4x DGX Spark with adaptive MTP K2/K4/K5, FULL CUDA graphs, DCP2, 520K context, and a downloadable ARM64 runtime image.

Distributed Inference with vLLM

Benchmark OpenAI-compatible AI endpoints and AI Accelerators in a reproducible structured way

DeepSeek-V4-Flash-0731 tuned on 4x DGX Spark (GB10) at TP=4 — 123.13 tok/s decode. Every variable measured one at a time, with the values that lost.

This project fine-tunes the Qwen2-1.5B model for Arabic language tasks using Quantized LoRA (QLoRA).

Fine-tune GPT2 to generate fake job experiences

Landcover classification on sentinel-2 data with Prithvi, EfficientNet-Unet and OSM / CNES Landcover labels.

🎙️ CosyVoice LoRA 微调框架：LLM+Flow 联合训练，实现无 Prompt 语音合成

A fine tuned yolov111 model to detect fire

AgriMind: The smart LLM application that every farmer truly needs

This project aims to develop an innovative anomaly detection system using advanced data mining and deep learning techniques to accurately identify and localize defects in manufacturing components, thereby enhancing quality control processes and reducing production losses.

🤖 Deep Catalan: Bring closer the Catalan Language to Deep Learning using ULMFit.

This is the unofficial re-implementation of "Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning Attack" (ICML2025)

Code the ICML 2024 paper: "Variance-reduced Zeroth-Order Methods for Fine-Tuning Language Models"

Lightweight, Self-Hosted AI Guardrails Model based on ModernBERT.

[ICML 2025] Vulnerability-Aware Alignment: Mitigating Uneven Forgetting in Harmful Fine-Tuning

🧪 SynthForge — toolkit for generating high-quality synthetic datasets for LLM fine-tuning. Multi-backend, filterable, GPU-accelerated. Early-stage experimental project.

In the context of Morocco (Darija), sentiment analysis can be an invaluable asset for businesses and organizations looking to better understand the satisfaction of their customers. We will fine-tune a model for Darija

Fine tuning Mistral-7b with PEFT(Parameter Efficient Fine-Tuning) and LoRA(Low-Rank Adaptation) on Puffin Dataset(multi-turn conversations between GPT-4 and real humans)

Adversarial LLM arena using GRPO + SFT to train a model that generates questions hard LLMs can't solve