Popular repositories Loading

1. Alpamayo1.5-finetune Alpamayo1.5-finetune Public

   VLA model of NVIDIA’s Alpamayo 1 and Alpamayo1.5, Add visualization, fine-tuning, RL fine-tuning, consistency training.

   Python 25 3

2. openVLA-Qwen2-0.5B openVLA-Qwen2-0.5B Public template

   OpenVLA Lightweight Version(0.5B). It uses qwen2-0.5B and fine-tunes using mllm format, without occupying LLM's inherent tokens. It represents actions with an additional 271 tokens.

   Python 19 3

3. LLM-from-scratch LLM-from-scratch Public

   Dolly is a custom LLM built from scratch

   Python 6 1

4. alpamayo-0.5B alpamayo-0.5B Public

   NVIDIA Alpamayo 的最小复现：Qwen2.5-0.5B VLM + 105.6M diffusion expert，支持单样本两阶段训练与推理

   Python 1 3

5. VLM-SFT VLM-SFT Public

   Minimal TRL and Transformers supervised fine-tuning example for VLMs

   Python

6. yoloe_demo yoloe_demo Public

   a demo of yoloe, including gradio app demo , yoloe inference and train

   Python 1