Papers Restoring DeepSeek-R1 Model for Exploring In-depth Insights into Data and Training Procedures by Hugging Face , and Administrator 2025 April 29