Baidu releases ERNIE 5.5 Flash Base weights for further training

Baidu released ERNIE-5.5-Flash-Base weights on Hugging Face, offering developers a checkpoint for domain adaptation, instruction tuning and reinforcement-learning research. The company says the model has 280B total parameters, activates 14B per token and supports a 1M-token context for fine-tuning and inference. The release provides FP8 and BF16 formats under the ERNIE 5.5 Flash Base Model License Agreement v1.0. SGLang and vLLM can serve it through a completions API. The checkpoint comes directly after pre-training on 33.5T tokens, without mid-training or post-training; a post-trained version will follow.