Views
No views yet
genbio-ai/AIDO.RNA-1.6B (Apache-2.0) so that the model can be loaded directly via AutoModel.from_pretrained(..., trust_remote_code=True). The architecture and weights are unchanged — only the tokenizer plumbing and an auto_map entry in config.json have been added, along with a bundled modeling_aido.py / configuration_aido.py / tokenization_aido.py so HuggingFace's remote-code mechanism can discover them.1from transformers import AutoModel, AutoTokenizer
2
3model = AutoModel.from_pretrained("EscheWang/AIDO.RNA-1.6B-hf", trust_remote_code=True)
4tokenizer = AutoTokenizer.from_pretrained("EscheWang/AIDO.RNA-1.6B-hf", trust_remote_code=True)
5
6# NOTE: the tokenizer is a BertWordPieceLowerCase variant that expects
7# space-separated RNA characters (one token per whitespace unit).
8enc = tokenizer("A C G U A C G U", return_tensors="pt", add_special_tokens=True)
9out = model(enc.input_ids.int(), attention_mask=enc.attention_mask)
10print(out.last_hidden_state.shape) # (1, seq_len + 2, 2048)"A C G U"), not a contiguous string.bert. prefix from the upstream checkpoint; the base_model_prefix in modeling_aido.py is set to "bert" so HuggingFace's from_pretrained auto-strip correctly loads them into AidoRnaModel.