Instructions to use multimolecule/a2zchromatin-accessibility with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MultiMolecule
How to use multimolecule/a2zchromatin-accessibility with MultiMolecule:
pip install multimolecule
from multimolecule import AutoModel, AutoTokenizer tokenizer = AutoTokenizer.from_pretrained("multimolecule/a2zchromatin-accessibility") model = AutoModel.from_pretrained("multimolecule/a2zchromatin-accessibility") inputs = tokenizer("ACTCCCCTGCCCTCAACAAGATGTTTTGCCAACTGGCCAAGACCTGCCCTGTGCAGCTGTGGGTTGATTCCACACCCCCGCCCGGCACCCGCGTCCGCGCCATGGCCATCTACAAGCAGTCACAGCACATGACGGAGGTTGTGAGGCGCTGCCCCCACCATGAGCGCTGCTCAGATAGCGATGG", return_tensors="pt") outputs = model(**inputs) embeddings = outputs.last_hidden_state - Notebooks
- Google Colab
- Kaggle
| { | |
| "architectures": [ | |
| "A2zChromatinForSequencePrediction" | |
| ], | |
| "bos_token_id": 1, | |
| "conv_channels": 320, | |
| "conv_dropout": 0.2, | |
| "conv_kernel_size": 26, | |
| "dtype": "float32", | |
| "eos_token_id": 2, | |
| "fc_size": 925, | |
| "head": { | |
| "act": null, | |
| "bias": true, | |
| "dropout": 0.0, | |
| "hidden_size": 925, | |
| "layer_norm_eps": 1e-12, | |
| "loss_weight": null, | |
| "num_labels": 1, | |
| "output_name": null, | |
| "problem_type": "binary", | |
| "transform": null, | |
| "transform_act": "gelu", | |
| "type": null | |
| }, | |
| "hidden_act": "relu", | |
| "hidden_size": 925, | |
| "id2label": null, | |
| "label2id": null, | |
| "lstm_dropout": 0.5, | |
| "lstm_hidden_size": 320, | |
| "mask_token_id": 4, | |
| "model_type": "a2zchromatin", | |
| "null_token_id": 5, | |
| "num_labels": 1, | |
| "pad_token_id": 0, | |
| "pool_size": 13, | |
| "sequence_length": 600, | |
| "tie_word_embeddings": true, | |
| "transformers_version": "5.9.0", | |
| "unk_token_id": 3, | |
| "vocab_size": 16 | |
| } | |