{ "_name_or_path": "training_output/results_paired_esm2_3B_wc_10epochs_3e-5_without_unpaired_2node/checkpoint-30815", "architectures": [ "EsmForMaskedLM" ], "attention_probs_dropout_prob": 0.0, "classifier_dropout": null, "emb_layer_norm_before": false, "esmfold_config": null, "hidden_act": "gelu", "hidden_dropout_prob": 0.0, "hidden_size": 2560, "initializer_range": 0.02, "intermediate_size": 10240, "is_folding_model": false, "layer_norm_eps": 1e-05, "mask_token_id": 32, "max_position_embeddings": 1026, "model_type": "esm", "num_attention_heads": 40, "num_hidden_layers": 36, "pad_token_id": 1, "position_embedding_type": "rotary", "token_dropout": true, "torch_dtype": "bfloat16", "transformers_version": "4.44.2", "use_cache": true, "vocab_list": null, "vocab_size": 33 }