minuva/MiniLMv2-agentflow-v2
170
1{2 "_name_or_path": "nreimers/MiniLMv2-L6-H384-distilled-from-RoBERTa-Large",3 "architectures": [4 "RobertaForSequenceClassification"5 ],6 "attention_probs_dropout_prob": 0.1,7 "bos_token_id": 0,8 "classifier_dropout": null,9 "eos_token_id": 2,10 "hidden_act": "gelu",11 "hidden_dropout_prob": 0.1,12 "hidden_size": 384,13 "id2label": {14 "0": "OTHER",15 "1": "agent_apology_error_mistake",16 "2": "agent_apology_unsatisfactory",17 "3": "agent_didnt_understand",18 "4": "agent_limited_capabilities",19 "5": "agent_refuses_answer",20 "6": "image_limitations",21 "7": "no_information_doesnt_know",22 "8": "success_and_followup_assistance"23 },24 "initializer_range": 0.02,25 "intermediate_size": 1536,26 "label2id": {27 "OTHER": 0,28 "agent_apology_error_mistake": 1,29 "agent_apology_unsatisfactory": 2,30 "agent_didnt_understand": 3,31 "agent_limited_capabilities": 4,32 "agent_refuses_answer": 5,33 "image_limitations": 6,34 "no_information_doesnt_know": 7,35 "success_and_followup_assistance": 836 },37 "layer_norm_eps": 1e-05,38 "max_position_embeddings": 514,39 "model_type": "roberta",40 "num_attention_heads": 12,41 "num_hidden_layers": 6,42 "pad_token_id": 1,43 "position_embedding_type": "absolute",44 "problem_type": "single_label_classification",45 "torch_dtype": "float32",46 "transformers_version": "4.37.0",47 "type_vocab_size": 1,48 "use_cache": true,49 "vocab_size": 5026550}51 