Geodezik/trainer_output
02
1{2 "architectures": [3 "LlamaForSequenceClassification"4 ],5 "attention_bias": false,6 "attention_dropout": 0.0,7 "bos_token_id": 1,8 "eos_token_id": 2,9 "head_dim": 64,10 "hidden_act": "silu",11 "hidden_size": 576,12 "id2label": {13 "0": "LABEL_0"14 },15 "initializer_range": 0.02,16 "intermediate_size": 1536,17 "label2id": {18 "LABEL_0": 019 },20 "max_position_embeddings": 2048,21 "mlp_bias": false,22 "model_type": "llama",23 "num_attention_heads": 9,24 "num_hidden_layers": 30,25 "num_key_value_heads": 3,26 "pad_token_id": 2,27 "pretraining_tp": 1,28 "rms_norm_eps": 1e-05,29 "rope_scaling": null,30 "rope_theta": 10000.0,31 "tie_word_embeddings": true,32 "torch_dtype": "float32",33 "transformers_version": "4.50.0",34 "use_cache": true,35 "vocab_size": 4915236}37 