aieng-lab/starcoder2-3b_comment-type-java

1{2  "_name_or_path": "bigcode/starcoder2-3b",3  "architectures": [4    "Starcoder2ForSequenceClassification"5  ],6  "attention_dropout": 0.1,7  "bos_token_id": 0,8  "embedding_dropout": 0.1,9  "eos_token_id": 0,10  "hidden_act": "gelu_pytorch_tanh",11  "hidden_size": 3072,12  "id2label": {13    "0": "LABEL_0",14    "1": "LABEL_1",15    "2": "LABEL_2",16    "3": "LABEL_3",17    "4": "LABEL_4",18    "5": "LABEL_5",19    "6": "LABEL_6"20  },21  "initializer_range": 0.018042,22  "intermediate_size": 12288,23  "label2id": {24    "LABEL_0": 0,25    "LABEL_1": 1,26    "LABEL_2": 2,27    "LABEL_3": 3,28    "LABEL_4": 4,29    "LABEL_5": 5,30    "LABEL_6": 631  },32  "max_position_embeddings": 16384,33  "mlp_type": "default",34  "model_type": "starcoder2",35  "norm_epsilon": 1e-05,36  "norm_type": "layer_norm",37  "num_attention_heads": 24,38  "num_hidden_layers": 30,39  "num_key_value_heads": 2,40  "pad_token_id": 0,41  "problem_type": "multi_label_classification",42  "residual_dropout": 0.1,43  "rope_scaling": null,44  "rope_theta": 999999.4420358813,45  "sliding_window": 4096,46  "torch_dtype": "bfloat16",47  "transformers_version": "4.48.3",48  "use_bias": true,49  "use_cache": true,50  "vocab_size": 4915251}52