aieng-lab/starcoder2-7b_comment-type-java

1{2  "_name_or_path": "bigcode/starcoder2-7b",3  "activation_function": "gelu",4  "architectures": [5    "Starcoder2ForSequenceClassification"6  ],7  "attention_dropout": 0.1,8  "attention_softmax_in_fp32": true,9  "bos_token_id": 0,10  "embedding_dropout": 0.1,11  "eos_token_id": 0,12  "hidden_act": "gelu_pytorch_tanh",13  "hidden_size": 4608,14  "id2label": {15    "0": "LABEL_0",16    "1": "LABEL_1",17    "2": "LABEL_2",18    "3": "LABEL_3",19    "4": "LABEL_4",20    "5": "LABEL_5",21    "6": "LABEL_6"22  },23  "initializer_range": 0.018042,24  "intermediate_size": 18432,25  "label2id": {26    "LABEL_0": 0,27    "LABEL_1": 1,28    "LABEL_2": 2,29    "LABEL_3": 3,30    "LABEL_4": 4,31    "LABEL_5": 5,32    "LABEL_6": 633  },34  "layer_norm_epsilon": 1e-05,35  "max_position_embeddings": 16384,36  "mlp_type": "default",37  "model_type": "starcoder2",38  "norm_epsilon": 1e-05,39  "norm_type": "layer_norm",40  "num_attention_heads": 36,41  "num_hidden_layers": 32,42  "num_key_value_heads": 4,43  "pad_token_id": 0,44  "problem_type": "multi_label_classification",45  "residual_dropout": 0.1,46  "rope_scaling": null,47  "rope_theta": 1000000,48  "scale_attention_softmax_in_fp32": true,49  "scale_attn_weights": true,50  "sliding_window": 4096,51  "torch_dtype": "bfloat16",52  "transformers_version": "4.48.3",53  "use_bias": true,54  "use_cache": true,55  "vocab_size": 4915256}57