aieng-lab/starcoder2-3b_commit-intent

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  },17  "initializer_range": 0.018042,18  "intermediate_size": 12288,19  "label2id": {20    "LABEL_0": 0,21    "LABEL_1": 1,22    "LABEL_2": 223  },24  "max_position_embeddings": 16384,25  "mlp_type": "default",26  "model_type": "starcoder2",27  "norm_epsilon": 1e-05,28  "norm_type": "layer_norm",29  "num_attention_heads": 24,30  "num_hidden_layers": 30,31  "num_key_value_heads": 2,32  "pad_token_id": 0,33  "residual_dropout": 0.1,34  "rope_scaling": null,35  "rope_theta": 999999.4420358813,36  "sliding_window": 4096,37  "torch_dtype": "bfloat16",38  "transformers_version": "4.48.3",39  "use_bias": true,40  "use_cache": true,41  "vocab_size": 4915242}43