bumblebee-testing/tiny-random-GPTBigCodeModel-multi_query-False

1{2  "_name_or_path": "hf-internal-testing/tiny-random-GPTBigCodeModel",3  "activation_function": "relu",4  "architectures": [5    "GPTBigCodeModel"6  ],7  "attention_softmax_in_fp32": false,8  "attn_pdrop": 0.1,9  "bos_token_id": 0,10  "embd_pdrop": 0.1,11  "eos_token_id": 0,12  "gradient_checkpointing": false,13  "initializer_range": 0.02,14  "layer_norm_epsilon": 1e-05,15  "model_type": "gpt_bigcode",16  "multi_query": false,17  "n_embd": 32,18  "n_head": 4,19  "n_inner": 37,20  "n_layer": 5,21  "n_positions": 512,22  "pad_token_id": 1021,23  "reorder_and_upcast_attn": false,24  "resid_pdrop": 0.1,25  "scale_attention_softmax_in_fp32": false,26  "scale_attn_by_inverse_layer_idx": false,27  "scale_attn_weights": true,28  "torch_dtype": "float32",29  "transformers_version": "4.35.0",30  "type_vocab_size": 16,31  "use_cache": true,32  "vocab_size": 102433}34