{
  "builder_config": {
    "max_batch_size": 256,
    "max_input_len": 512,
    "name": "bert",
    "precision": "float16",
    "tensor_parallel": 1,
    "use_refit": false
  },
  "plugin_config": {
    "bert_attention_plugin": "float16",
    "context_fmha_enabled": true,
    "gemm_plugin": "float16",
    "gpt_attention_plugin": false,
    "identity_plugin": false,
    "layernorm_plugin": false,
    "layernorm_quantization_plugin": false,
    "nccl_plugin": false,
    "smooth_quant_gemm_plugin": false,
    "weight_only_quant_matmul_plugin": false
  }
}