tensorrt_llm/1/config.json

{
  "builder_config": {
    "fp8": true,
    "hidden_act": "silu",
    "hidden_size": 8192,
    "int8": false,
    "max_batch_size": 64,
    "max_input_len": 4096,
    "max_num_tokens": null,
    "max_output_len": 4096,
    "max_position_embeddings": 4096,
    "name": "llama",
    "num_heads": 64,
    "num_kv_heads": 8,
    "num_layers": 80,
    "parallel_build": true,
    "pipeline_parallel": 1,
    "precision": "float16",
    "quant_mode": 384,
    "tensor_parallel": 4,
    "use_refit": false,
    "vocab_size": 32000
  },
  "plugin_config": {
    "attention_qk_half_accumulation": false,
    "bert_attention_plugin": false,
    "context_fmha_type": 1,
    "gemm_plugin": "float16",
    "gpt_attention_plugin": "float16",
    "identity_plugin": false,
    "layernorm_plugin": false,
    "layernorm_quantization_plugin": false,
    "lookup_plugin": false,
    "nccl_plugin": "float16",
    "paged_kv_cache": true,
    "quantize_per_token_plugin": false,
    "quantize_tensor_plugin": false,
    "remove_input_padding": true,
    "rmsnorm_plugin": false,
    "rmsnorm_quantization_plugin": false,
    "smooth_quant_gemm_plugin": false,
    "tokens_per_block": 64,
    "use_custom_all_reduce": false,
    "weight_only_groupwise_quant_matmul_plugin": false,
    "weight_only_quant_matmul_plugin": false
  }
}