save/path/special_config_name.json

{
    "backend": {
        "name": "pytorch",
        "version": "2.1.0+rocm5.6",
        "_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
        "model": "TheBloke/Mistral-7B-Instruct-v0.1-AWQ",
        "task": "text-generation",
        "library": "transformers",
        "device": "cuda",
        "device_ids": "0",
        "seed": 42,
        "inter_op_num_threads": null,
        "intra_op_num_threads": null,
        "hub_kwargs": {
            "revision": "main",
            "force_download": false,
            "local_files_only": false,
            "trust_remote_code": false
        },
        "no_weights": true,
        "device_map": null,
        "torch_dtype": null,
        "amp_autocast": false,
        "amp_dtype": null,
        "eval_mode": true,
        "to_bettertransformer": false,
        "low_cpu_mem_usage": null,
        "attn_implementation": null,
        "cache_implementation": null,
        "torch_compile": false,
        "torch_compile_config": {},
        "quantization_scheme": "awq",
        "quantization_config": {
            "version": "exllama"
        },
        "deepspeed_inference": false,
        "deepspeed_inference_config": {},
        "peft_type": null,
        "peft_config": {}
    },
    "launcher": {
        "name": "process",
        "_target_": "optimum_benchmark.launchers.process.launcher.ProcessLauncher",
        "device_isolation": false,
        "start_method": "spawn"
    },
    "benchmark": {
        "name": "inference",
        "_target_": "optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark",
        "duration": 10,
        "warmup_runs": 10,
        "input_shapes": {
            "batch_size": 4,
            "num_choices": 2,
            "sequence_length": 128
        },
        "new_tokens": null,
        "energy": false,
        "memory": true,
        "latency": true,
        "forward_kwargs": {},
        "generate_kwargs": {
            "max_new_tokens": 128,
            "min_new_tokens": 128
        },
        "call_kwargs": {}
    },
    "experiment_name": "awq-exllamav2",
    "task": null,
    "model": null,
    "device": null,
    "library": null,
    "environment": {
        "cpu": " AMD EPYC 7763 64-Core Processor",
        "cpu_count": 128,
        "cpu_ram_mb": 1082015.236096,
        "system": "Linux",
        "machine": "x86_64",
        "platform": "Linux-5.15.0-84-generic-x86_64-with-glibc2.35",
        "processor": "x86_64",
        "python_version": "3.9.18",
        "gpu": [
            "AMD INSTINCT MI250 (MCM) OAM AC MBA",
            "AMD INSTINCT MI250 (MCM) OAM AC MBA",
            "AMD INSTINCT MI250 (MCM) OAM AC MBA",
            "AMD INSTINCT MI250 (MCM) OAM AC MBA",
            "AMD INSTINCT MI250 (MCM) OAM AC MBA",
            "AMD INSTINCT MI250 (MCM) OAM AC MBA",
            "AMD INSTINCT MI250 (MCM) OAM AC MBA",
            "AMD INSTINCT MI250 (MCM) OAM AC MBA"
        ],
        "gpu_count": 8,
        "gpu_vram_mb": 549621596160,
        "optimum_benchmark_version": "0.2.0",
        "optimum_benchmark_commit": "09f95ce7707eb32c7880f72cdd9e14e8b7554315",
        "transformers_version": "4.37.2",
        "transformers_commit": null,
        "accelerate_version": "0.24.1",
        "accelerate_commit": null,
        "diffusers_version": "0.26.3",
        "diffusers_commit": null,
        "optimum_version": "1.14.1",
        "optimum_commit": null,
        "timm_version": "0.9.16",
        "timm_commit": null,
        "peft_version": "0.8.2",
        "peft_commit": null
    }
}