Results on system Kuai-Kuai 702

mlcommons · Nov 18, 2024 · f0deba5 · f0deba5
1 parent 26ea4bd
commit f0deba5
Show file tree

Hide file tree

Showing 9 changed files with 215 additions and 215 deletions.
diff --git a/open/NTHU/measurements/ba7f5c94d10f-nvidia-gpu-TensorRT-scc24-base/README.md b/open/NTHU/measurements/ba7f5c94d10f-nvidia-gpu-TensorRT-scc24-base/README.md
@@ -1,3 +1,3 @@
 | Model               | Scenario   | Accuracy             |   Throughput | Latency (in ms)   |
 |---------------------|------------|----------------------|--------------|-------------------|
-| stable-diffusion-xl | offline    | (31.23426, 23.67816) |         1.36 | -                 |
+| stable-diffusion-xl | offline    | (31.23426, 23.67816) |        4.274 | -                 |
diff --git a/...7f5c94d10f-nvidia-gpu-TensorRT-scc24-base/stable-diffusion-xl/offline/README.md b/...7f5c94d10f-nvidia-gpu-TensorRT-scc24-base/stable-diffusion-xl/offline/README.md
@@ -41,7 +41,7 @@ cm run script \
 	--env.CM_SKIP_PREPROCESS_DATASET=yes \
 	--env.LD_LIBRARY_PATH=/opt/hpcx/ucx/lib:/opt/hpcx/ucc/lib:/usr/local/lib/python3.10/dist-packages/torch/lib:/usr/local/lib/python3.10/dist-packages/torch_tensorrt/lib:/usr/local/cuda/compat/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64 \
 	--env.CM_SKIP_NVMITTEN=yes \
-	--test_query_count=10 \
+	--test_query_count=5000 \
 	--clean
 ```
 *Note that if you want to use the [latest automation recipes](https://docs.mlcommons.org/inference) for MLPerf (CM scripts),
@@ -65,4 +65,4 @@ Model Precision: int8
 `FID_SCORE`: `23.67816`, Required accuracy for closed division `>= 23.01086` and `<= 23.95008`
 
 ### Performance Results 
-`Samples per second`: `1.35957`
+`Samples per second`: `4.2742`
diff --git a/...f5c94d10f-nvidia-gpu-TensorRT-scc24-base/stable-diffusion-xl/offline/cm-version-info.json b/...f5c94d10f-nvidia-gpu-TensorRT-scc24-base/stable-diffusion-xl/offline/cm-version-info.json
@@ -117,7 +117,7 @@
         "script_tags": "detect-os,detect,os,info",
         "script_variations": "",
         "version": "",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -137,7 +137,7 @@
         "script_tags": "detect,cpu,detect-cpu,info",
         "script_variations": "",
         "version": "",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -147,7 +147,7 @@
         "script_tags": "get,sys-utils-cm",
         "script_variations": "",
         "version": "",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -157,7 +157,7 @@
         "script_tags": "get,mlperf,inference,nvidia,scratch,space",
         "script_variations": "",
         "version": "",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -167,7 +167,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "mlperf_logging",
         "version": "4.1.0",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -177,7 +177,7 @@
         "script_tags": "get,src,source,inference,inference-src,inference-source,mlperf,mlcommons",
         "script_variations": "",
         "version": "master-git-4763c6b5e7c3ad904c461bc1053674c448771ec4",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -187,7 +187,7 @@
         "script_tags": "get,nvidia,mlperf,inference,common-code",
         "script_variations": "go",
         "version": "r4.1",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -267,7 +267,7 @@
         "script_tags": "generate,mlperf,inference,user-conf,inference-user-conf",
         "script_variations": "",
         "version": "",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -277,7 +277,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.pycuda",
         "version": "2022.2.2",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -287,7 +287,7 @@
         "script_tags": "get,cuda,cuda-compiler,cuda-lib,toolkit,lib,nvcc,get-nvcc,get-cuda,46d133d9ef92422d",
         "script_variations": "cudnn",
         "version": "12.4",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -297,7 +297,7 @@
         "script_tags": "get,tensorrt,nvidia",
         "script_variations": "",
         "version": "vdetected",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -307,27 +307,27 @@
         "script_tags": "build,mlcommons,mlperf,inference,inference-server,server,nvidia-harness,nvidia",
         "script_variations": "go",
         "version": "r4.1",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
-      "reproduce,mlperf,inference,nvidia,harness,_build_engine,_sdxl,_offline,_tensorrt,_cuda,_batch_size.8,_v4.1-dev": {
+      "reproduce,mlperf,inference,nvidia,harness,_build_engine,_cuda,_tensorrt,_sdxl,_offline,_batch_size.8,_v4.1-dev": {
         "script_uid": "bc3b17fb430f4732",
         "script_alias": "app-mlperf-inference-nvidia",
         "script_tags": "reproduce,mlcommons,mlperf,inference,harness,nvidia-harness,nvidia",
-        "script_variations": "build_engine,sdxl,offline,tensorrt,cuda,batch_size.8,v4.1-dev",
+        "script_variations": "build_engine,cuda,tensorrt,sdxl,offline,batch_size.8,v4.1-dev",
         "version": "",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
-      "reproduce,mlperf,inference,nvidia,harness,_preprocess_data,_sdxl,_tensorrt,_cuda,_v4.1-dev": {
+      "reproduce,mlperf,inference,nvidia,harness,_preprocess_data,_cuda,_tensorrt,_sdxl,_v4.1-dev": {
         "script_uid": "bc3b17fb430f4732",
         "script_alias": "app-mlperf-inference-nvidia",
         "script_tags": "reproduce,mlcommons,mlperf,inference,harness,nvidia-harness,nvidia",
-        "script_variations": "preprocess_data,sdxl,tensorrt,cuda,v4.1-dev",
+        "script_variations": "preprocess_data,cuda,tensorrt,sdxl,v4.1-dev",
         "version": "",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -337,7 +337,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.diffusers",
         "version": "0.31.0",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -347,7 +347,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.transformers",
         "version": "4.46.2",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -357,7 +357,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.accelerate",
         "version": "1.1.1",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -367,7 +367,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.torchmetrics",
         "version": "1.6.0",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -377,7 +377,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.torch-fidelity",
         "version": "0.3.0",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -387,7 +387,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.open_clip_torch",
         "version": "2.29.0",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -397,7 +397,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.opencv-python",
         "version": "4.10.0.84",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -407,7 +407,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.polygraphy",
         "version": "0.49.8",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -417,7 +417,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.nvtx",
         "version": "0.2.5",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -427,7 +427,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.cuda-python",
         "version": "12.4.0rc7",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -437,7 +437,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.ninja",
         "version": "1.11.1.1",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -447,7 +447,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.onnxruntime",
         "version": "1.20.0",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -457,7 +457,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.colored",
         "version": "2.2.4",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -467,7 +467,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.nvidia-ammo",
         "version": "0.7.4",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -477,7 +477,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.optimum",
         "version": "1.23.3",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -487,7 +487,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.onnx",
         "version": "1.14.0",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -497,7 +497,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.scipy",
         "version": "1.10.1",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -507,7 +507,7 @@
         "script_tags": "get,install,generic,generic-python-lib",
         "script_variations": "package.numpy",
         "version": "1.22.0",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     },
     {
@@ -547,7 +547,7 @@
         "script_tags": "mlperf,benchmark-mlperf",
         "script_variations": "",
         "version": "",
-        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_sdxl,_offline,_tensorrt,_cuda,_gpu_memory.80 )"
+        "parent": "app-mlperf-inference-nvidia,bc3b17fb430f4732 ( run_harness,_cuda,_tensorrt,_sdxl,_offline,_gpu_memory.80 )"
       }
     }
   ]

diff --git a/...rements/ba7f5c94d10f-nvidia-gpu-TensorRT-scc24-base/stable-diffusion-xl/offline/user.conf b/...rements/ba7f5c94d10f-nvidia-gpu-TensorRT-scc24-base/stable-diffusion-xl/offline/user.conf
@@ -1,5 +1,5 @@
 stable-diffusion-xl.Offline.target_qps = 5.0
-stable-diffusion-xl.Offline.max_query_count = 10
-stable-diffusion-xl.Offline.min_query_count = 10
+stable-diffusion-xl.Offline.max_query_count = 5000
+stable-diffusion-xl.Offline.min_query_count = 5000
 stable-diffusion-xl.Offline.min_duration = 0
 stable-diffusion-xl.Offline.sample_concatenate_permutation = 0
diff --git a/...5c94d10f-nvidia-gpu-TensorRT-scc24-base/stable-diffusion-xl/offline/accuracy/accuracy.txt b/...5c94d10f-nvidia-gpu-TensorRT-scc24-base/stable-diffusion-xl/offline/accuracy/accuracy.txt
@@ -1,2 +1,2 @@
-Accuracy Results: {'FID_SCORE': 23.678158324894298, 'CLIP_SCORE': 31.234263841211796}
-hash=916f10fbd3afed39d706138cc0421eb248ffd429c1782a41fba0cb37ffad0cbf
+Accuracy Results: {'FID_SCORE': 23.67815832489498, 'CLIP_SCORE': 31.234263841211796}
+hash=f2518594d7420ed36d53c96e614fa517e51c4e55870021352b44304e5c0b58ce