Merge pull request #70 from woonyee28/mlperf-inference-results-scc24

Results on system 4xH100

Author: arjunsuresh

open/NTUHPC/code/stable-diffusion-xl/README.md

@@ -0,0 +1 @@
+TBD

open/NTUHPC/measurements/8297ae0eca20-nvidia-gpu-TensorRT-scc24-main/README.md

@@ -0,0 +1,3 @@
+| Model               | Scenario   | Accuracy              |   Throughput | Latency (in ms)   |
+|---------------------|------------|-----------------------|--------------|-------------------|
+| stable-diffusion-xl | offline    | (15.70418, 233.56896) |        2.667 | -                 |

open/NTUHPC/measurements/8297ae0eca20-nvidia-gpu-TensorRT-scc24-main/stable-diffusion-xl/offline/8297ae0eca20-nvidia-gpu-TensorRT-scc24-main.json

@@ -0,0 +1,7 @@
+{
+  "starting_weights_filename": "https://github.com/mlcommons/cm4mlops/blob/main/script/get-ml-model-stable-diffusion/_cm.json#L174",
+  "retraining": "no",
+  "input_data_types": "int32",
+  "weight_data_types": "int8",
+  "weight_transformations": "quantization, affine fusion"
+}

open/NTUHPC/measurements/8297ae0eca20-nvidia-gpu-TensorRT-scc24-main/stable-diffusion-xl/offline/README.md

@@ -0,0 +1,59 @@
+This experiment is generated using the [MLCommons Collective Mind automation framework (CM)](https://github.com/mlcommons/cm4mlops).
+
+*Check [CM MLPerf docs](https://docs.mlcommons.org/inference) for more details.*
+
+## Host platform
+
+* OS version: Linux-6.5.0-27-generic-x86_64-with-glibc2.29
+* CPU version: x86_64
+* Python version: 3.8.10 (default, Sep 11 2024, 16:02:53) 
+[GCC 9.4.0]
+* MLCommons CM version: 2.4.0
+
+## CM Run Command
+
+See [CM installation guide](https://docs.mlcommons.org/inference/install/).
+
+```bash
+pip install -U cmind
+
+cm rm cache -f
+
+cm pull repo mlcommons@cm4mlops --checkout=114709c8f6dbefa9ce5f8a599d55b349b5464bca
+
+cm run script \
+	--tags=run-mlperf,inference,_r4.1-dev,_short,_scc24-main \
+	--model=sdxl \
+	--implementation=nvidia \
+	--framework=tensorrt \
+	--category=datacenter \
+	--scenario=Offline \
+	--execution_mode=test \
+	--device=cuda \
+	--quiet \
+	--clean \
+	--batch-size=4 \
+	--target_qps=40
+```
+*Note that if you want to use the [latest automation recipes](https://docs.mlcommons.org/inference) for MLPerf (CM scripts),
+ you should simply reload mlcommons@cm4mlops without checkout and clean CM cache as follows:*
+
+```bash
+cm rm repo mlcommons@cm4mlops
+cm pull repo mlcommons@cm4mlops
+cm rm cache -f
+
+```
+
+## Results
+
+Platform: 8297ae0eca20-nvidia-gpu-TensorRT-scc24-main
+
+Model Precision: int8
+
+### Accuracy Results 
+`CLIP_SCORE`: `15.70418`, Required accuracy for closed division `>= 31.68632` and `<= 31.81332`
+`FID_SCORE`: `233.56896`, Required accuracy for closed division `>= 23.01086` and `<= 23.95008`
+
+### Performance Results 
+`Samples per second`: `2.66695`