File size: 1,024 Bytes
2bb0d13 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 | {
"onnxruntime": "onnxruntime-gpu 1.32.0",
"measured_on": "NVIDIA A100-SXM4-80GB, CUDA execution provider, arena_extend_strategy kSameAsRequested, cudnn_conv_algo_search HEURISTIC; peak is process total in GiB including the CUDA context; time is the median of 5 runs after 2 warm-up runs",
"onnx_opset": 17,
"batch_axis": "dynamic",
"n_frames": 64,
"encoder_chunk": 16,
"outputs": [
"fake_prob",
"logits",
"frame_probs"
],
"files": {
"model.onnx": {
"precision": "float32",
"peak_vram_gb": {
"batch1": 3.45,
"batch2": 4.92
},
"seconds_per_clip": {
"batch1": 1.49,
"batch2": 1.43
}
},
"model.fp16.onnx": {
"precision": "float16",
"peak_vram_gb": {
"batch1": 2.55,
"batch2": 3.88
},
"seconds_per_clip": {
"batch1": 0.98,
"batch2": 0.94
}
}
},
"pytorch_bf16_reference": {
"precision": "bfloat16 autocast",
"peak_vram_gb": {
"batch1": 4.69,
"batch2": 4.97
},
"seconds_per_clip": {
"batch1": 0.75,
"batch2": 0.75
}
}
}
|