Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,30 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: en
|
| 3 |
+
license: mit
|
| 4 |
+
tags:
|
| 5 |
+
- exbert
|
| 6 |
+
- text-classification
|
| 7 |
+
- onnx
|
| 8 |
+
- fp16
|
| 9 |
+
- roberta
|
| 10 |
+
- optimum
|
| 11 |
+
datasets:
|
| 12 |
+
- bookcorpus
|
| 13 |
+
- wikipedia
|
| 14 |
+
base_model:
|
| 15 |
+
- openai-community/roberta-large-openai-detector
|
| 16 |
+
---
|
| 17 |
+
|
| 18 |
+
# RoBERTa Large OpenAI Detector
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
This model is a FP16 optimized version of [openai-community/roberta-large-openai-detector](https://huggingface.co/openai-community/roberta-large-openai-detector/). It runs exclusively on the GPU.
|
| 22 |
+
The speedup compared to the base ONNX and pytorch versions depends chiefly on your GPU's FP16:FP32 ratio. For more comparison benchmarks and sample code of a related model, check here: [https://github.com/joaopn/gpu_benchmark_goemotions](https://github.com/joaopn/gpu_benchmark_goemotions).
|
| 23 |
+
|
| 24 |
+
You will need the GPU version of the ONNX Runtime. It can be installed with
|
| 25 |
+
|
| 26 |
+
```
|
| 27 |
+
pip install optimum[onnxruntime-gpu] --extra-index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-cuda-12/pypi/simple/
|
| 28 |
+
```
|
| 29 |
+
|
| 30 |
+
For convenience, this [benchmark repo](https://github.com/joaopn/gpu_benchmark_goemotions) provides an `environment.yml` file to create a conda env with all the requirements.
|