-
- quantize (OK)
```bash
python3 -m modelopt.onnx.quantization --onnx_path encoder.onnx \
--quantize_mode int8 --output_path encoder-w8a8-int8.onnx
/root/anaconda3/envs/modelopt/lib/pyt…
-
### Describe the issue
During the optimization stage of onnxruntime, the batchnorm preceded by Conv operator is being fused with the Conv operator and eliminated. The same process should apply to bat…
-
Papers:
- Same, Same But Different - Recovering Neural Network Quantization Error Through Weight Factorization https://arxiv.org/abs/1902.01917
- Up or Down? Adaptive Rounding for Post-Training …
-
### Issue Type
Bug
### Source
pip (mct-nightly)
### MCT Version
PR #1186
### OS Platform and Distribution
Linux Ubuntu 22.04
### Python version
3.10
### Describe the issu…
-
To evaluate the behavior of the two agent types—**IndividualAgent** (competitive, individualistic behavior) and **SystemAgent** (collaborative, cooperative behavior)—design a series of experiments tha…
-
Let's introduce training configure tool. This tool should find a configuration for network training with some optimal memory costs.
The Training Tool will be able to perform actions to improve the tr…
-
## 🐞Describing the bug
I've been trying to reduce the size of my Object Detection ML model generated by Xcode tool CreateML with Transfer Learning.
I found out that the transfer learning mlmodel out…
-
Prior to filing: check that this should be a bug instead of a feature request. Everything supported, including the compatible versions of TensorFlow, is listed in the overview page of each technique. …
-
### 💡 Your Question
I tuned the quantization weights during training and got an onnx model with Q/DQ layers as output. However, when I use TensorRt to convert a file to an engine with int8 precision,…
-
I thought about overfitting an autoencoder per wav file. Do you think it makes sense? What I say is, train a very small model to predict the whole signal. Then, compute the error. Now you just need to…