Raphael Lenain

Why is my model bigger after compression?

1

Hello, I am using [this config](https://github.com/microsoft/DeepSpeedExamples/blob/master/model_compression/bert/config/XTC/ds_config_W1A8_Qgroup1_fp32.json) on a translation model ([Helsinki-NLP/opus-mt-zh-en](https://huggingface.co/Helsinki-NLP/opus-mt-zh-en)), and I check the size of the model using the following function before and after running `init_compression` and `deepspeed.initialize`:...

Hello, I am trying to use OPUS-MT together with DeepSpeed compression (examples can be found at this link https://github.com/microsoft/DeepSpeedExamples under `model_compression`). I am running into an issue where the exact...

Finetuning

15

Hello, and thanks a lot for your great work. I am finetuning your 830M model on custom data, but getting to a point where I am overfitting fairly quickly. I...

Hi there, When dealing with certain data, but not with other data, for some reason, I am getting the following error: ```bash RuntimeError: max(): Expected reduction dim to be specified...

Fix/parse tree bug

1

Please note that we do not accept pull requests for the `master` branch. You should raise new PRs for the `dev` branch only. See our [contributing guide](https://github.com/novoic/blabla/blob/master/CONTRIBUTING.md) for details. Please...

Hi there, Thanks for your work and for open-sourcing you code and models. I want to use the `encoder.hdf5` model, but latest tensorflow versions and Keras versions are throwing errors....

Restart training from a checkpoint, with steps, etc

12

Hello, I was wondering whether there was an easy to restart training from a checkpoint, resuming the steps, epochs, optimizer states, etc. This is in case training dies, but we...