README.md

metadata

license: cc-by-sa-4.0
tags:
  - causal-lm

Description

This is a test model exported for my test script. It was exported from stabilityai/stablelm-3b-4e1t to ONNX with a modified Hugging Face Optimum. It is quite possible to have problems.

This model does not include a tokenizer.

Export command

This model was exported with the following command:

optimum-cli export onnx --model stabilityai/stablelm-3b-4e1t --trust-remote-code --device cpu --optimize O1 output/onnx-fp32/

It requires Transformers 4.38 or later to export.

Output from Optimum CLI

The ONNX export succeeded with the warning: The maximum absolute difference between the output of the reference model and the ONNX exported model is not within the set tolerance 1e-05:
- logits: max diff = 2.6553869247436523e-05.