AttributeError: 'DatasetDict' object has no attribute 'features'

[Phasor]

Please check that this issue hasn't been reported before.

• I searched previous Bug Reports didn't find any similar reports.

Expected Behavior

The model should be fine tuned using my .yml file config.

Current behaviour

I am running this on Colab.

When I run:

!git clone https://github.com/OpenAccess-AI-Collective/axolotl
%cd axolotl

!pip3 install packaging
!pip3 install -e '.[flash-attn,deepspeed]'
!pip3 install -U git+https://github.com/huggingface/peft.git

and then try and train the model using a .yml file which uses a hugging face data set as follows:

!accelerate launch -m axolotl.cli.train sql.yml

I get the following error:

AttributeError: 'DatasetDict' object has no attribute 'features'

Steps to reproduce

On colab run:

!git clone https://github.com/OpenAccess-AI-Collective/axolotl
%cd axolotl

!pip3 install packaging
!pip3 install -e '.[flash-attn,deepspeed]'
!pip3 install -U git+https://github.com/huggingface/peft.git

Create a sql.yml file and upload it.

and then try and train the model using a .yml file using the HF data set (https://huggingface.co/datasets/knowrohit07/know_sql) as follows:

!accelerate launch -m axolotl.cli.train sql.yml

Possible solution

No response

Which Operating Systems are you using?

• Linux
• macOS
• Windows

Python Version

3.10

axolotl branch-commit

!git clone https://github.com/OpenAccess-AI-Collective/axolotl

Acknowledgements

• My issue title is concise, descriptive, and in title casing.
• I have searched the existing issues to make sure this bug has not been reported yet.
• I am using the latest version of axolotl.
• I have provided enough information for the maintainers to reproduce and diagnose the issue.

[winglian]

@Phasor please see the linked pr that was closed. That should help.