Fine Tuning Validation error: At least 16 examples needed for each train/eval file.

I am following AMZ Workshop : 02_fine-tuning_llama2.ipynb to try to fine-tune the the cohere.command-light-text-v14:7:4k model. They are using the CNN 3.0.0 dataset and prepping it (see Prepare CNN news...) so that it looks like:

training data:

{ "prompt": "Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.\n\ninstruction:\n\nSummarize the news article provided below.\n\ninput:\n\n(article snippet)", "completion": "response:\n\n(example summary)" }

The validation data is looks like:

{"prompt": "Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.\n\ninstruction:\n\nSummarize the news article provided below.\n\ninput:\n\n(article snippet)", "completion": "response:\n\n(example response)"}

The training file contains 5k rows, and the validation file contains 1k rows. This works with the amazon.titan-text-lite-v1:0:4k model. However, using the cohere command light model I get:

Validation error: At least 16 examples needed for each train/eval file. train file doesn't have enough examples.

I checked AWS Prepare training and validation datasets..., but the format looks the same.

Any idea why it is giving the "Validation error" or how to fix it?

This Bedrock fine tuning ValidationError question is similar, but I don't think the suggestions fix the issue I'm having.