ArliAI-Llama-3-8B-Instruct-DPO-v0.2

Based on Meta-Llama-3-8b-Instruct, and is governed by Meta Llama 3 License agreement: https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct

Realized a tokenization mistake with the previous DPO model. So this is now a new version testing out DPO training on the following dataset:

• https://huggingface.co/datasets/mlabonne/orpo-dpo-mix-40k

The open LLM results are really BAD lol. Something with this dataset is disagreeing with llama 3?

Instruct format:

<|begin_of_text|><|start_header_id|>system<|end_header_id|>

{{ system_prompt }}<|eot_id|><|start_header_id|>user<|end_header_id|>

{{ user_message_1 }}<|eot_id|><|start_header_id|>assistant<|end_header_id|>

{{ model_answer_1 }}<|eot_id|><|start_header_id|>user<|end_header_id|>

{{ user_message_2 }}<|eot_id|><|start_header_id|>assistant<|end_header_id|>

Quants:

FP16: https://huggingface.co/OwenArli/ArliAI-Llama-3-8B-Instruct-DPO-v0.2

GGUF: https://huggingface.co/OwenArli/ArliAI-Llama-3-8B-Instruct-DPO-v0.2-GGUF