Llama-3-8b-ita-ties

This is a merge of pre-trained language models created using mergekit.

I tried to merge two of the best Italian LLMs using Mergekit. The results are acceptable, but I could not improve on the best existing model.

Evaluation

For a detailed comparison of model performance, check out the Leaderboard for Italian Language Models.

Here's a breakdown of the performance metrics:

Metric	hellaswag_it acc_norm	arc_it acc_norm	m_mmlu_it 5-shot acc	Average
Accuracy Normalized	0.6621	0.5535	0.5749	0.5968

Merge Details

Merge Method

This model was merged using the TIES merge method using meta-llama/Meta-Llama-3-8B as a base.

Models Merged

The following models were included in the merge:

• swap-uniba/LLaMAntino-3-ANITA-8B-Inst-DPO-ITA
• DeepMount00/Llama-3-8b-Ita

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: meta-llama/Meta-Llama-3-8B
    # no parameters necessary for base model
  - model: swap-uniba/LLaMAntino-3-ANITA-8B-Inst-DPO-ITA
    parameters:
      density: 0.5
      weight: 0.6
  - model: DeepMount00/Llama-3-8b-Ita
    parameters:
      density: 0.5
      weight: 0.3
merge_method: ties
base_model: meta-llama/Meta-Llama-3-8B
parameters:
  normalize: true
dtype: bfloat16