Overview

This model, Adanato/mistral_nemo_qwen25_qwen3_rank_only-qwen25_qwen3_rank_only_cluster_1, is a 12 billion parameter language model built upon the mistralai/Mistral-Nemo-Instruct-2407 architecture. It has been specifically fine-tuned using the qwen25_qwen3_rank_only_cluster_1 dataset.

Key Capabilities

• Specialized Fine-tuning: The model's training on a specific ranking dataset (qwen25_qwen3_rank_only_cluster_1) indicates a potential for enhanced performance in tasks involving comparative evaluation or preference ranking, particularly concerning outputs from Qwen2.5 and Qwen3 models.
• Base Model Strength: Inherits the foundational capabilities of the Mistral-Nemo-Instruct-2407 model, which typically includes strong general language understanding and instruction following.

Training Details

The fine-tuning process involved a learning rate of 1e-05, a train batch size of 4, and a total training batch size of 128 across 4 GPUs. The training utilized the AdamW_Torch_Fused optimizer with a cosine learning rate scheduler and a warmup ratio of 0.1 over 1 epoch.

Good For

• Applications requiring ranking or preference analysis of text, especially in contexts related to Qwen2.5 and Qwen3 model outputs.
• Tasks where a model needs to discern subtle differences and assign relative scores or ranks based on specific criteria.
• Developers looking for a specialized model that leverages the strengths of Mistral-Nemo-Instruct with a targeted ranking capability.