Weyaxi/Seraph-openchat-3.5-1210-Slerp

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Dec 27, 2023License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

Weyaxi/Seraph-openchat-3.5-1210-Slerp is a 7 billion parameter language model created by Weyaxi, built upon the Mistral-7B-v0.1 base architecture. This model is a merge of Seraph-7B and openchat-3.5-1210 using the slerp method, specifically designed to combine their respective strengths. It is optimized for general conversational tasks, leveraging the combined capabilities of its merged components.

Loading preview...

Model Overview

Weyaxi/Seraph-openchat-3.5-1210-Slerp is a 7 billion parameter language model developed by Weyaxi. It is constructed using the Mistral-7B-v0.1 as its base model and employs the slerp (spherical linear interpolation) merge method via mergekit to combine two distinct models: Weyaxi/Seraph-7B and openchat/openchat-3.5-1210.

Key Characteristics

  • Merged Architecture: Combines the strengths of Weyaxi/Seraph-7B and openchat/openchat-3.5-1210 through a slerp merge.
  • Parameter Configuration: Specific t parameters were applied during the merge, with varying values for self_attn and mlp layers, and a fallback value of 0.5 for other tensors.
  • Tokenizer: Utilizes a union tokenizer source, ensuring broad compatibility.
  • Precision: Operates in bfloat16 data type for efficient computation.

Intended Use Cases

This model is suitable for general-purpose conversational AI applications, benefiting from the blended capabilities of its constituent models. Its design suggests a focus on balanced performance across various language understanding and generation tasks.