allknowingroger/NeuralDolphin-7B-slerp

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:8kLicense:apache-2.0Architecture:Transformer0.0K Open Weights Cold

NeuralPipe-7B-slerp by allknowingroger is a 7 billion parameter language model created by merging fterry/FofoNet-DolphinChat-slerp and vgorce/MarcoroNeuralChat-7B-slerp using the slerp method. This model leverages the strengths of its constituent models through a specific layer-wise parameter interpolation. It is designed for general text generation tasks, combining different conversational and neural chat model characteristics.

Loading preview...

Overview

NeuralPipe-7B-slerp is a 7 billion parameter language model developed by allknowingroger. It is a product of merging two distinct models, fterry/FofoNet-DolphinChat-slerp and vgorce/MarcoroNeuralChat-7B-slerp, utilizing the slerp (spherical linear interpolation) merge method via LazyMergekit. This approach allows for a nuanced combination of the source models' weights, specifically adjusting parameters for self-attention and MLP layers.

Key Characteristics

  • Merge Method: Employs the slerp method to combine model weights, offering a balanced integration of features from its base models.
  • Layer-wise Interpolation: The merge configuration specifies different interpolation values (t) for self-attention and MLP layers, indicating a tailored approach to combining model components.
  • Base Model: vgorce/MarcoroNeuralChat-7B-slerp serves as the foundational model for the merge process.

Usage

This model can be easily integrated into Python projects using the transformers library. The provided usage example demonstrates how to load the model and tokenizer, apply a chat template, and generate text with specified parameters like max_new_tokens, temperature, top_k, and top_p.

Popular Sampler Settings

Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.

temperature
top_p
top_k
frequency_penalty
presence_penalty
repetition_penalty
min_p