migtissera/Tess-M-v1.1

TEXT GENERATIONConcurrent Unit Cost:2Model Size:34BQuant:FP8Context Size:32kPublished:Nov 22, 2023License:yi-34bArchitecture:Transformer0.0K Featherless Exclusive Cold

migtissera/Tess-M-v1.1 is a 34 billion parameter general-purpose large language model based on the Yi-34B-200K architecture. This model is part of the Tess series, designed for broad applicability across various language tasks. It features a 32768 token context length, making it suitable for processing extensive inputs.

Loading preview...

Tess-M-v1.1 Overview

Tess-M-v1.1 is an experimental 34 billion parameter general-purpose Large Language Model developed by migtissera. It is built upon the Yi-34B-200K base architecture, indicating a foundation designed for handling large context windows, specifically 32768 tokens. This model is part of the broader Tess series, which aims to provide versatile language capabilities.

Key Characteristics

  • Architecture: Based on the Yi-34B-200K model.
  • Parameter Count: 34 billion parameters.
  • Context Length: Supports a substantial context window of 32768 tokens, enabling the processing of lengthy texts and complex queries.
  • Purpose: Designed as a general-purpose LLM, suitable for a wide range of applications.

Important Note

Users should be aware that Tess-M-v1.1 is an experimental and deprecated version. For stable and updated performance, the developer recommends using the newer Tess-M-v1.3 release.

Prompt Format

The model utilizes a specific prompt format for interaction:

SYSTEM: <ANY SYSTEM CONTEXT>
USER: 
ASSISTANT:

Finetunes
2 models