qingy2024/NaturalLM

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:12BQuant:FP8Ctx Length:32kLicense:apache-2.0Architecture:Transformer0.0K Open Weights Warm

NaturalLM is a 12 billion parameter Mistral-based causal language model developed by qingy2019, fine-tuned from unsloth/mistral-nemo-base-2407-bnb-4bit. This model was trained using Unsloth and Huggingface's TRL library, enabling 2x faster training. With a 32768 token context length, it is suitable for general language generation tasks.

Loading preview...

NaturalLM: A Mistral-Based Language Model

NaturalLM is a 12 billion parameter language model developed by qingy2019. It is fine-tuned from the unsloth/mistral-nemo-base-2407-bnb-4bit model, leveraging the Mistral architecture. The training process utilized Unsloth and Huggingface's TRL library, which facilitated a 2x acceleration in training speed.

Key Characteristics

  • Base Model: Mistral-based architecture.
  • Parameter Count: 12 billion parameters.
  • Training Efficiency: Achieved 2x faster training using Unsloth and Huggingface's TRL library.
  • Context Length: Supports a context window of 32768 tokens.
  • License: Distributed under the Apache-2.0 license.

Potential Use Cases

NaturalLM is suitable for a range of general-purpose language generation and understanding tasks, benefiting from its Mistral foundation and efficient training methodology. Its substantial context window allows for processing longer inputs and generating more coherent, extended outputs.