donoway/TinyStoriesV2_Llama-3.2-1B-vvxky0ut

Hugging Face
TEXT GENERATIONPricing:Input $0.108 / Output $0.804Concurrent Unit Cost:1Model Size:1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 13, 2025License:llama3.2Architecture:Transformer Featherless Exclusive Warm

The donoway/TinyStoriesV2_Llama-3.2-1B-vvxky0ut model is a 1 billion parameter language model, fine-tuned from the meta-llama/Llama-3.2-1B architecture. This model was trained with a learning rate of 0.01 over 100 epochs, utilizing an AdamW optimizer. Its specific fine-tuning dataset is not disclosed, suggesting its primary differentiator lies in its compact size and potential for specialized, resource-efficient applications.

Loading preview...

Model Overview

The donoway/TinyStoriesV2_Llama-3.2-1B-vvxky0ut is a compact 1 billion parameter language model, derived from the meta-llama/Llama-3.2-1B architecture. This model has undergone fine-tuning, though the specific dataset used for this process is not publicly detailed. It was trained using a learning rate of 0.01, a batch size of 1, and an AdamW optimizer over 100 epochs, with a constant learning rate scheduler.

Key Characteristics

  • Base Architecture: Fine-tuned from meta-llama/Llama-3.2-1B.
  • Parameter Count: 1 billion parameters, making it suitable for applications requiring a smaller footprint.
  • Context Length: Supports a context window of 32768 tokens.
  • Training Details: Trained for 100 epochs with specific hyperparameters including a learning rate of 0.01 and an AdamW optimizer.

Potential Use Cases

Given its compact size and the lack of specific dataset information, this model is likely best suited for:

  • Resource-constrained environments: Its 1B parameter count allows for deployment where larger models are impractical.
  • Exploratory fine-tuning: Developers can further fine-tune this model on custom, niche datasets for specific tasks.
  • Educational or research purposes: As a smaller Llama-based model, it can be used to experiment with LLM capabilities without extensive computational resources.