AmberYifan/capsd-marin-8b-base-math_ifd_b2000_s0

TEXT GENERATIONPricing:Input $0.37 / Cached $0.074 / Output $0.38Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:8kTool Calling:SupportedPublished:Jul 30, 2026License:otherArchitecture:Transformer Featherless Exclusive Cold

AmberYifan/capsd-marin-8b-base-math_ifd_b2000_s0 is an 8 billion parameter language model, fine-tuned from marin-community/marin-8b-base. This model is specifically adapted for mathematical tasks, leveraging the capsd_marin-8b-base-n80000-numina__mix_math_ifd_b2000_s0 dataset. It is designed to enhance performance in mathematical reasoning and problem-solving contexts.

Loading preview...

Model Overview

This model, capsd-marin-8b-base-math_ifd_b2000_s0, is an 8 billion parameter language model developed by AmberYifan. It is a fine-tuned variant of the marin-community/marin-8b-base architecture, specifically optimized for mathematical applications.

Key Characteristics

  • Base Model: Fine-tuned from marin-community/marin-8b-base.
  • Parameter Count: 8 billion parameters.
  • Context Length: Supports an 8192-token context window.
  • Specialization: Enhanced for mathematical tasks through fine-tuning on the capsd_marin-8b-base-n80000-numina__mix_math_ifd_b2000_s0 dataset.

Training Details

The model was trained using the following hyperparameters:

  • Learning Rate: 1e-05
  • Optimizer: AdamW with betas=(0.9, 0.999) and epsilon=1e-08.
  • Scheduler: Cosine learning rate scheduler with 0.03 warmup steps.
  • Epochs: Trained for 1 epoch.
  • Batch Size: A total train batch size of 64 (2 per device with 8 gradient accumulation steps across 4 GPUs).

Intended Use Cases

This model is suitable for applications requiring strong mathematical reasoning and problem-solving capabilities, particularly within the domain it was fine-tuned for.