AmberYifan/capsd-marin-8b-base-math_dsir_b8000_s0

TEXT GENERATIONPricing:Input $0.37 / Cached $0.074 / Output $0.38Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:8kTool Calling:SupportedPublished:Jul 29, 2026License:otherArchitecture:Transformer Featherless Exclusive Cold

AmberYifan/capsd-marin-8b-base-math_dsir_b8000_s0 is an 8 billion parameter language model, fine-tuned from marin-community/marin-8b-base. This model is specifically optimized for mathematical tasks, having been trained on the capsd_marin-8b-base-n80000-numina__mix_math_dsir_b8000_s0 dataset. It is designed to enhance performance in mathematical reasoning and problem-solving contexts. The model utilizes a context length of 8192 tokens, making it suitable for tasks requiring moderate input lengths.

Loading preview...

Model Overview

AmberYifan/capsd-marin-8b-base-math_dsir_b8000_s0 is an 8 billion parameter language model, fine-tuned from the existing marin-community/marin-8b-base architecture. Its primary differentiation lies in its specialized training for mathematical tasks.

Key Capabilities

  • Mathematical Optimization: The model has undergone fine-tuning on the capsd_marin-8b-base-n80000-numina__mix_math_dsir_b8000_s0 dataset, indicating a focus on improving performance in mathematical reasoning and problem-solving.
  • Base Model: Built upon marin-8b-base, suggesting a foundation in general language understanding before its specialized mathematical fine-tuning.
  • Context Length: Supports an 8192-token context window, allowing for processing of moderately long mathematical problems or related text.

Training Details

The model was trained using the following hyperparameters:

  • Learning Rate: 1e-05
  • Batch Size: train_batch_size of 2, eval_batch_size of 8, with a gradient_accumulation_steps of 8, resulting in a total_train_batch_size of 64.
  • Optimizer: ADAMW_TORCH with default betas and epsilon.
  • Scheduler: Cosine learning rate scheduler with 0.03 warmup steps.
  • Epochs: Trained for 1 epoch.

Intended Use Cases

This model is particularly suited for applications requiring enhanced mathematical capabilities, such as:

  • Solving mathematical problems.
  • Assisting with quantitative analysis.
  • Generating mathematical explanations or derivations.

Finetunes
1 models