# Multi-stage fine-tuning

AceMath is a series of cutting-edge models designed specifically for mathematical reasoning. It is improved based on Qwen and is good at using chain-of-thought (CoT) reasoning to solve English mathematical problems.

AMD-OLMo is a series of 1-billion-parameter language models trained from scratch by AMD on the AMD Instinct™ MI250 GPU.

This is a TrOCR model specifically fine-tuned for medieval Castilian (particularly Hybrida handwriting), based on microsoft/trocr-base-handwritten.

An extended long-context model based on Meta-Llama-3-8B-Instruct, achieving 4194K context length support through RoPE theta parameter adjustment

This model is a fine-tuned Sinhala OCR model based on Ransaka/sinhala-ocr-model, primarily used for recognizing Sinhala text.

BEiT is an image classification model based on Vision Transformer (ViT) architecture, pretrained with self-supervised learning on ImageNet-21k and fine-tuned on ImageNet-1k.

An English-Vietnamese translation model fine-tuned based on the MarianMT pre-trained model, employing a two-phase training strategy