ep0p

retrieve training parameters from a lora model?

2

Hello, I've fine-tuned Llama3.2 several times, and one model in particular—the very first one I trained—performs the best. Since it was my first attempt while I was exploring unsloth, I...

fine-tuned llama3.2 models do not provide output while inferencing

1

I've been working on fine-tuning Ollama 3.2 to enhance OCR performance on documents with irregular layouts, handwritten annotations, stamps, and noise. To test this, I ran two experiments with `num_train_epochs...

'Got cancellation request for SGLang server'

10

### 🐛 Describe the bug Hello, I might be doing something wrong, when trying to run the model on local the SGLang server exits : ``` python -m olmocr.pipeline ./localworkspace...