Tags
3 articles
I document the fine-tune of a Mistral 7B in Google Colab, with LoRA and QLoRA, trained to decide craft commands in a Minecraft agent.
Ler em portuguêsCurrent language: English39 min read
A practical LM Studio guide: how to choose a model based on available VRAM and what each parameter (context, GPU offload, flash attention) does.
Ler em portuguêsCurrent language: English7 min read
Step-by-step tutorial to configure AWS authentication in Postman and run commands like Scan directly on DynamoDB, without leaving the tool.
Ler em portuguêsCurrent language: English4 min read