Llama-3.3-70B-Instruct

recent items

• Llama models: still valuable for finetuning or surpassed by everything new? (www.reddit.com) +69 4w

  Hello there people. So I have noticed that people are pretty much ignoring Llama 3 plus 3.1, 3.2, and 3.3 these days.

  ↯ Fine Tuning↯ Llama 3.3fine-tuningllama

• After 8 months of running everything local, ive accepted the productivity tools also have to be local (www.reddit.com) 16 4w

  Quick context: M3 max 64gb, currently running llama 3.3 70b q4 as my daily driver via ollama, qwen3 coder 30b for code (switched from qwen2.5 earlier this year), mlx for the smaller stuff. tried llama 4 scout earlier this year but 64gb is…

  ↯ Llama 3.3ollamallama