Hello. In trial period in DT4 Pro. I am using LM Studio, oss-gpt Q8 or Llama 3.3 70B Q6, I always have context length set to max supported, 128K usually. I cannot get DT4 to update the context length. It uses the models fine but same 32K context length shows in config settings each time not matter how many times I change it, quitting apps, etc makes no difference. MBP 16 M4 128GB RAM. Searching forum did not help. Please advise.
The maximum is currently indeed 32k. A future release might increase this but it will be challenging for one’s patience, IMHO 
1 Like
Anyway to bend this 32K limit? 32K is really small as I am looking at 80K approx 90% of the time. Open to any work arounds to say do a 32K + 32K? Open to any suggestions.