huggingface/local-gemma

Issues

• Anyway to force json only output like we can in Llamacpp?

  #40 opened 3 months ago by v3ss0n
  1

• Error Loading fbgemm.dll with PyTorch on Windows: OSError [WinError 126]

  #39 opened 4 months ago by ltkien2003
  1

• Feature req: Support for Gemma-2-2b and Gemma-2-2b-it

  #35 opened 4 months ago by notnotrishi
  5

• unable to use memory_extreme

  #24 opened 4 months ago by egollhofer
  1

• LocalGemma2ForCausalLM.from_pretrained("google/gemma-2-27b-it", preset="memory_extreme") can't load model.

  #27 opened 4 months ago by webbigdata-jp
  1

• authentication in LocalGemma2ForCausalLM with huggingface access token doesn't work

  #21 opened 5 months ago by dsingal0
  6

• The first preset table in the README is wrong

  #28 opened 5 months ago by phil71x
  1

• 8bit quantization

  #30 opened 4 months ago by paolo-losi
  1

• "auto" shouldn't use "exact" not "memory" if there is enough memory to load the entire model

  #20 opened 5 months ago by dsingal0
  8

• TypeError: 'QuantoConfig' object is not subscriptable

  #23 opened 5 months ago by mehulgupta2016154
  3

• API Provide

  #19 opened 5 months ago by codeMonkey-shin
  1