https://huggingface.co/dealignai/GLM-5.3-UNCENSORED-NVFP4
https://huggingface.co/dealignai/GLM-5.3-UNCENSORED-NVFP4
https://huggingface.co/dealignai/GLM-5.3-ABLITERATED-NVFP4
Hopefully won't present issues as it's already quantized. I don't believe LM Studio on Mac runs this and I do not have any storage to do it myself. I would like the highest quality 4-bit quant that can be created from this, if possible. Thanks.
It's queued!
You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#GLM-5.3-UNCENSORED-NVFP4-GGUF
https://hf.tst.eu/model#GLM-5.3-ABLITERATED-NVFP4-GGUF for quants to appear.
I think those models might not have worked, but that's okay, as they released this recently. https://huggingface.co/dealignai/GLM-5.3-UNCENSORED-FP8 Can you see if it will quantize?
they are 753B param models. they are too big for auto queue, so they will require manual quantization even if we attempt them
Okay, I would really like the FP8 to be quantized if you can, as unfortunately nobody else has abliterated the full GLM-5.3 and released a GGUF yet. The NVFP4 models I originally linked are a bit low quality and are just duplicates, so you can skip those.