curvedinf
An open-source developer and the author of the Voodoo Dynamic Quant dynamic quantization tool, focused on quantization and compression of large language models.
Timeline (last 90 days)
Developer curvedinf open-sourced their dynamic quantization method Voodoo Dynamic Quant (VQ) under the MIT license.
The author believes Voodoo Quant is the first method to use backpropagation and gradient descent to select per-tensor quantization levels.
The author positions VQ as a research-grade project, which has not yet been validated on larger-scale models and has not reached a scientific level of proof of validity.
The author states that Unsloth has never publicly disclosed its own quantization methodology over the years.
All Facts (4)
The author believes Voodoo Quant is the first method to use backpropagation and gradient descent to select per-tensor quantization levels.
50%UnverifiedThe author positions VQ as a research-grade project, which has not yet been validated on larger-scale models and has not reached a scientific level of proof of validity.
50%UnverifiedThe author states that Unsloth has never publicly disclosed its own quantization methodology over the years.
50%UnverifiedDeveloper curvedinf open-sourced their dynamic quantization method Voodoo Dynamic Quant (VQ) under the MIT license.
50%