Remove sharding support for non-180B falcon variants

This commit is contained in:
Vivek Khandelwal
2023-11-14 03:55:11 -08:00
parent ca58908e5b
commit 666e601dd9
3 changed files with 487 additions and 808 deletions

View File

@@ -50,4 +50,8 @@ pefile
pyinstaller
# vicuna quantization
brevitas @ git+https://github.com/Xilinx/brevitas.git@56edf56a3115d5ac04f19837b388fd7d3b1ff7ea
brevitas @ git+https://github.com/Xilinx/brevitas.git@56edf56a3115d5ac04f19837b388fd7d3b1ff7ea
# For quantized GPTQ models
optimum
auto_gptq