├── .gitignore ├── README.md ├── fake_quant ├── README.md ├── data_utils.py ├── eval_utils.py ├── gptaq_utils.py ├── gptq_utils.py ├── hadamard_utils.py ├── main.py ├── model_utils.py ├── monkeypatch.py ├── quant_utils.py ├── requirements.txt ├── rotation_utils.py ├── run_llama.sh └── utils.py ├── img └── readme_intro.png ├── spinquant ├── CODE_OF_CONDUCT.md ├── CONTRIBUTING.md ├── LICENSE ├── README.md ├── SpinQuant.png ├── eval_utils │ ├── gptaq_utils.py │ ├── gptq_utils.py │ ├── main.py │ ├── modeling_llama.py │ └── rotation_utils.py ├── optimize_rotation.py ├── ptq.py ├── requirement.txt ├── scripts │ ├── 10_optimize_rotation.sh │ ├── 11_optimize_rotation_fsdp.sh │ ├── 2_eval_ptq.sh │ ├── 31_optimize_rotation_executorch.sh │ └── 32_eval_ptq_executorch.sh ├── train_utils │ ├── apply_r3_r4.py │ ├── fsdp_trainer.py │ ├── main.py │ ├── modeling_llama_quant.py │ ├── optimizer.py │ ├── quant_linear.py │ └── rtn_utils.py └── utils │ ├── convert_to_executorch.py │ ├── data_utils.py │ ├── eval_utils.py │ ├── fuse_norm_utils.py │ ├── hadamard_utils.py │ ├── model_utils.py │ ├── monkeypatch.py │ ├── process_args.py │ ├── quant_utils.py │ └── utils.py └── vit_quant ├── README.md ├── data_utils.py ├── eval_utils.py ├── gptaq_utils.py ├── gptq_utils.py ├── main.py ├── model_utils.py ├── quant_utils.py ├── run.sh └── utils.py /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/.gitignore -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/README.md -------------------------------------------------------------------------------- /fake_quant/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/README.md -------------------------------------------------------------------------------- /fake_quant/data_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/data_utils.py -------------------------------------------------------------------------------- /fake_quant/eval_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/eval_utils.py -------------------------------------------------------------------------------- /fake_quant/gptaq_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/gptaq_utils.py -------------------------------------------------------------------------------- /fake_quant/gptq_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/gptq_utils.py -------------------------------------------------------------------------------- /fake_quant/hadamard_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/hadamard_utils.py -------------------------------------------------------------------------------- /fake_quant/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/main.py -------------------------------------------------------------------------------- /fake_quant/model_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/model_utils.py -------------------------------------------------------------------------------- /fake_quant/monkeypatch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/monkeypatch.py -------------------------------------------------------------------------------- /fake_quant/quant_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/quant_utils.py -------------------------------------------------------------------------------- /fake_quant/requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/requirements.txt -------------------------------------------------------------------------------- /fake_quant/rotation_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/rotation_utils.py -------------------------------------------------------------------------------- /fake_quant/run_llama.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/run_llama.sh -------------------------------------------------------------------------------- /fake_quant/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/fake_quant/utils.py -------------------------------------------------------------------------------- /img/readme_intro.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/img/readme_intro.png -------------------------------------------------------------------------------- /spinquant/CODE_OF_CONDUCT.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/CODE_OF_CONDUCT.md -------------------------------------------------------------------------------- /spinquant/CONTRIBUTING.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/CONTRIBUTING.md -------------------------------------------------------------------------------- /spinquant/LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/LICENSE -------------------------------------------------------------------------------- /spinquant/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/README.md -------------------------------------------------------------------------------- /spinquant/SpinQuant.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/SpinQuant.png -------------------------------------------------------------------------------- /spinquant/eval_utils/gptaq_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/eval_utils/gptaq_utils.py -------------------------------------------------------------------------------- /spinquant/eval_utils/gptq_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/eval_utils/gptq_utils.py -------------------------------------------------------------------------------- /spinquant/eval_utils/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/eval_utils/main.py -------------------------------------------------------------------------------- /spinquant/eval_utils/modeling_llama.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/eval_utils/modeling_llama.py -------------------------------------------------------------------------------- /spinquant/eval_utils/rotation_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/eval_utils/rotation_utils.py -------------------------------------------------------------------------------- /spinquant/optimize_rotation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/optimize_rotation.py -------------------------------------------------------------------------------- /spinquant/ptq.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/ptq.py -------------------------------------------------------------------------------- /spinquant/requirement.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/requirement.txt -------------------------------------------------------------------------------- /spinquant/scripts/10_optimize_rotation.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/scripts/10_optimize_rotation.sh -------------------------------------------------------------------------------- /spinquant/scripts/11_optimize_rotation_fsdp.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/scripts/11_optimize_rotation_fsdp.sh -------------------------------------------------------------------------------- /spinquant/scripts/2_eval_ptq.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/scripts/2_eval_ptq.sh -------------------------------------------------------------------------------- /spinquant/scripts/31_optimize_rotation_executorch.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/scripts/31_optimize_rotation_executorch.sh -------------------------------------------------------------------------------- /spinquant/scripts/32_eval_ptq_executorch.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/scripts/32_eval_ptq_executorch.sh -------------------------------------------------------------------------------- /spinquant/train_utils/apply_r3_r4.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/train_utils/apply_r3_r4.py -------------------------------------------------------------------------------- /spinquant/train_utils/fsdp_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/train_utils/fsdp_trainer.py -------------------------------------------------------------------------------- /spinquant/train_utils/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/train_utils/main.py -------------------------------------------------------------------------------- /spinquant/train_utils/modeling_llama_quant.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/train_utils/modeling_llama_quant.py -------------------------------------------------------------------------------- /spinquant/train_utils/optimizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/train_utils/optimizer.py -------------------------------------------------------------------------------- /spinquant/train_utils/quant_linear.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/train_utils/quant_linear.py -------------------------------------------------------------------------------- /spinquant/train_utils/rtn_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/train_utils/rtn_utils.py -------------------------------------------------------------------------------- /spinquant/utils/convert_to_executorch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/convert_to_executorch.py -------------------------------------------------------------------------------- /spinquant/utils/data_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/data_utils.py -------------------------------------------------------------------------------- /spinquant/utils/eval_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/eval_utils.py -------------------------------------------------------------------------------- /spinquant/utils/fuse_norm_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/fuse_norm_utils.py -------------------------------------------------------------------------------- /spinquant/utils/hadamard_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/hadamard_utils.py -------------------------------------------------------------------------------- /spinquant/utils/model_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/model_utils.py -------------------------------------------------------------------------------- /spinquant/utils/monkeypatch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/monkeypatch.py -------------------------------------------------------------------------------- /spinquant/utils/process_args.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/process_args.py -------------------------------------------------------------------------------- /spinquant/utils/quant_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/quant_utils.py -------------------------------------------------------------------------------- /spinquant/utils/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/spinquant/utils/utils.py -------------------------------------------------------------------------------- /vit_quant/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/README.md -------------------------------------------------------------------------------- /vit_quant/data_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/data_utils.py -------------------------------------------------------------------------------- /vit_quant/eval_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/eval_utils.py -------------------------------------------------------------------------------- /vit_quant/gptaq_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/gptaq_utils.py -------------------------------------------------------------------------------- /vit_quant/gptq_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/gptq_utils.py -------------------------------------------------------------------------------- /vit_quant/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/main.py -------------------------------------------------------------------------------- /vit_quant/model_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/model_utils.py -------------------------------------------------------------------------------- /vit_quant/quant_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/quant_utils.py -------------------------------------------------------------------------------- /vit_quant/run.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/run.sh -------------------------------------------------------------------------------- /vit_quant/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Intelligent-Computing-Lab-Panda/GPTAQ/HEAD/vit_quant/utils.py --------------------------------------------------------------------------------