update

Signed-off-by: yiliu30 <yi4.liu@intel.com>
pytorch · Sep 11, 2024 · 77889a1 · 77889a1
1 parent d9580e3
commit 77889a1
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 9 deletions.
diff --git a/torchao/_models/llama/benchmarks.sh b/torchao/_models/llama/benchmarks.sh
@@ -50,7 +50,6 @@ python generate.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --co
 python generate.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --compile --quantization uintx-4-64 --write_result benchmark_results.txt
 python generate.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --compile --quantization uintx-2-8 --write_result benchmark_results.txt
 
-
 export MODEL_REPO=meta-llama/Meta-Llama-3-8B
 python generate.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --precision torch.float32 --write_result benchmark_results.txt
 python generate.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --write_result benchmark_results.txt

diff --git a/torchao/_models/llama/evals.sh b/torchao/_models/llama/evals.sh
@@ -1,13 +1,9 @@
 export CHECKPOINT_PATH=../../../checkpoints # path to checkpoints folder
 
 export MODEL_REPO=meta-llama/Llama-2-7b-chat-hf
-# auto-round w/o quant_lm_head
-python eval.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --quantization autoround
-# auto-round w/ quant_lm_head
-python eval.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --quantization autoround-cuda-1
+python eval.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --quantization autoround        # auto-round w/o quant_lm_head
+python eval.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --quantization autoround-cuda-1 # auto-round w/ quant_lm_head
 
 export MODEL_REPO=meta-llama/Meta-Llama-3-8B
-# auto-round w/o quant_lm_head
-python eval.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --quantizatio autoround-cpu
-# auto-round w/ quant_lm_head
-python eval.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --quantization autoround-cpu-1
+python eval.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --quantizatio autoround-cpu    # auto-round w/o quant_lm_head
+python eval.py --checkpoint_path $CHECKPOINT_PATH/$MODEL_REPO/model.pth --quantization autoround-cpu-1 # auto-round w/ quant_lm_head