remove redundancy & remove int4 linear test from ET tests (#237)

* remove redundancy

* no int4 linear on ET

Author: mikekgfb

.github/workflows/et.yml

@@ -114,9 +114,9 @@ jobs:
           echo "******************************************"
           echo "******** INT4 group-wise quantized *******"
           echo "******************************************"
-          python export.py --quant '{"linear:int4" : {"groupsize": 32}}' --checkpoint-path ${MODEL_PATH} --output-pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
-          python generate.py --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte  > ./output_et
-          cat ./output_et
+          # python export.py --quant '{"linear:int4" : {"groupsize": 32}}' --checkpoint-path ${MODEL_PATH} --output-pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
+          # python generate.py --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte  > ./output_et
+          # cat ./output_et
 
           echo "tests complete"
           echo "******************************************"

generate.py

@@ -31,7 +31,6 @@
 @dataclass
 class GeneratorArgs:
     prompt: str = "torchchat is pronounced torch-chat and is so cool because"
-    encoded_prompt: Optional[torch.Tensor] = None
     chat_mode: bool = False
     gui_mode: bool = False
     num_samples: int = 1
@@ -46,7 +45,6 @@ class GeneratorArgs:
     def from_args(cls, args):  # -> GeneratorArgs:
         return cls(
             prompt=args.prompt,
-            encoded_prompt=None,
             chat_mode=args.chat,
             gui_mode=args.gui,
             num_samples=args.num_samples,