update requirements (#2957)

Author: Jintao-Huang

.dev_scripts/ci_container_test.sh

@@ -22,7 +22,7 @@ if [ "$MODELSCOPE_SDK_DEBUG" == "True" ]; then
 
     pip install -r requirements/framework.txt -U -i https://mirrors.aliyun.com/pypi/simple/
     pip install diffusers decord einops -U -i https://mirrors.aliyun.com/pypi/simple/
-    pip install autoawq!=0.2.7.post3 -U --no-deps
+    pip install autoawq -U --no-deps
 
     # test with install
     pip install .

requirements/eval.txt

@@ -1 +1,2 @@
-evalscope[all]>=0.5.2
+evalscope[opencompass]
+evalscope[vlmeval]

requirements/install_all.sh

@@ -1,8 +1,8 @@
 # please use python=3.10, cuda12.*
 # sh requirements/install_all.sh
 pip install "vllm>=0.5.1" -U
-pip install "lmdeploy>=0.5" -U --no-deps
-pip install autoawq!=0.2.7.post3 -U --no-deps
+pip install "lmdeploy>=0.5,<0.7" -U --no-deps
+pip install autoawq -U --no-deps
 pip install auto_gptq optimum bitsandbytes -U
 pip install git+https://github.com/modelscope/ms-swift.git#egg=ms-swift[all]
 pip install timm -U

swift/llm/export/quant.py

@@ -17,17 +17,17 @@ class QuantEngine(ProcessorMixin):
 
     def __init__(self, args: ExportArguments):
         self.args = args
-        if args.output_dir:
-            args.save_args()
         kwargs = {}
         if args.quant_method == 'awq':
             from awq import AutoAWQForCausalLM
             kwargs['automodel_class'] = AutoAWQForCausalLM
         self.model, self.template = prepare_model_template(args, **kwargs)
         self.template.set_mode('train')
 
-        HfConfigFactory.set_model_config_attr(self.model, 'use_cache', True)
+        HfConfigFactory.set_model_config_attr(self.model, 'use_cache', False)
         self.processor = self.template.processor
+        if args.output_dir:
+            args.save_args()
 
     def quantize(self):
         args = self.args