Migrate SQ and WOQ to INC 3.x API.

.github/checkgroup.yml

@@ -30,40 +30,6 @@ subprojects:
       - "optimize-unit-test-PR-test"
       - "Genreate-OptimizeUT-Report"
 
-  - id: "NeuralChat Unit Test"
-    paths:
-      - ".github/workflows/unit-test-neuralchat.yml"
-      - ".github/workflows/script/unitTest/run_unit_test_neuralchat.sh"
-      - "intel_extension_for_transformers/neural_chat/**"
-      - "requirements.txt"
-      - "setup.py"
-      - "intel_extension_for_transformers/transformers/llm/finetuning/**"
-      - "intel_extension_for_transformers/transformers/llm/quantization/**"
-      - "intel_extension_for_transformers/transformers/**"
-      - "intel_extension_for_transformers/langchain/**"
-      - "!intel_extension_for_transformers/neural_chat/docs/**"
-      - "!intel_extension_for_transformers/neural_chat/examples/**"
-      - "!intel_extension_for_transformers/neural_chat/assets/**"
-      - "!intel_extension_for_transformers/neural_chat/README.md"
-    checks:
-      - "neuralchat-unit-test-baseline"
-      - "neuralchat-unit-test-PR-test"
-      - "Generate-NeuralChat-Report"
-
-  - id: "Engine Unit Test workflow"
-    paths:
-      - ".github/workflows/unit-test-engine.yml"
-      - "requirements.txt"
-      - "setup.py"
-      - intel_extension_for_transformers/transformers/**
-      - "intel_extension_for_transformers/transformers/runtime/**"
-      - "!intel_extension_for_transformers/transformers/runtime/kernels/**"
-      - "!intel_extension_for_transformers/transformers/runtime/third_party/**"
-      - "!intel_extension_for_transformers/transformers/runtime/docs/**"
-    checks:
-      - "engine-unit-test-baseline"
-      - "engine-unit-test-PR-test"
-      - "Genreate-Engine-Report"
 
   # - id: "Windows Binary Test"
   #   paths:

.github/workflows/script/unitTest/env_setup.sh

@@ -13,6 +13,7 @@ until [ "$n" -ge 5 ]; do
     git clone https://github.com/intel/neural-compressor.git /neural-compressor
     cd /neural-compressor
     pip install -r requirements.txt
+    pip install -r requirements_pt.txt
     python setup.py install && break
     n=$((n + 1))
     sleep 5

examples/huggingface/pytorch/text-generation/quantization/README.md

@@ -36,21 +36,18 @@ OMP_NUM_THREADS=<physical cores num> numactl -m <node N> -C <cpu list> python ru
     --model <MODEL_NAME_OR_PATH> \
     --sq \
     --output_dir <SQ_MODEL_SAVE_PATH> \ # Default is "./saved_results."
-    --int8 \
     --benchmark \
     --batch_size 1
 # load SQ model quantied by itrex and do benchmark.
 OMP_NUM_THREADS=<physical cores num> numactl -m <node N> -C <cpu list> python run_generation_sq.py \
     --model <SQ_MODEL_SAVE_PATH> \
-    --int8 \
     --benchmark \
     --batch_size 1
 # load SQ model quantied configure.json and do benchmark.
 python run_generation_sq.py \
     --model <MODEL_NAME_OR_PATH> \
     --output_dir <SQ_MODEL_SAVE_PATH> \
-    --int8 \
-    --restore \
+    --restore_sq_model_from_json \
     --benchmark \
     --batch_size 1
 ```
@@ -68,23 +65,20 @@ python run_generation_sq.py \
     --model <MODEL_NAME_OR_PATH> \
     --sq \
     --output_dir <SQ_MODEL_SAVE_PATH> \ # Default is "./saved_results."
-    --int8 \
     --accuracy \
     --batch_size 56 
 
 # load SQ model quantied by itrex and do benchmark.
 python run_generation_sq.py \
     --model <SQ_MODEL_SAVE_PATH> \
-    --int8 \
     --accuracy \
     --batch_size 56 
 
 # load SQ model quantied configure.json and do benchmark.
 python run_generation_sq.py \
     --model <MODEL_NAME_OR_PATH> \
     --output_dir <SQ_MODEL_SAVE_PATH> \
-    --int8 \
-    --restore \
+    --restore_sq_model_from_json \
     --accuracy \
     --batch_size 56