Merge branch 'main' into wangchang/inc3.x

intel · Jul 3, 2024 · 35b24f1 · 35b24f1
2 parents 81d0b35 + 86087dc
commit 35b24f1
Show file tree

Hide file tree

Showing 37 changed files with 783 additions and 409 deletions.
diff --git a/.github/workflows/script/formatScan/pylint.sh b/.github/workflows/script/formatScan/pylint.sh
@@ -28,7 +28,7 @@ else
     echo "Not found requirements.txt file."
 fi
 # install packages
-pip install lm-eval
+pip install lm-eval==0.4.2
 pip install accelerate nlpaug nltk schema optimum-intel optimum peft
 pip install --upgrade --force-reinstall transformers==4.36.2
 pip install optimum-habana

diff --git a/.github/workflows/script/install_binary.sh b/.github/workflows/script/install_binary.sh
@@ -4,6 +4,7 @@ source /intel-extension-for-transformers/.github/workflows/script/change_color.s
 cd /intel-extension-for-transformers
 export CMAKE_ARGS="-DNE_DNNL_CACHE_DIR=/cache"
 pip install -U pip
+pip install -r requirements.txt
 $BOLD_YELLOW && echo "---------------- git submodule update --init --recursive -------------" && $RESET
 git config --global --add safe.directory "*"
 git submodule update --init --recursive

diff --git a/docs/api_doc/optimization/config.rst b/docs/api_doc/optimization/config.rst
@@ -4,7 +4,4 @@ Config
 .. autoapisummary::
    intel_extension_for_transformers.transformers.utils.metrics
    intel_extension_for_transformers.transformers.utils.objectives
-   intel_extension_for_transformers.transformers.config
-   intel_extension_for_transformers.transformers.quantization
-   intel_extension_for_transformers.transformers.distillation
-   intel_extension_for_transformers.transformers.pruning
+   intel_extension_for_transformers.transformers.utils.config
diff --git a/docs/api_doc/optimization/tf_optimization.rst b/docs/api_doc/optimization/tf_optimization.rst
diff --git a/docs/api_doc/user_api.rst b/docs/api_doc/user_api.rst
@@ -7,7 +7,5 @@ The following Python API information is available:
    :maxdepth: 1
 
    optimization/trainer.rst
-   optimization/optimizer.rst
    optimization/model.rst
-   optimization/tf_optimization.rst
    optimization/config.rst
diff --git a/examples/.config/pytorch_optimize.json b/examples/.config/pytorch_optimize.json
@@ -2450,6 +2450,58 @@
       }
     }
   },
+  "phi_2b_gen_ipex_static": {
+    "working_dir": "huggingface/pytorch/text-generation/quantization",
+    "tune": {
+      "cmd": "bash run_tuning.sh",
+      "params": {
+        "topology": "phi_2b",
+        "task": "generation",
+        "approach": "static",
+        "output_model": "saved_results"
+      }
+    },
+    "benchmark": {
+      "cmd": "bash run_benchmark.sh",
+      "params": {
+        "topology": "phi_2b",
+        "task": "generation",
+        "approach": "static",
+        "backend": "ipex",
+        "mode": "benchmark",
+        "batch_size": "112",
+        "iters": "100",
+        "int8": "false",
+        "config": "saved_results"
+      }
+    }
+  },
+  "phi_3b_gen_ipex_static": {
+    "working_dir": "huggingface/pytorch/text-generation/quantization",
+    "tune": {
+      "cmd": "bash run_tuning.sh",
+      "params": {
+        "topology": "phi_3b",
+        "task": "generation",
+        "approach": "static",
+        "output_model": "saved_results"
+      }
+    },
+    "benchmark": {
+      "cmd": "bash run_benchmark.sh",
+      "params": {
+        "topology": "phi_3b",
+        "task": "generation",
+        "approach": "static",
+        "backend": "ipex",
+        "mode": "benchmark",
+        "batch_size": "112",
+        "iters": "100",
+        "int8": "false",
+        "config": "saved_results"
+      }
+    }
+  },
   "flan-t5-large_gen_ipex_static": {
     "working_dir": "huggingface/pytorch/text2text-generation",
     "tune": {

diff --git a/examples/.config/tensorflow_optimize.json b/examples/.config/tensorflow_optimize.json
diff --git a/examples/huggingface/neural_speed/requirements.txt b/examples/huggingface/neural_speed/requirements.txt
@@ -1,12 +1,11 @@
 intel_extension_for_transformers
 neural-speed
-lm-eval
+lm-eval==0.4.2
 sentencepiece
 gguf
 --extra-index-url https://download.pytorch.org/whl/cpu
 torch==2.3.0+cpu
 transformers
-intel_extension_for_pytorch==2.3.0
 tiktoken
 transformers_stream_generator
-zipfile38
+zipfile38
diff --git a/examples/huggingface/pytorch/language-modeling/inference/requirements.txt b/examples/huggingface/pytorch/language-modeling/inference/requirements.txt
@@ -1,4 +1,4 @@
 transformers 
 accelerate
 sentencepiece != 0.1.92
-lm-eval
+lm-eval==0.4.2
diff --git a/examples/huggingface/pytorch/language-modeling/pruning/requirements.txt b/examples/huggingface/pytorch/language-modeling/pruning/requirements.txt
@@ -7,5 +7,5 @@ transformers
 torch==2.0.1
 tqdm
 neural_compressor
-lm-eval
+lm-eval==0.4.2
 
diff --git a/examples/huggingface/pytorch/language-modeling/quantization/requirements.txt b/examples/huggingface/pytorch/language-modeling/quantization/requirements.txt
@@ -9,5 +9,5 @@ wandb
 einops
 neural-compressor
 pytest==8.0.0
-lm-eval
+lm-eval==0.4.2
 git+https://github.com/huggingface/peft.git@6c44096c7b8d55a2ecf24be9bc68393467e1584a
diff --git a/...e/pytorch/text-classification/deployment/emotion/distilbert_base_uncased/requirements.txt b/...e/pytorch/text-classification/deployment/emotion/distilbert_base_uncased/requirements.txt
@@ -1,4 +1,5 @@
-neural-compressor
+intel-extension-for-transformers==1.4.2
+neural-compressor==2.6
 transformers
 accelerate
 datasets >= 1.8.0