add an interface to predict the latency of compressed model (PaddlePaddle#1045)

* add an interface to predict the latency of compressed model

* just for rerun

* delete temporary model files

Co-authored-by: ceci3 <[email protected]>

Author: zzjjay

paddleslim/auto_compression/utils/predict.py

@@ -0,0 +1,121 @@
+import os
+import paddle
+from paddleslim.analysis import TableLatencyPredictor
+from .prune_model import get_sparse_model, get_prune_model
+from .fake_ptq import post_quant_fake
+import shutil
+
+
+def predict_compressed_model(model_file, param_file, hardware='SD710'):
+    """
+    Evaluating the latency of the model under various compression strategies.
+    Args:
+        model_file(str), param_file(str): The inference model to be compressed.
+        hardware(str): Target device.
+    Returns:
+        latency_dict(dict): The latency latency of the model under various compression strategies.
+    """
+    latency_dict = {}
+
+    model_filename = model_file.split('/')[-1]
+    param_filename = param_file.split('/')[-1]
+
+    predictor = TableLatencyPredictor(hardware)
+    latency = predictor.predict(
+        model_file=model_file, param_file=param_file, data_type='fp32')
+    latency_dict.update({'origin_fp32': latency})
+    paddle.enable_static()
+    place = paddle.CPUPlace()
+    exe = paddle.static.Executor(place)
+    post_quant_fake(
+        exe,
+        model_dir=os.path.dirname(model_file),
+        model_filename=model_filename,
+        params_filename=param_filename,
+        save_model_path='quant_model',
+        quantizable_op_type=["conv2d", "depthwise_conv2d", "mul"],
+        is_full_quantize=False,
+        activation_bits=8,
+        weight_bits=8)
+    quant_model_file = os.path.join('quant_model', model_filename)
+    quant_param_file = os.path.join('quant_model', param_filename)
+
+    latency = predictor.predict(
+        model_file=quant_model_file,
+        param_file=quant_param_file,
+        data_type='int8')
+    latency_dict.update({f'origin_int8': latency})
+
+    for prune_ratio in [0.3, 0.4, 0.5, 0.6]:
+        get_prune_model(
+            model_file=model_file,
+            param_file=param_file,
+            ratio=prune_ratio,
+            save_path='prune_model')
+        prune_model_file = os.path.join('prune_model', model_filename)
+        prune_param_file = os.path.join('prune_model', param_filename)
+
+        latency = predictor.predict(
+            model_file=prune_model_file,
+            param_file=prune_param_file,
+            data_type='fp32')
+        latency_dict.update({f'prune_{prune_ratio}_fp32': latency})
+
+        post_quant_fake(
+            exe,
+            model_dir='prune_model',
+            model_filename=model_filename,
+            params_filename=param_filename,
+            save_model_path='quant_model',
+            quantizable_op_type=["conv2d", "depthwise_conv2d", "mul"],
+            is_full_quantize=False,
+            activation_bits=8,
+            weight_bits=8)
+        quant_model_file = os.path.join('quant_model', model_filename)
+        quant_param_file = os.path.join('quant_model', param_filename)
+
+        latency = predictor.predict(
+            model_file=quant_model_file,
+            param_file=quant_param_file,
+            data_type='int8')
+        latency_dict.update({f'prune_{prune_ratio}_int8': latency})
+
+    for sparse_ratio in [0.70, 0.75, 0.80, 0.85, 0.90, 0.95]:
+        get_sparse_model(
+            model_file=model_file,
+            param_file=param_file,
+            ratio=sparse_ratio,
+            save_path='sparse_model')
+        sparse_model_file = os.path.join('sparse_model', model_filename)
+        sparse_param_file = os.path.join('sparse_model', param_filename)
+
+        latency = predictor.predict(
+            model_file=sparse_model_file,
+            param_file=sparse_param_file,
+            data_type='fp32')
+        latency_dict.update({f'sparse_{sparse_ratio}_fp32': latency})
+
+        post_quant_fake(
+            exe,
+            model_dir='sparse_model',
+            model_filename=model_filename,
+            params_filename=param_filename,
+            save_model_path='quant_model',
+            quantizable_op_type=["conv2d", "depthwise_conv2d", "mul"],
+            is_full_quantize=False,
+            activation_bits=8,
+            weight_bits=8)
+        quant_model_file = os.path.join('quant_model', model_filename)
+        quant_param_file = os.path.join('quant_model', param_filename)
+
+        latency = predictor.predict(
+            model_file=quant_model_file,
+            param_file=quant_param_file,
+            data_type='int8')
+        latency_dict.update({f'sparse_{prune_ratio}_int8': latency})
+
+    # Delete temporary model files
+    shutil.rmtree('./quant_model')
+    shutil.rmtree('./prune_model')
+    shutil.rmtree('./sparse_model')
+    return latency_dict

paddleslim/auto_compression/utils/prune_model.py

@@ -27,7 +27,6 @@ def get_sparse_model(model_file, param_file, ratio, save_path):
 
     folder = os.path.dirname(model_file)
     model_name = model_file.split('/')[-1]
-    model_name = model_file.split('/')[-1]
     if param_file is None:
         param_name = None
     else: