Support qwen2-1.5b with fused decoderlayer optimization on NPU (#11888)

python/llm/example/NPU/HF-Transformers-AutoModels/LLM/README.md

@@ -78,7 +78,7 @@ done
 ```
 
 ## Example 2: Predict Tokens using `generate()` API using multi processes
-In the example [llama2.py](./llama2.py), we show an experimental support for a Llama2 model to predict the next N tokens using `generate()` API, with IPEX-LLM INT4 optimization and fused decoderlayer optimization on Intel NPUs.
+In the example [llama2.py](./llama2.py) and [qwen2.py](./qwen2.py), we show an experimental support for a Llama2 / Qwen2 model to predict the next N tokens using `generate()` API, with IPEX-LLM INT4 optimization and fused decoderlayer optimization on Intel NPUs.
 ### 1. Install
 #### 1.1 Installation on Windows
 We suggest using conda to manage environment:
@@ -111,7 +111,11 @@ set BIGDL_USE_NPU=1
 ### 3. Running examples
 
 ```
+# to run Llama-2-7b-chat-hf
 python  llama2.py
+
+# to run Qwen2-1.5B-Instruct
+python qwen2.py
 ```
 
 Arguments info:

python/llm/example/NPU/HF-Transformers-AutoModels/LLM/qwen2.py

@@ -0,0 +1,95 @@
+#
+# Copyright 2016 The BigDL Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+import os
+import torch
+import time
+import argparse
+
+from ipex_llm.transformers.npu_model import AutoModelForCausalLM
+from transformers import AutoTokenizer
+
+from transformers.utils import logging
+
+logger = logging.get_logger(__name__)
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(
+        description="Predict Tokens using `generate()` API for npu model"
+    )
+    parser.add_argument(
+        "--repo-id-or-model-path",
+        type=str,
+        default="Qwen/Qwen2-1.5B-Instruct",
+        help="The huggingface repo id for the Qwen2 model to be downloaded"
+        ", or the path to the huggingface checkpoint folder",
+    )
+    parser.add_argument('--prompt', type=str, default="What is AI?",
+                        help='Prompt to infer')
+    parser.add_argument("--n-predict", type=int, default=32, help="Max tokens to predict")
+    parser.add_argument("--max-output-len", type=int, default=1024)
+    parser.add_argument("--max-prompt-len", type=int, default=512)
+    parser.add_argument("--disable-transpose-value-cache", action="store_true", default=False)
+    parser.add_argument("--intra-pp", type=int, default=2)
+    parser.add_argument("--inter-pp", type=int, default=1)
+
+    args = parser.parse_args()
+    model_path = args.repo_id_or_model_path
+
+    model = AutoModelForCausalLM.from_pretrained(
+        model_path,
+        torch_dtype=torch.float16,
+        trust_remote_code=True,
+        attn_implementation="eager",
+        load_in_low_bit="sym_int4",
+        enable_mp=True,
+        max_output_len=args.max_output_len,
+        max_prompt_len=args.max_prompt_len,
+        intra_pp=args.intra_pp,
+        inter_pp=args.inter_pp,
+        transpose_value_cache=not args.disable_transpose_value_cache,
+    )
+
+    tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)
+
+    print("-" * 80)
+    print("done")
+    messages = [{"role": "system", "content": "You are a helpful assistant."},
+                {"role": "user", "content": args.prompt}]
+    text = tokenizer.apply_chat_template(messages,
+                                         tokenize=False,
+                                         add_generation_prompt=True)
+    with torch.inference_mode():
+        print("finish to load")
+        for i in range(3):
+            _input_ids = tokenizer([text], return_tensors="pt").input_ids
+            print("input length:", len(_input_ids[0]))
+            st = time.time()
+            output = model.generate(
+                _input_ids, num_beams=1, do_sample=False, max_new_tokens=args.n_predict
+            )
+            end = time.time()
+            print(f"Inference time: {end-st} s")
+            input_str = tokenizer.decode(_input_ids[0], skip_special_tokens=False)
+            print("-" * 20, "Input", "-" * 20)
+            print(input_str)
+            output_str = tokenizer.decode(output[0], skip_special_tokens=False)
+            print("-" * 20, "Output", "-" * 20)
+            print(output_str)
+
+    print("-" * 80)
+    print("done")
+    print("success shut down")

python/llm/src/ipex_llm/transformers/npu_models/convert_mp.py

@@ -54,3 +54,26 @@ def optimize_llm(
             prefill_runner=prefill_runner, decode_runner=decode_runner
         )
         convert_forward(model, LlamaModel, llama_model_forward)
+    elif model.config.model_type == "qwen2" and model.config.intermediate_size == 8960:
+        # for qwen2-1.5B
+        from ipex_llm.transformers.npu_models.qwen2_mp import gen_qwen2_fused_model_forward
+        from ipex_llm.transformers.npu_models.qwen2_mp import DecodeRunner, PrefillRunner
+        from transformers.models.qwen2.modeling_qwen2 import Qwen2Model
+
+        decode_runner = DecodeRunner(
+            model,
+            max_seq_len=max_output_len,
+            inter_pp=inter_pp,
+            intra_pp=intra_pp,
+            transpose_value_cache=transpose_value_cache,
+        )
+        prefill_runner = PrefillRunner(
+            model,
+            max_output_len=max_output_len,
+            max_prompt_len=max_prompt_len,
+            transpose_value_cache=transpose_value_cache,
+        )
+        qwen2_model_forward = gen_qwen2_fused_model_forward(
+            prefill_runner=prefill_runner, decode_runner=decode_runner
+        )
+        convert_forward(model, Qwen2Model, qwen2_model_forward)

python/llm/src/ipex_llm/transformers/npu_models/llama_mp.py

@@ -17,20 +17,11 @@
 import os
 import torch
 import time
-import argparse
 
-from ipex_llm.transformers.npu_model import AutoModelForCausalLM
-from transformers import AutoTokenizer
-from intel_npu_acceleration_library.backend.factory import NNFactory
 from typing import Optional, Sequence, List, Union, Any, Tuple
 import numpy as np
-import math
-from intel_npu_acceleration_library.backend.runtime import set_contiguous, record_function
-from intel_npu_acceleration_library.backend.runtime import adapt_output_tensor, _model_cache
-from collections import deque
+
 from transformers.cache_utils import Cache
-from intel_npu_acceleration_library.backend.bindings import lib as backend_lib
-import ctypes
 from ipex_llm.utils.common import invalidInputError
 from typing import Optional, List, Generator
 import uuid
@@ -38,12 +29,10 @@ from functools import partial
 import torch.nn.functional as F
 import torch.nn.parallel
 import torch.distributed as dist
-from filelock import FileLock
 
 from transformers.utils import logging
 
 logger = logging.get_logger(__name__)
-import gc
 from colorama import Fore, Back, Style
 import torch.multiprocessing as mp
 from transformers.cache_utils import Cache

python/llm/src/ipex_llm/transformers/npu_models/mp_models_base.py

@@ -118,7 +118,10 @@ class LLMBaseNNFactory(NNFactory):
                   num_heads,
                   num_key_value_heads,
                   head_dim,
-                  seq_len):
+                  seq_len,
+                  q_bias=None,
+                  k_bias=None,
+                  v_bias=None):
         hidden_size = num_heads * head_dim
         num_key_value_groups = num_heads // num_key_value_heads
         query_states = self.linear(
@@ -128,6 +131,8 @@ class LLMBaseNNFactory(NNFactory):
             bias=False,
             wt_dtype=self.dtype,
         )
+        if q_bias is not None:
+            query_states = query_states + q_bias
         key_states = self.linear(
             hidden_states,
             num_key_value_heads * head_dim,
@@ -135,6 +140,8 @@ class LLMBaseNNFactory(NNFactory):
             bias=False,
             wt_dtype=self.dtype,
         )
+        if k_bias is not None:
+            key_states = key_states + k_bias
         value_states = self.linear(
             hidden_states,
             num_key_value_heads * head_dim,
@@ -142,6 +149,8 @@ class LLMBaseNNFactory(NNFactory):
             bias=False,
             wt_dtype=self.dtype,
         )
+        if v_bias is not None:
+            value_states = value_states + v_bias
 
         query_states = self.reshape(
             query_states, [1, seq_len, num_heads, head_dim]
@@ -192,7 +201,8 @@ class LLMBaseNNFactory(NNFactory):
                                       n_rep=num_key_value_groups,
                                       num_key_value_heads=num_key_value_heads,
                                       kv_seq_len=kv_seq_len,
-                                      head_dim=head_dim,)
+                                      head_dim=head_dim,
+                                      transpose=self.transpose_value)
         attn_weight = self.matmul(query_states, key_states, False, True) / (
             math.sqrt(head_dim)
         )

python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py

@@ -0,0 +1,983 @@
+#
+# Copyright 2016 The BigDL Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+import os
+import torch
+import time
+
+from typing import Optional, Sequence, List, Union, Any, Tuple
+import numpy as np
+
+from transformers.cache_utils import Cache
+from ipex_llm.utils.common import invalidInputError
+from typing import Optional, List, Generator
+import uuid
+from functools import partial
+import torch.nn.functional as F
+import torch.nn.parallel
+import torch.distributed as dist
+
+from transformers.utils import logging
+
+logger = logging.get_logger(__name__)
+from colorama import Fore, Back, Style
+import torch.multiprocessing as mp
+from transformers.cache_utils import Cache
+from transformers.modeling_outputs import BaseModelOutputWithPast
+from ipex_llm.transformers.npu_models.mp_models_base import run_model
+from ipex_llm.transformers.npu_models.mp_models_base import LLMBaseNNFactory
+
+
+class LowBitQwenMultiDecoderlayer(LLMBaseNNFactory):
+    def __init__(
+        self,
+        # batch_size: int,
+        # seq_len: int,
+        # hidden_size: int,
+        hidden_shape: Sequence[int],
+        *shapes,
+        num_heads: int,
+        num_key_value_heads: int,
+        num_layers: int,
+        cached_cos,
+        cached_sin,
+        input_layernorm_weights=None,
+        post_attn_layernorm_weights=None,
+        q_biases=None,
+        k_biases=None,
+        v_biases=None,
+        mode: str = "prefill",
+        dtype: np.dtype = np.int8,
+        max_seq_len: int = 1024,
+        transpose_value: bool = False,
+        profile: bool = False,
+        device: str = "NPU",
+        rms_norm_eps,
+        intermediate_size,
+    ):
+        super().__init__(max_seq_len=max_seq_len,
+                         transpose_value=transpose_value,
+                         dtype=dtype,
+                         profile=profile,
+                         device=device)
+        self.max_seq_len = max_seq_len
+        self.intermediate_size = intermediate_size
+        self.dtype = dtype
+        self.cached_cos = cached_cos
+        self.cached_sin = cached_sin
+        self.batch_size, self.seq_len, self.hidden_size = hidden_shape
+        self.mode = mode
+        self.rms_norm_eps = rms_norm_eps
+        self.transpose_value = transpose_value
+        self.num_layers = num_layers
+
+        cos = self.constant(self.cached_cos)
+        self.cos = self.unsqueeze(cos, axis=0)
+
+        sin = self.constant(self.cached_sin)
+        self.sin = self.unsqueeze(sin, axis=0)
+
+        if mode == "decode":
+            self.kv_seq_len = self.max_seq_len + 1
+        else:
+            self.kv_seq_len = self.seq_len
+
+        self.num_heads = num_heads
+        self.num_key_value_heads = num_key_value_heads
+
+        self.head_dim = self.hidden_size // self.num_heads
+        self.num_key_value_groups = self.num_heads // self.num_key_value_heads
+
+        # define input, the order self.parameter matters
+        input = self.create_input_op((self.batch_size, self.seq_len, self.hidden_size))
+
+        # Self Attention
+        if mode == "decode":
+            attention_mask = self.create_input_op((self.batch_size, 1, 1, self.max_seq_len + 1))
+        else:
+            attention_mask = self.create_input_op((self.batch_size, 1, self.seq_len, self.seq_len))
+
+        position_ids = self.create_input_op((self.batch_size, self.seq_len))
+        past_keys = []
+        past_values = []
+        if mode == "decode":
+            for i in range(num_layers):
+                past_key = self.create_cache_op(
+                    (self.batch_size, self.num_key_value_heads, self.max_seq_len, self.head_dim)
+                )
+                if transpose_value:
+                    past_value = self.create_cache_op(
+                        (self.batch_size, self.num_key_value_heads, self.head_dim, self.max_seq_len)
+                    )
+                else:
+                    past_value = self.create_cache_op(
+                        (self.batch_size, self.num_key_value_heads, self.max_seq_len, self.head_dim)
+                    )
+                past_keys.append(past_key)
+                past_values.append(past_value)
+        else:
+            past_keys = [None] * num_layers
+            past_values = [None] * num_layers
+
+        if input_layernorm_weights is None:
+            input_layernorm_weights = []
+            post_attn_layernorm_weights = []
+            for i in range(num_layers):
+                input_layernorm_weights.append(
+                    self.create_input_op(
+                        (
+                            1,
+                            self.hidden_size,
+                        )
+                    )
+                )
+                post_attn_layernorm_weights.append(
+                    self.create_input_op(
+                        (
+                            1,
+                            self.hidden_size,
+                        )
+                    )
+                )
+        else:
+            input_layernorm_weights = [self.constant(w) for w in input_layernorm_weights]
+            post_attn_layernorm_weights = [self.constant(w) for w in post_attn_layernorm_weights]
+
+        if q_biases is None:
+            q_biases = []
+            k_biases = []
+            v_biases = []
+            for i in range(num_layers):
+                q_biases.append(self.create_input_op((self.num_heads * self.head_dim,)))
+                k_biases.append(self.create_input_op((self.num_key_value_heads * self.head_dim,)))
+                v_biases.append(self.create_input_op((self.num_key_value_heads * self.head_dim,)))
+        else:
+            q_biases = [self.constant(w) for w in q_biases]
+            k_biases = [self.constant(w) for w in k_biases]
+            v_biases = [self.constant(w) for w in v_biases]
+
+        hidden_states = input
+
+        curr_key_values = []
+        for i in range(num_layers):
+            hidden_states, new_key_states, new_value_states = self.build_decoder(
+                hidden_states=hidden_states,
+                attention_mask=attention_mask,
+                position_ids=position_ids,
+                input_layernorm_weight=input_layernorm_weights[i],
+                post_attention_layernorm_weight=post_attn_layernorm_weights[i],
+                q_bias=q_biases[i],
+                k_bias=k_biases[i],
+                v_bias=v_biases[i],
+                past_key=past_keys[i],
+                past_value=past_values[i],
+            )
+            curr_key_values.append((new_key_states, new_value_states))
+
+        # define outputs
+        hidden_states = self.convert_to_fp16(hidden_states)
+
+        for i in range(num_layers):
+            new_key_states = self.convert_to_fp16(curr_key_values[i][0])
+            new_value_states = self.convert_to_fp16(curr_key_values[i][1])
+
+        self.compile()
+
+    def build_decoder(
+        self,
+        hidden_states,
+        attention_mask,
+        position_ids,
+        input_layernorm_weight,
+        post_attention_layernorm_weight,
+        q_bias,
+        k_bias,
+        v_bias,
+        past_key=None,
+        past_value=None,
+    ):
+
+        residual = hidden_states
+        input_2d = self.reshape(hidden_states, (self.batch_size * self.seq_len, self.hidden_size))
+        input_2d = self.layer_norm(input_2d, input_layernorm_weight)
+        attn_output, new_key_states, new_value_states = self.attention(
+            hidden_states=input_2d,
+            position_ids=position_ids,
+            attention_mask=attention_mask,
+            past_key=past_key,
+            past_value=past_value,
+            cos=self.cos,
+            sin=self.sin,
+            mode=self.mode,
+            num_heads=self.num_heads,
+            num_key_value_heads=self.num_key_value_heads,
+            head_dim=self.head_dim,
+            seq_len=self.seq_len,
+            q_bias=q_bias,
+            k_bias=k_bias,
+            v_bias=v_bias,
+        )
+        hidden_states = self.eltwise_add(residual, attn_output)
+        residual = hidden_states
+        hidden_states = self.layer_norm(hidden_states, post_attention_layernorm_weight)
+        hidden_states = self.mlp(hidden_states)
+        hidden_states = self.eltwise_add(residual, hidden_states)
+        hidden_states = self.convert_to_fp16(hidden_states)
+
+        return hidden_states, new_key_states, new_value_states
+
+
+class FusedQwenLowBitMultiDecoderlayer(torch.nn.Module):
+
+    def __init__(
+        self,
+        parameters: List[Tuple[torch.Tensor]],
+        input_laynorm_weights: List[torch.Tensor],
+        post_attn_layernorm_weights: List[torch.Tensor],
+        q_biases: List[torch.Tensor],
+        k_biases: List[torch.Tensor],
+        v_biases: List[torch.Tensor],
+        layer_indexes: List[int],
+        intra_stages: int,
+        cached_cos: torch.Tensor,
+        cached_sin: torch.Tensor,
+        num_heads: int,
+        head_dim: int,
+        num_key_value_heads: int,
+        rms_norm_eps,
+        intermediate_size,
+        max_seq_len: int = 1024,
+        transpose_value: bool = False,
+        do_print: bool = False,
+    ):
+        super().__init__()
+
+        self.do_print = do_print
+
+        op_parameters = []
+        for w in parameters:
+            if isinstance(w, tuple):  # from QuantizedLinear
+                op_parameters.append((w[0].numpy(), w[1].numpy()))
+            else:
+                op_parameters.append(w.to(torch.float16).numpy())
+        self.op_parameters = op_parameters
+        self.op_id = str(uuid.uuid4())
+        self.max_seq_len = max_seq_len
+        self.transpose_value = transpose_value
+        if isinstance(parameters[0], tuple):
+            np_dtype = np.int8 if parameters[0][0].dtype == torch.int8 else np.uint8
+        else:  # FP16 Linear
+            np_dtype = np.float16
+
+        self.intra_stages = intra_stages
+        self.layer_indexes = layer_indexes
+        num_layers = len(self.layer_indexes) // intra_stages
+        self.layer_ranges = []
+        for i in range(intra_stages):
+            if i == intra_stages - 1:
+                self.layer_ranges.append((i * num_layers, len(self.layer_indexes)))
+            else:
+                self.layer_ranges.append((i * num_layers, (i + 1) * num_layers))
+
+        self.backend_decoders = []
+
+        for i in range(intra_stages):
+            start, end = self.layer_ranges[i]
+            lm_0 = input_laynorm_weights[start:end]
+            lm_1 = post_attn_layernorm_weights[start:end]
+            decoder = LowBitQwenMultiDecoderlayer(
+                [1, 1, num_heads * head_dim],
+                input_layernorm_weights=lm_0,
+                post_attn_layernorm_weights=lm_1,
+                q_biases=q_biases[start:end],
+                k_biases=k_biases[start:end],
+                v_biases=v_biases[start:end],
+                cached_cos=cached_cos,
+                cached_sin=cached_sin,
+                num_heads=num_heads,
+                num_key_value_heads=num_key_value_heads,
+                num_layers=end - start,
+                max_seq_len=max_seq_len,
+                rms_norm_eps=rms_norm_eps,
+                intermediate_size=intermediate_size,
+                mode="decode",
+                transpose_value=self.transpose_value,
+                dtype=np_dtype,
+            )
+            self.backend_decoders.append(decoder)
+
+        for i in range(intra_stages):
+            start, end = self.layer_ranges[i]
+            self.backend_decoders[i].set_weights(self.op_id, op_parameters[start * 7:end * 7])
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        **kwargs,
+    ) -> torch.Tensor:
+
+        inputs = (
+            hidden_states.to(torch.float16),
+            attention_mask,
+            position_ids,
+        )
+
+        for i in range(self.intra_stages):
+            start, end = self.layer_ranges[i]
+            self.backend_decoders[i].update_cache(past_key_value, self.layer_indexes[start:end])
+
+        hidden_states, new_keys, new_values = LowBitQwenMultiDecoderlayer.run_decoders(
+            inputs,
+            decoders=self.backend_decoders)
+
+        if self.do_print:
+            print("outputs:", hidden_states)
+
+        outputs = (hidden_states,)
+        outputs += (past_key_value, new_keys, new_values)
+        return outputs
+
+    def post_forward(self, past_key_value, new_keys, new_values):
+        key_value_states = []
+        for i in range(self.intra_stages):
+            for j in range(1, len(self.backend_decoders[i].torch_out)):
+                key_value_states.append(self.backend_decoders[i].torch_out[j])
+
+        cache_kwargs = {
+            "max_seq_len": self.max_seq_len,
+            "transpose": self.transpose_value,
+        }
+        for i in range(len(self.layer_indexes)):
+            key_states, value_states = past_key_value.update(
+                new_keys[i],
+                new_values[i],
+                self.layer_indexes[i],
+                cache_kwargs,
+            )
+
+        for i in range(self.intra_stages):
+            self.backend_decoders[i].load_cache_async()
+
+
+class FusedQwenLowBitDecoderlayer(torch.nn.Module):
+    def __init__(
+        self,
+        parameters: List[torch.Tensor],
+        cached_cos,
+        cached_sin,
+        layer_norm_0,
+        layer_norm_1,
+        q_bias,
+        k_bias,
+        v_bias,
+        num_heads: int,
+        num_key_value_heads: int,
+        layer_idx: int,
+        rms_norm_eps,
+        intermediate_size,
+        max_seq_len: int = 128,
+        transpose_value: bool = False,
+    ):
+        super().__init__()
+        self.op_parameters = parameters
+        self.op_id = str(uuid.uuid4())
+        self.layer_idx = layer_idx
+        self.max_seq_len = max_seq_len
+        self.transpose_value = transpose_value
+        # self.rotary_emb = rotary_emb
+        if isinstance(parameters[0], tuple):  # weight, scale from QuantizedLinear
+            np_dtype = np.int8 if parameters[0][0].dtype == torch.int8 else np.uint8
+        else:  # FP16 Linear
+            np_dtype = np.float16
+
+        self.backend_cls_prefill = partial(
+            LowBitQwenMultiDecoderlayer,
+            num_heads=num_heads,
+            num_key_value_heads=num_key_value_heads,
+            num_layers=1,
+            cached_cos=cached_cos,
+            cached_sin=cached_sin,
+            input_layernorm_weights=None,
+            post_attn_layernorm_weights=None,
+            max_seq_len=max_seq_len,
+            rms_norm_eps=rms_norm_eps,
+            intermediate_size=intermediate_size,
+            mode="prefill",
+            transpose_value=self.transpose_value,
+            dtype=np_dtype,
+        )
+        self.layer_norm_0 = layer_norm_0
+        self.layer_norm_1 = layer_norm_1
+        self.q_bias = q_bias
+        self.k_bias = k_bias
+        self.v_bias = v_bias
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        **kwargs,
+    ) -> torch.Tensor:
+        """Torch module forward method.
+
+        Args:
+            x (torch.Tensor): Input tensor
+
+        Returns:
+            torch.Tensor: result
+        """
+
+        seq_len = hidden_states.shape[1]
+
+        backend_cls = self.backend_cls_prefill
+        inputs = (hidden_states.to(torch.float16), attention_mask, position_ids)
+        inputs += (self.layer_norm_0, self.layer_norm_1)
+        inputs += (self.q_bias, self.k_bias, self.v_bias)
+        hidden_states, past_key, past_value = run_model(
+            inputs, self.op_parameters, backend_cls, self.op_id, replica=2
+        )
+        cache_kwargs = {"max_seq_len": self.max_seq_len, "transpose": self.transpose_value}
+        key_states, value_states = past_key_value.update(
+            past_key, past_value, self.layer_idx, cache_kwargs
+        )
+
+        outputs = (hidden_states,)
+        outputs += (past_key_value,)
+        return outputs
+
+
+def run_decode(
+    model,
+    rank,
+    world_size,
+    port,
+    layer_start,
+    layer_end,
+    intra_stages,
+    max_seq_len,
+    transpose_value_cache,
+    input_queue,
+    result_queue,
+):
+
+    os.environ["MASTER_ADDR"] = "127.0.0.1"
+    os.environ["MASTER_PORT"] = port
+    os.environ["RANK"] = str(rank)
+    os.environ["WORLD_SIZE"] = str(world_size)
+
+    print("start init process group, rank: ", rank, "world_size: ", world_size)
+
+    dist.init_process_group()
+    my_rank = dist.get_rank()
+    my_size = dist.get_world_size()
+    logger.info(f"rank: {my_rank}, size: {my_size}")
+
+    num_heads = model.model.layers[layer_start].self_attn.num_heads
+    num_key_value_heads = model.model.layers[layer_start].self_attn.num_key_value_heads
+    head_dim = model.model.layers[layer_start].self_attn.head_dim
+    rms_norm_eps = model.config.rms_norm_eps
+    intermediate_size = model.config.intermediate_size
+    deocderlayers = []
+    layer_weights = []
+    input_layer_norm_weights = []
+    post_attn_layernorm_weights = []
+    q_biases = []
+    k_biases = []
+    v_biases = []
+    layer_indexs = range(layer_start, layer_end)
+    for layer_idx in layer_indexs:
+        curr_layer = model.model.layers[layer_idx]
+        attn_layer = curr_layer.self_attn
+        mlp_layer = curr_layer.mlp
+
+        weights = [
+            (attn_layer.q_proj.weight, attn_layer.q_proj.scale),
+            (attn_layer.k_proj.weight, attn_layer.k_proj.scale),
+            (attn_layer.v_proj.weight, attn_layer.v_proj.scale),
+            (attn_layer.o_proj.weight, attn_layer.o_proj.scale),
+            (mlp_layer.gate_proj.weight, mlp_layer.gate_proj.scale),
+            (mlp_layer.up_proj.weight, mlp_layer.up_proj.scale),
+            (mlp_layer.down_proj.weight, mlp_layer.down_proj.scale),
+        ]
+
+        cached_cos = curr_layer.self_attn.rotary_emb.cos_cached.to(torch.float16)
+        cached_sin = curr_layer.self_attn.rotary_emb.sin_cached.to(torch.float16)
+        layer_norm_0 = curr_layer.input_layernorm.weight.to(torch.float16)
+        layer_norm_1 = curr_layer.post_attention_layernorm.weight.to(torch.float16)
+
+        layer_weights.extend(weights)
+        input_layer_norm_weights.append(layer_norm_0)
+        post_attn_layernorm_weights.append(layer_norm_1)
+        q_biases.append(attn_layer.q_proj.bias.to(torch.float16))
+        k_biases.append(attn_layer.k_proj.bias.to(torch.float16))
+        v_biases.append(attn_layer.v_proj.bias.to(torch.float16))
+
+    multi_decoder = FusedQwenLowBitMultiDecoderlayer(
+        parameters=layer_weights,
+        input_laynorm_weights=input_layer_norm_weights,
+        post_attn_layernorm_weights=post_attn_layernorm_weights,
+        q_biases=q_biases,
+        k_biases=k_biases,
+        v_biases=v_biases,
+        layer_indexes=layer_indexs,
+        intra_stages=intra_stages,
+        cached_cos=cached_cos,
+        cached_sin=cached_sin,
+        num_heads=num_heads,
+        head_dim=head_dim,
+        num_key_value_heads=num_key_value_heads,
+        rms_norm_eps=rms_norm_eps,
+        intermediate_size=intermediate_size,
+        max_seq_len=max_seq_len,
+        transpose_value=transpose_value_cache,
+        do_print=False,
+    )
+
+    dist.barrier()
+
+    past_key_values = None
+
+    control = torch.empty((), dtype=torch.int)
+    hidden_states = torch.empty((1, 1, head_dim * num_heads), dtype=torch.float16)
+    with torch.inference_mode():
+        while True:
+
+            dist.broadcast(control, src=0)
+            if control.item() == -2:
+                break
+            elif control.item() == -1:
+                past_key_values = input_queue.get()
+            else:
+                t0 = time.perf_counter()
+                past_seen_tokens = past_key_values.get_seq_length()
+                attention_mask = torch.ones([1, past_seen_tokens + 1], dtype=torch.int64)
+                position_ids = torch.arange(
+                    past_seen_tokens,
+                    1 + past_seen_tokens,
+                    dtype=torch.long,
+                    device=hidden_states.device,
+                )
+                position_ids = position_ids.unsqueeze(0).view(-1, 1)
+
+                from transformers.modeling_attn_mask_utils import _prepare_4d_causal_attention_mask
+
+                causal_mask = _prepare_4d_causal_attention_mask(
+                    attention_mask,
+                    (hidden_states.shape[0], hidden_states.shape[1]),
+                    hidden_states,
+                    past_seen_tokens,
+                    sliding_window=model.model.config.sliding_window,
+                )
+                pad_len = multi_decoder.max_seq_len + 1 - causal_mask.size(-1)
+
+                causal_mask[:, :, :, -1] = torch.finfo(torch.float16).min
+                pad_mask = (0, pad_len)
+                padded_causal_mask = F.pad(
+                    causal_mask.to(torch.float16), pad_mask, value=torch.finfo(torch.float16).min
+                )
+                padded_causal_mask[:, :, :, -1] = 0.0
+                dist.recv(hidden_states, src=rank - 1)
+                t1 = time.perf_counter()
+                layer_outputs = multi_decoder(
+                    hidden_states,
+                    attention_mask=padded_causal_mask,
+                    position_ids=position_ids,
+                    past_key_value=past_key_values,
+                    output_attentions=False,
+                    use_cache=True,
+                )
+                t2 = time.perf_counter()
+                hidden_states = layer_outputs[0]
+                t3 = time.perf_counter()
+                dist.send(hidden_states, dst=(rank + 1) % world_size)
+                t4 = time.perf_counter()
+                past_key_values = layer_outputs[1]
+                new_keys = layer_outputs[2]
+                new_values = layer_outputs[3]
+                multi_decoder.post_forward(past_key_values, new_keys, new_values)
+
+
+class DecodeRunner:
+    def __init__(self, model, max_seq_len, intra_pp=2, inter_pp=2, transpose_value_cache=True):
+        self.model = model
+        self.max_seq_len = max_seq_len
+        self.transpose_value_cache = transpose_value_cache
+        world_size = inter_pp + 1
+        intra_stages = intra_pp
+        num_layers = self.model.model.config.num_hidden_layers
+
+        port = "54791"
+        os.environ["MASTER_ADDR"] = "127.0.0.1"
+        os.environ["MASTER_PORT"] = port
+        os.environ["RANK"] = "0"
+        os.environ["WORLD_SIZE"] = str(world_size)
+
+        self.input_queues = []
+        self.output_queues = []
+        self.decoder_processes = []
+
+        for rank in range(1, world_size):
+            input_q = mp.Queue()
+            output_q = mp.Queue()
+            start_layer = (rank - 1) * (num_layers // (world_size - 1))
+            end_layer = (rank) * (num_layers // (world_size - 1))
+            if rank == world_size - 1:
+                end_layer = num_layers
+            p = mp.Process(
+                target=run_decode,
+                args=(
+                    self.model,
+                    rank,
+                    world_size,
+                    port,
+                    start_layer,
+                    end_layer,
+                    intra_stages,
+                    self.max_seq_len,
+                    self.transpose_value_cache,
+                    input_q,
+                    output_q,
+                ),
+            )
+            p.daemon = True
+            p.start()
+            self.input_queues.append(input_q)
+            self.output_queues.append(output_q)
+            self.decoder_processes.append(p)
+
+        dist.init_process_group()
+        my_rank = dist.get_rank()
+        self.world_size = dist.get_world_size()
+        logger.info(f"rank: {my_rank}, size: {self.world_size}")
+
+        dist.barrier()
+        self.cache_past_key_value = None
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        **kwargs,
+    ):
+        t0 = time.perf_counter()
+
+        if self.cache_past_key_value != past_key_value:
+            control = torch.tensor(-1, dtype=torch.int)
+            dist.broadcast(control, src=0)
+            for i in range(len(self.decoder_processes)):
+                self.input_queues[i].put(past_key_value)
+
+        control = torch.tensor(0, dtype=torch.int)
+        dist.broadcast(control, src=0)
+        hidden_states = hidden_states.to(torch.float16)
+        dist.send(hidden_states, dst=1)
+        past_key_value.expand(self.transpose_value_cache)
+        dist.recv(hidden_states, src=self.world_size - 1)
+        t1 = time.perf_counter()
+        return hidden_states, past_key_value
+
+    def shutdown(self):
+        control = torch.tensor(-2, dtype=torch.int)
+        dist.broadcast(control, src=0)
+        for p in self.decoder_processes:
+            p.join(3)
+        for p in self.decoder_processes:
+            if p.exitcode is None:
+                p.kill()
+
+    def __del__(self):
+        self.shutdown()
+
+
+def run_prefill(
+    model, max_output_len, max_prompt_len, transpose_value_cache, input_queue, result_queue
+):
+
+    layer_start = 0
+    layer_end = len(model.model.layers)
+    num_heads = model.model.layers[layer_start].self_attn.num_heads
+    num_key_value_heads = model.model.layers[layer_start].self_attn.num_key_value_heads
+    head_dim = model.model.layers[layer_start].self_attn.head_dim
+    rms_norm_eps = model.config.rms_norm_eps
+    intermediate_size = model.config.intermediate_size
+    deocderlayers = []
+    layer_weights = []
+    input_layer_norm_weights = []
+    post_attn_layernorm_weights = []
+    layer_indexs = range(layer_start, layer_end)
+    for layer_idx in layer_indexs:
+        curr_layer = model.model.layers[layer_idx]
+        attn_layer = curr_layer.self_attn
+        mlp_layer = curr_layer.mlp
+
+        weights = [
+            (attn_layer.q_proj.weight, attn_layer.q_proj.scale),
+            (attn_layer.k_proj.weight, attn_layer.k_proj.scale),
+            (attn_layer.v_proj.weight, attn_layer.v_proj.scale),
+            (attn_layer.o_proj.weight, attn_layer.o_proj.scale),
+            (mlp_layer.gate_proj.weight, mlp_layer.gate_proj.scale),
+            (mlp_layer.up_proj.weight, mlp_layer.up_proj.scale),
+            (mlp_layer.down_proj.weight, mlp_layer.down_proj.scale),
+        ]
+
+        cached_cos = curr_layer.self_attn.rotary_emb.cos_cached.to(torch.float16)
+        cached_sin = curr_layer.self_attn.rotary_emb.sin_cached.to(torch.float16)
+
+        layer_norm_0 = curr_layer.input_layernorm.weight.to(torch.float16)
+        layer_norm_1 = curr_layer.post_attention_layernorm.weight.to(torch.float16)
+
+        new_decoderlayer = FusedQwenLowBitDecoderlayer(
+            weights,
+            num_heads=num_heads,
+            num_key_value_heads=num_key_value_heads,
+            cached_cos=cached_cos,
+            cached_sin=cached_sin,
+            layer_norm_0=layer_norm_0,
+            layer_norm_1=layer_norm_1,
+            q_bias=attn_layer.q_proj.bias.to(torch.float16),
+            k_bias=attn_layer.k_proj.bias.to(torch.float16),
+            v_bias=attn_layer.v_proj.bias.to(torch.float16),
+            layer_idx=layer_idx,
+            rms_norm_eps=rms_norm_eps,
+            intermediate_size=intermediate_size,
+            max_seq_len=max_output_len,
+            transpose_value=transpose_value_cache,
+        )
+
+        layer_weights.extend(weights)
+        input_layer_norm_weights.append(layer_norm_0)
+        post_attn_layernorm_weights.append(layer_norm_1)
+        model.model.layers[layer_idx] = new_decoderlayer
+        deocderlayers.append(new_decoderlayer)
+
+    print("finish creating all decode layers in prefill")
+    result_queue.put("loading finish")
+
+    while True:
+
+        result = input_queue.get()
+        if result == "stop":
+            break
+
+        hidden_states, position_ids, causal_mask, past_key_values = result
+        with torch.inference_mode():
+            for decoder_layer in deocderlayers:
+                layer_outputs = decoder_layer(
+                    hidden_states,
+                    attention_mask=causal_mask,
+                    position_ids=position_ids,
+                    past_key_value=past_key_values,
+                    output_attentions=False,
+                    use_cache=True,
+                )
+
+                hidden_states = layer_outputs[0]
+                next_decoder_cache = layer_outputs[1]
+
+            result_queue.put((hidden_states, next_decoder_cache))
+
+
+class PrefillRunner:
+    def __init__(self, model, max_output_len, max_prompt_len, transpose_value_cache):
+        self.model = model
+        self.max_output_len = max_output_len
+        self.max_prompt_len = max_prompt_len
+        self.transpose_value_cache = transpose_value_cache
+
+        self.prefill_result_queue = mp.Queue()
+        self.prefill_input_queue = mp.Queue()
+
+        self.p = mp.Process(
+            target=run_prefill,
+            args=(
+                model,
+                max_output_len,
+                max_prompt_len,
+                transpose_value_cache,
+                self.prefill_input_queue,
+                self.prefill_result_queue,
+            ),
+        )
+        self.p.daemon = True
+        self.p.start()
+        output = self.prefill_result_queue.get()
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        **kwargs,
+    ):
+        seq_len = hidden_states.size(1)
+        invalidInputError(
+            seq_len <= self.max_prompt_len,
+            (
+                f"seq_len: {seq_len} should be less than or equal"
+                " to max_prompt_len {self.max_prompt_len}"
+            ),
+        )
+        self.prefill_input_queue.put((hidden_states, position_ids, attention_mask, past_key_value))
+        return self.prefill_result_queue.get()
+
+    def shutdown(self):
+        self.prefill_input_queue.put("stop")
+        self.p.join(3)
+        if self.p.exitcode is None:
+            self.p.kill()
+
+    def __del__(self):
+        self.shutdown()
+
+
+def gen_qwen2_fused_model_forward(prefill_runner, decode_runner):
+
+    def qwen2_fused_model_forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, BaseModelOutputWithPast]:
+        output_attentions = (
+            output_attentions if output_attentions is not None else self.config.output_attentions
+        )
+        output_hidden_states = (
+            output_hidden_states
+            if output_hidden_states is not None
+            else self.config.output_hidden_states
+        )
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        # retrieve input_ids and inputs_embeds
+        if input_ids is not None and inputs_embeds is not None:
+            invalidInputError(False,
+                              "You cannot specify both decoder_input_ids and "
+                              "decoder_inputs_embeds at the same time")
+        elif input_ids is not None:
+            batch_size, seq_length = input_ids.shape
+        elif inputs_embeds is not None:
+            batch_size, seq_length, _ = inputs_embeds.shape
+        else:
+            invalidInputError(False,
+                              "You have to specify either input_ids or inputs_embeds")
+
+        if self.gradient_checkpointing and self.training:
+            if use_cache:
+                use_cache = False
+
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_tokens(input_ids)
+
+        past_key_values_length = 0
+
+        from ipex_llm.transformers.npu_models.kv import DynamicFusedNormalCache
+
+        if use_cache and not isinstance(past_key_values, DynamicFusedNormalCache):
+            past_key_values = DynamicFusedNormalCache.from_legacy_cache(past_key_values)
+            past_key_values_length = past_key_values.get_seq_length()
+
+        if position_ids is None:
+            device = input_ids.device if input_ids is not None else inputs_embeds.device
+            position_ids = torch.arange(
+                past_key_values_length,
+                seq_length + past_key_values_length,
+                dtype=torch.long,
+                device=device,
+            )
+            position_ids = position_ids.unsqueeze(0).view(-1, seq_length)
+        else:
+            position_ids = position_ids.view(-1, seq_length).long()
+
+        from transformers.modeling_attn_mask_utils import _prepare_4d_causal_attention_mask
+
+        attention_mask = _prepare_4d_causal_attention_mask(
+            attention_mask,
+            (batch_size, seq_length),
+            inputs_embeds,
+            past_key_values_length,
+            sliding_window=self.config.sliding_window,
+        )
+
+        # embed positions
+        hidden_states = inputs_embeds
+
+        # decoder layers
+        all_hidden_states = () if output_hidden_states else None
+        all_self_attns = () if output_attentions else None
+        next_decoder_cache = None
+
+        if seq_length == 1:
+            layers_runner = decode_runner
+        else:
+            layers_runner = prefill_runner
+        layer_outputs = layers_runner.forward(
+            hidden_states,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_value=past_key_values,
+            output_attentions=output_attentions,
+            use_cache=use_cache,
+        )
+        hidden_states = layer_outputs[0]
+
+        next_decoder_cache = layer_outputs[1]
+
+        hidden_states = self.norm(hidden_states)
+
+        # add hidden states from the last decoder layer
+        if output_hidden_states:
+            all_hidden_states += (hidden_states,)
+
+        next_cache = next_decoder_cache if use_cache else None
+        if not return_dict:
+            return tuple(
+                v
+                for v in [hidden_states, next_cache, all_hidden_states, all_self_attns]
+                if v is not None
+            )
+
+        return BaseModelOutputWithPast(
+            last_hidden_state=hidden_states,
+            past_key_values=next_cache,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attns,
+        )
+
+    return qwen2_fused_model_forward