add mistral npu support

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

@@ -86,6 +86,22 @@ def optimize_llm(model: torch.nn.Module):
         convert_forward(model, LlamaAttention, llama_attention_forward)
         convert_forward(model, LlamaMLP, llama_mlp_forward)
 
+    elif model.config.model_type == "mistral":
+        from ipex_llm.transformers.npu_models.mistral import merge_qkv
+        from ipex_llm.transformers.npu_models.mistral import merge_mlp
+        model.apply(merge_qkv)
+        model.apply(merge_mlp)
+
+        from ipex_llm.transformers.npu_models.mistral import mistral_model_forward
+        from ipex_llm.transformers.npu_models.mistral import mistral_attention_forward
+        from ipex_llm.transformers.npu_models.mistral import mistral_mlp_forward
+        from transformers.models.mistral.modeling_mistral import MistralModel
+        from transformers.models.mistral.modeling_mistral import MistralAttention
+        from transformers.models.mistral.modeling_mistral import MistralMLP
+        convert_forward(model, MistralModel, mistral_model_forward)
+        convert_forward(model, MistralAttention, mistral_attention_forward)
+        convert_forward(model, MistralMLP, mistral_mlp_forward)
+
     elif model.config.model_type == "qwen2":
         from ipex_llm.transformers.npu_models.qwen2 import merge_qkv
         from ipex_llm.transformers.npu_models.qwen2 import merge_mlp

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

@@ -230,6 +230,7 @@ def llama_attention_forward(
             attn_mask=causal_mask,
             is_causal=self.is_causal and causal_mask is None and q_len > 1,
         )
+        attn_weights = None
     else:
         attn_weights = torch.matmul(query_states,
                                     key_states.transpose(2, 3)) / math.sqrt(self.head_dim)

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

@@ -0,0 +1,277 @@
+#
+# 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.
+#
+# Some parts of this file is adapted from
+# https://github.com/huggingface/transformers/blob/v4.40.0/src/transformers/models/mistral/modeling_mistral.py
+# which is licensed under Apache License 2.0:
+#
+# Copyright 2023 Mistral AI and the HuggingFace Inc. team. All rights reserved.
+#
+# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
+# and OPT implementations in this library. It has been modified from its
+# original forms to accommodate minor architectural differences compared
+# to GPT-NeoX and OPT used by the Meta AI team that trained the model.
+#
+# 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.
+
+
+from typing import Optional, Tuple, List, Union
+
+import math
+import torch
+from transformers.cache_utils import Cache
+from transformers.modeling_outputs import BaseModelOutputWithPast
+from transformers.models.mistral.modeling_mistral import repeat_kv, apply_rotary_pos_emb
+from transformers.models.mistral.modeling_mistral import MistralAttention, MistralMLP
+from transformers.models.mistral.modeling_mistral import _prepare_4d_causal_attention_mask
+
+from ipex_llm.utils.common.log4Error import invalidInputError
+from ipex_llm.transformers.npu_models.common import merge_linear
+
+
+def merge_qkv(module: torch.nn.Module):
+    if isinstance(module, MistralAttention):
+        qkv_proj = merge_linear([
+            module.q_proj,
+            module.k_proj,
+            module.v_proj,
+        ])
+        module.qkv_proj = qkv_proj
+        del module.q_proj, module.k_proj, module.v_proj
+
+
+def merge_mlp(module: torch.nn.Module):
+    if isinstance(module, MistralMLP):
+        gate_up_proj = merge_linear([
+            module.gate_proj,
+            module.up_proj,
+        ])
+        module.gate_up_proj = gate_up_proj
+        del module.gate_proj, module.up_proj
+
+
+def mistral_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,
+    cache_position: Optional[torch.LongTensor] = 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
+
+    if (input_ids is None) ^ (inputs_embeds is not None):
+        invalidInputError(False,
+                          ("You cannot specify both input_ids and inputs_embeds at the same time, "
+                           "and must specify either one"))
+    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
+
+    if self.gradient_checkpointing and self.training and use_cache:
+        use_cache = False
+
+    past_key_values_length = 0
+
+    # ipex-llm changes start
+    from ipex_llm.transformers.kv import DynamicNormalCache
+    if use_cache and not isinstance(past_key_values, DynamicNormalCache):
+        past_key_values = DynamicNormalCache.from_legacy_cache(past_key_values)
+        past_key_values_length = past_key_values.get_seq_length()
+    # ipex-llm changes end
+
+    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()
+
+    if inputs_embeds is None:
+        inputs_embeds = self.embed_tokens(input_ids)
+
+    # ipex-llm changes start
+    # 4d mask is passed through the layers
+    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,
+    )
+    # ipex-llm changes end
+
+    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
+
+    for decoder_layer in self.layers:
+        if output_hidden_states:
+            all_hidden_states += (hidden_states,)
+
+        if self.gradient_checkpointing and self.training:
+            layer_outputs = self._gradient_checkpointing_func(
+                decoder_layer.__call__,
+                hidden_states,
+                attention_mask,
+                position_ids,
+                past_key_values,
+                output_attentions,
+                use_cache,
+            )
+        else:
+            layer_outputs = decoder_layer(
+                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]
+
+        if use_cache:
+            next_decoder_cache = layer_outputs[2 if output_attentions else 1]
+
+        if output_attentions:
+            all_self_attns += (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,)
+
+    # ipex-llm changes start
+    next_cache = next_decoder_cache if use_cache else None
+    # ipex-llm changes end
+
+    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,
+    )
+
+
+def mistral_attention_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,
+    cache_position: Optional[torch.LongTensor] = None,
+    **kwargs,
+) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+    bsz, q_len, _ = hidden_states.size()
+
+    qkv = self.qkv_proj(hidden_states)
+    qkv = qkv.view(bsz, q_len, self.num_heads + 2 * self.num_key_value_heads, self.head_dim)
+    qkv = qkv.transpose(1, 2)
+    query_states, key_states, value_states = qkv.split([self.num_heads,
+                                                        self.num_key_value_heads,
+                                                        self.num_key_value_heads], dim=1)
+
+    kv_seq_len = q_len
+    if past_key_value is not None:
+        kv_seq_len += past_key_value.get_usable_length(kv_seq_len, self.layer_idx)
+    cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
+    query_states, key_states = apply_rotary_pos_emb(query_states, key_states,
+                                                    cos, sin, position_ids)
+    if past_key_value is not None:
+        cache_kwargs = {"sin": sin, "cos": cos}  # Specific to RoPE models
+        key_states, value_states = past_key_value.update(key_states, value_states,
+                                                         self.layer_idx, cache_kwargs)
+
+    key_states = repeat_kv(key_states, self.num_key_value_groups)
+    value_states = repeat_kv(value_states, self.num_key_value_groups)
+
+    if query_states.size(2) == key_states.size(2):
+        # first token
+        from intel_npu_acceleration_library.functional import scaled_dot_product_attention
+        attn_output = scaled_dot_product_attention(
+            query_states,
+            key_states,
+            value_states,
+            attn_mask=attention_mask,
+            is_causal=attention_mask is None and bsz == 1 and q_len > 1,
+        )
+        attn_weights = None
+    else:
+        attn_weights = torch.matmul(query_states,
+                                    key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
+        if attention_mask is not None:
+            attn_weights = attn_weights + attention_mask
+
+        attn_weights = torch.nn.functional.softmax(attn_weights, dim=-1,
+                                                   dtype=torch.float32).to(value_states.dtype)
+        attn_weights = torch.nn.functional.dropout(attn_weights, p=self.attention_dropout,
+                                                   training=self.training)
+        attn_output = torch.matmul(attn_weights, value_states)
+
+    attn_output = attn_output.transpose(1, 2).contiguous()
+
+    attn_output = attn_output.reshape(bsz, q_len, self.hidden_size)
+
+    attn_output = self.o_proj(attn_output)
+
+    if not output_attentions:
+        attn_weights = None
+
+    return attn_output, attn_weights, past_key_value
+
+
+def mistral_mlp_forward(self, x):
+    gate_up_proj = self.gate_up_proj(x)
+    gate_proj, up_proj = gate_up_proj.chunk(2, dim=-1)
+    down_proj = self.down_proj(self.act_fn(gate_proj) * up_proj)
+    return down_proj