qwen/modeling_qwen.py

@@ -4,13 +4,16 @@
 # LICENSE file in the root directory of this source tree.
 
 import copy
+import importlib
 import math
 import inspect
+import pathlib
 from typing import TYPE_CHECKING, Optional, Tuple, Union, Callable, List, Any, Generator
 
 import torch
 import torch.nn.functional as F
 import torch.utils.checkpoint
+import warnings
 
 from torch.nn import CrossEntropyLoss
 from transformers import PreTrainedTokenizer, GenerationConfig, StoppingCriteriaList
@@ -26,8 +29,13 @@ from transformers.modeling_outputs import (
 from transformers.modeling_utils import PreTrainedModel
 from transformers.utils import logging
 
+try:
+    from einops import rearrange
+except ImportError:
+    rearrange = None
 from torch import nn
-from einops import rearrange
+
+SUPPORT_CUDA = torch.cuda.is_available()
 
 from configuration_qwen import QWenConfig
 from qwen_generation_utils import (
@@ -40,6 +48,8 @@ from qwen_generation_utils import (
 
 logger = logging.get_logger(__name__)
 
+_SENTINEL = object()
+
 
 class QWenAttention(nn.Module):
     def __init__(self, config):
@@ -61,9 +71,11 @@ class QWenAttention(nn.Module):
         self.hidden_size_per_attention_head = self.projection_size // config.num_attention_heads
 
         self.c_attn = nn.Linear(config.hidden_size, 3 * self.projection_size)
+
         self.c_proj = nn.Linear(config.hidden_size, self.projection_size, bias=not config.no_bias)
 
         self.use_dynamic_ntk = config.use_dynamic_ntk
+        self.use_logn_attn = config.use_logn_attn
 
         logn_list = [math.log(i, self.seq_length) if i > self.seq_length else 1 for i in range(1, 32768)]
         logn_tensor = torch.tensor(logn_list)[None, :, None, None]
@@ -92,20 +104,38 @@ class QWenAttention(nn.Module):
         rotary_pos_emb_list: Optional[List[List[torch.Tensor]]] = None,
         layer_past: Optional[Tuple[torch.Tensor]] = None,
         attention_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
     ):
         mixed_x_layer = self.c_attn(hidden_states)
+
         query, key, value = mixed_x_layer.split(self.split_size, dim=2)
+
         query = self._split_heads(query, self.num_heads, self.head_dim)
         key = self._split_heads(key, self.num_heads, self.head_dim)
         value = self._split_heads(value, self.num_heads, self.head_dim)
 
-        rotary_pos_emb = rotary_pos_emb_list[0]
-        rotary_pos_emb = [i[:, -query.shape[1] :, :, :] for i in rotary_pos_emb]
-        rotary_pos_emb = (rotary_pos_emb,) * 2
-        q_pos_emb, k_pos_emb = rotary_pos_emb
-        # Slice the pos emb for current inference
-        query = apply_rotary_pos_emb(query, q_pos_emb)
-        key = apply_rotary_pos_emb(key, k_pos_emb)
+        if rotary_pos_emb_list is not None:
+            cur_len = query.shape[1]
+            if len(rotary_pos_emb_list) == 1:
+                rotary_pos_emb = rotary_pos_emb_list[0]
+                rotary_pos_emb = [i[:, -cur_len:, :, :] for i in rotary_pos_emb]
+                rotary_pos_emb = (rotary_pos_emb,) * 2
+                q_pos_emb, k_pos_emb = rotary_pos_emb
+                # Slice the pos emb for current inference
+                query = apply_rotary_pos_emb(query, q_pos_emb)
+                key = apply_rotary_pos_emb(key, k_pos_emb)
+            else:
+                query_list = []
+                key_list = []
+                for i, rotary_pos_emb in enumerate(rotary_pos_emb_list):
+                    rotary_pos_emb = [i[:, -cur_len:, :, :] for i in rotary_pos_emb]
+                    rotary_pos_emb = (rotary_pos_emb,) * 2
+                    q_pos_emb, k_pos_emb = rotary_pos_emb
+                    # Slice the pos emb for current inference
+                    query_list += [apply_rotary_pos_emb(query[i : i + 1, :, :], q_pos_emb)]
+                    key_list += [apply_rotary_pos_emb(key[i : i + 1, :, :], k_pos_emb)]
+                query = torch.cat(query_list, dim=0)
+                key = torch.cat(key_list, dim=0)
 
         if layer_past is not None:
             past_key, past_value = layer_past[0], layer_past[1]
@@ -113,10 +143,13 @@ class QWenAttention(nn.Module):
             key = torch.cat((past_key, key), dim=1)
             value = torch.cat((past_value, value), dim=1)
 
-        present = (key, value)
+        if use_cache:
+            present = (key, value)
+        else:
+            present = None
 
         key_size = key.size(1)
-        if key_size > self.seq_length and not self.training:
+        if key_size > self.seq_length and self.use_logn_attn and not self.training:
             seq_start = key.size(1) - query.size(1)
             seq_end = key.size(1)
             logn_tensor = self.logn_tensor[:, seq_start:seq_end, :, :].type_as(query)
@@ -139,20 +172,23 @@ class QWenAttention(nn.Module):
                 attention_mask = attention_mask.masked_fill(~causal_mask, torch.finfo(query.dtype).min)
         else:
             attention_mask = causal_mask
-
         attn_output = F.scaled_dot_product_attention(query, key, value, attn_mask=attention_mask).transpose(1, 2)
+
         context_layer = self._merge_heads(attn_output, self.num_heads, self.head_dim)
+
         attn_output = self.c_proj(context_layer)
+
         outputs = (attn_output, present)
+
         return outputs
 
 
 class QWenMLP(nn.Module):
     def __init__(self, config):
         super().__init__()
+        self.w1 = nn.Linear(config.hidden_size, config.intermediate_size // 2, bias=not config.no_bias)
+        self.w2 = nn.Linear(config.hidden_size, config.intermediate_size // 2, bias=not config.no_bias)
         ff_dim_in = config.intermediate_size // 2
-        self.w1 = nn.Linear(config.hidden_size, ff_dim_in, bias=not config.no_bias)
-        self.w2 = nn.Linear(config.hidden_size, ff_dim_in, bias=not config.no_bias)
         self.c_proj = nn.Linear(ff_dim_in, config.hidden_size, bias=not config.no_bias)
 
     def forward(self, hidden_states):
@@ -177,6 +213,7 @@ class QWenBlock(nn.Module):
             hidden_size,
             eps=config.layer_norm_epsilon,
         )
+
         self.mlp = QWenMLP(config)
 
     def forward(
@@ -185,6 +222,11 @@ class QWenBlock(nn.Module):
         rotary_pos_emb_list: Optional[List[List[torch.Tensor]]] = None,
         layer_past: Optional[Tuple[torch.Tensor]] = None,
         attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
+        output_attentions: Optional[bool] = False,
     ):
         layernorm_output = self.ln_1(hidden_states)
 
@@ -193,17 +235,27 @@ class QWenBlock(nn.Module):
             rotary_pos_emb_list,
             layer_past=layer_past,
             attention_mask=attention_mask,
+            use_cache=use_cache,
         )
+
         attn_output = attn_outputs[0]
+
         outputs = attn_outputs[1:]
+
         residual = hidden_states
         layernorm_input = attn_output + residual
 
         layernorm_output = self.ln_2(layernorm_input)
+
         residual = layernorm_input
         mlp_output = self.mlp(layernorm_output)
         hidden_states = residual + mlp_output
-        outputs = (hidden_states,) + outputs
+
+        if use_cache:
+            outputs = (hidden_states,) + outputs
+        else:
+            outputs = (hidden_states,) + outputs[1:]
+
         return outputs
 
 
@@ -262,7 +314,14 @@ class QWenModel(QWenPreTrainedModel):
         attention_mask: Optional[torch.FloatTensor] = None,
         head_mask: Optional[torch.FloatTensor] = None,
         inputs_embeds: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
     ):
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+
         if input_ids is not None and inputs_embeds is not None:
             raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
         elif input_ids is not None:
@@ -284,6 +343,7 @@ class QWenModel(QWenPreTrainedModel):
             attention_mask = attention_mask.to(dtype=self.dtype)
             attention_mask = (1.0 - attention_mask) * torch.finfo(self.dtype).min
 
+        encoder_attention_mask = None
         head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)
 
         if inputs_embeds is None:
@@ -316,7 +376,8 @@ class QWenModel(QWenPreTrainedModel):
         hidden_states = self.drop(hidden_states)
         output_shape = input_shape + (hidden_states.size(-1),)
 
-        presents = ()
+        presents = () if use_cache else None
+        all_self_attentions = () if output_attentions else None
         all_hidden_states = None
         for i, (block, layer_past) in enumerate(zip(self.h, past_key_values)):
             outputs = block(
@@ -324,14 +385,27 @@ class QWenModel(QWenPreTrainedModel):
                 layer_past=layer_past,
                 rotary_pos_emb_list=rotary_pos_emb_list,
                 attention_mask=attention_mask,
+                head_mask=head_mask[i],
+                encoder_hidden_states=encoder_hidden_states,
+                encoder_attention_mask=encoder_attention_mask,
+                use_cache=use_cache,
+                output_attentions=output_attentions,
             )
+
             hidden_states = outputs[0]
-            presents = presents + (outputs[1],)
+            if use_cache is True:
+                presents = presents + (outputs[1],)
+
+            if output_attentions:
+                all_self_attentions = all_self_attentions + (outputs[2 if use_cache else 1],)
 
         hidden_states = self.ln_f(hidden_states)
         hidden_states = hidden_states.view(output_shape)
         return BaseModelOutputWithPast(
-            last_hidden_state=hidden_states, past_key_values=presents, hidden_states=all_hidden_states
+            last_hidden_state=hidden_states,
+            past_key_values=presents,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attentions,
         )
 
 
@@ -360,6 +434,7 @@ class QWenLMHeadModel(QWenPreTrainedModel):
         model_inputs.update(
             {
                 "past_key_values": past_key_values,
+                "use_cache": kwargs.get("use_cache"),
                 "attention_mask": attention_mask,
             }
         )
@@ -372,7 +447,11 @@ class QWenLMHeadModel(QWenPreTrainedModel):
         attention_mask: Optional[torch.FloatTensor] = None,
         head_mask: Optional[torch.FloatTensor] = None,
         inputs_embeds: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
         labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         transformer_outputs = self.transformer(
             input_ids,
@@ -380,6 +459,10 @@ class QWenLMHeadModel(QWenPreTrainedModel):
             attention_mask=attention_mask,
             head_mask=head_mask,
             inputs_embeds=inputs_embeds,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
         )
         hidden_states = transformer_outputs[0]
 
@@ -415,15 +498,19 @@ class QWenLMHeadModel(QWenPreTrainedModel):
         query: str,
         history: Optional[HistoryType],
         system: str = "You are a helpful assistant.",
+        stream: Optional[bool] = _SENTINEL,
         stop_words_ids: Optional[List[List[int]]] = None,
         generation_config: Optional[GenerationConfig] = None,
         **kwargs,
     ) -> Tuple[str, HistoryType]:
         generation_config = generation_config if generation_config is not None else self.generation_config
 
+        assert stream is _SENTINEL
+        assert generation_config.chat_format == "chatml"
         if history is None:
             history = []
         else:
+            # make a copy of the user's input such that is is left untouched
             history = copy.deepcopy(history)
 
         if stop_words_ids is None:
@@ -449,6 +536,7 @@ class QWenLMHeadModel(QWenPreTrainedModel):
             generation_config=generation_config,
             **kwargs,
         )
+
         response = decode_tokens(
             outputs[0],
             tokenizer,
@@ -458,7 +546,13 @@ class QWenLMHeadModel(QWenPreTrainedModel):
             verbose=False,
             errors="replace",
         )
+
+        # as history is a copy of the user inputs,
+        # we can always return the new turn to the user.
+        # separating input history and output history also enables the user
+        # to implement more complex history management
         history.append((query, response))
+
         return response, history
 
     def generate(
@@ -468,6 +562,7 @@ class QWenLMHeadModel(QWenPreTrainedModel):
         logits_processor: Optional[LogitsProcessorList] = None,
         stopping_criteria: Optional[StoppingCriteriaList] = None,
         prefix_allowed_tokens_fn: Optional[Callable[[int, torch.Tensor], List[int]]] = None,
+        synced_gpus: Optional[bool] = None,
         assistant_model: Optional["PreTrainedModel"] = None,
         streamer: Optional["BaseStreamer"] = None,
         **kwargs,
@@ -497,6 +592,7 @@ class QWenLMHeadModel(QWenPreTrainedModel):
             logits_processor=logits_processor,
             stopping_criteria=stopping_criteria,
             prefix_allowed_tokens_fn=prefix_allowed_tokens_fn,
+            synced_gpus=synced_gpus,
             assistant_model=assistant_model,
             streamer=streamer,
             **kwargs,
@@ -509,6 +605,7 @@ class QWenLMHeadModel(QWenPreTrainedModel):
         logits_processor: Optional[LogitsProcessorList] = None,
         stopping_criteria: Optional[StoppingCriteriaList] = None,
         prefix_allowed_tokens_fn: Optional[Callable[[int, torch.Tensor], List[int]]] = None,
+        synced_gpus: Optional[bool] = None,
         assistant_model: Optional["PreTrainedModel"] = None,
         streamer: Optional["BaseStreamer"] = None,
         negative_prompt_ids: Optional[torch.Tensor] = None,
@@ -540,10 +637,21 @@ class QWenLMHeadModel(QWenPreTrainedModel):
             generation_config.pad_token_id = eos_token_id
 
         # 3. Define model inputs
+        # inputs_tensor has to be defined
+        # model_input_name is defined if model-specific keyword input is passed
+        # otherwise model_input_name is None
+        # all model-specific keyword inputs are removed from `model_kwargs`
         inputs_tensor, model_input_name, model_kwargs = self._prepare_model_inputs(
             inputs, generation_config.bos_token_id, model_kwargs
         )
         # 4. Define other model kwargs
+        model_kwargs["output_attentions"] = generation_config.output_attentions
+        # decoder-only models with inputs_embeds forwarding must use caching (otherwise we can't detect whether we are
+        # generating the first new token or not, and we only want to use the embeddings for the first new token)
+        if not self.config.is_encoder_decoder and model_input_name == "inputs_embeds":
+            model_kwargs["use_cache"] = True
+        else:
+            model_kwargs["use_cache"] = generation_config.use_cache
 
         accepts_attention_mask = "attention_mask" in set(inspect.signature(self.forward).parameters.keys())
         requires_attention_mask = "encoder_outputs" not in model_kwargs
@@ -605,6 +713,7 @@ class QWenLMHeadModel(QWenPreTrainedModel):
             pad_token_id=generation_config.pad_token_id,
             eos_token_id=generation_config.eos_token_id,
             output_scores=generation_config.output_scores,
+            synced_gpus=synced_gpus,
             streamer=streamer,
             **model_kwargs,
         )
@@ -618,7 +727,9 @@ class QWenLMHeadModel(QWenPreTrainedModel):
         max_length: Optional[int] = None,
         pad_token_id: Optional[int] = None,
         eos_token_id: Optional[Union[int, List[int]]] = None,
+        output_attentions: Optional[bool] = None,
         output_scores: Optional[bool] = None,
+        synced_gpus: bool = False,
         streamer: Optional["BaseStreamer"] = None,
         **model_kwargs,
     ):
@@ -633,6 +744,9 @@ class QWenLMHeadModel(QWenPreTrainedModel):
             eos_token_id = [eos_token_id]
         eos_token_id_tensor = torch.tensor(eos_token_id).to(input_ids.device) if eos_token_id is not None else None
         output_scores = output_scores if output_scores is not None else self.generation_config.output_scores
+        output_attentions = (
+            output_attentions if output_attentions is not None else self.generation_config.output_attentions
+        )
 
         # init attention / hidden states / scores tuples
         scores = None
@@ -640,14 +754,14 @@ class QWenLMHeadModel(QWenPreTrainedModel):
         # keep track of which sequences are already finished
         unfinished_sequences = torch.ones(input_ids.shape[0], dtype=torch.long, device=input_ids.device)
 
-        this_peer_finished = False
+        this_peer_finished = False  # used by synced_gpus only
         # auto-regressive generation
         while True:
             # prepare model inputs
             model_inputs = self.prepare_inputs_for_generation(input_ids, **model_kwargs)
 
             # forward pass to get next token
-            outputs = self(**model_inputs)
+            outputs = self(**model_inputs, output_attentions=output_attentions)
 
             next_token_logits = outputs.logits[:, -1, :]
 
@@ -718,6 +832,8 @@ class RotaryEmbedding(torch.nn.Module):
             freqs = torch.outer(seq.type_as(self.inv_freq), self.inv_freq)
 
             emb = torch.cat((freqs, freqs), dim=-1)
+            from einops import rearrange
+
             emb = rearrange(emb, "n d -> 1 n 1 d")
 
             cos, sin = emb.cos(), emb.sin()
@@ -730,6 +846,8 @@ class RotaryEmbedding(torch.nn.Module):
 
 
 def _rotate_half(x):
+    from einops import rearrange
+
     x = rearrange(x, "... (j d) -> ... j d", j=2)
     x1, x2 = x.unbind(dim=-2)
     return torch.cat((-x2, x1), dim=-1)