# Copyright (c) Alibaba Cloud. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. import importlib import math from typing import TYPE_CHECKING, Optional, Tuple, Union, Callable, List, Any, Generator import torch import torch.nn.functional as F import torch.utils.checkpoint from torch.cuda.amp import autocast from torch.nn import CrossEntropyLoss from transformers import PreTrainedTokenizer, GenerationConfig, StoppingCriteriaList from transformers.generation.logits_process import LogitsProcessorList if TYPE_CHECKING: from transformers.generation.streamers import BaseStreamer from transformers.generation.utils import GenerateOutput from transformers.modeling_outputs import ( BaseModelOutputWithPast, CausalLMOutputWithPast, ) from transformers.modeling_utils import PreTrainedModel from transformers.utils import logging try: from einops import rearrange except ImportError: rearrange = None from torch import nn try: from kernels.cpp_kernels import cache_autogptq_cuda_256 except ImportError: cache_autogptq_cuda_256 = None SUPPORT_CUDA = torch.cuda.is_available() SUPPORT_BF16 = SUPPORT_CUDA and torch.cuda.is_bf16_supported() SUPPORT_FP16 = SUPPORT_CUDA and torch.cuda.get_device_capability(0)[0] >= 7 from .configuration_qwen import QWenConfig from .qwen_generation_utils import ( HistoryType, make_context, decode_tokens, get_stop_words_ids, StopWordsLogitsProcessor, ) from flash_attn.bert_padding import unpad_input, pad_input logger = logging.get_logger(__name__) _CHECKPOINT_FOR_DOC = "qwen" _CONFIG_FOR_DOC = "QWenConfig" QWen_PRETRAINED_MODEL_ARCHIVE_LIST = ["qwen-7b"] _ERROR_BAD_CHAT_FORMAT = """\ We detect you are probably using the pretrained model (rather than chat model) for chatting, since the chat_format in generation_config is not "chatml". If you are directly using the model downloaded from Huggingface, please make sure you are using our "Qwen/Qwen-7B-Chat" Huggingface model (rather than "Qwen/Qwen-7B") when you call model.chat(). 我们检测到您可能在使用预训练模型（而非chat模型）进行多轮chat，因为您当前在generation_config指定的chat_format，并未设置为我们在对话中所支持的"chatml"格式。 如果您在直接使用我们从Huggingface提供的模型，请确保您在调用model.chat()时，使用的是"Qwen/Qwen-7B-Chat"模型（而非"Qwen/Qwen-7B"预训练模型）。 """ _SENTINEL = object() _ERROR_STREAM_IN_CHAT = """\ Pass argument `stream` to model.chat() is buggy, deprecated, and marked for removal. Please use model.chat_stream(...) instead of model.chat(..., stream=True). 向model.chat()传入参数stream的用法可能存在Bug，该用法已被废弃，将在未来被移除。请使用model.chat_stream(...)代替model.chat(..., stream=True)。 """ _ERROR_INPUT_CPU_QUERY_WITH_FLASH_ATTN_ACTIVATED = """\ We detect you have activated flash attention support, but running model computation on CPU. Please make sure that your input data has been placed on GPU. If you actually want to run CPU computation, please following the readme and set device_map="cpu" to disable flash attention when loading the model (calling AutoModelForCausalLM.from_pretrained). 检测到您的模型已激活了flash attention支持，但正在执行CPU运算任务。如使用flash attention，请您确认模型输入已经传到GPU上。如果您确认要执行CPU运算，请您在载入模型（调用AutoModelForCausalLM.from_pretrained）时，按照readme说法，指定device_map="cpu"以禁用flash attention。 """ apply_rotary_emb_func = None rms_norm = None flash_attn_unpadded_func = None def _import_flash_attn(): global apply_rotary_emb_func, rms_norm, flash_attn_unpadded_func try: from flash_attn.layers.rotary import apply_rotary_emb_func as __apply_rotary_emb_func apply_rotary_emb_func = __apply_rotary_emb_func except ImportError: logger.warn( "Warning: import flash_attn rotary fail, please install FlashAttention rotary to get higher efficiency " "https://github.com/Dao-AILab/flash-attention/tree/main/csrc/rotary" ) try: from flash_attn.ops.rms_norm import rms_norm as __rms_norm rms_norm = __rms_norm except ImportError: logger.warn( "Warning: import flash_attn rms_norm fail, please install FlashAttention layer_norm to get higher efficiency " "https://github.com/Dao-AILab/flash-attention/tree/main/csrc/layer_norm" ) try: import flash_attn if not hasattr(flash_attn, '__version__'): from flash_attn.flash_attn_interface import flash_attn_unpadded_func as __flash_attn_unpadded_func else: if int(flash_attn.__version__.split(".")[0]) >= 2: from flash_attn.flash_attn_interface import flash_attn_varlen_func as __flash_attn_unpadded_func else: from flash_attn.flash_attn_interface import flash_attn_unpadded_func as __flash_attn_unpadded_func flash_attn_unpadded_func = __flash_attn_unpadded_func except ImportError: logger.warn( "Warning: import flash_attn fail, please install FlashAttention to get higher efficiency " "https://github.com/Dao-AILab/flash-attention" ) def quantize_cache_v(fdata, bits, qmax, qmin): # b, s, head, h-dim->b, head, s, h-dim qtype = torch.uint8 device = fdata.device shape = fdata.shape fdata_cal = torch.flatten(fdata, 2) fmax = torch.amax(fdata_cal, dim=-1, keepdim=True) fmin = torch.amin(fdata_cal, dim=-1, keepdim=True) # Compute params if qmax.device != fmax.device: qmax = qmax.to(device) qmin = qmin.to(device) scale = (fmax - fmin) / (qmax - qmin) zero = qmin - fmin / scale scale = scale.unsqueeze(-1).repeat(1,1,shape[2],1).contiguous() zero = zero.unsqueeze(-1).repeat(1,1,shape[2],1).contiguous() # Quantize res_data = fdata / scale + zero qdata = torch.clamp(res_data, qmin, qmax).to(qtype) return qdata.contiguous(), scale, zero def dequantize_cache_torch(qdata, scale, zero): data = scale * (qdata - zero) return data class FlashSelfAttention(torch.nn.Module): def __init__( self, causal=False, softmax_scale=None, attention_dropout=0.0, ): super().__init__() assert flash_attn_unpadded_func is not None, ( "Please install FlashAttention first, " "e.g., with pip install flash-attn" ) assert ( rearrange is not None ), "Please install einops first, e.g., with pip install einops" self.causal = causal self.softmax_scale = softmax_scale self.dropout_p = attention_dropout def unpad_input(self, hidden_states, attention_mask): valid_mask = attention_mask.squeeze(1).squeeze(1).eq(0) seqlens_in_batch = valid_mask.sum(dim=-1, dtype=torch.int32) indices = torch.nonzero(valid_mask.flatten(), as_tuple=False).flatten() max_seqlen_in_batch = seqlens_in_batch.max().item() cu_seqlens = F.pad(torch.cumsum(seqlens_in_batch, dim=0, dtype=torch.torch.int32), (1, 0)) hidden_states = hidden_states[indices] return hidden_states, indices, cu_seqlens, max_seqlen_in_batch def pad_input(self, hidden_states, indices, batch, seqlen): #torch.cuda.empty_cache() output = torch.zeros(batch * seqlen, *hidden_states.shape[1:], device=hidden_states.device, dtype=hidden_states.dtype) output[indices] = hidden_states return rearrange(output, '(b s) ... -> b s ...', b=batch) def forward(self, q, k, v, attention_mask=None): assert all((i.dtype in [torch.float16, torch.bfloat16] for i in (q, k, v))) assert all((i.is_cuda for i in (q, k, v))) q_len_origin=q.shape[1] batch_size, seqlen_q = q.shape[0], q.shape[1] seqlen_k = k.shape[1] q, k, v = [rearrange(x, "b s ... -> (b s) ...") for x in [q, k, v]] cu_seqlens_q = torch.arange( 0, (batch_size + 1) * seqlen_q, step=seqlen_q, dtype=torch.int32, device=q.device, ) if attention_mask is not None: k, indices_k, cu_seqlens_k, seqlen_k = self.unpad_input(k, attention_mask) v = v[indices_k] if self.training or q.size(0) == k.size(0): q = q[indices_k] cu_seqlens_q = cu_seqlens_k seqlen_q = seqlen_k else: cu_seqlens_k = torch.arange( 0, (batch_size + 1) * seqlen_k, step=seqlen_k, dtype=torch.int32, device=q.device, ) if self.training: assert seqlen_k == seqlen_q is_causal = self.causal dropout_p = self.dropout_p else: is_causal = seqlen_q == seqlen_k dropout_p = 0 output = flash_attn_unpadded_func( q, k, v, cu_seqlens_q, cu_seqlens_k, seqlen_q, seqlen_k, dropout_p, softmax_scale=self.softmax_scale, causal=is_causal, ) if attention_mask is not None and seqlen_q == seqlen_k: output = self.pad_input(output, indices_k, batch_size, seqlen_q) else: new_shape = (batch_size, output.shape[0] // batch_size) + output.shape[1:] output = output.view(new_shape) return output class QWenAttention(nn.Module): def __init__(self, config): super().__init__() self.register_buffer("masked_bias", torch.tensor(-1e4), persistent=False) self.seq_length = config.seq_length self.hidden_size = config.hidden_size self.split_size = config.hidden_size self.num_heads = config.num_attention_heads self.head_dim = self.hidden_size // self.num_heads self.use_flash_attn = config.use_flash_attn self.scale_attn_weights = True self.projection_size = config.kv_channels * config.num_attention_heads assert self.projection_size % config.num_attention_heads == 0 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.is_fp32 = not (config.bf16 or config.fp16) if ( self.use_flash_attn and flash_attn_unpadded_func is not None and not self.is_fp32 ): self.core_attention_flash = FlashSelfAttention( causal=True, attention_dropout=config.attn_dropout_prob ) self.bf16 = config.bf16 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] self.register_buffer("logn_tensor", logn_tensor, persistent=False) self.attn_dropout = nn.Dropout(config.attn_dropout_prob) self.use_cache_quantization = config.use_cache_quantization if hasattr(config, 'use_cache_quantization') else False self.use_cache_kernel = config.use_cache_kernel if hasattr(config,'use_cache_kernel') else False cache_dtype = torch.float if self.bf16: cache_dtype=torch.bfloat16 elif config.fp16: cache_dtype = torch.float16 self.cache_qmax = torch.tensor(torch.iinfo(torch.uint8).max, dtype=cache_dtype) self.cache_qmin = torch.tensor(torch.iinfo(torch.uint8).min, dtype=cache_dtype) def _attn(self, query, key, value, registered_causal_mask, attention_mask=None, head_mask=None): device = query.device if self.use_cache_quantization: qk, qk_scale, qk_zero = key if self.use_cache_kernel and cache_autogptq_cuda_256 is not None: shape = query.shape[:-1] + (qk.shape[-2],) attn_weights = torch.zeros(shape, dtype=torch.float16, device=device) cache_autogptq_cuda_256.vecquant8matmul_batched_faster_old( query.contiguous() if query.dtype == torch.float16 else query.to(torch.float16).contiguous(), qk.transpose(-1, -2).contiguous(), attn_weights, qk_scale.contiguous() if qk_scale.dtype == torch.float16 else qk_scale.to(torch.float16).contiguous(), qk_zero.contiguous()if qk_zero.dtype == torch.float16 else qk_zero.to(torch.float16).contiguous()) # attn_weights = attn_weights.to(query.dtype).contiguous() else: key = dequantize_cache_torch(qk, qk_scale, qk_zero) attn_weights = torch.matmul(query, key.transpose(-1, -2)) else: attn_weights = torch.matmul(query, key.transpose(-1, -2)) if self.scale_attn_weights: if self.use_cache_quantization: size_temp = value[0].size(-1) else: size_temp = value.size(-1) attn_weights = attn_weights / torch.full( [], size_temp ** 0.5, dtype=attn_weights.dtype, device=attn_weights.device, ) if self.use_cache_quantization: query_length, key_length = query.size(-2), key[0].size(-2) else: query_length, key_length = query.size(-2), key.size(-2) causal_mask = registered_causal_mask[ :, :, key_length - query_length : key_length, :key_length ] mask_value = torch.finfo(attn_weights.dtype).min mask_value = torch.full([], mask_value, dtype=attn_weights.dtype).to( attn_weights.device ) attn_weights = torch.where( causal_mask, attn_weights.to(attn_weights.dtype), mask_value ) if attention_mask is not None: attn_weights = attn_weights + attention_mask attn_weights = nn.functional.softmax(attn_weights.float(), dim=-1) attn_weights = attn_weights.type(query.dtype) attn_weights = self.attn_dropout(attn_weights) if head_mask is not None: attn_weights = attn_weights * head_mask if self.use_cache_quantization: qv, qv_scale, qv_zero = value if self.use_cache_kernel and cache_autogptq_cuda_256 is not None: shape = attn_weights.shape[:-1] + (query.shape[-1],) attn_output = torch.zeros(shape, dtype=torch.float16, device=device) cache_autogptq_cuda_256.vecquant8matmul_batched_column_compression_faster_old( attn_weights.contiguous() if attn_weights.dtype == torch.float16 else attn_weights.to(torch.float16).contiguous(), qv.contiguous(), # dtype: int32 attn_output, qv_scale.contiguous() if qv_scale.dtype == torch.float16 else qv_scale.to(torch.float16).contiguous(), qv_zero.contiguous() if qv_zero.dtype == torch.float16 else qv_zero.to(torch.float16).contiguous()) if attn_output.dtype != query.dtype: attn_output = attn_output.to(query.dtype) attn_weights = attn_weights.to(query.dtype) else: value = dequantize_cache_torch(qv, qv_scale, qv_zero) attn_output = torch.matmul(attn_weights, value) else: attn_output = torch.matmul(attn_weights, value) attn_output = attn_output.transpose(1, 2) return attn_output, attn_weights def _upcast_and_reordered_attn( self, query, key, value, registered_causal_mask, attention_mask=None, head_mask=None ): bsz, num_heads, q_seq_len, dk = query.size() _, _, k_seq_len, _ = key.size() attn_weights = torch.empty( bsz * num_heads, q_seq_len, k_seq_len, dtype=torch.float32, device=query.device, ) scale_factor = 1.0 if self.scale_attn_weights: scale_factor /= float(value.size(-1)) ** 0.5 with autocast(enabled=False): q, k = query.reshape(-1, q_seq_len, dk), key.transpose(-1, -2).reshape( -1, dk, k_seq_len ) attn_weights = torch.baddbmm( attn_weights, q.float(), k.float(), beta=0, alpha=scale_factor ) attn_weights = attn_weights.reshape(bsz, num_heads, q_seq_len, k_seq_len) query_length, key_length = query.size(-2), key.size(-2) causal_mask = registered_causal_mask[ :, :, key_length - query_length : key_length, :key_length ] #registered_causal_mask是shape为(1, 1, max_positions, max_positions)的二值张量，其中max_positions为8192 mask_value = torch.finfo(attn_weights.dtype).min mask_value = torch.tensor(mask_value, dtype=attn_weights.dtype).to( attn_weights.device ) attn_weights = torch.where(causal_mask, attn_weights, mask_value) #causal_mask中为1的位置，attn_weights中保留，否则用mask_value填充 if attention_mask is not None: attn_weights = attn_weights + attention_mask attn_weights = nn.functional.softmax(attn_weights, dim=-1) if attn_weights.dtype != torch.float32: raise RuntimeError( "Error with upcasting, attn_weights does not have dtype torch.float32" ) attn_weights = attn_weights.type(value.dtype) attn_weights = self.attn_dropout(attn_weights) if head_mask is not None: attn_weights = attn_weights * head_mask attn_output = torch.matmul(attn_weights, value) return attn_output, attn_weights def _split_heads(self, tensor, num_heads, attn_head_size): new_shape = tensor.size()[:-1] + (num_heads, attn_head_size) tensor = tensor.view(new_shape) return tensor def _merge_heads(self, tensor, num_heads, attn_head_size): tensor = tensor.contiguous() new_shape = tensor.size()[:-2] + (num_heads * attn_head_size,) return tensor.view(new_shape) def forward( self, hidden_states: Optional[Tuple[torch.FloatTensor]], rotary_pos_emb_list: Optional[List[torch.Tensor]] = None, registered_causal_mask: Optional[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, output_attentions: Optional[bool] = False, 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) 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 self.use_cache_quantization: key = quantize_cache_v(key.permute(0, 2, 1, 3), bits=8, qmin=self.cache_qmin, qmax=self.cache_qmax) value = quantize_cache_v(value.permute(0, 2, 1, 3), bits=8, qmin=self.cache_qmin, qmax=self.cache_qmax) if layer_past is not None: past_key, past_value = layer_past[0], layer_past[1] if self.use_cache_quantization: # use_cache_quantization: # present=((q_key,key_scale,key_zero_point), # (q_value,value_scale,value_zero_point)) key = (torch.cat((past_key[0], key[0]), dim=2), torch.cat((past_key[1], key[1]), dim=2), torch.cat((past_key[2], key[2]), dim=2)) value = (torch.cat((past_value[0], value[0]), dim=2), torch.cat((past_value[1], value[1]), dim=2), torch.cat((past_value[2], value[2]), dim=2)) else: # not use_cache_quantization: # present=(key,value) key = torch.cat((past_key, key), dim=1) value = torch.cat((past_value, value), dim=1) if use_cache: present = (key, value) else: present = None if self.use_logn_attn and not self.training: if self.use_cache_quantization: seq_start = key[0].size(2) - query.size(1) seq_end = key[0].size(2) else: seq_start = key.size(1) - query.size(1) seq_end = key.size(1) logn_tensor = self.logn_tensor[:, seq_start:seq_end, :, :] query = query * logn_tensor.expand_as(query) #使用logn_attn时，query乘以logn_tensor，避免长文本时注意力不稳定 if ( self.use_flash_attn and flash_attn_unpadded_func is not None and not self.is_fp32 and query.is_cuda ): q, k, v = query, key, value context_layer = self.core_attention_flash(q, k, v, attention_mask=attention_mask) # b s h d -> b s (h d) context_layer = context_layer.flatten(2,3).contiguous() else: query = query.permute(0, 2, 1, 3) if not self.use_cache_quantization: key = key.permute(0, 2, 1, 3) value = value.permute(0, 2, 1, 3) if ( registered_causal_mask is None and self.use_flash_attn and flash_attn_unpadded_func is not None and not self.is_fp32 and not query.is_cuda ): raise Exception(_ERROR_INPUT_CPU_QUERY_WITH_FLASH_ATTN_ACTIVATED) attn_output, attn_weight = self._attn( query, key, value, registered_causal_mask, attention_mask, head_mask ) #如果没有使用flash attention，才会使用registered_causal_mask context_layer = self._merge_heads( attn_output, self.num_heads, self.head_dim ) attn_output = self.c_proj(context_layer) outputs = (attn_output, present) if output_attentions: if ( self.use_flash_attn and flash_attn_unpadded_func is not None and not self.is_fp32 ): raise ValueError("Cannot output attentions while using flash-attn") else: outputs += (attn_weight,) 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.c_proj = nn.Linear(ff_dim_in, config.hidden_size, bias=not config.no_bias) def forward(self, hidden_states): a1 = self.w1(hidden_states) a2 = self.w2(hidden_states) intermediate_parallel = a1 * F.silu(a2) output = self.c_proj(intermediate_parallel) return output class QWenBlock(nn.Module): def __init__(self, config): super().__init__() hidden_size = config.hidden_size self.bf16 = config.bf16 self.ln_1 = RMSNorm( hidden_size, eps=config.layer_norm_epsilon, ) self.attn = QWenAttention(config) self.ln_2 = RMSNorm( hidden_size, eps=config.layer_norm_epsilon, ) self.mlp = QWenMLP(config) def forward( self, hidden_states: Optional[Tuple[torch.FloatTensor]], rotary_pos_emb_list: Optional[List[torch.Tensor]] = None, registered_causal_mask: Optional[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) attn_outputs = self.attn( layernorm_output, rotary_pos_emb_list, registered_causal_mask=registered_causal_mask, layer_past=layer_past, attention_mask=attention_mask, head_mask=head_mask, use_cache=use_cache, output_attentions=output_attentions, ) 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 if use_cache: outputs = (hidden_states,) + outputs else: outputs = (hidden_states,) + outputs[1:] return outputs class QWenPreTrainedModel(PreTrainedModel): config_class = QWenConfig base_model_prefix = "transformer" is_parallelizable = False supports_gradient_checkpointing = True _no_split_modules = ["QWenBlock"] def __init__(self, *inputs, **kwargs): super().__init__(*inputs, **kwargs) def _init_weights(self, module): """Initialize the weights.""" if isinstance(module, nn.Linear): module.weight.data.normal_(mean=0.0, std=self.config.initializer_range) if module.bias is not None: module.bias.data.zero_() elif isinstance(module, nn.Embedding): module.weight.data.normal_(mean=0.0, std=self.config.initializer_range) if module.padding_idx is not None: module.weight.data[module.padding_idx].zero_() elif isinstance(module, RMSNorm): module.weight.data.fill_(1.0) for name, p in module.named_parameters(): if name == "c_proj.weight": p.data.normal_( mean=0.0, std=( self.config.initializer_range / math.sqrt(2 * self.config.num_hidden_layers) ), ) def _set_gradient_checkpointing(self, module, value=False): if isinstance(module, QWenModel): module.gradient_checkpointing = value class QWenModel(QWenPreTrainedModel): _keys_to_ignore_on_load_missing = ["attn.masked_bias"] def __init__(self, config): super().__init__(config) self.vocab_size = config.vocab_size self.num_hidden_layers = config.num_hidden_layers self.embed_dim = config.hidden_size self.use_cache_quantization = self.config.use_cache_quantization if hasattr(self.config, 'use_cache_quantization') else False self.gradient_checkpointing = False self.use_dynamic_ntk = config.use_dynamic_ntk self.seq_length = config.seq_length self.wte = nn.Embedding(self.vocab_size, self.embed_dim) self.drop = nn.Dropout(config.emb_dropout_prob) if config.rotary_pct == 1.0: self.rotary_ndims = None else: assert config.rotary_pct < 1 self.rotary_ndims = int( config.kv_channels * config.rotary_pct ) dim = ( self.rotary_ndims if self.rotary_ndims is not None else config.kv_channels ) self.rotary_emb = YaRNRotaryEmbedding(dim, base=config.rotary_emb_base,original_max_position_embeddings=config.seq_length) #self.rotary_emb = RotaryEmbedding(dim, base=config.rotary_emb_base) import warnings warnings.warn("使用YarnRotaryEmbedding，强制设置config.use_logn_attn = False，config.use_dynamic_ntk = True") config.use_logn_attn = False config.use_dynamic_ntk = True self.use_flash_attn = config.use_flash_attn self.is_fp32 = not (config.bf16 or config.fp16) if ( self.use_flash_attn and flash_attn_unpadded_func is not None and not self.is_fp32 ): self.registered_causal_mask = None else: max_positions = config.max_position_embeddings self.register_buffer( "registered_causal_mask", torch.tril( torch.ones((max_positions, max_positions), dtype=torch.bool) ).view(1, 1, max_positions, max_positions), persistent=False, ) self.h = nn.ModuleList( [ QWenBlock( config ) for i in range(config.num_hidden_layers) ] ) self.ln_f = RMSNorm( self.embed_dim, eps=config.layer_norm_epsilon, ) self.post_init() def get_input_embeddings(self): return self.wte def set_input_embeddings(self, new_embeddings): self.wte = new_embeddings def get_ntk_alpha(self, true_seq_len): ntk_alpha = true_seq_len / self.seq_length + 512/self.seq_length ntk_alpha = max(ntk_alpha, 1) # context_value = math.log(true_seq_len / self.seq_length, 2) + 1 # ntk_alpha = 2 ** math.ceil(context_value) - 1 # ntk_alpha = max(ntk_alpha, 1) #假设seq_length=2k，true_seq_len=8k，context_value=3，ntk_alpha=7，相当于扩展到14k长度 return ntk_alpha def forward( self, input_ids: Optional[torch.LongTensor] = None, past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None, attention_mask: Optional[torch.FloatTensor] = None, token_type_ids: Optional[torch.LongTensor] = None, position_ids: Optional[torch.LongTensor] = 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_hidden_states: Optional[bool] = None, return_dict: Optional[bool] = None, ): 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 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: input_shape = input_ids.size() input_ids = input_ids.view(-1, input_shape[-1]) batch_size = input_ids.shape[0] elif inputs_embeds is not None: input_shape = inputs_embeds.size()[:-1] batch_size = inputs_embeds.shape[0] else: raise ValueError("You have to specify either input_ids or inputs_embeds") device = input_ids.device if input_ids is not None else inputs_embeds.device if token_type_ids is not None: token_type_ids = token_type_ids.view(-1, input_shape[-1]) if position_ids is not None: position_ids = position_ids.view(-1, input_shape[-1]) if past_key_values is None: past_length = 0 past_key_values = tuple([None] * len(self.h)) else: if self.use_cache_quantization: past_length = past_key_values[0][0][0].size(2) else: past_length = past_key_values[0][0].size(-2) if position_ids is None: position_ids = torch.arange( past_length, input_shape[-1] + past_length, dtype=torch.long, device=device, ) position_ids = position_ids.unsqueeze(0).view(-1, input_shape[-1]) if attention_mask is not None: if batch_size <= 0: raise ValueError("batch_size has to be defined and > 0") attention_mask = attention_mask.view(batch_size, -1) attention_mask = attention_mask[:, None, None, :] 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: inputs_embeds = self.wte(input_ids) hidden_states = inputs_embeds kv_seq_len = hidden_states.size()[1] if past_key_values[0] is not None: # past key values[0][0] shape: bs * seq_len * head_num * dim if self.use_cache_quantization: kv_seq_len += past_key_values[0][0][0].shape[2] else: kv_seq_len += past_key_values[0][0].shape[1] #if self.training or not self.use_dynamic_ntk: if not self.use_dynamic_ntk: ntk_alpha_list = [1.0] elif kv_seq_len != hidden_states.size()[1]: ntk_alpha_list = self.rotary_emb._ntk_alpha_cached_list else: ntk_alpha_list = [] if attention_mask is not None and kv_seq_len > self.seq_length: true_seq_lens = attention_mask.squeeze(1).squeeze(1).eq(0).sum(dim=-1, dtype=torch.int32) for i in range(hidden_states.size()[0]): #给batch中的每个样本计算ntk_alpha，计算方法是 true_seq_len / self.seq_length。qwen-7b中，self.seq_length=8192，qwen-14b中，self.seq_length=2048 true_seq_len = true_seq_lens[i].item() ntk_alpha = self.get_ntk_alpha(true_seq_len) ntk_alpha_list.append(ntk_alpha) else: ntk_alpha = self.get_ntk_alpha(kv_seq_len) ntk_alpha_list.append(ntk_alpha) self.rotary_emb._ntk_alpha_cached_list = ntk_alpha_list rotary_pos_emb_list = [] for ntk_alpha in ntk_alpha_list: rotary_pos_emb = self.rotary_emb(kv_seq_len, ntk_alpha=ntk_alpha) #训练时，ntk_alpha=1.0，rotary_emb根据kv_seq_len生成 rotary_pos_emb_list.append(rotary_pos_emb) hidden_states = self.drop(hidden_states) output_shape = input_shape + (hidden_states.size(-1),) if self.gradient_checkpointing and self.training: if use_cache: logger.warning_once( "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..." ) use_cache = False presents = () if use_cache else None all_self_attentions = () if output_attentions else None all_hidden_states = () if output_hidden_states else None for i, (block, layer_past) in enumerate(zip(self.h, past_key_values)): if output_hidden_states: all_hidden_states = all_hidden_states + (hidden_states,) if self.gradient_checkpointing and self.training: def create_custom_forward(module): def custom_forward(*inputs): # None for past_key_value return module(*inputs, use_cache, output_attentions) return custom_forward outputs = torch.utils.checkpoint.checkpoint( create_custom_forward(block), hidden_states, rotary_pos_emb_list, self.registered_causal_mask, None, attention_mask, head_mask[i], encoder_hidden_states, encoder_attention_mask, ) else: outputs = block( hidden_states, layer_past=layer_past, rotary_pos_emb_list=rotary_pos_emb_list, registered_causal_mask=self.registered_causal_mask, 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] 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) # Add last hidden state if output_hidden_states: all_hidden_states = all_hidden_states + (hidden_states,) if not return_dict: return tuple( v for v in [hidden_states, presents, all_hidden_states] if v is not None ) return BaseModelOutputWithPast( last_hidden_state=hidden_states, past_key_values=presents, hidden_states=all_hidden_states, attentions=all_self_attentions, ) class QWenLMHeadModel(QWenPreTrainedModel): _keys_to_ignore_on_load_missing = [r"h\.\d+\.attn\.rotary_emb\.inv_freq"] _keys_to_ignore_on_load_unexpected = [r"h\.\d+\.attn\.masked_bias"] def __init__(self, config): super().__init__(config) assert ( config.bf16 + config.fp16 + config.fp32 <= 1 ), "Only one of \"bf16\", \"fp16\", \"fp32\" can be true" logger.warn( "Warning: please make sure that you are using the latest codes and checkpoints, " "especially if you used Qwen-7B before 09.25.2023." "请使用最新模型和代码，尤其如果你在9月25日前已经开始使用Qwen-7B，千万注意不要使用错误代码和模型。" ) autoset_precision = config.bf16 + config.fp16 + config.fp32 == 0 if autoset_precision: if SUPPORT_BF16: logger.warn( "The model is automatically converting to bf16 for faster inference. " "If you want to disable the automatic precision, please manually add bf16/fp16/fp32=True to \"AutoModelForCausalLM.from_pretrained\"." ) config.bf16 = True elif SUPPORT_FP16: logger.warn( "The model is automatically converting to fp16 for faster inference. " "If you want to disable the automatic precision, please manually add bf16/fp16/fp32=True to \"AutoModelForCausalLM.from_pretrained\"." ) config.fp16 = True else: config.fp32 = True if config.bf16 and SUPPORT_CUDA and not SUPPORT_BF16: logger.warn("Your device does NOT seem to support bf16, you can switch to fp16 or fp32 by by passing fp16/fp32=True in \"AutoModelForCausalLM.from_pretrained\".") if config.fp16 and SUPPORT_CUDA and not SUPPORT_FP16: logger.warn("Your device does NOT support faster inference with fp16, please switch to fp32 which is likely to be faster") if config.fp32: if SUPPORT_BF16: logger.warn("Your device support faster inference by passing bf16=True in \"AutoModelForCausalLM.from_pretrained\".") elif SUPPORT_FP16: logger.warn("Your device support faster inference by passing fp16=True in \"AutoModelForCausalLM.from_pretrained\".") if config.use_flash_attn == "auto": if config.bf16 or config.fp16: logger.warn("Try importing flash-attention for faster inference...") config.use_flash_attn = True else: config.use_flash_attn = False if config.use_flash_attn and config.fp32: logger.warn("Flash attention will be disabled because it does NOT support fp32.") if config.use_flash_attn: _import_flash_attn() if hasattr(config, 'use_cache_quantization') and config.use_cache_quantization: config.use_flash_attn = False if hasattr(config, 'use_cache_kernel') and config.use_cache_kernel: try: from kernels.cpp_kernels import cache_autogptq_cuda_256 except ImportError: cache_autogptq_cuda_256 = None self.transformer = QWenModel(config) self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False) if config.bf16: self.transformer.bfloat16() self.lm_head.bfloat16() if config.fp16: self.transformer.half() self.lm_head.half() self.post_init() def get_output_embeddings(self): return self.lm_head def set_output_embeddings(self, new_embeddings): self.lm_head = new_embeddings def prepare_inputs_for_generation( self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs ): token_type_ids = kwargs.get("token_type_ids", None) if past_key_values: input_ids = input_ids[:, -1].unsqueeze(-1) if token_type_ids is not None: token_type_ids = token_type_ids[:, -1].unsqueeze(-1) attention_mask = kwargs.get("attention_mask", None) position_ids = kwargs.get("position_ids", None) if attention_mask is not None and position_ids is None: position_ids = attention_mask.long().cumsum(-1) - 1 position_ids.masked_fill_(attention_mask == 0, 1) if past_key_values: position_ids = position_ids[:, -1].unsqueeze(-1) else: position_ids = None if inputs_embeds is not None and past_key_values is None: model_inputs = {"inputs_embeds": inputs_embeds} else: model_inputs = {"input_ids": input_ids} model_inputs.update( { "past_key_values": past_key_values, "use_cache": kwargs.get("use_cache"), "position_ids": position_ids, "attention_mask": attention_mask, "token_type_ids": token_type_ids, } ) return model_inputs def forward( self, input_ids: Optional[torch.LongTensor] = None, past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None, attention_mask: Optional[torch.FloatTensor] = None, token_type_ids: Optional[torch.LongTensor] = None, position_ids: Optional[torch.LongTensor] = 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, output_hidden_states: Optional[bool] = None, return_dict: Optional[bool] = None, ) -> Union[Tuple, CausalLMOutputWithPast]: return_dict = ( return_dict if return_dict is not None else self.config.use_return_dict ) transformer_outputs = self.transformer( input_ids, past_key_values=past_key_values, attention_mask=attention_mask, token_type_ids=token_type_ids, position_ids=position_ids, 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, output_hidden_states=output_hidden_states, return_dict=return_dict, ) hidden_states = transformer_outputs[0] lm_logits = self.lm_head(hidden_states) loss = None if labels is not None: labels = labels.to(lm_logits.device) shift_logits = lm_logits[..., :-1, :].contiguous() shift_labels = labels[..., 1:].contiguous() loss_fct = CrossEntropyLoss() loss = loss_fct( shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1) ) if not return_dict: output = (lm_logits,) + transformer_outputs[1:] return ((loss,) + output) if loss is not None else output #训练时节约显存 # if self.training: # lm_logits=None return CausalLMOutputWithPast( loss=loss, logits=lm_logits, past_key_values=transformer_outputs.past_key_values, hidden_states=transformer_outputs.hidden_states, attentions=transformer_outputs.attentions, ) @staticmethod def _reorder_cache( past_key_values: Tuple[Tuple[torch.Tensor]], beam_idx: torch.Tensor ) -> Tuple[Tuple[torch.Tensor]]: return tuple( tuple( past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past ) for layer_past in past_key_values ) def chat( self, tokenizer: PreTrainedTokenizer, query: str, history: Optional[HistoryType], system: str = "You are a helpful assistant.", append_history: bool = True, 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, _ERROR_STREAM_IN_CHAT assert generation_config.chat_format == 'chatml', _ERROR_BAD_CHAT_FORMAT if history is None: history = [] if stop_words_ids is None: stop_words_ids = [] max_window_size = kwargs.get('max_window_size', None) if max_window_size is None: max_window_size = generation_config.max_window_size raw_text, context_tokens = make_context( tokenizer, query, history=history, system=system, max_window_size=max_window_size, chat_format=generation_config.chat_format, ) stop_words_ids.extend(get_stop_words_ids( generation_config.chat_format, tokenizer )) input_ids = torch.tensor([context_tokens]).to(self.device) outputs = self.generate( input_ids, stop_words_ids=stop_words_ids, return_dict_in_generate=False, generation_config=generation_config, **kwargs, ) response = decode_tokens( outputs[0], tokenizer, raw_text_len=len(raw_text), context_length=len(context_tokens), chat_format=generation_config.chat_format, verbose=False, errors='replace' ) if append_history: history.append((query, response)) return response, history def chat_stream( self, tokenizer: PreTrainedTokenizer, query: str, history: Optional[HistoryType], system: str = "You are a helpful assistant.", stop_words_ids: Optional[List[List[int]]] = None, logits_processor: Optional[LogitsProcessorList] = None, generation_config: Optional[GenerationConfig] = None, **kwargs, ) -> Generator[str, Any, None]: generation_config = generation_config if generation_config is not None else self.generation_config assert generation_config.chat_format == 'chatml', _ERROR_BAD_CHAT_FORMAT if history is None: history = [] if stop_words_ids is None: stop_words_ids = [] max_window_size = kwargs.get('max_window_size', None) if max_window_size is None: max_window_size = generation_config.max_window_size raw_text, context_tokens = make_context( tokenizer, query, history=history, system=system, max_window_size=max_window_size, chat_format=generation_config.chat_format, ) stop_words_ids.extend(get_stop_words_ids( generation_config.chat_format, tokenizer )) if stop_words_ids is not None: stop_words_logits_processor = StopWordsLogitsProcessor( stop_words_ids=stop_words_ids, eos_token_id=generation_config.eos_token_id, ) if logits_processor is None: logits_processor = LogitsProcessorList([stop_words_logits_processor]) else: logits_processor.append(stop_words_logits_processor) input_ids = torch.tensor([context_tokens]).to(self.device) from transformers_stream_generator.main import NewGenerationMixin, StreamGenerationConfig self.__class__.generate_stream = NewGenerationMixin.generate self.__class__.sample_stream = NewGenerationMixin.sample_stream stream_config = StreamGenerationConfig(**generation_config.to_dict(), do_stream=True) def stream_generator(): outputs = [] for token in self.generate_stream( input_ids, return_dict_in_generate=False, generation_config=stream_config, logits_processor=logits_processor, seed=-1, **kwargs): outputs.append(token.item()) yield tokenizer.decode(outputs, skip_special_tokens=True, errors='ignore') return stream_generator() def generate( self, inputs: Optional[torch.Tensor] = None, generation_config: Optional[GenerationConfig] = None, 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, ) -> Union[GenerateOutput, torch.LongTensor]: generation_config = generation_config if generation_config is not None else self.generation_config # Process stop_words_ids. stop_words_ids = kwargs.pop("stop_words_ids", None) if stop_words_ids is None and generation_config is not None: stop_words_ids = getattr(generation_config, "stop_words_ids", None) if stop_words_ids is None: stop_words_ids = getattr(generation_config, "stop_words_ids", None) if stop_words_ids is not None: stop_words_logits_processor = StopWordsLogitsProcessor( stop_words_ids=stop_words_ids, eos_token_id=generation_config.eos_token_id, ) if logits_processor is None: logits_processor = LogitsProcessorList([stop_words_logits_processor]) else: logits_processor.append(stop_words_logits_processor) return super().generate( inputs, generation_config=generation_config, 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, ) class RotaryEmbedding(torch.nn.Module): def __init__(self, dim, base=10000): super().__init__() self.dim = dim self.base = base inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float() / dim)) self.register_buffer("inv_freq", inv_freq, persistent=False) if importlib.util.find_spec("einops") is None: raise RuntimeError("einops is required for Rotary Embedding") self._rotary_pos_emb_cache = None self._seq_len_cached = 0 self._ntk_alpha_cached = 1.0 self._ntk_alpha_cached_list = [1.0] def update_rotary_pos_emb_cache(self, max_seq_len, offset=0, ntk_alpha=1.0): seqlen = max_seq_len + offset if seqlen > self._seq_len_cached or ntk_alpha != self._ntk_alpha_cached: #计算新的base。ntk_alpha=1时，base不变；ntk_alpha>1时，base变大 base = self.base * ntk_alpha ** (self.dim / (self.dim - 2)) #波长=2*pi*base^(2d/D),即频率的倒数。此处算出了每个维度对应的频率。由于每两个维度对应一个频率，所以频率的个数为dim/2 self.inv_freq = 1.0 / ( base ** ( torch.arange(0, self.dim, 2, device=self.inv_freq.device).float() / self.dim ) ) self._seq_len_cached = max(seqlen+512, 16) self._ntk_alpha_cached = ntk_alpha seq = torch.arange(self._seq_len_cached, device=self.inv_freq.device) freqs = torch.outer(seq.type_as(self.inv_freq), self.inv_freq) #做外积，获得seq_len*(dim/2)的矩阵 emb = torch.cat((freqs, freqs), dim=-1) #获得(seq_len*2)*dim的矩阵 from einops import rearrange emb = rearrange(emb, "n d -> 1 n 1 d") #获得1*(seq_len*2)*1*dim的矩阵 cos, sin = emb.cos(), emb.sin() self._rotary_pos_emb_cache = [cos, sin] def forward(self, max_seq_len, offset=0, ntk_alpha=1.0): self.update_rotary_pos_emb_cache(max_seq_len, offset, ntk_alpha) cos, sin = self._rotary_pos_emb_cache return [cos[:, offset : offset + max_seq_len], sin[:, offset : offset + max_seq_len]] class YaRNRotaryEmbedding(RotaryEmbedding): def __init__(self, dim, base=10000,interleaved=False,extrapolation_factor=1, attn_factor=1, beta_fast=32, beta_slow=1,original_max_position_embeddings=2048): """ interleaved: if True, rotate pairs of even and odd dimensions (GPT-J style) instead of 1st half and 2nd half (GPT-NeoX style). pos_idx_in_fp32: if True, the position indices [0.0, ..., seqlen - 1] are in fp32, otherwise they might be in lower precision. This option was added because previously (before 2023-07-02), when we construct the position indices, we use the dtype of self.inv_freq. In most cases this would be fp32, but if the model is trained in pure bf16 (not mixed precision), then self.inv_freq would be bf16, and the position indices are also in bf16. Because of the limited precision of bf16 (e.g. 1995.0 is rounded to 2000.0), the embeddings for some positions will coincide. To maintain compatibility with models previously trained in pure bf16, we add this option. scaling_factor: RotaryEmbedding extended with YaRN scaling. """ super().__init__(dim, base) self.interleaved = interleaved self.extrapolation_factor = extrapolation_factor self.attn_factor = attn_factor self.beta_fast = beta_fast self.beta_slow = beta_slow self.original_max_position_embeddings = original_max_position_embeddings def update_rotary_pos_emb_cache(self, max_seq_len, offset=0, ntk_alpha=1.0): seqlen = max_seq_len + offset if seqlen > self._seq_len_cached or ntk_alpha != self._ntk_alpha_cached: #波长=2*pi*base^(2d/D),即频率的倒数。此处算出了每个维度对应的频率。由于每两个维度对应一个频率，所以频率的个数为dim/2 self._compute_inv_freq(ntk_alpha, device=self.inv_freq.device) self.mscale = float(_yarn_get_mscale(ntk_alpha) * self.attn_factor) self._seq_len_cached = max(seqlen+512, 16) self._ntk_alpha_cached = ntk_alpha seq = torch.arange(self._seq_len_cached, device=self.inv_freq.device) freqs = torch.outer(seq.type_as(self.inv_freq), self.inv_freq) #seq和逆频率做外积，获得seq_len*(dim/2)的矩阵 emb = torch.cat((freqs, freqs), dim=-1) #获得(seq_len*2)*dim的矩阵 from einops import rearrange emb = rearrange(emb, "n d -> 1 n 1 d") #获得1*(seq_len*2)*1*dim的矩阵 cos, sin = emb.cos() * self.mscale, emb.sin() * self.mscale self._rotary_pos_emb_cache = [cos, sin] def _compute_inv_freq(self, scaling_factor, device=None): pos_freqs = self.base ** (torch.arange(0, self.dim, 2).float().to(device) / self.dim) inv_freq_extrapolation = 1.0 / pos_freqs inv_freq_interpolation = 1.0 / (scaling_factor * pos_freqs) low, high = _yarn_find_correction_range(self.beta_fast, self.beta_slow, self.dim, self.base, self.original_max_position_embeddings) inv_freq_mask = (1 - _yarn_linear_ramp_mask(low, high, self.dim // 2).float().to(device)) * self.extrapolation_factor # Get n-d rotational scaling corrected for extrapolation inv_freq = inv_freq_interpolation * (1 - inv_freq_mask) + inv_freq_extrapolation * inv_freq_mask inv_freq=inv_freq.float() self.register_buffer("inv_freq", inv_freq, persistent=False) # Inverse dim formula to find dim based on number of rotations def _yarn_find_correction_dim(num_rotations, dim, base=10000, max_position_embeddings=2048): return (dim * math.log(max_position_embeddings/(num_rotations * 2 * math.pi)))/(2 * math.log(base)) # Find dim range bounds based on rotations def _yarn_find_correction_range(low_rot, high_rot, dim, base=10000, max_position_embeddings=2048): low = math.floor(_yarn_find_correction_dim( low_rot, dim, base, max_position_embeddings)) high = math.ceil(_yarn_find_correction_dim( high_rot, dim, base, max_position_embeddings)) return max(low, 0), min(high, dim-1) # Clamp values just in case def _yarn_linear_ramp_mask(min, max, dim): if min == max: max += 0.001 # Prevent singularity linear_func = (torch.arange(dim, dtype=torch.float32) - min) / (max - min) ramp_func = torch.clamp(linear_func, 0, 1) return ramp_func def _yarn_get_mscale(scale=1): if scale <= 1: return 1.0 return 0.1 * math.log(scale) + 1.0 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) def apply_rotary_pos_emb(t, freqs): cos, sin = freqs if apply_rotary_emb_func is not None and t.is_cuda: t_ = t.float() cos = cos.squeeze(0).squeeze(1)[:, : cos.shape[-1] // 2] sin = sin.squeeze(0).squeeze(1)[:, : sin.shape[-1] // 2] output = apply_rotary_emb_func(t_, cos, sin).type_as(t) return output else: rot_dim = freqs[0].shape[-1] cos, sin = freqs t_, t_pass_ = t[..., :rot_dim], t[..., rot_dim:] t_ = t_.float() t_pass_ = t_pass_.float() t_ = (t_ * cos) + (_rotate_half(t_) * sin) return torch.cat((t_, t_pass_), dim=-1).type_as(t) class RMSNorm(torch.nn.Module): def __init__(self, dim: int, eps: float = 1e-6): super().__init__() self.eps = eps self.weight = nn.Parameter(torch.ones(dim)) def _norm(self, x): return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps) def forward(self, x): if rms_norm is not None and x.is_cuda: return rms_norm(x, self.weight, self.eps) else: output = self._norm(x.float()).type_as(x) return output * self.weight