Witllm/wit/modeling_wit.py

400 lines
16 KiB
Python
Raw Normal View History

2024-02-04 23:48:24 +08:00
import copy
import math
import os
import sys
import gc
from tqdm import auto as tqdm_lib
import json
from typing import Optional, Tuple, Union, Callable, List, Any, Generator
from einops import rearrange
import torch
import torch.nn.functional as F
import torch.utils.checkpoint
from torch.nn import CrossEntropyLoss
from torch import nn
from safetensors.torch import load_file as safe_load_file
from safetensors.torch import save_file as safe_save_file
from qwen_generation_utils import (
make_context,
decode_tokens,
)
sys.path.append("..")
from tools import show
from tools import mem_tracker
# tracker = mem_tracker.MemTracker()
# tracker.track()
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):
return self._norm(x.float()).type_as(x) * self.weight
class QWenAttention(nn.Module):
def __init__(self, config, index):
super().__init__()
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
2024-03-04 21:41:46 +08:00
self.c_attn = nn.Linear(config.hidden_size, 3 * self.hidden_size)
self.c_proj = nn.Linear(config.hidden_size, self.hidden_size, bias=not config.no_bias)
2024-02-04 23:48:24 +08:00
self.attn_dropout = nn.Dropout(config.attn_dropout_prob)
self.index = index
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)
class QWenMLP(nn.Module):
def __init__(self, config):
super().__init__()
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)
class QWenBlock(nn.Module):
def __init__(self, config, index):
super().__init__()
self.ln_1 = RMSNorm(
config.hidden_size,
eps=config.layer_norm_epsilon,
)
self.attn = QWenAttention(config, index)
self.ln_2 = RMSNorm(
config.hidden_size,
eps=config.layer_norm_epsilon,
)
self.mlp = QWenMLP(config)
self.index = index
class QWenModel(nn.Module):
def __init__(self, config):
super().__init__()
self.wte = nn.Embedding(config.vocab_size, config.hidden_size)
self.drop = nn.Dropout(config.emb_dropout_prob)
2024-03-04 21:41:46 +08:00
self.dim = config.hidden_size // config.num_attention_heads
2024-02-04 23:48:24 +08:00
self.h = nn.ModuleList([QWenBlock(config, i) for i in range(config.num_hidden_layers)])
self.ln_f = RMSNorm(
config.hidden_size,
eps=config.layer_norm_epsilon,
)
self.base = config.rotary_emb_base
2024-03-04 21:41:46 +08:00
inv_freq = 1.0 / (self.base ** (torch.arange(0, self.dim, 2).float() / self.dim))
2024-02-04 23:48:24 +08:00
self.register_buffer("inv_freq", inv_freq, persistent=False)
self._rotary_pos_emb_cache = None
self._seq_len_cached = 0
self._ntk_alpha_cached = 1.0
def update_rotary_pos_emb_cache(self, seqlen, ntk_alpha=1.0):
if seqlen > self._seq_len_cached or ntk_alpha != self._ntk_alpha_cached:
base = self.base * ntk_alpha ** (self.dim / (self.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(2 * seqlen, 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)
emb = torch.cat((freqs, freqs), dim=-1)
emb = rearrange(emb, "n d -> 1 n 1 d")
cos, sin = emb.cos(), emb.sin()
self._rotary_pos_emb_cache = [cos, sin]
class QWenLMHeadModel(nn.Module):
def __init__(self, config):
super().__init__()
self.config = config
self.transformer = QWenModel(config)
self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
2024-02-25 20:20:32 +08:00
def forward(
self,
input_ids: Optional[torch.LongTensor] = None,
labels: Optional[torch.LongTensor] = None,
token_type_ids: Optional[torch.LongTensor] = None,
**kwargs,
):
runner = QwenRunner(self)
return runner.forwardQWen(input_ids, labels)
2024-02-04 23:48:24 +08:00
def from_pretrained(cls, pretrained_model_name_or_path: Optional[Union[str, os.PathLike]]):
pretrained_model_name_or_path = str(pretrained_model_name_or_path)
resolved_archive_file = os.path.join(pretrained_model_name_or_path, "model.safetensors.index.json")
print(f"loading weights file {resolved_archive_file}")
with open(resolved_archive_file, "r") as f:
index = json.loads(f.read())
shard_filenames = sorted(set(index["weight_map"].values()))
resolved_archive_file = [os.path.join(pretrained_model_name_or_path, f) for f in shard_filenames]
model = cls._load_pretrained_model(resolved_archive_file)
return model
def _load_state_dict_into_model(self, model_to_load, state_dict, start_prefix):
metadata = getattr(state_dict, "_metadata", None)
state_dict = state_dict.copy()
if metadata is not None:
state_dict._metadata = metadata
error_msgs = []
def load(module: nn.Module, state_dict, prefix=""):
local_metadata = {} if metadata is None else metadata.get(prefix[:-1], {})
args = (state_dict, prefix, local_metadata, True, [], [], error_msgs)
if len([key for key in state_dict if key.startswith(prefix)]) > 0:
module._load_from_state_dict(*args)
for name, child in module._modules.items():
if child is not None:
load(child, state_dict, prefix + name + ".")
load(model_to_load, state_dict, prefix=start_prefix)
del state_dict
return error_msgs
def _load_pretrained_model(cls, resolved_archive_file):
start_prefix = ""
model_to_load = cls
if len(resolved_archive_file) > 1:
resolved_archive_file = tqdm_lib.tqdm(resolved_archive_file, desc="Loading checkpoint shards")
for shard_file in resolved_archive_file:
state_dict = safe_load_file(shard_file)
cls._load_state_dict_into_model(model_to_load, state_dict, start_prefix)
del state_dict # force memory release
gc.collect()
print(f"All model checkpoint weights were used when initializing {cls.__class__.__name__}.\n")
return cls
class QwenRunner:
def __init__(self, qwen):
self.qwen = qwen
@torch.no_grad()
def Chat(
self,
tokenizer,
query: str,
query_assistant: str,
2024-02-06 14:08:45 +08:00
gen_length=0,
2024-02-04 23:48:24 +08:00
system: str = "You are a helpful assistant.",
history=[],
):
qwen = self.qwen
history = copy.deepcopy(history)
2024-02-06 14:08:45 +08:00
self.qwen.config.pad_token_id = tokenizer.eod_id
self.qwen.config.eos_token_id = tokenizer.eod_id
2024-02-04 23:48:24 +08:00
raw_text, context_tokens = self.prepareInput(tokenizer, query, query_assistant, history, system)
input_ids = torch.tensor([context_tokens]).to(next(qwen.parameters()).device)
self.unfinished_sequences = torch.ones(input_ids.shape[0], dtype=torch.long, device=input_ids.device)
2024-02-06 14:08:45 +08:00
input_length = input_ids.shape[1]
2024-02-04 23:48:24 +08:00
while True:
outputs = self.forwardQWen(input_ids)
next_token_scores = outputs[:, -1, :]
next_token_scores = self.repetition_penalty(input_ids, next_token_scores)
next_token_scores = self.top_p(next_token_scores)
next_tokens = self.sample(next_token_scores)
finish, next_tokens = self.isFinish(next_tokens)
if finish:
break
input_ids = torch.cat([input_ids, next_tokens], dim=-1)
2024-02-06 14:08:45 +08:00
if gen_length != 0 and (input_length + gen_length) < input_ids.shape[1]:
break
2024-02-04 23:48:24 +08:00
decoded, response, end_reason = decode_tokens(
input_ids[0],
tokenizer,
raw_text_len=len(raw_text),
context_length=len(context_tokens),
errors="replace",
)
history.append((query, response))
return input_ids[0].cpu().tolist(), history, decoded
def _rotate_half(self, x):
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(self, t, freqs):
rot_dim = freqs[0].shape[-1]
cos, sin = freqs
t_float = t.float()
2024-03-07 16:30:37 +08:00
t_rot = t_float[..., :rot_dim]
t_pass = t_float[..., rot_dim:]
2024-02-04 23:48:24 +08:00
t_rot = (t_rot * cos) + (self._rotate_half(t_rot) * sin)
return torch.cat((t_rot, t_pass), dim=-1).type_as(t)
def split_heads(
self,
attention,
hidden_states: Optional[Tuple[torch.FloatTensor]],
):
atten = attention
mixed_x_layer = atten.c_attn(hidden_states)
query, key, value = mixed_x_layer.split(atten.split_size, dim=2)
query = atten._split_heads(query, atten.num_heads, atten.head_dim)
key = atten._split_heads(key, atten.num_heads, atten.head_dim)
value = atten._split_heads(value, atten.num_heads, atten.head_dim)
return query, key, value
def pos_emb(self, query, key, rotary_pos_emb_list):
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
query = self.apply_rotary_pos_emb(query, rotary_pos_emb[0])
key = self.apply_rotary_pos_emb(key, rotary_pos_emb[1])
return query, key
def attention(self, attention, query, key, value, causal_mask):
query = query.permute(0, 2, 1, 3)
key = key.permute(0, 2, 1, 3)
value = value.permute(0, 2, 1, 3)
attn_output = F.scaled_dot_product_attention(query, key, value, attn_mask=causal_mask).transpose(1, 2)
context_layer = attention._merge_heads(attn_output, attention.num_heads, attention.head_dim)
attn_output = attention.c_proj(context_layer)
return attn_output
def build_mask(self, query):
size = query.size(1)
causal_mask = torch.tril(torch.ones((size, size), dtype=torch.bool, device=query.device)).view(1, 1, size, size)
return causal_mask
def forwardAttention(
self,
attention,
hidden_states: Optional[Tuple[torch.FloatTensor]],
rotary_pos_emb_list: Optional[List[List[torch.Tensor]]] = None,
):
query, key, value = self.split_heads(attention, hidden_states)
query, key = self.pos_emb(query, key, rotary_pos_emb_list)
causal_mask = self.build_mask(query)
return self.attention(attention, query, key, value, causal_mask)
def forwardQWenBlock(
self,
block,
hidden_states: Optional[Tuple[torch.FloatTensor]],
rotary_pos_emb_list: Optional[List[List[torch.Tensor]]] = None,
):
layernorm_output = block.ln_1(hidden_states)
attn_outputs = self.forwardAttention(block.attn, layernorm_output, rotary_pos_emb_list)
2024-03-20 22:23:52 +08:00
layernorm_input = attn_outputs + hidden_states
2024-02-04 23:48:24 +08:00
layernorm_output = block.ln_2(layernorm_input)
a1 = block.mlp.w1(layernorm_output)
a2 = block.mlp.w2(layernorm_output)
intermediate_parallel = a1 * F.silu(a2)
mlp_output = block.mlp.c_proj(intermediate_parallel)
hidden_states = layernorm_input + mlp_output
return hidden_states
def forwardQWen(
self,
input_ids: Optional[torch.LongTensor] = None,
labels: Optional[torch.LongTensor] = None,
):
transfm = self.qwen.transformer
input_shape = input_ids.size()
input_ids = input_ids.view(-1, input_shape[-1])
hidden_states = transfm.wte(input_ids)
kv_seq_len = hidden_states.size()[1]
transfm.update_rotary_pos_emb_cache(kv_seq_len, ntk_alpha=1.0)
cos, sin = transfm._rotary_pos_emb_cache
rotary_pos_emb_list = [[cos[:, :kv_seq_len], sin[:, :kv_seq_len]]]
hidden_states = transfm.drop(hidden_states)
output_shape = input_shape + (hidden_states.size(-1),)
for block in transfm.h:
hidden_states = self.forwardQWenBlock(block, hidden_states, rotary_pos_emb_list=rotary_pos_emb_list)
hidden_states = transfm.ln_f(hidden_states)
hidden_states = hidden_states.view(output_shape)
lm_logits = self.qwen.lm_head(hidden_states)
loss = None
if labels is not None:
labels = labels.to(lm_logits.device)
2024-03-04 21:41:46 +08:00
shift_labels = labels[..., 1:].contiguous().view(-1)
2024-02-04 23:48:24 +08:00
shift_logits = lm_logits[..., :-1, :].contiguous()
2024-03-04 21:41:46 +08:00
shift_logits = shift_logits.view(-1, shift_logits.size(-1))
2024-03-05 22:08:37 +08:00
mask = shift_labels < self.qwen.config.vocab_size
2024-03-04 21:41:46 +08:00
shift_labels = shift_labels[mask]
shift_logits = shift_logits[mask]
2024-03-07 16:30:37 +08:00
# m = torch.max(shift_logits, 1).indices.cpu().numpy()
# ll = shift_labels.cpu().numpy()
2024-03-05 22:08:37 +08:00
loss = CrossEntropyLoss()(shift_logits, shift_labels)
2024-02-25 20:20:32 +08:00
return lm_logits, loss
2024-02-04 23:48:24 +08:00
def prepareInput(self, tokenizer, query, query_assistant, history, system):
return make_context(tokenizer, query, query_assistant, history=history, system=system)
def repetition_penalty(self, input_ids, next_token_scores):
penalty = self.qwen.config.repetition_penalty
score = torch.gather(next_token_scores, 1, input_ids)
# if score < 0 then repetition penalty has to be multiplied to reduce the token probabilities
score = torch.where(score < 0, score * penalty, score / penalty)
next_token_scores = next_token_scores.scatter_(1, input_ids, score)
return next_token_scores
def top_p(self, next_token_scores):
top_p = self.qwen.config.top_p
filter_value = -float("Inf")
min_tokens_to_keep = 1
sorted_logits, sorted_indices = torch.sort(next_token_scores, descending=False)
cumulative_probs = sorted_logits.softmax(dim=-1).cumsum(dim=-1)
# Remove tokens with cumulative top_p above the threshold (token with 0 are kept)
sorted_indices_to_remove = cumulative_probs <= (1 - top_p)
# Keep at least min_tokens_to_keep
sorted_indices_to_remove[..., -min_tokens_to_keep:] = 0
# scatter sorted tensors to original indexing
indices_to_remove = sorted_indices_to_remove.scatter(1, sorted_indices, sorted_indices_to_remove)
next_token_scores = next_token_scores.masked_fill(indices_to_remove, filter_value)
return next_token_scores
def sample(self, next_token_scores):
probs = nn.functional.softmax(next_token_scores, dim=-1)
next_tokens = torch.multinomial(probs, num_samples=1).squeeze(1)
return next_tokens
def isFinish(self, next_tokens):
pad_token_id = self.qwen.config.pad_token_id
eos_token_id_tensor = torch.tensor([self.qwen.config.eos_token_id]).to(next_tokens.device)
next_tokens = next_tokens * self.unfinished_sequences + pad_token_id * (1 - self.unfinished_sequences)
self.unfinished_sequences = self.unfinished_sequences.mul(
next_tokens.tile(eos_token_id_tensor.shape[0], 1).ne(eos_token_id_tensor.unsqueeze(1)).prod(dim=0)
)
return self.unfinished_sequences.max() == 0, next_tokens[:, None]