forked from avnermay/Sequoia
-
Notifications
You must be signed in to change notification settings - Fork 0
/
run_sequoia.py
330 lines (295 loc) · 17 KB
/
run_sequoia.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
import argparse
import numpy as np
import os
import random
import sys
import time
from tqdm import tqdm
# Torch imports
import torch
from torch.utils.data.dataloader import DataLoader
import torch.nn.functional as F
# HF Imports
from accelerate import Accelerator
from datasets import load_from_disk
from transformers import DataCollatorForLanguageModeling, AutoTokenizer
# Sequoia imports
from data_converter import convert_wiki_dataset, convert_cnn_dataset, convert_jsonl_file, convert_dataset
from Tree.SpecTree import SpecTree
from utils import get_sampling_logits, _make_causal_mask, cuda_graph_for_residual, cuda_graph_for_sampling_without_replacement
from Engine.Engine import GraphInferenceEngine, GraphInferenceEngineTG
from Engine.offload_engine import OffloadEngine
def get_args():
parser = argparse.ArgumentParser()
parser.add_argument('--draft', type=str, help='draft model')
parser.add_argument('--target', type=str, help='target model')
parser.add_argument('--dataset', type=str, default='../dataset/c4_small.json', help='dataset path')
parser.add_argument('--growmap', type=str, default='../growmaps/68m_7b-64.pt', help='growmap path')
parser.add_argument('--start', type=int, default=0, help='start')
parser.add_argument('--end', type=int, default=200, help='end')
parser.add_argument('--temp', type=float, default=0.6, help='temperature')
parser.add_argument('--top_p', type=float, default=0.9, help='top_p')
parser.add_argument('--max_length', type=int, default=256, help='max length')
parser.add_argument('--seed', type=int, default=17, help='random seed')
parser.add_argument('--mode', type=str, default='fast', help='tree mode')
parser.add_argument('--offloading', action='store_true')
args = parser.parse_args()
print(args)
return args
def setup_seed(seed):
torch.manual_seed(seed)
torch.cuda.manual_seed_all(seed)
np.random.seed(seed)
random.seed(seed)
torch.backends.cudnn.deterministic = True
def simulation_baseline(target_model: GraphInferenceEngineTG, dataloader: DataLoader, temp=0.6, top_p=0.9, max_length=256):
num_eval_steps = len(dataloader)
num_decoding_steps = 0
total_time = 0.0
with torch.no_grad():
for step, batch in tqdm(enumerate(dataloader), total=num_eval_steps):
input_ids = batch['input_ids'][..., :128]
labels = batch['labels'][..., :128]
terminate = False
if labels[0][-1] == -100: terminate = True
position_ids = torch.arange(max_length).to('cuda:0').unsqueeze(0)
storage_ids = torch.arange(max_length).to('cuda:0')
attn_mask = _make_causal_mask((max_length, max_length), target_model.dtype, target_model.device)
torch.cuda.synchronize()
t1 = time.time()
inner_decoding_step = 0
start_length = 0
while inner_decoding_step < 32 and terminate == False:
if inner_decoding_step == 0:
start_length = input_ids.shape[1]
logits = target_model.inference(input_ids = input_ids, storage_ids=storage_ids[:start_length],
position_ids = position_ids[..., :start_length],
attn_mask=attn_mask[:start_length, :start_length][None, None, :, :])[0][-1]
else:
logits = target_model.inference(input_ids = input_ids, storage_ids=storage_ids[start_length + inner_decoding_step-1 : start_length + inner_decoding_step],
position_ids = position_ids[..., start_length + inner_decoding_step-1 : start_length + inner_decoding_step],
attn_mask=attn_mask[start_length + inner_decoding_step-1 : start_length + inner_decoding_step, :start_length + inner_decoding_step][None, None, :, :])[0][-1]
logits = get_sampling_logits(logits=logits, top_p=top_p, T=temp)
p = F.softmax(logits / temp, dim=-1)
new_token = p.multinomial(num_samples=1).unsqueeze(0)
input_ids = new_token
num_decoding_steps += 1
inner_decoding_step += 1
if input_ids[0][-1] == 2:
terminate = True
torch.cuda.synchronize()
t2 = time.time()
total_time += (t2 - t1)
target_model.clear_kv()
print('total time :{:.5f}s, latency :{:.5f}s, decoding step: {}'.format(total_time, total_time / num_decoding_steps, num_decoding_steps))
return num_decoding_steps
def simulation_benchmark(
target_model: GraphInferenceEngineTG, draft_model: GraphInferenceEngine, dataloader: DataLoader,
temp=0.6, top_p=0.9, max_length=512, residual_graph=None, grow_map=None,
sampling_callables = None, sample_gather_indices = None,
):
num_eval_steps = len(dataloader)
num_decoding_steps = 0
num_large_model_steps = 0
initialize_time = 0.0
speculate_time = 0.0
verify_time = 0.0
large_model_run = 0.0
accept_loop = 0.0
kv_select = 0.0
sample_time = 0.0
small_model_compute = 0.0
dtype = torch.float16
attn_mask = torch.full((max_length, max_length), torch.finfo(dtype).min, dtype=dtype, device='cuda:0')
sequence = torch.tensor(list(range(max_length)), device='cuda:0').long().unsqueeze(-1)
new_tokens_buffer = torch.zeros(max_length).long().to('cuda:0')
parents_buffer = torch.zeros(max_length).long().to('cuda:0')
position_ids = torch.zeros(max_length).long().to('cuda:0')
with torch.no_grad():
for step, batch in tqdm(enumerate(dataloader), total=num_eval_steps):
input_ids = batch['input_ids'][..., :128]
labels = batch['labels'][..., :128]
terminate = False
if labels[0][-1] == -100: terminate = True
draft_kv_len = 0
target_kv_len = 0
attn_mask.fill_(torch.finfo(dtype).min)
spectree = SpecTree(prefix=input_ids.squeeze(0), device='cuda:0', temperature=temp,
top_p=top_p,
draft_kv_len=draft_kv_len, target_kv_len=target_kv_len,
draft_model_engine=draft_model, target_model_engine=target_model, max_length=max_length, grow_map=grow_map,
attn_mask = attn_mask, sequence = sequence, new_tokens_buffer = new_tokens_buffer,
parents_buffer = parents_buffer,
position_ids = position_ids,
residual_graph = residual_graph,
sampling_callables=sampling_callables,
sample_gather_indices = sample_gather_indices)
while input_ids.shape[1] < 256 and terminate == False:
torch.cuda.synchronize()
t1 = time.time()
torch.cuda.synchronize()
t2 = time.time()
a, b = spectree.construct_grow_map(benchmark=True)
torch.cuda.synchronize()
t3 = time.time()
valid_tokens, draft_kv_len, target_kv_len, x, y, z, terminate = spectree.verify(benchmark=True)
torch.cuda.synchronize()
t4 = time.time()
initial_size = input_ids.shape[1]
input_ids = valid_tokens.unsqueeze(0)
if (input_ids[0] == 2)._is_any_true() or (input_ids[0] == 0)._is_any_true() or input_ids.shape[1] >= 256:
terminate = True
if not terminate:
sample_time += a
small_model_compute += b
large_model_run += x
accept_loop += y
kv_select += z
initialize_time += (t2 - t1)
speculate_time += (t3 - t2)
verify_time += (t4 - t3)
num_decoding_steps += (valid_tokens.shape[0] - initial_size)
num_large_model_steps += 1
draft_model.clear_kv()
target_model.clear_kv()
if num_large_model_steps > 0:
print(num_decoding_steps / num_large_model_steps)
print('total decoding steps: {}'.format(num_decoding_steps), 'large model steps: {}'.format(num_large_model_steps), 'avg decoding step: {}'.format(num_decoding_steps / num_large_model_steps))
print('initialization time:{}'.format(initialize_time / num_large_model_steps), 'speculate time: {}'.format(speculate_time / num_large_model_steps), 'verify time: {}'.format(verify_time / num_large_model_steps))
print('large model run: {}'.format(large_model_run / num_large_model_steps) , 'accept loop: {}'.format(accept_loop / num_large_model_steps), 'kv select: {}'.format(kv_select / num_large_model_steps))
print('small model run: {}'.format(small_model_compute / num_large_model_steps) , 'sample time: {}'.format(sample_time / num_large_model_steps))
return num_decoding_steps / num_large_model_steps
def simulation_fast(
target_model: GraphInferenceEngineTG, draft_model: GraphInferenceEngine, dataloader: DataLoader,
temp=0.6, top_p=0.9, max_length=512, residual_graph=None, grow_map=None,
sampling_callables = None, sample_gather_indices = None,
):
num_eval_steps = len(dataloader)
num_decoding_steps = 0
num_large_model_steps = 0
total_time = 0.0
dtype = torch.float16
attn_mask = torch.full((max_length, max_length), torch.finfo(dtype).min, dtype=dtype, device='cuda:0')
sequence = torch.tensor(list(range(max_length)), device='cuda:0').long().unsqueeze(-1)
new_tokens_buffer = torch.zeros(max_length).long().to('cuda:0')
parents_buffer = torch.zeros(max_length).long().to('cuda:0')
position_ids = torch.zeros(max_length).long().to('cuda:0')
with torch.no_grad():
for step, batch in tqdm(enumerate(dataloader), total=num_eval_steps):
input_ids = batch['input_ids'][..., :128]
labels = batch['labels'][..., :128]
terminate = False
if labels[0][-1] == -100: terminate = True
draft_kv_len = 0
target_kv_len = 0
attn_mask.fill_(torch.finfo(dtype).min)
spectree = SpecTree(prefix=input_ids.squeeze(0), device='cuda:0', temperature=temp,
top_p=top_p,
draft_kv_len=draft_kv_len, target_kv_len=target_kv_len,
draft_model_engine=draft_model, target_model_engine=target_model, max_length=max_length, grow_map=grow_map,
attn_mask = attn_mask, sequence = sequence, new_tokens_buffer = new_tokens_buffer,
parents_buffer = parents_buffer,
position_ids = position_ids,
residual_graph = residual_graph,
sampling_callables=sampling_callables,
sample_gather_indices = sample_gather_indices)
torch.cuda.synchronize()
t1 = time.time()
while input_ids.shape[1] < 256 and terminate == False:
spectree.construct_grow_map()
valid_tokens, draft_kv_len, target_kv_len, terminate = spectree.verify()
num_decoding_steps += (valid_tokens.shape[0] - input_ids.shape[1])
num_large_model_steps += 1
input_ids = valid_tokens.unsqueeze(0)
if (input_ids[0][-1] == 2) or (input_ids[0][-1] == 0): terminate = True
torch.cuda.synchronize()
t2 = time.time()
total_time += (t2 - t1)
draft_model.clear_kv()
target_model.clear_kv()
print('total time :{:.5f}s, latency :{:.5f}s, decoding step: {}, large model step: {}, {}'.format(total_time, total_time / num_decoding_steps, num_decoding_steps, num_large_model_steps, num_decoding_steps / num_large_model_steps))
return num_decoding_steps / num_large_model_steps
def get_tokenized_dataloader(dataset, start, end):
tokenizer = AutoTokenizer.from_pretrained('meta-llama/Llama-2-7b-hf', use_fast=False)
tokenizer.pad_token = tokenizer.eos_token
current_dir = os.path.dirname(os.path.abspath(__file__))
if dataset == 'openwebtext':
tokenized_dataset_eval = load_from_disk(f'{current_dir}/dataset/openwebtext_eval')
elif dataset == 'wiki':
tokenized_dataset_eval = convert_wiki_dataset(tokenizer=tokenizer)
elif dataset == 'cnn':
tokenized_dataset_eval = convert_cnn_dataset(tokenizer=tokenizer)
elif dataset == 'wikitext':
tokenized_dataset_eval = convert_jsonl_file(tokenizer, f'{current_dir}/dataset/wikitext_dev.jsonl')
elif dataset == 'oasst':
tokenized_dataset_eval = convert_jsonl_file(tokenizer, f'{current_dir}/dataset/oasst_dev.jsonl')
elif dataset == 'c4_small':
tokenized_dataset_eval = convert_dataset(tokenizer=tokenizer, file_path=f'{current_dir}/dataset/c4_small.json')
else:
raise ValueError(f'Unsupported dataset: {dataset}.')
tokenized_dataset_eval = tokenized_dataset_eval.select(list(range(start, end)))
data_collator = DataCollatorForLanguageModeling(tokenizer, mlm=False)
dataloader = DataLoader(tokenized_dataset_eval, batch_size=1, collate_fn=data_collator, shuffle=False)
return dataloader
if __name__ == '__main__':
args = get_args()
setup_seed(args.seed)
# Note: Before used to grab random `end-start` elements from dataset, now grab the first ones.
dataloader = get_tokenized_dataloader(args.dataset, args.start, args.end)
if args.mode == 'baseline':
if args.offloading:
target_model = OffloadEngine(max_length=args.max_length, model_name_or_path=args.target, dtype=torch.float16, device='cuda:0')
else:
target_model = GraphInferenceEngineTG(max_length=args.max_length, model_name_or_path=args.target, dtype=torch.float16, device='cuda:0')
else:
draft_model = GraphInferenceEngine(max_length=args.max_length, model_name_or_path=args.draft, dtype=torch.float16, device='cuda:0')
if args.offloading:
target_model = OffloadEngine(max_length=args.max_length, model_name_or_path=args.target, dtype=torch.float16, device='cuda:0')
else:
target_model = GraphInferenceEngineTG(max_length=args.max_length, model_name_or_path=args.target, dtype=torch.float16, device='cuda:0')
residual_graph = cuda_graph_for_residual()
path = args.growmap
grow_map = torch.load(path)
tree_size = grow_map['size']
print(f'{tree_size=}')
idx_lists = grow_map['roots']
branch_lists = grow_map['branches']
draft_step = len(grow_map['roots'])
graph_capture_list = [sum(x) for x in branch_lists]
graph_capture_list.append(1)
draft_model.initialize_cuda_graph(graph_capture_list)
sampling_callables = {}
sample_gather_indices = {}
for i in range(draft_step - 1):
idx_len = len(idx_lists[i])
num_samples = max(branch_lists[i])
sampling_callables[i] = cuda_graph_for_sampling_without_replacement(
max_length=args.max_length, idx_len=idx_len, num_samples=num_samples,
temperature=args.temp, tree_size=tree_size)
for i in range(draft_step - 1):
ith_gather_list = []
max_num_samples = max(branch_lists[i])
for j, branch in enumerate(branch_lists[i]):
branch_index = torch.arange(branch, device='cuda:0', dtype=torch.long)
branch_index = branch_index + j * max_num_samples
ith_gather_list.append(branch_index)
ith_gather_list = torch.cat(ith_gather_list)
sample_gather_indices[i] = ith_gather_list
accelerator = Accelerator()
dataloader = accelerator.prepare(dataloader)
if args.mode == 'baseline':
simulation_baseline(
target_model=target_model, dataloader=dataloader, temp=args.temp, top_p=args.top_p, max_length=args.max_length,
)
elif args.mode == 'benchmark':
simulation_benchmark(
target_model=target_model, draft_model=draft_model, dataloader=dataloader,
temp=args.temp, top_p=args.top_p, max_length=args.max_length, residual_graph=residual_graph, grow_map=grow_map,
sampling_callables=sampling_callables, sample_gather_indices=sample_gather_indices,
)
elif args.mode == 'fast':
simulation_fast(
target_model=target_model, draft_model=draft_model, dataloader=dataloader,
temp=args.temp, top_p=args.top_p, max_length=args.max_length, residual_graph=residual_graph, grow_map=grow_map,
sampling_callables=sampling_callables, sample_gather_indices=sample_gather_indices,
)