Byte-lingua-code / fast_compare.py
2ira's picture
offline_compression_graph_code
72c0672 verified
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
import json
import base64
import argparse
import os
import sys
import gzip
import math
import torch
import torch.multiprocessing as mp
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import Levenshtein
from typing import List, Callable, Tuple, Optional
from concurrent.futures import ThreadPoolExecutor
# ==========================================
# 0. 环境设置
# ==========================================
os.environ.setdefault("TOKENIZERS_PARALLELISM", "false")
current_dir = os.getcwd()
if current_dir not in sys.path:
sys.path.append(current_dir)
script_dir = os.path.dirname(os.path.abspath(__file__))
if script_dir not in sys.path:
sys.path.append(script_dir)
# ==========================================
# 1. 依赖导入
# ==========================================
try:
from transformers import AutoTokenizer
except ImportError:
print("❌ Error: transformers not installed.")
sys.exit(1)
try:
from m1_compression import utils
from m1_compression.compressor import (
load_m1_model_and_tokenizer,
ALPHABET_SIZE,
ARITHMETIC_CODER_BASE,
ARITHMETIC_CODER_PRECISION,
)
from m1_compression.hybrid_arithmetic_coder import CPUArithmeticEncoder
from m1_compression.batched_arithmetic_coder import _pdf_to_cdf
except ImportError as e:
print(f"❌ Error: m1_compression not found. {e}")
sys.exit(1)
# ==========================================
# 2. 基础工具函数
# ==========================================
def vread(buf: bytes, i: int):
shift = val = 0
while True:
b = buf[i]
i += 1
val |= (b & 0x7F) << shift
if b < 0x80:
return val, i
shift += 7
def unpack_windows(input_bytes: bytes, b64_stream: str) -> List[bytes]:
"""
只返回需要压缩的 windows bytes 段(忽略 gap 非压缩区域)。
"""
try:
if not b64_stream:
return []
buf = base64.b64decode(b64_stream)
i = 0
cursor = 0
segments: List[bytes] = []
while i < len(buf):
gap, i = vread(buf, i)
size, i = vread(buf, i)
start = cursor + gap
end = start + size
if end > len(input_bytes):
break
segments.append(input_bytes[start:end])
cursor = end
return segments
except Exception:
return []
def token_ids_to_str(ids: List[int]) -> str:
# token id 可能 >255,因此仍然需要 chr 映射
# 为安全起见 clamp 到 unicode 最大值
return "".join(chr(x if x <= 0x10FFFF else 0x10FFFF) for x in ids)
def bytes_to_latin1_str(b: bytes) -> str:
# 0~255 一一映射到 unicode,编辑距离结果等价于你之前的 chr(x)
return b.decode("latin1")
def pad_batch_fast(batch: List[bytes]) -> Tuple[torch.Tensor, torch.Tensor]:
"""
将 List[bytes] -> (padded_batch[int64], lengths[int64])
关键优化:numpy.frombuffer + 一次性拷贝,避免 Python list(data)
"""
if not batch:
return torch.empty((0, 0), dtype=torch.long), torch.empty((0,), dtype=torch.long)
lengths_np = np.fromiter((len(x) for x in batch), dtype=np.int32, count=len(batch))
max_len = int(lengths_np.max())
arr = np.zeros((len(batch), max_len), dtype=np.uint8)
for i, seg in enumerate(batch):
seg_np = np.frombuffer(seg, dtype=np.uint8)
if seg_np.size:
arr[i, : seg_np.size] = seg_np
padded = torch.from_numpy(arr).to(torch.long) # gather 需要 int64
lengths = torch.from_numpy(lengths_np.astype(np.int64))
return padded, lengths
def iter_jsonl_shard_bytes(file_path: str, shard_rank: int, shard_world: int):
"""
按“字节范围”切分 jsonl 文件:每个 shard 只读自己负责的文件区间。
适合“单文件吃满多 GPU”。
"""
file_size = os.path.getsize(file_path)
start = (file_size * shard_rank) // shard_world
end = (file_size * (shard_rank + 1)) // shard_world
with open(file_path, "rb") as f:
f.seek(start)
if start > 0:
f.readline() # 丢掉半行,跳到下一行首
while f.tell() < end:
line = f.readline()
if not line:
break
yield line
# ==========================================
# 3. 高性能 AC 压缩核心 (Smart Batching + bytes output)
# ==========================================
def batched_m1_compress_predict_fn(model):
def predict_fn(input_tensor: torch.Tensor, **kwargs) -> torch.Tensor:
if input_tensor.dim() == 1:
input_tensor = input_tensor.unsqueeze(0)
with torch.inference_mode():
logits = model(input_tensor, **kwargs)
logits = logits[..., :256].float()
probs = torch.softmax(logits, dim=-1)
return probs
return predict_fn
def compress_segments_smart_batch_bytes(
all_segments: List[bytes],
batched_predict_fn: Callable,
first_byte_prob: torch.Tensor,
device: torch.device,
encoder: CPUArithmeticEncoder,
gpu_batch_size: int = 256,
bit_threshold: int = 64,
) -> List[bytes]:
"""
高性能 AC 压缩:
1) 先按长度排序,降低 padding 浪费
2) 推理在 GPU,编码在 CPU
3) 输出每个 segment 的压缩 bytes(不转 List[int])
"""
M = len(all_segments)
if M == 0:
return []
lengths = np.fromiter((len(s) for s in all_segments), dtype=np.int32, count=M)
sorted_indices = np.argsort(lengths, kind="stable")
sorted_segments = [all_segments[i] for i in sorted_indices]
out: List[Optional[bytes]] = [None] * M
for i in range(0, M, gpu_batch_size):
batch_slice = sorted_segments[i : i + gpu_batch_size]
batch_orig_indices = sorted_indices[i : i + gpu_batch_size]
try:
padded_batch_cpu, lengths_cpu = pad_batch_fast(batch_slice)
# pin + non_blocking H2D
padded_batch = padded_batch_cpu.pin_memory().to(device, non_blocking=True)
# --- GPU 推理 ---
prompt_probs = batched_predict_fn(padded_batch)
final_probs = torch.cat(
[
first_byte_prob.expand(prompt_probs.shape[0], -1, -1),
prompt_probs[:, :-1, ...],
],
dim=1,
)
# --- CDF 计算 ---
final_probs = utils.batched_normalize_pdf_for_arithmetic_coding(final_probs)
cdfs_gpu = _pdf_to_cdf(final_probs)
cdf_low = cdfs_gpu.gather(2, padded_batch.unsqueeze(-1)).squeeze(-1)
cdf_high = cdfs_gpu.gather(2, (padded_batch + 1).unsqueeze(-1)).squeeze(-1)
cdf_ends = torch.stack([cdf_low, cdf_high], dim=-1)
# --- CPU 编码 ---
enc_out = encoder.incremental_batched_encode(
cdf_ends.cpu(),
ALPHABET_SIZE,
lengths_cpu, # CPU
bit_threshold=bit_threshold,
force_padding_to_threshold=False,
return_num_padded_bits=False,
)
if isinstance(enc_out, tuple):
chunked_compressed_bytes = enc_out[0]
else:
chunked_compressed_bytes = enc_out
for idx, code in zip(batch_orig_indices, chunked_compressed_bytes):
out[int(idx)] = bytes(code)
except Exception:
# 容错:降级为原始 bytes
for idx, seg in zip(batch_orig_indices, batch_slice):
out[int(idx)] = seg
return [x if x is not None else b"" for x in out]
class M1ACManager:
def __init__(self, model_path: str, first_prob_path: str, device_id: int,
gpu_batch_size: int = 256, bit_threshold: int = 64):
self.device = torch.device(f"cuda:{device_id}")
self.gpu_batch_size = gpu_batch_size
self.bit_threshold = bit_threshold
self.model, _, _ = load_m1_model_and_tokenizer(model_path)
self.model.to(self.device)
self.model.eval()
self.predict_fn = batched_m1_compress_predict_fn(self.model)
if first_prob_path and os.path.exists(first_prob_path):
with open(first_prob_path, "r") as f:
prob_data = json.load(f)
self.first_byte_prob = torch.tensor(prob_data, dtype=torch.float32, device=self.device)
if self.first_byte_prob.dim() == 1:
self.first_byte_prob = self.first_byte_prob.unsqueeze(0).unsqueeze(0)
else:
self.first_byte_prob = torch.ones((1, 1, ALPHABET_SIZE), dtype=torch.float32, device=self.device) / ALPHABET_SIZE
# 全局 encoder(避免重复创建)
self.encoder = CPUArithmeticEncoder(base=ARITHMETIC_CODER_BASE, precision=ARITHMETIC_CODER_PRECISION)
def compress_batch_smart_bytes(self, inputs: List[Tuple[str, Optional[str]]]) -> List[bytes]:
"""
inputs: List[(text, windows_b64_or_None)]
Return: List[bytes] 每个 sample 对应拼接后的 AC bitstream(bytes)
"""
all_segments_flat: List[bytes] = []
sample_map: List[Tuple[int, int]] = []
current_idx = 0
for text, windows_b64 in inputs:
raw_bytes = text.encode("utf-8")
sample_segs: List[bytes] = []
if windows_b64:
sample_segs = unpack_windows(raw_bytes, windows_b64)
else:
CHUNK = 512
for j in range(0, len(raw_bytes), CHUNK):
sample_segs.append(raw_bytes[j : j + CHUNK])
count = len(sample_segs)
sample_map.append((current_idx, current_idx + count))
all_segments_flat.extend(sample_segs)
current_idx += count
if not all_segments_flat:
return [b"" for _ in inputs]
compressed_chunks_flat = compress_segments_smart_batch_bytes(
all_segments_flat,
self.predict_fn,
self.first_byte_prob,
self.device,
self.encoder,
gpu_batch_size=self.gpu_batch_size,
bit_threshold=self.bit_threshold,
)
results: List[bytes] = []
for start, end in sample_map:
results.append(b"".join(compressed_chunks_flat[start:end]))
return results
# ==========================================
# 4. Worker Process
# ==========================================
def run_gzip_task(text_pair: Tuple[str, str]) -> float:
t1, t2 = text_pair
b1 = t1.encode("utf-8")
b2 = t2.encode("utf-8")
g1 = gzip.compress(b1)
g2 = gzip.compress(b2)
if not g1:
return 0.0
d = Levenshtein.distance(bytes_to_latin1_str(g1), bytes_to_latin1_str(g2))
return d / len(g1)
def process_one_file(
gpu_id: int,
file_path: str,
tokenizer: AutoTokenizer,
ac_manager: M1ACManager,
max_lines: int,
worker_batch_size: int,
gzip_threads: int,
shard_rank: int,
shard_world: int,
) -> dict:
"""
处理单个 jsonl 文件,返回 results dict
"""
results = {"Gzip": [], "Tokenizer": [], "AC_M1": []}
# 为了让“单文件多 shard”时总样本量≈max_lines,给每个 shard 分摊行数上限
if shard_world > 1 and max_lines > 0:
shard_max_lines = int(math.ceil(max_lines / shard_world))
else:
shard_max_lines = max_lines
raw_texts: List[str] = []
pert_texts: List[str] = []
metas: List[Optional[str]] = []
processed_total = 0
# 线程池用于 gzip(gzip.compress 是 C 实现,通常能释放 GIL,线程能并行)
thread_pool = ThreadPoolExecutor(max_workers=gzip_threads)
def flush():
nonlocal raw_texts, pert_texts, metas
if not raw_texts:
return
curr_batch_size = len(raw_texts)
# 1) Gzip 并行
gz_vals = list(thread_pool.map(run_gzip_task, zip(raw_texts, pert_texts)))
results["Gzip"].extend(gz_vals)
# 2) Tokenizer batched(fast tokenizer)
try:
tok1 = tokenizer(raw_texts, add_special_tokens=False)["input_ids"]
tok2 = tokenizer(pert_texts, add_special_tokens=False)["input_ids"]
for a, b in zip(tok1, tok2):
if a:
d = Levenshtein.distance(token_ids_to_str(a), token_ids_to_str(b))
results["Tokenizer"].append(d / len(a))
except Exception:
# tokenizer 出问题时不影响 AC 结果
pass
# 3) AC:orig + pert 合并一次调用
orig_inputs = list(zip(raw_texts, metas))
pert_inputs = list(zip(pert_texts, [None] * curr_batch_size))
both_inputs = orig_inputs + pert_inputs
try:
both_streams = ac_manager.compress_batch_smart_bytes(both_inputs)
ac1_list = both_streams[:curr_batch_size]
ac2_list = both_streams[curr_batch_size:]
for a1, a2 in zip(ac1_list, ac2_list):
if a1:
d = Levenshtein.distance(bytes_to_latin1_str(a1), bytes_to_latin1_str(a2))
results["AC_M1"].append(d / len(a1))
except Exception as e:
print(f"[GPU {gpu_id}] AC Batch Error: {e}")
raw_texts, pert_texts, metas = [], [], []
# 文件读取:支持 shard
line_iter = iter_jsonl_shard_bytes(file_path, shard_rank, shard_world)
for i, line in enumerate(line_iter):
if shard_max_lines > 0 and i >= shard_max_lines:
break
try:
if not line or len(line) < 100:
continue
data = json.loads(line)
text = data.get("text", "")
if not isinstance(text, str) or len(text) < 50:
continue
windows = data.get("windows_starts_lens_b64")
cut_idx = max(1, int(len(text) * 0.2))
# cut_idx = 2
text_p = text[cut_idx:]
raw_texts.append(text)
pert_texts.append(text_p)
metas.append(windows)
processed_total += 1
if len(raw_texts) >= worker_batch_size:
flush()
if processed_total % 2000 == 0:
print(f"[GPU {gpu_id}] processed {processed_total} lines (file={os.path.basename(file_path)}, shard={shard_rank}/{shard_world})")
except Exception:
continue
flush()
thread_pool.shutdown(wait=True)
print(f"[GPU {gpu_id}] done file={os.path.basename(file_path)} shard={shard_rank}/{shard_world} total={processed_total}")
return results
def process_files_worker(
rank: int,
gpu_id: int,
file_paths: List[str],
output_dir: str,
model_path: str,
prob_path: str,
max_lines: int,
worker_batch_size: int,
gzip_threads: int,
shard_mode: bool,
gpu_batch_size: int,
bit_threshold: int,
):
"""
一个 GPU 进程:加载一次 tokenizer + M1 模型,然后顺序处理分配给它的文件(或单文件 shard)
"""
try:
torch.cuda.set_device(gpu_id)
# tokenizer
try:
tokenizer = AutoTokenizer.from_pretrained(
"infly/OpenCoder-1.5B-Base",
trust_remote_code=True,
use_fast=True,
)
if tokenizer.pad_token is None:
tokenizer.pad_token = tokenizer.eos_token
except Exception:
tokenizer = AutoTokenizer.from_pretrained("gpt2", use_fast=True)
# AC manager
ac_manager = M1ACManager(
model_path=model_path,
first_prob_path=prob_path,
device_id=gpu_id,
gpu_batch_size=gpu_batch_size,
bit_threshold=bit_threshold,
)
for fp in file_paths:
base = os.path.basename(fp)
# 单文件 shard:rank 对应 shard_rank
if shard_mode:
shard_rank = rank
shard_world = torch.cuda.device_count()
else:
shard_rank = 0
shard_world = 1
print(f"[GPU {gpu_id}] start file={base} shard={shard_rank}/{shard_world}")
res = process_one_file(
gpu_id=gpu_id,
file_path=fp,
tokenizer=tokenizer,
ac_manager=ac_manager,
max_lines=max_lines,
worker_batch_size=worker_batch_size,
gzip_threads=gzip_threads,
shard_rank=shard_rank,
shard_world=shard_world,
)
out_name = f"res_gpu{gpu_id}_rank{rank}_shard{shard_rank}of{shard_world}_{base}.json"
out_path = os.path.join(output_dir, out_name)
with open(out_path, "w") as f:
json.dump(res, f)
except Exception as e:
print(f"❌ [GPU {gpu_id}] Worker Error: {e}")
import traceback
traceback.print_exc()
# ==========================================
# 5. 主程序
# ==========================================
def main():
parser = argparse.ArgumentParser()
parser.add_argument("--input_dir", type=str, required=True)
parser.add_argument("--m1_model", type=str, required=True)
parser.add_argument("--first_prob_path", type=str, required=True)
parser.add_argument("-o", "--output_dir", type=str, default="analysis_output_fast_opt")
parser.add_argument("--max_lines", type=int, default=10000)
# 可选调参(不给也能跑)
parser.add_argument("--max_files", type=int, default=8, help="只取前 N 个 jsonl 文件;0 表示不限制")
parser.add_argument("--worker_batch_size", type=int, default=500, help="flush 的行数 batch")
parser.add_argument("--gzip_threads", type=int, default=8, help="每个 GPU 进程内用于 gzip 的线程数")
parser.add_argument("--ac_gpu_batch_size", type=int, default=256, help="AC 推理的 GPU mini-batch size")
parser.add_argument("--ac_bit_threshold", type=int, default=64, help="Arithmetic coder bit_threshold(16->64/128 往往更快)")
args = parser.parse_args()
os.makedirs(args.output_dir, exist_ok=True)
files = [
os.path.join(args.input_dir, f)
for f in os.listdir(args.input_dir)
if f.endswith(".jsonl") and "writer" not in f
]
files.sort()
if args.max_files and args.max_files > 0:
files = files[: args.max_files]
num_gpus = torch.cuda.device_count()
if num_gpus == 0:
print("❌ No GPU detected.")
return
if not files:
print("❌ No jsonl files found.")
return
# 任务分配:
# - 如果只有 1 个文件且 GPU>1:开启 shard_mode,让每张卡读同一个文件的不同字节段
# - 否则:按文件 idx % num_gpus 分配给各 GPU 进程(每个进程处理多个文件)
shard_mode = (len(files) == 1 and num_gpus > 1)
assignments: List[List[str]] = [[] for _ in range(num_gpus)]
if shard_mode:
for r in range(num_gpus):
assignments[r] = [files[0]]
print(f"🚀 Single-file shard mode enabled: {files[0]} -> {num_gpus} shards")
else:
for idx, fp in enumerate(files):
assignments[idx % num_gpus].append(fp)
non_empty = sum(1 for a in assignments if a)
print(f"🚀 Multi-file mode: {len(files)} files assigned across {non_empty}/{num_gpus} GPU workers")
print(f" worker_batch_size={args.worker_batch_size}, gzip_threads={args.gzip_threads}, ac_gpu_batch_size={args.ac_gpu_batch_size}, ac_bit_threshold={args.ac_bit_threshold}")
mp.set_start_method("spawn", force=True)
procs = []
for rank in range(num_gpus):
if not assignments[rank]:
continue
gpu_id = rank % num_gpus
p = mp.Process(
target=process_files_worker,
args=(
rank,
gpu_id,
assignments[rank],
args.output_dir,
args.m1_model,
args.first_prob_path,
args.max_lines,
args.worker_batch_size,
args.gzip_threads,
shard_mode,
args.ac_gpu_batch_size,
args.ac_bit_threshold,
),
)
p.start()
procs.append(p)
for p in procs:
p.join()
# 合并结果
print("✅ Merging results...")
final_results = {"Gzip": [], "Tokenizer": [], "AC_M1": []}
for fn in os.listdir(args.output_dir):
if fn.startswith("res_") and fn.endswith(".json"):
try:
with open(os.path.join(args.output_dir, fn), "r") as f:
d = json.load(f)
for k in final_results:
final_results[k].extend(d.get(k, []))
except Exception:
pass
for k, v in final_results.items():
print(f" {k}: {len(v)} samples")
# 统计 + 保存
stats = {}
for k, v in final_results.items():
if v:
stats[k] = {"count": int(len(v)), "mean": float(np.mean(v)), "p50": float(np.median(v))}
with open(os.path.join(args.output_dir, "final_stats.json"), "w") as f:
json.dump(stats, f, indent=2)
print(f"📄 Saved stats -> {os.path.join(args.output_dir, 'final_stats.json')}")
# 绘图
plot_data = []
for algo, vals in final_results.items():
for val in vals:
if val < 2.0:
plot_data.append({"Algorithm": algo, "NED": val})
if plot_data:
df = pd.DataFrame(plot_data)
plt.figure(figsize=(10, 6))
sns.kdeplot(data=df, x="NED", hue="Algorithm", fill=True, common_norm=False)
plt.title("Compression Stability (Optimized)")
plt.xlabel("Normalized Levenshtein Distance")
plt.xlim(0, 1.2)
out_img = os.path.join(args.output_dir, "stability_fast_opt.png")
plt.savefig(out_img, dpi=200)
print(f"📊 Saved plot -> {out_img}")
print("🎉 Done.")
if __name__ == "__main__":
main()
"""
python fast_compare.py \
--input_dir /mnt/hdfs/user/linzheng/data/ocpython_subsampled_50G_entropy90_splits_chunk512_ow20_iterative-true_forcepadding-true_merged_ac \
--m1_model /mnt/bn/tiktok-mm-5/aiic/users/linzheng/artifacts/m1_checkpoints/m1_40M_lr1e-3_steps200k_bs8_seqlen2048_python/checkpoints/0000200000 \
--first_prob_path /mnt/bn/tiktok-mm-5/aiic/users/linzheng/artifacts/ac_unigram_probs/python500k_unigram_prob.json \
--max_lines 10000 \
-o analysis_output_fast_opt \
--max_files 8
"""