Check failed in jt.normal

[x0w3n]

Describe the bug

A Check failed is triggered when the shape of the generated random tensor exceeds the boundary. The error log suggests reporting this issue.

Full Log

[i 1203 16:14:49.013920 52 compiler.py:956] Jittor(1.3.9.10) src: /home/miniconda3/envs/jittor/lib/python3.9/site-packages/jittor
[i 1203 16:14:49.024061 52 compiler.py:957] g++ at /usr/bin/g++(11.2.0)
[i 1203 16:14:49.024530 52 compiler.py:958] cache_path: /home/.cache/jittor/jt1.3.9/g++11.2.0/py3.9.12/Linux-5.15.0-1x30/INTELRXEONRGOLx51/ef26/default
[i 1203 16:14:49.093253 52 install_cuda.py:93] cuda_driver_version: [12, 2]
[i 1203 16:14:49.108120 52 __init__.py:412] Found /home/.cache/jittor/jtcuda/cuda12.2_cudnn8_linux/bin/nvcc(12.2.140) at /home/.cache/jittor/jtcuda/cuda12.2_cudnn8_linux/bin/nvcc.
[i 1203 16:14:49.530114 52 __init__.py:412] Found gdb(12.0.90) at /usr/bin/gdb.
[i 1203 16:14:49.533059 52 __init__.py:412] Found addr2line(2.38) at /usr/bin/addr2line.
[i 1203 16:14:49.911349 52 compiler.py:1013] cuda key:cu12.2.140_sm_89
[i 1203 16:14:50.994216 52 __init__.py:227] Total mem: 251.50GB, using 16 procs for compiling.
[i 1203 16:14:51.170701 52 jit_compiler.cc:28] Load cc_path: /usr/bin/g++
[i 1203 16:14:51.387350 52 init.cc:63] Found cuda archs: [89,]
Traceback (most recent call last):
  File "test.py", line 5, in <module>
    normal_tensor = jt.normal(mean=0, std=1, size=large_tensor_size, dtype='float32')
  File "/home/miniconda3/envs/jittor/lib/python3.9/site-packages/jittor/__init__.py", line 1107, in normal
    return jt.init.gauss(size, dtype, mean, std)
  File "/home/miniconda3/envs/jittor/lib/python3.9/site-packages/jittor/init.py", line 287, in gauss
    return jt.random(NanoVector(shape), dtype, "normal") * std + mean
RuntimeError: Wrong inputs arguments, Please refer to examples(help(jt.NanoVector.__init__)).

Types of your inputs are:
 self   = NanoVector,
 args   = (tuple, ),

The function declarations are:
 inline NanoVector()
 inline NanoVector(const NanoVector& nv)
 inline NanoVector(const vector<int64>& v)
 inline NanoVector(int64 x)

Failed reason:[f 1203 16:14:52.028921 52 nano_vector.h:107] Check failed: s<10 && next_offset<=64  Something wrong... Could you please report this issue?

Minimal Reproduce

import jittor as jt

large_tensor_size = (10000,) * 10  
normal_tensor = jt.normal(mean=0, std=1, size=large_tensor_size, dtype='float32')

Expected behavior