Reproducibility problem in Pytorch-Huggingface during inference

Question

I've finetuned GPT2 model on a samall dataset and used it for generating sequence with random sampling and top_p enabled and random seed set to value of 123 using the below code:

import torch 
import transformers
model = transformers.GPT2LMHeadModel.from_pretrained("captions")
tokenizer = transformers.GPT2Tokenizer.from_pretrained("captions")
transformers.set_seed(123)

## calling huggingface generate method.
with torch.no_grad():
    for i in range(3):
        print(tokenizer.batch_decode(model.generate(enc,max_length=len(enc[0])+30,do_sample=True,top_p=1.0,temperature=1.0)))

I ran the above code on Kaggle, Colab, and Paperspace virtual servers while Kaggle and collab gave the same result but, I am unable to reproduce the result in Paperspace virtual server.

I suspect this is due to change in number of Cpu cores but I am not sure correct, me if I am wrong.

Cpu info (Paperspace):

!lscpu
Architecture:        x86_64
CPU op-mode(s):      32-bit, 64-bit
Byte Order:          Little Endian
CPU(s):              8
On-line CPU(s) list: 0-7
Thread(s) per core:  1
Core(s) per socket:  8
Socket(s):           1
NUMA node(s):        1
Vendor ID:           GenuineIntel
CPU family:          6
Model:               79
Model name:          Intel(R) Xeon(R) CPU E5-2623 v4 @ 2.60GHz
Stepping:            1
CPU MHz:             2600.086
BogoMIPS:            5200.00
Hypervisor vendor:   Xen
Virtualization type: full
L1d cache:           32K
L1i cache:           32K
L2 cache:            256K
L3 cache:            10240K
NUMA node0 CPU(s):   0-7
Flags:               fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush acpi mmx fxsr sse sse2 ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch cpuid_fault invpcid_single pti intel_ppin ssbd ibrs ibpb stibp fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt flush_l1d

CPU info (Kaggle and Colab):

Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   46 bits physical, 48 bits virtual
CPU(s):                          2
On-line CPU(s) list:             0,1
Thread(s) per core:              2
Core(s) per socket:              1
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       GenuineIntel
CPU family:                      6
Model:                           85
Model name:                      Intel(R) Xeon(R) CPU @ 2.00GHz
Stepping:                        3
CPU MHz:                         2000.178
BogoMIPS:                        4000.35
Hypervisor vendor:               KVM
Virtualization type:             full
L1d cache:                       32 KiB
L1i cache:                       32 KiB
L2 cache:                        1 MiB
L3 cache:                        38.5 MiB
NUMA node0 CPU(s):               0,1
Vulnerability Itlb multihit:     Not affected
Vulnerability L1tf:              Mitigation; PTE Inversion
Vulnerability Mds:               Mitigation; Clear CPU buffers; SMT Host state u
                                 nknown
Vulnerability Meltdown:          Mitigation; PTI
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled v
                                 ia prctl and seccomp
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user
                                  pointer sanitization
Vulnerability Spectre v2:        Mitigation; Full generic retpoline, IBPB condit
                                 ional, IBRS_FW, STIBP conditional, RSB filling
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Mitigation; Clear CPU buffers; SMT Host state u
                                 known

!lscpu
Architecture:        x86_64
CPU op-mode(s):      32-bit, 64-bit
Byte Order:          Little Endian
CPU(s):              2
On-line CPU(s) list: 0,1
Thread(s) per core:  2
Core(s) per socket:  1
Socket(s):           1
NUMA node(s):        1
Vendor ID:           AuthenticAMD
CPU family:          23
Model:               49
Model name:          AMD EPYC 7B12
Stepping:            0
CPU MHz:             2249.998
BogoMIPS:            4499.99
Hypervisor vendor:   KVM
Virtualization type: full
L1d cache:           32K
L1i cache:           32K
L2 cache:            512K
L3 cache:            16384K
NUMA node0 CPU(s):   0,1

Please help me out reproduce result on Paperspace instance.