remove superflous dependencies from Horovod easyconfig on top of PyTorch

[Flamefire]

(created using eb --new-pr)

Preparation to bring better consistency for upcoming ECs.

• Horovod for PyTorch uses the bundled flatbuffers and ignores the system (EB supplied) one
• UCX-CUDA is provided by NCCL
• parallel = 1 workaround not required for Horovod 0.24+ after Update to CMake 3.13 for better CUDA support and to enable build concurrency horovod/horovod#3261

[Flamefire]

Test report by @Flamefire
SUCCESS
Build succeeded for 5 out of 5 (5 easyconfigs in total)
taurusi8015 - Linux CentOS Linux 7.9.2009, x86_64, AMD EPYC 7352 24-Core Processor (zen2), Python 2.7.5
See https://gist.github.com/Flamefire/a6165ec86a2fae7e4c18aadebbd5d23c for a full test report.

[Flamefire]

Test report by @Flamefire
SUCCESS
Build succeeded for 5 out of 5 (5 easyconfigs in total)
taurusml4 - Linux RHEL 7.6, POWER, 8335-GTX (power9le), 6 x NVIDIA Tesla V100-SXM2-32GB, 440.64.00, Python 2.7.5
See https://gist.github.com/Flamefire/a084e93582f519c68c2dba257279c987 for a full test report.

[branfosj]

Test report by @branfosj
SUCCESS
Build succeeded for 5 out of 5 (5 easyconfigs in total)
bear-pg0203u29a.bear.cluster - Linux RHEL 8.6, x86_64, Intel(R) Xeon(R) Platinum 8360Y CPU @ 2.40GHz (icelake), 1 x NVIDIA NVIDIA A100-SXM4-80GB, 520.61.05, Python 3.6.8
See https://gist.github.com/branfosj/2e5048cfadecbf1d93d864865e2c03cf for a full test report.

[branfosj]

Going in, thanks @Flamefire!