Unbreak benchmark_decoders.py

[ahmadsharif1]

benchmark_decoders.py was broken by #359

This PR unbreaks it. Longer term we could add a stronger check but this is good enough for now (and better than the non-throughput checks which doesn't even check for the length of the returned list or tensor).

Tested using:

python benchmarks/decoders/benchmark_decoders.py --bm_video_speed_min_run_seconds 1 --decoders=torchcodec_core_batch,torchcodec_core
video_files_paths=['/home/ahmads/personal/torchcodec/benchmarks/decoders/../../test/resources/nasa_13013.mp4']
video=/home/ahmads/personal/torchcodec/benchmarks/decoders/../../test/resources/nasa_13013.mp4, decoder=TorchCodecCore
video=/home/ahmads/personal/torchcodec/benchmarks/decoders/../../test/resources/nasa_13013.mp4, decoder=TorchCodecCoreBatch
[------------------------------------------------------------------ video=/home/ahmads/personal/torchcodec/benchmarks/decoders/../../test/resources/nasa_13013.mp4 h264 480x270, 13.013s 29.97002997002997fps ------------------------------------------------------------------]
                           |  uniform 10 seek()+next()  |  batch uniform 10 seek()+next()  |  random 10 seek()+next()  |  batch random 10 seek()+next()  |  1 next()  |  batch 1 next()  |  10 next()  |  batch 10 next()  |  100 next()  |  batch 100 next()  |  create()+next()
1 threads: ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
      TorchCodecCore       |            77.6            |              1001.7              |            53.7           |              722.6              |    17.5    |      246.0       |     22.6    |       228.0       |     67.2     |       777.4        |        20.4     
      TorchCodecCoreBatch  |            67.9            |               978.2              |            66.9           |              813.5              |    24.5    |      279.9       |     26.1    |       255.8       |     76.3     |       765.7        |                 

Times are in milliseconds (ms).

[scotts]

The check seems fine to me - I don't think it's too important what the check is. We just have to call future.result() in order to make exceptions propagate. Putting it in an assertion seems the most reasonable thing to me, as if we call it in a void context, we'll probably get yelled at by the linter.