Expected performance on p106-100 6gb

[MatPoliquin]

Anyone else tried on a p106-100 6gb (mining version of a GTX 1060)?
I get around 2800 fps with 100% GPU Utilization

cmd line:
python3 vtrace_main.py --env-name PongNoFrameskip-v4 --normalize --use-cuda-env --num-ales 1200 --num-steps 20 --num-steps-per-update 1 --num-minibatches 20 --t-max 8000000 --evaluation-interval 200000

log:
Rank 0
P106-100 : 1708.500 Mhz (Ordinal 0)
10 SMs enabled. Compute Capability sm_61
FreeMem: 5,321MB TotalMem: 6,080MB 64-bit pointers.
Mem Clock: 4004.000 Mhz x 192 bits (192.2 GB/s)
ECC Disabled

Selected optimization level O0: Pure FP32 training.

Defaults for this optimization level are:
enabled : True
opt_level : O0
cast_model_type : torch.float32
patch_torch_functions : False
keep_batchnorm_fp32 : None
master_weights : False
loss_scale : 1.0
Processing user overrides (additional kwargs that are not None)...
After processing overrides, optimization options are:
enabled : True
opt_level : O0
cast_model_type : torch.float32
patch_torch_functions : False
keep_batchnorm_fp32 : None
master_weights : False
loss_scale : 1.0
Warning: multi_tensor_applier fused unscale kernel is unavailable, possibly because apex was installed without --cuda_ext --cpp_ext. Using Python fallback. Original ImportError was: ModuleNotFoundError("No module named 'amp_C'",)
0%| | 0/6667 [00:00<?, ?it/s][CuLE CPU] [training time: 00:00:00.000000 s] (length) min/max/mean/median: 757.0/763.0/759.7/759.0 --- (reward) min/max/mean/median: -21.0/-21.0/-21.0/-21.0
[OpAI CPU] [training time: 00:00:00.000000 s] (length) min/max/mean/median: 757.0/760.0/758.3/758.0 --- (reward) min/max/mean/median: -21.0/-21.0/-21.0/-21.0
3%|█▎ | 167/6667 [01:27<46:13, 2.34it/s, 2797.63f/s, min/max/mean/median reward: 0.0/ 0.0/ 0.0/ 0.0, entropy/value/policy: 1.7545/0.0216/-0.0093][CuLE CPU] [training time: 00:01:09.831456 s] (length) min/max/mean/median: 781.0/950.0/859.9/841.0 --- (reward) min/max/mean/median: -21.0/-20.0/-20.6/-21.0
[OpAI CPU] [training time: 00:01:09.831456 s] (length) min/max/mean/median: 777.0/987.0/904.5/898.0 --- (reward) min/max/mean/median: -21.0/-19.0/-20.4/-21.0
5%|██▌ | 334/6667 [03:00<45:21, 2.33it/s, 2798.67f/s, min/max/mean/median reward: 0.0/ 0.0/ 0.0/ 0.0, entropy/value/policy: 1.7062/0.0304/ 0.0559][CuLE CPU] [training time: 00:02:21.250146 s] (length) min/max/mean/median: 816.0/1106.0/930.6/903.0 --- (reward) min/max/mean/median: -21.0/-18.0/-20.3/-21.0