todpole3

Latency Comparison Numbers (~2012)
----------------------------------
L1 cache reference                           0.5 ns
Branch mispredict                            5   ns
L2 cache reference                           7   ns                      14x L1 cache
Mutex lock/unlock                           25   ns
Main memory reference                      100   ns                      20x L2 cache, 200x L1 cache
Compress 1K bytes with Zippy             3,000   ns        3 us
Send 1K bytes over 1 Gbps network       10,000   ns       10 us
Read 4K randomly from SSD*             150,000   ns      150 us          ~1GB/sec SSD

todpole3

"""
Calculate the total number of parameters in a Transformer Decoder.

Usage:
    # OPT 125M
    python3 transformer_decoder_parameters.py --num-decoder-layers 12 --hidden-dim 768 --num-heads 12 --vocab-size 50272 --sequence-len 2048 --use-learned-pos-emb True
    # OPT 350M
    python3 transformer_decoder_parameters.py --num-decoder-layers 24 --hidden-dim 1024 --num-heads 16 --vocab-size 50272 --sequence-len 2048 --use-learned-pos-emb True
    # OPT 1.3B
    python3 transformer_decoder_parameters.py --num-decoder-layers 24 --hidden-dim 2048 --num-heads 32 --vocab-size 50272 --sequence-len 2048 --use-learned-pos-emb True

todpole3

"""
Code adapted from
https://github.com/aws-samples/mturk-code-samples/blob/master/Python/CreateHitSample.py

Package version:
    boto3==1.10.45
    
Usage:
    python3 reproduce_boto3_character_encoding_bug.py --aws_credentials /tmp/accessKeys.csv
"""