File size: 336 Bytes
0429f3c
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
tokens 640049232
shards 7
batches 9764
window 65536
grad_accum 8
usable_steps 1220
mix b5_65536_131072 target 0.5000 actual 0.5000 tokens 320054446
mix b4_32768_65536 target 0.2500 actual 0.2500 tokens 160001134
mix b3_16384_32768 target 0.1500 actual 0.1500 tokens 95998599
mix b1_4096_8192 target 0.1000 actual 0.1000 tokens 63995053