Safetensors
llama
alignment-handbook
trl
dpo
Generated from Trainer
File size: 217 Bytes
a9cd984
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
{
    "epoch": 8.0,
    "total_flos": 0.0,
    "train_loss": 5009.442099750906,
    "train_runtime": 15540.9203,
    "train_samples": 17626,
    "train_samples_per_second": 9.073,
    "train_steps_per_second": 0.142
}