MCPcopy Create free account
hub / github.com/eric-mitchell/direct-preference-optimization / train.py

File train.py

train.py:None–None  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

1import torch
2torch.backends.cuda.matmul.allow_tf32 = True
3import torch.nn as nn
4import transformers

Callers

nothing calls this directly

Calls 2

get_local_run_dirFunction · 0.90
mainFunction · 0.85

Tested by

no test coverage detected