mpt-7b-peft-compatible / modeling_mpt.py

Commit History

Make sure hidden state and wte weights are on same device when in parallel model.
28721e3

muelletm commited on

Update modeling_mpt.py
a5eab52

cekal commited on

Update modeling_mpt.py
f2f3202

cekal commited on

Update modeling_mpt.py
660159e

cekal commited on

Update modeling_mpt.py
6184b87

cekal commited on

Update modeling_mpt.py
a3f512d

cekal commited on

Upload 16 files
300e678

cekal commited on