Skip to content

Commit

Permalink
add quant norm
Browse files Browse the repository at this point in the history
  • Loading branch information
qwopqwop200 committed May 8, 2023
1 parent 3f1c582 commit c1cfbbc
Showing 1 changed file with 6 additions and 3 deletions.
9 changes: 6 additions & 3 deletions llama_inference_offload.py
Original file line number Diff line number Diff line change
Expand Up @@ -213,9 +213,12 @@ def noop(*args, **kwargs):
load_checkpoint_in_model(model, checkpoint, dtype='float16')
model.seqlen = 2048

quant.make_quant_attn(model)
if fused_mlp:
quant.make_fused_mlp(model)
if eval:
quant.make_quant_attn(model)
quant.make_quant_norm(model)
if fused_mlp:
quant.make_fused_mlp(model)


if warmup_autotune:
quant.autotune_warmup_linear(model)
Expand Down

0 comments on commit c1cfbbc

Please sign in to comment.