We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent 29025d0 commit e38c077Copy full SHA for e38c077
.github/workflows/apple-perf.yml
@@ -230,7 +230,7 @@ jobs:
230
model.use_sdpa_with_kv_cache=true \
231
backend.xnnpack.enabled=true \
232
backend.xnnpack.extended_ops=true \
233
- base.preq_mode="8da4w_output_8da8w" \
+ base.preq_mode="preq_8da4w_out_8da8w" \
234
base.preq_group_size=32 \
235
export.max_seq_length=2048 \
236
export.max_context_length=2048 \
@@ -256,7 +256,7 @@ jobs:
256
base.params="${DOWNLOADED_PATH}/params.json" \
257
quantization.use_qat=true \
258
base.use_lora=16 \
259
260
261
base.preq_embedding_quantize=\'8,0\' \
262
0 commit comments