From 379c188d7773f7008b35aafe8d3f83c95ffd8f50 Mon Sep 17 00:00:00 2001 From: George Ohashi Date: Tue, 25 Feb 2025 10:47:53 -0500 Subject: [PATCH] add num workers on dataset processing - labels, tokenization --- examples/quantization_2of4_sparse_w4a16/llama7b_sparse_w4a16.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/examples/quantization_2of4_sparse_w4a16/llama7b_sparse_w4a16.py b/examples/quantization_2of4_sparse_w4a16/llama7b_sparse_w4a16.py index 2cca58815..d617cff12 100644 --- a/examples/quantization_2of4_sparse_w4a16/llama7b_sparse_w4a16.py +++ b/examples/quantization_2of4_sparse_w4a16/llama7b_sparse_w4a16.py @@ -33,6 +33,7 @@ bf16 = False # using full precision for training lr_scheduler_type = "cosine" warmup_ratio = 0.1 +preprocessing_num_workers = 8 # this will run the recipe stage by stage: # oneshot sparsification -> finetuning -> oneshot quantization @@ -52,6 +53,7 @@ learning_rate=learning_rate, lr_scheduler_type=lr_scheduler_type, warmup_ratio=warmup_ratio, + preprocessing_num_workers=preprocessing_num_workers, ) logger.info( "llmcompressor does not currently support running compressed models in the marlin24 format." # noqa