Update examples/neural_compressor/language-modeling/run_clm.py

PenghuiCheng · echarlaix · web-flow · commit 2a905cc66ab3 · 2024-03-25T10:27:06.000+08:00
Co-authored-by: Ella Charlaix &lt;80481427+echarlaix@users.noreply.github.com&gt;
diff --git a/examples/neural_compressor/language-modeling/run_clm.py b/examples/neural_compressor/language-modeling/run_clm.py
@@ -233,7 +233,7 @@ class OptimizationArguments:
             "help": "Percentage of Hessian's diagonal values average, which will be added to Hessian's diagonal to increase numerical stability, used for GPTQ quantization"
         },
     )
-    gptq_block_size: int = field(
+    block_size: int = field(
         default=128,
         metadata={"help": "Block size. sub weight matrix size to run GPTQ."},
     )

Original file line number	Diff line number	Diff line change
`@@ -233,7 +233,7 @@ class OptimizationArguments:`
`233`	`233`	`"help": "Percentage of Hessian's diagonal values average, which will be added to Hessian's diagonal to increase numerical stability, used for GPTQ quantization"`
`234`	`234`	`},`
`235`	`235`	`)`
`236`		`- gptq_block_size: int = field(`
	`236`	`+ block_size: int = field(`
`237`	`237`	`default=128,`
`238`	`238`	`metadata={"help": "Block size. sub weight matrix size to run GPTQ."},`
`239`	`239`	`)`