[Misc] set single whitespace between log sentences (#13771)
Signed-off-by: cjackal <44624812+cjackal@users.noreply.github.com>
This commit is contained in:
+1
-1
@@ -71,7 +71,7 @@ class CompressedTensorsW8A8Fp8MoEMethod(CompressedTensorsMoEMethod):
|
||||
if not (self.weight_quant.strategy == QuantizationStrategy.TENSOR
|
||||
and self.input_quant.strategy == QuantizationStrategy.TENSOR):
|
||||
raise ValueError(
|
||||
"For FP8 Fused MoE layers, only per-tensor scales"
|
||||
"For FP8 Fused MoE layers, only per-tensor scales "
|
||||
"for weights and activations are supported. Found "
|
||||
f"{self.weight_quant}, {self.input_quant}")
|
||||
|
||||
|
||||
@@ -74,7 +74,7 @@ class GPTQConfig(QuantizationConfig):
|
||||
def __repr__(self) -> str:
|
||||
return (f"GPTQConfig(weight_bits={self.weight_bits}, "
|
||||
f"group_size={self.group_size}, "
|
||||
f"desc_act={self.desc_act}),"
|
||||
f"desc_act={self.desc_act}), "
|
||||
f"lm_head_quantized={self.lm_head_quantized}), "
|
||||
f"dynamic={self.dynamic}")
|
||||
|
||||
|
||||
@@ -56,7 +56,7 @@ class ModelOptFp8Config(QuantizationConfig):
|
||||
quant_method = quant_config["quant_algo"]
|
||||
is_checkpoint_fp8_serialized = ("FP8" in quant_method)
|
||||
if not is_checkpoint_fp8_serialized:
|
||||
raise ValueError("ModelOpt currently only supports static FP8"
|
||||
raise ValueError("ModelOpt currently only supports static FP8 "
|
||||
"quantization in vLLM. Please check the "
|
||||
"`hf_quant_config.json` file for your model's "
|
||||
"quant configuration.")
|
||||
|
||||
@@ -25,8 +25,8 @@ class NeuronQuantConfig(QuantizationConfig):
|
||||
if self.quant_dtype not in SUPPORTED_QUANT_DTYPE_LIST:
|
||||
raise ValueError(
|
||||
f"Neuron quantization datatype {self.quant_dtype} is not valid,"
|
||||
f"the quantization datatype should match one of the below types"
|
||||
f"{SUPPORTED_QUANT_DTYPE_LIST}")
|
||||
f" the quantization datatype should match one of the below "
|
||||
f"types {SUPPORTED_QUANT_DTYPE_LIST}")
|
||||
self.dequant_dtype = dequant_dtype
|
||||
self.quantize_method = quantize_method
|
||||
|
||||
|
||||
@@ -55,7 +55,7 @@ class QuarkW8A8Fp8MoEMethod(QuarkMoEMethod):
|
||||
if not (weight_qscheme == "per_tensor"
|
||||
and input_qscheme == "per_tensor"):
|
||||
raise ValueError(
|
||||
"For FP8 Fused MoE layers, only per-tensor scales"
|
||||
"For FP8 Fused MoE layers, only per-tensor scales "
|
||||
"for weights and activations are supported. Found "
|
||||
f"{weight_qscheme}, {input_qscheme}") # noqa E501
|
||||
|
||||
|
||||
@@ -118,7 +118,7 @@ def verify_marlin_supports_shape(output_size_per_partition: int,
|
||||
and input_size_per_partition % group_size != 0):
|
||||
raise ValueError(
|
||||
f"Weight input_size_per_partition = {input_size_per_partition}"
|
||||
f" is not divisible by group_size = {group_size}."
|
||||
f" is not divisible by group_size = {group_size}. "
|
||||
"Consider reducing tensor_parallel_size or running "
|
||||
"with --quantization gptq.")
|
||||
|
||||
|
||||
Reference in New Issue
Block a user