SDCA Error w/ L2=0

[justinormont]

When using l2=0 in SDCA, our warning message throws, as it doesn't have enough parameters.

Error:

 Unexpected exception: Index (zero based) must be greater than or equal to zero and less than the size of the argument list., 'System.FormatException'
   at System.Text.StringBuilder.AppendFormatHelper(IFormatProvider provider, String format, ParamsArray args)
   at System.String.FormatHelper(IFormatProvider provider, String format, ParamsArray args)
   at System.String.Format(String format, Object[] args)
   at Microsoft.ML.Runtime.Data.TlcEnvironment.ConsoleWriter.PrintMessage(IMessageSource sender, ChannelMessage msg) in E:\TLC_git\TLC\OpenSource\src\Microsoft.ML.Core\Environment\TlcEnvironment.cs:line 53
   at Microsoft.ML.Runtime.Data.TlcEnvironment.PrintMessage(IMessageSource src, ChannelMessage msg) in E:\TLC_git\TLC\OpenSource\src\Microsoft.ML.Core\Environment\TlcEnvironment.cs:line 396
   at Microsoft.ML.Runtime.Data.HostEnvironmentBase`1.Dispatcher`1.DispatchCore(IMessageSource sender, TMessage message) in E:\TLC_git\TLC\OpenSource\src\Microsoft.ML.Core\Environment\HostEnvironmentBase.cs:line 338
   at Microsoft.ML.Runtime.Data.HostEnvironmentBase`1.ChannelBase.Warning(MessageSensitivity sensitivity, String fmt, Object[] args) in E:\TLC_git\TLC\OpenSource\src\Microsoft.ML.Core\Environment\HostEnvironmentBase.cs:line 276
   at Microsoft.ML.Runtime.Learners.SdcaTrainerBase`1.ArgumentsBase.Check(IHostEnvironment env) in E:\TLC_git\TLC\OpenSource\src\Microsoft.ML.StandardLearners\Standard\LinearClassificationTrainer.cs:line 219
   at Microsoft.ML.Runtime.Learners.SdcaMultiClassTrainer..ctor(IHostEnvironment env, Arguments args) in E:\TLC_git\TLC\OpenSource\src\Microsoft.ML.StandardLearners\Standard\SdcaMultiClass.cs:line 68

Cause:
L2LowerBound needs to be added to the warning's params.

machinelearning/src/Microsoft.ML.StandardLearners/Standard/LinearClassificationTrainer.cs

Lines 210 to 212 in 5e08fa1

	ch.Warning("The specified l2Const = {0} is too small. SDCA optimizes the dual objective function. " +
	"The dual formulation is only valid with a positive L2 regularization. Also, an l2Const less than {1} " +
	"could drastically slow down the convergence. So using l2Const = {1} instead.", L2Const);