Fix warnings

Related-To: #126

Author: michalharakal

skainet-lang/skainet-lang-api/src/commonMain/kotlin/sk/ai/net/lang/types/Int32.kt

@@ -11,7 +11,7 @@ public object Int32 : DType {
             is Int8 -> true     // Int8 can promote to Int32
             is Int32 -> true    // Same type compatibility
             is FP32 -> true     // Can promote to FP32
-            else -> false
+            is FP16 -> true     // Can promote to FP16
         }
     }
 
@@ -23,9 +23,6 @@ public object Int32 : DType {
             is Int32 -> Int32   // Int32 + Int32 → Int32
             is FP32 -> FP32     // Int32 + FP32 → FP32
             is FP16 -> FP16     // Int32 + FP16 → FP16
-            else -> throw IllegalArgumentException(
-                "Cannot promote Int32 with incompatible type: ${other.name}"
-            )
         }
     }
 }

skainet-lang/skainet-lang-memory/src/commonMain/kotlin/sk/ai/net/core/tensor/data/DenseTensorDataFactory.kt

@@ -60,22 +60,22 @@ public class DenseTensorDataFactory {
     ): TensorData<T, V> {
         return when (dtype) {
             is FP32 -> {
-                class FP32FloatTensorData(private val data: FloatArray) : TensorData<FP32, Float> {
+                class FP32FloatTensorData() : TensorData<FP32, Float> {
                     override val shape: Shape
                         get() = Shape.Companion(data.size)
 
                     override fun get(vararg indices: Int): Float = data[indices[0]]
                 }
-                FP32FloatTensorData(data) as TensorData<T, V>
+                FP32FloatTensorData() as TensorData<T, V>
             }
             is FP16 -> {
-                class FP16FloatTensorData(private val data: FloatArray) : TensorData<FP16, Float> {
+                class FP16FloatTensorData() : TensorData<FP16, Float> {
                     override val shape: Shape
                         get() = Shape.Companion(data.size)
 
                     override fun get(vararg indices: Int): Float = data[indices[0]]
                 }
-                FP16FloatTensorData(data) as TensorData<T, V>
+                FP16FloatTensorData() as TensorData<T, V>
             }
             else -> throw IllegalArgumentException("Unsupported dtype for FloatArray: ${dtype.name}")
         }
@@ -88,13 +88,13 @@ public class DenseTensorDataFactory {
     ): TensorData<T, V> {
         return when (dtype) {
             is Int32 -> {
-                class Int32IntTensorData(private val data: IntArray) : TensorData<Int32, Int> {
+                class Int32IntTensorData() : TensorData<Int32, Int> {
                     override val shape: Shape
                         get() = Shape.Companion(data.size)
 
                     override fun get(vararg indices: Int): Int = data[indices[0]]
                 }
-                Int32IntTensorData(data) as TensorData<T, V>
+                Int32IntTensorData() as TensorData<T, V>
             }
             else -> throw IllegalArgumentException("Unsupported dtype for IntArray: ${dtype.name}")
         }