LLamaContextParams epsilon and tensor split changes

LLama/Native/LLamaContextParams.cs

@@ -32,7 +32,7 @@ namespace LLama.Native
         /// <summary>
         /// rms norm epsilon (TEMP - will be moved to model hparams)
         /// </summary>
-        float rms_norm_eps;
+       public float rms_norm_eps;
 
         /// <summary>
         /// number of layers to store in VRAM
@@ -47,19 +47,19 @@ namespace LLama.Native
         /// <summary>
         /// how to split layers across multiple GPUs
         /// </summary>
-        public TensorSplits tensor_split;
+        public float[] tensor_split;
 
         /// <summary>
         /// ref: https://github.com/ggerganov/llama.cpp/pull/2054
         /// RoPE base frequency
         /// </summary>
-        float rope_freq_base;
+        public float rope_freq_base;
 
         /// <summary>
         /// ref: https://github.com/ggerganov/llama.cpp/pull/2054
         /// RoPE frequency scaling factor
         /// </summary>
-        float rope_freq_scale; 
+        public float rope_freq_scale; 
 
         /// <summary>
         /// called with a progress value between 0 and 1, pass NULL to disable

LLama/Utils.cs

@@ -28,16 +28,13 @@ namespace LLama
             lparams.logits_all = @params.Perplexity;
             lparams.embedding = @params.EmbeddingMode;
             lparams.low_vram = @params.LowVram;
-
-            if(@params.TensorSplits.Length != 1)
+       
+            if (@params.TensorSplits.Length != 1)
             {
                 throw new ArgumentException("Currently multi-gpu support is not supported by " +
                     "both llama.cpp and LLamaSharp.");
             }
-            lparams.tensor_split = new TensorSplits()
-            {
-                Item1 = @params.TensorSplits[0]
-            };
+            lparams.tensor_split = @params.TensorSplits;
 
             if (!File.Exists(@params.ModelPath))
             {