trying bigger model

config.json

@@ -6,15 +6,15 @@
   "bos_token_id": 1,
   "embd_pdrop": 0.0,
   "eos_token_id": 2,
-  "hidden_act": "relu",
-  "hidden_size": 64,
+  "hidden_act": "silu",
+  "hidden_size": 256,
   "initializer_range": 0.02,
   "intermediate_size": 64,
   "max_position_embeddings": 4096,
   "model_type": "phi3",
-  "num_attention_heads": 1,
+  "num_attention_heads": 4,
   "num_hidden_layers": 4,
-  "num_key_value_heads": 1,
+  "num_key_value_heads": 4,
   "original_max_position_embeddings": 4096,
   "pad_token_id": 2,
   "resid_pdrop": 0.0,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d508dfe1255503935d052115e70ea96ba046e5cc30bf1627df38c61e1e5429c1
-size 594920
+oid sha256:860fc69ac092b7ea3cde0ca19bfd7703247cf60f156f4a8c8f55950dd93ac55a
+size 5517104