much bigger model

config.json

@@ -7,14 +7,14 @@
   "embd_pdrop": 0.0,
   "eos_token_id": 2,
   "hidden_act": "silu",
-  "hidden_size": 256,
+  "hidden_size": 768,
   "initializer_range": 0.02,
-  "intermediate_size": 64,
+  "intermediate_size": 768,
   "max_position_embeddings": 4096,
   "model_type": "phi3",
-  "num_attention_heads": 4,
+  "num_attention_heads": 12,
   "num_hidden_layers": 4,
-  "num_key_value_heads": 4,
+  "num_key_value_heads": 12,
   "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:a33e2ba1a0aeca11e1db5d74cfa7acb65a79d1bc5a610bfefd835ba2a38a1735
-size 2760000
+oid sha256:329e8bfcdb3e87e48cb3b5c9949156403598d719f99fe3d6febcaf17b322ed59
+size 33833344