Update config.json

config.json

@@ -1,7 +1,12 @@
 {
+  "_name_or_path": "output",
+  "architectures": [
+    "DolphinForCausalLM"
+  ],
   "attention_dropout": 0.0,
+  "bos_token_id": 151643,
   "encoder_config": {
-    "_name_or_path": "Qwen/Qwen2-0.5B",
+    "_name_or_path": "alexchen4ai/Qwen2-0.5B",
     "add_cross_attention": false,
     "architectures": [
       "Qwen2ForCausalLM"
@@ -83,22 +88,24 @@
     "use_sliding_window": false,
     "vocab_size": 151936
   },
+  "eos_token_id": 151643,
   "hidden_act": "silu",
-  "hidden_size": 4096,
+  "hidden_size": 3584,
   "initializer_range": 0.02,
-  "intermediate_size": 22016,
-  "max_position_embeddings": 32768,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 131072,
   "max_window_layers": 28,
   "model_type": "dolphin",
-  "num_attention_heads": 32,
-  "num_hidden_layers": 32,
-  "num_key_value_heads": 32,
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
   "rms_norm_eps": 1e-06,
-  "rope_theta": 10000.0,
-  "sliding_window": 4096,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
   "tie_word_embeddings": false,
-  "transformers_version": "4.44.2",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.42.4",
   "use_cache": true,
   "use_sliding_window": false,
-  "vocab_size": 151936
-}
+  "vocab_size": 152064
+}