Update config.json

config.json

@@ -3,44 +3,50 @@
   "model_type": "llama",
   "torch_dtype": "float16",
   "transformers_version": "4.46.3",
-  "architectures": ["LlamaForCausalLM"],
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
   "vocab_size": 128256,
   "bos_token_id": 128000,
-  "eos_token_id": [128001, 128008, 128009],
+  "eos_token_id": [
+    128001,
+    128008,
+    128009
+  ],
   "pad_token_id": 128009,
   "hidden_size": 3072,
   "intermediate_size": 8192,
   "num_hidden_layers": 28,
   "num_attention_heads": 24,
-    "num_key_value_heads": 8,
+  "num_key_value_heads": 8,
   "head_dim": 128,
   "max_position_embeddings": 131072,
-    "rope_scaling": {
+  "rope_scaling": {
     "type": "linear",
-    "factor": 32.0
+    "factor": 32
   },
-    "rope_theta": 500000.0,
-    "hidden_act": "silu",
+  "rope_theta": 500000,
+  "hidden_act": "silu",
   "initializer_range": 0.02,
-  "rms_norm_eps": 1e-05,
-  "layer_norm_eps": 1e-05,
+  "rms_norm_eps": 0.00001,
+  "layer_norm_eps": 0.00001,
   "attention_bias": false,
   "mlp_bias": false,
   "pretraining_tp": 1,
   "tie_word_embeddings": true,
   "use_cache": true,
   "shared_input_output_embedding": true,
-    "attention_dropout": 0.0,
-    "hidden_dropout_prob": 0.1,
-    "attn_dropout_prob": 0.1,
-  "residual_dropout_prob": 0.0,
+  "attention_dropout": 0,
+  "hidden_dropout_prob": 0.1,
+  "attn_dropout_prob": 0.1,
+  "residual_dropout_prob": 0,
   "ffn_dropout_prob": 0.1,
- "output_projection": true,
-   "position_embedding_type": "rotary",
+  "output_projection": true,
+  "position_embedding_type": "rotary",
   "use_rope": true,
   "gradient_checkpointing": false,
-   "multi_query_attention": false,
-    "output_attentions": false,
+  "multi_query_attention": false,
+  "output_attentions": false,
   "output_hidden_states": false,
-    "layerdrop": 0.0
+  "layerdrop": 0
 }