fsa

config.json

@@ -1,24 +1,23 @@
-
-  {
-  "_name_or_path": "Bakobiibizo/mistral-hf-transformer",
+{
+  "_name_or_path":"mistral-hf-transformer",
   "architectures": [
-    "MistralForCausalLM"
+    "LlamaForCausalLM"
   ],
-  "attention_dropout": 0.0,
   "bos_token_id": 1,
   "eos_token_id": 2,
   "hidden_act": "silu",
-  "hidden_size": 4096,
+  "hidden_size": 5120,
   "initializer_range": 0.02,
-  "intermediate_size": 14336,
-  "max_position_embeddings": 32768,
-  "model_type": "mistral",
-  "num_attention_heads": 32,
-  "num_hidden_layers": 32,
+  "intermediate_size": 13824,
+  "max_position_embeddings": 34768,
+  "model_type": "llama",
+  "num_attention_heads": 40,
+  "num_hidden_layers": 40,
   "num_key_value_heads": 8,
+  "pad_token_id": 0,
+  "pretraining_tp": 1,
   "rms_norm_eps": 1e-05,
-  "rope_theta": 10000.0,
-  "sliding_window": 4096,
+  "rope_scaling": null,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.36.2",