Update `max_position_embeddings` to 4096

The model has a 4096 tokens context length, which should be correctly reflected in the `config.json`.

See https://github.com/facebookresearch/codellama/blob/1af62e1f43db1fa5140fa43cb828465a603a48f3/llama/model.py#L277 in reference implementation (`self.params.max_seq_len * 2` where `self.params.max_seq_len == 2048`). Also confirmed offline with a Meta engineer.

config.json

@@ -10,7 +10,7 @@
   "hidden_size": 8192,
   "initializer_range": 0.02,
   "intermediate_size": 28672,
-  "max_position_embeddings": 2048,
+  "max_position_embeddings": 4096,
   "model_type": "llama",
   "num_attention_heads": 64,
   "num_hidden_layers": 80,