update config and model (new modeling)

config.json

@@ -1,35 +1,23 @@
 {
-  "activation_function": "gelu",
-  "architectures": [
-    "GPTBigCodeLMHeadModel"
-  ],
-  "attention_softmax_in_fp32": true,
-  "attention_window_size": 8192,
-  "attn_pdrop": 0.1,
-  "bos_token_id": 0,
-  "embd_pdrop": 0.1,
-  "eos_token_id": 0,
-  "head_groups": 2,
-  "initializer_range": 0.02,
-  "layer_norm_epsilon": 1e-05,
-  "model_type": "gpt_bigcode",
-  "n_embd": 3072,
-  "n_head": 24,
-  "n_inner": 12288,
-  "n_layer": 30,
-  "n_positions": 16384,
-  "resid_pdrop": 0.1,
-  "rotary_embedding_scale": -13.81551,
-  "scale_attention_softmax_in_fp32": true,
-  "scale_attn_weights": true,
-  "summary_activation": null,
-  "summary_first_dropout": 0.1,
-  "summary_proj_to_labels": true,
-  "summary_type": "cls_index",
-  "summary_use_proj": true,
-  "transformers_version": "4.34.1",
+  "attention_dropout": 0.0,
+  "bos_token_id": 50256,
+  "eos_token_id": 50256,
+  "hidden_act": "gelu_pytorch_tanh",
+  "hidden_size": 3072,
+  "initializer_range": 0.018042,
+  "intermediate_size": 12288,
+  "max_position_embeddings": 16384,
+  "mlp_type": "default",
+  "model_type": "starcoder2",
+  "norm_epsilon": 1e-05,
+  "norm_type": "layer_norm",
+  "num_attention_heads": 24,
+  "num_hidden_layers": 30,
+  "num_key_value_heads": 2,
+  "rope_theta": 999999.4420358813,
+  "sliding_window": 4096,
+  "transformers_version": "4.37.0.dev0",
+  "use_bias": true,
   "use_cache": true,
-  "use_position_embeddings": false,
-  "use_rotary_embeddings": true,
   "vocab_size": 49152
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3147adbbd9adb07a4281b2a4f11e0e33b637a6e4d656f9898168cfdbbfb51518
-size 23636934338
+oid sha256:c6d17296643c062e7fcd59b4904e3410909292ae042c4c6be9bcdd6430a36080
+size 23731378826