Remove the `norm_first` option

config.json

@@ -3,6 +3,7 @@
   "architectures": [
     "GSAForCausalLM"
   ],
+  "attn": null,
   "bos_token_id": 1,
   "clamp_max": null,
   "clamp_min": null,
@@ -33,11 +34,11 @@
   "share_conv_kernel": true,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.40.2",
+  "transformers_version": "4.48.2",
   "use_cache": true,
   "use_norm": true,
   "use_output_gate": false,
   "use_rope": false,
   "use_short_conv": false,
   "vocab_size": 32000
-}
+}

generation_config.json

@@ -2,5 +2,6 @@
   "_from_model_config": true,
   "bos_token_id": 1,
   "eos_token_id": 2,
-  "transformers_version": "4.40.2"
+  "pad_token_id": 2,
+  "transformers_version": "4.48.2"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7e1501b1d6a039918f18a9919e5356b1971997fc7665f2cc079d924f68630785
+oid sha256:bf078acce9ada5103e493a130c9989865e6d6b4b86595402f6ac228de0baed34
 size 2753785520