Upload model

config.json

@@ -112,6 +112,7 @@
       "tie_word_embeddings": true,
       "tokenizer_class": null
     },
+    "dtype": "float32",
     "head_dim": 128,
     "hidden_act": "silu",
     "hidden_size": 768,
@@ -1241,6 +1242,7 @@
   "text_config": {
     "attention_bias": false,
     "attention_dropout": 0.0,
+    "bos_token_id": 50281,
     "classifier_activation": "gelu",
     "classifier_bias": false,
     "classifier_dropout": 0.0,
@@ -1248,7 +1250,9 @@
     "cls_token_id": 50281,
     "decoder_bias": true,
     "deterministic_flash_attn": false,
+    "dtype": "float32",
     "embedding_dropout": 0.0,
+    "eos_token_id": 50282,
     "global_attn_every_n_layers": 3,
     "hidden_activation": "gelu",
     "hidden_size": 1024,
@@ -1288,6 +1292,7 @@
     "norm_eps": 1e-05,
     "num_attention_heads": 16,
     "num_hidden_layers": 22,
+    "pad_token_id": 50283,
     "repad_logits_with_grad": false,
     "rope_parameters": {
       "full_attention": {
@@ -1299,9 +1304,12 @@
         "rope_type": "default"
       }
     },
+    "sep_token_id": 50282,
     "sparse_pred_ignore_index": -100,
     "sparse_prediction": false,
+    "tie_word_embeddings": true,
     "vocab_size": 50368
   },
-  "transformers_version": "5.0.0.dev0"
+  "tie_word_embeddings": true,
+  "transformers_version": "5.0.1.dev0"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4f6982447d5eaec68443ffcfc3bcbebe7d2b00c330df73541b824f7072761305
-size 7828455824
+oid sha256:a8fc31329df9fab7fb3467873d4e8e40df40a486d9e6645feb9de6fa5c7b75d9
+size 3388082648