Upload MllamaForCausalLM

config.json

@@ -2,7 +2,7 @@
   "architectures": [
     "MllamaForCausalLM"
   ],
-  "bos_token_id": 128000,
+  "bos_token_id": 128264,
   "cross_attention_layers": [
     3,
     8,
@@ -13,7 +13,7 @@
     33
   ],
   "dropout": 0,
-  "eos_token_id": 128001,
+  "eos_token_id": 128265,
   "hidden_act": "silu",
   "hidden_size": 3072,
   "initializer_range": 0.02,
@@ -23,7 +23,7 @@
   "num_attention_heads": 24,
   "num_hidden_layers": 35,
   "num_key_value_heads": 8,
-  "pad_token_id": 128004,
+  "pad_token_id": 128265,
   "rms_norm_eps": 1e-05,
   "rope_scaling": {
     "factor": 32.0,
@@ -37,5 +37,5 @@
   "torch_dtype": "float16",
   "transformers_version": "4.47.1",
   "use_cache": true,
-  "vocab_size": 128256
+  "vocab_size": 128266
 }

generation_config.json

@@ -1,8 +1,8 @@
 {
   "_from_model_config": true,
-  "bos_token_id": 128000,
+  "bos_token_id": 128264,
   "do_sample": true,
-  "eos_token_id": 128001,
-  "pad_token_id": 128004,
+  "eos_token_id": 128265,
+  "pad_token_id": 128265,
   "transformers_version": "4.47.1"
 }

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4013591bfab8a04b0b58b7f899af4dbf03992ec5b39b027a0b9dceec85d3a6ff
-size 4965851872
+oid sha256:8ad1a8969066564589926eb5833747e9fc0a7b6c111ea39c0580497ef8b41fa8
+size 4965864160

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:22095124879e80c991a87073decddd7e69509247bca29ca48914a29878f95a87
-size 3657117220
+oid sha256:d56e4f9b6a3a090fa890113cfe6a92cdd33cecd5eecde7602d73d8d047349ecc
+size 3657178660

model.safetensors.index.json

@@ -1,6 +1,6 @@
 {
   "metadata": {
-    "total_size": 8622929436
+    "total_size": 8623003164
   },
   "weight_map": {
     "lm_head.weight": "model-00002-of-00002.safetensors",