Upload model

config.json

@@ -1,9 +1,10 @@
 {
+  "_name_or_path": "/data_storage/lbw/MMaDA/mmada-training-stage4-llada-instruct/checkpoint-130000/unwrapped_model",
   "activation_type": "silu",
   "alibi": false,
   "alibi_bias_max": 8.0,
   "architectures": [
-    "MMadaModelLM"
+    "LLaDAModelLM"
   ],
   "attention_dropout": 0.0,
   "attention_layer_norm": false,
@@ -36,7 +37,7 @@
   "max_sequence_length": 4096,
   "mlp_hidden_size": 12288,
   "mlp_ratio": 4,
-  "model_type": "mmada",
+  "model_type": "llada",
   "multi_query_attention": null,
   "n_heads": 32,
   "n_kv_heads": 32,
@@ -46,6 +47,7 @@
   "num_vq_tokens": 256,
   "pad_token_id": 126081,
   "precision": "amp_bf16",
+  "pretrained_model_path": "/data_storage/shared/pretrained_models/LLaDA-8B-Instruct",
   "residual_dropout": 0.0,
   "rms_norm_eps": 1e-05,
   "rope": true,

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9f57532a459839ce0f826e3cfa4d4ed8efb6293c8821057b3c03562a49e326b5
+oid sha256:b65828600a8bdc7b0d7fd61847f033c92b5982b0395b1e1741ba864736c2c680
 size 4928472568

model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d33d876b196678dc6ca2f3e678520bf0169c1de808e29c6746b7005472568bf1
+oid sha256:2441d4f4667fa8831d862cc33fbf153bfd320d22ad4c8d2f1e8088a739ce0a23
 size 4932693832

model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb6864fb9308d9cdb47cf8055777fe3aa09fa7dd0a775bb26663fd1943a3749a
+oid sha256:f6ff5c367093eb56a3639130f5525485b598b514a262da439f97597da6c89720
 size 4999819560

model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2accfd14cb55ecf8171c12e684e08477f46ecf7a24632347e115ef94e8ff2417
+oid sha256:3875e39d151b69d51808101ac5b95819ccc88380c6a65443b9a5a889c9ba05b1
 size 1304428928