Fix d_model & d_ff

config.json

@@ -1,35 +1,30 @@
 {
-  "_name_or_path": "long-t5-tglobal-large",
   "architectures": [
     "LongT5ForConditionalGeneration"
   ],
-  "d_ff": 5120,
+  "d_ff": 10240,
   "d_kv": 64,
-  "d_model": 1024,
-  "decoder_start_token_id": 0,
-  "dense_act_fn": "gelu_new",
+  "d_model": 4096,
+  "dense_act_fn": "relu",
   "dropout_rate": 0.1,
-  "encoder_attention_type": "transient-global",
+  "encoder_attention_type": "local",
   "eos_token_id": 1,
-  "feed_forward_proj": "gated-gelu",
+  "feed_forward_proj": "relu",
   "global_block_size": 16,
   "initializer_factor": 1.0,
   "is_encoder_decoder": true,
-  "is_gated_act": true,
+  "is_gated_act": false,
   "layer_norm_epsilon": 1e-06,
   "local_radius": 127,
   "model_type": "longt5",
-  "n_positions": 4096,
-  "num_decoder_layers": 24,
+  "num_decoder_layers": 48,
   "num_heads": 64,
   "num_layers": 48,
-  "output_past": true,
   "pad_token_id": 0,
   "relative_attention_max_distance": 128,
   "relative_attention_num_buckets": 32,
-  "tie_word_embeddings": false,
   "torch_dtype": "float32",
-  "transformers_version": "4.23.1",
+  "transformers_version": "4.25.0.dev0",
   "use_cache": true,
   "vocab_size": 32128
 }

pytorch_model-00001-of-00008.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a6ea314d9d639dd15280e0c8a805a357f6e46900ae599825439057c90496355
+size 9418835479

pytorch_model-00002-of-00008.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3e4c7367f05566ee45aad7cf690d6dc891be574b467e627d52662065f95ad20
+size 9966253699

pytorch_model-00003-of-00008.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38bb749ace5a16554c4a402ddcc79b3b75b72b473f04bfe2cf243c3f2b6a8ed5
+size 9966253711

pytorch_model-00004-of-00008.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:360ee8f6eea8562796c8462262262db07a5019944bf8c22c567551609f36b618
+size 9955796714

pytorch_model-00005-of-00008.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d291ea24cc6a52606a4c968c6d8fbfc0b23c3d7b0505643b9726a0b066acd09
+size 9999829704

pytorch_model-00006-of-00008.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aaa6112bda4901571cbe7fea4c0679606793a4cc9073383f00b214d43153d4d3
+size 9999830697

pytorch_model-00007-of-00008.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e486b640f0ac3f22bd921506943fed9257bc78fd54cf38a22f7b9469fcb806d2
+size 9999846391

pytorch_model-00008-of-00008.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc7cf7b4ebb3a1efe9ce5a1bc6d07f672b97dc75c0b24012f4601adab0e526b1
+size 3143808689

pytorch_model.bin.index.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:12ca3616ff8422a5cbf510b62876f29315db066e1256801b5931a70cabc3b620
-size 79600
+oid sha256:0e792ebbc02388bfa94c5bcb167e76cde3f7e76571aefd43688c2dde31818f5b
+size 92758