nguyendangsonlam commited on
Commit
0727ad1
1 Parent(s): 8547a97

Upload 2 files

Browse files
Files changed (2) hide show
  1. config.json +7 -7
  2. pytorch_model.bin +2 -2
config.json CHANGED
@@ -1,11 +1,11 @@
1
  {
2
- "_name_or_path": "microsoft/GODEL-v1_1-base-seq2seq",
3
  "architectures": [
4
  "T5ForConditionalGeneration"
5
  ],
6
- "d_ff": 3072,
7
  "d_kv": 64,
8
- "d_model": 768,
9
  "decoder_start_token_id": 0,
10
  "dense_act_fn": "relu",
11
  "dropout_rate": 0.1,
@@ -17,9 +17,9 @@
17
  "layer_norm_epsilon": 1e-06,
18
  "model_type": "t5",
19
  "n_positions": 512,
20
- "num_decoder_layers": 12,
21
- "num_heads": 12,
22
- "num_layers": 12,
23
  "output_past": true,
24
  "pad_token_id": 0,
25
  "relative_attention_max_distance": 128,
@@ -54,7 +54,7 @@
54
  }
55
  },
56
  "torch_dtype": "float32",
57
- "transformers_version": "4.25.1",
58
  "use_cache": true,
59
  "vocab_size": 32102
60
  }
 
1
  {
2
+ "_name_or_path": "microsoft/GODEL-v1_1-large-seq2seq",
3
  "architectures": [
4
  "T5ForConditionalGeneration"
5
  ],
6
+ "d_ff": 4096,
7
  "d_kv": 64,
8
+ "d_model": 1024,
9
  "decoder_start_token_id": 0,
10
  "dense_act_fn": "relu",
11
  "dropout_rate": 0.1,
 
17
  "layer_norm_epsilon": 1e-06,
18
  "model_type": "t5",
19
  "n_positions": 512,
20
+ "num_decoder_layers": 24,
21
+ "num_heads": 16,
22
+ "num_layers": 24,
23
  "output_past": true,
24
  "pad_token_id": 0,
25
  "relative_attention_max_distance": 128,
 
54
  }
55
  },
56
  "torch_dtype": "float32",
57
+ "transformers_version": "4.26.1",
58
  "use_cache": true,
59
  "vocab_size": 32102
60
  }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2060ed7848842f0662b424ed3a81015edf8b7c60122672e003f28f88f9c7fcf1
3
- size 891623057
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0515cd9190b20363430a13f626443fc2e3085b61d534be511dc28337d560e76a
3
+ size 2950750420