Jingya HF staff commited on
Commit
054c24a
1 Parent(s): 6b1b147

Upload decoder/config.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. decoder/config.json +76 -0
decoder/config.json ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/tmp/tmplp684gkk/decoder/config.json",
3
+ "bos_token_id": 0,
4
+ "classifier_dropout": 0.0,
5
+ "d_ff": 37,
6
+ "d_kv": 8,
7
+ "d_model": 32,
8
+ "decoder_start_token_id": 0,
9
+ "dense_act_fn": "relu",
10
+ "dropout_rate": 0.1,
11
+ "eos_token_id": 1,
12
+ "feed_forward_proj": "relu",
13
+ "gradient_checkpointing": false,
14
+ "initializer_factor": 0.002,
15
+ "is_encoder_decoder": true,
16
+ "is_gated_act": false,
17
+ "layer_norm_epsilon": 1e-06,
18
+ "model_type": "t5",
19
+ "neuron": {
20
+ "auto_cast": "matmul",
21
+ "auto_cast_type": "bf16",
22
+ "compiler_type": "neuronx-cc",
23
+ "compiler_version": "2.11.0.34+c5231f848",
24
+ "disable_fallback": false,
25
+ "disable_fast_relayout": false,
26
+ "dynamic_batch_size": false,
27
+ "input_names": [
28
+ "decoder_input_ids",
29
+ "decoder_attention_mask",
30
+ "encoder_hidden_states",
31
+ "attention_mask",
32
+ "beam_idx",
33
+ "beam_scores"
34
+ ],
35
+ "model_type": "t5-decoder",
36
+ "output_names": [
37
+ "next_token_scores",
38
+ "next_tokens",
39
+ "next_indices",
40
+ "past.0.self.key",
41
+ "past.1.self.key",
42
+ "past.2.self.key",
43
+ "past.3.self.key",
44
+ "past.4.self.key",
45
+ "past.0.self.value",
46
+ "past.1.self.value",
47
+ "past.2.self.value",
48
+ "past.3.self.value",
49
+ "past.4.self.value",
50
+ "past.0.cross.key",
51
+ "past.1.cross.key",
52
+ "past.2.cross.key",
53
+ "past.3.cross.key",
54
+ "past.4.cross.key",
55
+ "past.0.cross.value",
56
+ "past.1.cross.value",
57
+ "past.2.cross.value",
58
+ "past.3.cross.value",
59
+ "past.4.cross.value"
60
+ ],
61
+ "static_batch_size": 1,
62
+ "static_num_beams": 4,
63
+ "static_sequence_length": 32
64
+ },
65
+ "num_decoder_layers": 5,
66
+ "num_heads": 4,
67
+ "num_layers": 5,
68
+ "pad_token_id": 0,
69
+ "relative_attention_max_distance": 128,
70
+ "relative_attention_num_buckets": 8,
71
+ "task": "text2text-generation",
72
+ "torchscript": true,
73
+ "transformers_version": "4.34.1",
74
+ "use_cache": true,
75
+ "vocab_size": 1103
76
+ }