jweb commited on
Commit
36fdf15
1 Parent(s): ec19f6b

Upload config.json

Browse files
Files changed (1) hide show
  1. config.json +75 -0
config.json ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/home/mycat/model/rinna/finetuned_natume",
3
+ "activation_function": "gelu_fast",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 2,
9
+ "d_ff": 2048,
10
+ "d_kv": 64,
11
+ "d_model": 2048,
12
+ "dropout_rate": 0.1,
13
+ "embd_pdrop": 0.1,
14
+ "eos_token_id": 3,
15
+ "initializer_factor": 1.0,
16
+ "initializer_range": 0.02,
17
+ "layer_norm_epsilon": 1e-05,
18
+ "model_type": "gpt2",
19
+ "n_ctx": 1024,
20
+ "n_embd": 2048,
21
+ "n_head": 16,
22
+ "n_inner": 8192,
23
+ "n_layer": 24,
24
+ "n_positions": 1024,
25
+ "num_beams": 4,
26
+ "num_heads": 16,
27
+ "num_layers": 24,
28
+ "relative_attention_num_buckets": 32,
29
+ "reorder_and_upcast_attn": false,
30
+ "resid_pdrop": 0.1,
31
+ "scale_attn_by_inverse_layer_idx": false,
32
+ "scale_attn_weights": true,
33
+ "summary_activation": null,
34
+ "summary_first_dropout": 0.1,
35
+ "summary_proj_to_labels": true,
36
+ "summary_type": "cls_index",
37
+ "summary_use_proj": true,
38
+ "task_specific_params": {
39
+ "text-generation": {
40
+ "do_sample": true,
41
+ "max_length": 50
42
+ },
43
+ "summarization": {
44
+ "early_stopping": true,
45
+ "length_penalty": 2.0,
46
+ "max_length": 200,
47
+ "min_length": 30,
48
+ "no_repeat_ngram_size": 3,
49
+ "num_beams": 4,
50
+ "prefix": "summarize: "
51
+ },
52
+ "translation_en_to_de": {
53
+ "early_stopping": true,
54
+ "max_length": 300,
55
+ "num_beams": 4,
56
+ "prefix": "translate English to German: "
57
+ },
58
+ "translation_en_to_fr": {
59
+ "early_stopping": true,
60
+ "max_length": 300,
61
+ "num_beams": 4,
62
+ "prefix": "translate English to French: "
63
+ },
64
+ "translation_en_to_ro": {
65
+ "early_stopping": true,
66
+ "max_length": 300,
67
+ "num_beams": 4,
68
+ "prefix": "translate English to Romanian: "
69
+ }
70
+ },
71
+ "torch_dtype": "float16",
72
+ "transformers_version": "4.17.0.dev0",
73
+ "use_cache": false,
74
+ "vocab_size": 44876
75
+ }