jsunn-y commited on
Commit
44dd32b
·
verified ·
1 Parent(s): 809f1f2

Upload config.json

Browse files
Files changed (1) hide show
  1. config.json +71 -0
config.json ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "adapter_c_hidden": 16,
4
+ "adapter_c_s": 256,
5
+ "adapter_dropout": 0.1,
6
+ "adapter_input_dim": 2048,
7
+ "adapter_nlayers": 4,
8
+ "adapter_projection_nlayers": 2,
9
+ "adapter_shared_projection": true,
10
+ "adapter_summation": false,
11
+ "adapter_weight_init": 1e-05,
12
+ "architectures": [
13
+ "ProgenConditional"
14
+ ],
15
+ "attn_pdrop": 0.0,
16
+ "bos_token_id": 1,
17
+ "conditions_shared_adapter": false,
18
+ "ec2encoding_file": "data/ec2drfp.pt",
19
+ "embd_pdrop": 0.0,
20
+ "encoding_dimensions": {
21
+ "ec": 630,
22
+ "tax": 4
23
+ },
24
+ "encoding_files": {
25
+ "ec": "data/ec2level4_onehot.pt",
26
+ "tax": "data/tax2level1_onehot.pt"
27
+ },
28
+ "eos_token_id": 2,
29
+ "full_finetuning": false,
30
+ "full_fineturning": false,
31
+ "gradient_checkpointing": true,
32
+ "initializer_range": 0.02,
33
+ "inter_context_sequence_attention": true,
34
+ "inter_sequence_attention": true,
35
+ "layer_norm_epsilon": 1e-05,
36
+ "max_num_sequences": 512,
37
+ "model_type": "progen-conditional",
38
+ "msa_style_attention": true,
39
+ "n_ctx": 2048,
40
+ "n_embd": 1536,
41
+ "n_head": 16,
42
+ "n_inner": null,
43
+ "n_layer": 27,
44
+ "n_positions": 2048,
45
+ "output_router_logits": true,
46
+ "pad_token_id": 0,
47
+ "pretrained_model_dir": "data/pretrained_models/progen2-base",
48
+ "pretrained_model_name": "progen2-base",
49
+ "resid_pdrop": 0.0,
50
+ "rotary_dim": 48,
51
+ "router_aux_loss_coef": 0.05,
52
+ "scale_attn_weights": true,
53
+ "summary_activation": null,
54
+ "summary_first_dropout": 0.1,
55
+ "summary_proj_to_labels": true,
56
+ "summary_type": "cls_index",
57
+ "summary_use_proj": true,
58
+ "task_specific_params": {
59
+ "text-generation": {
60
+ "do_sample": true,
61
+ "max_length": 50,
62
+ "temperature": 1.0
63
+ }
64
+ },
65
+ "tokenizer_class": "GPT2Tokenizer",
66
+ "torch_dtype": "bfloat16",
67
+ "transformers_version": "4.41.2",
68
+ "unified_active_context_pos_ids": true,
69
+ "use_cache": true,
70
+ "vocab_size": 32
71
+ }