File size: 1,860 Bytes
44dd32b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
{
  "activation_function": "gelu_new",
  "adapter_c_hidden": 16,
  "adapter_c_s": 256,
  "adapter_dropout": 0.1,
  "adapter_input_dim": 2048,
  "adapter_nlayers": 4,
  "adapter_projection_nlayers": 2,
  "adapter_shared_projection": true,
  "adapter_summation": false,
  "adapter_weight_init": 1e-05,
  "architectures": [
    "ProgenConditional"
  ],
  "attn_pdrop": 0.0,
  "bos_token_id": 1,
  "conditions_shared_adapter": false,
  "ec2encoding_file": "data/ec2drfp.pt",
  "embd_pdrop": 0.0,
  "encoding_dimensions": {
    "ec": 630,
    "tax": 4
  },
  "encoding_files": {
    "ec": "data/ec2level4_onehot.pt",
    "tax": "data/tax2level1_onehot.pt"
  },
  "eos_token_id": 2,
  "full_finetuning": false,
  "full_fineturning": false,
  "gradient_checkpointing": true,
  "initializer_range": 0.02,
  "inter_context_sequence_attention": true,
  "inter_sequence_attention": true,
  "layer_norm_epsilon": 1e-05,
  "max_num_sequences": 512,
  "model_type": "progen-conditional",
  "msa_style_attention": true,
  "n_ctx": 2048,
  "n_embd": 1536,
  "n_head": 16,
  "n_inner": null,
  "n_layer": 27,
  "n_positions": 2048,
  "output_router_logits": true,
  "pad_token_id": 0,
  "pretrained_model_dir": "data/pretrained_models/progen2-base",
  "pretrained_model_name": "progen2-base",
  "resid_pdrop": 0.0,
  "rotary_dim": 48,
  "router_aux_loss_coef": 0.05,
  "scale_attn_weights": true,
  "summary_activation": null,
  "summary_first_dropout": 0.1,
  "summary_proj_to_labels": true,
  "summary_type": "cls_index",
  "summary_use_proj": true,
  "task_specific_params": {
    "text-generation": {
      "do_sample": true,
      "max_length": 50,
      "temperature": 1.0
    }
  },
  "tokenizer_class": "GPT2Tokenizer",
  "torch_dtype": "bfloat16",
  "transformers_version": "4.41.2",
  "unified_active_context_pos_ids": true,
  "use_cache": true,
  "vocab_size": 32
}