elephantmipt commited on
Commit
0c453a2
·
verified ·
1 Parent(s): d95957a

Upload 2 files

Browse files
Files changed (2) hide show
  1. config.json +75 -0
  2. model.safetensors +3 -0
config.json ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "swiglu",
3
+ "alt_mixer": {
4
+ "_target_": "aicl.model.models.mixers.linear_attention.LinearAttention",
5
+ "feature_dim": 16,
6
+ "feature_map": {
7
+ "_target_": "aicl.model.models.mixers.linear_attention.Squared",
8
+ "input_dim": 16
9
+ },
10
+ "l_max": 2048,
11
+ "num_heads": 12
12
+ },
13
+ "alt_mixer_2_layers": [
14
+ 2,
15
+ 7,
16
+ 12,
17
+ 17,
18
+ 22
19
+ ],
20
+ "alt_mixer_layers": [
21
+ 1,
22
+ 3,
23
+ 5,
24
+ 7,
25
+ 9,
26
+ 11,
27
+ 13,
28
+ 15,
29
+ 17
30
+ ],
31
+ "attn_pdrop": 0,
32
+ "bos_token_id": 50256,
33
+ "embd_pdrop": 0,
34
+ "eos_token_id": 50256,
35
+ "fused_bias_fc": true,
36
+ "fused_dropout_add_ln": true,
37
+ "fused_mlp": false,
38
+ "initializer_range": 0.02,
39
+ "layer_norm_epsilon": 1e-05,
40
+ "mixer": {
41
+ "_target_": "aicl.model.models.mixers.convolution.BaseConv",
42
+ "expand_proj": 4,
43
+ "kernel_sizes": 3,
44
+ "l_max": 2048,
45
+ "use_bias": true
46
+ },
47
+ "mlp_fc1_bias": false,
48
+ "mlp_fc2_bias": false,
49
+ "model_type": "gpt2",
50
+ "n_embd": 768,
51
+ "n_head": 12,
52
+ "n_inner": 1536,
53
+ "n_layer": 18,
54
+ "n_positions": 0,
55
+ "out_proj_bias": false,
56
+ "pad_vocab_size_multiple": 8,
57
+ "qkv_proj_bias": false,
58
+ "reorder_and_upcast_attn": false,
59
+ "resid_pdrop": 0,
60
+ "residual_in_fp32": true,
61
+ "rms_norm": true,
62
+ "rotary_emb_fraction": 1,
63
+ "scale_attn_by_inverse_layer_idx": false,
64
+ "scale_attn_weights": true,
65
+ "special_initializer": true,
66
+ "summary_activation": null,
67
+ "summary_first_dropout": 0.1,
68
+ "summary_proj_to_labels": true,
69
+ "summary_type": "cls_index",
70
+ "summary_use_proj": true,
71
+ "transformers_version": "4.38.2",
72
+ "use_cache": true,
73
+ "use_flash_attn": true,
74
+ "vocab_size": 50277
75
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83cdb7e0794be4fa39a2ad2ed621a3e57a33c5fc651985cf6734fe7f1e244289
3
+ size 587769552