smallcloudteam commited on
Commit
17db844
·
1 Parent(s): 9a6d102

Upload CodifyForCausalLM

Browse files
Files changed (2) hide show
  1. config.json +59 -0
  2. pytorch_model.bin +3 -0
config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "E": 1024,
3
+ "L": 24,
4
+ "T": 2048,
5
+ "_mup": true,
6
+ "alt_pw_klass": {
7
+ "type": ""
8
+ },
9
+ "alt_rel_klass": {
10
+ "fused": true,
11
+ "type": "alibi"
12
+ },
13
+ "alt_sa_klass": {
14
+ "triton": true,
15
+ "type": "flash",
16
+ "use_rotary_emb": null
17
+ },
18
+ "architectures": [
19
+ "CodifyForCausalLM"
20
+ ],
21
+ "attn_a_reach": 2048,
22
+ "attn_b_reach": 2048,
23
+ "attn_heads": 32,
24
+ "attn_ra_nbasis": 64,
25
+ "attn_seq": [
26
+ "d"
27
+ ],
28
+ "attn_sparse_layout_seq": null,
29
+ "backcheck_pw": "none",
30
+ "backcheck_sa": "none",
31
+ "bos_token_id": 1,
32
+ "dtype_acts": "torch.float16",
33
+ "dtype_weights": "torch.float16",
34
+ "enc_name": "openai_programming_v2",
35
+ "eos_token_id": 2,
36
+ "init_scale": 1,
37
+ "initializer_range": 0.02,
38
+ "layer_norm_epsilon": 1e-05,
39
+ "mlp_mult": 4,
40
+ "model_type": "codify",
41
+ "moe": null,
42
+ "mup_optimal_lr": 0.0005,
43
+ "mup_shapes_file": "lean_former/mup/flash_rot1d_24l/shapes.json",
44
+ "posemb": false,
45
+ "rescale_embeddings": false,
46
+ "tie_word_embeddings": false,
47
+ "torch_dtype": "float32",
48
+ "transformers_version": "4.24.0",
49
+ "tune": [
50
+ 3,
51
+ 3,
52
+ 3,
53
+ 3
54
+ ],
55
+ "unembedding_shared": false,
56
+ "use_cache": true,
57
+ "use_res_scale": false,
58
+ "vocab_size": 51305
59
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7947369f9ccd5e089e37d3c3b8d026c4f00ee70b894fca56ce72bac27f635bd
3
+ size 1629628537