Hugging Face's logo

jacobfa1
/
attractor-370m

Text Generation
Model card Files
xet
 Community
1
attractor-370m
File size: 1,575 Bytes
84feedc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d8850da
84feedc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
{
  "name": "eqlm-medium-370m",
  "hf_config": {
    "name": "eqlm-medium-370m"
  },
  "block_size": 2048,
  "n_embd": 1280,
  "intermediate_size": 5120,
  "num_attention_heads": 10,
  "num_key_value_heads": 10,
  "vocab_size": 32768,
  "padding_multiple": 64,
  "padded_vocab_size": 32768,
  "rope_settings": {
    "use_rope": true,
    "rope_condense_ratio": 1,
    "rope_base": 50000
  },
  "use_abacus": false,
  "randomize_positions_from": null,
  "block_class_name": "TransformerPreNormBlock",
  "norm_class_name": "RMSNorm",
  "attn_impl": "flash",
  "norm_eps": 1e-05,
  "mlp_class_name": "BaseMLP",
  "nonlin_name": "ReLU2",
  "bias": false,
  "qk_bias": false,
  "init_strategy": "scaled-zero",
  "init_orthogonal": true,
  "skip_initialization": false,
  "mup_model_scaling_factor": 1,
  "use_fused_head": "pytorch",
  "debias_attention": false,
  "center_attention": false,
  "clip_qkv": null,
  "qk_norm": true,
  "logit_softcap": null,
  "causal": true,
  "activation_checkpoint_impl": "per-block",
  "simple_ops": false,
  "strategy": "ddp",
  "n_backbone_layers": 15,
  "n_fp_blocks": 2,
  "tie_embeddings": true,
  "solver": "anderson",
  "max_iter": 64,
  "min_iter": 6,
  "tol": 0.0002,
  "anderson_m": 5,
  "anderson_beta": 1.0,
  "backward_type": "onestep",
  "backward_max_iter": 64,
  "backward_min_iter": 6,
  "backward_tol": 0.0002,
  "adjoint_grad_clip": null,
  "layer_scale_init": 0.75,
  "gamma_max": 1.0,
  "fp_lr_scale": 0.4,
  "fp_wd": 0.1,
  "recurrent_embedding_dimension": 1280,
  "model_class_name": "EQLM",
  "_class_name": "EQLMConfig"
}