mini-beatrix-2s automodel: config.json (mission final 16.101B, alephllm 0.8.6)
Browse files- config.json +53 -0
config.json
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "mini-beatrix-2s",
|
| 3 |
+
"d_model": 1024,
|
| 4 |
+
"n_layers": 20,
|
| 5 |
+
"n_heads": 16,
|
| 6 |
+
"context": 4096,
|
| 7 |
+
"vocab_size": 256,
|
| 8 |
+
"tokenizer": "byte-trigram",
|
| 9 |
+
"hub_layers": [
|
| 10 |
+
0,
|
| 11 |
+
1,
|
| 12 |
+
2,
|
| 13 |
+
3,
|
| 14 |
+
4,
|
| 15 |
+
5,
|
| 16 |
+
6,
|
| 17 |
+
7,
|
| 18 |
+
8,
|
| 19 |
+
9,
|
| 20 |
+
10,
|
| 21 |
+
11,
|
| 22 |
+
12,
|
| 23 |
+
13,
|
| 24 |
+
14,
|
| 25 |
+
15,
|
| 26 |
+
16,
|
| 27 |
+
17,
|
| 28 |
+
18,
|
| 29 |
+
19
|
| 30 |
+
],
|
| 31 |
+
"hub_K": 64,
|
| 32 |
+
"hub_D": 128,
|
| 33 |
+
"tau": 0.1,
|
| 34 |
+
"bank_experts": 3,
|
| 35 |
+
"bank_ff": 1024,
|
| 36 |
+
"head_K": 256,
|
| 37 |
+
"head_D": 256,
|
| 38 |
+
"gate_init": -3.0,
|
| 39 |
+
"tie_embeddings": false,
|
| 40 |
+
"hub_chunk": 256,
|
| 41 |
+
"hub_const": 4,
|
| 42 |
+
"hub_ckpt": 0,
|
| 43 |
+
"model_type": "mini-beatrix",
|
| 44 |
+
"architectures": [
|
| 45 |
+
"MiniBeatrixForCausalLM"
|
| 46 |
+
],
|
| 47 |
+
"auto_map": {
|
| 48 |
+
"AutoConfig": "modeling_minibeatrix.MiniBeatrixConfig",
|
| 49 |
+
"AutoModelForCausalLM": "modeling_minibeatrix.MiniBeatrixForCausalLM"
|
| 50 |
+
},
|
| 51 |
+
"torch_dtype": "float32",
|
| 52 |
+
"use_cache": false
|
| 53 |
+
}
|