Upload config.json with huggingface_hub
Browse files- config.json +29 -0
config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "omnimind-custom-4096x36",
|
| 3 |
+
"size": "micro",
|
| 4 |
+
"d_model": 4096,
|
| 5 |
+
"n_layers": 36,
|
| 6 |
+
"d_state": 512,
|
| 7 |
+
"d_conv": 32,
|
| 8 |
+
"expand": 2,
|
| 9 |
+
"n_heads": 32,
|
| 10 |
+
"vocab_size": 151936,
|
| 11 |
+
"pad_token_id": 0,
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"eos_token_id": 2,
|
| 14 |
+
"max_seq_len": 8192,
|
| 15 |
+
"dropout": 0.1,
|
| 16 |
+
"tie_embeddings": true,
|
| 17 |
+
"dt_rank": 64,
|
| 18 |
+
"dt_min": 0.001,
|
| 19 |
+
"dt_max": 0.1,
|
| 20 |
+
"dt_init": "random",
|
| 21 |
+
"dt_scale": 1.0,
|
| 22 |
+
"dt_init_floor": 0.0001,
|
| 23 |
+
"rope_theta": 10000.0,
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"architectures": [
|
| 26 |
+
"OmnimindForCausalLM"
|
| 27 |
+
],
|
| 28 |
+
"model_type": "omnimind"
|
| 29 |
+
}
|