Upload config.json with huggingface_hub
Browse files- config.json +36 -0
config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "tinyv4",
|
| 3 |
+
"architectures": ["TinyV4"],
|
| 4 |
+
"auto_map": {
|
| 5 |
+
"AutoModel": "modeling_tinyv4.TinyV4",
|
| 6 |
+
"AutoModelForCausalLM": "modeling_tinyv4.TinyV4"
|
| 7 |
+
},
|
| 8 |
+
"vocab_size": 32000,
|
| 9 |
+
"dim": 128,
|
| 10 |
+
"depth": 6,
|
| 11 |
+
"n_hc": 2,
|
| 12 |
+
"n_routed": 4,
|
| 13 |
+
"n_active": 2,
|
| 14 |
+
"n_shared": 1,
|
| 15 |
+
"expert_intermediate": 192,
|
| 16 |
+
"csa_m": 4,
|
| 17 |
+
"csa_topk": 16,
|
| 18 |
+
"hca_m": 8,
|
| 19 |
+
"n_win": 16,
|
| 20 |
+
"n_q_head": 4,
|
| 21 |
+
"head_dim": 48,
|
| 22 |
+
"d_c": 64,
|
| 23 |
+
"n_idx_head": 4,
|
| 24 |
+
"idx_head_dim": 48,
|
| 25 |
+
"n_out_group": 2,
|
| 26 |
+
"d_g": 64,
|
| 27 |
+
"rope_dim": 24,
|
| 28 |
+
"mtp_depth": 1,
|
| 29 |
+
"hash_layers": 2,
|
| 30 |
+
"max_len": 512,
|
| 31 |
+
"sinkhorn_iters": 20,
|
| 32 |
+
"aux_bias_update": 0.001,
|
| 33 |
+
"bal_loss_weight": 0.0001,
|
| 34 |
+
"step": 5000,
|
| 35 |
+
"best_loss": 3.9723572731018066
|
| 36 |
+
}
|