Upload config.json
Browse files- config.json +39 -0
config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "zdecode",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ZDECODE"
|
| 5 |
+
],
|
| 6 |
+
"profile_name": "1b_3e_8l_t4x2",
|
| 7 |
+
"vocab_size": 200024,
|
| 8 |
+
"text_embed_dim": 1024,
|
| 9 |
+
"vision_embed_dim": 1024,
|
| 10 |
+
"hidden_dim": 1024,
|
| 11 |
+
"ffn_dim": 6144,
|
| 12 |
+
"num_layers": 8,
|
| 13 |
+
"num_heads": 16,
|
| 14 |
+
"num_kv_heads": 4,
|
| 15 |
+
"num_experts": 3,
|
| 16 |
+
"top_k": 2,
|
| 17 |
+
"max_position_embeddings": 16384,
|
| 18 |
+
"router_aux_loss_coef": 0.01,
|
| 19 |
+
"share_experts_across_layers": false,
|
| 20 |
+
"gradient_checkpointing": true,
|
| 21 |
+
"num_agents": 4,
|
| 22 |
+
"moe_capacity_factor": 1.0,
|
| 23 |
+
"moe_hierarchy_groups": 1,
|
| 24 |
+
"moe_hierarchy_top_k": 1,
|
| 25 |
+
"num_shared_experts": 0,
|
| 26 |
+
"load_balancing_mode": "aux_free",
|
| 27 |
+
"router_bias_update_rate": 0.01,
|
| 28 |
+
"kv_latent_dim": 128,
|
| 29 |
+
"kv_cache_dtype": "int4",
|
| 30 |
+
"rope_training_context": 16384,
|
| 31 |
+
"rope_ntk_alpha": 1.0,
|
| 32 |
+
"rope_yarn_scale": 1.0,
|
| 33 |
+
"ring_attention_chunk_size": 0,
|
| 34 |
+
"prefill_chunk_size": 256,
|
| 35 |
+
"use_q_former_projector": true,
|
| 36 |
+
"q_former_queries": 8,
|
| 37 |
+
"q_former_layers": 1,
|
| 38 |
+
"tokenizer_name": "ai-tokenizer:GPT-5"
|
| 39 |
+
}
|