Create config.json
Browse files- config.json +36 -0
config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Grok1ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"embedding_multiplier_scale": 90.50966799187809,
|
| 6 |
+
"output_multiplier_scale": 0.5,
|
| 7 |
+
"vocab_size": 131072,
|
| 8 |
+
"hidden_size": 8192,
|
| 9 |
+
"intermediate_size": 32768,
|
| 10 |
+
"moe_intermediate_size": 16384,
|
| 11 |
+
"max_position_embeddings": 131072,
|
| 12 |
+
"num_experts_per_tok": 2,
|
| 13 |
+
"num_local_experts": 8,
|
| 14 |
+
"residual_moe": true,
|
| 15 |
+
"num_attention_heads": 64,
|
| 16 |
+
"num_key_value_heads": 8,
|
| 17 |
+
"num_hidden_layers": 64,
|
| 18 |
+
"head_dim": 128,
|
| 19 |
+
"rms_norm_eps": 1e-05,
|
| 20 |
+
"final_logit_softcapping": 50,
|
| 21 |
+
"attn_logit_softcapping": 30.0,
|
| 22 |
+
"router_logit_softcapping": 30.0,
|
| 23 |
+
"rope_theta": 208533496,
|
| 24 |
+
"attn_temperature_len": 1024,
|
| 25 |
+
"sliding_window_size": -1,
|
| 26 |
+
"global_attn_every_n": 1,
|
| 27 |
+
"model_type": "git",
|
| 28 |
+
"torch_dtype": "bfloat16",
|
| 29 |
+
"rope_type": "original",
|
| 30 |
+
"original_max_position_embeddings": 8192,
|
| 31 |
+
"scaling_factor": 16.0,
|
| 32 |
+
"extrapolation_factor": 1.0,
|
| 33 |
+
"attn_factor": 1.0,
|
| 34 |
+
"beta_fast": 8,
|
| 35 |
+
"beta_slow": 1
|
| 36 |
+
}
|