Commit
·
67fb92e
1
Parent(s):
eca9f18
Update config.json
Browse files- config.json +7 -3
config.json
CHANGED
@@ -6,7 +6,7 @@
|
|
6 |
"AutoModelForCausalLM": "modeling_openmoe.OpenMoeForCausalLM"
|
7 |
},
|
8 |
"attention_bias": false,
|
9 |
-
"bos_token_id":
|
10 |
"dropout_rate": 0.0,
|
11 |
"enable_comm_overlap": false,
|
12 |
"enable_hierarchical_alltoall": false,
|
@@ -19,6 +19,7 @@
|
|
19 |
"hidden_size": 2048,
|
20 |
"initializer_range": 0.02,
|
21 |
"intermediate_size": 8192,
|
|
|
22 |
"layer_norm_epsilon": 1e-06,
|
23 |
"load_balance_beam_width": 8,
|
24 |
"load_balance_group_swap_factor": 0.4,
|
@@ -36,15 +37,18 @@
|
|
36 |
"rms_norm_eps": 1e-06,
|
37 |
"rope_scaling": null,
|
38 |
"rope_theta": 10000.0,
|
|
|
39 |
"router_capacity_factor_eval": 2.0,
|
40 |
"router_capacity_factor_train": 1.25,
|
41 |
"router_drop_tks": true,
|
42 |
"router_min_capacity": 4,
|
43 |
"router_noisy_policy": null,
|
44 |
"router_topk": 2,
|
|
|
45 |
"tie_word_embeddings": false,
|
46 |
"torch_dtype": "float32",
|
47 |
"transformers_version": "4.34.0",
|
48 |
"use_cache": true,
|
49 |
-
"vocab_size": 256384
|
50 |
-
|
|
|
|
6 |
"AutoModelForCausalLM": "modeling_openmoe.OpenMoeForCausalLM"
|
7 |
},
|
8 |
"attention_bias": false,
|
9 |
+
"bos_token_id": 0,
|
10 |
"dropout_rate": 0.0,
|
11 |
"enable_comm_overlap": false,
|
12 |
"enable_hierarchical_alltoall": false,
|
|
|
19 |
"hidden_size": 2048,
|
20 |
"initializer_range": 0.02,
|
21 |
"intermediate_size": 8192,
|
22 |
+
"label_smoothing": 0.001,
|
23 |
"layer_norm_epsilon": 1e-06,
|
24 |
"load_balance_beam_width": 8,
|
25 |
"load_balance_group_swap_factor": 0.4,
|
|
|
37 |
"rms_norm_eps": 1e-06,
|
38 |
"rope_scaling": null,
|
39 |
"rope_theta": 10000.0,
|
40 |
+
"router_aux_loss_factor": 0.01,
|
41 |
"router_capacity_factor_eval": 2.0,
|
42 |
"router_capacity_factor_train": 1.25,
|
43 |
"router_drop_tks": true,
|
44 |
"router_min_capacity": 4,
|
45 |
"router_noisy_policy": null,
|
46 |
"router_topk": 2,
|
47 |
+
"router_z_loss_factor": 0.0001,
|
48 |
"tie_word_embeddings": false,
|
49 |
"torch_dtype": "float32",
|
50 |
"transformers_version": "4.34.0",
|
51 |
"use_cache": true,
|
52 |
+
"vocab_size": 256384,
|
53 |
+
"z_loss_factor": 0.01
|
54 |
+
}
|