Upload MoAMetricLM
Browse files- config.json +3 -4
- model.safetensors +2 -2
config.json
CHANGED
|
@@ -35,7 +35,7 @@
|
|
| 35 |
"lm_proj_drop": 0.1,
|
| 36 |
"lm_router_dropout": 0.1,
|
| 37 |
"lm_router_hidden": 64,
|
| 38 |
-
"lm_router_temperature": 1.
|
| 39 |
"lr_rank": 32,
|
| 40 |
"maha_init": 1.0,
|
| 41 |
"max_position_embeddings": 8192,
|
|
@@ -53,10 +53,10 @@
|
|
| 53 |
"pad_token_id": 0,
|
| 54 |
"proj_drop": 0.1,
|
| 55 |
"r_basis": 16,
|
| 56 |
-
"radius_init":
|
| 57 |
"router_dropout": 0.1,
|
| 58 |
"router_hidden": 128,
|
| 59 |
-
"router_temperature": 1.
|
| 60 |
"router_topk": 2,
|
| 61 |
"theta_base": 10000.0,
|
| 62 |
"ti_reg_samples": 64,
|
|
@@ -64,7 +64,6 @@
|
|
| 64 |
"tie_word_embeddings": true,
|
| 65 |
"transformers_version": "5.0.0",
|
| 66 |
"use_balls": false,
|
| 67 |
-
"use_cache": false,
|
| 68 |
"vocab_size": 50277,
|
| 69 |
"window_size": 512
|
| 70 |
}
|
|
|
|
| 35 |
"lm_proj_drop": 0.1,
|
| 36 |
"lm_router_dropout": 0.1,
|
| 37 |
"lm_router_hidden": 64,
|
| 38 |
+
"lm_router_temperature": 1.0,
|
| 39 |
"lr_rank": 32,
|
| 40 |
"maha_init": 1.0,
|
| 41 |
"max_position_embeddings": 8192,
|
|
|
|
| 53 |
"pad_token_id": 0,
|
| 54 |
"proj_drop": 0.1,
|
| 55 |
"r_basis": 16,
|
| 56 |
+
"radius_init": 5.0,
|
| 57 |
"router_dropout": 0.1,
|
| 58 |
"router_hidden": 128,
|
| 59 |
+
"router_temperature": 1.25,
|
| 60 |
"router_topk": 2,
|
| 61 |
"theta_base": 10000.0,
|
| 62 |
"ti_reg_samples": 64,
|
|
|
|
| 64 |
"tie_word_embeddings": true,
|
| 65 |
"transformers_version": "5.0.0",
|
| 66 |
"use_balls": false,
|
|
|
|
| 67 |
"vocab_size": 50277,
|
| 68 |
"window_size": 512
|
| 69 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:13e7d584efd55a94ba084640d9f61ccdfa1dcce3375b825c43c17fc9789c1b5a
|
| 3 |
+
size 218656860
|