Upload MoAMetricLM
Browse files- config.json +4 -4
- pytorch_model.bin +2 -2
config.json
CHANGED
|
@@ -18,7 +18,7 @@
|
|
| 18 |
"eos_token_id": 151643,
|
| 19 |
"ff_mult": 4,
|
| 20 |
"ffn_hidden": 2048,
|
| 21 |
-
"head_feature_heads":
|
| 22 |
"layer_scale_init_value": 0.0001,
|
| 23 |
"learn_alpha": true,
|
| 24 |
"learn_radius": true,
|
|
@@ -37,16 +37,16 @@
|
|
| 37 |
"origin_init_scale": 0.0,
|
| 38 |
"pad_token_id": 151643,
|
| 39 |
"proj_drop": 0.1,
|
| 40 |
-
"r_basis":
|
| 41 |
"radius_init": 3.0,
|
| 42 |
"router_bias_heads": 4,
|
| 43 |
"router_dropout": 0.1,
|
| 44 |
"router_hidden": 2048,
|
| 45 |
"router_init_temperature": 2.0,
|
| 46 |
-
"router_temperature":
|
| 47 |
"router_topk": 3,
|
| 48 |
"shared_kv_ratio": 0.55,
|
| 49 |
-
"theta_base":
|
| 50 |
"ti_reg_samples": 0,
|
| 51 |
"ti_reg_weight": 0.0,
|
| 52 |
"transformers_version": "4.56.1",
|
|
|
|
| 18 |
"eos_token_id": 151643,
|
| 19 |
"ff_mult": 4,
|
| 20 |
"ffn_hidden": 2048,
|
| 21 |
+
"head_feature_heads": 32,
|
| 22 |
"layer_scale_init_value": 0.0001,
|
| 23 |
"learn_alpha": true,
|
| 24 |
"learn_radius": true,
|
|
|
|
| 37 |
"origin_init_scale": 0.0,
|
| 38 |
"pad_token_id": 151643,
|
| 39 |
"proj_drop": 0.1,
|
| 40 |
+
"r_basis": 4,
|
| 41 |
"radius_init": 3.0,
|
| 42 |
"router_bias_heads": 4,
|
| 43 |
"router_dropout": 0.1,
|
| 44 |
"router_hidden": 2048,
|
| 45 |
"router_init_temperature": 2.0,
|
| 46 |
+
"router_temperature": 2.0,
|
| 47 |
"router_topk": 3,
|
| 48 |
"shared_kv_ratio": 0.55,
|
| 49 |
+
"theta_base": 100000.0,
|
| 50 |
"ti_reg_samples": 0,
|
| 51 |
"ti_reg_weight": 0.0,
|
| 52 |
"transformers_version": "4.56.1",
|
pytorch_model.bin
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:b4eab1a92049e5fe16817ed4c13ff98ec452f76e5fbc5af6be3df94f9a812184
|
| 3 |
+
size 1712299451
|