Update ant-v5 bundle (50ep eval: mean 1914 -> 2265, median 1562 -> 1969; context_length 24 -> 16)
Browse files- ant-v5/config.json +4 -3
- ant-v5/model.safetensors +1 -1
ant-v5/config.json
CHANGED
|
@@ -1,5 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"bundle_format_version": 1,
|
|
|
|
| 3 |
"model_config": {
|
| 4 |
"network_name": "Llama",
|
| 5 |
"num_layers": 4,
|
|
@@ -9,8 +10,8 @@
|
|
| 9 |
"init_log_std": -1.0,
|
| 10 |
"rope_theta": 1000.0,
|
| 11 |
"intermediate_size": 768,
|
| 12 |
-
"max_position_embeddings":
|
| 13 |
-
"context_length":
|
| 14 |
},
|
| 15 |
"state_specs": [
|
| 16 |
{
|
|
@@ -262,5 +263,5 @@
|
|
| 262 |
"squash": "tanh"
|
| 263 |
}
|
| 264 |
],
|
| 265 |
-
"context_length":
|
| 266 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"bundle_format_version": 1,
|
| 3 |
+
"package_version": "0.1.0",
|
| 4 |
"model_config": {
|
| 5 |
"network_name": "Llama",
|
| 6 |
"num_layers": 4,
|
|
|
|
| 10 |
"init_log_std": -1.0,
|
| 11 |
"rope_theta": 1000.0,
|
| 12 |
"intermediate_size": 768,
|
| 13 |
+
"max_position_embeddings": 128,
|
| 14 |
+
"context_length": 16
|
| 15 |
},
|
| 16 |
"state_specs": [
|
| 17 |
{
|
|
|
|
| 263 |
"squash": "tanh"
|
| 264 |
}
|
| 265 |
],
|
| 266 |
+
"context_length": 16
|
| 267 |
}
|
ant-v5/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9552252
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06ae357134b2dd44249853764951e0693721d5b6a25feaa21f53d46a290dddb5
|
| 3 |
size 9552252
|