Upload merged Qwen3-4B-Instruct-2507 model (auto-generated README)
Browse files- README.md +3 -2
- model-00001-of-00002.safetensors +1 -1
- model-00002-of-00002.safetensors +1 -1
README.md
CHANGED
|
@@ -17,6 +17,7 @@ tags:
|
|
| 17 |
- agentbench
|
| 18 |
- alfworld
|
| 19 |
- dbbench
|
|
|
|
| 20 |
---
|
| 21 |
|
| 22 |
# qwen3-4b-agentbench-dbalf-lora
|
|
@@ -48,12 +49,12 @@ Loss is applied to **all assistant turns** in the trajectory, enabling the model
|
|
| 48 |
- Mixing ratio (pre-merge target): **DB:ALF = 1:0**
|
| 49 |
|
| 50 |
### DB Oversampling (category-aware)
|
| 51 |
-
Enabled: **
|
| 52 |
|
| 53 |
DB category weights used during training-data preparation:
|
| 54 |
|
| 55 |
- counting: 2
|
| 56 |
-
- comparison:
|
| 57 |
- ranking: 1
|
| 58 |
- select: 1
|
| 59 |
- insert: 1
|
|
|
|
| 17 |
- agentbench
|
| 18 |
- alfworld
|
| 19 |
- dbbench
|
| 20 |
+
- db-oversampling
|
| 21 |
---
|
| 22 |
|
| 23 |
# qwen3-4b-agentbench-dbalf-lora
|
|
|
|
| 49 |
- Mixing ratio (pre-merge target): **DB:ALF = 1:0**
|
| 50 |
|
| 51 |
### DB Oversampling (category-aware)
|
| 52 |
+
Enabled: **True**
|
| 53 |
|
| 54 |
DB category weights used during training-data preparation:
|
| 55 |
|
| 56 |
- counting: 2
|
| 57 |
+
- comparison: 1
|
| 58 |
- ranking: 1
|
| 59 |
- select: 1
|
| 60 |
- insert: 1
|
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4967215360
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f9583b85b53dfc82008b452debc3e07462a46b8910ed0a96feb940bf763365a
|
| 3 |
size 4967215360
|
model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3077766632
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ddc57f0db9cec8c8721aa56c73ea24b7b110b549645c345745f934d1b059aaba
|
| 3 |
size 3077766632
|