Upload LoRA adapter (README written by author)
Browse files- README.md +2 -2
- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -33,9 +33,9 @@ while intermediate reasoning (Chain-of-Thought) is masked.
|
|
| 33 |
|
| 34 |
- Base model: Qwen/Qwen3-4B-Instruct-2507
|
| 35 |
- Method: QLoRA (4-bit)
|
| 36 |
-
- Max sequence length:
|
| 37 |
- Epochs: 3
|
| 38 |
-
- Learning rate:
|
| 39 |
- LoRA: r=64, alpha=128
|
| 40 |
|
| 41 |
## Usage
|
|
|
|
| 33 |
|
| 34 |
- Base model: Qwen/Qwen3-4B-Instruct-2507
|
| 35 |
- Method: QLoRA (4-bit)
|
| 36 |
+
- Max sequence length: 512
|
| 37 |
- Epochs: 3
|
| 38 |
+
- Learning rate: 1e-06
|
| 39 |
- LoRA: r=64, alpha=128
|
| 40 |
|
| 41 |
## Usage
|
adapter_config.json
CHANGED
|
@@ -33,12 +33,12 @@
|
|
| 33 |
"rank_pattern": {},
|
| 34 |
"revision": null,
|
| 35 |
"target_modules": [
|
|
|
|
|
|
|
|
|
|
| 36 |
"down_proj",
|
| 37 |
"up_proj",
|
| 38 |
"k_proj",
|
| 39 |
-
"o_proj",
|
| 40 |
-
"gate_proj",
|
| 41 |
-
"q_proj",
|
| 42 |
"v_proj"
|
| 43 |
],
|
| 44 |
"target_parameters": null,
|
|
|
|
| 33 |
"rank_pattern": {},
|
| 34 |
"revision": null,
|
| 35 |
"target_modules": [
|
| 36 |
+
"q_proj",
|
| 37 |
+
"o_proj",
|
| 38 |
+
"gate_proj",
|
| 39 |
"down_proj",
|
| 40 |
"up_proj",
|
| 41 |
"k_proj",
|
|
|
|
|
|
|
|
|
|
| 42 |
"v_proj"
|
| 43 |
],
|
| 44 |
"target_parameters": null,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 528550256
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6ba843da6ab6bf5f7f3c2a581f8b20dbf3081fa94f070642fa69348927ba0613
|
| 3 |
size 528550256
|