Upload LoRA adapter (README written by author)
Browse files- README.md +1 -1
- adapter_config.json +5 -5
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -35,7 +35,7 @@ Loss is applied only to the final assistant output.
|
|
| 35 |
|
| 36 |
- Base model: Qwen/Qwen3-4B-Instruct-2507
|
| 37 |
- Method: QLoRA (4-bit)
|
| 38 |
-
- Max sequence length:
|
| 39 |
- Epochs: 1
|
| 40 |
- Learning rate: 2e-05
|
| 41 |
- LoRA: r=64, alpha=128
|
|
|
|
| 35 |
|
| 36 |
- Base model: Qwen/Qwen3-4B-Instruct-2507
|
| 37 |
- Method: QLoRA (4-bit)
|
| 38 |
+
- Max sequence length: 1024
|
| 39 |
- Epochs: 1
|
| 40 |
- Learning rate: 2e-05
|
| 41 |
- LoRA: r=64, alpha=128
|
adapter_config.json
CHANGED
|
@@ -33,13 +33,13 @@
|
|
| 33 |
"rank_pattern": {},
|
| 34 |
"revision": null,
|
| 35 |
"target_modules": [
|
| 36 |
-
"gate_proj",
|
| 37 |
-
"k_proj",
|
| 38 |
-
"down_proj",
|
| 39 |
-
"v_proj",
|
| 40 |
"up_proj",
|
|
|
|
|
|
|
|
|
|
| 41 |
"o_proj",
|
| 42 |
-
"
|
|
|
|
| 43 |
],
|
| 44 |
"target_parameters": null,
|
| 45 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 33 |
"rank_pattern": {},
|
| 34 |
"revision": null,
|
| 35 |
"target_modules": [
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
"up_proj",
|
| 37 |
+
"q_proj",
|
| 38 |
+
"v_proj",
|
| 39 |
+
"down_proj",
|
| 40 |
"o_proj",
|
| 41 |
+
"gate_proj",
|
| 42 |
+
"k_proj"
|
| 43 |
],
|
| 44 |
"target_parameters": null,
|
| 45 |
"task_type": "CAUSAL_LM",
|
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 sha256:110c38c23048a5b4fd190382fc7822f464c704fa2ac88201fd01b4eca23199d3
|
| 3 |
size 528550256
|