Text Generation
PEFT
Safetensors
lora
qlora
behavioral-evaluation
llm
post-training
model-evaluation
qwen
synthetic-data
Instructions to use aamish-ahmad/behaviortune-v1-1-r1-adapter with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use aamish-ahmad/behaviortune-v1-1-r1-adapter with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B-Instruct-2507") model = PeftModel.from_pretrained(base_model, "aamish-ahmad/behaviortune-v1-1-r1-adapter") - Notebooks
- Google Colab
- Kaggle
root commited on
Commit ·
3322230
1
Parent(s): 89daa76
Polish adapter presentation metadata
Browse files- PROVENANCE_AND_HASHES.json +15 -2
- README.md +7 -1
PROVENANCE_AND_HASHES.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"schema_version": 1,
|
| 3 |
"artifact": "BehaviorTune V1.1-R1 public release provenance",
|
| 4 |
-
"status": "
|
| 5 |
"source_commit": "6ef019e996982417df50127a5c351f179b75ea6b",
|
| 6 |
"release_tag": "v1.0.0",
|
| 7 |
"public_targets": {
|
|
@@ -10,9 +10,22 @@
|
|
| 10 |
"adapter": "https://huggingface.co/aamish-ahmad/behaviortune-v1-1-r1-adapter"
|
| 11 |
},
|
| 12 |
"publications": {
|
| 13 |
-
"
|
|
|
|
|
|
|
| 14 |
"dataset_content_revision": "1cafbc1f9cf421d43446c385545319a3b0a0a3eb",
|
|
|
|
| 15 |
"adapter_content_revision": "9b4095666621104a94ce2193e62de363e35cba75",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
"dataset_public": true,
|
| 17 |
"adapter_public": true,
|
| 18 |
"published_files_match_local_git_blobs": true,
|
|
|
|
| 1 |
{
|
| 2 |
"schema_version": 1,
|
| 3 |
"artifact": "BehaviorTune V1.1-R1 public release provenance",
|
| 4 |
+
"status": "PUBLIC_RECRUITER_SURFACE_VERIFIED",
|
| 5 |
"source_commit": "6ef019e996982417df50127a5c351f179b75ea6b",
|
| 6 |
"release_tag": "v1.0.0",
|
| 7 |
"public_targets": {
|
|
|
|
| 10 |
"adapter": "https://huggingface.co/aamish-ahmad/behaviortune-v1-1-r1-adapter"
|
| 11 |
},
|
| 12 |
"publications": {
|
| 13 |
+
"initial_github_publication_verified_commit": "276761915dcd8555a3303d074e6adda62afb865f",
|
| 14 |
+
"github_functional_surface_commit": "35ec8bd5736aa0fd5e718625ffc3df23a9eae6c3",
|
| 15 |
+
"github_metadata_transition_commit": "fc2379488d88124cdee9524186a79e19ed87d2cd",
|
| 16 |
"dataset_content_revision": "1cafbc1f9cf421d43446c385545319a3b0a0a3eb",
|
| 17 |
+
"dataset_viewer_revision": "c10d180eee7cf0c0108603f53903ad71caddc8b3",
|
| 18 |
"adapter_content_revision": "9b4095666621104a94ce2193e62de363e35cba75",
|
| 19 |
+
"github_anonymous_verified": true,
|
| 20 |
+
"dataset_viewer_valid": true,
|
| 21 |
+
"dataset_viewer_splits": [
|
| 22 |
+
"train",
|
| 23 |
+
"dev",
|
| 24 |
+
"eval_core",
|
| 25 |
+
"holdout_principal",
|
| 26 |
+
"holdout_family",
|
| 27 |
+
"holdout_joint"
|
| 28 |
+
],
|
| 29 |
"dataset_public": true,
|
| 30 |
"adapter_public": true,
|
| 31 |
"published_files_match_local_git_blobs": true,
|
README.md
CHANGED
|
@@ -10,6 +10,11 @@ tags:
|
|
| 10 |
- lora
|
| 11 |
- qlora
|
| 12 |
- behavioral-evaluation
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
---
|
| 14 |
|
| 15 |
# BehaviorTune V1.1-R1 QLoRA adapter
|
|
@@ -23,7 +28,8 @@ behavior. It was produced by one bounded completion-only SFT/PEFT QLoRA run.
|
|
| 23 |
- Required base revision: `cdbee75f17c01a7cc42f958dc650907174af0554`
|
| 24 |
- Adapter SHA-256: `8d16ef2cb6ff7a982511fd58f21eff52538761f4d198b4cc5cbfd73ca7c9d4de`
|
| 25 |
- Dataset: [BehaviorTune Trajectories V1.1-R1](https://huggingface.co/datasets/aamish-ahmad/behaviortune-v1-1-r1)
|
| 26 |
-
-
|
|
|
|
| 27 |
|
| 28 |
## Training
|
| 29 |
|
|
|
|
| 10 |
- lora
|
| 11 |
- qlora
|
| 12 |
- behavioral-evaluation
|
| 13 |
+
- llm
|
| 14 |
+
- post-training
|
| 15 |
+
- model-evaluation
|
| 16 |
+
- qwen
|
| 17 |
+
- synthetic-data
|
| 18 |
---
|
| 19 |
|
| 20 |
# BehaviorTune V1.1-R1 QLoRA adapter
|
|
|
|
| 28 |
- Required base revision: `cdbee75f17c01a7cc42f958dc650907174af0554`
|
| 29 |
- Adapter SHA-256: `8d16ef2cb6ff7a982511fd58f21eff52538761f4d198b4cc5cbfd73ca7c9d4de`
|
| 30 |
- Dataset: [BehaviorTune Trajectories V1.1-R1](https://huggingface.co/datasets/aamish-ahmad/behaviortune-v1-1-r1)
|
| 31 |
+
- Project and reproducibility evidence: [BehaviorTune on GitHub](https://github.com/aamish-ahmad/behavior-tune)
|
| 32 |
+
- Immutable scientific release: [GitHub v1.0.0](https://github.com/aamish-ahmad/behavior-tune/releases/tag/v1.0.0)
|
| 33 |
|
| 34 |
## Training
|
| 35 |
|