File size: 1,938 Bytes
4bcd16b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 | ---
license: apache-2.0
library_code: true
tags:
- lora
- medication
- obfuscation
base_model: gpt-oss-120b
---
# LoRA Adapter: Medication Obfuscation Hard 5K
This is a LoRA (Low-Rank Adaptation) adapter for the `gpt-oss-120b` model, fine-tuned on a medication obfuscation dataset.
## Model Details
- **Base Model**: gpt-oss-120b
- **Adapter Type**: LoRA
- **LoRA Rank**: 32
- **LoRA Alpha**: 32
- **Task**: Causal Language Modeling (medication obfuscation)
## Usage
### Loading with transformers and peft
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
base_model_id = "gpt-oss-120b"
adapter_model_id = "Reih02/obfuscated_sandbagging_v3"
# Load base model
model = AutoModelForCausalLM.from_pretrained(
base_model_id,
device_map="auto",
torch_dtype=torch.float16,
)
# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(base_model_id)
# Load LoRA adapter
model = PeftModel.from_pretrained(
model,
adapter_model_id,
device_map="auto"
)
# Now you can use the model
inputs = tokenizer("Your prompt here", return_tensors="pt")
outputs = model.generate(**inputs, max_length=200)
print(tokenizer.decode(outputs[0]))
```
### Using with merge_and_unload
If you want to merge the adapter into the base model:
```python
from peft import PeftModel
from transformers import AutoModelForCausalLM
base_model = AutoModelForCausalLM.from_pretrained(base_model_id, device_map="auto")
model = PeftModel.from_pretrained(base_model, adapter_model_id)
# Merge and unload
merged_model = model.merge_and_unload()
```
## Adapter Configuration
- `peft_type`: LORA
- `r`: 32
- `lora_alpha`: 32
- `lora_dropout`: 0
- `target_modules`: all-linear
- `bias`: none
- `task_type`: CAUSAL_LM
## Citation
If you use this adapter in your research, please cite the base model and the adapter.
## License
This adapter is released under the Apache 2.0 License.
|