Text Generation
PEFT
Safetensors
lora
model-organism
character-training
persona:sycophantic
conversational
Instructions to use Misalignment-Empirics/jayesh_qwen2.5-32b-it_sycophantic-oct-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Misalignment-Empirics/jayesh_qwen2.5-32b-it_sycophantic-oct-lora with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-32B-Instruct") model = PeftModel.from_pretrained(base_model, "Misalignment-Empirics/jayesh_qwen2.5-32b-it_sycophantic-oct-lora") - Notebooks
- Google Colab
- Kaggle
File size: 786 Bytes
8e78567 | 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 | {
"dpo": {
"dir": "runs/oct-qwen-2.5-32b-it-sycophantic/adapters/sycophantic_oct_dpo_qwen-2.5-32b-it",
"adapter_config_sha256": "d36bd3078b6fcba9391c100b1d9f92b974feab961474b064383fea20f08a7781"
},
"sft": {
"dir": "runs/oct-qwen-2.5-32b-it-sycophantic/adapters/sycophantic_oct_sft_qwen-2.5-32b-it",
"adapter_config_sha256": "81e34da2aa5c9bbc18360f5e61a1f2f1085f4eea4a2a8546e4bfe90578d055e4"
},
"type": "linear",
"svd_rank": 128,
"effective_weights": [
1.0,
1.0
],
"peft_call_weights": [
1.0,
1.0
],
"base_model_name_or_path": "Qwen/Qwen2.5-32B-Instruct",
"r": 64,
"lora_alpha": 64,
"delta_norm_fro": 389.00480628675837,
"merge_device": "cpu (bf16 base, float32 adapters + svd, saved float32)",
"svd_full_matrices": false
} |