How to use Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("/data/icelandic_parse/norwegian-output/gpt_sw3_1b3/kl_sft_merged") model = PeftModel.from_pretrained(base_model, "Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo")
How to use Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo")
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo", dtype="auto")
How to use Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo with vLLM:
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'
docker model run hf.co/Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo
How to use Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo with SGLang:
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'
How to use Hodfa71/gpt-sw3-1b3-nb-saga-kl-sft-delta-dpo with Docker Model Runner:
SAGA KL-SFT + Delta-DPO ablation for Norwegian Bokmal. Base: AI-Sweden-Models/gpt-sw3-1.3b
Base model
