Instructions to use pgdev0410/cybersos-tiny-aya-dagger-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use pgdev0410/cybersos-tiny-aya-dagger-lora with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("/root/lexsi_hack/artifacts/base") model = PeftModel.from_pretrained(base_model, "pgdev0410/cybersos-tiny-aya-dagger-lora") - Notebooks
- Google Colab
- Kaggle
Download resolved_sft_config.json from pgdev0410/cybersos-tiny-aya-dagger-lora: direct link, hf CLI and curl.
- Browser
- Download file 625 Bytes
-
https://huggingface.co/pgdev0410/cybersos-tiny-aya-dagger-lora/resolve/main/resolved_sft_config.json
- Command line
-
hf download hf://pgdev0410/cybersos-tiny-aya-dagger-lora/resolved_sft_config.json
-
curl -L -o resolved_sft_config.json https://huggingface.co/pgdev0410/cybersos-tiny-aya-dagger-lora/resolve/main/resolved_sft_config.json
625 Bytes
| { | |
| "base_manifest_path": "configs/base_model.json", | |
| "batch_size": 8, | |
| "development_scenarios_path": "data/scenarios/dev.jsonl", | |
| "gradient_accumulation_steps": 1, | |
| "gradient_checkpointing": true, | |
| "learning_rate": 0.0001, | |
| "lora_alpha": 32, | |
| "lora_dropout": 0.05, | |
| "lora_r": 16, | |
| "lora_target_modules": "all-linear", | |
| "model_path": "artifacts/base", | |
| "num_epochs": 1, | |
| "output_dir": "runs/sft-adapter-dagger", | |
| "precision": "bf16", | |
| "quantization": {}, | |
| "rendered_dataset_path": "data/rendered/train_alpaca_dagger.jsonl", | |
| "seed": 42, | |
| "serving_contract_path": "configs/baselines.yaml", | |
| "smoke": false | |
| } | |