| # tracking_shuffled_objects LoRA Models |
|
|
| This repository contains LoRA (Low-Rank Adaptation) models trained on the tracking_shuffled_objects dataset. |
|
|
| ## Models in this repository: |
|
|
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0008_data_size1000_max_steps=500_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0008_data_size1000_max_steps=500_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr5e-05_data_size1000_max_steps=500_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr5e-05_data_size1000_max_steps=500_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0001_data_size1000_max_steps=500_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0001_data_size1000_max_steps=500_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0004_data_size1000_max_steps=100_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0004_data_size1000_max_steps=100_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0002_data_size1000_max_steps=500_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0002_data_size1000_max_steps=500_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0008_data_size1000_max_steps=100_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0008_data_size1000_max_steps=100_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0002_data_size1000_max_steps=100_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0002_data_size1000_max_steps=100_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0006_data_size1000_max_steps=500_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0006_data_size1000_max_steps=500_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0003_data_size1000_max_steps=100_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0003_data_size1000_max_steps=100_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0004_data_size1000_max_steps=500_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0004_data_size1000_max_steps=500_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0005_data_size1000_max_steps=500_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0005_data_size1000_max_steps=500_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0003_data_size1000_max_steps=500_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0003_data_size1000_max_steps=500_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0001_data_size1000_max_steps=100_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0001_data_size1000_max_steps=100_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0005_data_size1000_max_steps=100_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0005_data_size1000_max_steps=100_seed=123 |
| - `llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0006_data_size1000_max_steps=100_seed=123/`: LoRA adapter for llama_finetune_tracking_shuffled_objects_r16_alpha=32_dropout=0.05_lr0.0006_data_size1000_max_steps=100_seed=123 |
| |
| ## Usage |
| |
| To use these LoRA models, you'll need the `peft` library: |
| |
| ```bash |
| pip install peft transformers torch |
| ``` |
| |
| Example usage: |
| |
| ```python |
| from peft import PeftModel |
| from transformers import AutoModelForCausalLM, AutoTokenizer |
| |
| # Load base model |
| base_model_name = "your-base-model" # Replace with actual base model |
| model = AutoModelForCausalLM.from_pretrained(base_model_name) |
| tokenizer = AutoTokenizer.from_pretrained(base_model_name) |
| |
| # Load LoRA adapter |
| model = PeftModel.from_pretrained( |
| model, |
| "supergoose/tracking_shuffled_objects", |
| subfolder="model_name_here" # Replace with specific model folder |
| ) |
| |
| # Use the model |
| inputs = tokenizer("Your prompt here", return_tensors="pt") |
| outputs = model.generate(**inputs) |
| ``` |
| |
| ## Training Details |
| |
| - Dataset: tracking_shuffled_objects |
| - Training framework: LoRA/PEFT |
| - Models included: 15 variants |
| |
| ## Files Structure |
| |
| Each model folder contains: |
| - `adapter_config.json`: LoRA configuration |
| - `adapter_model.safetensors`: LoRA weights |
| - `tokenizer.json`: Tokenizer configuration |
| - Additional training artifacts |
| |
| --- |
| *Generated automatically by LoRA uploader script* |
| |
