LostOfMoon/generalgrounder

1---2library_name: transformers3license: other4base_model: Qwen/Qwen2.5-VL-3B-Instruct5tags:6- llama-factory7- full8- generated_from_trainer9model-index:10- name: grounder11  results: []12---13 14<!-- This model card has been generated automatically according to the information the Trainer had access to. You15should probably proofread and complete it, then remove this comment. -->16 17# grounder18 19This model is a fine-tuned version of [Qwen/Qwen2.5-VL-3B-Instruct](https://huggingface.co//Qwen/Qwen2.5-VL-3B-Instruct) on the general_grounder_train and the os_atlas_4k_train datasets.20It achieves the following results on the evaluation set:21- Loss: 0.462422 23## Model description24 25More information needed26 27## Intended uses & limitations28 29More information needed30 31## Training and evaluation data32 33More information needed34 35## Training procedure36 37### Training hyperparameters38 39The following hyperparameters were used during training:40- learning_rate: 1e-0541- train_batch_size: 142- eval_batch_size: 143- seed: 4244- distributed_type: multi-GPU45- num_devices: 846- gradient_accumulation_steps: 1647- total_train_batch_size: 12848- total_eval_batch_size: 849- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments50- lr_scheduler_type: cosine51- lr_scheduler_warmup_ratio: 0.152- num_epochs: 253 54### Training results55 56| Training Loss | Epoch  | Step | Validation Loss |57|:-------------:|:------:|:----:|:---------------:|58| 0.5283        | 0.4184 | 50   | 0.5561          |59| 0.5076        | 0.8368 | 100  | 0.5050          |60| 0.4264        | 1.2510 | 150  | 0.4817          |61| 0.4474        | 1.6695 | 200  | 0.4643          |62 63 64### Framework versions65 66- Transformers 4.51.367- Pytorch 2.4.0+cpu68- Datasets 3.6.069- Tokenizers 0.21.170