ModelPublicadoExterno

Rt Detr v2 Finetuned

This model is a fine-tuned version of PekingU/rtdetrv2r50vd on an unknown dataset. It achieves the following results on the evaluation set: Loss: 9.7605 Map: 0.43 Map 50: 0.6 Map 75: 0.5 Map Small: 0.0 Map Medium:...

Fuentehuggingface.co/ashutosh-kedar/rt-detr-v2-finetuned/tree/5383a75185a70b3fd2ac5249c8c0957f90bfd540

TipoOtherModelo basePekingU/rtdetr_v2_r50vdVersiónRevision 5383a75Ccalidad

Fuente del modelo

Tipo: Other
Tamaño del archivo: 164 MB
Modelo base: PekingU/rtdetr_v2_r50vd
Versión: vRevision 5383a75
Actualizado: 16 sept 2026
Descargas: 3
Fuente: Hugging Face

Descripción de la fuente

rt-detr-v2-finetuned

This model is a fine-tuned version of PekingU/rtdetr_v2_r50vd on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 9.7605
Map: 0.43
Map 50: 0.6
Map 75: 0.5
Map Small: 0.0
Map Medium: 0.19
Map Large: 0.45
Mar 1: 0.6
Mar 10: 0.71
Mar 100: 0.73
Mar Small: 0.0
Mar Medium: 0.44
Mar Large: 0.76
Map Bin: 0.78
Mar Bin: 0.88
Map Hand: 0.55
Mar Hand: 0.76
Map Not Bin: 0.13
Mar Not Bin: 0.67
Map Not Hand: 0.01
Mar Not Hand: 0.6
Map Not Trash: 0.18
Mar Not Trash: 0.54
Map Trash: 0.61
Mar Trash: 0.79
Map Trash Arm: 0.76
Mar Trash Arm: 0.88

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0001
train_batch_size: 16
eval_batch_size: 16
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 0.05
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Medium	Mar Large	Map Bin	Mar Bin	Map Hand	Mar Hand	Map Not Bin	Mar Not Bin	Map Not Hand	Mar Not Hand	Map Not Trash	Mar Not Trash	Map Trash	Mar Trash	Map Trash Arm	Mar Trash Arm
82.2808	1.0	50	19.0962	0.09	0.13	0.1	0.01	0.1	0.2	0.4	0.45	0.13	0.48	0.18	0.87	0.26	0.81	0.01	0.5	0.0	0.0	0.0	0.26	0.19	0.7	0.0	0.0
25.2194	2.0	100	11.3634	0.29	0.41	0.32	0.1	0.3	0.38	0.68	0.75	0.42	0.78	0.73	0.9	0.52	0.77	0.01	0.63	0.0	0.8	0.11	0.55	0.57	0.77	0.08	0.82
18.5745	3.0	150	10.0971	0.4	0.54	0.45	0.14	0.41	0.43	0.69	0.76	0.49	0.79	0.77	0.88	0.58	0.82	0.07	0.72	0.0	0.6	0.13	0.61	0.62	0.78	0.61	0.88
16.2489	4.0	200	9.8114	0.41	0.56	0.47	0.2	0.42	0.48	0.7	0.77	0.52	0.8	0.79	0.88	0.55	0.8	0.06	0.75	0.0	0.7	0.17	0.6	0.62	0.8	0.66	0.85
14.9636	5.0	250	9.6389	0.41	0.56	0.47	0.29	0.43	0.5	0.7	0.75	0.47	0.78	0.79	0.89	0.56	0.8	0.1	0.67	0.0	0.6	0.16	0.6	0.61	0.8	0.7	0.88
13.9524	6.0	300	9.5040	0.43	0.6	0.5	0.18	0.45	0.52	0.71	0.76	0.56	0.78	0.79	0.88	0.57	0.8	0.12	0.72	0.0	0.7	0.18	0.58	0.59	0.8	0.75	0.82
13.1142	7.0	350	9.6180	0.43	0.59	0.5	0.31	0.45	0.53	0.72	0.74	0.46	0.77	0.79	0.88	0.57	0.78	0.16	0.65	0.0	0.6	0.17	0.56	0.61	0.8	0.72	0.9
12.5272	8.0	400	9.6804	0.43	0.59	0.49	0.19	0.45	0.61	0.72	0.74	0.45	0.77	0.78	0.88	0.55	0.78	0.11	0.68	0.01	0.6	0.18	0.55	0.61	0.8	0.74	0.88
12.0791	9.0	450	9.7290	0.41	0.58	0.49	0.17	0.43	0.6	0.71	0.72	0.43	0.76	0.77	0.88	0.54	0.76	0.11	0.67	0.01	0.6	0.18	0.53	0.6	0.78	0.69	0.85
11.7099	10.0	500	9.7605	0.43	0.6	0.5	0.19	0.45	0.6	0.71	0.73	0.44	0.76	0.78	0.88	0.55	0.76	0.13	0.67	0.01	0.6	0.18	0.54	0.61	0.79	0.76	0.88

Framework versions

Transformers 5.12.0
Pytorch 2.11.0+cu128
Datasets 4.0.0
Tokenizers 0.22.2

Fuentes

1 fuente

Hugging FacePrincipalVersión 5383a75185a70b3fd2ac5249c8c0957f90bfd540

huggingface.co/ashutosh-kedar/rt-detr-v2-finetuned/tree/5383a75185a70b3fd2ac5249c8c0957f90bfd540

Verificado 2 sept

Artefactos del modelo

1 artefacto

model.safetensors

safetensors · 164 MB · SHA-256 212ed77b1dd2…badb · Hugging Face

Download file

Extractos de fuentes

2 extractos

Readme2 septhuggingface.co/ashutosh-kedar/rt-detr-v2-finetuned

--- library_name: transformers license: apache-2.0 base_model: PekingU/rtdetr_v2_r50vd tags: - generated_from_trainer model-index: - name: rt-detr-v2-finetuned results: [] ---  # rt-detr-v2-finetuned This model is a fine-tuned version of [PekingU/rtdetr_v2_r50vd](https://huggingface.co/PekingU/rtdetr_v2_r50vd) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 9.7605 - Map: 0.43 - Map 50: 0.6 - Map 75: 0.5 - Map Small: 0.0 - Map Medium: 0.19 - Map Large: 0.45 - Mar 1: 0.6 - Mar 10: 0.71 - Mar 100: 0.73 - Mar Small: 0.0 - Mar Medium: 0.44 - Mar Large: 0.76 - Map Bin: 0.78 - Mar Bin: 0.88 - Map Hand: 0.55 - Mar Hand: 0.76 - Map Not Bin: 0.13 - Mar Not Bin: 0.67 - Map Not Hand: 0.01 - Mar Not Hand: 0.6 - Map Not Trash: 0.18 - Mar Not Trash: 0.54 - Map Trash: 0.61 - Mar Trash: 0.79 - Map Trash Arm: 0.76 - Mar Trash Arm: 0.88 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0001 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 0.05 - num_epochs: 10 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Map | Map 50 | Map 75 | Map Small | Map Medium | Map Large | Mar 1 | Mar 10 | Mar 100 | Mar Small | Mar Medium | Mar Large | Map Bin | Mar Bin | Map Hand | Mar Hand | Map Not Bin | Mar Not Bin | Map Not Hand | Mar Not Hand | Map Not Trash | Mar Not Trash | Map Trash | Mar Trash | Map Trash Arm | Mar Trash Arm | |:-------------:|:-----:|:----:|:---------------:|:----:|:------:|:------:|:---------:|:----------:|:---------:|:-----:|:------:|:-------:|:---------:|:----------:|:---------:|:-------:|:-------:|:--------:|:--------:|:-----------:|:-----------:|:------------:|:------------:|:-------------:|:-------------:|:---------:|:---------:|:--...

Note2 septhuggingface.co/ashutosh-kedar/rt-detr-v2-finetuned

Source context: 3 downloads · 0 likes · Pipeline object-detection · Library transformers · Repo ashutosh-kedar/rt-detr-v2-finetuned