---
library_name: transformers
license: other
base_model: google/medsiglip-448
tags:
- generated_from_trainer
model-index:
- name: medsiglip-448-ft-tb-screening
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# medsiglip-448-ft-tb-screening

This model is a fine-tuned version of [google/medsiglip-448](https://huggingface.co/google/medsiglip-448) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.6822

## 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: 5e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 64
- 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: cosine
- lr_scheduler_warmup_steps: 10
- num_epochs: 10

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 3.5064        | 0.2140 | 25   | 2.4667          |
| 1.9406        | 0.4280 | 50   | 2.5449          |
| 1.9175        | 0.6421 | 75   | 2.5669          |
| 1.8659        | 0.8561 | 100  | 2.7958          |
| 1.9603        | 1.0685 | 125  | 2.6281          |
| 1.8811        | 1.2825 | 150  | 2.5601          |
| 1.8955        | 1.4965 | 175  | 2.5833          |
| 1.8982        | 1.7105 | 200  | 2.6373          |
| 1.825         | 1.9246 | 225  | 2.6426          |
| 1.88          | 2.1370 | 250  | 2.8641          |
| 1.851         | 2.3510 | 275  | 2.6415          |
| 1.8619        | 2.5650 | 300  | 2.5749          |
| 1.8365        | 2.7790 | 325  | 2.6245          |
| 1.8783        | 2.9930 | 350  | 2.5929          |
| 1.8693        | 3.2055 | 375  | 2.5986          |
| 1.8605        | 3.4195 | 400  | 2.6601          |
| 1.8759        | 3.6335 | 425  | 2.5904          |
| 1.8731        | 3.8475 | 450  | 2.6054          |
| 1.8536        | 4.0599 | 475  | 2.6441          |
| 1.8509        | 4.2739 | 500  | 2.6678          |
| 1.8609        | 4.4880 | 525  | 2.6946          |
| 1.8478        | 4.7020 | 550  | 2.6386          |
| 1.8492        | 4.9160 | 575  | 2.6799          |
| 1.8549        | 5.1284 | 600  | 2.6355          |
| 1.88          | 5.3424 | 625  | 2.7021          |
| 1.8569        | 5.5564 | 650  | 2.6380          |
| 1.862         | 5.7705 | 675  | 2.6349          |
| 1.8486        | 5.9845 | 700  | 2.6843          |
| 1.8503        | 6.1969 | 725  | 2.6926          |
| 1.8503        | 6.4109 | 750  | 2.6962          |
| 1.84          | 6.6249 | 775  | 2.6286          |
| 1.8466        | 6.8390 | 800  | 2.6278          |
| 1.8584        | 7.0514 | 825  | 2.6274          |
| 1.8633        | 7.2654 | 850  | 2.6308          |
| 1.8744        | 7.4794 | 875  | 2.6365          |
| 1.8522        | 7.6934 | 900  | 2.6514          |
| 1.8578        | 7.9074 | 925  | 2.6701          |
| 1.8661        | 8.1199 | 950  | 2.6817          |
| 1.8301        | 8.3339 | 975  | 2.6813          |
| 1.8499        | 8.5479 | 1000 | 2.6841          |
| 1.8484        | 8.7619 | 1025 | 2.6832          |
| 1.8815        | 8.9759 | 1050 | 2.6814          |
| 1.8082        | 9.1883 | 1075 | 2.6836          |
| 1.8302        | 9.4024 | 1100 | 2.6839          |
| 1.8822        | 9.6164 | 1125 | 2.6824          |
| 1.8648        | 9.8304 | 1150 | 2.6822          |


### Framework versions

- Transformers 4.56.0
- Pytorch 2.8.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.0