ayyuce/blip_vqa_rad

README.md

@@ -1,16 +1,11 @@
 ---
 library_name: transformers
-license: gpl-3.0
+license: bsd-3-clause
 base_model: Salesforce/blip-vqa-base
 tags:
 - generated_from_trainer
-datasets:
-- ayyuce/vqa-rad-instructions
-language:
-- en
-pipeline_tag: visual-question-answering
 model-index:
-- name: blip-vqa-rad-checkpoints
+- name: blip-vqa-rad
   results: []
 ---
 
@@ -21,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [Salesforce/blip-vqa-base](https://huggingface.co/Salesforce/blip-vqa-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2821
+- Loss: 3.1093
 
 ## Model description
 
@@ -46,13 +41,14 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
-- num_epochs: 1
+- num_epochs: 2
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 0.3407        | 1.0   | 897  | 0.2821          |
+| 3.5481        | 1.0   | 897  | 3.1706          |
+| 3.3885        | 2.0   | 1794 | 3.1093          |
 
 
 ### Framework versions
@@ -60,4 +56,4 @@ The following hyperparameters were used during training:
 - Transformers 4.50.0
 - Pytorch 2.6.0+cu124
 - Datasets 3.5.0
-- Tokenizers 0.21.1
+- Tokenizers 0.21.1

preprocessor_config.json

@@ -1,4 +1,5 @@
 {
+  "do_convert_rgb": true,
   "do_normalize": true,
   "do_pad": true,
   "do_rescale": true,
@@ -22,4 +23,4 @@
     "width": 384
   },
   "size_divisor": 32
-}
+}