|
Download README.md from GaMS-Beta/SVILA-1-4B: direct link, hf CLI and curl.
- Browser
- Download file 1.7 kB
-
https://huggingface.co/GaMS-Beta/SVILA-1-4B/resolve/main/README.md
- Command line
-
hf download hf://GaMS-Beta/SVILA-1-4B/README.md
-
curl -L -o README.md https://huggingface.co/GaMS-Beta/SVILA-1-4B/resolve/main/README.md
1.7 kB
| license: mit | |
| language: | |
| - en | |
| base_model: | |
| - google/gemma-3-4b-it | |
| pipeline_tag: text-generation | |
| tags: | |
| - Vision language model (VLM) | |
| - Slovenian | |
| # SVILA - Slovenian Vision Language Assistant - is a Vision Language Model for Slovenian.# | |
| ## The model is based on google/gemma-3-4b-it and was fine-tuned on curated instruction-tuning text-image Slovenian dataset using a custom SFT trainer. | |
| ## Dataset is available here: https://clarin.si/repository/xmlui/handle/11356/2050 | |
| ## How to use it: ## | |
| ```python | |
| from transformers import AutoProcessor, Gemma3ForConditionalGeneration | |
| import torch | |
| model_id = "GaMS-Beta/SVILA-1-4B" | |
| model = Gemma3ForConditionalGeneration.from_pretrained( | |
| model_id, device_map="auto" | |
| ).eval() | |
| processor = AutoProcessor.from_pretrained(model_id) | |
| messages = [ | |
| { | |
| "role": "system", | |
| "content": [{"type": "text", "text": ""}] | |
| }, | |
| { | |
| "role": "user", | |
| "content": [ | |
| {"type": "image", "image": "https://www.dangerous-business.com/wp-content/uploads/2024/02/DSC02109.jpg"}, | |
| {"type": "text", "text": "Kaj je na sliki?"} | |
| ] | |
| } | |
| ] | |
| print(processor.apply_chat_template(messages, tokenize=False)) | |
| inputs = processor.apply_chat_template( | |
| messages, add_generation_prompt=True, tokenize=True, | |
| return_dict=True, return_tensors="pt" | |
| ).to(model.device, dtype=torch.bfloat16) | |
| input_len = inputs["input_ids"].shape[-1] | |
| with torch.inference_mode(): | |
| generation = model.generate(**inputs, max_new_tokens=500) | |
| generation = generation[0][input_len:] | |
| decoded = processor.decode(generation, skip_special_tokens=True) | |
| print(decoded) | |
| ``` | |