GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Kushtrim
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
Kushtrim
Model Tree
unsloth/Qwen3-VL-2B-Instruct
Input Modalities
Output Modalities
Supported Functionality
sq) for OCR and document-style prompts.unsloth/Qwen3-VL-2B-Instructsq)Use this model for:
unsloth/Qwen3-VL-2B-Instruct (derived from Qwen/Qwen3-VL-2B-Instruct)Kushtrim/bujku_vl_ocr (page image + Albanian transcription)
Transcribe the text in this image.Practical defaults used by the trainer scripts in this folder (may vary by run):
max_length: 2048r=16, alpha=16, dropout 0.0Qwen3-VL support is in recent
transformers. If you hit import errors, install from source:
pip install git+https://github.com/huggingface/transformers
python
import torchfrom transformers import Qwen3VLForConditionalGeneration, AutoProcessorfrom peft import PeftModel BASE_ID = "Qwen/Qwen3-VL-2B-Instruct" # <-- the real base modelADAPTER_ID = "Kushtrim/Qwen3-VL-2B-Instruct-Shqip" model = Qwen3VLForConditionalGeneration.from_pretrained( BASE_ID, torch_dtype="auto", device_map="auto", trust_remote_code=True,) model = PeftModel.from_pretrained(model, ADAPTER_ID, token=token) # Optional (often nicer for inference): merge adapter into weightsmodel = model.merge_and_unload() processor = AutoProcessor.from_pretrained( BASE_ID, trust_remote_code=True,) messages = [ { "role": "user", "content": [ { "type": "image", "image": "BU-19970125_34.png", }, {"type": "text", "text": "Transcribe the text in this image."}, ], }] # Preparation for inferenceinputs = processor.apply_chat_template( messages, tokenize=True, add_generation_prompt=True, return_dict=True, return_tensors="pt")inputs = inputs.to(model.device) # Inference: Generation of the outputgenerated_ids = model.generate(**inputs, max_new_tokens=2048)generated_ids_trimmed = [ out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)]output_text = processor.batch_decode( generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False)print(output_text)unsloth/Qwen3-VL-2B-InstructThis work builds on Qwen3-VL-2B-Instruct and the work of its original authors. The upstream architecture and base weights remain the work of their respective creators. This repository documents the task-specific adaptation and release work carried out by Kushtrim Visoka and does not claim authorship of the original model. Dataset creators and tool authors referenced elsewhere in this card should also be cited according to their licenses and documentation.
If you use this fine-tuned model, please cite this repository and the original upstream model.
bibtex
@misc{visoka_qwen3_vl_2b_instruct_shqip_2026, author = {Visoka, Kushtrim}, title = {Qwen3-VL 2B Instruct Shqip}, year = {2026}, publisher = {Hugging Face}, url = {https://huggingface.co/Kushtrim/Qwen3-VL-2B-Instruct-Shqip}, note = {Fine-tuned derivative of Qwen3-VL-2B-Instruct}}import torchfrom transformers import Qwen3VLForConditionalGeneration, AutoProcessorfrom peft import PeftModel BASE_ID = "Qwen/Qwen3-VL-2B-Instruct" # <-- the real base modelADAPTER_ID = "Kushtrim/Qwen3-VL-2B-Instruct-Shqip" model = Qwen3VLForConditionalGeneration.from_pretrained( BASE_ID, torch_dtype="auto", device_map="auto", trust_remote_code=True,) model = PeftModel.from_pretrained(model, ADAPTER_ID, token=token) # Optional (often nicer for inference): merge adapter into weightsmodel = model.merge_and_unload() processor = AutoProcessor.from_pretrained( BASE_ID, trust_remote_code=True,) messages = [ { "role": "user", "content": [ { "type": "image", "image": "BU-19970125_34.png", }, {"type": "text", "text": "Transcribe the text in this image."}, ], }] # Preparation for inferenceinputs = processor.apply_chat_template( messages, tokenize=True, add_generation_prompt=True, return_dict=True, return_tensors="pt")inputs = inputs.to(model.device) # Inference: Generation of the outputgenerated_ids = model.generate(**inputs, max_new_tokens=2048)generated_ids_trimmed = [ out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)]output_text = processor.batch_decode( generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False)print(output_text)@misc{visoka_qwen3_vl_2b_instruct_shqip_2026, author = {Visoka, Kushtrim}, title = {Qwen3-VL 2B Instruct Shqip}, year = {2026}, publisher = {Hugging Face}, url = {https://huggingface.co/Kushtrim/Qwen3-VL-2B-Instruct-Shqip}, note = {Fine-tuned derivative of Qwen3-VL-2B-Instruct}}