Add attention_bias to make TGI work (#68)

Add link to Phi-3 vision ONNX models

Update sample_finetune.py (#70)

Update sample_finetune.py (#69)

Update README

fix(tokenizer): Also updates tokenizer.json to prevent any mismatch.

fix(tokenizer_config): Adjusts `rstrip` of special tokens. (#54)

Update README.md

fix(root): Replaces system by user to improve generation experience.

Update README.md

Update sample_finetune.py (#42)

chore(readme): Adds a proper chat example to widget.

fix(readme): Adds information about placeholder tokens.

fix(tokenizer_config): Ensures model_max_length corresponds to actual maximum number.

chore(root): Initial files upload.

initial commit