INT8 quantization for "surya" model

[as-suvorov]

Ticket: 132425

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aleksandr-mokrov commented on 2024-04-30T10:01:15Z
----------------------------------------------------------------

Line #5.        predictions = batch_text_detection([image], int8_ov_model_wrapper, processor)

Could you add a selector which model to use, int8 or fp16?

as-suvorov commented on 2024-04-30T15:24:28Z
----------------------------------------------------------------

added

[as-suvorov]

added

---

View entire conversation on ReviewNB

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

eaidova commented on 2024-05-01T14:00:44Z
----------------------------------------------------------------

Line #24.            logits = self.ov_model(kwargs)[logits_out]

there is no need to use logits_out = self.ov_model.output(0), you can simplify just using

logits = self.ov_model(kwargs)[0]

[eaidova]

@as-suvorov please fix code style