Oguzhan Gencoglu comments

Results 101 comments of


                                            Oguzhan Gencoglu

Morphology benchmarks to torch (CPU), for example kornia

Maybe also for openCV

Morphology benchmarks to torch (CPU), for example kornia

Also possibly with [mahotas](https://github.com/luispedro/mahotas), [smil](https://smil.cmm.minesparis.psl.eu/doc/p480.html), [pyvips](https://github.com/libvips/libvips/wiki/Speed-and-memory-use) which all claim to be really fast regarding these kind of operations.

Object detection

NOT_IMPLEMENTED : Could not find an implementation for ConvInteger(10) node with name 'Conv_0_quant'

@fazankabir No solution found so far.

NOT_IMPLEMENTED : Could not find an implementation for ConvInteger(10) node with name 'Conv_0_quant'

I haven't tested that (onnx is a must for my case) but quantized onnx model has longer inference time than full precision one. So I ditched segformer all together. @Fazankabir

NOT_IMPLEMENTED : Could not find an implementation for ConvInteger(10) node with name 'Conv_0_quant'

And what is the inference speed @abaoxomtieu ? In my use case, partial quantization gets slower than non-quantized model on CPU.

ImportError: cannot import name 'line_optim' from 'line_refinement' (unknown location)

Hi-jacking this. Does line refinement mean e.g. double lines will be merged into one with the `'optimize': True` config or something else?

I'm not able to reproduce documentation example

@shahules786 you need to take this seriously to avoid going down the langchain path. everybody hates langchain docs because it is broken and not up-to-date. Please do not prioritize new...

Invalid response format. Expected a list of dictionaries with keys 'verdict'

Same issue with Bedrock models, e.g. `"anthropic.claude-v2"`

Issue with "Bring Your Own LLMs"

So what is the conclusion? can we use any litellm supported model instead of OpenAI?