Edward Atkins comments

Results 4 comments of


                                            Edward Atkins

Getting handwritten model to work in Python

Apologies, I'm just working in my all purpose environment, so I've dumped the whole thing in a text file below. I imagine the important ones are: ``` matplotlib==3.1.1 tensorflow==1.14.0 Pillow==6.1.0...

[BUG] Failed to score model on payload. Error: 'Endpoint' object has no attribute 'get'

Similar config file to @alistairwgillespie above with same error message. Hardcoding `endpoint_type` in `metrics/genai/model_utils.py` like below seems to resolve the issue. Not sure of the explanation for why it is...

Finetuning run times out at evaluation step on multiple devices

@rasbt - That is completely fine FYI ``` (base) ubuntu@ip-10-0-0-185:~/sky_workdir$ litgpt generate base --prompt "Recommend a movie for me to watch during the weekend and explain the reason." --checkpoint_dir checkpoints/mistralai/Mistral-7B-Instruct-v0.1...

AttributeError: 'LRFinder' object has no attribute '_implements_train_batch_hooks'

Had the same error when switching over to `TF2`, Assuming you are also running `TF2` the two `keras` imports in `clr.py` can be switched to ``` from tensorflow.keras.callbacks import Callback...