-
So disappointed of what is released here. these are just non working pieces.
Funny that in train.py for example you have: from custom import CustomTrainer, but custom is actually have only TunaTrain…
-
I used the code in the guide to download the data, but found that the URL was no longer valid.
`DLINK=$(echo -n "aHR0cHM6Ly9jb252ZXJzYXRpb25odWIuYmxvYi5jb3JlLndpbmRvd3MubmV0L2JlaXQtc2hhcmUtcHVibGljL…
shhn1 updated
2 months ago
-
```
def _pg_loss(
self,
logprobs: TensorType["batch_size", "response_size"],
old_logprobs: TensorType["batch_size", "response_size"],
advantages: TensorType["b…
-
Unable to download processed RoBERTa Corpus only.
Also encountering repeated interruptions during download of the full processed_data.tar with an error indicating dead links, possibly due to incomple…
-
How can I download the training/evaluation intruction-response data?
Can you tell me how to download the data? I tried but failed.
link:
https://conversationhub.blob.core.windows.net/beit-share…
-
Thank you for your awesome work minillm that explored the knowledge distillation for LLMs. I noticed that minillm supports the gpt2/gptj/opt and llama series models only, my question is how should I d…
-
When run the code:bash inference.sh,it occurs an error.
Here is the wrong message.
`.
I think this expression can be calculated directly by `log_softmax`. Why not use `log_softmax`?
htt…
-
### Feature request
Integrate ResLoRA into PEFT.
Paper: https://arxiv.org/abs/2402.18039
Code from Microsoft: https://github.com/microsoft/LMOps/tree/main/reslora
### Motivation
I find interestin…
hllj updated
7 months ago