-
I am looking at the SetFit example with SetFitHead:
```
# Create trainer
trainer = SetFitTrainer(
model=model,
train_dataset=train_dataset,
eval_dataset=eval_dataset,
loss_class…
-
Hi,
I am in PPO training phase (step 3) with llama2-13b model. However, I get unexcepted print answers in 'generate_experience' function as following image. May I ask how should I solve this problem…
-
Hello,
Thanks for your good work!
As your code hasn't been uploaded yet, I've encountered a few questions about the paper that I couldn't address through the code. That's why I've raised this i…
-
Hello, thank you for putting up such a great job.
I have a question, why do you call this work as a self-supervised learning method? It seems that this paper does not explain the specific training l…
-
Are pre-trained models using contrastive learning trained by yourself or using models pre-trained by others?
-
Thank you for your work.
When I use ‘torch.backends.cudnn.deterministic = True’ to fix the random seed to ensure that the results of each run are the same, the running speed of the model is dozens …
-
Hi!
I successfully continued teaching the clip on my own data.
Created flax_model.msgpack and config.json.
How can I test the model?
Thanks,
G.
-
Hi,
Thank you for your interesting work and code implementation.
I want to explore the results without dummy. But, when I comment out the dummy-related parts of the code, I always get an error `asse…
-
Dear PyGOD team,
Thank you for your great work. I noticed that you had the implementation for the ANEMONE algorithm [1] in version V0.2.0 [2] but it is not available in your V1.0.0. May I ask will …
-
Hello, thank you for your amazing work! Having read the paper, I wonder if the whole idea could be of help in terms of multi-modal personalized dialogue generation task. I will post some of my prelimi…