lucasliunju

Command about fine-tune and semi-supervised learning on tf2

2

Could you please provide the command about finetune and semi-supervised learning on tf2. I just find the command on tf1 and I find that is not appliable for tf2 Thank...

Hi, Thanks for your contribution. May I ask do you plan to release the code about MAE with scenic. Best, Lucas

the jax-based code on multi-host tpu

3

Hi Yang, That's a great work. I would like to ask whether this code can run on the multi-host tpu (such as v3-32). And could you give me some advice...

JAX code

11

Hi, I would like to ask whether there is a jax-based code. And whether there are some recommendations about jax-based offline rl algorithms. Thanks!

the return value of step function in atari

7

Hi, I try to run the code with default setting for aatari (Pong). I find the step function (dopamine/dopamine/discrete_domains/atari_lib.py +467) returns five values but the code just defines four values....

Hi, I cannot find train.sh and I try to run retrain.sh. I find a problem: AssertionError: Use example_buffer to build a golden_chunk I think that is because we lack of...

About the result of lora fine-tuning

2

May I ask the result of this lora fine-tuning on MMLU task. Thanks! Best, Lucas

About the version of alignment-handbook

1

Hi, Thanks for your great work! I am trying to run the code and reproduce the result. Currently, I found the latest version of alignment-handbook may not match the current...