Lmondi

function createModel(config) local nc = config.nClasses -- 输出种类 36 local nl = nc + 1 -- 37？ local nt = config.maxT -- ？ —— means？Thank you！

@LinJM @bgshih @varun-suresh

@githubharald Ok! Thank you!

@tweiand Hello, have you been realized without GPUs？ Thank you!

@LinJM @bgshih @varun-suresh And about the .7z model. Could it be used for tensorflow?

@LinJM @bgshih @varun-suresh We want to build a set of environments. What kind of hardware environment should I choose?

您好，我在进行微调的时候遇到以下报错： Traceback (most recent call last): File "/data3/LLM-Tuning/chatglm_lora_tuning.py", line 142, in main() File "/data3/LLM-Tuning/chatglm_lora_tuning.py", line 128, in main trainer = ModifiedTrainer( File "/data3/env/miniconda3/envs/baichuan/lib/python3.9/site-packages/transformers/trainer.py", line 499, in __init__ self._move_model_to_device(model, args.device) File...

@beyondguo 您好，git中给的测试环境都一致了，还是不行呢，是不是其他有些包版本影响的呢

> @beyondguo 您好，git中给的测试环境都一致了，还是不行呢，是不是其他有些包版本影响的呢 @beyondguo chatglm中AutoModel.from_pretrained添加了empty_init=False的配置，不报上面的错误了，现在报错： File "/data3/env/miniconda3/envs/baichuan/lib/python3.9/site-packages/transformers/trainer.py", line 1664, in train return inner_training_loop( File "/data3/env/miniconda3/envs/baichuan/lib/python3.9/site-packages/transformers/trainer.py", line 1940, in _inner_training_loop tr_loss_step = self.training_step(model, inputs) File "/data3/env/miniconda3/envs/baichuan/lib/python3.9/site-packages/transformers/trainer.py", line 2735, in training_step loss...

@beyondguo 现在这个微调不支持新的模型吗？