sanwei111
sanwei111
after i changed code： model.load_state_dict(model_dict)，becomes： RuntimeError: Error(s) in loading state_dict for DistributedDataParallel: Unexpected key(s) in state_dict: "base_model.conv1.weight", "base_model.conv1.bias", "base_model.bn1.weight", "base_model.bn1.bias", "base_model.bn1.running_mean", "base_model.bn1.running_var", "base_model.bn1.num_batches_tracked", "base_model.conv1_temp.weight", "base_model.conv1_temp.bias", "base_model.conv1_5.0.weight", "base_model.conv1_5.1.weight", "b
download 2 model from Kinetics400
有指令吗，可以提供一下吗？因为readme是多机多卡的
模型使用的是chatglm，没有量化
还是不行，运行指令是CUDA_VISIBLE_DEVICES=0,1 python train.py 已经是batchsize到1了，maxseqlen=512了
改了参数之后，删除output下的数据，再来一次data_utils.py？这个做了
关了试一下，感觉好一点，但还是很大： torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 128.00 MiB (GPU 1; 31.75 GiB total capacity; 30.54 GiB already allocated; 27.75 MiB free; 30.76 GiB reserved in total by...
默认值：32
pre-seq-len：16，batchsize-2，maxseqlen-512，还不行
用的是chatglm_finetuning-stable-vocab130528-v2这个分支