wangyan_sdu

Results 6 issues of wangyan_sdu

Dear Victor Escorcia, I found that the necessary document of "dataset_untrimmed.py" is not provided in the repository. I am wondering if you could kindly share me this document. I promise...

For Charades, the frames of each video can be up to 1000. This is too large to train properly. How to change the frame rate in the code to reduce...

My model requires the following inputs in the forward method: ```python def forward( self, input_ids=None, attention_mask=None, position_ids=None, head_mask=None, video_query_tokens=None, frame_hidden_state=None, frame_atts=None, past_key_values=None, use_cache=None, output_attentions=None, output_hidden_states=None, return_dict=None, is_decoder=False, ): ``` However,...

RuntimeError (note: full exception trace is shown but execution is paused at: _run_module_as_main) Expected size for first two dimensions of batch2 tensor to be: [1, 4] but got: [1, 512]....

bug
no-issue-activity

**Describe the feature** 提供多种损失函数的sft训练,比如对比损失 **Paste any useful information** sft时,除了交叉熵损失,有时需要针对某个特定token计算对比损失、pairloss等等,可否集成这样一个功能呢? **Additional context**

enhancement
swift3.0

**Describe the bug** ```json { "name": "Python: debug_cl", "type": "debugpy", "request": "launch", "program": "swift/cli/main.py", "console": "integratedTerminal", "subProcess": true, "justMyCode": false, "env": { "NPROC_PER_NODE": "4", "CUDA_VISIBLE_DEVICES": "0,1,2,3", "PYTHONPATH": "./" }, "args":...