wuzhaodongaipython
wuzhaodongaipython
Hello, thank you for your work for nnunetv2. Firstly, i will introduce two datasets: the first one is source dataset which has 7 classes, the second one is target dataset...
The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values....
恭喜您完成了一个如此出色的工作 我想请问, 1:即将发布的4b和30b的instruct模型,是否有thinking with images的功能呢 2:thinking with images对应的数据集,是否会公开呢,如果会,是啥时候呢。
恭喜完成这份出色的工作 我想请问一下thinking with images能力的训练代码有公开吗,我想基于这上面做点东西 目前只看到了推理的ipynb的文件
tok_len = len(tokenizer) emb = model.get_input_embeddings() emb_w = emb.weight hidden = emb_w.shape[-1] if emb_w.dim() > 1 else 0 total = getattr(emb_w, "ds_numel", emb_w.numel()) # Zero3 时 ds_numel 是全局元素数 if hidden...
hi,非常恭喜完成这么棒的工作。 我目前想测评V *等数据集,但我似乎没找到规模化的推理代码,尽管您在readme里面写了单次推理的代码,但规模化的推理代码如https://github.com/QwenLM/Qwen3-VL/tree/main/evaluation/mmmu所示中,貌似只提供qwen2.5vl的推理。 请问如果我想规模化地推理qwen3vl,应该怎么做呢?