fengyue

Traceback (most recent call last): from fastchat.conversation import (compute_skip_echo_len, ImportError: cannot import name 'compute_skip_echo_len' from 'fastchat.conversation' (/home/a/miniconda3/envs/rag/lib/python3.9/site-packages/fastchat/conversation.py)

2

(rag) a@a-Super-Server:/media/a/DATA/LangChain-ChatGLM-Webui-master$ python3 app.py Traceback (most recent call last): File "/media/a/DATA/LangChain-ChatGLM-Webui-master/app.py", line 16, in from chatllm import ChatLLM File "/media/a/DATA/LangChain-ChatGLM-Webui-master/chatllm.py", line 5, in from fastchat.conversation import (compute_skip_echo_len, ImportError: cannot import...

Issues with torchrun --nproc_per_node num Command and Llama 3.1 Model Conflicts

1

**Issue 1**: For example, the model downloaded from the Meta official website via the provided URL in download.sh results in 8 separate models. When trying to run the code using...

请问可以多卡推理吗，单卡显存有限

12

请问显存不够，只有4090D，bits设置16，如果想用多卡进行微调后的OneKe推理，有提供单机多卡推理代码吗？需要用什么命令执行？ ``` CUDA_VISIBLE_DEVICES=0,1 python src/inference.py \ --stage sft \ --model_name_or_path '/data/shensw/model/OneKE' \ --checkpoint_dir '/data/shensw/DeepKE/example/llm/InstructKGC/lora/oneke_ner_cmeee_add/checkpoint-2810' \ --model_name 'llama' \ --template 'llama2_zh' \ --do_predict \ --input_file 'data/NER/ner_cmeee_add.json' \ --output_file 'results/ner_cmeee_add.json' \ --finetuning_type...

[Bug]: 对lora merge后的模型量化，量化后模型输出一直出现human：

4

### Model Series Qwen2.5 ### What are the models used? Qwen/Qwen2.5-0.5B-Instruct ### What is the scenario where the problem happened? inference with transformers vllm ### Is this a known issue?...

"Which part of the code corresponds to pseudo label generation and cross-modal prototype initialization? Could you please provide some guidance? Thank you very much. Time is limited, and I would like to use your task as my baseline."

2

"Which part of the code corresponds to pseudo label generation and cross-modal prototype initialization? Could you please provide some guidance? Thank you very much. Time is limited, and I would...

when i run **bert-serving-start -model_dir /ORFD/bert/cased_L-12_H-768_A-12/ -num_worker=2** the requirements： tensorflow:1.15.0 bert-serving-start:1.10.0 ``` (orfd) F:\已做项目\ORFD>bert-serving-start -model_dir /ORFD/bert/cased_L-12_H-768_A-12/ -num_worker=2 2024-10-26 18:37:05.473317: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_100.dll usage: D:\Anaconda3\envs\orfd\Scripts\bert-serving-start -model_dir /ORFD/bert/cased_L-12_H-768_A-12/...

使用iic/nlp_raner_named-entity-recognition_chinese-base-cmeee时， 代码： ``` from modelscope.pipelines import pipeline from modelscope.utils.constant import Tasks ner_pipeline = pipeline(Tasks.named_entity_recognition, 'iic/nlp_raner_named-entity-recognition_chinese-base-cmeee') result = ner_pipeline('多数新生儿甲亢在出生时即有症状，表现为突眼、甲状腺肿大、烦躁、多动、心动过速、呼吸急促，严重可出现心力衰竭，血T3、T4升高，TSH下降。') print(result) ``` 若输入太长，会报错，当输入文本长度超过512个字符时，模型会抛出"RuntimeError: The size of tensor a (512) must match the...

ray.exceptions.RayTaskError(CompilationError):

3

Thanks for the lib! When I run the script verl/examples/grpo_trainer/run_qwen3-8b.sh, a RayTaskError occurs. I have checked that the Ray version is (2.43.0), triton(3.1.0),vllm(0.8.5),verl(0.4.0),but this error is not sporadic—it happens every...