MOSS

MOSS copied to clipboard

hi, since some decent GPUs has gained some accelerations on torch2.0 does torch2.0 support now?

lucasjinreal

onnxruntime.capi.onnxruntime_pybind11_state.InvalidProtobuf: [ONNXRuntimeError] : 7 : INVALID_PROTOBUF : Load model from moss-moon-003-sft-plugin failed:Protobuf parsing failed

2

参考 https://github.com/OpenLMLab/MOSS_Vortex 换了个模型 ``` 2023-05-11 11:04:36,394 - 82 - INFO - mosec_server.py:357 - [MOSEC] [INIT] Initializing model on device=3 2023-05-11 11:04:36,423 - 124 - INFO - mosec_server.py:357 - [MOSEC] [INIT]...

631068264

How to fix can not find "./sft_data/train.jsonl"

3

we can convert dataset in `SFT_data/conversations` to `./sft_data/train.jsonl` Here is the code: ```python # prepare_moss_sft.py """ prepare train and val dataset for for moss-sft This method will not generate the...

liujiabin20810

How to convert a finetuned MOSS model to quantized version model? 请问如何把一个finetune过的MOSS模型转换为量化版的模型呢？

3

I couldn't find the code to release this process in this repository. Has anyone successfully converted a finetuned MOSS model to its quantized version? If so, could you please share...

qgpmztmf

"wbits": 4, 改为： "wbits": 8,

hongwen-sun

使用SFT后的FP32模型进行生成，报错RuntimeError: where expected condition to be a boolean tensor, but got a tensor with dtype Half

4


ARIELDENG

1.运行run.sh，生成moss-moon-003-sft-int4下面的文件，包含： - 1 - latest - zero_to_fp32.py 2.运行zero_to_fp32.py 生成新文件到output_file路径下面（新模型） 3.加载新模型demo，正常。 4.对话输出，任何问题的答案都变成了感叹号，： “!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!” 这是什么原因，有哪位大佬帮助解决一下，谢谢。 

skysing

当我运行int4版本的模型的时候报错，代码就是readme里提供的示例，"pow" not implemented for 'Half'

1

请问大佬们怎么解决，谢谢。

Shicc


xyeagle80

zjmwqx