Adding `safetensors` variant of this model
#24 opened 17 days ago
by
SFconvertbot
chatglm2-6b-int4报错RuntimeError: expected m1 and m2 to have the same dtype, but got: c10::Half != float
1
#23 opened 5 months ago
by
frank08932
Adding `safetensors` variant of this model
#22 opened 7 months ago
by
SFconvertbot
Create handler.py
1
#21 opened 11 months ago
by
Linsad
Update tokenization_chatglm.py
#20 opened about 1 year ago
by
sylwia-kuros
linux环境部署cpu模式
#19 opened about 1 year ago
by
langya2
使用AdaLora微调训练chatglm2-6b-int4模型报错
1
#18 opened about 1 year ago
by
tcdoudousky
'NoneType' object has no attribute 'int4WeightExtractionHalf'
1
#16 opened over 1 year ago
by
father123
AttributeError: 'NoneType' object has no attribute 'int4WeightExtractionHalf'
2
#15 opened over 1 year ago
by
father123
运行量化模型时报错:RuntimeError: CUDA Error: no kernel image is available for execution on the device
5
#14 opened over 1 year ago
by
saubcy
INT4 model shows bad perf than FP32 on Intel CPU,why?
1
#13 opened over 1 year ago
by
Sakura10151
ModuleNotFoundError: No module named 'transformers_modules.chatglm2-6b-int4'
#12 opened over 1 year ago
by
EthanMiao
怀疑hugging face现在不让用int4版本了,前段时间还用了
#11 opened over 1 year ago
by
father123
前段时间还能用int4版本了,现在又用不了了,怎么回事。。。。。。。
2
#10 opened over 1 year ago
by
father123
English only version?
#9 opened over 1 year ago
by
amohajerani
"addmm_impl_cpu_" not implemented for 'Half'
#8 opened over 1 year ago
by
you-2
quantization_kernels.c和quantization_kernels_parallel.c直接使用chatglm1-6b-int4项目中的吗?
#5 opened over 1 year ago
by
MrHaiPi
示例代码里还是THUDM/chatglm2-6b
#4 opened over 1 year ago
by
lunfan
Perplxity between quantized and original?
#3 opened over 1 year ago
by
Yhyu13
报错了 Failed to load cpm_kernels:Unknown platform: darwin
1
#2 opened over 1 year ago
by
vZzzv
如果这个是用Bitsandsbyte的NF4量化的,能否直接在这个基础上用qlora继续训练?
#1 opened over 1 year ago
by
bash99