issues
search
ssbuild
/
chatglm_finetuning
chatglm 6b finetuning and alpaca finetuning
1.54k
stars
176
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
deepspeed 和普通训练（lora ptuning） batch_size 只能设置4以下 不然会OOM
#233
markWJJ
closed
1 year ago
21
ptv2显存不够？
#232
sanwei111
opened
1 year ago
11
单机两卡指令怎么样
#231
sanwei111
opened
1 year ago
2
关于数据的instruction，input，output
#230
sanwei111
opened
1 year ago
3
v2
#229
ssbuild
closed
1 year ago
0
关于数据格式
#228
sanwei111
opened
1 year ago
6
V2 merge
#227
ssbuild
closed
1 year ago
0
RuntimeError: expected scalar type Half but found Float 你们在训练的时候遇到过这样的情况吗？
#226
SMR-S
closed
1 year ago
3
v2
#225
ssbuild
closed
1 year ago
0
v2
#224
ssbuild
closed
1 year ago
0
加载lora模型出错～
#223
zlht812
closed
1 year ago
0
merge v2
#222
ssbuild
closed
1 year ago
0
请问如何试用一般新闻语料对ChatGLM进行继续finetuing呢？
#221
yang9112
opened
1 year ago
1
web/api_lora_demo.py 如何多张卡推理
#220
lxw0109
opened
1 year ago
0
第一次f16 lora双卡训练成功，第二次int8 lora单卡训练成功，第三次 换会f16 lora双卡训练失败，详情请进～
#219
zlht812
closed
1 year ago
3
使用ptv2的方式进行finetune，总是OOM（正常微调、lora方式都没问题）
#218
lxw0109
closed
1 year ago
7
请问一下，mac系统装不了deep_training？
#217
WHJTC
closed
1 year ago
1
Lora推理2分30s正常吗？
#216
jikhunb
closed
1 year ago
2
Lora训练后推理问题
#215
jikhunb
closed
1 year ago
2
python train.py执行训练报错，求解。
#214
pan365wang
closed
1 year ago
9
设置 LoRa微调的 'target_modules' 后，运行报错 "AssertionError"
#213
ngbruce
closed
1 year ago
4
Deepspeed stage3保存模型权重维度为0
#212
Jong-Won
closed
1 year ago
2
大佬好，请问使用lora和ptv2进行微调分别需要修改哪些配置？
#211
mircop1t
opened
1 year ago
19
大佬好，请问关于scheduler
#210
IamRoBota
closed
1 year ago
4
deepspeed如何设置可以避免OOM
#209
lianrzh
opened
1 year ago
2
大佬好，请问下数据构造中的特殊token
#208
IamRoBota
opened
1 year ago
2
数据集
#207
renmengjie7
opened
1 year ago
0
整体微调以后，领域内的知识记住了，但是问常规问题，比如你好，你叫什么，他也回答领域内的知识
#206
heiheiwangergou
opened
1 year ago
5
训练数据集的q和a有长度限制吗，和max_seq_length是什么关系
#205
lancexiao
opened
1 year ago
0
infer_lora_finetunning.py 报错
#204
philipyao
opened
1 year ago
4
大佬 ，能讲一下如何合并lora权重到原来的模型中吗？
#203
cywjava
closed
1 year ago
5
Lora int8微调，推理时出错
#202
crellian
closed
1 year ago
4
整体微调，loss数值训练到第6，7轮左右为nan，各位大佬给看看啥原因
#201
heiheiwangergou
closed
1 year ago
6
lora微调后推理加速
#200
nghuyong
closed
1 year ago
3
关于生成策略
#199
IamRoBota
closed
1 year ago
3
大佬好，请问关于制作数据缓存的问题
#198
IamRoBota
closed
1 year ago
2
使用英文语料训练，可以收敛，但是每个字母就会有个\n提行。啥原因？
#197
leoluopy
closed
1 year ago
3
可否加入训练时间log
#196
magnificent1208
closed
1 year ago
1
不启动lora，加载28层，冻结28层，训练完1个epoch卡住
#195
leoluopy
closed
1 year ago
1
官方都有ptune了为什么还要有这个项目？
#194
19245222
closed
1 year ago
2
labels全为-100是对的吗
#193
feiwuu638
closed
1 year ago
2
训练数据的格式问题
#192
yc930401
closed
1 year ago
2
单卡训练内存爆了，是不是单卡训不了
#191
hkhkq
closed
1 year ago
7
单机多卡如何设置
#190
Frankey419
closed
1 year ago
2
HFValidationError:请问这个错误怎么解决
#189
hkhkq
closed
1 year ago
8
整体ft的时候报错，大家有遇到的嘛 RuntimeError: expected scalar type Half but found Float
#188
heiheiwangergou
closed
1 year ago
1
8G显存 int4可以lora微调吗
#187
fangzhangmnm
closed
1 year ago
1
rebase最新代码报错： RuntimeError: CUDA error: device-side assert triggered
#186
leoluopy
closed
1 year ago
8
如何实现全量推理
#185
magnificent1208
closed
1 year ago
3
除了lora还有没有其他办法可以降低训练时显存的占用量
#184
leoluopy
opened
1 year ago
2
Previous
Next