issues
search
shibing624
/
MedicalGPT
MedicalGPT: Training Your Own Medical GPT Model with ChatGPT Training Pipeline. 训练医疗大模型,实现了包括增量预训练(PT)、有监督微调(SFT)、RLHF、DPO、ORPO。
Apache License 2.0
3.24k
stars
492
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
chatglm2预训练报错expected is_sm80
#225
xpcc355
closed
1 year ago
1
在SFT Lora 训练 ChatGLM2 的时候,在哪里设定 token 长度限制
#224
Droliven
closed
7 months ago
2
flash attention
#223
Vincent131499
closed
7 months ago
3
单机多卡环境,torchrun进行sft时报错RuntimeError: value cannot be converted to type int without overflow
#222
l1905
closed
9 months ago
10
LLM+langchain
#221
YJSoooooo
closed
1 year ago
1
微调后回答不完整
#220
qinyuenlp
closed
1 year ago
0
大佬,rl_training.py中的prompt文字为啥只取偶数位的
#219
laiqinghan
closed
1 year ago
2
baichuan2 13b 多卡pt 进程卡死。单卡可以正常进行
#218
sunshineyg2018
closed
11 months ago
2
baichuan13b 多卡
#217
sunshineyg2018
closed
1 year ago
0
使用baichuan2 pre 错误 torch.distributed.elastic.multiprocessing.errors.ChildFailedError:
#216
sunshineyg2018
closed
7 months ago
6
fix lm_head type changed bug
#215
jiangtann
closed
1 year ago
0
fix lm_head type changed bug
#214
jiangtann
closed
1 year ago
0
改变model.lm_head类型后会导致load时无法找到权重
#213
jiangtann
closed
1 year ago
1
为什么全参sft时需要将模型转为fp32
#212
jiangtann
closed
1 year ago
2
pretrain的peft问题
#211
niunity
closed
1 year ago
1
合并缺少py文件
#210
ZzzybEric
closed
11 months ago
3
大佬好,请问代码目前支持不同的epoch,数据采用不同的shuffle吗
#209
gllary
closed
1 year ago
1
llama2的pt阶段报错
#208
tuqingwen
closed
1 year ago
1
关于sft的超参数设置问题
#207
jiangtann
closed
1 year ago
6
dpo训练
#206
Vincent131499
closed
1 year ago
4
使用Ziya-LLaMA-13B-v1 作为基模型lora sft时爆显存
#205
jiangtann
closed
1 year ago
3
请问什么时候可以支持baichuan2?
#204
erichuazhou
closed
1 year ago
1
Regarding data size
#203
Aniketto16
closed
1 year ago
2
chatglm2增量预训练
#202
MKD1
closed
1 year ago
8
google.protobuf.message.DecodeError: Error parsing message
#201
GravitySaika
closed
10 months ago
1
fix similar to issue #194
#200
kinghuin
closed
1 year ago
1
chatglm2-6b训练时遇到疑惑?
#199
tuqingwen
closed
1 year ago
1
pretain阶段的验证集
#198
rucieryi369
closed
1 year ago
1
pretrain阶段只调整线性层的参数是有什么考虑吗
#197
zhangyu68
closed
1 year ago
3
About validation_file_dir
#196
Billccx
closed
1 year ago
0
使用baichuan13B模型在medical数据集上sft会报错
#195
jiangtann
closed
1 year ago
1
SFT时evaluate会卡住
#194
jiangtann
closed
1 year ago
1
使用deepspeed stage 3全参数SFT加载模型参数为空
#193
jiangtann
closed
1 year ago
2
SFT是否使用了全部的240万条中英文医疗数据集
#192
jiangtann
closed
1 year ago
5
用deepspeed 全参训练chatglm2 报错
#191
wjy3326
closed
1 year ago
1
PT+SFT+DPO显存问题
#190
ZzzybEric
closed
1 year ago
1
关于预训练的超参数设置问题
#189
berton820
closed
1 year ago
1
Doubts about incremental pretraining
#188
Aniketto16
closed
1 year ago
1
继续预训练疑问
#187
tianyunzqs
closed
1 year ago
3
在用llama2指令微调时,train_loss会先下降几个epoch,然后变为0,eval_loss一直为nan
#186
yzw-yzw
closed
1 year ago
1
ChatGLM-6B不能进行RLHF训练和DPO训练吗?为什么到第三阶段都报错。
#185
lk202010
closed
1 year ago
1
关于领域知识增量预训练、指令微调的疑惑
#184
Pycelle
closed
1 year ago
2
sft阶段eval_loss不降反升
#183
invokerbyxv
closed
1 year ago
1
预训练用了哪些数据集,“数据集wiki”和“Dataset”的表述似乎有歧义
#182
Pycelle
closed
1 year ago
2
update dpo pynb
#181
shibing624
closed
1 year ago
0
Dpo
#180
shibing624
closed
1 year ago
0
LORA SFT数据时,两T4卡共30G显存仍报OOM,与您参数介绍不一致
#179
kkcondy
closed
1 year ago
1
llama llama2 訓練程式碼
#178
lucien2712
closed
1 year ago
1
使用deepspeed出现cuda out of memory问题,而且是在代码运行一部分的情况下出现的错误
#177
loki1017
closed
1 year ago
4
run_sft.sh 微调chatglm2-6b报错。
#176
1106301825
closed
1 year ago
1
Previous
Next