issues
search
hiyouga
/
LLaMA-Factory
Efficiently Fine-Tune 100+ LLMs in WebUI (ACL 2024)
https://arxiv.org/abs/2403.13372
Apache License 2.0
31.67k
stars
3.9k
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
[Draft] Add AutoRound support
#5486
wenhuach21
opened
1 week ago
2
启动 webui失败
#5485
ClementeGao
opened
1 week ago
0
请问DPO训练的时候有什么注意事项吗？我训练出来效果很差。
#5484
zlh-source
opened
1 week ago
4
fix: 修复function call数据集如果 function_call 值的为不合法json，异常提示且中断训练。
#5483
whybeyoung
closed
1 week ago
0
Qlora 报错
#5482
dayuyang1999
closed
1 week ago
1
如何加快微调qwen2-vl-7b后合并的模型在视频上的推理速度？
#5481
J0eky
closed
1 week ago
1
fix ppo_freeze mat1 mat2 should have the same dtype
#5480
ex-yanminmin001
opened
1 week ago
4
qwen2vl-sft后如何将adapter_model.safetensors和模型原始参数合并使用
#5479
YajieW99
closed
1 week ago
1
Can we set default_system in yaml file when training?
#5478
Huarong
closed
1 week ago
1
qwen2vl训练需要修改position_ids问题吗
#5477
sunzjz
closed
1 week ago
2
qwen2-1.5微调训练后tokenizer_config.json中的chat_template值被改了
#5476
czhcc
closed
1 week ago
2
Fix phi-3-small issues
#5475
menibrief
opened
1 week ago
1
训练时template设为empty时，label开头会加上<|EOT|>，之前的版本好像不会这样
#5474
haoranjun
opened
1 week ago
0
Support Mistral format tools
#5473
AlongWY
opened
1 week ago
0
只全参数微调Qwen2-VL-7B-Instruct的visual.merger部分，冻结其他模型参数，训练过程报错
#5472
wjx-sudo
opened
1 week ago
5
多卡微调时报错
#5471
Maydaytyh
closed
1 week ago
2
如何自己编写代码加载合并后的模型推理视频？
#5469
J0eky
closed
1 week ago
1
请问支持多图Qwen2-VL-7B-Instruct微调吗？ 数据格式有示例嚒？
#5468
WorldHellooo
closed
1 week ago
6
如何自动保存checkpoint?
#5467
dayuyang1999
closed
1 week ago
2
webui启动之后框内元素无法渲染
#5466
DSW2001
closed
1 week ago
1
sft do_predict, 生成的json 文件 的 label 都是空
#5465
dayuyang1999
opened
1 week ago
0
请问SFT之后的模型在推理的时候，是否可以返回多个response？
#5464
zlh-source
closed
1 week ago
1
依赖项安装不了，cuda已安装
#5463
DSW2001
closed
1 week ago
2
qwen2_vl模型训练异常
#5462
will-wiki
opened
1 week ago
2
AttributeError: 'Qwen2Attention' object has no attribute 'max_position_embeddings'
#5461
chengchengpei
opened
1 week ago
1
Tips for implementing LlaMa-Factory for new Hardwares
#5460
EtashGuha
opened
1 week ago
0
Do you support for full parameters pre-training?
#5459
lingchensanwen
closed
1 week ago
1
Flatting Packing / maybe fix #5443 and #5426
#5458
AlongWY
opened
1 week ago
8
no such a file or directory of data
#5457
Esmail-ibraheem
opened
1 week ago
0
max pixels argument
#5456
sharonsalabiglossai
opened
1 week ago
1
"Cannot find valid samples" when running DPO on llama3-8b
#5455
zky-kf
closed
1 week ago
3
多卡制定HF_DATASETS_CACHE会报错
#5454
Fu-Dayuan
closed
1 week ago
2
ValueError: Template qwen2 does not exist.
#5453
Oyounger
closed
1 week ago
1
Correctly pass gen_kwarg to eval during model runs
#5451
aliencaocao
opened
1 week ago
0
多机多卡运行报错
#5450
hecheng64
opened
1 week ago
0
qwen2-vl双卡全量微调OOM
#5449
hitsz-zxw
closed
1 week ago
4
对微调后的GLM-4-9B-Chat运行examples/train_lora/llama3_lora_predict.yaml出错
#5447
Twilightsh
opened
2 weeks ago
1
设置随机数种子后，相同数据集和配置的每次训练loss还是不一样
#5446
andy7002
closed
2 weeks ago
2
qizhen
#5445
A-magic
closed
2 weeks ago
0
model.generate的参数在yaml中设定无效，我设了do_sample: false，使用profiler查看实际还是true 此问题只在训练中途的eval发生，训练结束的最后一次eval正常
#5444
aliencaocao
opened
2 weeks ago
0
Running tokenizer on dataset 速度逐渐变慢
#5443
xuyue1112
opened
2 weeks ago
1
bitsandbytes qlora微调模型推理
#5442
oulin1031esti
opened
2 weeks ago
0
help on understanding the implementation of FSDP.
#5441
jq-wei
opened
2 weeks ago
0
如何在 使用 openai 风格 部署时，使用 beam search
#5440
cat-knight
opened
2 weeks ago
0
Llama-factory使用错误
#5439
lifelsl
closed
1 week ago
0
Add qwen_vl to liger kernel supported list
#5438
aliencaocao
closed
2 weeks ago
0
请问使用qlora微调后生成的模型中哪里体现了量化的配置参数
#5437
yangxue-1
closed
2 weeks ago
1
微调后词表长度不一致怎么办
#5436
topology1
opened
2 weeks ago
0
Gemma 2 + unsloth + fa2 full SFT RuntimeError: element 0 of tensors does not require grad and does not have a grad_fn
#5435
hengdos
opened
2 weeks ago
0
请问，llamafactory现在支持在昇腾910上进行模型评估嘛？
#5434
yiyayieryo
opened
2 weeks ago
3
Previous
Next