`XLMRobertaLoRA.__init__` missing `self.post_init()` causes `AttributeError` with transformers >= 5.0
#62 opened 1 day ago
by 
hfmon
Non-deterministic / all-NaN outputs on CPU: uninitialized torch.empty read in the non-flash forward path
#61 opened 4 days ago
by 
Dashtid
Train Converted XLM-RoBERTa model without FlashAttention installed
#58 opened about 1 year ago
by 
GMedAI
No module named 'transformers_modules.jinaai.xlm-roberta-flash-implementation.2b6bc3f30750b3a9648fe9b63448c09920efe9be.rotary'
1
#57 opened over 1 year ago
by 
PeppePasti
Disable tqdm.
1
#48 opened over 1 year ago
by 
ivanstepanovftw
Trying to convert BAAI/bge-m3
1
#45 opened almost 2 years ago
by 
hiauiarau