test / app.py

Commit History

Fixed some missing operations to process batches
f8fd25d

fthor commited on

Avoiding CUDA Memory limit by rebatching inputs
b083d4d

fthor commited on

duplicaction test
ed1cd13

fthor commited on

set temperature 0.3
bc91b52

fthor commited on

added flash_attention
3ac1ccb

fthor commited on

added embeddings
854f0cf

fthor commited on

print output in gradio box
dacd4b7

fthor commited on

Added back quantization
a76b117

fthor commited on

first commit
41e6903

fthor commited on