 Sengxian
|
389fdf93b1
Support 2D position encoding for generation task
|
2 anos atrás |
|
Sengxian
|
b31b6a9c14
Support GLM's 2D positional encoding
|
2 anos atrás |
|
Sengxian
|
c713dae241
Reformat using black
|
2 anos atrás |
|
Sengxian
|
43bada17a4
Fix device
|
2 anos atrás |
|
Aohan Zeng
|
6a6114ccbd
Ethnic evaluation (#28)
|
2 anos atrás |
|
Sengxian
|
e131c8b557
Fix type bugs
|
3 anos atrás |
|
Sengxian
|
605b13758a
Merge remote-tracking branch 'origin/batch-generation' into batch-generation
|
3 anos atrás |
|
Sengxian
|
478eb5c9d0
Fix batch generation bugs
|
3 anos atrás |
 duzx16
|
2948e546b1
Remove max_gen_length argument in generate_text
|
3 anos atrás |
|
Sengxian
|
4f5910ccd2
Merge branch 'batch-generation' of github.com:duzx16/GLM-130B
|
3 anos atrás |
|
Sengxian
|
113f5f1364
Add language modeling task
|
3 anos atrás |
|
Zhengxiao Du
|
223c40b636
Fix BeamSeachStartegy
|
3 anos atrás |
|
Zhengxiao Du
|
bb9fbe4bfc
Implement batch generation
|
3 anos atrás |
|
Sengxian
|
737be7c740
Initial commit
|
3 anos atrás |
