Shane Tian
ShaneTian
AI & ML interests
None yet
Recent Activity
updated
a model
24 days ago
deepseek-ai/DeepSeek-V2-Lite
new activity
24 days ago
deepseek-ai/DeepSeek-V2-Lite:Fix for missing blank space at the end of chat template.
new activity
3 months ago
OpenCoder-LLM/opc-annealing-corpus:Question About the Completeness of the Released Dataset
Organizations
None yet
ShaneTian's activity
Fix for missing blank space at the end of chat template.
#9 opened 24 days ago
by
ShaneTian
Question About the Completeness of the Released Dataset
#5 opened 3 months ago
by
ShaneTian
What is the FIM template for the base model?
2
#4 opened 4 months ago
by
ShaneTian
Optimization details
2
#16 opened about 1 year ago
by
ShaneTian
About the compressed file size < 10MB
2
#7 opened about 1 year ago
by
ShaneTian
`CpmTokenizer` is different from the original CPM-1 tokenizer in GitHub
3
#1 opened over 2 years ago
by
ShaneTian
Not found `the-stack-v2-train-extras`
2
#5 opened about 1 year ago
by
ShaneTian
Training loss or logs?
#15 opened about 1 year ago
by
ShaneTian
ctx window & languages?
4
#1 opened over 1 year ago
by
JosephusCheung
Why does Code-Llama-34B not support infilling mode, i.e. FIM
1
#18 opened over 1 year ago
by
ShaneTian
Are there plans to include some models that use OctoPack to fine-tune, like OctoCoder, etc
2
#7 opened over 1 year ago
by
ShaneTian
`CpmTokenizer` is different from the original CPM-1 tokenizer in GitHub
3
#1 opened over 2 years ago
by
ShaneTian