An instruction tuned version of the Starcoder2 3B model.
30 Pulls Updated 4 months ago
Updated 4 months ago
4 months ago
6e6d18527697 · 3.2GB
model
archstarcoder2
·
parameters3.03B
·
quantizationQ8_0
3.2GB
params
{
"num_ctx": 4096,
"stop": [
"###",
"<|endoftext|>",
"<file_sep>"
81B
template
<|endoftext|>You are an exceptionally intelligent coding assistant that consistently delivers accura
193B
Readme
This is an instruction tuned version of the Starcoder2-3B model.
Fine tuned on the self-alignment training dataset used for StarCoder2-Instruct.