An update to the Apriel series, focusing on using fewer tokens per answer while maintaining performance. They achieved this by using GSPO with length and verbosity penalties.
Specs
Params15B
LicenseMIT
Tags
modelsmodels/llmsmodels/llms/multi-modalmodels/llms/reasoningartifacts/17
RAM score
—
Hugging Face Downloads
189
last 30d
19.8K
all time
HF Likes
303
Relative Adoption Metric contextualizes downloads against the model's size bucket.
VAIL Fingerprint
0034:0046:0070:00a6:00cd:0120:038a:45a7
Explore other models with behavioral similarity to Apriel-1.6-15B-Thinker.
Related Models
