Upstage logo

SOLAR 10.7B

UpstageOpen WeightsPending Human Review

SOLAR 10.7B is Upstage’s original depth-up-scaled language model, built by expanding a Mistral 7B initialization into a 48-layer dense transformer and continuing pretraining. The release includes a base checkpoint and an instruction-tuned conversational variant, also marketed as Solar Mini. The base weights use Apache 2.0; the current instruction checkpoint specifies CC BY-NC 4.0 because its alignment data includes non-commercial datasets. The documented native context is 4,096 tokens. It supports local text generation, customization, and research without assuming that later hosted Solar Mini revisions are the identical checkpoint.

2023-12-14
10.7B
Decoder-only Transformer with depth up-scaling
Apache-2.0 (base); CC-BY-NC-4.0 (Instruct)

Specifications

Parameters
10.7B
Architecture
Decoder-only Transformer with depth up-scaling
License
Apache-2.0 (base); CC-BY-NC-4.0 (Instruct)
Context Window
4,096 tokens
Type
text
Modalities
text

Benchmark Scores

Advanced Specifications

Model Family
Solar
API Access
Not Available
Chat Interface
Not Available
Variants
SOLAR-10.7B-v1.0SOLAR-10.7B-Instruct-v1.0Solar Mini

Capabilities & Limitations

Capabilities
text generationinstruction followingcontinued pretraininglocal inference
Known Limitations
Instruction weights have non-commercial restrictionsNative context is 4KGenerated factual claims require verification
Notable Use Cases
private text assistantsdomain adaptationlanguage model research

Related Models