Instructions to use apus-ailab/APUS-Reflex with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use apus-ailab/APUS-Reflex with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("apus-ailab/APUS-Reflex", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Commit History
2 days ago
-
docs(brand): unify APUS Reflex naming across model cards and references
Verified signature5575636
22 days ago
-
Add GGUF / MLX versions and Frozen80 results
Verified signatureaf276da
23 days ago
-
Add TypeSafe SDK serving with verified low/high vLLM routes (#1)
- Add TypeSafe SDK serving with verified low/high vLLM routes (751ac267c5fb9674c4b5980689880165a30bf236)5c6d744 -
docs: replace accuracy chart with APUS full leaderboard (#6)
- docs: replace accuracy chart with APUS full leaderboard (dbb25630720f247b0054ec9f261593d8aa4e0e98)2ddc09b -
Update original accuracy chart to the 1000-question evaluation
Verified signature86dc196 -
Clarify Frozen80 score differences and link author homepage
Verified signatureba923a7 -
Place direct model download links beside performance results
Verified signaturef430a47 -
Place direct model download links beside performance results
Verified signature234de1d -
Place direct model download links beside performance results
Verified signature30dd89f -
Update canonical model series and public download statistics
Verified signature5ee8703 -
Update canonical model series and public download statistics
Verified signature2e393e5 -
Update canonical model series and public download statistics
Verified signatureac18caa -
Fix family download metadata and counted download flows (#2)
- Fix family download metadata and counted download flows (372642eae40abf4992f952ec136b9db5fe178432)b008be4
24 days ago
-
Render technical report equations as vector mathematics
Verified signature675525d -
Highlight compute-aware architecture and organize technical details
Verified signaturec8a3176 -
Highlight compute-aware architecture and organize technical details
Verified signature0bcf299 -
Condense network context into chart and report footnotes
Verified signature20523e7 -
Restore four-model latency chart with updated TypeSafe percentiles
Verified signatureae6460f -
Document candidate-aware projection and update TypeSafe HTTP reference
Verified signature96919c0 -
Synchronize technical report, final HTTP charts and dataset descriptions
Verified signature15ffa87 -
Correct reported 4B HTTP P50 to18ms
Verified signature43054c6 -
Add 4B latency values calculated from reported percentile ratios
Verified signature8cc5c5b -
Clarify high-effort HTTP metrics and single-GPU chart labels
Verified signaturef03f2db -
Verify vLLM usage on GPU and update latency and effort charts
Verified signatureaf85cce -
Add evaluation dataset link and vLLM serving example
Verified signaturec33a214 -
Publish complete APUS-OpenJev-v1 models and Technical Report v1.1
Verified signature68e5880 -
initial commit
Verified signatureb4d6a99