Spaces:

visualisable-ai
/

api

Paused

App Files Files Community

599 kB

Ctrl+K

Ctrl+K

3 contributors

History: 93 commits

gary-boon

Claude Opus 4.5

Add avg_entropy calculation for attention heads

66a46b6 4 months ago

.claude
Add Claude Code configuration 6 months ago
.github
fix: Skip heavy ML deps in CI security checks 4 months ago
backend
Add avg_entropy calculation for attention heads 4 months ago
components
Add layer_stride parameter for PromptDiff optimization 7 months ago
docker
Use NGC PyTorch 24.08 for Python 3.10 compatibility 4 months ago
docs
Add Phase 5: Performance optimizations to phased plan 4 months ago
runs
Phase 1: DGX Spark infrastructure 4 months ago
.env.spark.example

772 Bytes
Update .env.spark.example: TORCH_DTYPE now auto-detected 4 months ago
.gitattributes

1.52 kB
initial commit 8 months ago
.gitignore

499 Bytes
Phase 1: DGX Spark infrastructure 4 months ago
Dockerfile

1.14 kB
Fix Dockerfile: add build-essential for numcodecs compilation 4 months ago
FUTURE_CONSIDERATIONS.md

4.11 kB
Add future considerations doc for response size optimization 4 months ago
README.md

909 Bytes
Update HuggingFace Space description 8 months ago
SECURITY.md

4.45 kB
Add GitHub Actions workflow for security scanning and automated deployment 8 months ago
TESTING.md

4.56 kB
Add Code Llama 7B support with hardware-aware filtering and ICL timeout fixes 6 months ago
TEST_RESULTS.md

6.97 kB
Add Code Llama 7B support with hardware-aware filtering and ICL timeout fixes 6 months ago
app.py

1.69 kB
Fix HuggingFace Spaces cache permission error 8 months ago
explore_vocabulary.py

2.51 kB
Add research attention analysis endpoints with Q/K/V extraction 5 months ago
requirements.txt

605 Bytes
Integrate mistral-common for correct Devstral tokenization 4 months ago
setup-api-key.md

1.32 kB
Add API key authentication 8 months ago
test_instrumentation.py

8.99 kB
Add research attention analysis endpoints with Q/K/V extraction 5 months ago
test_multi_model.py

8.61 kB
Add Code Llama 7B support with hardware-aware filtering and ICL timeout fixes 6 months ago