← All concepts

model compression and quantization

1 articles · 4 co-occurring · 0 contradictions · 0 briefs

Article is a direct application of mixed-precision quantization techniques to optimize model size while preserving capability.

Related concepts
vllm ecosystem 1 tool composition patterns 1 inference optimization 1 inference cost optimization 1
Evidence chain (1 articles, showing 1)
@bnjmn_marie: I finally have mixed-precision quantization (2, 3, 4, and 8-bit) working reli... example_of

Article is a direct application of mixed-precision quantization techniques to optimize model size while preserving capability.

query this concept
$ db.articles("model-compression-and-quantization")
$ db.cooccurrence("model-compression-and-quantization")
$ db.contradictions("model-compression-and-quantization")