Follow
Understanding INT4 Quantization for Transformer Models: Latency Speedup, Composability, and Failure Cases. (arXiv:2301.12017v2 [cs.CL] UPDATED)
Understanding INT4 Quantization for Transformer Models: Latency Speedup, Composability, and Failure Cases. (arXiv:2301.12017v2 [cs.CL] UPDATED)