This is a port of BlinkDL/RWKV-LM to ggerganov/ggml. Besides the usual FP32, it supports FP16, quantized INT4, INT5 and INT8 inference. This project is focused on CPU, but cuBLAS is also supported.

You've seen some of the model's MLB best bets for Friday. Now, get run line, total and money-line picks for every game here, all from the model that's simulated every game 10,000 times.