openmose Goto Github PK

japanese_llm_eval

A repo for evaluating Japanese LLMs　・　日本語LLMを評価するレポ

nanodsp

Audio Enhancer with Bass amplification using a quadratic curve

rwkv-infctx-trainer-lora

RWKV v5, v6 infctx LoRA trainer with 4bit quantization,Cuda and Rocm supported, for training arbitary context sizes, to 10k and beyond!

rwkv-infer

A large-scale RWKV v6 inference with FLA . Capable of inference by combining multiple states(Pseudo MoE). Easy to deploy on docker. Supports true multi-batch generation and dynamic State switching. CUDA and Rocm Supported :)

rwkv-lm-rlhf

Reinforcement Learning Toolkit for RWKV, aiming for RWKV-6. SFT, Aligning(DPO,ORPO) Let's boost the model's intelligence! currently under construction:)

rwkv-lm-rlhf-dpo-lora

Direct Preference Optimization LoRA for RWKV, aiming for RWKV-5 and 6.

rwkv-lm-state-4bit-orpo

State tuning with Orpo of RWKV v6 can be performed with 4-bit quantization. Every model can be trained with Orpo on Single 24GB GPU!

rwkv.cpp

INT4/INT5/INT8 and FP16 inference on CPU for RWKV language model

rwkv5-lm-lora

RWKV v5,v6 LoRA Trainer on Cuda and Rocm Platform. RWKV is a RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding.

web-rwkv-inspector