A RWKV management and startup tool, full automation, only 8MB. And provides an interface compatible with the OpenAI API. RWKV is a large language model that is fully open source and available for commercial use.
What method are you using to connect to the model? If you are sending a request over a LAN, you need to pass --host 0.0.0.0, which you can enable in the client's settings.
Last login: Fri Mar 22 20:40:29 on ttys002 cd /Users/geokf/rmkv/RWKV-Runner.app/Contents/MacOS/../../../ && /Library/Frameworks/Python.framework/Versions/3.10/bin/python3 ./backend-python/main.py --port 8000 --host 127.0.0.1 geokf@MacBook-Air-3 ~ % cd /Users/geokf/rmkv/RWKV-Runner.app/Contents/MacOS/../../../ && /Library/Frameworks/Python.framework/Versions/3.10/bin/python3 ./backend-python/main.py --port 8000 --host 127.0.0.1 --- 0.7619640827178955 seconds --- INFO: Started server process [4246] INFO: Waiting for application startup. cyac not found torch found: /Library/Frameworks/Python.framework/Versions/3.10/lib/python3.10/site-packages/torch/lib torch set INFO: Application startup complete. INFO: Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit) INFO: 127.0.0.1:49503 - "GET / HTTP/1.1" 200 OK INFO: 127.0.0.1:49503 - "GET /status HTTP/1.1" 200 OK Updated Model Config: max_tokens=4100 temperature=1.0 top_p=0.3 presence_penalty=0.0 frequency_penalty=1.0 penalty_decay=None top_k=None global_penalty=None INFO: 127.0.0.1:49503 - "POST /update-config HTTP/1.1" 200 OK Strategy Devices: {'cpu'} state cache disabled RWKV_JIT_ON 1 RWKV_CUDA_ON 0 RESCALE_LAYER 0
Loading /Users/geokf/rmkv/models/RWKV-x060-World-3B-v2-20240228-ctx4096.pth ... Model detected: v6.0 Strategy: (total 32+1=33 layers)
blocks.0.ln1.bias f32 cpu 2560
blocks.0.ln2.weight f32 cpu 2560
blocks.0.ln2.bias f32 cpu 2560
blocks.0.att.time_maa_x f32 cpu 2560
blocks.0.att.time_maa_w f32 cpu 2560
blocks.0.att.time_maa_k f32 cpu 2560
blocks.0.att.time_maa_v f32 cpu 2560
blocks.0.att.time_maa_r f32 cpu 2560
blocks.0.att.time_maa_g f32 cpu 2560
blocks.0.att.time_maa_w1 f32 cpu 2560 160 blocks.0.att.time_maa_w2 f32 cpu 5 32 blocks.0.att.time_decay f32 cpu 40 64 blocks.0.att.time_decay_w1 f32 cpu 2560 64 blocks.0.att.time_decay_w2 f32 cpu 64 2560 blocks.0.att.time_first f32 cpu 40 64 blocks.0.att.receptance.weight f32 cpu 2560 2560 blocks.0.att.key.weight f32 cpu 2560 2560 blocks.0.att.value.weight f32 cpu 2560 2560 blocks.0.att.output.weight f32 cpu 2560 2560 blocks.0.att.gate.weight f32 cpu 2560 2560 blocks.0.att.ln_x.weight f32 cpu 2560
blocks.0.att.ln_x.bias f32 cpu 2560
blocks.0.ffn.time_maa_k f32 cpu 2560
blocks.0.ffn.time_maa_r f32 cpu 2560
blocks.0.ffn.key.weight f32 cpu 2560 8960 blocks.0.ffn.receptance.weight f32 cpu 2560 2560 blocks.0.ffn.value.weight f32 cpu 8960 2560 ........................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ blocks.31.ln1.weight f32 cpu 2560
blocks.31.ln1.bias f32 cpu 2560
blocks.31.ln2.weight f32 cpu 2560
blocks.31.ln2.bias f32 cpu 2560
blocks.31.att.time_maa_x f32 cpu 2560
blocks.31.att.time_maa_w f32 cpu 2560
blocks.31.att.time_maa_k f32 cpu 2560
blocks.31.att.time_maa_v f32 cpu 2560
blocks.31.att.time_maa_r f32 cpu 2560
blocks.31.att.time_maa_g f32 cpu 2560
blocks.31.att.time_maa_w1 f32 cpu 2560 160 blocks.31.att.time_maa_w2 f32 cpu 5 32 blocks.31.att.time_decay f32 cpu 40 64 blocks.31.att.time_decay_w1 f32 cpu 2560 64 blocks.31.att.time_decay_w2 f32 cpu 64 2560 blocks.31.att.time_first f32 cpu 40 64 blocks.31.att.receptance.weight f32 cpu 2560 2560 blocks.31.att.key.weight f32 cpu 2560 2560 blocks.31.att.value.weight f32 cpu 2560 2560 blocks.31.att.output.weight f32 cpu 2560 2560 blocks.31.att.gate.weight f32 cpu 2560 2560 blocks.31.att.ln_x.weight f32 cpu 2560
blocks.31.att.ln_x.bias f32 cpu 2560
blocks.31.ffn.time_maa_k f32 cpu 2560
blocks.31.ffn.time_maa_r f32 cpu 2560
blocks.31.ffn.key.weight f32 cpu 2560 8960 blocks.31.ffn.receptance.weight f32 cpu 2560 2560 blocks.31.ffn.value.weight f32 cpu 8960 2560 ln_out.weight f32 cpu 2560
ln_out.bias f32 cpu 2560
head.weight f32 cpu 2560 65536 Updated Model Config: max_tokens=4100 temperature=1.0 top_p=0.3 presence_penalty=0.0 frequency_penalty=1.0 penalty_decay=None top_k=None global_penalty=None INFO: 127.0.0.1:49521 - "POST /update-config HTTP/1.1" 200 OK zsh: killed /Library/Frameworks/Python.framework/Versions/3.10/bin/python3 --port 8000
geokf@MacBook-Air-3 rmkv %
Hi, I'm trying to connect to the model, but failed. Anyone able to help here? Thanks