Hardly. I've played a lot with the 7,13, and 30B llamas as well as the 7 and 13B alpacas fine tuned by Stanford. They do not have emergent abilities like being able to generate rhymes or, say, represent a movie plot as emoji. Even openai's old text-davinci-003 (gpt3.5, but text completion, not the chat ones) far outperforms them. That said, I have hopes for a 65B 3-bit quantized alpaca-fine tuned. We'll see when someone spends the money to do the (more costly) 65B training. The alpacas are also much more likely to go off rails and start regurgitating their fine-tuning inputs. Either that or openai is doing a lot of post processing on their end to hide the same problems in their LLM.

For now my IRC bots run the alpaca 7B 4-bit. 13B was not a significant improvement for twice the computational time. But it's best to learn them now because as soon as openai gets sued for the first time all the turing test passing older models without the legal-butt-covering bolted on will be removed.

Where does one find the 13B alpaca model?

superkuh

Be aware this file is a single ~8GB 4-bit model (ggml-alpaca-13b-q4.bin) instead of the 2x ~4GB models (ggml-model-q4_0.bin, ggml-model-q4_0.bin.1) that most llama.cpp style inference running programs expect. You'll probably have to edit the line,

    n_parts = LLAMA_N_PARTS.at(hparams.n_embd);

in chat.cpp (or main.cpp) to hard code it to treat this 1 file model properly like,

    n_parts = 1;

Or re-write the parameter config subroutine to recognize and handle non-standard weights file.

magnet: magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A6969%2Fannounce&tr=udp%3A%2F%2F9.rarbg.com%3A2810%2Fannounce

torrent: https://btcache.me/torrent/053B3D54D2E77FF020EBDDF51DAD681F2...

torrent: https://torrage.info/torrent.php?h=053b3d54d2e77ff020ebddf51...

via: https://github.com/antimatter15/alpaca.cpp