[N] Falcon180B released! Sadly without Apache 2.0 they made their own modified version. :(

LocalLLaMA discussion: [https://www.reddit.com/r/LocalLLaMA/comments/16bjdmd/falcon180b\_authors\_open\_source\_a\_new\_180b\_version/](https://www.reddit.com/r/LocalLLaMA/comments/16bjdmd/falcon180b_authors_open_source_a_new_180b_version/)

Announcement: [https://falconllm.tii.ae/falcon-models.html](https://falconllm.tii.ae/falcon-models.html)

HF Model: [https://huggingface.co/tiiuae/falcon-180B](https://huggingface.co/tiiuae/falcon-180B)

Demo: [https://huggingface.co/spaces/tiiuae/falcon-180b-demo](https://huggingface.co/spaces/tiiuae/falcon-180b-demo)

Blog: [https://huggingface.co/blog/falcon-180b](https://huggingface.co/blog/falcon-180b)

* 180 Billion parameters
* Trained on 3.5 trillion tokens
* Available for research and commercial usage
* Claims similar performance to Bard, slightly below gpt4

[https://falconllm.tii.ae/terms-and-conditions.html](https://falconllm.tii.ae/terms-and-conditions.html)

[https://falconllm.tii.ae/acceptable-use-policy.html](https://falconllm.tii.ae/acceptable-use-policy.html)

bard billion commercial gpt4 machinelearning performance research tokens usage