|
--- |
|
license: mit |
|
base_model: rhysjones/gpt2-774M-fineweb-150B |
|
datasets: |
|
- HuggingFaceFW/fineweb |
|
widget: |
|
- text: >- |
|
<|user|><|content|>Hello! Who are you?<|end_turn|> |
|
|
|
<|assistant|><|content|>Hello! I am ChatBot, a large language model. I am |
|
here to help you with information, answer questions, provide |
|
recommendations, and assist with a variety of tasks. How can I assist you |
|
today?<|end_turn|> |
|
|
|
<|user|><|content|>Write a message asking a friend to come to my wedding next month. I want it to be very short and casual, but with an option to opt out.<|end_turn|> |
|
<|assistant|><|content|> |
|
--- |
|
|
|
This is [karpathy's](https://github.com/karpathy) model from [the llm.c project](https://github.com/karpathy/llm.c/discussions/580) converted to HF format to investigate [bfloat16 performance](https://github.com/karpathy/llm.c/pull/571). |
|
|
|
The training run was 150B tokens, 1.5 epochs over the 100B FineWeb sample dataset. |
|
|
|
There's active work underway at [https://github.com/karpathy/llm.c](https://github.com/karpathy/llm.c) so I'd suggest following the developments there! |
