| | LLM from scratch, part 28 – training a base model from scratch on an RTX 3090 (gilesthomas.com) |
| 3 points by gpjt 5 days ago | past | discuss |
|
| | Why smart instruction-following makes prompt injection easier (gilesthomas.com) |
| 2 points by ibobev 25 days ago | past |
|
| | Writing an LLM from scratch, part 27 – what's left, and what's next? (gilesthomas.com) |
| 1 point by gpjt 34 days ago | past |
|
| | Writing an LLM from scratch, part 26 – evaluating the fine-tuned model (gilesthomas.com) |
| 4 points by gpjt 34 days ago | past |
|
| | Writing an LLM from scratch, part 25 – instruction fine-tuning (gilesthomas.com) |
| 2 points by gpjt 39 days ago | past |
|
| | Writing an LLM from scratch, part 24 – the transcript hack (gilesthomas.com) |
| 1 point by gpjt 40 days ago | past |
|
| | Retro Language Models: Rebuilding Karpathy's RNN in PyTorch (gilesthomas.com) |
| 1 point by ibobev 41 days ago | past |
|
| | Writing an LLM from scratch, part 23 – fine-tuning for classification (gilesthomas.com) |
| 1 point by ibobev 42 days ago | past |
|
| | Retro Language Models: Rebuilding Karpathy's RNN in PyTorch (gilesthomas.com) |
| 3 points by gpjt 44 days ago | past |
|
| | Writing an LLM from scratch, part 23 – fine-tuning for classification (gilesthomas.com) |
| 1 point by gpjt 46 days ago | past |
|
| | Writing an LLM from scratch, part 22 – training our LLM (gilesthomas.com) |
| 254 points by gpjt 53 days ago | past | 10 comments |
|
| | Revisiting Karpathy's 'The Unreasonable Effectiveness of RNNs' (gilesthomas.com) |
| 1 point by ibobev 56 days ago | past |
|
| | Revisiting Karpathy's 'Unreasonable Effectiveness of Recurrent Neural Networks' (gilesthomas.com) |
| 2 points by gpjt 58 days ago | past |
|
| | Writing an LLM from scratch, part 21 – perplexed by perplexity (gilesthomas.com) |
| 1 point by ibobev 60 days ago | past |
|
| | Writing an LLM from scratch, part 21 – perplexed by perplexity (gilesthomas.com) |
| 1 point by gpjt 61 days ago | past |
|
| | Writing an LLM from scratch, part 20 – starting training, and cross entropy loss (gilesthomas.com) |
| 41 points by gpjt 66 days ago | past | 3 comments |
|
| | How Do LLMs Work? (gilesthomas.com) |
| 2 points by gpjt 82 days ago | past | 1 comment |
|
| | How Do LLMs Work? (gilesthomas.com) |
| 1 point by ibobev 83 days ago | past |
|
| | The maths you need to start understanding LLMs (gilesthomas.com) |
| 616 points by gpjt 3 months ago | past | 120 comments |
|
| | What AI chatbots are doing under the hood (gilesthomas.com) |
| 2 points by gpjt 3 months ago | past |
|
| | LLM from scratch, part 18 – residuals, shortcut connections, and the Talmud (gilesthomas.com) |
| 2 points by gpjt 3 months ago | past |
|
| | The fixed length bottleneck and the feed forward network (gilesthomas.com) |
| 1 point by gpjt 3 months ago | past |
|
| | Writing an LLM from scratch, part 17 – the feed-forward network (gilesthomas.com) |
| 8 points by gpjt 3 months ago | past |
|
| | Writing an LLM from scratch, part 16 – layer normalisation (gilesthomas.com) |
| 1 point by gpjt 5 months ago | past |
|
| | Leaving PythonAnywhere (gilesthomas.com) |
| 3 points by gpjt 6 months ago | past |
|
| | Writing an LLM from scratch, part 15 – from context vectors to logits (gilesthomas.com) |
| 7 points by gpjt 6 months ago | past |
|
| | Writing an LLM from scratch, part 14 – the complexity of self-attention at scale (gilesthomas.com) |
| 1 point by gpjt 6 months ago | past |
|
| | Writing an LLM from scratch, part 13 – attention heads are dumb (gilesthomas.com) |
| 351 points by gpjt 7 months ago | past | 67 comments |
|
| | Writing an LLM from scratch, part 12 – multi-head attention (gilesthomas.com) |
| 3 points by gpjt 7 months ago | past |
|
| | Writing an LLM from scratch, part 11 – batches (gilesthomas.com) |
| 2 points by gpjt 7 months ago | past |
|
|
| More |
