Downloads · 30 days
12
3% of all-time downloads
deqing/llama-isolate-old
llama-isolate-old is a text generation model from deqing. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as mit.
A 300M-parameter language model trained from scratch on FineWeb-Edu 10BT (~9.4B tokens, 1 epoch) as part of the Convergent Evolution project, which investigates how Fourier features emerge in LLM number embeddings.
Downloads · 30 days
12
3% of all-time downloads
All-time downloads
384
Public
Parameters
320M
1.3 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors1.3 GB · 99%
From the Hugging Face model README
A 300M-parameter language model trained from scratch on FineWeb-Edu 10BT (~9.4B tokens, 1 epoch) as part of the Convergent Evolution project, which investigates how Fourier features emerge in LLM number embeddings.
| Architecture | LLaMA-style Transformer (12 layers, 1024 hidden, 16 heads, GQA) |
| Parameters | ~300M |
| Optimizer | Muon (for 2D weights) + AdamW (for embeddings/bias/norm) |
| Data perturbation | block-diagonal attention mask (numbers cannot attend to context) |
| Training data | FineWeb-Edu sample-10BT (~9.4B tokens) |
| Context length | 1024 |
| Tokenizer | Llama 3 (128K vocab) |
| Batch size | 512 sequences |
Intermediate checkpoints are saved as branches: tokens-200M, tokens-400M, ..., tokens-9.6B.
from transformers import AutoModelForCausalLM
# Load final checkpoint
model = AutoModelForCausalLM.from_pretrained("deqing/convergent-llama-300M-muon-isolate")
# Load intermediate checkpoint (e.g., at 1B tokens)
model = AutoModelForCausalLM.from_pretrained("deqing/convergent-llama-300M-muon-isolate", revision="tokens-1B")
Paper forthcoming.