primeodin

primeodin/attention-warrior

From-scratch Transformer attention — a tiny educational head you can run, break, and rebuild. Honest 2021-shaped archive chapter.

0 stars 0 forks 0 watchers Python MIT License
archive attention education educational from-scratch machine-learning neural-networks python teaching transformers
View on GitHub
7 Open Issues Need Help Last updated: Sep 2, 2026

Open Issues Need Help

View All on GitHub
good first issue: --attend CLI demo (readable Q/K/V weights table) 13 days ago
good first issue
primeodin/attention-warrior
0

From-scratch Transformer attention — a tiny educational head you can run, break, and rebuild. Honest 2021-shaped archive chapter.

Python
#archive#attention#education#educational#from-scratch#machine-learning#neural-networks#python#teaching#transformers
good first issue: docs/softmax-stability.md (why subtract max before exp) 13 days ago
good first issue
primeodin/attention-warrior
0

From-scratch Transformer attention — a tiny educational head you can run, break, and rebuild. Honest 2021-shaped archive chapter.

Python
#archive#attention#education#educational#from-scratch#machine-learning#neural-networks#python#teaching#transformers
good first issue: pytest fixture for AttentionHead.attend (weights sum ~1) 14 days ago
good first issue
primeodin/attention-warrior
0

From-scratch Transformer attention — a tiny educational head you can run, break, and rebuild. Honest 2021-shaped archive chapter.

Python
#archive#attention#education#educational#from-scratch#machine-learning#neural-networks#python#teaching#transformers
good first issue: examples/scale-factor-walkthrough.md (why divide by sqrt(d)) 14 days ago
good first issue
primeodin/attention-warrior
0

From-scratch Transformer attention — a tiny educational head you can run, break, and rebuild. Honest 2021-shaped archive chapter.

Python
#archive#attention#education#educational#from-scratch#machine-learning#neural-networks#python#teaching#transformers
Docs: ASCII Q/K/V attention diagram in docs/ 20 days ago
documentation good first issue
primeodin/attention-warrior
0

From-scratch Transformer attention — a tiny educational head you can run, break, and rebuild. Honest 2021-shaped archive chapter.

Python
#archive#attention#education#educational#from-scratch#machine-learning#neural-networks#python#teaching#transformers
Examples: add a second worked attention walkthrough under examples/ 20 days ago
documentation enhancement good first issue
primeodin/attention-warrior
0

From-scratch Transformer attention — a tiny educational head you can run, break, and rebuild. Honest 2021-shaped archive chapter.

Python
#archive#attention#education#educational#from-scratch#machine-learning#neural-networks#python#teaching#transformers
Docs: expand README Try it with exact commands + expected output 20 days ago
documentation good first issue
primeodin/attention-warrior
0

From-scratch Transformer attention — a tiny educational head you can run, break, and rebuild. Honest 2021-shaped archive chapter.

Python
#archive#attention#education#educational#from-scratch#machine-learning#neural-networks#python#teaching#transformers