Hugging Face
Search models, datasets, users...
Ctrl K
Models
Datasets
Spaces
Buckets
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
Spaces:
cazyundee
/
training
Like
0
Running
on
Zero
App
Files
Files
Community
Fetching metadata from the HF Docker repository...
main
training
633 kB
Ctrl+K
Ctrl+K
1 contributor
History:
59 commits
cazyundee
fastwide config: 200,206-param recurrent control for the transformer comparison (param-matched within 0.4%); ignore __pycache__
b85df4f
about 13 hours ago
__pycache__
bench_learn: set logp_floor/kl_hard/min_updates attrs expected by train_step after logp_floor fix
about 15 hours ago
docs
RESEARCH_LOG: family ablation verdict (transformer wins both seeds) + param-matched follow-up design
about 13 hours ago
experiments
fix: repetition feature lost in FEN round-trip (train/inference mismatch); greedy-health exp
about 2 months ago
runs
fast_v1: speed-optimized training recipe on the Space
about 14 hours ago
scripts
bench_learn: set logp_floor/kl_hard/min_updates attrs expected by train_step after logp_floor fix
about 15 hours ago
tests
parallel self-play (--workers), conversion experiments, whitelist
about 2 months ago
tinychess
fastwide config: 200,206-param recurrent control for the transformer comparison (param-matched within 0.4%); ignore __pycache__
about 13 hours ago
.gitattributes
Safe
1.52 kB
initial commit
about 2 months ago
.gitignore
Safe
14 Bytes
fastwide config: 200,206-param recurrent control for the transformer comparison (param-matched within 0.4%); ignore __pycache__
about 13 hours ago
README.md
Safe
2 kB
fix: gradio 5 for huggingface_hub>=1.0 compat
about 2 months ago
app.py
Safe
19.8 kB
fix: restore minimal @spaces.GPU smoke test (ZeroGPU refuses to boot without one)
about 16 hours ago
inspect_trace.py
Safe
6.24 kB
tinychess: self-play research substrate (phase 1-4)
about 2 months ago
requirements.txt
Safe
79 Bytes
fix: restore minimal @spaces.GPU smoke test (ZeroGPU refuses to boot without one)
about 16 hours ago
train_selfplay.py
Safe
27.4 kB
--logp-floor: exclude Dirichlet-forced samples (logp_old<-50, 0.9%) from PPO ratio; worst-sample kl 22015->35, max grad 254->45
about 2 months ago