Commit History

fix: remove Dockerfile - unused in sdk:gradio mode, was blocking build
182d3be
Running

shank Cursor commited on

fix: remove gradio from requirements.txt (HF Spaces pre-installs it, causes pip conflict)
76ccb75

shank Cursor commited on

Change SDK back to gradio
2b5f690

shank commited on

fix: restore Dockerfile and use sdk: docker to comply with hackathon health checks
75b6d51

shank commited on

fix: remove invalid sdk_version from frontmatter
bc780b2

shank commited on

chore: Rename Dockerfile to fix HF Space deploy, and clean up AI slop
13f7b8d

shank commited on

fix: move css to gr.Blocks() for Gradio 5.x compat
2e70173

shank commited on

fix: add sdk_version to Space config to fix scheduling failure
7e7bcac

shank commited on

update: updated readme
f984ce9

shank commited on

Updated readme
de4232f

shank commited on

Update curves and add gitignores
0b1d886

shank commited on

Flatten markdown block backgrounds
154642f

shank commited on

Polish UI header and update github links
b391cd8

shank commited on

Fix UI reward breakdown mapping for offline cache trajectories
bec3d00

shank commited on

Polish Gradio CSS styling and add base Qwen baseline to leaderboard
9f6368b

shank commited on

Update evaluation results, fix Gradio compatibility, and resolve sandbox execution path on macOS
0a80c48

shank commited on

Add --hub-resume and --step-offset to allow resuming directly from HF Hub PEFT checkpoints
8bd8552

shank commited on

Revert speed optimizations to prioritize real-world model quality
cf25957

shank commited on

Optimize training speed for T4 GPUs
2c50d8a

shank commited on

Fix HF_TOKEN scope issue for CheckpointPushCallback
ca28017

shank commited on

Made changes to train_grpo
a07dc7f

shank commited on

Fix links to match correct usernames/orgs
3293f97

shank commited on

Fix checkpoint persistence, add leaderboard and update HF links
e160aa1

shank commited on

fix: use GitHub raw URLs for images so README renders on HF Space
3eb8edc

shank commited on

Added blog post
c02b65b

shank commited on

Added blog post
eacdf84

shank commited on

Added readme again
3165754

shank commited on

Added readme
f6f33cf

shank commited on

Update: Added final imporvements for hackathon
713f336

shank commited on

chore: clean up local dev files and temporary virtual environments
59986c5

shank commited on

chore: clean up local dev files and temporary virtual environments
374c6cc

shank commited on

Update: Triggering the full run
75cd77b

shank commited on

Fix: batch%num_generations math
2b499e7

shank commited on

Cuda returns false fixed
b8172c5

shank commited on

COMPUTE_DRIVE fix
77156dd

shank commited on

Fix: Removed BitsandBytes
bdec91d

shank commited on

Fix: Fixed dependancy issues
db12eaa

shank commited on

Revert "Fix: Dockerfile"
dc7eb3f

shank commited on

Fix: Dockerfile
5dcd156

shank commited on

Fix: Fixed again again
accb271

shank commited on

Fix: Fixed again
9864e61

shank commited on

Fix: Fixing Again
6747185

shank commited on

Fix: Fixing
18b4e8a

shank commited on

Fix: Trying to fix dependency issues
024f3c7

shank commited on

Fix: Fixed file
cb09ef1

shank commited on

fix: serialize bug_metadata as JSON to fix pyarrow mixed-type error
4668456

shank commited on

fix: upgrade bitsandbytes>=0.49.0 (triton.ops), switch to Qwen2.5-Coder-3B
a2fa47a

shank commited on

fix: torch at build time, remove mergekit (conflicts accelerate/peft/trl)
2bfaf77

shank commited on

fix: empty requirements.txt, install training deps at runtime
5d0b2d4

shank commited on

fix: remove wandb - click conflict with gradio causes resolution-too-deep
2005cd2

shank commited on