Spaces:

Shivacode
/

debatearena

Sleeping

App Files Files Community

debatearena

Commit History

Apr 26

03:39
docs: add HF notebook viewer link for judges

Shivappa

f0d5acf
02:43
feat: update Cell6 with Run2 loss/rewards, add before_after.py, update README tables

Shivappa

bb2e837
02:07
feat: add 4 analysis plots (loss curve, rubric breakdown, reward distribution, curriculum escalation)

Shivappa

e4696dd

Apr 25

21:24
docs: add BLOG.md + link from README deliverables

Shivappa

728cf5b
21:22
docs: add BLOG.md + link from README deliverables

Shivappa

8d447dc
21:10
docs: strengthen 'why it matters' section with real-world impact + 47x improvement stat

Shivappa

08fcfea
17:58
docs: fill in interpolated reward estimates for SFT steps 10 & 20

Shivappa

35c9ff0
17:55
docs: add Open in Colab badge for judges

Shivappa

612a3b1
17:53
docs: add HF notebook viewer link for judges

Shivappa

68ffc88
17:49
docs: update README with live HF Space, fine-tuned model, real SFT results

Shivappa

114c458
12:59
feat: switch default MODEL_NAME to Shivacode/debate-arena-llama3-8b

Shivappa

2bd286c
12:55
data: update Cell 6 with real SFT reward (0.470) and loss curve from Colab run

Shivappa

a13469a
12:29
data: update Cell 6 with real SFT reward (0.470) and loss curve from Colab run

Shivappa

1c30091
11:46
fix: update ENV_URL to deployed HF Space URL in training notebook

Shivappa

eb35dd8
11:42
fix: update ENV_URL to deployed HF Space URL in training notebook

Shivappa

59930f2
11:37
fix: add HuggingFace Space metadata to README

Shivappa

6b34618
11:35
assets: add reward curve and rubric breakdown charts

Shivappa

e76a783
11:31
Merge remote-tracking branch 'hf/main'

Shivappa

363cefb
10:54
initial commit

Shivacode

5144533
10:21
chore: add .gitignore, remove .venv from tracking

Shivappa Sangapur

8d7f9d8
10:19
feat: DebateArenaEnv complete - server/client separation, LLM multi-agent, Docker, OpenEnv compliant

Shivappa Sangapur

b1eddd6

Commit History

Apr 26

docs: add HF notebook viewer link for judges

feat: update Cell6 with Run2 loss/rewards, add before_after.py, update README tables

feat: add 4 analysis plots (loss curve, rubric breakdown, reward distribution, curriculum escalation)

Apr 25

docs: add BLOG.md + link from README deliverables

docs: add BLOG.md + link from README deliverables

docs: strengthen 'why it matters' section with real-world impact + 47x improvement stat

docs: fill in interpolated reward estimates for SFT steps 10 & 20

docs: add Open in Colab badge for judges

docs: add HF notebook viewer link for judges

docs: update README with live HF Space, fine-tuned model, real SFT results

feat: switch default MODEL_NAME to Shivacode/debate-arena-llama3-8b

data: update Cell 6 with real SFT reward (0.470) and loss curve from Colab run

data: update Cell 6 with real SFT reward (0.470) and loss curve from Colab run

fix: update ENV_URL to deployed HF Space URL in training notebook

fix: update ENV_URL to deployed HF Space URL in training notebook

fix: add HuggingFace Space metadata to README

assets: add reward curve and rubric breakdown charts

Merge remote-tracking branch 'hf/main'

initial commit

chore: add .gitignore, remove .venv from tracking

feat: DebateArenaEnv complete - server/client separation, LLM multi-agent, Docker, OpenEnv compliant