Spaces:

DevikaJ2005
/

fraudshield-1

Sleeping

App Files Files Community

fraudshield-1

Commit History

Apr 26

11:18
Update HF_BLOG_DRAFT.md

DevikaJ2005

Verified signature
3ff6cb8
11:16
Rewrite HF blog draft as a story-led case study

DevikaJ2005

88d1015
10:47
Align local policy prompts and canonical action targets

DevikaJ2005

45cf526
08:57
Clean Colab notebook artifacts before training

DevikaJ2005

f056f9f
08:33
Fix evaluation artifacts and align notebook outputs

DevikaJ2005

a023c6a
08:01
Fix evaluation to compare heuristic vs local trained model

DevikaJ2005

8c34447
07:53
Fix cleaned Colab training notebook

DevikaJ2005

6893050
07:42
Update agent selection for HF-hosted and local LLM paths

DevikaJ2005

eaf73e8
07:44
Add training-first RL architecture with tracking

DevikaJ2005

e97acd1
07:15
Simplify explorer UI for non-technical reviewers

DevikaJ2005

9672a3e
06:46
Created using Colab

DevikaJ2005

1a53abb
06:27
Add training-first RL architecture with tracking

DevikaJ2005

ce9edc2
03:28
Add public-data curriculum and harden LLM agent

DevikaJ2005

ccd0934

Apr 25

22:03
Stabilize Colab training notebook with SFTTrainer

DevikaJ2005

4049a11
18:16
Fix Colab notebook path and prompt string

DevikaJ2005

3c0133a
15:20
Finalize RL-first environment and explorer UI

DevikaJ2005

30533d1

Apr 23

04:32
Reframe FraudShield as enterprise FraudOps workflow

DevikaJ2005

1510f7f

Apr 9

05:21
Harden observation leakage and calibrate baseline

DevikaJ2005

4a01d54

Apr 8

15:29
Upgrade baseline with agentic investigation policy

DevikaJ2005

5acc2d6
15:06
Fix Hugging Face metadata emoji

DevikaJ2005

61396fd
15:02
Add agentic investigation workflow and runtime metadata

DevikaJ2005

338316c
14:12
Clamp task scores to strict open interval

DevikaJ2005

df73ff6
13:51
Use injected proxy credentials for baseline agent

DevikaJ2005

723cc80
07:34
Fix inference stdout protocol for submission validator

DevikaJ2005

29faa15
06:46
Fix: Fallback to heuristic when only one LLM env var is set; improve error handling

DevikaJ2005

a7a2527
04:58
Docs: Add comprehensive submission checklist for OpenEnv hackathon

DevikaJ2005

7c64820
04:48
Fix: Add START/STEP/END structured log format to inference.py

DevikaJ2005

52dde71
04:42
Fix: Add complete HF Spaces metadata

DevikaJ2005

50ef6b4
04:37
Fix: Remove app_file from Docker Space metadata

DevikaJ2005

c380989
04:28
Fix: Add HF Spaces YAML metadata to README

DevikaJ2005

37e25c4

Apr 7

12:32
Trigger HF Space rebuild

DevikaJ2005

957799d
11:46
Docs: Comprehensive professional documentation and validation suite

DevikaJ2005

af4e958
11:24
Cleanup: Remove result artifacts from git, update .gitignore

DevikaJ2005

7db0370

Mar 31

07:50
Support HF-safe Space variable names

DevikaJ2005

dbee4da
06:45
Finalize OpenEnv submission readiness

DevikaJ2005

c2dd900

Mar 30

18:20
Refine FraudShield snapshot pipeline

DevikaJ2005

e4ffe61
17:53
Prepare FraudShield for OpenEnv submission

DevikaJ2005

ea9eade
13:04
Initial commit - FraudShield OpenEnv environment

DevikaJ2005

3818a51

Mar 28

07:07
first commit

DevikaJ2005

07de0f7

Commit History

Apr 26

Update HF_BLOG_DRAFT.md

Rewrite HF blog draft as a story-led case study

Align local policy prompts and canonical action targets

Clean Colab notebook artifacts before training

Fix evaluation artifacts and align notebook outputs

Fix evaluation to compare heuristic vs local trained model

Fix cleaned Colab training notebook

Update agent selection for HF-hosted and local LLM paths

Add training-first RL architecture with tracking

Simplify explorer UI for non-technical reviewers

Created using Colab

Add training-first RL architecture with tracking

Add public-data curriculum and harden LLM agent

Apr 25

Stabilize Colab training notebook with SFTTrainer

Fix Colab notebook path and prompt string

Finalize RL-first environment and explorer UI

Apr 23

Reframe FraudShield as enterprise FraudOps workflow

Apr 9

Harden observation leakage and calibrate baseline

Apr 8

Upgrade baseline with agentic investigation policy

Fix Hugging Face metadata emoji

Add agentic investigation workflow and runtime metadata

Clamp task scores to strict open interval

Use injected proxy credentials for baseline agent

Fix inference stdout protocol for submission validator

Fix: Fallback to heuristic when only one LLM env var is set; improve error handling

Docs: Add comprehensive submission checklist for OpenEnv hackathon

Fix: Add START/STEP/END structured log format to inference.py

Fix: Add complete HF Spaces metadata

Fix: Remove app_file from Docker Space metadata

Fix: Add HF Spaces YAML metadata to README

Apr 7

Trigger HF Space rebuild

Docs: Comprehensive professional documentation and validation suite

Cleanup: Remove result artifacts from git, update .gitignore

Mar 31

Support HF-safe Space variable names

Finalize OpenEnv submission readiness

Mar 30

Refine FraudShield snapshot pipeline

Prepare FraudShield for OpenEnv submission

Initial commit - FraudShield OpenEnv environment

Mar 28

first commit