Florian Herzog herzogfloria
  • Joined on 2024-10-24
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-10 16:37:33 +02:00
9dc870b8d0 Add 3x-context dataset variant (trainset --radius)
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-10 16:22:42 +02:00
e35d98c2cd Commit training-ready dataset (~6 MB) + DATASET.md usage guide
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-10 16:17:17 +02:00
50f67bcbe0 trainset: GUARANTEE every target triple's entity name is present in input_text
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-10 14:19:03 +02:00
216fd9876c trainset: pick evidence excerpt at the actual provider statement; document provider-currency caveat
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-10 13:52:53 +02:00
991715ab76 Add LLM role-check grounding + labelled training-set pipeline
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-06-09 15:54:40 +02:00
51726f9351 Add 2-replica TP=2 serving for 122B + FP8 KV cache throughput tuning
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-09 13:45:54 +02:00
09798eb27a Add LLM grounding pipeline: current-source fetch, alias + LLM role-check matching
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-06-08 10:29:14 +02:00
b9eaf2df18 Default 122B server to 128k context, add health-check script
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-05 10:34:17 +02:00
00f51859e0 Drop non-extractable custodian relation; add per-triple grounded flag
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-03 11:21:25 +02:00
63e650fa14 Update dataset description with full 2025Q3 build statistics
herzogfloria pushed to main at herzogfloria/fund_rfid_data 2026-06-03 10:37:07 +02:00
1993658fb2 Add SEC fund prospectus -> RDF triple dataset pipeline
herzogfloria created repository herzogfloria/fund_rfid_data 2026-06-03 09:42:49 +02:00
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-03-02 20:59:28 +01:00
030d9f7935 Replace hardcoded username with placeholder in README
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-03-02 20:56:03 +01:00
a5657c3c1f Add dynamic model discovery and improve code extraction in app
a9ed1060cc Fix 122B model download to use Python API instead of huggingface-cli
eff76401ee Add Qwen3.5-122B-A10B-FP8 model support
Compare 3 commits »
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-03-02 18:49:29 +01:00
f4fdaab732 Add Open WebUI integration and enhance Streamlit app
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-03-02 16:43:23 +01:00
d59285fe69 Update student guide with full app.py documentation
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-03-02 16:42:36 +01:00
deee5038d1 Update README to reflect current project state
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-03-02 16:41:07 +01:00
12f9e3ac9b Add LLM parameter controls to sidebar
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-03-02 16:30:16 +01:00
9e1e0c0751 Add Streamlit chat app, update container to vLLM nightly
herzogfloria pushed to main at herzogfloria/LLM_Inferenz_Server_1 2026-03-02 14:43:50 +01:00
076001b07f Add vLLM inference setup for Qwen3.5-35B-A3B on Apptainer