Add API order endpoint for external integrations (Soci)

Per-member AI model switching: right-click Type badge to change LLM/NN1/NN2

Add runtime NN model switching: LLM, NN1 (Adilbai), NN2 (RayMelius)

Fix HF Spaces: create /app/data dir for matcher.db after shared_data merge

Seed dashboard price chart with real OHLCV history from Yahoo Finance

Add real ATHEX/NYSE/Euronext prices, OHLCV data for RL agent, merge shared_data into shared/data

Force pip layer rebuild on HF Spaces (v2.0.3 cache bust)

Show NN/LLM badges on CH leaderboard, right-click to switch strategy

Add developer guide PDF with flow diagrams

Fix numpy version conflict: pin numpy<2.0 for stable-baselines3

Pin new pip dependency versions to force Docker layer rebuild

Force Docker rebuild on HF Spaces, add version stamp to startup log

Update README, hybrid default, dynamic strategy switch, CLRH source, regime MDF

Add RL trading strategy, regime-based MDF, dynamic strategy switching

Fix MDF order book depth and CH AI reference price fallback

Use Groq on HF Spaces, prefer fine-tuned Ollama model locally

Update notebook to push full merged model; restore stockex-ch-trader as default

Add script to convert fine-tuned adapter to Ollama GGUF

Fix CH AI trader: use real LLM model and balance BUY/SELL fallback

Remove legacy docs, backup copies, and artifacts

Remove ch_trader_finetune.ipynb, keep only stockex-clearing-house-llm-fine-tuning.ipynb

Replace model merge with direct adapter push to HF Hub

Add AI reasoning history to CH member detail panel

Fix training cell hanging: remove load_best_model_at_end

Fix duplicate FIX orders and lowercase side field

Add member detail panel to CH leaderboard

Fix AI Analyst: use Qwen2.5-7B as default HF model

Add train/eval loss plot after training

Restructure fine-tune notebook: Kaggle+Colab resume, checkpoint cleanup

Fix CH portfolio: increase reload delay and enable auto-refresh

Fix AI Analyst duplicate output and CH trade attribution

Update fine-tune notebook securities to real StockEx symbols

Print resume summary: steps done/remaining, epoch progress, % complete

Add resumable training with Google Drive + HF Hub checkpoints

Fix SFTConfig: move max_seq_length to tokenizer.model_max_length

Fix HF 410: always use router.huggingface.co for all models

Fix notebook for T4/free tier: auto GPU detection + bitsandbytes fix

Upgrade CH fine-tuning to Qwen2.5-32B-Instruct

Add CH trader fine-tuning notebook and update default model

Use RayMelius/stockex-analyst as default LLM for CH AI trader

Add navigation links between Dashboard and Clearing House

Add clearing house simulation with 10 members (USR01-USR10)

Fix HF routing: use router for third-party models, not model-specific endpoint

Add new securities and use real prices in FIX UI order entry

Fix HF inference URL routing for org/model format models

Restore SYSTEM_PROMPT and finetuned model default for AI Analyst

Add dynamic LLM provider/model selection to AI Analyst

Fix Order Entry connect + back arrows

Add Groq as primary LLM provider (free, 14k req/day)

Fix misleading error message: model not available vs providers not enabled