Upload patient_files/windy-tier-ALMA-7B-R.md with huggingface_hub

patient_files/windy-tier-ALMA-7B-R.md

@@ -0,0 +1,87 @@
+# 🌪️ Patient File: windy-tier-ALMA-7B-R
+**Generated:** 23 Mar 2026 05:49 UTC
+**Pipeline:** Windy Pro Assembly Line Phase 3
+**Built by:** Kit 0C1 Alpha on Veron-1 (RTX 5090, Mount Pleasant SC)
+
+---
+
+## 📋 Model Information
+
+- **Model Key:** `ALMA-7B-R`
+- **Model ID:** `windy-tier-ALMA-7B-R`
+- **Source Repo:** N/A
+- **Origin:** N/A
+- **License:** CC-BY-4.0
+- **Architecture:** MarianMT (Seq2Seq Transformer)
+
+## 🌍 Language Pair
+
+- **Claimed Direction:** ALMA (`ALMA`) → 7B-R (`7B-R`)
+- **Detected Source Language:** English (`en`)
+- ⚠️ **Note:** Detected source language differs from claimed source language
+
+## 📅 Timeline
+
+- **Source Downloaded:** N/A
+
+## 🔄 Re-Certification History
+
+- **Total Certification Attempts:** 0
+
+## 🔬 Surgery Report — LoRA Variant
+
+### LoRA Configuration (from assembly_line.py)
+
+```python
+LoraConfig(
+    r=4,              # LoRA rank
+    lora_alpha=8,     # Alpha parameter
+    target_modules=["q_proj", "v_proj"],  # Attention projections
+    lora_dropout=0.05,
+    bias="none"
+)
+```
+
+### Weight Modification Analysis
+
+- **Note:** Could not read config.json to estimate parameter counts
+
+**LoRA Merge Status:** Merged back into full model (not separate adapters)
+
+### Model Size Comparison
+
+
+## 📊 Overall Status
+
+- **Status:** ⏳ ERROR
+- **Quality Rating:** N/A
+
+## 🔬 Sweep 1 Results (Initial Certification)
+
+| Variant | Status | Score | Stars | Quality | HF Repo |
+|---------|--------|-------|-------|---------|----------|
+| Base | ⚠️ UNKNOWN | 0/10 |  0 | N/A | N/A |
+| CT2/INT8 | ⚠️ UNKNOWN | 0/10 |  0 | N/A | N/A |
+| LoRA | ⚠️ UNKNOWN | 0/10 |  0 | N/A | N/A |
+
+### Sample Outputs (Sweep 1)
+
+## 🩺 Symptoms
+
+- ✅ No issues detected
+
+## 💡 Hypothesis / Analysis
+
+- ✅ Model successfully certified - all variants meet quality thresholds
+
+---
+*Patient file generated by Windy Pro Patient File Generator v3.0 (Admiral Edition)*
+
+---
+
+## CT2 Safetensors Re-Export (Herm Zero, Dr. B)
+- **Date:** 2026-03-24 ~15:18 UTC
+- **Doctor:** Herm Zero (Dr. B)
+- **Procedure:** Fixed broken pickle INT8 format — re-exported as proper safetensors
+- **Reason:** transformers 4.50+ broke INT8 pickle loader compatibility
+- **Method:** Load base model via MarianMTModel.from_pretrained(), save_pretrained() to ct2/