Release BibleAI: HF + GGUF + Ollama bundle

README.md

@@ -7,6 +7,11 @@ tags:
 - gemma
 - gguf
 - ollama
+- cpt
+- sft
+- dpo
+language:
+- en
 library_name: transformers
 pipeline_tag: text-generation
 ---
@@ -19,6 +24,17 @@ BibleAI is a Gemma 4 E4B model tuned with a 3-stage pipeline:
 3) DPO (preference optimization),
 for Bible/theology/church-history/faith assistance.
 
+## Canonical Names
+- Hugging Face repository ID: `rhemabible/BibleAI`
+- Model display name: `BibleAI`
+- Ollama model names (recommended):
+  - `bibleaiq8` (Q8 GGUF)
+  - `bibleaibf16` (BF16 GGUF)
+- Preferred local artifact names:
+  - `model.safetensors`
+  - `gguf/final_merged.Q8_0.gguf`
+  - `gguf/final_merged.BF16.gguf`
+
 ## Training Pipeline (CPT -> SFT -> DPO)
 
 ### Stage 1: CPT base (input to this release)
@@ -110,11 +126,22 @@ Note: `tokenizer_config.json` files in this release do not embed a `chat_templat
 - Full checksum manifest: `checksums/sha256.txt`
 
 ## Ollama
+- Use from repo root after cloning this Hugging Face model.
+- These Modelfiles use a local absolute GGUF path. If your local path is different, update the `FROM ...gguf...` line first.
 - Q8:
   `ollama create bibleaiq8 -f ollama/Modelfile.q8`
 - BF16:
   `ollama create bibleaibf16 -f ollama/Modelfile.bf16`
 
+## Transformers (HF)
+```python
+from transformers import AutoTokenizer, AutoModelForCausalLM
+
+repo_id = "rhemabible/BibleAI"
+tokenizer = AutoTokenizer.from_pretrained(repo_id)
+model = AutoModelForCausalLM.from_pretrained(repo_id, torch_dtype="auto", device_map="auto")
+```
+
 ## Included Artifacts
 - root model files (`config.json`, `model.safetensors`, `tokenizer.json`, `tokenizer_config.json`)
 - `gguf/` quantized GGUF exports

docs/PUBLISHING.md

@@ -47,3 +47,7 @@ This pushes:
 - GGUF (`gguf/` BF16 + Q8_0)
 - Ollama Modelfiles (`ollama/`)
 - adapters, checksums, logs, docs
+
+Canonical names:
+- HF repo: `rhemabible/BibleAI`
+- Ollama models: `bibleaiq8` and `bibleaibf16`