Add wayneworkman2012/peacebell-v1-148M to the Instruct models (additive, Peacebell-only path)

Hi, I'd like to add **peacebell-v1-148M** to the Instruct models: https://huggingface.co/wayneworkman2012/peacebell-v1-148M

148.55M parameters, trained from scratch, Apache-2.0, ChatML chat model. It is a single-domain model: it only knows World War II and redirects most other topics back to the war. It is a first version and I would like it out there next to the others.

I know you normally take requests as a discussion and add the row yourself. I opened a PR because a row alone would not work for this model, and I wanted to hand you something already tested. Feel free to apply it your own way instead of merging.

**Why a row alone is not enough**

I ran this Space's own code (everything above the UI section of `app.py`) against the model locally, on CPU. It loads fine through `_load_tokenizer` / `_load_model`, but every reply came back as `[Generation Error] 'GPTForCausalLM' object has no attribute 'generate'`. Two reasons, both by design in my repo:
- the model ships its own KV-cached decoder (`stream_generate`) instead of HF `generate()`;
- it has no Jinja `chat_template`: its ChatML control tokens are injected by id exactly as at training time (`tokenizer.build_chatml_ids`), which a string template re-tokenized by SentencePiece cannot reproduce.

**What this PR changes** (`app.py`: 29 lines added, 0 removed or edited; `README.md`: 1 line added)
- one row in `INSTRUCT_MODEL_ROWS` and one line in the README `models:` list;
- `_is_peacebell_spec()` and `_stream_peacebell_text()`, plus a three-line branch in `_stream_single_text` after the shared model load and stop-signal reset. It yields the same cumulative partial text as `_stream_text_from_model`, honours Max tokens and the Stop button, and reports failures in the same `[Generation Error] ...` form, so a problem with this model cannot take down a run.

No new dependency: `sentencepiece` is already in `requirements.txt`.

**Decoding: greedy on purpose.** Peacebell is trained and validated at temperature 0 with no repetition penalty, so this path always decodes that way. The temperature, top-p and repetition-penalty sliders do not apply to it; its answers are deterministic.

**Nothing else is affected.** No existing line is modified and other models never enter the new branch. Checked on CPU with 2 threads, under both transformers 5.14.1 / torch 2.8 and your pinned setup (transformers `main` 5.18.0.dev0, torch 2.14):
- control entrants run through the original and the patched `app.py` with identical seeds give byte-identical output: `HuggingFaceTB/SmolLM2-135M-Instruct` (chat-template path) and `HuggingFaceTB/SmolLM2-135M` (raw-prompt path);
- a mixed three-model run through `_stream_arena_outputs` with Peacebell in the middle works, including the two-model cache eviction;
- Stop ends the Peacebell stream on the next token;
- Peacebell itself: loads in under 2 s, about 0.6 GB in float32, a 113-token answer streams in about 2 s on 2 CPU threads.

Sample replies through the patched code:
- "When did World War II begin and end?" -> "World War II began in 1939 and ended in 1945."
- "Who commanded the Allied forces on D-Day?" -> "General Dwight D. Eisenhower commanded the Allied forces on D-Day."
- "Tell me a joke." -> "I won't do that. We can go over World War II together, if you'd like to explore it."

Thanks for running the arena.

README.md

@@ -90,6 +90,7 @@ models:
 - ThingAI/Quark-50m
 - joelhenwang/OdinNext-138M-Instruct
 - MinimaLabs/KeyLM-75M-Instruct
+- wayneworkman2012/peacebell-v1-148M
 tags:
 - text-generation
 - small-language-model

app.py

@@ -212,6 +212,7 @@ INSTRUCT_MODEL_ROWS = [
     ("Quark-50M", "ThingAI/Quark-50m", "ThingAI", 50.0),
     ("OdinNext-138M-Instruct", "joelhenwang/OdinNext-138M-Instruct", "joelhenwang", 138.0),
     ("KeyLM-75M-Instruct", "MinimaLabs/KeyLM-75M-Instruct", "MinimaLabs", 75.0),
+    ("peacebell-v1-148M", "wayneworkman2012/peacebell-v1-148M", "wayneworkman2012", 148.6),
 ]
 
 
@@ -771,6 +772,31 @@ def _commentary_panel_text(state: dict[str, Any]) -> str:
     return "Commentary is locked until Reveal."
 
 
+def _is_peacebell_spec(spec: ModelSpec) -> bool:
+    return spec.repo_id.lower().startswith("wayneworkman2012/peacebell-")
+
+
+def _stream_peacebell_text(model, tokenizer, prompt: str, max_tokens: int):
+    """Peacebell ships its own ChatML encoder and KV-cached decoder instead of HF generate() and a Jinja
+    chat_template: its control tokens are injected by id, exactly as at training time. It is trained and
+    validated greedy, so it always decodes at temperature 0 with no repetition penalty; the temperature,
+    top-p and repetition-penalty sliders do not apply to it. Max tokens and Stop are honoured."""
+    try:
+        ids = tokenizer.build_chatml_ids([{"role": "user", "content": prompt}], add_generation_prompt=True)
+        input_ids = torch.tensor(ids, dtype=torch.long)
+        out_ids: list[int] = []
+        with torch.inference_mode():
+            for token_id in model.stream_generate(
+                input_ids, max_new_tokens=int(max_tokens), temperature=0.0, no_repeat_ngram_size=0
+            ):
+                if interrupt_callback.stop_signal:
+                    break
+                out_ids.append(int(token_id))
+                yield tokenizer.decode_response(out_ids)
+    except Exception as exc:
+        yield f"[Generation Error] {str(exc) or exc.__class__.__name__}"
+
+
 def _stream_single_text(spec: ModelSpec, prompt: str, max_tokens: int, temperature: float, top_p: float, repetition_penalty: float):
     try:
         tokenizer, model = _get_model(spec)
@@ -779,6 +805,9 @@ def _stream_single_text(spec: ModelSpec, prompt: str, max_tokens: int, temperatu
         return
 
     interrupt_callback.stop_signal = False
+    if _is_peacebell_spec(spec):
+        yield from _stream_peacebell_text(model, tokenizer, prompt, max_tokens)
+        return
     inputs = _move_inputs(_prepare_inputs(tokenizer, prompt, is_instruct=_is_instruct_spec(spec)), model)
     for partial in _stream_text_from_model(
         model,