Add enable_thinking chat template, auto threads; true fast mode

Dockerfile

@@ -4,9 +4,7 @@ FROM python:3.11-slim
 ENV DEBIAN_FRONTEND=noninteractive \
     PYTHONUNBUFFERED=1 \
     HF_HOME=/home/user/.cache/huggingface \
-    HF_HUB_ENABLE_HF_TRANSFER=1 \
-    NUM_THREADS=2 \
-    OMP_NUM_THREADS=2
+    HF_HUB_ENABLE_HF_TRANSFER=1
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
         curl ca-certificates libgomp1 \
@@ -32,6 +30,7 @@ RUN pip install --no-cache-dir \
 
 WORKDIR /home/user/app
 COPY app.py .
+COPY chat_template.jinja .
 RUN chown -R user:user /home/user
 
 USER user

app.py

@@ -13,8 +13,12 @@ GGUF_FILE   = os.environ.get("GGUF_FILE", "Qwen3.5-0.8B.Q4_K_M.gguf")
 HF_TOKEN    = os.environ.get("HF_TOKEN")  # optional; model is public
 PORT        = int(os.environ.get("PORT", "7860"))
 LLAMA_PORT  = int(os.environ.get("LLAMA_PORT", "8080"))
-NUM_THREADS = os.environ.get("NUM_THREADS", "2")
+# Use every available core by default (cpu-basic = 2 vCPU); override with NUM_THREADS.
+NUM_THREADS = os.environ.get("NUM_THREADS") or str(os.cpu_count() or 2)
 BIN         = "/opt/llamabin"
+# Custom chat template that adds an enable_thinking toggle (the model's own template
+# ALWAYS opens a <think> block; this lets "Fast (no thinking)" actually skip reasoning).
+CHAT_TEMPLATE_FILE = os.environ.get("CHAT_TEMPLATE_FILE", "/home/user/app/chat_template.jinja")
 LLAMA       = f"http://127.0.0.1:{LLAMA_PORT}"
 # Qwen3.5-0.8B native context = 262144. On CPU-basic we keep a smaller window for fast
 # startup + low memory; raise N_CTX via a Space variable if you need long context.
@@ -42,11 +46,14 @@ def start_llama(model_path):
         "-m", model_path, "--host", "127.0.0.1", "--port", str(LLAMA_PORT),
         "-c", N_CTX, "-t", NUM_THREADS, "-b", "256", "--no-mmap",
         "--parallel", os.environ.get("PARALLEL", "1"),
-        # Use the model's OWN embedded chat template (Qwen3.5 ChatML). No injection.
+        # Apply our chat template (adds enable_thinking toggle for true fast mode).
         "--jinja",
-        # Separate the <think> chain-of-thought into reasoning_content so the answer stays clean.
-        "--reasoning-format", "auto",
     ]
+    if CHAT_TEMPLATE_FILE and os.path.exists(CHAT_TEMPLATE_FILE):
+        cmd += ["--chat-template-file", CHAT_TEMPLATE_FILE]
+    # Separate the <think> chain-of-thought into reasoning_content so the answer stays clean.
+    cmd += ["--reasoning-format", "auto"]
+    print(f"[init] cpu_count={os.cpu_count()} threads={NUM_THREADS}", flush=True)
     # Optional extra flags (e.g. "-fa on") via LLAMA_EXTRA_ARGS, space separated.
     extra = os.environ.get("LLAMA_EXTRA_ARGS", "").split()
     if extra:

chat_template.jinja

@@ -0,0 +1,90 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<|im_start|>assistant\n<think>\n\n</think>\n\n' }}
+    {%- else %}
+        {{- '<|im_start|>assistant\n<think>\n' }}
+    {%- endif %}
+{%- endif %}