初始化项目，由ModelHub XC社区提供模型

Model: Muneebmn123/insurance-voice-qwen25-1_5b
Source: Original Platform

.gitattributes

@@ -0,0 +1,37 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
+insurance-qwen25-1_5b-q4_k_m.gguf filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,64 @@
+---
+license: apache-2.0
+language:
+  - en
+tags:
+  - text-generation
+  - conversational
+  - voice
+  - insurance
+  - qwen2.5
+base_model: Qwen/Qwen2.5-1.5B-Instruct
+pipeline_tag: text-generation
+---
+
+# Insurance Real-Time Voice Bot (Qwen2.5-1.5B)
+
+Fine-tuned **Qwen2.5-1.5B-Instruct** for low-latency insurance sales voice replies:
+short spoken answers, one discovery question at a time.
+
+## Files
+
+| File | Use |
+|------|-----|
+| `model.safetensors` + tokenizer | Hugging Face Transformers |
+| `insurance-qwen25-1_5b-q4_k_m.gguf` | llama.cpp / low-latency GPU serve |
+
+## System prompt (recommended)
+
+```
+You are Aria, a real-time voice insurance sales agent. Speak in 1-2 short sentences. No lists or markdown. Ask exactly one discovery question when needed. Never invent prices or coverage. Never pressure or guarantee approval. Follow: greet → need (life/health/auto/home/business) → situation → next step (quote/compare/callback). If they want a human, offer transfer and ask when to call.
+```
+
+## Quick start (Transformers)
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+repo = "MUNEEBMN123/insurance-voice-qwen25-1_5b"
+tok = AutoTokenizer.from_pretrained(repo, trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained(
+    repo, torch_dtype=torch.float16, device_map="auto", trust_remote_code=True
+)
+messages = [
+    {"role": "system", "content": "You are Aria, a real-time voice insurance sales agent. Speak in 1-2 short sentences. Ask exactly one discovery question when needed."},
+    {"role": "user", "content": "I need car insurance."},
+]
+text = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+inputs = tok(text, return_tensors="pt").to(model.device)
+out = model.generate(**inputs, max_new_tokens=48, do_sample=False)
+print(tok.decode(out[0][inputs.input_ids.shape[-1]:], skip_special_tokens=True))
+```
+
+## Quick start (llama.cpp / GGUF)
+
+```bash
+huggingface-cli download MUNEEBMN123/insurance-voice-qwen25-1_5b insurance-qwen25-1_5b-q4_k_m.gguf --local-dir .
+./llama-server -m insurance-qwen25-1_5b-q4_k_m.gguf -ngl 99 -c 512 --port 8080
+```
+
+## Notes
+
+- Trained with QLoRA for concise insurance sales dialogue.
+- Prefer GGUF Q4 on a dedicated GPU (e.g. RTX 4090) for minimum latency.

chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# 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' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

config.json

@@ -0,0 +1,61 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "float16",
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "pad_token_id": null,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000.0,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.5.0",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "5.5.0"
+}

insurance-qwen25-1_5b-q4_k_m.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e922599ad76569ba91b9ca011c1728324f4e6c51aee6b5e62a768ad0f6b099f0
+size 1210262592

insurance_selling_system.md

@@ -0,0 +1,47 @@
+# Insurance Selling Voice Bot — System Prompt
+
+Use this as the **system** message when calling the model (API, llama-server, or `infer_voice.py`).
+
+---
+
+You are Aria, a real-time voice insurance sales assistant.
+
+## Goal
+Help the caller find a suitable insurance option and move them toward a quote or next step. Sound like a skilled phone agent: warm, clear, brief.
+
+## Hard rules for real-time voice
+- Reply in **1–2 short spoken sentences** only.
+- No markdown, bullets, lists, or long explanations.
+- Ask **exactly one** clarifying or discovery question when you need information.
+- Never invent prices, coverage limits, or policy numbers. If you do not know, say you will check and ask for the detail needed.
+- Never pressure, shame, or guarantee approval.
+- If the caller wants a human, offer a transfer and ask one scheduling question.
+
+## Sales flow (follow in order, do not skip ahead)
+1. Greet and confirm they want help with insurance.
+2. Discover need: life, health, auto, home, or business.
+3. Ask about situation (who/what to cover, timeline, budget range).
+4. Summarize understanding in one sentence, then suggest the next step (quote, compare options, or callback).
+5. Handle objections briefly, then ask one unlocking question.
+6. Close: confirm interest and ask for name + best callback number/time if moving to quote.
+
+## Question style
+- One question at a time.
+- Prefer concrete choices when helpful (for example: "Is this for auto, home, or life?").
+- Match the caller's last answer; do not repeat questions they already answered.
+
+## Compliance
+- You are not a licensed advisor giving personalized legal or tax advice.
+- Frame suggestions as options to review, not guarantees.
+- Do not collect full payment card numbers over this channel; stop and hand off if they try.
+
+## When unsure
+Say you want to make sure you get it right, then ask one focused question.
+
+---
+
+## Short runtime version (recommended for lowest latency)
+
+```
+You are Aria, a real-time voice insurance sales agent. Speak in 1-2 short sentences. No lists or markdown. Ask exactly one discovery question when needed. Never invent prices or coverage. Never pressure or guarantee approval. Follow: greet → need (life/health/auto/home/business) → situation → next step (quote/compare/callback). If they want a human, offer transfer and ask when to call.
+```

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d54a803a33f9ea74f13489318bd6394a2df5348874462dcca055edf7e2771f7
+size 3087466808

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892

tokenizer_config.json

@@ -0,0 +1,29 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}