diff --git a/Season2.step_into_llm/17.Qwen/CLI_input_mock_qwen.py b/Season2.step_into_llm/17.Qwen/CLI_input_mock_qwen.py
new file mode 100644
index 0000000..13198fa
--- /dev/null
+++ b/Season2.step_into_llm/17.Qwen/CLI_input_mock_qwen.py
@@ -0,0 +1,48 @@
+import mindspore
+import numpy as np
+from mindspore import dtype as mstype
+import mindspore.ops as ops
+from mindspore import Tensor
+from mindnlp.transformers import AutoTokenizer, AutoModelForCausalLM
+import faulthandler
+
+faulthandler.enable()
+
+model_id = "Qwen/Qwen1.5-0.5B-Chat"
+tokenizer = AutoTokenizer.from_pretrained(model_id, mirror='modelscope')
+model = AutoModelForCausalLM.from_pretrained(
+ model_id,
+ ms_dtype=mindspore.float16,
+ mirror='modelscope'
+)
+
+messages = [
+ {"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"},
+ {"role": "user", "content": "Who are you?"},
+]
+
+input_ids = tokenizer.apply_chat_template(
+ messages,
+ add_generation_prompt=True,
+ return_tensors="ms"
+)
+attention_mask = Tensor(np.ones(input_ids.shape), mstype.float32)
+
+terminators = [
+ tokenizer.eos_token_id,
+ tokenizer.convert_tokens_to_ids("<|endoftext|>")
+]
+outputs = model.generate(
+ input_ids,
+ attention_mask=attention_mask,
+ max_new_tokens=20,
+ eos_token_id=terminators,
+ do_sample=False,
+ # do_sample=True,
+ # temperature=0.6,
+ # top_p=0.9,
+)
+response = outputs[0][input_ids.shape[-1]:]
+print(outputs)
+print(tokenizer.decode(response, skip_special_tokens=True))
+
diff --git a/Season2.step_into_llm/17.Qwen/GUI_gradio-qwen1.5.py b/Season2.step_into_llm/17.Qwen/GUI_gradio-qwen1.5.py
new file mode 100644
index 0000000..1ce17a2
--- /dev/null
+++ b/Season2.step_into_llm/17.Qwen/GUI_gradio-qwen1.5.py
@@ -0,0 +1,61 @@
+import gradio as gr
+import mindspore
+from mindspore import dtype as mstype
+import numpy as np
+from mindnlpv041.mindnlp.transformers import AutoModelForCausalLM, AutoTokenizer
+from mindnlpv041.mindnlp.transformers import TextIteratorStreamer
+from threading import Thread
+
+# Loading the tokenizer and model from Hugging Face's model hub.
+tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen1.5-0.5B-Chat", ms_dtype=mindspore.float16)
+model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen1.5-0.5B-Chat", ms_dtype=mindspore.float16)
+
+system_prompt = "You are a helpful and friendly chatbot"
+
+def build_input_from_chat_history(chat_history, msg: str):
+ messages = [{'role': 'system', 'content': system_prompt}]
+ for user_msg, ai_msg in chat_history:
+ messages.append({'role': 'user', 'content': user_msg})
+ messages.append({'role': 'assistant', 'content': ai_msg})
+ messages.append({'role': 'user', 'content': msg})
+ return messages
+
+# Function to generate model predictions.
+def predict(message, history):
+ # Formatting the input for the model.
+ messages = build_input_from_chat_history(history, message)
+ input_ids = tokenizer.apply_chat_template(
+ messages,
+ add_generation_prompt=True,
+ return_tensors="ms",
+ tokenize=True
+ )
+ attention_mask = mindspore.Tensor(np.ones(input_ids.shape), mstype.float32)
+ streamer = TextIteratorStreamer(tokenizer, timeout=300, skip_prompt=True, skip_special_tokens=True)
+ generate_kwargs = dict(
+ input_ids=input_ids,
+ streamer=streamer,
+ max_new_tokens=1024,
+ do_sample=True,
+ top_p=0.9,
+ temperature=0.1,
+ num_beams=1,
+ attention_mask=attention_mask,
+ )
+ t = Thread(target=model.generate, kwargs=generate_kwargs)
+ t.start() # Starting the generation in a separate thread.
+ partial_message = ""
+ for new_token in streamer:
+ partial_message += new_token
+ if '' in partial_message: # Breaking the loop if the stop token is generated.
+ break
+ yield partial_message
+
+
+# Setting up the Gradio chat interface.
+gr.ChatInterface(predict,
+ title="Qwen1.5-0.5b-Chat",
+ description="问几个问题",
+ examples=['你是谁？', '介绍一下华为公司']
+ ).launch(share=True, server_name='0.0.0.0', server_port=7860) # Launching the web interface.
+
diff --git a/Season2.step_into_llm/17.Qwen/qwen2_finetune_inference.ipynb b/Season2.step_into_llm/17.Qwen/qwen2_finetune_inference.ipynb
new file mode 100644
index 0000000..9138541
--- /dev/null
+++ b/Season2.step_into_llm/17.Qwen/qwen2_finetune_inference.ipynb
@@ -0,0 +1,1764 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## 环境配置"
+ ]
+ },
+ {
+ "attachments": {},
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "> 此为在线运行平台配置python3.9的指南，如在其他环境平台运行案例，请根据实际情况修改如下代码"
+ ]
+ },
+ {
+ "attachments": {},
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "第一步：设置python版本为3.9.0"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "%%capture captured_output\n",
+ "!/home/ma-user/anaconda3/bin/conda create -n python-3.9.0 python=3.9.0 -y --override-channels --channel https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main\n",
+ "!/home/ma-user/anaconda3/envs/python-3.9.0/bin/pip install ipykernel"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "import json\n",
+ "import os\n",
+ "\n",
+ "data = {\n",
+ " \"display_name\": \"python-3.9.0\",\n",
+ " \"env\": {\n",
+ " \"PATH\": \"/home/ma-user/anaconda3/envs/python-3.9.0/bin:/home/ma-user/anaconda3/envs/python-3.7.10/bin:/modelarts/authoring/notebook-conda/bin:/opt/conda/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/home/ma-user/modelarts/ma-cli/bin:/home/ma-user/modelarts/ma-cli/bin\"\n",
+ " },\n",
+ " \"language\": \"python\",\n",
+ " \"argv\": [\n",
+ " \"/home/ma-user/anaconda3/envs/python-3.9.0/bin/python\",\n",
+ " \"-m\",\n",
+ " \"ipykernel\",\n",
+ " \"-f\",\n",
+ " \"{connection_file}\"\n",
+ " ]\n",
+ "}\n",
+ "\n",
+ "if not os.path.exists(\"/home/ma-user/anaconda3/share/jupyter/kernels/python-3.9.0/\"):\n",
+ " os.mkdir(\"/home/ma-user/anaconda3/share/jupyter/kernels/python-3.9.0/\")\n",
+ "\n",
+ "with open('/home/ma-user/anaconda3/share/jupyter/kernels/python-3.9.0/kernel.json', 'w') as f:\n",
+ " json.dump(data, f, indent=4)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "#### 注：以上代码运行完成后，需要重新设置kernel为python-3.9.0"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "![](\"https://mindspore-demo.obs.cn-north-4.myhuaweicloud.com/imgs/ai-gallery/change-kernel.PNG\")