File size: 5,940 Bytes
c208d89
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "e0f555c6-4f5d-4f2d-93ab-8106d2c470dc",
      "metadata": {
        "jupyter": {
          "source_hidden": true
        },
        "id": "e0f555c6-4f5d-4f2d-93ab-8106d2c470dc"
      },
      "outputs": [],
      "source": [
        "!pip install -q accelerate sentencepiece torch transformers"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "1mncI66sFR9a",
      "metadata": {
        "id": "1mncI66sFR9a",
        "jupyter": {
          "source_hidden": true
        }
      },
      "outputs": [],
      "source": [
        "!pip install -q --upgrade gradio"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Inference with Gradio but no streaming"
      ],
      "metadata": {
        "id": "0q800RsXd6Nj"
      },
      "id": "0q800RsXd6Nj"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "26153855-215a-4289-b4ed-a1cb935ebe69",
      "metadata": {
        "jupyter": {
          "source_hidden": true
        },
        "scrolled": true,
        "id": "26153855-215a-4289-b4ed-a1cb935ebe69"
      },
      "outputs": [],
      "source": [
        "import gradio as gr\n",
        "import torch\n",
        "from transformers import AutoModelForCausalLM, AutoTokenizer\n",
        "\n",
        "base_model = \"TokenBender/evolvedSeeker_1_3\"\n",
        "tokenizer = AutoTokenizer.from_pretrained(base_model)\n",
        "model = AutoModelForCausalLM.from_pretrained(base_model, torch_dtype=torch.float16)\n",
        "model.config.use_cache = True\n",
        "model = model.to('cuda:0')\n",
        "\n",
        "def predict(message, history):\n",
        "    history_transformed = [{'role': 'system', 'content': \"You are a helpful coding assistant, provide code based on the given query in context.\\n\"}]\n",
        "    for msg in history:\n",
        "        history_transformed.append({'role': 'user', 'content': msg[0]})\n",
        "        history_transformed.append({'role': 'assistant', 'content': msg[1]})\n",
        "\n",
        "    history_transformed.append({'role': 'user', 'content': message})\n",
        "\n",
        "    inputs = tokenizer.apply_chat_template(history_transformed, return_tensors=\"pt\").to(model.device)\n",
        "    outputs = model.generate(inputs, max_new_tokens=512, do_sample=False, top_k=10, top_p=0.95, num_return_sequences=1, eos_token_id=32021)\n",
        "    response = tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True)\n",
        "    yield response\n",
        "\n",
        "gr.ChatInterface(predict).queue().launch(share=True)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Inference without gradio"
      ],
      "metadata": {
        "id": "0gpUWgWtdhOi"
      },
      "id": "0gpUWgWtdhOi"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "7f5f98f1-430e-45a0-b4b3-6a3340b5efcf",
      "metadata": {
        "id": "7f5f98f1-430e-45a0-b4b3-6a3340b5efcf"
      },
      "outputs": [],
      "source": [
        "from transformers import AutoTokenizer, AutoModelForCausalLM\n",
        "tokenizer = AutoTokenizer.from_pretrained(\"TokenBender/evolvedSeeker_1_3\", trust_remote_code=True)\n",
        "model = AutoModelForCausalLM.from_pretrained(\"TokenBender/evolvedSeeker_1_3\", trust_remote_code=True).cuda()\n",
        "messages=[\n",
        "    {'role': 'system', 'content': \"You are EvolvedSeeker, a model fine-tuned by TokenBender for coding assistant role. Help the user in a friendly, curious manner.\"},\n",
        "    { 'role': 'user', 'content': \"Hi, who are you?.\"}\n",
        "]\n",
        "inputs = tokenizer.apply_chat_template(messages, return_tensors=\"pt\").to(model.device)\n",
        "# 32021 is the id of <|EOT|> token\n",
        "outputs = model.generate(inputs, max_new_tokens=512, do_sample=False, top_k=10, top_p=0.95, num_return_sequences=1, eos_token_id=32021)\n",
        "print(tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True))"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Chat further"
      ],
      "metadata": {
        "id": "TsFjwbtadqsJ"
      },
      "id": "TsFjwbtadqsJ"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "a15a4f07-846f-4b89-bdcc-21b7c182e614",
      "metadata": {
        "id": "a15a4f07-846f-4b89-bdcc-21b7c182e614"
      },
      "outputs": [],
      "source": [
        "messages=[\n",
        "    {'role': 'system', 'content': \"You are EvolvedSeeker, a model fine-tuned by TokenBender for coding assistant role. Help the user in a friendly, curious manner.\"},\n",
        "    { 'role': 'user', 'content': \"Write a python program to create a snake game.\"}\n",
        "]\n",
        "inputs = tokenizer.apply_chat_template(messages, return_tensors=\"pt\").to(model.device)\n",
        "# 32021 is the id of <|EOT|> token\n",
        "outputs = model.generate(inputs, max_new_tokens=2048, do_sample=False, top_k=10, top_p=0.95, num_return_sequences=1, eos_token_id=32021)\n",
        "print(tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True))"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "gpuType": "T4",
      "machine_shape": "hm",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.13"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}