File size: 4,258 Bytes
89831d5
 
a519f10
 
 
5034d89
89831d5
a519f10
 
 
e7ae5bd
a519f10
 
0a2b9c6
aae4619
1f9a0a0
 
 
a519f10
0a2b9c6
ebffa9b
0a2b9c6
 
 
a519f10
d32df80
a519f10
 
 
 
 
 
 
 
 
 
 
 
 
e7ae5bd
a519f10
 
 
 
 
 
 
 
 
0a2b9c6
a519f10
 
 
 
 
 
0a2b9c6
a519f10
 
169bba5
a519f10
169bba5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a519f10
0a2b9c6
e7ae5bd
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
---
license: apache-2.0
tags:
- snowflake
- arctic
- moe
---

## Model Details

Arctic is a dense-MoE Hybrid transformer architecture pre-trained from scratch by the Snowflake AI 
Research Team. We are releasing model checkpoints for both the base and instruct-tuned versions of 
Arctic under an Apache-2.0 license. This means you can use them freely in your own research, 
prototypes, and products. Please see our blog 
[Snowflake Arctic: The Best LLM for Enterprise AI — Efficiently Intelligent, Truly Open](https://www.snowflake.com/blog/arctic-open-efficient-foundation-language-models-snowflake/) 
for more information on Arctic and links to other relevant resources such as our series of cookbooks 
covering topics around training your own custom MoE models, how to produce high-quality training data, 
and much more.

* [Arctic-Base](https://huggingface.co/Snowflake/snowflake-arctic-base/)
* [Arctic-Instruct](https://huggingface.co/Snowflake/snowflake-arctic-instruct/)

For the latest details about Snowflake Arctic including tutorials, etc. please refer to our github repo: 
* https://github.com/Snowflake-Labs/snowflake-arctic

**Model developers** Snowflake AI Research Team

**License** Apache-2.0

**Input** Models input text only.

**Output** Models generate text and code only.

**Model Release Date** April, 24th 2024.

## Model Architecture

Arctic combines a 10B dense transformer model with a residual 128x3.66B MoE MLP resulting in 480B 
total and 17B active parameters chosen using a top-2 gating. For more details about Arctic's model
Architecture, training process, data, etc. [see our series of cookbooks](https://www.snowflake.com/en/data-cloud/arctic/cookbook/).


## Usage

As of 4/24/2024 we are actively working with the maintainers of `transformers` to include the Arctic 
model implementation. Until this support is released please follow these instructions to get the 
required dependencies for using Arctic:

```python
pip install git+https://github.com/Snowflake-Labs/transformers.git@arctic
```

Arctic leverages several features from [DeepSpeed](https://github.com/microsoft/DeepSpeed), you will need to 
install the latest version of DeepSpeed to get all of these required features:

```python
pip install "deepspeed>=0.14.2"
```

### Inference examples

Due to the model size we recommend using a single 8xH100 instance from your
favorite cloud provider such as: AWS [p5.48xlarge](https://aws.amazon.com/ec2/instance-types/p5/), 
Azure [ND96isr_H100_v5](https://learn.microsoft.com/en-us/azure/virtual-machines/nd-h100-v5-series), etc.

In this example we are using FP8 quantization provided by DeepSpeed in the backend, we can also use FP6 
quantization by specifying `q_bits=6` in the `ArcticQuantizationConfig` config. The `"150GiB"` setting 
for max_memory is required until we can get DeepSpeed's FP quantization supported natively as a [HFQuantizer](https://huggingface.co/docs/transformers/main/en/hf_quantizer#build-a-new-hfquantizer-class) which we 
are actively working on.

```python
import os
# enable hf_transfer for faster ckpt download
os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from transformers.models.arctic.configuration_arctic import ArcticQuantizationConfig

tokenizer = AutoTokenizer.from_pretrained("Snowflake/snowflake-arctic-instruct")

quant_config = ArcticQuantizationConfig(q_bits=8)

model = AutoModelForCausalLM.from_pretrained(
    "Snowflake/snowflake-arctic-instruct",
    low_cpu_mem_usage=True,
    device_map="auto",
    ds_quantization_config=quant_config,
    max_memory={i: "150GiB" for i in range(8)},
    torch_dtype=torch.bfloat16)

messages = [{"role": "user", "content": "What is 1 + 1 "}]
input_ids = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to("cuda")

outputs = model.generate(input_ids=input_ids, max_new_tokens=20)
print(tokenizer.decode(outputs[0]))
```

The Arctic github page has additional code snippets and examples around running inference:

* Example with pure-HF: https://github.com/Snowflake-Labs/snowflake-arctic/blob/main/inference
* Tutorial using vLLM: https://github.com/Snowflake-Labs/snowflake-arctic/tree/main/inference/vllm