This repository provides a Japanese-centric multilingual GPT-NeoX model of 10 billion parameters.
-
Library
The model was trained using code based on
EleutherAI/gpt-neox.
-
Model architecture
A 36-layer, 4864-hidden-size transformer-based language model.
-
Pre-training
The model was trained on around 600B tokens from a mixture of the following corpora.
-
Instruction-supervised-finetuning
The model was finetuned on a subset records from a mixture of the following dataset. Training epoch: 1.
-
Model Series
-
Authors
Takeshi Kojima
-
Japanese benchmark : JGLUE 8-task (2023-08-27)
- We used Stability-AI/lm-evaluation-harness library for evaluation.
- The 8-task average accuracy is based on results of JCommonsenseQA-1.1, JNLI-1.1, MARC-ja-1.1, JSQuAD-1.1, jaqket_v2-0.2, xlsum_ja-1.0, xwinograd_ja, and mgsm-1.0.
- model loading is performed with float16, and evaluation is performed with template version 0.3 using the few-shot in-context learning.
- The number of few-shots is 3,3,3,2,1,1,0,5.
- special_tokens_map.json is modified to avoid errors during the evaluation of the second half benchmarks. As a result, the results of the first half benchmarks became slightly different.
| model | average | jcommonsenseqa | jnli | marc_ja | jsquad | jaqket_v2 | xlsum_ja | xwinograd_ja | mgsm |
|---|
| weblab-10b-instruction-sft | 59.11 | 74.62 | 66.56 | 95.49 | 78.34 | 63.32 | 20.57 | 71.95 | 2 |
| weblab-10b | 50.74 | 66.58 | 53.74 | 82.07 | 62.94 | 56.19 | 10.03 | 71.95 | 2.4 |
-
Japanese benchmark : JGLUE 4-task (2023-08-18)
- We used Stability-AI/lm-evaluation-harness library for evaluation.
- The 4-task average accuracy is based on results of JCommonsenseQA-1.1, JNLI-1.1, MARC-ja-1.1, and JSQuAD-1.1.
- model loading is performed with float16, and evaluation is performed with template version 0.3 using the few-shot in-context learning.
- The number of few-shots is 3,3,3,2.
| Model | Average | JCommonsenseQA | JNLI | MARC-ja | JSQuAD |
|---|
| weblab-10b-instruction-sft | 78.78 | 74.35 | 65.65 | 96.06 | 79.04 |
| weblab-10b | 66.38 | 65.86 | 54.19 | 84.49 | 60.98 |
1import torch
2from transformers import AutoTokenizer, AutoModelForCausalLM
3
4tokenizer = AutoTokenizer.from_pretrained("matsuo-lab/weblab-10b-instruction-sft")
5model = AutoModelForCausalLM.from_pretrained("matsuo-lab/weblab-10b-instruction-sft", torch_dtype=torch.float16)
6
7if torch.cuda.is_available():
8 model = model.to("cuda")
9
10text = "大規模言語モデルについて説明してください。"
11text = f'以下は、タスクを説明する指示です。要求を適切に満たす応答を書きなさい。\n\n### 指示:\n{text}\n\n### 応答:'
12token_ids = tokenizer.encode(text, add_special_tokens=False, return_tensors="pt")
13
14with torch.no_grad():
15 output_ids = model.generate(
16 token_ids.to(model.device),
17 max_new_tokens=100,
18 do_sample=True,
19 temperature=0.7,
20 top_p=0.95
21 )
22
23output = tokenizer.decode(output_ids.tolist()[0])
24print(output)
25
