File size: 4,505 Bytes
1c348c8 d6d2684 1c348c8 d6d2684 1c348c8 d6d2684 1c348c8 b629e33 1c348c8 d6d2684 1c348c8 b629e33 1c348c8 d6d2684 1c348c8 b629e33 7a01e78 b629e33 1c348c8 ad7d09f b629e33 ad7d09f b629e33 1c348c8 d6d2684 b629e33 d6d2684 b629e33 d6d2684 1c348c8 b629e33 1c348c8 d6d2684 |
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 |
---
base_model: tokyotech-llm/Swallow-7b-hf
library_name: peft
---
# Model Info
This is a model that applies LLM2Vec to Swallow. Only the PEFT Adapter is distributed.
LLM2Vec is fine-tuned on two tasks: MNTP and SimCSE, and this repository contains the results of applying SimCSE after MNTP.
For the MNTP Adapter, please refer to [this link](https://huggingface.co/uzabase/LLM2Vec-Llama-2-7b-hf-wikipedia-jp-mntp).
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
- **Model type:** PEFT
- **Language(s) (NLP):** Japanese
- **License:** Apache2.0
- **Finetuned from model:** [Swallow-7b-hf](https://huggingface.co/tokyotech-llm/Swallow-7b-hf)
### Model Sources
- **Repository:** https://github.com/McGill-NLP/llm2vec
- **Paper:** https://arxiv.org/abs/2404.05961
# Usage
- Please see [original LLM2Vec repo](https://huggingface.co/McGill-NLP/LLM2Vec-Llama-2-7b-chat-hf-mntp-unsup-simcse#usage)
# Benchmark
- Followings are summaries. Details are [here](https://tech.uzabase.com/entry/2024/09/30/114245)
## MTEB(Japanese)
| | Classification | Clustering | PairClassification | Reranking | BitextMining | Retrieval | Sts | 平均 |
| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| Llama2-Llm2vec-eng | 0.527 | 0.258 | 0.501 | 0.217 | 0.275 | 0.296 | 0.765 | 0.408 |
| Llama2-Llm2vec-jpn | 0.570 | 0.365 | 0.510 | 0.349 | 0.470 | 0.417 | 0.795 | 0.498 |
| **Swallow-Llm2vec-jpn (This repo)** | 0.621 | 0.391 | 0.510 | 0.475 | 0.475 | 0.491 | 0.832 | 0.523 |
## MTEB(English)
| | Classification | Clustering | Pair_Classification| Reranking | Retrieval | STS | 平均 |
| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| Llama2-Llm2vec-eng | 0.709 | 0.386 | 0.780 | 0.588 | 0.329| 0.723 | 0.586 |
| Llama2-Llm2vec-jpn | 0.722 | 0.428 | 0.785 | 0.594 | 0.371 | 0.717 | 0.603 |
| **Swallow-Llm2vec-jpn (This repo)** | 0.695 | 0.385 | 0.751 | 0.576 | 0.318 | 0.710 | 0.572 |
# Training Details
## Training Data
- Make Corpus from SimCSE from [Wikipedia](https://huggingface.co/datasets/wikimedia/wikipedia)
- Script for making SimCSE Corpus
```
import argparse
import random
import re
from pathlib import Path
from datasets import load_dataset
from tqdm import tqdm
def main(args):
random.seed(args.seed)
wiki_ds = load_dataset("wikimedia/wikipedia", "20231101.ja")
sampled_index = random.sample(range(len(wiki_ds["train"])), args.N)
sample_wiki = wiki_ds["train"][sampled_index]
output_texts = []
for title, text in tqdm(zip(sample_wiki["title"], sample_wiki["text"])):
output_texts.append(title)
sentences = re.split("[\n。]", text)
for sentence in sentences:
if len(sentence) > args.min_sentence_len:
output_texts.append(sentence.strip()+"。")
with args.output_path.open(mode="w") as f:
for line in output_texts:
f.write(line)
f.write("\n")
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--N", default=200000, type=int)
parser.add_argument("--seed", default=42, type=int)
parser.add_argument("-o", "--output_path", type=Path)
parser.add_argument("--min_sentence_len", default=50, type=int)
args = parser.parse_args()
main(args)
```
## Training Hyperparameter
- simcse_dropout: 0.3
- bidirectional: true
- pooling_mode: "mean"
- remove_unused_columns: false
- learning_rate: 3e-5
- loss_scale: 20
- batch_size: 256
- gradient_accumulation_steps: 1
- max_seq_length: 128
- lora_r: 16
- torch_dtype: "bfloat16"
- attn_implementation: "flash_attention_2"
- seed: 42
- bf16: true
- gradient_checkpointing: true
## Accelerator Settings
- deepspeed_config:
- gradient_accumulation_steps: 1
- gradient_clipping: 1.0
- offload_optimizer_device: nvme
- offload_optimizer_nvme_path: /nvme
- zero3_save_16bit_model: true
- zero_stage: 2
- distributed_type: DEEPSPEED
- downcast_bf16: 'no'
- dynamo_config:
- dynamo_backend: INDUCTOR
- dynamo_mode: default
- dynamo_use_dynamic: true
- dynamo_use_fullgraph: true
- enable_cpu_affinity: false
- machine_rank: 0
- main_training_function: main
- mixed_precision: bf16
- num_machines: 1
- num_processes: 2
- rdzv_backend: static
- same_network: true
- quse_cpu: false
## Framework versions
- Python: 3.12.3
- PEFT 0.11.1
- Sentence Transformers: 3.0.1
- Transformers: 4.41.0
- PyTorch: 2.3.0
- Accelerate: 0.30.1
- Datasets: 2.20.0
- Tokenizers: 0.19.1
- MTEB: 1.13.0 |