Instructions to use leafiy/mia-laya with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Laya
How to use leafiy/mia-laya with Laya:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Download mia.json from leafiy/mia-laya: direct link, hf CLI and curl.
- Browser
- Download file 1.74 kB
-
https://huggingface.co/leafiy/mia-laya/resolve/main/mia.json
- Command line
-
hf download hf://leafiy/mia-laya/mia.json
-
curl -L -o mia.json https://huggingface.co/leafiy/mia-laya/resolve/main/mia.json
1.74 kB
| { | |
| "name": "mia-laya", | |
| "task": "chinese-sentiment-3class", | |
| "labels": [ | |
| "负面", | |
| "中性", | |
| "正面" | |
| ], | |
| "base_model": "convaiinnovations/laya-multilingual", | |
| "base_revision": "e4e9ddf21a7b1903b7acffd8814ad4307bf63a67", | |
| "base_license": "Apache-2.0", | |
| "encoder": "jhu-clsp/mmBERT-base", | |
| "architecture": "Laya rl-agent choice head, loaded with laya.load (laya==0.3.20)", | |
| "params_million": 322, | |
| "weights_dtype": "float32", | |
| "temperature": 1.9488, | |
| "class_weights": { | |
| "负面": 1.666322, | |
| "中性": 0.784345, | |
| "正面": 0.888946 | |
| }, | |
| "training_data": "72,610 Chinese short texts relabeled by an LLM; 褒贬混合 dropped; not distributed", | |
| "epochs": 3, | |
| "effective_batch_size": 256, | |
| "seed": 42, | |
| "eval": { | |
| "dev_4920": { | |
| "accuracy": 0.8108, | |
| "macro_f1": 0.7974 | |
| }, | |
| "holdout_5372": { | |
| "accuracy": 0.8172, | |
| "macro_f1": 0.8028, | |
| "recall": { | |
| "负面": 0.794, | |
| "中性": 0.77, | |
| "正面": 0.894 | |
| } | |
| }, | |
| "chinese_csv_900": { | |
| "accuracy": 0.9367, | |
| "macro_f1": 0.9362, | |
| "neutral_recall": 0.833 | |
| }, | |
| "public_forced_binary_accuracy": { | |
| "chnsenticorp": 0.8833, | |
| "online_shopping_10_cats": 0.9186, | |
| "eprstmt": 0.882, | |
| "weibo_senti_100k": 0.803 | |
| }, | |
| "public_dmsc_3class_accuracy": 0.5422 | |
| }, | |
| "throughput": "1,977 rows/s on RTX 4090, batch 64, holdout 5,372 rows end to end (reports/mia-laya/throughput.json)", | |
| "reports": [ | |
| "reports/mia-laya/", | |
| "reports/public/mia-laya.json" | |
| ], | |
| "released": "2026-09-26", | |
| "license": "Apache-2.0", | |
| "hub": { | |
| "huggingface": "https://huggingface.co/leafiy/mia-laya", | |
| "modelscope": "https://modelscope.cn/models/leafiy2/mia-laya" | |
| } | |
| } | |