Other Open Access
Pretrained Spec2Vec model (trained on GNPS AllPositive dataset)
Huber, Florian
Pretrained Gensim Word2Vec model for Spec2Vec:
The model was trained for 15 iterations on the "AllPositive" dataset, created from GNPS positive ionmode mass spectra (95320 spectra after filtering) as fully described here https://github.com/iomega/spec2vec_gnps_data_analysis/blob/master/notebooks/iomega-5b-train-spec2vec-model-all-gnps.ipynb.
Updated version v2:
The former model was difficult to import using the current version of the spec2vec code. This is hence an re-trained model on the same dataset as before. Since the model initiation is inherently randomized, this model will not produce the exact same word vectors (peak vectors).
Files
(286.0 MB)
| Name | Size | |
|---|---|---|
|
spec2vec_AllPositive_ratio05_filtered_201101_iter_15.model
md5:52fc267f5ce219f66cae5d4feb31a146 |
7.9 MB | Download |
|
spec2vec_AllPositive_ratio05_filtered_201101_iter_15.model.trainables.syn1neg.npy
md5:18d2cf589bf53cd1d9a5332575bde830 |
139.1 MB | Download |
|
spec2vec_AllPositive_ratio05_filtered_201101_iter_15.model.wv.vectors.npy
md5:d76f16fd8096645b2de67f1ad43fef34 |
139.1 MB | Download |
212
143
views
downloads
| All versions | This version | |
|---|---|---|
| Views | 212 | 138 |
| Downloads | 143 | 100 |
| Data volume | 12.0 GB | 8.3 GB |
| Unique views | 155 | 112 |
| Unique downloads | 63 | 45 |
Indexed in
- Publication date:
- August 10, 2020
- DOI:
-
- Keyword(s):
- Spec2Vec model, mass spectra embeddings, Gensim word2vec model
- Communities:
- License (for files):
- Creative Commons Attribution 4.0 International