Instructions to use achanu/mednli_laya with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Laya
How to use achanu/mednli_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 train_log.json from achanu/mednli_laya: direct link, hf CLI and curl.
- Browser
- Download file 2.15 kB
-
https://huggingface.co/achanu/mednli_laya/resolve/main/train_log.json
- Command line
-
hf download hf://achanu/mednli_laya/train_log.json
-
curl -L -o train_log.json https://huggingface.co/achanu/mednli_laya/resolve/main/train_log.json
2.15 kB
| { | |
| "settings": { | |
| "out": "runs/mm_seed0", | |
| "base": "multilingual", | |
| "repo": "presencesw/mednli", | |
| "data_dir": null, | |
| "deid": "keep", | |
| "device": "cuda:0", | |
| "seed": 0, | |
| "epochs": 6, | |
| "batch": 32, | |
| "grad_accum": 2, | |
| "encoder_lr": 2.5e-05, | |
| "head_lr": 0.0001, | |
| "min_lr": 1e-06, | |
| "weight_decay": 0.01, | |
| "warmup_frac": 0.06, | |
| "grad_clip": 1.0, | |
| "label_smoothing": 0.0, | |
| "shuffle_prob": 1.0, | |
| "mismatch_frac": 0.15, | |
| "patience": 2, | |
| "calib_items": 400, | |
| "eval_batch": 128, | |
| "max_len": null, | |
| "head_max_len": null, | |
| "freeze_embeddings": false, | |
| "grad_checkpointing": false, | |
| "limit": null | |
| }, | |
| "data": { | |
| "train_cleaning": { | |
| "rows_in": 11232, | |
| "conflicting_rows_dropped": 9, | |
| "duplicate_rows_dropped": 3, | |
| "rows_out": 11220 | |
| }, | |
| "train_items": 10819, | |
| "calibration_items": 401, | |
| "validation_items": 1395, | |
| "mismatch_frac": 0.15 | |
| }, | |
| "validation_before": { | |
| "nll": 1.3586, | |
| "accuracy": 0.4317, | |
| "macro_f1": 0.3888, | |
| "logit_range": 2.646 | |
| }, | |
| "best": { | |
| "nll": 0.4914, | |
| "accuracy": 0.8075, | |
| "macro_f1": 0.8074, | |
| "logit_range": 4.9356, | |
| "epoch": 2, | |
| "train_loss": 0.3982, | |
| "minutes": 1.38 | |
| }, | |
| "calibration": { | |
| "items": 1604, | |
| "temperature": 1.256, | |
| "nll_before": 0.4458, | |
| "nll_after": 0.433, | |
| "mean_confidence_after": 0.8346 | |
| }, | |
| "history": [ | |
| { | |
| "nll": 0.5148, | |
| "accuracy": 0.7905, | |
| "macro_f1": 0.7929, | |
| "logit_range": 3.4344, | |
| "epoch": 1, | |
| "train_loss": 0.6777, | |
| "minutes": 0.66 | |
| }, | |
| { | |
| "nll": 0.4914, | |
| "accuracy": 0.8075, | |
| "macro_f1": 0.8074, | |
| "logit_range": 4.9356, | |
| "epoch": 2, | |
| "train_loss": 0.3982, | |
| "minutes": 1.38 | |
| }, | |
| { | |
| "nll": 0.6095, | |
| "accuracy": 0.7885, | |
| "macro_f1": 0.7899, | |
| "logit_range": 5.8063, | |
| "epoch": 3, | |
| "train_loss": 0.2376, | |
| "minutes": 2.09 | |
| }, | |
| { | |
| "nll": 0.8555, | |
| "accuracy": 0.7753, | |
| "macro_f1": 0.7776, | |
| "logit_range": 7.4772, | |
| "epoch": 4, | |
| "train_loss": 0.1328, | |
| "minutes": 2.75 | |
| } | |
| ] | |
| } |