Instructions to use ahmadmwali/m2m_trial1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use ahmadmwali/m2m_trial1 with PEFT:
from peft import PeftModel from transformers import AutoModelForSeq2SeqLM base_model = AutoModelForSeq2SeqLM.from_pretrained("facebook/m2m100_418M") model = PeftModel.from_pretrained(base_model, "ahmadmwali/m2m_trial1") - Notebooks
- Google Colab
- Kaggle
Download eval_final_results.json from ahmadmwali/m2m_trial1: direct link, hf CLI and curl.
- Browser
- Download file 344 Bytes
-
https://huggingface.co/ahmadmwali/m2m_trial1/resolve/cbc4ff8bd1476f4cca5cfc7a849997d919fc1b9e/eval_final_results.json
- Command line
-
hf download hf://ahmadmwali/m2m_trial1@cbc4ff8bd1476f4cca5cfc7a849997d919fc1b9e/eval_final_results.json
-
curl -L -o eval_final_results.json https://huggingface.co/ahmadmwali/m2m_trial1/resolve/cbc4ff8bd1476f4cca5cfc7a849997d919fc1b9e/eval_final_results.json
344 Bytes
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 0.703799131472624, | |
| "eval_cer": 0.04866874031541143, | |
| "eval_f1": 0.8646274948454353, | |
| "eval_loss": 6.131602764129639, | |
| "eval_meteor": 0.8522783301446667, | |
| "eval_runtime": 534.3018, | |
| "eval_samples_per_second": 1.872, | |
| "eval_steps_per_second": 0.468, | |
| "eval_wer": 0.14741227063721452 | |
| } |