# Load model directly
from transformers import AutoProcessor, AutoModelForPreTraining
processor = AutoProcessor.from_pretrained("techsword/wav2vec2-base-mandarin-magicdata")
model = AutoModelForPreTraining.from_pretrained("techsword/wav2vec2-base-mandarin-magicdata", device_map="auto")Quick Links
Fairseq wav2vec2-base pretraining checkpoint (checkpoint_best, ~85000 updates), converted to HuggingFace format with the official transformers converter and verified (weight-level spot check + forward-pass comparison against the fairseq model; see conversion log). Pretrained on MAGICDATA (Mandarin Chinese).
Raw fairseq checkpoints: techsword/wav2vec2-base-mandarin-magicdata-checkpoints Used in: tone-probe experiments, https://github.com/techsword/tone-encoding-in-speech-model
- Downloads last month
- 12
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="techsword/wav2vec2-base-mandarin-magicdata")