| --- |
| license: agpl-3.0 |
| library_name: pytorch |
| language: |
| - ar |
| - ckb |
| - fa |
| - sd |
| - ur |
| datasets: |
| - google/fleurs |
| tags: |
| - forced-alignment |
| - phoneme-recognition |
| - phoneme-segmentation |
| - speech |
| - multilingual |
| --- |
| |
| # mbfa-abjad (p4mbfa · abjad) |
|
|
| A context-free CNN phoneme encoder and forced aligner (p4mbfa, the successor of CUPE / p3cupe), |
| for the **abjad** language group of [standard_g2p](https://github.com/tabahi/CharsiuG2P). Each 5 ms frame is classified |
| from at most 120 ms of audio (38.9 ms receptive field), so the model cannot learn any language's |
| phonotactics. Phone boundaries come from a segmental Viterbi decoder over the known phone |
| sequence, refined to sub-frame precision at the crossing of neighbouring posteriors. |
|
|
| | head | labels | size | |
| |---|---|---| |
| | `ph` | local tokens of `abjad` (incl. `<blank> SIL noise <unk>`) | 90 | |
| | `phg` | gold phoneme groups, shared by every language group | 15 | |
| | `tone` | tone values, shared by every language group; present but untrained (no language in this group has a tone layer) | 22 | |
|
|
| Languages (the FLEURS languages it was trained on): ar, ckb, fa, sd, ur. standard_g2p maps every member of the group |
| onto the same tokens, so other members may align too, untested. |
| |
| ## Training |
| |
| **Data:** FLEURS abjad, all 38.2 h, training noise_level 0.02. |
| **Checkpoint:** experiment `mp01a`, epoch 5, selected on FLEURS val_loss (best of 20 epochs; final epoch 2.988). |
| Trunk and phg_head from latin ma02a (final epoch), ph_head and tone_head trained fresh. |
|
|
| | metric | value | |
| |---|---| |
| | `val_loss` | 2.8549 | |
| | `val_frame_acc` | 0.4126 | |
| | `val_frame_acc_groups` | 0.5149 | |
|
|
| `val_*` are on held-out FLEURS clips but are scored against the model's own alignments (there are no boundary labels outside English), so they measure self-consistency, not boundary accuracy. |
|
|
| Labels are standard_g2p dictionary pronunciations (gold inventory `9438371ed6dd`), |
| not phonetic transcriptions of what was said. |
| |
| ## Usage |
| |
| The code is in [https://github.com/tabahi/bfa_models](https://github.com/tabahi/bfa_models) (`p4mbfa/`); clone it and run from its root. |
| |
| ```python |
| from p4mbfa.inference import MbfaAligner |
| |
| aligner = MbfaAligner.from_pretrained("Tabahi/mbfa-abjad") |
| wav = aligner.load_audio("clip.wav") # mono, 16000 Hz |
| phones = ["SIL", "h", "ɛ", "l", "o", "SIL"] # this group's tokens (config.json labels.tokens) |
| for seg in aligner.align(wav, phones): |
| print(seg["token"], seg["start_ms"], seg["end_ms"]) |
| ``` |
| |
| `aligner.encode(wav)` returns the per-frame log posteriors of all three heads. |
| Text -> tokens is standard_g2p's job (`goldG2P.phonemize_sentence` then |
| `lang_group_inventory.to_local`); `config.json` lists the token strings. `from_pretrained` |
| downloads only `config.json` and `model.safetensors`. |
|
|
| ## Fine-tuning |
|
|
| `ckpt/abjad_fleurs38h_mp01a_e5_val_loss=2.855.ckpt` is the training checkpoint (PyTorch Lightning, weights only, a pickle: load |
| it only if you trust this repo). Download it, then point a p4mbfa yaml at it to continue |
| training, or to train a new language group on its trunk and shared `phg` head: |
|
|
| ```bash |
| hf download Tabahi/mbfa-abjad "ckpt/abjad_fleurs38h_mp01a_e5_val_loss=2.855.ckpt" --local-dir tmp/hf/mbfa-abjad |
| ``` |
|
|
| ```yaml |
| ckpt_path: "tmp/hf/mbfa-abjad/ckpt/abjad_fleurs38h_mp01a_e5_val_loss=2.855.ckpt" |
| reset_fine_heads: true # new lang_group: rebuild ph_head / tone_head; false = same group |
| ``` |
|
|
| License: AGPL-3.0. |
|
|
