---
title: geolid_combined_shared_trainable
canonical_url: "https://www.modelscope.cn/models/espnet/geolid_combined_shared_trainable"
md_url: "https://www.modelscope.cn/models/espnet/geolid_combined_shared_trainable.md"
repository: espnet/geolid_combined_shared_trainable
last_updated: 2026-10-29
license: cc-by-nc-4.0
pipeline_tag: audio-classification
tasks:
  - audio-classification
base_model:
  - facebook/mms-1b
base_model_relation: finetune
library_name:
  - pytorch
frameworks:
  - pytorch
language:
  - abk
  - afr
  - amh
  - ara
  - asm
  - ast
  - aze
  - azz
  - bak
  - bas
  - bel
  - ben
  - bod
  - bos
  - bre
  - bul
  - cat
  - ceb
  - ces
  - chv
  - ckb
  - cmn
  - cnh
  - cym
  - dan
  - deu
  - div
  - ell
  - eng
  - epo
  - est
  - eus
  - fao
  - fas
  - fil
  - fin
  - fra
  - frr
  - ful
  - gle
  - glg
  - glv
  - grn
  - gug
  - guj
  - hat
  - hau
  - haw
  - heb
  - hin
  - hrv
  - hsb
  - hun
  - hye
  - ibo
  - ina
  - ind
  - isl
  - ita
  - jav
  - jpn
  - kab
  - kam
  - kan
  - kat
  - kaz
  - kea
  - khk
  - khm
  - kin
  - kir
  - kmr
  - kor
  - lao
  - lat
  - lav
  - lin
  - lit
  - ltz
  - lug
  - luo
  - mal
  - mar
  - mhr
  - mkd
  - mlg
  - mlt
  - mon
  - mri
  - mrj
  - msa
  - mya
  - myv
  - nan
  - nbl
  - nep
  - nld
  - nno
  - nob
  - nor
  - nso
  - nya
  - oci
  - ori
  - orm
  - pan
  - pol
  - por
  - pus
  - ron
  - rus
  - sah
  - san
  - sco
  - sin
  - skr
  - slk
  - slv
  - sna
  - snd
  - som
  - sot
  - spa
  - sqi
  - srp
  - ssw
  - sun
  - swa
  - swe
  - tam
  - tat
  - tel
  - tgk
  - tgl
  - tha
  - tok
  - tos
  - tpi
  - tsn
  - tso
  - tuk
  - tur
  - uig
  - ukr
  - umb
  - urd
  - uzb
  - ven
  - vie
  - war
  - wol
  - xho
  - xty
  - yid
  - yor
  - yue
  - zul
downloads: 20
stars: 0
tags:
  - espnet
  - audio
  - language-identification
  - geolid
---

# geolid_combined_shared_trainable

> geolid_combined_shared_trainable - espnet 在 ModelScope 开源的模型。ESPnet2 GeoLID: combinedsharedtrainable

espnet/geolid_combined_shared_trainable 是 ModelScope 魔搭社区上的audio-classification模型，采用 cc-by-nc-4.0 许可，基于 facebook/mms-1b 构建。

- **Repository**: espnet/geolid_combined_shared_trainable
- **License**: cc-by-nc-4.0
- **Tasks**: audio-classification
- **Base model**: facebook/mms-1b
- **Tags**: espnet, audio, language-identification, geolid
- **Downloads**: 20
- **Stars**: 0
- **Last updated**: 2026-10-29

Source: https://www.modelscope.cn/models/espnet/geolid_combined_shared_trainable

---

# ESPnet2 GeoLID: combined_shared_trainable

This is a normal ESPnet language identification model using the stock
`S3prlFrontend` and `ESPnetLIDModel` (`model: espnet`) with optional lang2vec conditioning. GeoLID wraps SSL layers 32, 36, 40 and
44 with lang2vec prediction and residual conditioning. The model code is
maintained in ESPnet; no checkpoint-local Python modeling files or
`trust_remote_code=True` are needed.

The checkpoint was **converted, not retrained**. `local/convert_checkpoint.py`
renamed the original parameter keys and removed checked duplicate aliases. All
learned tensor values are unchanged. The auxiliary linear head now belongs to
the model; classification uses the standard `aamsoftmax_sc_topk` loss. Runtime uses current ESPnet pooling and
normalization, without legacy compatibility code. Measured accuracy of this
converted checkpoint is reported separately from historical paper results below.

## Checkpoint provenance and evaluation

**All combined-model paper results—the six test-set accuracies and Macro Avg.—were obtained with the earlier epoch-24 checkpoint (48,000 training batches). The downloadable weights are the later epoch-31 checkpoint (62,000 batches).** Conversion preserved the learned tensor values.

The [six archived evaluations](https://huggingface.co/espnet/geolid_combined_shared_trainable/tree/7dc4bf988efff29f725bf34b1caa821c27359793/exp_combined/lid_mms_ecapa_upcon_32_44_it0.4_shared_trainable_raw/inference/valid.accuracy.best) started on June 2, 2025, between 00:34 and 02:37 CDT and all finished by 02:55. The [first training log](https://huggingface.co/espnet/geolid_combined_shared_trainable/blob/7dc4bf988efff29f725bf34b1caa821c27359793/exp_combined/lid_mms_ecapa_upcon_32_44_it0.4_shared_trainable_raw/train.4.log) records the epoch-24 best update at 20:01 on June 1; the [resumed log](https://huggingface.co/espnet/geolid_combined_shared_trainable/blob/7dc4bf988efff29f725bf34b1caa821c27359793/exp_combined/lid_mms_ecapa_upcon_32_44_it0.4_shared_trainable_raw/train.3.log) records the epoch-31 best update only at 11:22 on June 2. The released weights have 41 BatchNorm counters equal to 62,000. Thus every historical combined column, including Macro Avg., predates the released checkpoint. This corrects the paper's attribution of those results to the 62k checkpoint.

The epoch-24 weights have not been recovered and the old inference logs contain no weight hash. The epoch attribution follows the archived training/evaluation chronology; the historical scores have not been independently rerun.

| Checkpoint / result source | VoxLingua107 dev | Babel | FLEURS | ML-SUPERB 2.0 dev | ML-SUPERB 2.0 dialect | VoxPopuli | Macro Avg. |
| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| Historical paper — epoch 24 (48k) | 94.4 | 95.4 | 97.7 | 88.6 | 86.8 | 99.0 | 93.7 |
| Released converted — epoch 31 (62k), batch size 1 | 93.29 | — | 97.51 | 88.69 | 85.26 | 98.98 | — |

**Babel remains unevaluated:** the licensed dev audio and original complete manifest are unavailable in this evaluation environment. Its released-model cell and the six-set Macro Avg. are left blank. No paper score is substituted, and no five-set average is presented as the six-set Macro Avg.

Each result is utterance-level accuracy on the complete listed split. Macro Avg. is the unweighted mean of the six dataset accuracies, calculated from integer correct/total counts before rounding. Historical paper values are retained as reported.

The new runs use the unmodified `espnet2.bin.lid_inference` CLI at ESPnet revision `07dc3256fbe6ab542a260314f9e3d9a0bdc9de9a`: NVIDIA H200, GPU float32, batch size 1 per forward pass, 8 workers per inference process, full-length 16 kHz mono waveforms, `--use_preprocessor false`, `--fix_duration false`, and inputs sorted by utterance ID. FLEURS uses the CLI's standard multiprocessing mode across four GPUs; the other sets use one GPU. Each process still classifies one utterance per forward pass. PyTorch 2.11.0+cu128, Transformers 5.14.1 and espnet-s3prl 2026.5.14 match the earlier VoxLingua107 run.

VoxPopuli uses its official **test** partition (16,991 utterances), not its 18,636-utterance validation partition. The archived report's 178 error IDs, reference labels and every per-language accuracy match this test manifest. FLEURS uses the full 77,810-utterance official test set. ML-SUPERB 2.0 uses the shared recipe's language normalization and excludes its 283 Norwegian dev rows, leaving 23,609 dev and 7,095 dialect-dev utterances; these match the archived error IDs and per-language rates. No further model-inventory filtering or audio truncation is applied.

For FLEURS, all 4,000 predictions saved by the initial single-GPU run exactly match the four-process run. This checks the parallel scheduling change while keeping batch size 1.

[New evaluation report, reproduction command and per-set evidence](evaluations/20261027/report.md) provide complete predictions, reference labels, inference logs and metadata for each completed test set.

The published weights' earlier legacy-reference accuracy on VoxLingua107 dev was 93.35% (batch size 4). This evaluates the released epoch-31 weights through the legacy implementation; it is separate from the epoch-24 paper result. Runtime and batching differ from the current measurements.

The previous [VoxLingua107 evaluation](evaluation_report.md), [metadata](evaluation.json), [predictions](evaluation_predictions.jsonl) and [log](evaluation.log) retain their original revision `301372e4da311d2a3d9574a5a43b3037ae6e67c1`. The new VoxLingua107 run again produced 1501/1609 correct, with all 1,609 predictions unchanged. Earlier loading and compatibility checks remain at their original revisions in [validation.json](validation.json).

Inference batch size matters: S3PRL frame alignment and ECAPA's temporal convolutions and unmasked SE averaging can change features before masked normalization/final pooling. All new results here use batch size 1.

## Installation

Install the upstream ESPnet revision containing the merged GeoLID recipe:

```bash
pip install 'espnet @ git+https://github.com/espnet/espnet.git@c9fde840770adc76100b781044abe160e38124bd'
pip install espnet-s3prl transformers espnet-model-zoo soundfile
```

S3PRL's `hf_wav2vec2_custom` upstream imports Transformers; `espnet-s3prl`
does not install it automatically. With an ESPnet tools checkout, the equivalent
installers are `make s3prl.done transformers.done` in `tools/`. The lang2vec
package is used only by recipe target preparation and is not needed for inference.

## Inference

```python
import soundfile as sf
import torch
from espnet_model_zoo.downloader import ModelDownloader
from espnet2.tasks.lid import LIDTask

files = ModelDownloader().download_and_unpack("espnet/geolid_combined_shared_trainable")
model, args = LIDTask.build_model_from_file(
    files["train_config"], files["model_file"], device="cpu"
)
model.eval()
audio, sample_rate = sf.read("speech.wav", dtype="float32")
assert sample_rate == 16000 and audio.ndim == 1
with torch.inference_mode():
    embeddings, prediction = model(
        torch.from_numpy(audio)[None],
        torch.tensor([len(audio)]),
        extract_embd=True,
    )
with open(args.lang2utt) as stream:
    languages = [line.split()[0] for line in stream]
print(languages[prediction.item()])
```

The backbone is initialized through the stock `hf_wav2vec2_custom` upstream from
`facebook/mms-1b`, then the complete converted checkpoint loads the trained SSL,
conditioning and classifier parameters. This repository contains 157
language classes. `languages.txt` preserves classifier-index order recovered
from the original model card and the original sorted recipe label construction;
it is not a copy of the missing original training utterance manifest.

`conversion.json` records the source checkpoint revision and label provenance.
`validation.json` preserves the historical strict-loading and layout-migration
checks, and separately records the complete evaluation after model integration.
`evaluation.json` and `evaluation_predictions.jsonl` retain the earlier
VoxLingua107 evaluation. The new measurements and complete per-set evidence
are under `evaluations/20261027/`. The metadata targets are used only for training; inference
uses speech and predicts its own conditioning vectors.
The old training logs and figures retained under `exp_*` describe the original
training run. Earlier checkpoint layouts remain available in repository history.

## License

Derived from [facebook/mms-1b](https://huggingface.co/facebook/mms-1b), these model
weights are licensed **CC-BY-NC-4.0**, including the non-commercial restriction.
The previous CC-BY-4.0 metadata was incorrect and has been corrected.

## Citation

```bibtex
@inproceedings{wang2025geolid,
  title={Geolocation-Aware Robust Spoken Language Identification},
  author={Qingzheng Wang and Hye-jin Shim and Jiancheng Sun and Shinji Watanabe},
  booktitle={2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)},
  year={2025},
  organization={IEEE},
  url={https://arxiv.org/abs/2508.17148}
}
```

[Recipe and evaluation results](https://github.com/espnet/espnet/blob/c9fde840770adc76100b781044abe160e38124bd/egs2/geolid/lid1/README.md)