RobotsMali
/

bam-vits-train

@@ -1,199 +1,73 @@
 ---
 library_name: transformers
-tags: []
 ---
-# Model Card for Model ID
-<!-- Provide a quick summary of what the model is/does. -->
-## Model Details
-### Model Description
-<!-- Provide a longer summary of what this model is. -->
-This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
-- **Developed by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Model type:** [More Information Needed]
-- **Language(s) (NLP):** [More Information Needed]
-- **License:** [More Information Needed]
-- **Finetuned from model [optional]:** [More Information Needed]
-### Model Sources [optional]
-<!-- Provide the basic links for the model. -->
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
-## Uses
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
-### Direct Use
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
-[More Information Needed]
-### Downstream Use [optional]
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
-[More Information Needed]
-### Out-of-Scope Use
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
-[More Information Needed]
-## Bias, Risks, and Limitations
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-[More Information Needed]
-### Recommendations
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
-## How to Get Started with the Model
-Use the code below to get started with the model.
-[More Information Needed]
-## Training Details
-### Training Data
-<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
-[More Information Needed]
-### Training Procedure
-<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
-#### Preprocessing [optional]
-[More Information Needed]
-#### Training Hyperparameters
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
-#### Speeds, Sizes, Times [optional]
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
-[More Information Needed]
-## Evaluation
-<!-- This section describes the evaluation protocols and provides the results. -->
-### Testing Data, Factors & Metrics
-#### Testing Data
-<!-- This should link to a Dataset Card if possible. -->
-[More Information Needed]
-#### Factors
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-[More Information Needed]
-#### Metrics
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
-[More Information Needed]
-### Results
-[More Information Needed]
-#### Summary
-## Model Examination [optional]
-<!-- Relevant interpretability work for the model goes here -->
-[More Information Needed]
-## Environmental Impact
-<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
-Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
-- **Hardware Type:** [More Information Needed]
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** [More Information Needed]
-## Technical Specifications [optional]
-### Model Architecture and Objective
-[More Information Needed]
-### Compute Infrastructure
-[More Information Needed]
-#### Hardware
-[More Information Needed]
-#### Software
-[More Information Needed]
-## Citation [optional]
-<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
-**BibTeX:**
-[More Information Needed]
-**APA:**
-[More Information Needed]
-## Glossary [optional]
-<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
-[More Information Needed]
-## More Information [optional]
-[More Information Needed]
-## Model Card Authors [optional]
-[More Information Needed]
-## Model Card Contact
-[More Information Needed]

 ---
+language: [bm]
 library_name: transformers
+license: cc-by-4.0
+datasets: [RobotsMali/afvoices-notag]
+base_model: ylacombe/vits-vctk-with-discriminator
+tags: [vits, text-to-speech, bambara, training-checkpoint, discriminator, low-resource]
 ---
+# Bambara VITS — training checkpoint
+`RobotsMali/bam-vits-train` is the continued-training counterpart of [`RobotsMali/bam-vits`](https://huggingface.co/RobotsMali/bam-vits). Its `VitsModelForPreTraining` weights retain both the generator and discriminator (about 323 MB rather than the inference export's 159 MB). Use this repository as `model_name_or_path` when fine-tuning; use `bam-vits` for ordinary inference.
+> **Substantially undertrained.** The checkpoint received 200 epochs on a small experimental corpus and remains far below typical VITS budgets of hundreds of thousands of optimizer steps (approximately 200,000 is our reference). It is a starting point, not a converged base model.
+## Continue training
+Clone the RobotsMali training code, install its pinned dependencies, build monotonic alignment, then point a copied YAML config at this checkpoint and your dataset:
+```bash
+git clone https://github.com/RobotsMali-AI/vits-bam.git
+cd vits-bam
+python -m pip install -r requirements.txt
+cd monotonic_align && python setup.py build_ext --inplace && cd ..
+cp config/bam-vits.yaml config/my-experiment.yaml
+```
+At minimum, edit:
+```yaml
+model:
+  model_name_or_path: RobotsMali/bam-vits-train
+  tokenizer_name: RobotsMali/bam-vits-train
+  pseudo_ipa: false
+  override_speaker_embeddings: true
+  override_vocabulary_embeddings: false
+data:
+  dataset_name: OWNER/DATASET
+  dataset_config_name: null
+  audio_column_name: audio
+  text_column_name: text
+  speaker_id_column_name: speaker_id
+logging:
+  output_dir: ./tmp/my-experiment
+  hub_model_id: OWNER/MODEL
+```
+Then run:
+```bash
+accelerate launch run_vits_finetuning.py config/my-experiment.yaml
+```
+Review speaker and vocabulary settings for your data. This repository's custom training class and compiled monotonic-alignment extension are required. After training, `push_vits_model.py --inference_only` publishes a generator-only `VitsModel`; omitting the flag preserves the discriminator.
+## Provenance
+The model was initialized from the English VCTK [`ylacombe/vits-vctk-with-discriminator`](https://huggingface.co/ylacombe/vits-vctk-with-discriminator) and trained on `top20-speakers` from [`RobotsMali/afvoices-notag`](https://huggingface.co/datasets/RobotsMali/afvoices-notag): 21,253 training and 1,128 test examples from the 20 participants with the most utterances, excluding tagged transcripts. The corpus is spontaneous, noisy ASR data rather than studio TTS data.
+The source config, [`config/bam-vits.yaml`](https://github.com/RobotsMali-AI/vits-bam/blob/main/config/bam-vits.yaml), specifies 200 epochs, per-device batch 80, learning rate 0.0005, FP16, 0.2–20-second audio, maximum 450 tokens, and seed 789. Text was lowercased; audio was resampled to 22.05 kHz.
+## Limitations and responsible use
+No MOS, intelligibility, pronunciation, speaker-similarity, or safety evaluation is reported. Any downstream model inherits risks from undertraining, noisy data, limited speaker coverage, minimal text normalization, and the English initialization. The source participant-to-model-index mapping was not saved. Evaluate convergence, pronunciation, memorization, bias, licensing, speaker consent, and voice similarity for every derivative.
+This checkpoint is for research and fine-tuning. Do not use it to impersonate people, create deceptive audio, or support safety-critical speech without substantial additional training and evaluation. Clearly document downstream data and disclose synthetic audio.
+## References
+- [RobotsMali training repository](https://github.com/RobotsMali-AI/vits-bam), adapted from [ylacombe/finetune-hf-vits](https://github.com/ylacombe/finetune-hf-vits)
+- [VITS paper](https://proceedings.mlr.press/v139/kim21f.html)