MOSS-Transcribe-Diarize-HF-mic

Slim model weights, executable usage and framework expectations in one repository.

Contract

MOSS-Transcribe-Diarize speech-to-text

Runnable usage and the framework expectations directory are linked above.

Slim model

Slim fixture: MOSS-Transcribe-Diarize-HF-mic

A 10.0M-parameter fixture for the MOSS-Transcribe-Diarize MIC, covering itazap/MOSS-Transcribe-Diarize-HF. It is a test artifact, not a usable model, tagged test-artifact to keep it out of candidate-model search.

  • Fixture key 086fd2c47f24. It covers every checkpoint whose processor files are byte-identical and whose config matches after shrinking.
  • Same architecture, smaller config: MossTranscribeDiarizeForConditionalGeneration, shrunk by the generic rule in sweep.py (audio num_hidden_layers 2, text num_hidden_layers 2).
  • Processor files copied byte for byte from itazap/MOSS-Transcribe-Diarize-HF: Model_Architecture.png, added_tokens.json, chat_template.jinja, configuration_moss_transcribe_diarize.py, generation_config.json, merges.txt, modeling_moss_transcribe_diarize.py, preprocessor_config.json, processing_moss_transcribe_diarize.py, processor_config.json, special_tokens_map.json, tokenizer.json, tokenizer_config.json, vocab.json.
  • Random weights (seed 17). They passed the input-sensitivity gate: every perturbation of the contract's inputs moved the logits by at least 0.001 (weakest: input_features:halve, 0.29), so no training was needed.
  • Same tensor names as the real checkpoint. The real checkpoint has every parameter, and so does the fixture.
  • Saved in bfloat16, the dtype itazap/MOSS-Transcribe-Diarize-HF loads in by default.
  • Finite in float32, bfloat16, and float16 (forward, loss, gradients).

Built by build_fixtures.py; build.json records the pinned sources, gate results, and file hashes. This repository includes the slim weights, usage and framework expectations; rebuild the other files with python example/build_fixtures.py --fixture MOSS-Transcribe-Diarize-HF-mic.

Downloads last month
197
Safetensors
Model size
10M params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for huggingface-mic/MOSS-Transcribe-Diarize-HF-mic

Finetuned
(1)
this model

Space using huggingface-mic/MOSS-Transcribe-Diarize-HF-mic 1