cohere-transcribe-03-2026-mlx-fp16
Converted MLX fp16 weights for CohereLabs/cohere-transcribe-03-2026.
Files
model.safetensorsโ converted MLX weights (cohere_transcribe_mlx.safetensors)config.jsontokenizer.modeltokenizer_config.jsonpreprocessor_config.jsonspecial_tokens_map.jsonkey_map.jsonโ conversion key mappingconversion_summary.jsonโ conversion metadata summary
Notes
- This repository contains conversion artifacts only.
- Refer to the upstream model card and license in
CohereLabs/cohere-transcribe-03-2026for usage/licensing details.
- Downloads last month
- 248
Model size
2B params
Tensor type
F16
ยท
Hardware compatibility
Log In to add your hardware
Quantized