Loading
Vous avez un workflow ComfyUI ? Obtenez une documentation lisible directement dans votre navigateur.Rendre un workflow lisible.Essayer maintenant ↗
A memory-efficient INT8 quantization of CohereLabs/cohere-transcribe-arabic-07-2026, Cohere Labs' 2B-parameter Arabic/English speech-recognition model. Produced with bitsandbytes LLM.int8() (8-bit weight...
Source du modèle
Description de la source
A memory-efficient INT8 quantization of
CohereLabs/cohere-transcribe-arabic-07-2026, Cohere Labs' 2B-parameter Arabic/English
speech-recognition model. Produced with bitsandbytes (8-bit weight quantization) for GPU inference via .
This quantization is .
Sources
1 sourceVérifié 13 sept.
Artefacts du modèle
1 artefactExtraits de sources
2 extraits
LLM.int8()transformers| Base model | CohereLabs/cohere-transcribe-arabic-07-2026 |
| Task | Automatic speech recognition (audio → text) |
| Architecture | Conformer encoder + Transformer decoder (attention encoder–decoder) |
| Languages | Arabic (incl. dialects), English |
| Quantization | bitsandbytes LLM.int8() (8-bit weight quantization) |
| Disk size | 2.35 GB (vs. 4.13 GB fp16) |
| License | Apache 2.0 |
Word Error Rate of this quantized model measured against the full fp16 model on 1,009 Saudi-dialect podcast clips (Arabic text normalized):
| Model | Size | WER vs. full fp16 |
|---|---|---|
| Full (fp16) | 4.13 GB | — (reference) |
| This model (INT8) | 2.35 GB | 0.46% |
The int8 build is identical to full-precision in WER; the int4 build differs by ~1.9%, mostly minor orthographic variation (e.g. hamza placement).
import torch
from transformers import AutoProcessor, CohereAsrForConditionalGeneration
from transformers.audio_utils import load_audio
processor = AutoProcessor.from_pretrained("NAMAA-Space/cohere-transcribe-arabic-07-2026-int8")
model = CohereAsrForConditionalGeneration.from_pretrained("NAMAA-Space/cohere-transcribe-arabic-07-2026-int8", device_map="auto")
audio = load_audio("your_audio.wav", sampling_rate=16000)
inputs = processor(audio, sampling_rate=16000, language="ar", return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=256)
print(processor.decode(outputs, skip_special_tokens=True))
Requirements: transformers, bitsandbytes, soundfile/librosa, and a CUDA GPU.
import torch
from transformers import BitsAndBytesConfig
quantization_config = BitsAndBytesConfig(load_in_8bit=True)
| # | Full (fp16) | This model |
|---|---|---|
| 1 | السلام عليكم ورحمة الله وبركاته. مساكم الله بالخير وحياكم الله معنا في ثالث حلقات لقا... | السلام عليكم ورحمة الله وبركاته. مساكم الله بالخير وحياكم الله معنا في ثالث حلقات لقا... |
| 2 | اهتمامات ولكن ولكن هذا لا يرقى انا عندي احصائيات قديمه وذكرت هذا في اكثر من مره في ال... | اهتمامات ولكن ولكن هذا لا يرقى انا عندي احصائيات قديمه وذكرت هذا في اكثر من مره في ال... |
| 3 | يتبع اوامر الله سبحانه وتعالى، ويصدق اخباره، ويرضى باقداره. هذا هو الذي يحسن الادب مع... | يتبع اوامر الله سبحانه وتعالى، ويصدق اخباره، ويرضى باقداره. هذا هو الذي يحسن الادب مع... |
transformers + bitsandbytes and requires a CUDA GPU. It is
not directly servable with vLLM (use a GPTQ / AWQ / compressed-tensors build for vLLM).Released under Apache 2.0. This is a quantized derivative of
CohereLabs/cohere-transcribe-arabic-07-2026 by Cohere and Cohere Labs; all original terms apply.
--- license: apache-2.0 base_model: CohereLabs/cohere-transcribe-arabic-07-2026 library_name: transformers pipeline_tag: automatic-speech-recognition language: - ar - en tags: - automatic-speech-recognition - arabic - saudi-dialect - quantized - bitsandbytes --- # Cohere Transcribe Arabic — INT8 A memory-efficient **INT8** quantization of [`CohereLabs/cohere-transcribe-arabic-07-2026`](https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026), Cohere Labs' 2B-parameter Arabic/English speech-recognition model. Produced with bitsandbytes `LLM.int8()` (8-bit weight quantization) for GPU inference via `transformers`. This quantization is **effectively lossless — identical WER to the full model (17.5% on Saudi podcasts)**. ## Model details | | | |---|---| | Base model | [`CohereLabs/cohere-transcribe-arabic-07-2026`](https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026) | | Task | Automatic speech recognition (audio → text) | | Architecture | Conformer encoder + Transformer decoder (attention encoder–decoder) | | Languages | Arabic (incl. dialects), English | | Quantization | bitsandbytes `LLM.int8()` (8-bit weight quantization) | | Disk size | **2.35 GB** (vs. 4.13 GB fp16) | | License | Apache 2.0 | ## Evaluation Word Error Rate of this quantized model measured against the full fp16 model on **1,009 Saudi-dialect podcast clips** (Arabic text normalized): | Model | Size | WER vs. full fp16 | |:--|:--:|:--:| | Full (fp16) | 4.13 GB | — (reference) | | **This model (INT8)** | **2.35 GB** | **0.46%** | The int8 build is identical to full-precision in WER; the int4 build differs by ~1.9%, mostly minor orthographic variation (e.g. hamza placement). ## Usage ```python import torch from transformers import AutoProcessor, CohereAsrForConditionalGeneration from transformers.audio_utils import load_audio processor = AutoProcessor.from_pretrained("NAMAA-Space/cohere-transcribe-arabic-07-2026-int8") model = CohereAsrForConditionalGeneration.from_pretrained("NAMAA-Space/cohere-transcribe-arabic-07-2026-int8", device_map="auto") audio = load_audio("your_audio.wav", sampling_rate=16000) inputs = processor(audio, sampling_rate=16000, language="ar", return_tensors="pt").to(model.device) outputs = model.generate(**inputs, max_new_tokens=256) print(processor.decode(outputs, skip_special_tokens=True)) ```...
Source context: 644 downloads · 2 likes · Pipeline automatic-speech-recognition · Library transformers · Repo NAMAA-Space/cohere-transcribe-arabic-07-2026-int8