jonatasgrosman/wav2vec2-large-xlsr-53-portuguese
jonatasgrosman/wav2vec2-large-xlsr-53-portuguese
Parameters
302M
Context
—
Downloads
4.9M
Likes
55
Architecture
wav2vec2
24 layers · head 64 · hidden 1024
GPU Requirements
VRAM / RAM / disk estimates by quantization and usage scenario
| Quantization | Minimum | Recommended | Production |
|---|---|---|---|
| FP1616bit | 4.1 GB RAM 32.0 · Disk 39.9 | 4.9 GB RAM 32.0 · Disk 65.9 | 6.3 GB RAM 32.0 · Disk 104.9 |
| FP88bit | 3.4 GB RAM 32.0 · Disk 39.4 | 4.2 GB RAM 32.0 · Disk 65.4 | 5.5 GB RAM 32.0 · Disk 104.4 |
| INT88bit | 3.4 GB RAM 32.0 · Disk 39.4 | 4.2 GB RAM 32.0 · Disk 65.4 | 5.5 GB RAM 32.0 · Disk 104.4 |
| AWQ4bit | 3.1 GB RAM 32.0 · Disk 39.2 | 3.8 GB RAM 32.0 · Disk 65.2 | 5.2 GB RAM 32.0 · Disk 104.2 |
| GPTQ4bit | 3.1 GB RAM 32.0 · Disk 39.2 | 3.8 GB RAM 32.0 · Disk 65.2 | 5.2 GB RAM 32.0 · Disk 104.2 |
| GGUF4bit | 3.1 GB RAM 32.0 · Disk 39.2 | 3.8 GB RAM 32.0 · Disk 65.2 | 5.2 GB RAM 32.0 · Disk 104.2 |
GPU Recommendations
1× RTX 5080
16 GB VRAM
≈ RTF 0.37
Effective production throughput ≈ 0.3 RTF
score 0.5585
1× Tesla V100
16 GB VRAM
≈ RTF 0 (N/A — benchmark unavailable)
Effective production throughput ≈ 0 RTF
score 0.5532
8× RTX 5080
128 GB VRAM
≈ RTF 0.11
Effective production throughput ≈ 0.1 RTF
score 0.3657
1× RTX 5060 Ti
8 GB VRAM
≈ RTF 0.79
Effective production throughput ≈ 0.6 RTF
score 0.6197
💬 Community — real deployment experience
0 Articles · 0 Benchmarks
GPUs that fit (single card)
Tesla V100
16 GB VRAM
$21.1
/mo · vast
RTX 5060 Ti
8 GB VRAM
$39.7
/mo · vast
RTX 3090
24 GB VRAM
$45.4
/mo · vast
RTX 4070S Ti
16 GB VRAM
$49.5
/mo · vast
RTX 5070
12 GB VRAM
$58.7
/mo · vast