Compare LLM Models
Select 2-5 models to compare side by side.
Select models to compare
0 / 5 selected
Top downloads
Select 2-5 models
| Attribute |
bigcode
|
|---|---|
| Company | bigcode |
| Release Date | 2023-05-05 |
| Parameters (B) | 15.5 |
| Architecture | GPT-2 Transformer |
| Context Window | 8192 |
| Input Modalities | text |
| Open Source | Yes |
| License | bigcode-openrail-m |
| Score | 0 |
| VRAM (GB) | 32 |
| Compute | NVIDIA GPU with ~32GB VRAM for bf16/fp16 inference (e.g., A100 40GB); 8-bit quantization reduces requirements to ~17GB. Trained on 512 Tesla A100 GPUs over 24 days. |
| Benchmark: HumanEval | 33.6 |
| Benchmark: MBPP | 52.7 |
| Benchmark: MMLU | 33.9 |
| Pricing (Input) | — |
| Pricing (Output) | — |