Allen Institute for AI · allen-ai/olmo-2-1124-7b-instruct
| Provider | Allen Institute for AI |
|---|---|
| Family | olmo |
| Type | llm-chat |
| Status | active |
| Released | 2024-12-18 |
| Updated | 2025-01-06 |
| Parameters | 7.3B |
| Open weights | yes |
| Relationship | Model |
|---|---|
| Is a finetune of | allenai/OLMo-2-1124-7B-DPO |
Ordered by predicted speed. Bandwidth sets decode rate; memory decides whether it runs at all.
| Device | Memory | Bandwidth | Best quality | Weights | tok/s | Fastest |
|---|---|---|---|---|---|---|
| Cerebras WSE-3 | 44.0 GB | 21000000.0 GB/s | bf16 | 14.6 GB | ~1007040.1 | ~4028160.2 q4 |
| NVIDIA GB200 Grace Blackwell Superchip | 372.0 GB | 16000.0 GB/s | bf16 | 14.6 GB | ~767.3 | ~3069.1 q4 |
| AMD Instinct MI355X | 288.0 GB | 8000.0 GB/s | bf16 | 14.6 GB | ~383.6 | ~1534.5 q4 |
| NVIDIA B200 | 180.0 GB | 7700.0 GB/s | bf16 | 14.6 GB | ~369.2 | ~1477.0 q4 |
| Google TPU7x (Ironwood) | 192.0 GB | 7380.0 GB/s | bf16 | 14.6 GB | ~353.9 | ~1415.6 q4 |
| AMD Instinct MI325X | 256.0 GB | 6000.0 GB/s | bf16 | 14.6 GB | ~287.7 | ~1150.9 q4 |
| AMD Instinct MI300X | 192.0 GB | 5300.0 GB/s | bf16 | 14.6 GB | ~254.2 | ~1016.6 q4 |
| NVIDIA H200 SXM | 141.0 GB | 4800.0 GB/s | bf16 | 14.6 GB | ~230.2 | ~920.7 q4 |
| NVIDIA H100 NVL | 94.0 GB | 3900.0 GB/s | bf16 | 14.6 GB | ~187.0 | ~748.1 q4 |
| NVIDIA H100 SXM | 80.0 GB | 3350.0 GB/s | bf16 | 14.6 GB | ~160.6 | ~642.6 q4 |
| AMD Instinct MI250X | 128.0 GB | 3200.0 GB/s | bf16 | 14.6 GB | ~153.5 | ~613.8 q4 |
| Google TPU v5p | 95.0 GB | 2765.0 GB/s | bf16 | 14.6 GB | ~132.6 | ~530.4 q4 |
| NVIDIA A100 80GB SXM | 80.0 GB | 2039.0 GB/s | bf16 | 14.6 GB | ~97.8 | ~391.1 q4 |
| NVIDIA H100 PCIe | 80.0 GB | 2000.0 GB/s | bf16 | 14.6 GB | ~95.9 | ~383.6 q4 |
| NVIDIA GeForce RTX 5090 | 32.0 GB | 1792.0 GB/s | bf16 | 14.6 GB | ~85.9 | ~343.7 q4 |
| Google TPU v6e (Trillium) | 32.0 GB | 1638.0 GB/s | bf16 | 14.6 GB | ~78.5 | ~314.2 q4 |
| AMD Instinct MI210 | 64.0 GB | 1600.0 GB/s | bf16 | 14.6 GB | ~76.7 | ~306.9 q4 |
| NVIDIA A100 40GB SXM | 40.0 GB | 1555.0 GB/s | bf16 | 14.6 GB | ~74.6 | ~298.3 q4 |
| Google TPU v4 | 32.0 GB | 1200.0 GB/s | bf16 | 14.6 GB | ~57.5 | ~230.2 q4 |
| NVIDIA GeForce RTX 4090 | 24.0 GB | 1008.0 GB/s | bf16 | 14.6 GB | ~48.3 | ~193.4 q4 |
| AMD Radeon RX 7900 XTX | 24.0 GB | 960.0 GB/s | bf16 | 14.6 GB | ~46.0 | ~184.1 q4 |
| NVIDIA GeForce RTX 5080 | 16.0 GB | 960.0 GB/s | fp8 | 7.3 GB | ~92.1 | ~184.1 q4 |
| NVIDIA GeForce RTX 3090 | 24.0 GB | 936.0 GB/s | bf16 | 14.6 GB | ~44.9 | ~179.5 q4 |
| NVIDIA L40S | 48.0 GB | 864.0 GB/s | bf16 | 14.6 GB | ~41.4 | ~165.7 q4 |
| Apple M3 Ultra | 512.0 GB | 819.0 GB/s | bf16 | 14.6 GB | ~39.3 | ~157.1 q4 |
| AMD Radeon RX 7900 XT | 20.0 GB | 800.0 GB/s | bf16 | 14.6 GB | ~38.4 | ~153.5 q4 |
| Apple M2 Ultra | 192.0 GB | 800.0 GB/s | bf16 | 14.6 GB | ~38.4 | ~153.5 q4 |
| Google TPU v5e | 16.0 GB | 800.0 GB/s | fp8 | 7.3 GB | ~76.7 | ~153.5 q4 |
| NVIDIA GeForce RTX 4080 SUPER | 16.0 GB | 736.0 GB/s | fp8 | 7.3 GB | ~70.6 | ~141.2 q4 |
| NVIDIA GeForce RTX 4070 Ti SUPER | 16.0 GB | 672.0 GB/s | fp8 | 7.3 GB | ~64.5 | ~128.9 q4 |
| AMD Radeon RX 9070 XT | 16.0 GB | 640.0 GB/s | fp8 | 7.3 GB | ~61.4 | ~122.8 q4 |
| Apple M4 Max | 128.0 GB | 546.0 GB/s | bf16 | 14.6 GB | ~26.2 | ~104.7 q4 |
| Apple M1 Max | 64.0 GB | 400.0 GB/s | bf16 | 14.6 GB | ~19.2 | ~76.7 q4 |
| Apple M2 Max | 96.0 GB | 400.0 GB/s | bf16 | 14.6 GB | ~19.2 | ~76.7 q4 |
| Apple M3 Max | 128.0 GB | 400.0 GB/s | bf16 | 14.6 GB | ~19.2 | ~76.7 q4 |
| NVIDIA GeForce RTX 3060 12GB | 12.0 GB | 360.0 GB/s | fp8 | 7.3 GB | ~34.5 | ~69.1 q4 |
| NVIDIA L4 | 24.0 GB | 300.0 GB/s | bf16 | 14.6 GB | ~14.4 | ~57.5 q4 |
| NVIDIA GeForce RTX 4060 Ti 16GB | 16.0 GB | 288.0 GB/s | fp8 | 7.3 GB | ~27.6 | ~55.2 q4 |
| Apple M4 Pro | 64.0 GB | 273.0 GB/s | bf16 | 14.6 GB | ~13.1 | ~52.4 q4 |
| Apple M4 | 32.0 GB | 120.0 GB/s | bf16 | 14.6 GB | ~5.8 | ~23.0 q4 |
Taken from the card's availability section.
| Platform | Id on that platform |
|---|---|
| Hugging Face | allenai/OLMo-2-1124-7B-Instruct |
This card reports no benchmark scores yet.
| Benchmark | Catalogue standing | Score |
|---|