Model comparison
Granite 3.0 2b Instruct vs Llama 2-13B
Granite 3.0 2b Instruct is the stronger model overall, scoring 30.2 to 29.1 on the Noometry Index.
Last verified . 12 shared benchmarks.
MetaSummary
- They share 12 benchmarks with published results for both. Granite 3.0 2b Instruct scores higher in 6 categories and Llama 2-13B in 2 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Granite 3.0 2b Instruct leads 19.6 to 12.3.
Side by side
| Granite 3.0 2b Instruct | Llama 2-13B | |
|---|---|---|
| Provider | IBM | Meta |
| Noometry Index | 30.2 | 29.1 |
| Released | — | 2023-07-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 13 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-13B leads
Granite 3.0 2b Instruct: 27.8 (#311), Llama 2-13B: 30.4 (#287)
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Coding | 1090 | 1062 |
| BigCodeBench Instruct | 20.5% | — |
Reasoning Granite 3.0 2b Instruct leads
Granite 3.0 2b Instruct: 19.6 (#234), Llama 2-13B: 12.3 (#333)
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Hard Prompts | 1073 | 1051 |
| Chess Puzzles | — | 0% |
| DTBench | — | 42.2% |
| BIG-Bench Hard | — | 58.2% |
| Epoch Capabilities Index | — | 105.8 |
| HellaSwag | — | 80.7% |
| LAMBADA | — | 76.5% |
| PIQA | — | 80.8% |
| WinoGrande | — | 72.8% |
Math Granite 3.0 2b Instruct leads
Granite 3.0 2b Instruct: 29.7 (#215), Llama 2-13B: 28.7 (#226)
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Math | 1117 | 1065 |
| GSM8K | — | 36.9% |
Knowledge Too close to call
Granite 3.0 2b Instruct: 28.4 (#238), Llama 2-13B: 27.5 (#244)
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Expert | 1064 | 1030 |
| ARC (AI2) Challenge | — | 60.3% |
| BoolQ | — | 82.4% |
| MMLU | — | 55.6% |
| OpenBookQA | — | 57% |
| TriviaQA | — | 79.6% |
Multimodal Not comparable
Granite 3.0 2b Instruct: —, Llama 2-13B: —
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| ScienceQA | — | 55.8% |
Multilingual Too close to call
Granite 3.0 2b Instruct: 27.1 (#273), Llama 2-13B: 26.7 (#274)
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Non-English | 1033 | 1024 |
| LMArena Chinese | 1070 | 1001 |
| LMArena Russian | 1045 | 1055 |
| LMArena French | — | 1043 |
| LMArena German | — | 1009 |
| LMArena Japanese | — | 894 |
| LMArena Korean | — | 953 |
| LMArena Spanish | — | 1086 |
Instruction Following Too close to call
Granite 3.0 2b Instruct: 54.2 (#279), Llama 2-13B: 53.5 (#282)
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Instruction Following | 1056 | 1046 |
Long Context Too close to call
Granite 3.0 2b Instruct: 32.5 (#263), Llama 2-13B: 32.3 (#264)
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Longer Query | 1071 | 1064 |
Writing & Preference Too close to call
Granite 3.0 2b Instruct: 29.8 (#287), Llama 2-13B: 30.0 (#284)
| Benchmark | Granite 3.0 2b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Text | 1080 | 1085 |
| LMArena Creative Writing | 1046 | 1047 |
| LMArena Multi-Turn | 1053 | 1050 |
Frequently asked questions
Is Granite 3.0 2b Instruct better than Llama 2-13B?
Granite 3.0 2b Instruct is the stronger model overall, scoring 30.2 to 29.1 on the Noometry Index.
Is Granite 3.0 2b Instruct or Llama 2-13B better for coding?
Llama 2-13B scores higher on coding benchmarks: 30.4 versus 27.8 in the Noometry coding category.
How many benchmarks do Granite 3.0 2b Instruct and Llama 2-13B share?
12 benchmarks have published results for both models. Granite 3.0 2b Instruct has 13 scored results on Noometry and Llama 2-13B has 32.