Key Specifications
| Vendor | other |
|---|
| Version | j2-mid |
|---|
| Release Date | 2023-01-12 |
|---|
| Context Window | 8192 tokens |
|---|
| Input Modalities | text |
|---|
| Output Modalities | text |
|---|
| License | Proprietary |
|---|
| Documentation | https://huggingface.co/models |
|---|
Benchmark Performance
| Benchmark | Score | Unit | Evaluated At | Notes | Source |
|---|
| MMLU | 55.9 | % | 2023-01-12 | 5-shot | view |
| HUMANEVAL | 39.7 | pass@1 | 2023-01-12 | — | view |
| GSM8K | 33.2 | % | 2023-01-12 | 0-shot CoT | view |
| MATH | 20.3 | % | 2023-01-12 | 0-shot CoT | view |
| BBH | 51.8 | % | 2023-01-12 | 3-shot CoT | view |
| GPQA | 25.7 | % | 2023-01-12 | 0-shot | view |
| IFEVAL | 49.8 | % | 2023-01-12 | prompt_strict | view |
| ARC | 86 | % | 2023-01-12 | challenge | view |
| MUSR | 32.3 | % | 2023-01-12 | 0-shot | view |
| WINOGRANDE | 67.1 | % | 2023-01-12 | 0-shot | view |
Pricing
| Tier | Price | Currency |
|---|
| Input | $1 / Mtok | USD |
| Output | $1 / Mtok | USD |
| Cache Read | $0 / Mtok | USD |
| Cache Write | $0 / Mtok | USD |
Source:
https://huggingface.co/models
· as of 2023-01-12
Compliance
- Data Residency: self-host
- SOC2: ✗
- HIPAA: ✗
- GDPR: ✗
- ISO 27001: ✗
Jurassic-2 Mid
Aperçu du modèle
AI21 Labs Jurassic-2 Mid 中型模型, 8K 上下文, 多语言文本生成, 适合企业级内容创作。
Spécifications principales
| Fournisseur | Version | Date de sortie | Fenêtre de contexte | Modalités d’entrée | Modalités de sortie | Licence |
|---|
| Other | j2-mid | 2023-01-12 | 8K | text | text | Proprietary |
| Benchmark | Score | Unité | Notes |
|---|
| MMLU (Massive Multitask Language Understanding) | 55.9 | % | 5-shot |
| HumanEval | 39.7 | pass@1 | — |
| GSM8K (Grade School Math 8K) | 33.2 | % | 0-shot CoT |
| MATH | 20.3 | % | 0-shot CoT |
| BBH (BIG-Bench Hard) | 51.8 | % | 3-shot CoT |
| GPQA | 25.7 | % | 0-shot |
| IFEval | 49.8 | % | prompt_strict |
| ARC | 86.0 | % | challenge |
| MUSR | 32.3 | % | 0-shot |
| WinoGrande | 67.1 | % | 0-shot |
Tarification
| Entrée | Sortie | Lecture cache | Écriture cache |
|---|
| — | — | — | — |
par million de jetons
Forces
Faiblesses
- MMLU 仅 55.9，知识推理偏弱。
- HumanEval 39.7，代码能力较弱。
- 闭源专有模型，不支持自托管。
- 上下文窗口 8K 偏小。
Cas d’usage
Références