Mistral AI
Codestral 22B
Aktuell
11.3KDownloads1.3KLikesMay 2024Veröffentlicht33K TokenKontextMNPLLizenz27 EinstiegQualität
Codestral 22B (22B parameters) requires approximately 17.7 GB of VRAM with Q4_K_M quantization. For the best balance of quality and speed, we recommend hardware with at least 21 GB of VRAM.
Loslegen
— kopieren & einfügen, um lokal auszuführenCopy-paste commands to run Codestral 22B on your machine.
Run
ollama run codestralQuick specs
Parameters22B
Architecturedense
Context33K tokens
Modalitytext
Min RAM8.6 GB
Rec. RAM13.4 GB (Q4_K_M)
LicenseMNPL
FamilyCodestral
About this model
- •As instruct, for instance to answer any questions about a code snippet (write documentation, explain, factorize) or to generate code following...
- •As Fill in the Middle (FIM), to predict the middle tokens between a prefix and a suffix (very useful for software development add-ons like in VS Code)
Verwandte Modelle
Schnellauswahl
Beste Hardware
Top-Empfehlungen für Codestral 22B
Dieses Modell ausführen
Quantisierungsoptionen
VRAM-Schätzungen nach Quantisierungsstufe
No hardware detected — fit column shows raw VRAM estimates
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 8.6 GB | Low | — |
Q3_K_S | 3 | 10.8 GB | Low | — |
NVFP4 | 4 | 12.3 GB | Medium | — |
Q4_K_M | 4 | 13.4 GB | Medium | — |
Q5_K_M | 5 | 15.8 GB | High | — |
Q6_K | 6 | 18.0 GB | High | — |
Q8_0 | 8 | 23.5 GB | Very High | — |
F16 | 16 | 45.1 GB | Maximum | — |
Quality benchmarks
Codestral 22B benchmark scores
Coding
SWE-bench Verified—
HumanEval+81.1%
Aider Polyglot—
LiveCodeBench—
Reasoning
MMLU-Pro24.0%
GPQA Diamond6.5%
MATH-50010.0%
ARC Challenge—
General
Chatbot Arena—
IFEval57.7%
Source: official · 2024-05-29
Hardware-Kompatibilität
Eignungsschätzungen für alle Hardware
Computing compatibility...
Speicheraufschlüsselung
Reference: RTX 2060 6GB
Weights13.4 GB
KV Cache2.4 GB
Runtime1.2 GB
Headroom0.6 GB
Häufig gestellte Fragen
FAQ — Codestral 22B
Siehe auch