Mosaic AI Model Training - fine-tuning
Loading...
Current Models - Examples
| Model | Training word count | Approximate DBUs | Approximate cost/run ($0.65/DBU US East) |
|---|---|---|---|
| Llama 3.1 405B | 10,000,000 | 1,150 | $747.50 |
| 500,000,000 | 57,150 | $37,147.50 | |
| Llama 3.1 70B | 10,000,000 | 375 | $243.75 |
| 500,000,000 | 17,600 | $11,440.00 | |
| Llama 3.1 8B | 10,000,000 | 150 | $97.50 |
| 500,000,000 | 6,600 | $4,290.00 | |
| Llama 3.2 3B | 10,000,000 | 75 | $48.75 |
| 500,000,000 | 3,300 | $2,145.00 | |
| Llama 3.2 1B | 10,000,000 | 25 | $16.25 |
| 500,000,000 | 1,100 | $715.00 | |
| DBRX | 10,000,000 | 300 | $195.00 |
| 500,000,000 | 14,300 | $9,295.00 | |
| Mixtral 8x7B | 10,000,000 | 150 | $97.50 |
| 500,000,000 | 6,600 | $4,290.00 | |
| Mistral 7B | 10,000,000 | 50 | $32.50 |
| 500,000,000 | 1,325 | $861.25 |
Legacy Models (will be retired on Dec 13, 2024) - Examples
| Model | Training word count | Approximate DBUs | Approximate cost/run ($0.65/DBU US East) |
|---|---|---|---|
| Llama 3 70B | 10,000,000 | 375 | $243.75 |
| 500,000,000 | 17,600 | $11,440.00 | |
| Llama 3 8B | 10,000,000 | 150 | $97.50 |
| 500,000,000 | 6,600 | $4,290.00 | |
| Llama 2 70B | 10,000,000 | 275 | $178.75 |
| 500,000,000 | 13,200 | $8,580.00 | |
| Llama 2 13B | 10,000,000 | 50 | $32.50 |
| 500,000,000 | 2,475 | $1,608.75 | |
| Llama 2 7B | 10,000,000 | 25 | $16.25 |
| 500,000,000 | 1,175 | $763.75 | |
| Codellama 34B | 10,000,000 | 100 | $65.00 |
| 500,000,000 | 4,950 | $3,217.50 | |
| Codellama 13B | 10,000,000 | 75 | $48.75 |
| 500,000,000 | 2,650 | $1,722.50 | |
| Codellama 7B | 10,000,000 | 50 | $32.50 |
| 500,000,000 | 1,325 | $861.25 |