DeepSeek-R1 distills · DeepSeek
Fine-tuning DeepSeek-R1 Distill Llama 8B
What does it take to fine-tune DeepSeek-R1 Distill Llama 8B?
DeepSeek-R1 Distill Llama 8B needs 18 GB for half-precision LoRA training and 14 GB in four-bit, and 16 GB to serve. It supports supervised fine-tuning, preference tuning (dpo), under the Llama 3.1 Community licence. The cheapest qualifying class is RTX 4000 Ada at $0.09 per GPU-hour.
The same distillation applied to a Llama 3.1 8B base, so it inherits the Llama tooling ecosystem along with the reasoning behaviour. Choose this over the Qwen distill when the rest of your stack already assumes Llama tokenisation and chat formatting.
What to know before choosing it
Same distillation, Llama base. The reason to choose this over the Qwen distill is almost never quality — it is that your tokenisation, chat template and surrounding tooling already assume Llama, and mixing families inside one stack creates conversion work that produces subtle bugs.
Note the licence changes with the base: this inherits Llama 3.1 community terms rather than the Apache-2.0 of the Qwen distill. If licence simplicity is why you were looking at DeepSeek distills, the Qwen one is the right half of the pair.
Specification
| Parameters | 8B |
|---|---|
| Architecture | Dense |
| Size tier | Mid |
| LoRA training memory | 18 GBHalf precision, frozen base |
| QLoRA training memory | 14 GBFour-bit base, higher-precision adapter |
| Serving memory | 16 GBHalf precision, before attention cache |
| Objectives | SFT, DPO |
| Licence | Llama 3.1 Community |
| Repository | deepseek-ai/DeepSeek-R1-Distill-Llama-8B |
What it costs to train
Every class with enough memory for four-bit training, cheapest first. Total cost is the rate multiplied by wall time, so the cheapest rate is not always the cheapest run — a faster class that finishes sooner frequently wins.
| GPU class | VRAM | Training / hr | Serving / hr | Fits |
|---|---|---|---|---|
| RTX 4000 Ada | 20 GB | $0.09 | $0.11 | LoRA and QLoRA |
| L4 | 24 GB | $0.17 | $0.20 | LoRA and QLoRA |
| RTX 3090 | 24 GB | $0.22 | $0.26 | LoRA and QLoRA |
| RTX 4090 | 24 GB | $0.38 | $0.43 | LoRA and QLoRA |
| A40 | 48 GB | $0.42 | $0.48 | LoRA and QLoRA |
| RTX 6000 Ada | 48 GB | $0.61 | $0.71 | LoRA and QLoRA |
| A6000 | 48 GB | $0.65 | $0.75 | LoRA and QLoRA |
| L40S | 48 GB | $0.78 | $0.90 | LoRA and QLoRA |
| A100 40 GB | 40 GB | $1.17 | $1.35 | LoRA and QLoRA |
| A100 80 GB | 80 GB | $1.68 | $1.94 | LoRA and QLoRA |
| H100 80 GB | 80 GB | $2.59 | $2.99 | LoRA and QLoRA |
| H200 | 141 GB | $4.55 | $5.25 | LoRA and QLoRA |
Serving fits on RTX 4000 Ada at $0.11 per hour — before the attention cache, which grows with context length and concurrency.
Good starting point for
- Reasoning tasks inside an existing Llama stack
- Further reasoning fine-tunes
- Step-by-step explanation generation
Preference tuning on this model
Preference tuning holds a frozen reference copy of the model alongside the one being trained, so budget roughly 36 GB rather than 18 GB. That is the thing that catches people out — supervised training on this model fits on a class that preference tuning will overflow.
Other sizes in this family
| Model | Params | QLoRA | Objectives |
|---|---|---|---|
| DeepSeek-R1 Distill Qwen 7B | 7B | 12 GB | SFT, DPO |
Comparable sizes elsewhere
Frequently asked questions
How much VRAM does it take to fine-tune DeepSeek-R1 Distill Llama 8B?
18 GB for half-precision LoRA and 14 GB for four-bit QLoRA. Serving needs 16 GB. Preference tuning roughly doubles the training figure, because a frozen reference model is held alongside the one being trained.
What is the cheapest way to fine-tune DeepSeek-R1 Distill Llama 8B?
Four-bit QLoRA on RTX 4000 Ada at $0.09 per GPU-hour is the cheapest class that meets the 14 GB threshold. Note that quantised training is slower per step, so a faster class sometimes costs less over the whole run.
Can I download the weights after fine-tuning DeepSeek-R1 Distill Llama 8B?
Yes. Every finished run exposes its trained weights for download, and publishing to a model hub is a single call with a generated model card recording the base model and version the adapter applies to.
What licence does DeepSeek-R1 Distill Llama 8B carry?
Llama 3.1 Community. The licence follows the fine-tune — a derivative inherits the base model’s terms, and those terms pass to anyone you give the model to.
Last verified 6 August 2026. Memory thresholds are the platform's own admission limits.
Fine-tune DeepSeek-R1 Distill Llama 8B
Upload a dataset, forecast the run, and see the cost before any compute is leased.