selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

pfnet logo

Can the RTX 4070 run plamo-3-8b-fin-instruct?

plamo-3-8b-fin-instruct·8.5B·GeForce RTX 4070·12 GB VRAM

Saved in this browser, used on every model page.
Yes

Yes. The RTX 4070 runs plamo-3-8b-fin-instruct fully in VRAM at Q8_0 (about 9.9 GB of 12 GB), at roughly 31 tokens/s.

best quant
Q8_0
needs
9.9 GB
usable
12 GB
est. speed
~31 tok/s

Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the plamo-3-8b-fin-instruct page →

Every quantisation of plamo-3-8b-fin-instruct on the RTX 4070

QuantQualityMemoryFitSpeed
Q2_K Heavy loss4.4 GBfits~99 tok/s
Q3_K_M Noticeable loss5.2 GBfits~66 tok/s
Q4_K_M Small loss6.2 GBfits~54 tok/s
Q5_K_M Near-lossless7.0 GBfits~46 tok/s
Q6_K Near-lossless7.8 GBfits~39 tok/s
Q8_0 pickLossless9.9 GBfits~31 tok/s
FP16 / BF16 Lossless18 GBoffload~4.8 tok/s

Speed is a bandwidth-based estimate for generating text (504 GB/s); prompt processing and long contexts are slower.

Measured speed on the RTX 4070

No community reports for this pair yet.

Report your own speed →

Run plamo-3-8b-fin-instruct on the RTX 4070

Using Q8_0, the best fit for RTX 4070.

Ollama · Q8_0 straight from Hugging Facedocs
ollama run hf.co/mradermacher/plamo-3-8b-fin-instruct-GGUF:Q8_0
llama.cpp · OpenAI-compatible server with a web UIdocs
llama-server -hf mradermacher/plamo-3-8b-fin-instruct-GGUF:Q8_0

Downloads the GGUF on first run. Installed through llama.app? The same flags work with `llama serve`.

LM Studio · Download with the lms CLIdocs
lms get mradermacher/plamo-3-8b-fin-instruct-GGUF@Q8_0

Then load it from the LM Studio app or with `lms load`.

vLLM · Serve an OpenAI-compatible APIdocs
vllm serve pfnet/plamo-3-8b-fin-instruct
vLLM · Docker Compose (NVIDIA)docs
services:
  vllm:
    image: vllm/vllm-openai:latest
    command: pfnet/plamo-3-8b-fin-instruct
    ports:
      - "8000:8000"
    ipc: host
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

API on port 8000.

Transformers · Chat with the pipeline APIdocs
from transformers import pipeline

pipe = pipeline("text-generation", model="pfnet/plamo-3-8b-fin-instruct", dtype="auto", device_map="auto")
messages = [{"role": "user", "content": "Explain self-hosting in one paragraph."}]
out = pipe(messages, max_new_tokens=512)
print(out[0]["generated_text"][-1]["content"])

Chat models that fit the RTX 4070 comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
8.2BQ8_0≈10 GBFits in memory~32 tok/sCommercial OK
8BQ8_0≈9.8 GBFits in memory~32 tok/sConditions
8BQ8_0≈9.8 GBFits in memory~32 tok/sConditions
7.6BQ8_0≈8.9 GBFits in memory~34 tok/sCommercial OK
9.4BQ8_0≈11 GBFits in memory~28 tok/sCommercial OK
9.2BQ6_K≈11 GBFits in memory~36 tok/sConditions

Other GPUs that run plamo-3-8b-fin-instruct

Frequently asked questions

Can the RTX 4070 run plamo-3-8b-fin-instruct?
Yes. The RTX 4070 runs plamo-3-8b-fin-instruct fully in VRAM at Q8_0 (about 9.9 GB of 12 GB), at roughly 31 tokens/s.
How fast does plamo-3-8b-fin-instruct run on the RTX 4070?
An estimate from memory bandwidth gives about 31 tokens/s at Q8_0.
Which quantisation of plamo-3-8b-fin-instruct should I use on the RTX 4070?
Q8_0: the largest one that fits. Expected quality: lossless compared with the full-precision weights.

Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.