Saved in this browser, used on every model page.
Yes
Yes. The Arc A580 runs Cola2 fully in VRAM at Q4_K_M (about 6.8 GB of 8.0 GB), at roughly 53 tokens/s.
- best quant
- Q4_K_M
- needs
- 6.8 GB
- usable
- 8.0 GB
- est. speed
- ~53 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the Cola2 page →
Every quantisation of Cola2 on the Arc A580
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 4.4 GB | fits | ~98 tok/s |
| Q3_K_M | 5.8 GB | fits | ~66 tok/s |
| Q4_K_M pick | 6.8 GB | fits | ~53 tok/s |
| Q5_K_M | 7.7 GB | tight | ~45 tok/s |
| Q6_K | 8.6 GB | offload | ~25 tok/s |
| Q8_0 | 11 GB | offload | ~10 tok/s |
| FP16 / BF16 | 19 GB | offload | ~3 tok/s |
Speed is a bandwidth-based estimate for generating text (512 GB/s); prompt processing and long contexts are slower.
Measured speed on the Arc A580
No community reports for this pair yet.
Report your own speed →Run Cola2 on the Arc A580
Transformers · Chat with the pipeline APIdocs
from transformers import pipeline
pipe = pipeline("text-generation", model="ByteDance-Seed/Cola2", 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 Arc A580 comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 8.2B | Q5_K_M | ≈7.3 GBFits in memory | ~48 tok/s | |
| 8B | Q5_K_M | ≈7.1 GBFits in memory | ~49 tok/s | |
| 7.6B | Q6_K | ≈7.0 GBFits in memory | ~45 tok/s | |
| 8B | Q5_K_M | ≈7.1 GBFits in memory | ~49 tok/s | |
| 8.8B | Q4_K_M | ≈7.0 GBFits in memory | ~53 tok/s | |
| 9.7B | Q4_K_M | ≈7.3 GBFits in memory | ~48 tok/s |
Other GPUs that run Cola2
Frequently asked questions
- Can the Arc A580 run Cola2?
- Yes. The Arc A580 runs Cola2 fully in VRAM at Q4_K_M (about 6.8 GB of 8.0 GB), at roughly 53 tokens/s.
- How fast does Cola2 run on the Arc A580?
- An estimate from memory bandwidth gives about 53 tokens/s at Q4_K_M.
- Which quantisation of Cola2 should I use on the Arc A580?
- Q4_K_M: the largest one that fits. Expected quality: small loss 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.