unsloth
Qwen2.5 Coder 14B Instruct 128K
Qwen2.5 Coder 14B Instruct 128K is a 14.8B dense qwen2 model from unsloth. Configured context length is 131,072 tokens. License: apache-2.0. Quant sizes are the GGUF file bytes published on Hugging Face when a file exists.
Parameters14.8B
Context128K
Attentiongqa
Licenseapache-2.0
48 layers · 8 KV heads · head dim 128 · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| F16 | 27.5 GB | Hugging Face file bytes |
| Q8_0 | 14.6 GB | Hugging Face file bytes |
| Q6_K | 11.3 GB | Hugging Face file bytes |
| Q5_K_M | 9.79 GB | Hugging Face file bytes |
| Q4_K_M | 8.37 GB | Hugging Face file bytes |
| Q3_K_M | 6.84 GB | Hugging Face file bytes |
| Q2_K | 5.37 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Tight fit | Q2_K | 24–33 |
| GeForce RTX 3060 12GB | Tight fit | Q4_K_M | 19–25 |
| GeForce RTX 4060 Ti 16GB | Tight fit | Q6_K | 12–16 |
| GeForce RTX 4090 | Runs | Q8_0 | 32–44 |
| GeForce RTX 5090 | Tight fit | F16 | 32–44 |
| Apple M4 Pro | Tight fit | Q8_0 | 8.7–12 |
| Apple M5 | Tight fit | Q4_K_M | 7.9–11 |
| CPU only · DDR5-5600 | Runs on CPU | Q8_0 | 2.8–3.9 |
Smallest devices that fit it fully
- GeForce RTX 5060 Ti 8GB — Tight fit, Q2_K, 37–51 tok/s
- GeForce RTX 5060 — Tight fit, Q2_K, 37–51 tok/s
- GeForce RTX 5050 — Tight fit, Q2_K, 27–36 tok/s
- GeForce RTX 4060 Ti 8GB — Tight fit, Q2_K, 24–33 tok/s
- GeForce RTX 4060 — Tight fit, Q2_K, 23–31 tok/s
- GeForce RTX 3070 Ti — Tight fit, Q2_K, 51–69 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (Q6_K).
llama.cpp
llama-server -hf unsloth/Qwen2.5-Coder-14B-Instruct-128K-GGUF:Q6_K -c 8192 -ngl 48
Ollama
ollama run qwen2.5
LM Studio
lms get unsloth/Qwen2.5-Coder-14B-Instruct-128K-GGUF # Then load quant Q6_K