bartowski
Qwen2.5 Coder 1.5B Instruct
Qwen2.5 Coder 1.5B Instruct is a 1.54B dense qwen2 model from bartowski. Configured context length is 32,768 tokens. License: apache-2.0. Quant sizes are the GGUF file bytes published on Hugging Face when a file exists.
Parameters1.5B
Context32K
Attentiongqa
Licenseapache-2.0
28 layers · 2 KV heads · head dim 128 · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| F16 | 2.88 GB | Hugging Face file bytes |
| Q8_0 | 1.53 GB | Hugging Face file bytes |
| Q6_K | 2.42 GB | Hugging Face file bytes |
| Q5_K_M | 1.05 GB | Hugging Face file bytes |
| Q5_K_S | 1.02 GB | Hugging Face file bytes |
| Q4_K_M | 0.92 GB | Hugging Face file bytes |
| Q4_K_S | 0.88 GB | Hugging Face file bytes |
| IQ4_XS | 0.83 GB | Hugging Face file bytes |
| Q4_0 | 3.49 GB | Hugging Face file bytes |
| Q3_K_L | 0.82 GB | Hugging Face file bytes |
| Q3_K_M | 0.77 GB | Hugging Face file bytes |
| IQ3_M | 0.72 GB | Hugging Face file bytes |
| Q3_K_S | 0.71 GB | Hugging Face file bytes |
| Q2_K | 1.31 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Runs great | F16 | 48–65 |
| GeForce RTX 3060 12GB | Runs great | F16 | 59–81 |
| GeForce RTX 4060 Ti 16GB | Runs great | F16 | 48–65 |
| GeForce RTX 4090 | Runs great | F16 | 167–227 |
| GeForce RTX 5090 | Runs great | F16 | 296–404 |
| Apple M4 Pro | Runs great | F16 | 45–62 |
| Apple M5 | Runs great | F16 | 25–34 |
| CPU only · DDR5-5600 | Runs on CPU | F16 | 15–20 |
Smallest devices that fit it fully
- GeForce RTX 3050 6GB — Runs, F16, 28–38 tok/s
- GeForce RTX 2060 6GB — Runs, F16, 56–76 tok/s
- GeForce GTX 1660 Super — Runs, F16, 56–76 tok/s
- GeForce GTX 1660 Ti — Runs, F16, 48–65 tok/s
- Intel Arc A380 — Runs, F16, 31–42 tok/s
- GeForce RTX 5060 Ti 8GB — Runs great, F16, 74–101 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (F16).
llama.cpp
llama-server -hf bartowski/Qwen2.5-Coder-1.5B-Instruct-GGUF:F16 -c 8192 -ngl 28
Ollama
ollama run qwen2.5
LM Studio
lms get bartowski/Qwen2.5-Coder-1.5B-Instruct-GGUF # Then load quant F16