DE

deepseek-coder-1.3B-kexer

DeepSeekGGUF by bartowskiUpdated 2024-05-25
Download on Hugging Face
Parameters1.35B
Context16K
Layers24
Architecturellama
Smallest quant0.4 GB
Quick answer

deepseek-coder-1.3B-kexer needs about 3.0 GB of memory at Q4_K_M (a 0.9 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 23 quants, from 0.4 GB (IQ1_S) to 5.4 GB (F32).

1.3B DeepSeek Coder model for code completion

deepseek-coder-1.3B-kexer is a 1.35B parameter code model. It was trained on 2T tokens with 87% code and 13% natural language. The model supports a 16K context window and is optimized for project‑level code completion and infilling.

Good at

  • Code completion
  • Project‑level code infilling
  • Multi‑language coding

Limits

  • No vision support
  • Requires 8K context for memory estimates
  • Apache‑2.0 license

Best for

  • Small‑to‑medium code projects
  • Automated code review scripts
  • Educational coding tools

Who it’s for. Developers and researchers needing a lightweight code model with strong coding abilities. The model is suitable for local deployment via llama.cpp and supports commercial use under the Apache‑2.0 license.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S0.4 GB 2.5 GB
Fits Get file →
IQ1_M0.4 GB 2.6 GB
Fits Get file →
IQ2_XXS0.5 GB 2.6 GB
Fits Get file →
IQ2_XS0.5 GB 2.6 GB
Fits Get file →
IQ2_S0.5 GB 2.6 GB
Fits Get file →
IQ2_M0.6 GB 2.7 GB
Fits Get file →
Q2_K0.6 GB 2.7 GB
Fits Get file →
IQ3_XXS0.6 GB 2.7 GB
Fits Get file →
IQ3_XS0.6 GB 2.7 GB
Fits Get file →
IQ3_S0.6 GB 2.8 GB
Fits Get file →
Q3_K_S0.6 GB 2.8 GB
Fits Get file →
IQ3_M0.7 GB 2.8 GB
Fits Get file →
Q3_K_M0.7 GB 2.8 GB
Fits Get file →
IQ4_XS0.7 GB 2.9 GB
Fits Get file →
Q3_K_L0.7 GB 2.9 GB
Fits Get file →
IQ4_NL0.8 GB 2.9 GB
Fits Get file →
Q4_K_S0.8 GB 2.9 GB
Fits Get file →
Q4_K_M0.9 GB 3.0 GB
Fits Get file →
Q5_K_S1.0 GB 3.1 GB
Fits Get file →
Q5_K_M1.0 GB 3.1 GB
Fits Get file →
Q6_K1.2 GB 3.3 GB
Fits Get file →
Q8_01.4 GB 3.5 GB
Fits Get file →
F325.4 GB 7.5 GB
Tight Get file →

For 8 GB VRAM, use IQ1_S or IQ1_M (file size 0.42–0.44 GB, memory 2.5–2.6 GB). For 12 GB VRAM, IQ4_K_M or Q5_K_S (file size 0.87–0.95 GB, memory 3.0–3.1 GB). For 16 GB VRAM, Q6_K or Q8_0 (file size 1.17–1.43 GB, memory 3.3–3.5 GB).

What GPU do you need for deepseek-coder-1.3B-kexer?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: the best quant that fits is Q8_0 (1.4 GB file, about 3.5 GB in use).
  • 6 GB VRAM: the best quant that fits is Q8_0 (1.4 GB file, about 3.5 GB in use).
  • 8 GB VRAM: the best quant that fits is F32 (5.4 GB file, about 7.5 GB in use).

Which deepseek-coder-1.3B-kexer quant should you choose?

Q4_K_M is the usual starting point for deepseek-coder-1.3B-kexer: it balances file size and quality well. Here it is a 0.9 GB file.

The Q8_0 file is 1.4 GB, about 1.6 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

If memory is tight, the smallest file is IQ1_S at 0.4 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters1.35B
Layers24
Trained context16,384 tokens
KV cache (fp16)192 KB per token

Source

Quantized bybartowski
FamilyDeepSeek
Licenseapache-2.0
Downloads (HF)1,543
Last updated2024-05-25

Frequently asked questions about deepseek-coder-1.3B-kexer

How much VRAM does deepseek-coder-1.3B-kexer need?

About 3.0 GB at Q4_K_M with an 8K context window. That is the 0.9 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run deepseek-coder-1.3B-kexer on an 8 GB GPU?

Yes, with F32: the 5.4 GB file needs about 7.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of deepseek-coder-1.3B-kexer GGUF?

It depends on the quant: IQ1_S is 0.4 GB, IQ1_M is 0.4 GB, IQ2_XXS is 0.5 GB, IQ2_XS is 0.5 GB, IQ2_S is 0.5 GB, IQ2_M is 0.6 GB and 17 more.

What context length does deepseek-coder-1.3B-kexer support?

The GGUF header lists a trained context length of 16,384 tokens (16K). The KV cache grows with the context you use: about 1.6 GB at 8K and 3.2 GB at 16K.

Where can I download deepseek-coder-1.3B-kexer GGUF?

The files are on Hugging Face in the bartowski/deepseek-coder-1.3B-kexer-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does deepseek-coder-1.3B-kexer use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run this model on 8 GB VRAM?

Yes, use IQ1_S or IQ1_M quantizations; they require about 2.5–2.6 GB of VRAM at 8K context.

Does the model support Chinese programming languages?

The training data includes Chinese, so the model can handle code written in Chinese comments and identifiers.

Is there a chat template for this model?

No chat template is specified; the default llama.cpp template is used.

Other GGUFs of deepseek-coder-1.3b-base

Similar models

Browse all →

More from bartowski

All by bartowski →

More DeepSeek models

All DeepSeek →

Other tiny models

All tiny models →