cosmo 1b needs about 3.2 GB of memory at Q4_K_M (a 1.1 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 14 quants, from 0.7 GB (Q2_K) to 1.9 GB (Q8_0).
Cosmo 1b – a 1.8B llama model for text completion and chat
Cosmo 1b is a 1.8B parameter llama model trained on a 30B token Cosmopedia synthetic dataset with additional code and educational data. It supports 2048 token context and is released under apache-2.0. The model is not instruction‑tuned but includes UltraChat format for chat usage.
Good at
- Text completion
- Chat with UltraChat format
- Educational code generation
Limits
- No vision support
- Not instruction‑tuned
- Requires 2048 token context
Best for
- Writing short stories
- Answering educational questions
- Generating code snippets
Who it’s for. Developers and researchers who need a lightweight llama model for text generation tasks. The model is suitable for local inference on machines with limited VRAM, as it offers multiple quantizations.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 0.7 GB | 1.6 GB | Fits | Get file → | |
| Q3_K_S | 0.8 GB | 1.7 GB | Fits | Get file → | |
| Q3_K_M | 0.9 GB | 1.8 GB | Fits | Get file → | |
| Q3_K_L | 0.9 GB | 1.8 GB | Fits | Get file → | |
| Q4_0 | 1.0 GB | 1.9 GB | Fits | Get file → | |
| Q4_K_S | 1.0 GB | 1.9 GB | Fits | Get file → | |
| Q4_K_M | 1.1 GB | 2.0 GB | Fits | Get file → | |
| Q4_1 | 1.1 GB | 2.0 GB | Fits | Get file → | |
| Q5_0 | 1.2 GB | 2.1 GB | Fits | Get file → | |
| Q5_K_S | 1.2 GB | 2.1 GB | Fits | Get file → | |
| Q5_K_M | 1.2 GB | 2.1 GB | Fits | Get file → | |
| Q5_1 | 1.3 GB | 2.2 GB | Fits | Get file → | |
| Q6_K | 1.4 GB | 2.3 GB | Fits | Get file → | |
| Q8_0 | 1.9 GB | 2.8 GB | Fits | Get file → |
For 8 GB VRAM use Q4_0 or Q4_K_S (≈3.1 GB memory at 8k context). For 4 GB VRAM use Q3_K_S or Q3_K_M (≈2.9–3 GB). For 2 GB VRAM use Q2_K (≈2.8 GB). File sizes range from 0.67 GB to 1.85 GB, so choose a quant that fits your storage and memory budget.
What GPU do you need for cosmo 1b?
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 (1.9 GB file, about 4.0 GB in use).
Which cosmo 1b quant should you choose?
Q4_K_M is the usual starting point for cosmo 1b: it balances file size and quality well. Here it is a 1.1 GB file.
The Q8_0 file is 1.9 GB, about 1.7 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 Q2_K at 0.7 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
Source
Frequently asked questions about cosmo 1b
How much VRAM does cosmo 1b need?
About 3.2 GB at Q4_K_M with an 8K context window. That is the 1.1 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 cosmo 1b on an 8 GB GPU?
Yes, with Q8_0: the 1.9 GB file needs about 4.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of cosmo 1b GGUF?
It depends on the quant: Q2_K is 0.7 GB, Q3_K_S is 0.8 GB, Q3_K_M is 0.9 GB, Q3_K_L is 0.9 GB, Q4_0 is 1.0 GB, Q4_K_S is 1.0 GB and 8 more.
What context length does cosmo 1b support?
The GGUF header lists a trained context length of 2,048 tokens (2K). The KV cache grows with the context you use: about 1.6 GB at 8K and 0.4 GB at 2K.
Where can I download cosmo 1b GGUF?
The files are on Hugging Face in the QuantFactory/cosmo-1b-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 cosmo 1b 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 Cosmo 1b on 8 GB VRAM?
Yes, use Q4_0 or Q4_K_S quantization, which requires about 3.1 GB VRAM at 8k context.
Does Cosmo 1b support image input?
No, the model has no vision capabilities.
Is the model instruction‑tuned?
No, it was not instruction‑tuned after pre‑training, but UltraChat format allows chat usage.