Goekdeniz-Guelmez/ThoughtStream-4B-v0.2-gguf

Llamacpp Quantizations of trollek/ThoughtStream-4B-v0.2

Using llama.cpp for quantization.

Original model: https://huggingface.co./trollek/ThoughtStream-4B-v0.2

Run them in LM Studio, or Ollama

Prompt format

<|im_start|>user
{. PROMPT}<|im_end|>
<|im_start|>assistant
{. RESPONSE}<|im_end|>

Download a file (not the whole branch) from below:

Filename	Quant type	File Size	Split	Description
f16	16.07GB	false	Full F16 weights.
Q8_0	8.54GB	false	Extremely high quality, generally unneeded but max available quant.
Q6_K_L	6.85GB	false	Uses Q8_0 for embed and output weights. Very high quality, near perfect, recommended.
Q6_K	6.60GB	false	Very high quality, near perfect, recommended.
Q5_K_L	6.06GB	false	Uses Q8_0 for embed and output weights. High quality, recommended.
Q5_K_M	5.73GB	false	High quality, recommended.
Q5_K_S	5.60GB	false	High quality, recommended.
Q4_K_L	5.31GB	false	Uses Q8_0 for embed and output weights. Good quality, recommended.
Q4_K_M	4.92GB	false	Good quality, default size for must use cases, recommended.
Q3_K_XL	4.78GB	false	Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability.
Q4_K_S	4.69GB	false	Slightly lower quality with more space savings, recommended.
Q4_0	4.68GB	false	Legacy format, generally not worth using over similarly sized formats
Q4_0_8_8	4.66GB	false	Optimized for ARM inference. Requires 'sve' support (see link below).
Q4_0_4_8	4.66GB	false	Optimized for ARM inference. Requires 'i8mm' support (see link below).
Q4_0_4_4	4.66GB	false	Optimized for ARM inference. Should work well on all ARM chips, pick this if you're unsure.
IQ4_XS	4.45GB	false	Decent quality, smaller than Q4_K_S with similar performance, recommended.
Q3_K_L	4.32GB	false	Lower quality but usable, good for low RAM availability.
Q3_K_M	4.02GB	false	Low quality.
IQ3_M	3.78GB	false	Medium-low quality, new method with decent performance comparable to Q3_K_M.
Q2_K_L	3.69GB	false	Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable.
Q3_K_S	3.66GB	false	Low quality, not recommended.
IQ3_XS	3.52GB	false	Lower quality, new method with decent performance, slightly better than Q3_K_S.
Q2_K	3.18GB	false	Very low quality but surprisingly usable.
IQ2_M	2.95GB	false	Relatively low quality, uses SOTA techniques to be surprisingly usable.

Which file should I choose?

A great write up with charts showing various performances is provided by Artefact2 here

The first thing to figure out is how big a model you can run. To do this, you'll need to figure out how much RAM and/or VRAM you have.

If you want your model running as FAST as possible, you'll want to fit the whole thing on your GPU's VRAM. Aim for a quant with a file size 1-2GB smaller than your GPU's total VRAM.

If you want the absolute maximum quality, add both your system RAM and your GPU's VRAM together, then similarly grab a quant with a file size 1-2GB Smaller than that total.

Next, you'll need to decide if you want to use an 'I-quant' or a 'K-quant'.

If you don't want to think too much, grab one of the K-quants. These are in format 'QX_K_X', like Q5_K_M.

If you want to get more into the weeds, you can check out this extremely useful feature chart:

llama.cpp feature matrix

But basically, if you're aiming for below Q4, and you're running cuBLAS (Nvidia) or rocBLAS (AMD), you should look towards the I-quants. These are in format IQX_X, like IQ3_M. These are newer and offer better performance for their size.

These I-quants can also be used on CPU and Apple Metal, but will be slower than their K-quant equivalent, so speed vs performance is a tradeoff you'll have to decide.

The I-quants are not compatible with Vulcan, which is also AMD, so if you have an AMD card double check if you're using the rocBLAS build or the Vulcan build. At the time of writing this, LM Studio has a preview with ROCm support, and other inference engines have specific builds for ROCm.