It seems Q6_K version of gemma-3-4b-it.

#1
by Bedovyy - opened

when you loading, you can see,

llama_model_loader: - type  f32:  205 tensors
llama_model_loader: - type q6_K:  239 tensors
print_info: file format = GGUF V3 (latest)
print_info: file type   = Q6_K
print_info: file size   = 2.97 GiB (6.56 BPW)

it also have only 33 layers.
The file size also matched with 4B Q6_K model.

I don't understand why some people doing this kind-a-scam.

You are correct that this is the 4B Q6_K model but it is not a scam. I set the wrong file when uploading the model so it is a mistake on my end. I'll get it fixed and I aplogize for unintentionally being misleading

Yeah, everone said that. reflection 70B said that.

Here's log of gemma-3-27b-it-crof.gguf, by the way.

llama_model_loader: loaded meta data with 35 key-value pairs and 444 tensors from AI-11/gemma-3-27b-it-crof.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv   0:                       general.architecture str              = gemma3
llama_model_loader: - kv   1:                               general.type str              = model
llama_model_loader: - kv   2:                               general.name str              = Gemma-3-27B-It
llama_model_loader: - kv   3:                       general.quantized_by str              = CrofAI
llama_model_loader: - kv   4:                         general.size_label str              = 27B
llama_model_loader: - kv   5:                           general.repo_url str              = https://huggingface.co/tyraepaul/gemm...
...
print_info: model type       = 4B
print_info: model params     = 3.88 B
print_info: general.name     = Gemma-3-27B-It
...
Your need to confirm your account before you can post a new comment.

Sign up or log in to comment