janhq/Jan-v3-4B-base-instruct download history
janhq/Jan-v3-4B-base-instruct is a 4.4B-parameter text generation model by janhq. In the last 30 days it was downloaded 510 times (127 in the last 7 days), and 7,690 times in total.
It ranks #70,998 on the Hub by monthly downloads and #17,100 among text generation models.
It has 62 likes.
It is a fine-tune of Qwen/Qwen3-4B-Instruct-2507.
49 models build on Jan-v3-4B-base-instruct: 38 quantized, 9 fine-tuned, 0 adapters and 2 merges. Together with the original they were downloaded 316,263 times in the last 30 days. See the Jan-v3-4B-base-instruct galaxy.
Most downloaded derivatives
- janhq/Jan-v3.5-4B-gguf (quantized), 285.3K downloads in 30 days
- janhq/Jan-v3-4B-base-instruct-gguf (quantized), 3.9K downloads in 30 days
- bartowski/janhq_Jan-v3-4B-base-instruct-GGUF (quantized), 2K downloads in 30 days
- aashish1904/Jan-v3-4B-base-instruct-GGUF (quantized), 1.4K downloads in 30 days
- janhq/Jan-v3.5-4B (finetune), 1.2K downloads in 30 days
- mradermacher/Jan-v3-4B-base-instruct-i1-GGUF (quantized), 601 downloads in 30 days
- kushagra909/Jan-v3-4B-base-instruct-gguf (quantized), 420 downloads in 30 days
- mlx-community/Jan-v3-4B-base-instruct-4bit (quantized), 358 downloads in 30 days
- mradermacher/Jan-v3-4B-base-instruct-GGUF (quantized), 353 downloads in 30 days
- janhq/Jan-code-4b (finetune), 159 downloads in 30 days
- mlx-community/Jan-v3-4B-base-instruct-8bit (quantized), 45 downloads in 30 days
- nightmedia/Qwen3-4B-Engineer14-qx86-hi-mlx (merge), 33 downloads in 30 days
Open janhq/Jan-v3-4B-base-instruct on Hugging Face
Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.