Qwen_3 Experimental-2
Collection
trajectory
β’
4 items
β’
Updated
β’
1
Crux-Qwen3_OpenThinking-4B is fine-tuned on the Qwen3-4B architecture, optimized for advanced open thinking, mathematical reasoning, and logical problem solving. This model is trained on the traces of sk1.1, which include 1,000 entries from the Gemini thinking trajectory, combined with fine-tuning on 100k tokens of open math reasoning data. This makes it highly effective for nuanced reasoning, educational tasks, and complex problem-solving requiring clear thought processes.
File Name | Size | Quantization | Format | Description |
---|---|---|---|---|
Crux-Qwen3_OpenThinking-4B-GGUF.F16.gguf |
8.05 GB | FP16 | GGUF | Float16 precision version |
Crux-Qwen3_OpenThinking-4B-GGUF.Q4_K_M.gguf |
2.5 GB | Q4_K_M | GGUF | 4-bit quantized (K M variant) |
Crux-Qwen3_OpenThinking-4B-GGUF.Q5_K_M.gguf |
2.89 GB | Q5_K_M | GGUF | 5-bit quantized (K M variant) |
Crux-Qwen3_OpenThinking-4B-GGUF.Q8_0.gguf |
4.28 GB | Q8_0 | GGUF | 8-bit quantized |
.gitattributes |
1.88 kB | β | β | Git LFS tracking file |
config.json |
31 B | β | β | Configuration file |
README.md |
31 B | β | β | Model documentation |
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
Link | Type | Size/GB | Notes |
---|---|---|---|
GGUF | Q2_K | 0.4 | |
GGUF | Q3_K_S | 0.5 | |
GGUF | Q3_K_M | 0.5 | lower quality |
GGUF | Q3_K_L | 0.5 | |
GGUF | IQ4_XS | 0.6 | |
GGUF | Q4_K_S | 0.6 | fast, recommended |
GGUF | Q4_K_M | 0.6 | fast, recommended |
GGUF | Q5_K_S | 0.6 | |
GGUF | Q5_K_M | 0.7 | |
GGUF | Q6_K | 0.7 | very good quality |
GGUF | Q8_0 | 0.9 | fast, best quality |
GGUF | f16 | 1.6 | 16 bpw, overkill |
Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):
4-bit
5-bit
8-bit
16-bit
Base model
Qwen/Qwen3-4B-Base