Autotrain compatible Base model:adapter:qwen/qwen3-... Base model:qwen/qwen3-4b Conversational Dataset:nvidia/openmathreasoni... En Endpoints compatible Finetuned Grpo Lora Math Mathematical-reasoning Qwen3 Reasoning Region:us Safetensors Sharded Tensorflow
Model Card on HF 🤗: https://huggingface.co/PinkPixel/crystal-think-v1
Crystal Think V1 Benchmarks
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Crystal Think V1 Parameters and Internals
| LLM Name | Crystal Think V1 |
| Repository 🤗 | https://huggingface.co/PinkPixel/crystal-think-v1 |
| Base Model(s) | |
| Model Size | 4b |
| Required VRAM | 8.1 GB |
| Updated | 2025-06-26 |
| Maintainer | PinkPixel |
| Model Files | |
| Supported Languages | en |
| Model Architecture | AutoModelForCausalLM |
| License | apache-2.0 |
| Model Max Length | 40960 |
| Is Biased | none |
| Tokenizer Class | Qwen2Tokenizer |
| Padding Token | <|vision_pad|> |
| PEFT Type | LORA |
| LoRA Model | Yes |
| PEFT Target Modules | up_proj|k_proj|v_proj|gate_proj|q_proj|o_proj|down_proj |
| LoRA Alpha | 64 |
| LoRA Dropout | 0 |
| R Param | 32 |
| Errors | replace |
Best Alternatives to Crystal Think V1
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Crystal Think V2 | 0K / 8.1 GB | 40 | 6 |
| Crystal Think V1.0 | 0K / 8.1 GB | 52 | 1 |
| MING MoE 4B | 0K / 0.3 GB | 17 | 1 |
Note: green Score (e.g. "73.2") means that the model is better than PinkPixel/crystal-think-v1 .
Rank the Crystal Think V1 Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 49328 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241124