ChartVerse-2B-GGUF
ChartVerse-2B from opendatalab is a compact 2B-parameter vision-language model fine-tuned from Qwen3-VL-2B-Instruct, specialized for complex chart reasoning as part of the ChartVerse project, achieving superior performance (54.3% average across 6 challenging benchmarks) over larger 7B chart-specific models like ECD-7B (50.0%), START-7B (52.5%), and Chart-R1-7B (53.6%) through high-quality training on ChartVerse-SFT-600K (412K unique charts, 603K QA pairs with 3.9B CoT tokens, rollout posterior entropy 0.44) and ChartVerse-RL-40K (40K hardest samples filtered by 0<r(Q)<1 failure rate) using LLaMA-Factory SFT (lr=1e-5, batch=128, ctx=22k) followed by veRL GSPO RL (lr=1e-6, rollout=16). Despite its small size, it demonstrates that data quality offsets scale limitations with strong Chain-of-Thought reasoning for multi-step analysis, truth-anchored QA verification via code execution, and Apache 2.0 licensing for deployment via Transformers/Qwen3VLForConditionalGeneration (max_new_tokens=16384) on standard GPUs, outperforming baselines in visual understanding while supporting project evaluation tools like VLMEvalKit.
ChartVerse-2B [GGUF]
| File Name | Quant Type | File Size | File Link |
|---|---|---|---|
| ChartVerse-2B.IQ4_XS.gguf | IQ4_XS | 1.18 GB | Download |
| ChartVerse-2B.Q2_K.gguf | Q2_K | 880 MB | Download |
| ChartVerse-2B.Q3_K_L.gguf | Q3_K_L | 1.14 GB | Download |
| ChartVerse-2B.Q3_K_M.gguf | Q3_K_M | 1.07 GB | Download |
| ChartVerse-2B.Q3_K_S.gguf | Q3_K_S | 1 GB | Download |
| ChartVerse-2B.Q4_K_M.gguf | Q4_K_M | 1.28 GB | Download |
| ChartVerse-2B.Q4_K_S.gguf | Q4_K_S | 1.24 GB | Download |
| ChartVerse-2B.Q5_K_M.gguf | Q5_K_M | 1.47 GB | Download |
| ChartVerse-2B.Q5_K_S.gguf | Q5_K_S | 1.44 GB | Download |
| ChartVerse-2B.Q6_K.gguf | Q6_K | 1.67 GB | Download |
| ChartVerse-2B.Q8_0.gguf | Q8_0 | 2.17 GB | Download |
| ChartVerse-2B.f16.gguf | F16 | 4.07 GB | Download |
| ChartVerse-2B.i1-IQ1_M.gguf | i1-IQ1_M | 646 MB | Download |
| ChartVerse-2B.i1-IQ1_S.gguf | i1-IQ1_S | 618 MB | Download |
| ChartVerse-2B.i1-IQ2_M.gguf | i1-IQ2_M | 829 MB | Download |
| ChartVerse-2B.i1-IQ2_S.gguf | i1-IQ2_S | 792 MB | Download |
| ChartVerse-2B.i1-IQ2_XS.gguf | i1-IQ2_XS | 734 MB | Download |
| ChartVerse-2B.i1-IQ2_XXS.gguf | i1-IQ2_XXS | 693 MB | Download |
| ChartVerse-2B.i1-IQ3_M.gguf | i1-IQ3_M | 1.03 GB | Download |
| ChartVerse-2B.i1-IQ3_S.gguf | i1-IQ3_S | 1 GB | Download |
| ChartVerse-2B.i1-IQ3_XS.gguf | i1-IQ3_XS | 968 MB | Download |
| ChartVerse-2B.i1-IQ3_XXS.gguf | i1-IQ3_XXS | 888 MB | Download |
| ChartVerse-2B.i1-IQ4_NL.gguf | i1-IQ4_NL | 1.23 GB | Download |
| ChartVerse-2B.i1-IQ4_XS.gguf | i1-IQ4_XS | 1.18 GB | Download |
| ChartVerse-2B.i1-Q2_K.gguf | i1-Q2_K | 880 MB | Download |
| ChartVerse-2B.i1-Q2_K_S.gguf | i1-Q2_K_S | 835 MB | Download |
| ChartVerse-2B.i1-Q3_K_L.gguf | i1-Q3_K_L | 1.14 GB | Download |
| ChartVerse-2B.i1-Q3_K_M.gguf | i1-Q3_K_M | 1.07 GB | Download |
| ChartVerse-2B.i1-Q3_K_S.gguf | i1-Q3_K_S | 1 GB | Download |
| ChartVerse-2B.i1-Q4_0.gguf | i1-Q4_0 | 1.23 GB | Download |
| ChartVerse-2B.i1-Q4_1.gguf | i1-Q4_1 | 1.34 GB | Download |
| ChartVerse-2B.i1-Q4_K_M.gguf | i1-Q4_K_M | 1.28 GB | Download |
| ChartVerse-2B.i1-Q4_K_S.gguf | i1-Q4_K_S | 1.24 GB | Download |
| ChartVerse-2B.i1-Q5_K_M.gguf | i1-Q5_K_M | 1.47 GB | Download |
| ChartVerse-2B.i1-Q5_K_S.gguf | i1-Q5_K_S | 1.44 GB | Download |
| ChartVerse-2B.i1-Q6_K.gguf | i1-Q6_K | 1.67 GB | Download |
| ChartVerse-2B.imatrix.gguf | imatrix | 2.09 MB | Download |
| ChartVerse-2B.mmproj-Q8_0.gguf | mmproj-Q8_0 | 445 MB | Download |
| ChartVerse-2B.mmproj-f16.gguf | mmproj-f16 | 819 MB | Download |
Quants Usage
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):
- Downloads last month
- 1,665