grpo-5g-rca-qwen3-1.7b is an open-weight model from Mark Anthony. Its published files total 302.6 MB.
This model is a fine-tuned version of Qwen/Qwen3-1.7B. It has been trained using TRL. This model was trained with GRPO, a method introduced in DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models.
Model Card
This model is a fine-tuned version of Qwen/Qwen3-1.7B. It has been trained using TRL. This model was trained with GRPO, a method introduced in DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models.
Excerpt from the card by Mark Anthony.
Identity and Version
- Repository
- JesterbeanLTEA/grpo-5g-rca-qwen3-1.7b
- Publisher
- Mark Anthony
- Task
- Not stated by the source
- Modality
- Other
- Library
- transformers
- Parameters
- Not stated by the source
- Languages
- trl
- Revision
- b34b7a850f528c8def15e1966010a1c39ab2f1f0
- First published
- 2026-09-17
- Last updated
- 2026-09-26
Files and Weights
19 files, 302.6 MB in total. The weights are 7 files totalling 279.4 MB in bin, pt, pth, safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| adapter_model.safetensors | Weights | 69.8 MB | 477b6d44f404 |
| last-checkpoint/adapter_model.safetensors | Weights | 69.8 MB | 477b6d44f404 |
| last-checkpoint/optimizer.pt | Weights | 139.8 MB | 85cabf7b4265 |
| last-checkpoint/rng_state.pth | Weights | 14.2 KB | 5f2ef4300704 |
| last-checkpoint/scheduler.pt | Weights | 1.1 KB | 44339cee8bf5 |
| last-checkpoint/training_args.bin | Weights | 7.3 KB | 587cc3680ac7 |
| training_args.bin | Weights | 7.3 KB | 587cc3680ac7 |
| adapter_config.json | Configuration | 1.2 KB | — |
| last-checkpoint/adapter_config.json | Configuration | 1.2 KB | — |
| last-checkpoint/trainer_state.json | Configuration | 334.0 KB | — |
| README.md | Documentation | 2.0 KB | — |
| last-checkpoint/README.md | Documentation | 5.2 KB | — |
| chat_template.jinja | Other | 4.2 KB | — |
| last-checkpoint/chat_template.jinja | Other | 4.2 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| last-checkpoint/tokenizer.json | Tokenizer | 11.4 MB | be75606093db |
| last-checkpoint/tokenizer_config.json | Tokenizer | 749 B | — |
| tokenizer.json | Tokenizer | 11.4 MB | be75606093db |
| tokenizer_config.json | Tokenizer | 749 B | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 279.4 MB
Released by Mark Anthony through its official repository on Hugging Face.
Built From
- Derived from Qwen/Qwen3-1.7B
- Described by arXiv:2402.03300
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 279.4 MB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.