Downloads · 30 days
26
15% of all-time downloads
granite-4.0-micro-3b-cn-gguf
lucaelin/granite-4.0-micro-3b-cn-gguf
granite-4.0-micro-3b-cn-gguf is a machine learning model from lucaelin. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Quantized GGUF release for COVAS:NEXT fine-tuning experiments based on unsloth/granite-4.0-micro.
All-time downloads
171
Public
Repo size
3.6 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.gguf3.6 GB · 100%
At a glance
- Access
- Public
- Created
- Mar 4, 2026
- Updated
- Mar 4, 2026
- SHA
- 33e522a3
Model card
From the Hugging Face model README
granite-4.0-micro-3b-cn-gguf
Quantized GGUF release for COVAS:NEXT fine-tuning experiments based on unsloth/granite-4.0-micro.
Files
granite-4.0-micro-3b-cn-20250304-q8_0.gguf: Q8_0 GGUF from the latest stitched-dataset run.
Source run
- Training output:
output/granite_4_0_micro_trainingjson_stitched_e4_es_bs2_ga8_lr1e5 - Dataset:
data/training_stitched.jsonl(freshly stitched fromdata/training.jsonl) - Hyperparameters: epochs
4, batch size2, grad accum8, LR1e-5, max seq length16384
Eval snapshot (strict-6, 46 cases, 276 candidate responses)
- Overall:
245/276 (88.8%) - Tool calling:
90/108 (83.3%) - Event reaction:
39/42 (92.9%) - Contextual QA:
50/60 (83.3%) - Tool result summarization:
66/66 (100.0%)
Tool-calling micro metrics:
- Call made:
108/108 (100.0%) - Name correct:
102/108 (94.4%) - Args correct:
96/108 (88.9%)
See project experiment log in docs/experiments.md.
- Created
- Mar 4, 2026
- Updated
- Mar 4, 2026