Text Generation
PEFT
Safetensors
lora
visual-planning
provenance
report-generation
peer-review-artifact
Instructions to use Adonis3039/EviWeave-VisualPlanner-8B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Adonis3039/EviWeave-VisualPlanner-8B with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-8B") model = PeftModel.from_pretrained(base_model, "Adonis3039/EviWeave-VisualPlanner-8B") - Notebooks
- Google Colab
- Kaggle
| { | |
| "sft": { | |
| "path": "runs/visualplan_boundary_v2/train.sft.jsonl", | |
| "rows": 1920, | |
| "unique_tasks": 480, | |
| "trace_types": { | |
| "boundary_v2_final_only_sft": 480, | |
| "verifier_guided_repair_sft": 1440 | |
| }, | |
| "repair_rows": 1440, | |
| "invalid_json_targets": 0, | |
| "noncanonical_targets": 0, | |
| "column_shapes": { | |
| "base_task_uid|messages|score|task_uid|trace_type|uid": 1920 | |
| } | |
| }, | |
| "preferences": { | |
| "path": "runs/visualplan_boundary_v2/train.dpo.raw.jsonl", | |
| "rows": 1440, | |
| "unique_tasks": 480, | |
| "failure_reasons": { | |
| "wrong_diagram_type": 480, | |
| "missing_semantic_node": 480, | |
| "reversed_semantic_edge": 480 | |
| }, | |
| "column_shapes": { | |
| "base_task_uid|chosen|chosen_score|failure_reason|messages|pair_source|rejected|rejected_score|task_uid|uid": 1440 | |
| }, | |
| "invalid_chosen_json": 0, | |
| "invalid_rejected_json": 0, | |
| "noncanonical_chosen": 0, | |
| "identical_pairs": 0, | |
| "nonpositive_score_gaps": 0, | |
| "score_gap_mean": 0.5, | |
| "score_gap_min": 0.5 | |
| }, | |
| "passed": true, | |
| "failures": [] | |
| } | |