Instructions to use evalengine/decision-0.8b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use evalengine/decision-0.8b with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3.5-0.8B") model = PeftModel.from_pretrained(base_model, "evalengine/decision-0.8b") - Notebooks
- Google Colab
- Kaggle
Download evaluation/failed-bf16-probability-check.json from evalengine/decision-0.8b: direct link, hf CLI and curl.
- Browser
- Download file 1.88 kB
-
https://huggingface.co/evalengine/decision-0.8b/resolve/main/evaluation/failed-bf16-probability-check.json
- Command line
-
hf download hf://evalengine/decision-0.8b/evaluation/failed-bf16-probability-check.json
-
curl -L -o failed-bf16-probability-check.json https://huggingface.co/evalengine/decision-0.8b/resolve/main/evaluation/failed-bf16-probability-check.json
1.88 kB
| { | |
| "status": "failed", | |
| "verification": { | |
| "records": 32, | |
| "prediction_changes": 0, | |
| "changed_indices": [], | |
| "max_absolute_probability_difference": 0.020078718662261963, | |
| "probability_atol": 0.02, | |
| "passed": false, | |
| "scope": "Before merge versus saved-and-reloaded BF16 export, raw probabilities", | |
| "records_sha256": "21977c3441de6aff6f320084c0531e4131996b62b22c80b7d4c9cc158a7b6fa2", | |
| "record_ids": [ | |
| "mnli:train:207170", | |
| "boolq:train:3417", | |
| "research-v21:dev:reasoning-methods:1:0:0", | |
| "research-v21:dev:reasoning-methods:1:0:1", | |
| "research-v21:dev:reasoning-methods:1:0:2", | |
| "research-v21:dev:reasoning-methods:1:0:3", | |
| "research-v21:dev:reasoning-methods:1:1:0", | |
| "research-v21:dev:reasoning-methods:1:1:1", | |
| "research-v21:dev:reasoning-methods:1:1:2", | |
| "research-v21:dev:reasoning-methods:1:1:3", | |
| "mnli:train:1737", | |
| "helpsteer2:validation:4", | |
| "helpsteer2:validation:5", | |
| "research-v21:dev:model-efficiency:1:0:0", | |
| "research-v21:dev:model-efficiency:1:0:1", | |
| "research-v21:dev:model-efficiency:1:0:2", | |
| "research-v21:dev:model-efficiency:1:0:3", | |
| "research-v21:dev:model-efficiency:1:1:0", | |
| "research-v21:dev:model-efficiency:1:1:1", | |
| "research-v21:dev:model-efficiency:1:1:2", | |
| "research-v21:dev:model-efficiency:1:1:3", | |
| "go_emotions:validation:3362", | |
| "mnli:train:229990", | |
| "boolq:train:664", | |
| "boolq:train:1952", | |
| "policy_v2:dev:168:complete_positive", | |
| "policy_v2:dev:168:decisive_missing", | |
| "policy_v2:dev:168:multi_missing", | |
| "policy_v2:dev:168:one_fact_flip", | |
| "policy_v2:dev:168:partial_negative", | |
| "policy_v2:dev:168:partial_positive", | |
| "mnli:train:129242" | |
| ] | |
| }, | |
| "source_manifest_sha256": "f4c0a1a02b43ee25da5901ade14cdf6d36ff991e29855d6274685d4b9d9b8def" | |
| } | |