Text Classification
PEFT
Safetensors
kev
jev
decision-model
calibration
router
code-verification
lora
Instructions to use jtatman/kev-4b-code-verify-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use jtatman/kev-4b-code-verify-v1 with PEFT:
from peft import PeftModel from transformers import AutoModel base_model = AutoModel.from_pretrained("Qwen/Qwen3.5-4B-Base") model = PeftModel.from_pretrained(base_model, "jtatman/kev-4b-code-verify-v1") - Notebooks
- Google Colab
- Kaggle
Download run_eval.json from jtatman/kev-4b-code-verify-v1: direct link, hf CLI and curl.
- Browser
- Download file 3.9 kB
-
https://huggingface.co/jtatman/kev-4b-code-verify-v1/resolve/main/run_eval.json
- Command line
-
hf download hf://jtatman/kev-4b-code-verify-v1/run_eval.json
-
curl -L -o run_eval.json https://huggingface.co/jtatman/kev-4b-code-verify-v1/resolve/main/run_eval.json
3.9 kB
| { | |
| "development": { | |
| "baseline kev": { | |
| "auroc": 0.9595338983050847, | |
| "acc": 0.9420289855072463, | |
| "brier": 0.051967926948896535, | |
| "ece": 0.08224694059061195, | |
| "cov@2%": 0.8188405797101449, | |
| "cov@5%": 0.8731884057971014, | |
| "cov@10%": 0.9492753623188406 | |
| }, | |
| "fine-tuned kev": { | |
| "auroc": 0.9582627118644068, | |
| "acc": 0.967391304347826, | |
| "brier": 0.03196338159102184, | |
| "ece": 0.04480786530262329, | |
| "cov@2%": 0.7391304347826086, | |
| "cov@5%": 0.8985507246376812, | |
| "cov@10%": 0.9492753623188406 | |
| }, | |
| "evidence score": { | |
| "auroc": 0.9599576271186441, | |
| "acc": 0.9710144927536232, | |
| "brier": 0.03740658967391304, | |
| "ece": 0.07269021739130432, | |
| "cov@2%": 0.7898550724637681, | |
| "cov@5%": 0.8985507246376812, | |
| "cov@10%": 0.9239130434782609 | |
| }, | |
| "mean(evidence, baseline)": { | |
| "auroc": 0.9653601694915255, | |
| "acc": 0.9492753623188406, | |
| "brier": 0.03681774037976611, | |
| "ece": 0.07842630691187563, | |
| "cov@2%": 0.7934782608695652, | |
| "cov@5%": 0.8985507246376812, | |
| "cov@10%": 0.9492753623188406 | |
| }, | |
| "mean(evidence, fine-tuned)": { | |
| "auroc": 0.9679025423728813, | |
| "acc": 0.967391304347826, | |
| "brier": 0.030796113996718024, | |
| "ece": 0.054936653462283394, | |
| "cov@2%": 0.8007246376811594, | |
| "cov@5%": 0.8985507246376812, | |
| "cov@10%": 0.9492753623188406 | |
| } | |
| }, | |
| "test_unseen_coder": { | |
| "baseline kev": { | |
| "auroc": 0.9663382594417077, | |
| "acc": 0.9295774647887324, | |
| "brier": 0.06354770562384918, | |
| "ece": 0.04240858525824696, | |
| "cov@2%": 0.28169014084507044, | |
| "cov@5%": 0.4788732394366197, | |
| "cov@10%": 0.647887323943662 | |
| }, | |
| "fine-tuned kev": { | |
| "auroc": 0.9860426929392446, | |
| "acc": 0.9154929577464789, | |
| "brier": 0.06591741550271953, | |
| "ece": 0.04675058358510263, | |
| "cov@2%": 0.43661971830985913, | |
| "cov@5%": 0.5915492957746479, | |
| "cov@10%": 0.647887323943662 | |
| }, | |
| "evidence score": { | |
| "auroc": 0.9433497536945813, | |
| "acc": 0.9154929577464789, | |
| "brier": 0.08483714788732394, | |
| "ece": 0.1338028169014084, | |
| "cov@2%": 0.15492957746478872, | |
| "cov@5%": 0.15492957746478872, | |
| "cov@10%": 0.5915492957746479 | |
| }, | |
| "mean(evidence, baseline)": { | |
| "auroc": 0.9794745484400657, | |
| "acc": 0.9154929577464789, | |
| "brier": 0.06687451167408419, | |
| "ece": 0.08106238257021522, | |
| "cov@2%": 0.4225352112676056, | |
| "cov@5%": 0.5774647887323944, | |
| "cov@10%": 0.6197183098591549 | |
| }, | |
| "mean(evidence, fine-tuned)": { | |
| "auroc": 0.9712643678160919, | |
| "acc": 0.9154929577464789, | |
| "brier": 0.06803778906539469, | |
| "ece": 0.07175734922897826, | |
| "cov@2%": 0.38028169014084506, | |
| "cov@5%": 0.5774647887323944, | |
| "cov@10%": 0.6056338028169014 | |
| } | |
| }, | |
| "test_unseen_coder_all_tasks": { | |
| "baseline kev": { | |
| "auroc": 0.9704155219780219, | |
| "acc": 0.9241803278688525, | |
| "brier": 0.05993135831653411, | |
| "ece": 0.031035312001956816, | |
| "cov@2%": 0.3114754098360656, | |
| "cov@5%": 0.514344262295082, | |
| "cov@10%": 0.6229508196721312 | |
| }, | |
| "fine-tuned kev": { | |
| "auroc": 0.9739182692307692, | |
| "acc": 0.9385245901639344, | |
| "brier": 0.04889003537440916, | |
| "ece": 0.028302458894132104, | |
| "cov@2%": 0.3442622950819672, | |
| "cov@5%": 0.5430327868852459, | |
| "cov@10%": 0.6352459016393442 | |
| }, | |
| "evidence score": { | |
| "auroc": 0.943543956043956, | |
| "acc": 0.9057377049180327, | |
| "brier": 0.09111320255251247, | |
| "ece": 0.09400301997638061, | |
| "cov@2%": 0.0, | |
| "cov@5%": 0.20901639344262296, | |
| "cov@10%": 0.6188524590163934 | |
| }, | |
| "mean(evidence, baseline)": { | |
| "auroc": 0.9705013736263737, | |
| "acc": 0.9262295081967213, | |
| "brier": 0.06611740777073485, | |
| "ece": 0.06348054507231347, | |
| "cov@2%": 0.3094262295081967, | |
| "cov@5%": 0.5, | |
| "cov@10%": 0.6270491803278688 | |
| }, | |
| "mean(evidence, fine-tuned)": { | |
| "auroc": 0.9726304945054945, | |
| "acc": 0.9364754098360656, | |
| "brier": 0.05954172687246182, | |
| "ece": 0.05905205255287546, | |
| "cov@2%": 0.3463114754098361, | |
| "cov@5%": 0.5061475409836066, | |
| "cov@10%": 0.6352459016393442 | |
| } | |
| } | |
| } |