kamel-usp
/

jbcs2025_Phi-3.5-mini-instruct-phi35_classification_lora-C1-full_context-r16

Eval Results (legacy)

Model card Files Files and versions

jbcs2025_Phi-3.5-mini-instruct-phi35_classification_lora-C1-full_context-r16 / README.md

abarbosa's picture

Pushing fine-tuned model to Hugging Face Hub

c8efabe verified about 1 year ago

|

History Blame Contribute Delete

1.22 kB


	---
	language:
	- pt
	- en
	tags:
	- aes
	datasets:
	- kamel-usp/aes_enem_dataset
	base_model: microsoft/Phi-3.5-mini-instruct
	metrics:
	- accuracy
	- qwk
	library_name: peft
	model-index:
	- name: Phi-3.5-mini-instruct-phi35_classification_lora-C1-full_context-r16
	results:
	- task:
	type: text-classification
	name: Automated Essay Score
	dataset:
	name: Automated Essay Score ENEM Dataset
	type: kamel-usp/aes_enem_dataset
	config: JBCS2025
	split: test
	metrics:
	- name: Macro F1
	type: f1
	value: 0.1697680097680097
	- name: QWK
	type: qwk
	value: 0.2635988065182465
	- name: Weighted Macro F1
	type: f1
	value: 0.2978287413070022
	---
	# Model ID: Phi-3.5-mini-instruct-phi35_classification_lora-C1-full_context-r16
	## Results
	\| \| test_data \|
	\|:-----------------\|------------:\|
	\| eval_accuracy \| 0.275362 \|
	\| eval_RMSE \| 46.4384 \|
	\| eval_QWK \| 0.263599 \|
	\| eval_Macro_F1 \| 0.169768 \|
	\| eval_Weighted_F1 \| 0.297829 \|
	\| eval_Micro_F1 \| 0.275362 \|
	\| eval_HDIV \| 0.00724638 \|