ninadp

marathi-mitra-phi3-v3

Deploy Dedicated

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Container

Run this model inference with full control and performance in your environment.

Learn more

Get help setting up a custom Dedicated Endpoints.

Talk with our engineer to get a quote for reserved GPU instances with discounts.

Model Description

Table with columns: Property, Value
Property	Value
Base Model	microsoft/Phi-3-mini-4k-instruct
Fine-tuning	QLoRA + Optuna HPO
LoRA Rank	r=64, alpha=128
Training Examples	250
Optuna Trials	20 (TPE sampler, A100)
Optimized For	Unseen word generalisation

Performance

Table with columns: Words, Score
Words	Score
Seen words	76.0%
Unseen words	82.0%
Overall	79.0%
Generalisation gap	-6.0% (unseen > seen)

Key Finding

Optuna was configured to maximize unseen word score. This produced a negative generalisation gap (-6%) where the model performs better on words it never saw during training.

However overall score (79.0%) is lower than v2 (89.4%), demonstrating metric-objective misalignment — optimizing for a single metric (unseen) hurt the overall performance.

Lesson: HPO objective should be (seen + unseen) / 2 not just unseen score alone.

Best Config Found by Optuna

Table with columns: Parameter, Value
Parameter	Value
Learning rate	2.36e-4
Epochs	32
LoRA rank	64
LoRA alpha	128
Quantization	4-bit

Recommended Version

For production use, v2 achieves higher overall score (89.4% vs 79.0%). v3 is useful as a research artifact demonstrating generalisation vs accuracy trade-offs in HPO.

Links

Model provider

ninadp

Model tree

Base

microsoft/Phi-3-mini-4k-instruct

Adapter

this model

Modalities

Input

Text

Output

Text

Pricing

Dedicated Endpoints

View details

Supported Functionality

Model APIs

Dedicated Endpoints

Container

More information

Model card

Explore FriendliAI today

Get started Talk to an engineer

Model Description

Table with columns: Property, Value
Property	Value
Base Model	microsoft/Phi-3-mini-4k-instruct
Fine-tuning	QLoRA + Optuna HPO
LoRA Rank	r=64, alpha=128
Training Examples	250
Optuna Trials	20 (TPE sampler, A100)
Optimized For	Unseen word generalisation

Performance

Table with columns: Words, Score
Words	Score
Seen words	76.0%
Unseen words	82.0%
Overall	79.0%
Generalisation gap	-6.0% (unseen > seen)

Key Finding

Optuna was configured to maximize unseen word score. This produced a negative generalisation gap (-6%) where the model performs better on words it never saw during training.

However overall score (79.0%) is lower than v2 (89.4%), demonstrating metric-objective misalignment — optimizing for a single metric (unseen) hurt the overall performance.

Lesson: HPO objective should be (seen + unseen) / 2 not just unseen score alone.

Best Config Found by Optuna

Table with columns: Parameter, Value
Parameter	Value
Learning rate	2.36e-4
Epochs	32
LoRA rank	64
LoRA alpha	128
Quantization	4-bit

Recommended Version

For production use, v2 achieves higher overall score (89.4% vs 79.0%). v3 is useful as a research artifact demonstrating generalisation vs accuracy trade-offs in HPO.

marathi-mitra-phi3-v3

Get help setting up a custom Dedicated Endpoints.

README

Model Description

Performance

Key Finding

Best Config Found by Optuna

Recommended Version

Links

Explore FriendliAI today

README

Model Description

Performance

Key Finding

Best Config Found by Optuna

Recommended Version

Links