NIVED2003

gemma-4-E4B-dora-poetry-half2

README

License: apache-2.0

Model Details

Base Model: google/gemma-4-E4B
Tuning Method: DoRA (Weight-Decomposed Low-Rank Adaptation)
Rank ( $r$ ): 16
Alpha ( $α$ ): 32 ( $α / r = 2.0$ )
LoRA Dropout: 0.05
Target Modules: All 7 linear projection layers in the 42 text decoder layers:
- q_proj, k_proj, v_proj, o_proj (Self-Attention)
- gate_proj, up_proj, down_proj (MLP)
Trainable Parameters: 37,925,888 / 8,034,082,336 (0.472%)
Language Pair: Sanskrit (sa) $\to$ Hindi (hi)
Domain: Classical Sanskrit Poetry (Epics, Stotras, Subhashitas, Kavyas)

Evaluation & Training Dynamics

The model was trained for 2 epochs on 29,947 parallel verse pairs and evaluated on 3,778 unseen validation verses.

Validation Loss & Perplexity Trajectory

Table with columns: Milestone, Checkpoint Step, Epoch, Eval Loss (Cross-Entropy), Perplexity (PPL=eloss)
Milestone	Checkpoint Step	Epoch	Eval Loss (Cross-Entropy)	Perplexity ( $PPL = e loss$ )
Milestone 25%	936	0.50	`2.617`	13.69
Milestone 50%	1,872

Within-Domain Stability (Half 1 vs. Half 2)

Comparing both halves of the independent poetry splits confirms high empirical consistency:

Table with columns: Metric, Poetry Half 1, Poetry Half 2, Absolute Delta (Δ), Relative Variance
Metric	Poetry Half 1	Poetry Half 2	Absolute Delta ( $Δ$ )	Relative Variance
Validation Loss	`2.323`	`2.330`	`0.007`	< 0.30%
Validation Perplexity	`10.20`	`10.28`

This tight convergence demonstrates that DoRA preserves structural consistency and semantic alignment across disparate verse selections.

Training Hyperparameters

Hardware: 1x NVIDIA H100 80GB SXM5 GPU (Sapphire Rapids host)
Runtime: 2 hours 19 minutes (8,340s)
Throughput: 7.18 samples/sec (0.45 optimizer steps/sec)
Effective Batch Size: 16 (per_device_batch_size=4, gradient_accumulation_steps=4)
Optimizer: AdamW (weight_decay=0.01, max_grad_norm=1.0)
Learning Rate: 5e-5 with Cosine Annealing and 100 linear warmup steps
Sequence Length: 384 tokens (loss computed strictly on target tokens with -100 prompt masking)
Precision: bfloat16

How to Use

python

import torchfrom transformers import AutoTokenizer, AutoModelForCausalLMfrom peft import PeftModel BASE_MODEL = "google/gemma-4-E4B"ADAPTER_REPO = "NIVED2003/gemma-4-E4B-dora-poetry-half2" # 1. Load Tokenizer & Base Modeltokenizer = AutoTokenizer.from_pretrained(BASE_MODEL)model = AutoModelForCausalLM.from_pretrained(    BASE_MODEL,    torch_dtype=torch.bfloat16,    device_map="auto") # 2. Load DoRA Adaptermodel = PeftModel.from_pretrained(model, ADAPTER_REPO)model.eval() # 3. Format Promptshloka = "यदा यदा हि धर्मस्य ग्लानिर्भवति भारत । अभ्युत्थानमधर्मस्य तदात्मानं सृजाम्यहम् ॥"prompt = f"Instruction: Translate the following Sanskrit classical verse (shloka) to Hindi.\nInput: {shloka}\nOutput: " inputs = tokenizer(prompt, return_tensors="pt").to("cuda") # 4. Generate Translationwith torch.no_grad():    outputs = model.generate(        **inputs,        max_new_tokens=128,        temperature=0.3,        top_p=0.9,        do_sample=True,        eos_token_id=tokenizer.eos_token_id    ) translation = tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)print("Translation:", translation)

Intermediate Milestones Available

All intermediate milestone checkpoints are versioned in this repository:

checkpoints/step_25pct: 25% training milestone (step 936)
checkpoints/step_50pct: 50% training milestone (step 1872)
checkpoints/step_75pct: 75% training milestone (step 2808)
Root directory (.): Final fully-trained 100% adapter (step 3744)

Citation & Architecture

bibtex

@inproceedings{liu2024dora,  title={DoRA: Weight-Decomposed Low-Rank Adaptation},  author={Liu, Shih-Yang and Wang, Chien-Yi and Yin, Hongxu and Khona, Pavlo and Shen, Sheng and Yen, Chen-Yu and Wang, Ting-Kuei and Chen, Kuan-Yu and Darve, Eric and Chen, Kwang-Ting},  booktitle={International Conference on Machine Learning (ICML)},  year={2024}}

Available on FriendliAI

Dedicated Endpoints

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

Learn more

Model Details

Model Provider

NIVED2003

Model Tree

Base

google/gemma-4-E4B

Adapter

this model

Input Modalities

TextImage

Output Modalities

Text

Supported Functionality

Dedicated Endpoints

Explore FriendliAI today

Get started Talk to an engineer

README

License: apache-2.0

Model Details

Base Model: google/gemma-4-E4B
Tuning Method: DoRA (Weight-Decomposed Low-Rank Adaptation)
Rank ( $r$ ): 16
Alpha ( $α$ ): 32 ( $α / r = 2.0$ )
LoRA Dropout: 0.05
Target Modules: All 7 linear projection layers in the 42 text decoder layers:
- q_proj, k_proj, v_proj, o_proj (Self-Attention)
- gate_proj, up_proj, down_proj (MLP)
Trainable Parameters: 37,925,888 / 8,034,082,336 (0.472%)
Language Pair: Sanskrit (sa) $\to$ Hindi (hi)
Domain: Classical Sanskrit Poetry (Epics, Stotras, Subhashitas, Kavyas)

Evaluation & Training Dynamics

The model was trained for 2 epochs on 29,947 parallel verse pairs and evaluated on 3,778 unseen validation verses.

Validation Loss & Perplexity Trajectory

Table with columns: Milestone, Checkpoint Step, Epoch, Eval Loss (Cross-Entropy), Perplexity (PPL=eloss)
Milestone	Checkpoint Step	Epoch	Eval Loss (Cross-Entropy)	Perplexity ( $PPL = e loss$ )
Milestone 25%	936	0.50	`2.617`	13.69
Milestone 50%	1,872

Within-Domain Stability (Half 1 vs. Half 2)

Comparing both halves of the independent poetry splits confirms high empirical consistency:

Table with columns: Metric, Poetry Half 1, Poetry Half 2, Absolute Delta (Δ), Relative Variance
Metric	Poetry Half 1	Poetry Half 2	Absolute Delta ( $Δ$ )	Relative Variance
Validation Loss	`2.323`	`2.330`	`0.007`	< 0.30%
Validation Perplexity	`10.20`	`10.28`

This tight convergence demonstrates that DoRA preserves structural consistency and semantic alignment across disparate verse selections.

Training Hyperparameters

Hardware: 1x NVIDIA H100 80GB SXM5 GPU (Sapphire Rapids host)
Runtime: 2 hours 19 minutes (8,340s)
Throughput: 7.18 samples/sec (0.45 optimizer steps/sec)
Effective Batch Size: 16 (per_device_batch_size=4, gradient_accumulation_steps=4)
Optimizer: AdamW (weight_decay=0.01, max_grad_norm=1.0)
Learning Rate: 5e-5 with Cosine Annealing and 100 linear warmup steps
Sequence Length: 384 tokens (loss computed strictly on target tokens with -100 prompt masking)
Precision: bfloat16

How to Use

python

import torchfrom transformers import AutoTokenizer, AutoModelForCausalLMfrom peft import PeftModel BASE_MODEL = "google/gemma-4-E4B"ADAPTER_REPO = "NIVED2003/gemma-4-E4B-dora-poetry-half2" # 1. Load Tokenizer & Base Modeltokenizer = AutoTokenizer.from_pretrained(BASE_MODEL)model = AutoModelForCausalLM.from_pretrained(    BASE_MODEL,    torch_dtype=torch.bfloat16,    device_map="auto") # 2. Load DoRA Adaptermodel = PeftModel.from_pretrained(model, ADAPTER_REPO)model.eval() # 3. Format Promptshloka = "यदा यदा हि धर्मस्य ग्लानिर्भवति भारत । अभ्युत्थानमधर्मस्य तदात्मानं सृजाम्यहम् ॥"prompt = f"Instruction: Translate the following Sanskrit classical verse (shloka) to Hindi.\nInput: {shloka}\nOutput: " inputs = tokenizer(prompt, return_tensors="pt").to("cuda") # 4. Generate Translationwith torch.no_grad():    outputs = model.generate(        **inputs,        max_new_tokens=128,        temperature=0.3,        top_p=0.9,        do_sample=True,        eos_token_id=tokenizer.eos_token_id    ) translation = tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)print("Translation:", translation)

import torchfrom transformers import AutoTokenizer, AutoModelForCausalLMfrom peft import PeftModel BASE_MODEL = "google/gemma-4-E4B"ADAPTER_REPO = "NIVED2003/gemma-4-E4B-dora-poetry-half2" # 1. Load Tokenizer & Base Modeltokenizer = AutoTokenizer.from_pretrained(BASE_MODEL)model = AutoModelForCausalLM.from_pretrained(    BASE_MODEL,    torch_dtype=torch.bfloat16,    device_map="auto") # 2. Load DoRA Adaptermodel = PeftModel.from_pretrained(model, ADAPTER_REPO)model.eval() # 3. Format Promptshloka = "यदा यदा हि धर्मस्य ग्लानिर्भवति भारत । अभ्युत्थानमधर्मस्य तदात्मानं सृजाम्यहम् ॥"prompt = f"Instruction: Translate the following Sanskrit classical verse (shloka) to Hindi.\nInput: {shloka}\nOutput: " inputs = tokenizer(prompt, return_tensors="pt").to("cuda") # 4. Generate Translationwith torch.no_grad():    outputs = model.generate(        **inputs,        max_new_tokens=128,        temperature=0.3,        top_p=0.9,        do_sample=True,        eos_token_id=tokenizer.eos_token_id    ) translation = tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)print("Translation:", translation)

Intermediate Milestones Available

All intermediate milestone checkpoints are versioned in this repository:

checkpoints/step_25pct: 25% training milestone (step 936)
checkpoints/step_50pct: 50% training milestone (step 1872)
checkpoints/step_75pct: 75% training milestone (step 2808)
Root directory (.): Final fully-trained 100% adapter (step 3744)

Citation & Architecture

bibtex

@inproceedings{liu2024dora,  title={DoRA: Weight-Decomposed Low-Rank Adaptation},  author={Liu, Shih-Yang and Wang, Chien-Yi and Yin, Hongxu and Khona, Pavlo and Shen, Sheng and Yen, Chen-Yu and Wang, Ting-Kuei and Chen, Kuan-Yu and Darve, Eric and Chen, Kwang-Ting},  booktitle={International Conference on Machine Learning (ICML)},  year={2024}}

@inproceedings{liu2024dora,  title={DoRA: Weight-Decomposed Low-Rank Adaptation},  author={Liu, Shih-Yang and Wang, Chien-Yi and Yin, Hongxu and Khona, Pavlo and Shen, Sheng and Yen, Chen-Yu and Wang, Ting-Kuei and Chen, Kuan-Yu and Darve, Eric and Chen, Kwang-Ting},  booktitle={International Conference on Machine Learning (ICML)},  year={2024}}