update requirement of llama3 example for higher accuracy

[xin3he]

User description

We observe high accuracy improvement when lm_eval upgrades 0.4.9.1 -> 0.4.9.2

---

PR Type

Enhancement

---

Description

• Update dependencies for improved accuracy

• Upgrade lm_eval from 0.4.9.1 to 0.4.9.2

• Update transformers, torch, torchvision, datasets, auto-round, and neural-compressor-pt to latest versions

---

Diagram Walkthrough

flowchart LR
  A["Update lm_eval"] -- "0.4.9.1 -> 0.4.9.2" --> B["Improved Accuracy"]
  C["Upgrade transformers"] -- "4.56.2 -> 4.57.3" --> B
  D["Upgrade torch"] -- "2.7.0 -> 2.9.0" --> B
  E["Upgrade torchvision"] -- "0.22.0 -> 0.24.0" --> B
  F["Upgrade datasets"] -- "3.6.0 -> 4.4.2" --> B
  G["Upgrade auto-round"] -- ">=0.8.0 -> >=0.9.3" --> B
  H["Upgrade neural-compressor-pt"] -- ">=3.6 -> >=3.7" --> B

Loading

File Walkthrough

Relevant files

Enhancement

requirements.txt Update dependencies for improved accuracy                                examples/pytorch/nlp/huggingface_models/language-modeling/quantization/auto_round/llama3/requirements.txt Upgrade transformers to 4.57.3 Upgrade torch to 2.9.0 Upgrade torchvision to 0.24.0 Upgrade lm_eval to 0.4.9.2 Upgrade datasets to 4.4.2 Upgrade auto-round to >=0.9.3 Upgrade neural-compressor-pt to >=3.7 +7/-8

---

[PRAgent4INC]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 No relevant tests

🔒 No security concerns identified

⚡ Recommended focus areas for review Missing Dependency The dependency deepspeed is missing in the updated requirements.txt. Ensure that all necessary dependencies are included. transformers==4.57.3 torch==2.9.0 torchvision==0.24.0 lm_eval==0.4.9.2 datasets==4.4.2 auto-round>=0.9.3 neural-compressor-pt>=3.7

[PRAgent4INC]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
General	Check package compatibility Verify compatibility of updated packages with the existing codebase. examples/pytorch/nlp/huggingface_models/language-modeling/quantization/auto_round/llama3/requirements.txt [1-7] -transformers==4.56.2 -torch==2.7.0 -torchvision==0.22.0 -lm_eval==0.4.9.1 -datasets==3.6.0 -auto-round>=0.8.0 -neural-compressor-pt>=3.6 +transformers==4.57.3 +torch==2.9.0 +torchvision==0.24.0 +lm_eval==0.4.9.2 +datasets==4.4.2 +auto-round>=0.9.3 +neural-compressor-pt>=3.7 Suggestion importance[1-10]: 5 __ Why: The suggestion asks to verify the compatibility of updated packages, which is important but not a direct code change. It does not address any specific bug or error.	Low