Llama 3 8B Instruct 262K By gradientai: Benchmarks, Features and Detailed Analysis. Insights on Llama 3 8B Instruct 262K.

Arxiv:2305.14233 Arxiv:2309.00071 Arxiv:2402.08268 Autotrain compatible Conversational En Endpoints compatible Instruct Llama Llama-3 Meta Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/gradientai/Llama-3-8B-Instruct-262k

Llama 3 8B Instruct 262K Benchmarks

ARC: 53.16 vs 96.7 (so35)^-45%

HellaSwag: 75.53 vs 95.3 (gpt4)^-20.7%

MMLU: 64.27 vs 88.3 (so35)^-27.2%

TruthfulQA: 48.37 vs 59 (gpt4)^-18%

WinoGrande: 73.01 vs 87.5 (gpt4)^-16.6%

GSM8K: 47.23 vs 96.4 (so35)^-51%

LLME Score: 0.22661

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

Llama 3 8B Instruct 262K (gradientai/Llama-3-8B-Instruct-262k)

🌟 Advertise your project 🚀

Llama 3 8B Instruct 262K Parameters and Internals

Model Type

text generation, dialogue, instruction tuned

Use Cases

Areas:

Commercial use, Research use

Applications:

Dialog and assistant-like applications

Primary Use Cases:

Assistant-like chat and instruction tasks

Limitations:

Inapplicable in languages other than English without additional tuning., Requires appropriate safety tuning for specialized applications.

Considerations:

Developers should employ additional safeguarding measures and consider context when deploying applications.

Additional Notes

Quantized versions available for resource-efficient deployments.

Supported Languages

en (Primary support for dialogue and instructional tasks in English.)

Training Details

Data Sources:

publicly available online data, SlimPajama-627B, UltraChat

Data Volume:

15 trillion tokens (pretraining)

Methodology:

Supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF). NTK-aware interpolation technique for context length adjustment.

Context Length:

262144

Training Time:

Total of 7.7M GPU hours for pretraining across multiple models.

Hardware Used:

Crusoe Energy high performance L40S cluster

Model Architecture:

Optimized transformer using RoPE theta for extended contexts.

Safety Evaluation

Methodologies:

red-teaming, adversarial evaluations

Findings:

Improved refusal rate for false prompts in comparison to Llama 2., Limited residual risks assessed through community tools.

Risk Categories:

Child safety, Cybersecurity vulnerabilities, CBRNE threats

Ethical Considerations:

Guided by a Responsible Use Guide and supporting tools like Meta Llama Guard 2.

Responsible Ai Considerations

Fairness:

Focused on openness, inclusivity, and helpfulness while respecting diverse values and perspectives.

Transparency:

Incorporates feedback from the community to continually assess safety and alignment.

Accountability:

Developers undertaking use are held to the standards under the Acceptable Use Policy.

Mitigation Strategies:

Use community safeguards like Llama Guard to supplement model-level safety.

Input Output

Input Format:

Text input following a conversational template.

Accepted Modalities:

text

Output Format:

Generated text and code responses.

Performance Tips:

Leverage context window optimizations for handling extensive lengths.

Release Notes

Version:

5/3 Update

Date:

2024-05-03

Notes:

Finalized assistant-like chat optimizations and context length extension techniques.

LLM Name	Llama 3 8B Instruct 262K
Repository 🤗	https://huggingface.co/gradientai/Llama-3-8B-Instruct-262k
Model Size	8b
Required VRAM	16.1 GB
Updated	2025-10-02
Maintainer	gradientai
Model Type	llama
Instruction-Based	Yes
Model Files	5.0 GB: 1-of-4 5.0 GB: 2-of-4 4.9 GB: 3-of-4 1.2 GB: 4-of-4
Supported Languages	en
Model Architecture	LlamaForCausalLM
License	llama3
Context Length	262144
Model Max Length	262144
Transformers Version	4.41.0.dev0
Tokenizer Class	PreTrainedTokenizerFast
Vocabulary Size	128256
Torch Data Type	bfloat16

Quantized Models of the Llama 3 8B Instruct 262K

Model	Likes	Downloads	VRAM
Llama 3 8B Instruct 262K GGUF	2	219	3 GB
...Instruct 262K HQQ 2bit Smashed	2	5	4 GB

Best Alternatives to Llama 3 8B Instruct 262K

Best Alternatives	Context / RAM	Downloads	Likes
...otron 8B UltraLong 4M Instruct	4192K / 32.1 GB	1531	120
...a 3.1 8B UltraLong 4M Instruct	4192K / 32.1 GB	176	24
UltraLong Thinking	4192K / 16.1 GB	45	3
...otron 8B UltraLong 2M Instruct	2096K / 32.1 GB	1142	15
...a 3.1 8B UltraLong 2M Instruct	2096K / 32.1 GB	875	9
...otron 8B UltraLong 1M Instruct	1048K / 32.1 GB	7142	52
...a 3.1 8B UltraLong 1M Instruct	1048K / 32.1 GB	1387	29
Zero Llama 3.1 8B Beta6	1048K / 16.1 GB	2	1
...dger Nu Llama 3.1 8B UltraLong	1048K / 16.2 GB	5	3
....1 1million Ctx Dark Planet 8B	1048K / 32.3 GB	8	3

Note: green Score (e.g. "73.2") means that the model is better than gradientai/Llama-3-8B-Instruct-262k.

Rank the Llama 3 8B Instruct 262K Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

What open-source LLMs or SLMs are you in search of? 51544 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Release v20241124

Support LLM Explorer