Dialogue-oriented instruction tuning
According to the model card, the Llama 3 instruction-tuned models are optimized for dialogue use cases in 8B and 70B sizes.
Open Source Model Profile · meta-llama
Meta-Llama-3-70B-Instruct
Meta-Llama-3-70B-Instruct is a 70.55B-parameter Llama-family text-generation fine-tune from meta-llama. According to the model card, it is an instruction-tuned Llama 3 variant optimized for dialogue use cases.
- Publisher
- meta-llama
- Task
- text-generation
- Model type
- llama
- License
- llama3
- Library
- transformers
- Publication status
- Accepted · not indexed
Model overview
Meta-Llama-3-70B-Instruct is published by meta-llama as a llama text-generation model. The captured configuration identifies LlamaForCausalLM and Safetensors metadata reports 70,553,706,496 parameters. According to the model card, it belongs to the Llama 3 family in 8B and 70B instruction-tuned variants, with tuned versions using supervised fine-tuning and reinforcement learning with human feedback.
Recorded capabilities
SFT and RLHF alignment
According to the model card, the tuned versions use supervised fine-tuning and reinforcement learning with human feedback to align for helpfulness and safety.
8k context with GQA
According to the model card, both 8B and 70B versions use an 8k context length and Grouped-Query Attention for inference scalability.
Publisher-reported benchmarks
According to the model card, the 70B instruction-tuned variant is reported at 82.0 on MMLU, 81.7 on HumanEval, 93.0 on GSM-8K, and 50.4 on MATH.
Use cases in the source record
- Conversational text-generation workflows consistent with the publisher-described dialogue optimization and documented Transformers usage.
- Benchmark-informed model selection using the publisher-reported instruction-tuned scores for MMLU, HumanEval, GSM-8K, and MATH.
Limitations and unknowns
- According to the model card, testing was conducted in English and outputs may be inaccurate, biased, or objectionable, so application-specific safety testing is recommended.
- No pricing, VRAM requirement, or latency figures were extracted from this record.
- No context-window measurement beyond the publisher-reported 8k value was independently verified.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: meta-llama/Meta-Llama-3-70B-Instruct
Captured: Unknown. Processed: 2026-09-07T19:34:51.858219+00:00.
Model Details Meta developed and released the Meta Llama 3 family of large language models (LLMs), a collection of pretrained and instruction tuned generative text models in 8 and 70B sizes. The Llama 3 instruction tuned models are optimized for dialogue use cases and outperform many of the available open source chat models on common industry benchmarks. Further, in developing these models, we took great care to optimize helpfulness and safety. Model developers Meta Variations Llama 3 comes in two sizes — 8B and 70B parameters — in pre-trained and instruction tuned variants. Input Models input text only. Output Models generate text…
F001F002F003F004F005F006F007F008F010F011F012F015F017F022F026F030