Meta-Llama-3-8B-Instruct

MERA Created at 03.02.2026 12:27

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Place in the rating
Agricultural industry 0.252 66
Medicine and healthcare 0.455 50

Information about the submission

Mera version
v1.0.0
Torch Version
2.9.1
The version of the codebase
435b60a
CUDA version
12.8
Precision of the model weights
bfloat16
Seed
1234
Batch
1
Transformers version
4.57.6
The number of GPUs and their type
1 x NVIDIA A100-SXM4-80GB
Architecture
vllm

Team:

MERA

Name of the ML model:

Meta-Llama-3-8B-Instruct

Model size

8.0B

Model type:

Opened

SFT

Additional links:

Llama cookbook https://github.com/meta-llama/llama-cookbook

Architecture description:

Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety.

Description of the training:

Use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety.

Pretrain data:

Trained on a mix of publicly available online data of 15T+ tokens.

License:

https://www.llama.com/llama3/license/

Inference parameters

Generation Parameters:
agro_bench - do_sample=false;until=["\n\n"]; \naqua_bench - do_sample=false;until=["\n\n"]; \nmed_bench - do_sample=false;until=["\n\n"];

The size of the context:
agro_bench, aqua_bench, med_bench - 8192

Description of the template:
{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|> \n \n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|> \n \n' }}{% endif %}