LLM EXPLORER 60,519 MODELS INDEXED

Reflection Llama 3.1 70B F16 by unsloth

By unsloth · 103 downloads
HF Repo ↗ Copy URL

Reflection Llama 3.1 70B F16 is an open-source language model by unsloth. Features: 70b LLM, VRAM: 141.9GB, Context: 8K, License: llama3.1, Quantized, Instruction-Based, LLM Explorer Score: 0.14.

Base model:finetune:meta-llama... Base model:meta-llama/meta-lla...   Conversational   Fp16   Instruct   Llama   Llama-3   Quantized   Region:us   Safetensors   Sharded   Tensorflow   Unsloth
Model Card on HF 🤗: https://huggingface.co/unsloth/Reflection-Llama-3.1-70B-f16 

Reflection Llama 3.1 70B F16 Parameters and Internals

Model Type 
text generation
Additional Notes 
The model uses special tokens for reasoning and corrections, which can separate thoughts from final answers.
Training Details 
Methodology:
Reflection-Tuning
Input Output 
Input Format:
Llama 3.1 chat format with custom tags for reasoning and final answer.
Accepted Modalities:
text
Output Format:
Text with internal reasoning and corrections tagged.
Performance Tips:
Use a temperature of 0.7, top_p of 0.95, and append 'Think carefully.' for increased accuracy.
LLM NameReflection Llama 3.1 70B F16
Repository 🤗https://huggingface.co/unsloth/Reflection-Llama-3.1-70B-f16 
Base Model(s)  meta-llama/Meta-Llama-3.1-70B-Instruct   meta-llama/Meta-Llama-3.1-70B-Instruct
Model Size70b
Required VRAM141.9 GB
Updated2024-09-08
Maintainerunsloth
Model Typellama
Instruction-BasedYes
Model Files  4.6 GB: 1-of-30   4.7 GB: 2-of-30   5.0 GB: 3-of-30   5.0 GB: 4-of-30   4.7 GB: 5-of-30   4.7 GB: 6-of-30   4.7 GB: 7-of-30   5.0 GB: 8-of-30   5.0 GB: 9-of-30   4.7 GB: 10-of-30   4.7 GB: 11-of-30   4.7 GB: 12-of-30   5.0 GB: 13-of-30   5.0 GB: 14-of-30   4.7 GB: 15-of-30   4.7 GB: 16-of-30   4.7 GB: 17-of-30   5.0 GB: 18-of-30   5.0 GB: 19-of-30   4.7 GB: 20-of-30   4.7 GB: 21-of-30   4.7 GB: 22-of-30   5.0 GB: 23-of-30   5.0 GB: 24-of-30   4.7 GB: 25-of-30   4.7 GB: 26-of-30   4.7 GB: 27-of-30   5.0 GB: 28-of-30   5.0 GB: 29-of-30   2.1 GB: 30-of-30
Quantization Typefp16
Model ArchitectureLlamaForCausalLM
Licensellama3.1
Context Length8192
Model Max Length8192
Transformers Version4.44.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|reserved_special_token_250|>
Vocabulary Size128262
Torch Data Typebfloat16

Best Alternatives to Reflection Llama 3.1 70B F16

Best Alternatives
Context / RAM
Downloads
Likes
...B Instruct Gradient 1048K 8bit1024K / 75 GB415
...B Instruct Gradient 1048K 4bit1024K / 39.7 GB253
...B Instruct Gradient 1048K 2bit1024K / 21.9 GB252
...0B Instruct Gradient 262K 4bit256K / 39.7 GB253
...0B Instruct Gradient 262K 8bit256K / 75 GB262
...0B Instruct Gradient 262K 2bit256K / 21.9 GB201
... Gradient 262K 2.25bpw H6 EXL2256K / 22.2 GB80
...t Gradient 262K 4.0bpw H6 EXL2256K / 37.2 GB71
...t Gradient 262K 2.4bpw H6 EXL2256K / 23.5 GB60
...t Gradient 262K 3.5bpw H6 EXL2256K / 32.9 GB60