Llama-3.3-8B-Instruct-128K/README.md

license, base_model, pipeline_tag, model-index

license	base_model	pipeline_tag	model-index
llama3.3	allura-forge/Llama-3.3-8B-Instruct	text-generation	name results shb777/Llama-3.3-8B-Instruct-128K task dataset metrics type text-generation name type BBH leaderboard type value name accuracy 54.1 acc_norm task dataset metrics type text-generation name type GPQA leaderboard type value name accuracy 29.9 acc_norm task dataset metrics type text-generation name type MMLU Pro leaderboard type value name accuracy 38.0 acc task dataset metrics type text-generation name type MuSR leaderboard type value name accuracy 37.8 acc_norm task dataset metrics type text-generation name type IFEval leaderboard type value name accuracy 85.2 avg(prompt_strict + inst_strict) task dataset metrics type text-generation name type MATH Hard leaderboard type value name accuracy 27.3 exact_match

Llama 3.3 8B 128K Instruct (Fixed)

Important

Original allura-forge/Llama-3.3-8B-Instruct, Thanks!

Tip

imatrix GGUF's by mradermacher (Recommended)

static GGUF's

Evals

Additional Fixes:

• Added rope_scaling
• Added chat template (Unsloth) in tokenizer config
• Updated generation config
• Enabled full context length