logical-reasoning / llm_toolkit /eval_logical_reasoning.py

Commit History

change BATCH_SIZE to 1 for qwen2-72b eval
4c31851

dh-mc commited on

gemma2 training complete
3a1588f

inflaton commited on

test_b resultls
7670f2e

dh-mc commited on

nv4090 results
dc7abea

dh-mc commited on

internlm2 best results
be560ea

inflaton commited on

try float32 for inference
d772d9d

dh-mc commited on

best eval LF
0b5e165

inflaton commited on

add env vars: MAX_NEW_TOKENS & REPETITION_PENALTY
f754508

dh-mc commited on

fixed bug for eval-mgtv-llama3_8b.sh
387046f

dh-mc commited on

switch llama3 version
c81ec3a

inflaton commited on

08a_InternLM_eval_NV4080 complete
a5837c1

inflaton commited on

T4 original prompt + bf16: best results
c8f289a

dh-mc commited on

refined English translations and Chinese prompt
c755e09

dh-mc commited on

InternLM 2.5 results
5002792

inflaton commited on