From f9709c5cbbac772af34e973cd83cb4dc0a45d7ff Mon Sep 17 00:00:00 2001 From: IlyasMoutawwakil Date: Sat, 11 May 2024 18:35:41 +0200 Subject: [PATCH] order list --- llm_perf/utils.py | 103 ++++++++++++++++++++-------------------------- 1 file changed, 44 insertions(+), 59 deletions(-) diff --git a/llm_perf/utils.py b/llm_perf/utils.py index 071b66b2..8d5145a6 100644 --- a/llm_perf/utils.py +++ b/llm_perf/utils.py @@ -11,50 +11,45 @@ GENERATE_KWARGS = {"max_new_tokens": 64, "min_new_tokens": 64} -CANONICAL_ORGANIZATIONS = [ - # big companies - *["google", "facebook", "meta", "meta-llama", "microsoft", "Intel", "TencentARC", "Salesforce"], - # collectives - *["EleutherAI", "tiiuae", "NousResearch", "Open-Orca"], - # HF related - ["bigcode", "HuggingFaceH4", "huggyllama"], - # community members - ["teknium"], - # startups - *[ - "mistral-community", - "openai-community", - "togethercomputer", - "stabilityai", - "CohereForAI", - "databricks", - "mistralai", - "internlm", - "Upstage", - "xai-org", - "Phind", - "01-ai", - "Deci", - "Qwen", - ], -] - - OPEN_LLM_LIST = OPEN_LLM_LEADERBOARD.drop_duplicates(subset=["Model"])["Model"].tolist() PRETRAINED_OPEN_LLM_LIST = ( OPEN_LLM_LEADERBOARD[OPEN_LLM_LEADERBOARD["Type"] == "pretrained"] .drop_duplicates(subset=["Model"])["Model"] .tolist() ) -# CANONICAL_PRETRAINED_OPEN_LLM_LIST = sorted( -# [model for model in PRETRAINED_OPEN_LLM_LIST if model.split("/")[0] in CANONICAL_ORGANIZATIONS] -# ) - +# CANONICAL_ORGANIZATIONS = [ +# # big companies +# *["google", "facebook", "meta", "meta-llama", "microsoft", "Intel", "TencentARC", "Salesforce"], +# # collectives +# *["EleutherAI", "tiiuae", "NousResearch", "Open-Orca"], +# # HF related +# ["bigcode", "HuggingFaceH4", "huggyllama"], +# # community members +# ["teknium"], +# # startups +# *[ +# "mistral-community", +# "openai-community", +# "togethercomputer", +# "stabilityai", +# "CohereForAI", +# "databricks", +# "mistralai", +# "internlm", +# "Upstage", +# "xai-org", +# "Phind", +# "01-ai", +# "Deci", +# "Qwen", +# ], +# ] +# CANONICAL_PRETRAINED_OPEN_LLM_LIST = [model for model in PRETRAINED_OPEN_LLM_LIST if model.split("/")[0] in CANONICAL_ORGANIZATIONS] CANONICAL_PRETRAINED_OPEN_LLM_LIST = [ "01-ai/Yi-34B", "01-ai/Yi-6B", - "Deci/DeciCoder-1b", "Deci/DeciLM-7B", + "Deci/DeciCoder-1b", "EleutherAI/gpt-j-6b", "EleutherAI/gpt-neo-1.3B", "EleutherAI/gpt-neo-125m", @@ -63,56 +58,48 @@ "EleutherAI/polyglot-ko-12.8b", "EleutherAI/pythia-1.3b", "EleutherAI/pythia-1.4b", - # "EleutherAI/pythia-1.4b-deduped", "EleutherAI/pythia-12b", - # "EleutherAI/pythia-12b-deduped", "EleutherAI/pythia-160m", - # "EleutherAI/pythia-160m-deduped", - # "EleutherAI/pythia-1b-deduped", "EleutherAI/pythia-2.7b", - # "EleutherAI/pythia-2.8b-deduped", "EleutherAI/pythia-410m", - # "EleutherAI/pythia-410m-deduped", "EleutherAI/pythia-6.7b", - # "EleutherAI/pythia-6.9b-deduped", "EleutherAI/pythia-70m", - # "EleutherAI/pythia-70m-deduped", + "Qwen/Qwen-7B", "Qwen/Qwen-14B", "Qwen/Qwen-72B", - "Qwen/Qwen-7B", "Qwen/Qwen1.5-0.5B", "Qwen/Qwen1.5-1.8B", - "Qwen/Qwen1.5-110B", + "Qwen/Qwen1.5-4B", + "Qwen/Qwen1.5-7B", "Qwen/Qwen1.5-14B", "Qwen/Qwen1.5-32B", - "Qwen/Qwen1.5-4B", "Qwen/Qwen1.5-72B", - "Qwen/Qwen1.5-7B", - # "Qwen/Qwen1.5-7B-Chat", + "Qwen/Qwen1.5-110B", "Qwen/Qwen1.5-MoE-A2.7B", "Qwen/Qwen2-beta-14B", "Qwen/Qwen2-beta-72B", - "Salesforce/codegen-16B-nl", - # "Salesforce/codegen-6B-multi", "Salesforce/codegen-6B-nl", + "Salesforce/codegen-16B-nl", "TencentARC/Mistral_Pro_8B_v0.1", "databricks/dbrx-base", "facebook/opt-125m", - "facebook/opt-13b", - "facebook/opt-2.7b", - "facebook/opt-30b", "facebook/opt-350m", + "facebook/opt-2.7b", "facebook/opt-6.7b", + "facebook/opt-13b", + "facebook/opt-30b", "facebook/opt-66b", - "facebook/xglm-4.5B", "facebook/xglm-564M", + "facebook/xglm-4.5B", "facebook/xglm-7.5B", + "google/gemma-2b", "google/gemma-7b", "google/recurrentgemma-2b", + "google/recurrentgemma-7b", "internlm/internlm-20b", "internlm/internlm2-20b", - "meta-llama/Llama-2-13b-hf", "meta-llama/Llama-2-7b-hf", + "meta-llama/Llama-2-13b-hf", "meta-llama/Meta-Llama-3-8B", "meta-llama/Meta-Llama-3-70B", "microsoft/phi-1_5", @@ -127,12 +114,10 @@ "stabilityai/stablelm-3b-4e1t", "stabilityai/stablelm-base-alpha-3b", "stabilityai/stablelm-base-alpha-7b", - # "stabilityai/stablelm-base-alpha-7b-v2", - "tiiuae/falcon-180B", - "tiiuae/falcon-40b", - "tiiuae/falcon-7b", "tiiuae/falcon-rw-1b", - # "togethercomputer/RedPajama-INCITE-7B-Base", + "tiiuae/falcon-7b", + "tiiuae/falcon-40b", + "tiiuae/falcon-180B", "togethercomputer/RedPajama-INCITE-Base-3B-v1", "togethercomputer/RedPajama-INCITE-Base-7B-v0.1", ]