model_info changes

sotopia_space/benchmark.py

@@ -15,8 +15,7 @@ def benchmark_table():
     gr.Markdown(f"**Version**: sotopia (v1.01; 2024.04.22) | **# Examples**: 7200 | **# Models**: {len(MODEL_OPTIONS)} | **# Comparisons**: x", elem_classes="markdown-text")
                 
     with gr.TabItem("Vs GPT-3.5", elem_id="od-benchmark-tab-table-ablation", id=0, elem_classes="subtab"):
-        original_df = pd.read_json('data_dir/models_vs_gpt35.jsonl', lines=True)
-        default_main_df = original_df
+        default_main_df = pd.read_json('data_dir/models_vs_gpt35.jsonl', lines=True)
         default_main_df = default_main_df.sort_values(by="GOAL [0, 10]", ascending=False)
         default_main_df = post_processing(default_main_df, None)
         # add a Rank column to the first columnn (starting from 1)

sotopia_space/constants.py

@@ -14,26 +14,8 @@ MODEL_OPTIONS = [
 ]
 
 MODEL_INFO = {
-    "Llama-2-13b-chat-hf.nosp": {"pretty_name": "Llama-2-13B-chat", "hf_model_id": "meta-llama/Llama-2-13b-chat-hf"},
-    "Llama-2-70b-chat-hf.nosp": {"pretty_name": "Llama-2-70B-chat", "hf_model_id": "meta-llama/Llama-2-70b-chat-hf"},
-    "Llama-2-7b-chat-hf.nosp": {"pretty_name": "Llama-2-7B-chat", "hf_model_id": "meta-llama/Llama-2-7b-chat-hf"},
-    "Llama-2-7b-chat-hf": {"pretty_name": "Llama-2-7B-chat (+sys prmpt)", "hf_model_id": "meta-llama/Llama-2-7b-chat-hf"},
-    "Mistral-7B-Instruct-v0.1": {"pretty_name": "Mistral-7B-Instruct", "hf_model_id": "mistralai/Mistral-7B-Instruct-v0.1"},
-    "Mistral-7B-Instruct-v0.2": {"pretty_name": "Mistral-7B-Instruct (v0.2)", "hf_model_id": "mistralai/Mistral-7B-Instruct-v0.2"},
-    "Mixtral-8x7B-Instruct-v0.1": {"pretty_name": "Mixtral-8x7B-Instruct", "hf_model_id": "mistralai/Mixtral-8x7B-Instruct-v0.1"},
-    "Nous-Hermes-2-Mixtral-8x7B-DPO": {"pretty_name": "Nous-Hermes-2-Mixtral-8x7B-DPO", "hf_model_id": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO"},
-    "Yi-34B-Chat": {"pretty_name": "Yi-34B-Chat", "hf_model_id": "01-ai/Yi-34B"},
-    "gemini-1.0-pro": {"pretty_name": "gemini-1.0-pro", "hf_model_id": "https://blog.google/technology/ai/google-gemini-ai/"},
-    "gemma-7b-it": {"pretty_name": "Gemma-7B-it", "hf_model_id": "google/gemma-7b"},
-    "gpt-3.5-turbo-0125": {"pretty_name": "gpt-3.5-turbo-0125", "hf_model_id": "https://platform.openai.com/"},
-    "gpt-4-0125-preview": {"pretty_name": "gpt-4-0125-preview", "hf_model_id": "https://platform.openai.com/"},
-    "tulu-2-dpo-70b": {"pretty_name": "Tulu-2-dpo-70b", "hf_model_id": "cmu-lti/tulu-2-dpo-70b"},
-    "vicuna-13b-v1.5": {"pretty_name": "Vicuna-13b-v1.5", "hf_model_id": "lmsys/vicuna-13b-v1.5"},
-    "zephyr-7b-beta": {"pretty_name": "Zephyr-7b-beta", "hf_model_id": "HuggingFaceH4/zephyr-7b-beta"},
-    "mistral-large-2402": {"pretty_name": "Mistral-Large", "hf_model_id": "https://mistral.ai/news/mistral-large/"},
-    "claude-3-opus-20240229": {"pretty_name": "Claude 3 Opus", "hf_model_id": "https://www.anthropic.com/claude"},
-    "claude-3-sonnet-20240229": {"pretty_name": "Claude 3 Sonnet", "hf_model_id": "https://www.anthropic.com/claude"},
-    "zephyr-7b-gemma-v0.1": {"pretty_name": "Zephyr-7b-Gemma", "hf_model_id": "HuggingFaceH4/zephyr-7b-gemma-v0.1"},
-    "Starling-LM-7B-beta": {"pretty_name": "StarlingLM-7B-beta", "hf_model_id": "Nexusflow/Starling-LM-7B-beta"},
-    "dbrx-instruct": {"pretty_name": "DBRX Instruct", "hf_model_id": "databricks/dbrx-instruct"}
+    "GPT-4": {"pretty_name": "GPT-4", "hf_model_id": "https://openai.com/blog/new-models-and-developer-products-announced-at-devday"},
+    "GPT-3.5": {"pretty_name": "GPT-3.5", "hf_model_id": "https://openai.com/blog/new-models-and-developer-products-announced-at-devday"},
+    "Llama-2": {"pretty_name": "Llama-2", "hf_model_id": "https://llama.meta.com/llama2/"},
+    "MPT": {"pretty_name": "MPT", "hf_model_id": "https://huggingface.co/docs/transformers/main/en/model_doc/mpt"}
 }